Ashington Piggeries Ltd. and Another Appellants v. Christopher Hill -1972- AC 441

FOR EDUCATIONAL USE ONLY 

*441 Ashington Piggeries Ltd. and Another Appellants v. Christopher Hill

Ltd. Respondents

Christopher Hill Ltd. Appellants v. Norsildmel Respondents

House of Lords

HL

Lord HodsonLord Guest, Viscount Dilhorne , Lord Wilberforce and Lord Diplock

1970 Nov. 10, 11, 12, 16, 18, 19, 23, 24, 25, 26, 30; Dec. 1, 2; 1971 Feb. 24

[On Appeal from Christopher Hill Ltd. v. Ashington Piggeries Ltd. (Conjoined

Appeals)]

Sale of Goods--Description--Merchantable quality--Mink food--Agreed formula-- Mixed by compounders--Norwegian herring meal--Ingredient toxic to mink--Not detectable on delivery--Liability of compounders to customers--Liability of suppliers to compounders--Defects clause--Effect--Description clause--Meaning of quality--Sale of Goods Act 1893 (56 & 57 Vict. c. 71), ss. 13, 14

By section 13 of the Sale of Goods Act 1893:
"Where there is a contract for the sale of goods by description there is an implied condition that the good shall correspond with description, ..."
By section 14:
"Subject to the provisions of this Act ... there is no implied warranty or condition as to the quality or fitness for any particular purpose of goods supplied under a contract of sale, except as follows:- (1) Where the buyer, expressly or by implication, makes known to the seller the particular purpose for which the goods are required, so as to show that the buyer relies on the *442 seller's skill or judgment, and the goods are of a description which it is in the course of the seller's business to supply ... there is an implied condition that the goods shall be reasonably fit for such purpose, ... (2) Where goods are bought by description from a seller who deals in goods of that description ... there is an implied condition that the goods shall be of merchantable quality; provided that if the buyer has examined the goods, there shall be no implied condition as regards defects which such examination ought to have revealed: ..."
The plaintiffs, who were feeding-stuff compounders, contracted with the defendants, who were mink breeders, to compound for and deliver to them an animal foodstuff known as "King Size." The foodstuff so supplied caused thousands of mink to die because an ingredient, Norwegian herring meal, supplied to the plaintiffs by a third party, a Norwegian organisation, contained a toxic agent, a chemical "DMNA" which, however, was not proved to be unfit for animals other than mink. The contamination resulted from the use of a preservative. The contract under which the herring meal was supplied to the plaintiffs provided as to quality and description that it was to be "fair average quality of the season, expected to analyse not less than 70 per cent. protein, not more than 12 per cent. fat and not more than 4 per cent salt. " Under sampling and analysis it was provided: "Official Norwegian certificates of analysis of samples taken officially during loading will be supplied. Such shipping analysis shall be final ..." By condition 3: "The goods shall be taken with all faults and defects, damaged or inferior, if any, at valuation to be arranged mutually or by arbitration." The defendants had provided for the compound a formula which, had it been adhered to, would not have contained the toxic agent.
The plaintiffs sued the defendants for sums due in respect of the supply of the feeding stuff. The defendants counterclaimed, relying on sections 13 and 14 of the Sale of Goods Act 1893. The plaintiffs joined as a third party the Norwegian sellers of the herring meal:-

Held: 

(1) that the plaintiffs, acting within the field in which they were called upon to exercise their skill and judgment, were in breach of the term implied under section 14 (1) of the Sale of Goods Act 1893 that the compound should be reasonably fit for the purpose of feeding mink and were liable in damages to the defendants.
Cammell Laird & Co. Ltd. v. Manganese Bronze & Brass Co. Ltd. [1934] A.C. 402, H.L.(E.) applied.
(2) (Lord Hodson and Lord Diplock dissenting) that, inasmuch as the goods were admittedly unmerchantable, the plaintiffs were also in breach of section 14 (2) of the Act since, being dealers in animal feeding stuffs, they dealt in "goods of that description," which words should not be restricted in their scope to the contract description of the goods.
(3) (Lord Diplock dissenting) that the plaintiffs could recover from the third party by way of indemnity what they had to pay to the defendants.
Held, further (Viscount Dilhorne dissenting), that the plaintiffs were not in breach of section 13 of the Act, inasmuch as there was no "misdescription, " since the fact that the herring meal was contaminated and poisonous to mink did not render it erroneous to describe it as herring meal.
*443 Pinnock Brothers v. Lewis & Peat Ltd. [1923] 1 K.B. 690 distinguished.
Manchester Liners Ltd. v. Rea Ltd. [1922] 2 A.C. 74, 89, H.L.(E.) applied.
Decisions of the Court of Appeal [1969] 3 All E.R. 1496; [1969] 2 Lloyd's Rep. 425 reversed.

The following cases are referred to in their Lordships' opinions:
Arcos Ltd. v. E. A. Ronaasen & Son [1933] A.C. 470, H.L.(E.).
Bowes v. Shand (1877) 2 App.Cas. 455, H.L.(E.).
British oil & Cake Co. Ltd. v. J. Burstall & Co. Ltd. (1923) 15 Ll.L.Rep. 46.
Brown v. Edgington (1841) 2 Man. &; G. 279.
Brown (B. S.) & Son Ltd. v. Craiks Ltd. [1970] 1 W.L.R. 752; [1970] 1 All E.R. 823, H.L.(Sc.).
Burnby v. Bollett (1847) 16 M. &; W. 644.
Cammell Laird & Co. Ltd. v. Manganese Bronze and Brass Co. Ltd. [1934] A.C. 402, H.L.(E.).
Czarnikow (C.) Ltd. v. Koufos [1969] 1 A.C. 350; [1967] 3 W.L.R. 1491; [1967] 3 All E.R. 686, H.L.(E.).
Drummond (James) & Sons v. E. H. Van Ingen & Co. (1887) 12 App.Cas. 284, H.L.(E.).
Griffiths v. Peter Conway Ltd. [1939] 1 All E.R. 685 , C.A..
Hardwick Game Farm v. Suffolk Agricultural Poultry Producers Association [1969] 2 A.C. 31; [1968] 3 W.L.R. 110; [1968] 2 All E.R. 444, H.L.(E.).
Hookway (F. E.) & Co. Ltd. v. Alfred Isaacs & Sons [1954] 1 Lloyd's Rep. 491.
Jones v. Bright (1829) 15 Bing. 533.
Jones v. Just (1868) L.R. 3 Q.B. 197, D.C..
MacPherson Train & Co. Ltd. v. Howard Ross & Co. Ltd. [1955] 1 W.L.R. 640; [1955] 2 All E.R. 445.
Manchester Liners Ltd. v. Rea Ltd. [1922] 2 A.C. 74, H.L.(E.).
Mash & Murrell Ltd. v. Joseph I. Emanuel Ltd. [1961] 1 W.L.R. 862; [1961] 1 All E.R. 485; [1962] 1 W.L.R. 16; [1962] 1 All E.R. 77, C.A..
Medway Oil & Storage Co. Ltd. v. Silica Gel Corporation (1928) 33 Com. Cas. 195, H.L.(E.).
Munro (Robert A.) & Co. Ltd. v. Meyer [1930] 2 K.B. 312.
Pacific Trading Co. Ltd. v. Wiener (1923) 14 Ll.L.Rep. 51.
Pinnock Brothers v. Lewis & Peat Ltd. [1923] 1 K.B. 690.
Spencer Trading Co. Ltd. v. Devon (Fixol and Stickphast Ltd., Third Parties) [1947] 1 All E.R. 284.
Steels & Busks Ltd. v. Bleecker Bik & Co. Ltd. [1956] 1 Lloyd's Rep. 228.
Teheran-Europe Co. Ltd. v. S. T. Belton (Tractors) Ltd. [1968] 2 Q.B. 545; [1968] 3 W.L.R. 205; [1968] 2 All E.R. 886, C.A..
Varley v. Whipp [1900] 1 Q.B. 513.
The following additional cases were cited in argument:
Couchman v. Hill [1947] K.B. 554; [1947] 1 All E.R. 103, C.A..
Frost v. Aylesbury Dairy Co. Ltd. [1905] 1 K.B. 608, C.A..
Grant v. Australian Knitting Mills Ltd. [1936] A.C. 85, P.C..
Groupement National d'Achat des Tourteaux v. Sociedad Industrial Financiera Argentina (The Milton B. Medary) [1962] 2 Lloyd's Rep. 192.
Jones v. Padgett (1890) 24 Q.B.D. 650, C.A..
*444 Lindsay (W. N.) & Co. Ltd. v. European Grain & Shipping Agency Ltd. [1963] 1 Lloyd's Rep. 437, C.A..
Lloyd del Pacifico v. Board of Trade (1929) 35 Ll.L.Rep. 217, C.A..
Meyer Ltd. v. Osakeyhtio Carelia Timber Co. Ltd. (1930) 37 Ll.L.Rep. 76.
Mihalis Angelos, The [1971] 1 Q.B. 164; [1970] 3 W.L.R. 601; [1970] 3 All E.R. 125, C.A..
North Western Rubber Co. Ltd. and Hüttenbach & Co.'s Arbitration, In re [1908] 2 K.B. 907, C.A..
Randall v. Newson (1877) 2 Q.B.D. 102, C.A..
Sanday & Co. v. Keighley, Maxted & Co. (1922) 91 L.J.K.B. 624, C.A..
Shepherd v. Pybus (1842) 3 M. &; G. 868.
Sumner, Permain & Co. v. Webb & Co. [1922] 1 K.B. 55, C.A..
Ward v. Hobbs (1878) 4 App.Cas. 13, H.L.(E.).
Wimble, Sons & Co. v. Lillico & Son (London) (1922) 38 T.L.R. 296.

APPEALS from the Court of Appeal (Davies, Russell and Megaw L.JJ.).
The first appeal was from an order of the Court of Appeal dated July 22, 1969, allowing the appeal of the present respondents herein, Christopher Hill Ltd. (the plaintiffs in the action), from the judgment of Milmo J. in the Queen's Bench Division, given on January 30, 1968, by which he held that the present appellants, Ashington Piggeries Ltd. and Fur Farm Supplies Ltd. (the defendants in the action), were entitled to recover damages from the respondents for breaches of condition in the sale of goods. (The two companies were controlled by a Mr. W. T. Udall.)
The second appeal was from a judgment of the Court of Appeal dated July 22, 1969, whereby so much of the judgment of Milmo J. was reversed as adjudged that Norsildmel, the respondents herein, had committed a breach of their contract with Christopher Hill Ltd., the appellants in this appeal, and were liable to pay to these appellants any part of such damages as might be found to be payable by the appellants to the defendants in the action; and that the respondents were further liable to pay to the appellants (1) the amount of costs ordered to be paid by the appellants to the defendants, being 75 per cent. of the defendants' costs and (2) 75 per cent. of the appellants' costs.
The plaintiffs in the action (Christopher Hill Ltd.) sued the defendants (Ashington Piggeries Ltd. and Fur Farm Supplies Ltd.) for the price of a feeding stuff supplied to them. The defendants denied liability and counterclaimed, contending that the goods supplied did not correspond to description, were not reasonably fit for the required purpose and were not of merchantable quality, so that mink which ate the food died. The plaintiffs joined as third parties Norsildmel, the successors to the Norwegian sellers from whom they had purchased herring meal which was an ingredient in the feeding stuff.
The dispute between Christopher Hill Ltd. and Norsildmel arose out of a contract dated February 14, 1961, by which Christopher Hill Ltd. agreed to buy and the predecessors of Norsildmel agreed to sell a quantity of Norwegian herring meal. The contract contained, inter alia, the following provisions:
"QUANTITY AND DESCRIPTION
About 300/350 tons at sellers' option of 2240 lb./1016 kilos of Norwegian herring meal fair average quality of the season, expected to *445 analyse not less than 70% protein, not more than 12% fat and not more than 4% salt.
"PRICE
12s. 3d. per unit of protein (fractions pro rata) per ton ... In the event of any surpluses of fat and/or salt, sellers willl make an allowance to buyers at the rate of 1% surplus = 1% of the contract price, fractions pro rata. If the fat contents exceed 13% or the salt contents exceed 5% the buyers will have the right of rejection. ...
"SHIPMENT
From Norway during March 1961. ...
"SAMPLING AND ANALYSIS
Official Norwegian certificates of analysis of samples taken officially during loading will be supplied. Such shipping analyses shall be final unless the buyer elects, prior to the arrival of the steamer(s) at port(s) of destination, to have samples drawn by independent parties on discharge of the vessel(s) at their expense, under the supervision of sellers' representative, if required. ...
"SPECIAL CONDITIONS
Guaranteed for consumption in U.K.
"GENERAL CONDITIONS
... (3) The goods to be taken with all faults and defects, damaged or inferior, if any, at valuation to be arranged mutually or by arbitration. ..."

[FIRST APPEAL]

G. R. F. Morris Q.C. and A. G. Guest (Michael Spencer with them) for the appellants, Ashington Piggeries Ltd. and Fur Farm Supplies Ltd. It is immaterial whether the contamination of the foodstuffs supplied to the appellants occurred by the addition of some toxic substance or by a processed chemical change: see Pinnock Brothers v. Lewis & Peat Ltd. [1923] 1 K.B. 690. The important thing was that there was a formula and a departure from the formula because DMNA was not an ingredient of it: see Arcos Ltd. v. E. A. Ronaasen & Son [1933] A.C. 471.
Under section 14 (1) of the Sale of Goods Act 1893 two questions arise: (1) Was the particular purpose for which the goods were required made known in such a way as to show that the buyers relied on the sellers' skill and judgment? (2) Were the goods of a description which it was in the course of the sellers' business to supply?
The selection and acquisition of foodstuffs was within the sellers' province, and this was relied on by the buyers. The position would be otherwise if mink were idiosyncratic and DMNA did not affect other animals. Reliance is placed on Cammell Laird & Co. Ltd. v. Manganese Bronze & Brass Co. Ltd. [1934] A.C. 402.
The buyers provided the formula specifying the ingredients and (as in Cammell Laird) the sellers, produced the specification and had the obligation of selecting the source of the ingredients and compounding them. That was a province which the buyers at no stage invaded. They had *446 no knowledge of the sources of the ingredients nor did they know if and when they were changed. There is no evidence that they ever asked about it. The Court of Appeal did not treat the sellers' province in the way laid down in Cammell Laird. One of the most striking ways of demonstrating the boundaries of that, province is to consider the case where a seller in the course of negotiations gives advice as to the contents of the formula presented to him. The sellers here suggested herring meal instead of fish meal.
Because of the imponderable variables which operated it was not possible on the evidence before Milmo J. to make any valid findings in respect of the balance of the consignment: (i) whether it contained any DMNA, or, if it did, (ii) into which feed or parcel the DMNA went or, (iii) the type of animal which received the toxic meal. Moreover, there was no evidence that mink had an exclusive reaction to DMNA; they merely died sooner than other animals. On that basis the province in which the respondents exercised their skill and judgment was the subjecting of the feed mix to herring meal containing DMNA which affected the other animals for whom they compounded feeding stuffs, apart from mink, and therefore they are not entitled to exclude liability, as the Court of Appeal held. The question of fault is irrelevant to knowledge on the part of the respondents for they are liable for latent defects: see Frost v. Aylesbury Dairy Co. Ltd. [1905] 1 K.B. 608, at ,pp. 612, 613, per Collins M.R. and Hardwick Game Farm v. Suffolk Agricultural Poultry Producers Association [1969] 2 A.C. 31, 84, per Lord Reid.
The appellants' argument falls under the following heads: (1) Breach by the respondents of the implied condition that where there is a contract for the sale of goods by description there is an implied condition that the goods shall correspond with the description pursuant to section 13 of the Sale of Goods Act or, alternatively, that the goods should correspond with the contractual description. (2) Breach by the respondents of an implied condition that the relevant consignments of "King Size" should be reasonably fit for the particular purpose for which they were required pursuant to section 14 (1) of the Sale of Goods Act. (3) Breach by the respondents of an implied condition that the relevant consignments of "King Size" should be of merchantable quality pursuant to section 14 (2) of the Sale of Goods Act.
Section 14 (1) applies to "goods supplied under a contract of sale." By section 62 (1) "'contract of sale' includes an agreement to sell as well as a sale." The agreement of May 1960 may have been an agreement to sell, but it was not a sale pursuant to section 1 (3). Contrary to the view of the Court of Appeal, each order and delivery constituted a "sale," and was therefore within the ambit of section 14 (1). By March 1961 it was unquestionably in the course of the respondents' business to supply mink food.
Further, or in the alternative, "King Size" was, in any event, goods of a description which it was in the course of the respondents' business to supply. Whether it be held that "description," related to the circumstances of this case, means "King Size mink food as described in the agreed formula" or "mink food" or animal food, the supply of such goods by the respondents was "in the course of their business" and not *447 otherwise: see Spencer Trading Co. Ltd. v. Devon (Fixol and Stickphast Ltd., Third parties) [1947] 1 All E.R. 284. The agreement was made in the course of business between corporations with a view to profit. The respondents, in accepting the contractual obligation to supply goods to the particular description, did so in the course of their business and not otherwise, and the negotiations leading up to the making of the agreement are a relevant factor in this respect.
The issue under section 14 (2) is whether the goods were bought by description "from a seller who deals in goods of that description." The absence of authority on this point is not confined to the United Kingdom but extends to New Zealand, several of the States of Australia and to Canada.
The essence of the Court of Appeal's finding was that the words "that description" meant precisely the contractual description of the goods: see [1969] 2 Ll.L.Rep. 425, 464. The Court of Appeal were wrong in approaching this matter solely with reference to contractual goods; the court should also have considered the status or capacity of the dealer. The word "deals" involves trade. A "dealer" in goods in the strict sense is one who buys goods and resells them without having himself altered such goods. But section 14 (2) embraces manufacturers, who are not in that strict sense "dealers" in goods. "Deals" in section 14 (2), therefore, necessarily possesses a wider meaning.
A person "deals" in goods within the meaning of the subsection when he is prepared to sell or handle, or holds himself out as selling or handling, goods in the way of trade or business whenever ordered. Similarly, a person deals in goods of a particular description when he is prepared to sell or order goods of that description whenever ordered. If he has only one customer for goods of a particular description he remains a dealer; so also if he offers goods of a particular description but has no purchaser for them, he remains a dealer in goods of that description.
Let it be supposed that a shopkeeper has on his shelves a camera, a compass, and a clock, none of which type of article he has sold before. On selling one of them he is not excluded by the subsection from being a dealer in goods of "that description." If a manufacturer should make a tripod for the camera, a box for the compass and a case for the clock similar considerations apply. Thus, if a chemist were to open a new business and make up during his first week several prescriptions for different physical conditions and for persons unknown to him, each prescription containing the Same ingredients in different proportions (as here), it could not validly be maintained that any or some or all of those prescriptions were not goods of a description which it was in the course of the chemist's business to supply.
The respondents deal in various animal feeding stuffs. The fact that different feeding stuffs and compounds have different labels does not limit the status of the respondents in their capacity as dealers and sellers of feeding stuffs.
The fact that these goods were labelled after discussion with the respondents shows that the respondents were dealing in goods of "that description" or of that kind: see Benjamin on Sale, 8th ed. (1950), p. 644n.
If the Court of Appeal's construction be adopted then there is excluded *448 a contract, which is meticulous in its terms, because its subject- matter is a product which is being produced for the first time. This cannot have been contemplated by the Act.
Section 13. It has been the appellants' contention that: (i) The formula identified generically the ingredients and specifically the chemical additives and precisely quantified the proportion of each ingredient from which "King Size" was to be manufactured. (ii) DMNA was not a constituent of the formula. (iii) Furthermore, it was a substance which was not to be found in any of the ingredients included in the formula in their ordinary and natural condition. (iv) The addition of any other substance and in, particular DMNA was a departure from the contractual description; (v) DMNA was present in "King Size " in significant quantities, such as to render it toxic to mink.
The trial judge accepted the appellant's submissions and held that both the quantity and the quality of the contaminating substance can be relevant considerations, and that the test must in each case be whether the presence of the contaminating substance, either because of its quantity or its quality, renders the resultant admixture something to which the original description can no longer be properly applied. On this test the judge held that "King Size " which contained DMNA in quantities which rendered it toxic to mink did not comply with the description in Mr. Udall's formula, and that it was goods so described, that the respondents had undertaken to manufacture, sell and deliver. Accordingly, the respondents were in breach of section 13.
The Court of Appeal held first that there was no evidence which enabled them to form any clear view as to what might be described as the physical nature of DMNA. DMNA resulted from a chemical reaction when sodium nitrate was applied to herring and when heat was applied in certain processing conditions. There was no more justification for describing the result as "herring meal plus DMNA " than there would be for describing iron which had been oxidised as "iron plus rust."
But, in so far as the physical nature of DMNA was relevant, there was unchallenged evidence that DMNA is a liquid, freely soluble in water. In any event the physical nature of DMNA is irrelevant to the correspondence or otherwise of "King Size" to the contractual description. It is immaterial whether it be soluble or insoluble. The analogy with iron is inexact, but does not in any event support the proposition which it is used to illustrate. Iron which has been oxidised is not iron plus rust. It is not iron at all. It is rust, which is a different substance with different properties.
The Court of Appeal held, secondly: (a) That the appellants' argument involved the proposition that the balance of 325 tons of the material consignment of herring meal from the third, parties to the respondents was not properly described as herring meal. (b) That the balance of the consignment was "perfectly suitable" for use as herring meal.
But, (a) the extent to which the herring meal incorporated by the respondents in "King Size" corresponded to the description "herring meal" was no part of the appellants' case. The appellants' case was directed to showing that "King Size" mink food which poisoned mink, did not correspond with the contractual description of food. (b) There *449 was evidence, which was accepted below, first, that the "King Size" supplied by the respondents to the appellants was contaminated with DMNA, and, secondly, that the DMNA was present by virtue of the incorporation of Norwegian herring meal. (c) There was no evidence concerning the extent to which the balance of the consignment was contaminated by DMNA, if it was contaminated at all. (d) There was no evidence that the balance of the consignment was "perfectly suitable" for use as herring meal.
The appellants therefore submit that: (i) The contractual description described the goods in detail. (ii) The respondents were under a contractual obligation to supply goods corresponding with that description. (iii) The goods supplied did not correspond with the description in that they included DMNA in such quantity or strength as could not be dismissed as insignificant. (iv) The respondents were for the above reasons in breach of an implied condition pursuant to section 13 that the goods should correspond with the description, or, alternatively, that the goods should correspond with the contractual description.
A. G. Guest following. The points of law raised by section 14 are points of considerable difficulty and importance. If the history of section 14 is examined, it is apparent that at common law a broad meaning was given to the words that were embodied in the section and they were not limited in the way they have been by the Court of Appeal in the present case.
The following propositions derive from the authorities: (i) At common law the principles now embodied in section 14 (1) and (2) applied where goods were supplied either by a manufacturer or by a dealer in the way of his business as opposed to the situation where the seller sold the goods in a personal capacity. (ii) No distinction was drawn in, this respect between the application of the principle now embodied in section 14 (1) and that embodied in section 14 (2). (iii) The Sale of Goods Act 1893 was a codifying Act and was intended to and did embody the principles of the common law relating to sale of goods: Chalmers on Sale of Goods Act (1890) p. iv: "The Bill is almost entirely a reproduction of common law."
The particular print that it is wished to emphasise is that it is apparent from the 19th century cases that at common law there was no distinction between the word "description" as it subsequently appeared in section 14 (1) and (2): see Jones v. Bright (1829) 15 Bing. 533; Brown v. Edgington (1841) 2 Man. & G. 279; and Burnby v. Bollett (1847) 16 M. & W. 644. Strong reliance is placed on the judgment of Mellor J. in Jones v. Just (1868) L.R. 3 Q.B. 197, 202-203 where the question at issue is considered from the point of view of the status or capacity of the seller.
David Croom-Johnson Q.C. and Andrew Bateson for the respondents, Christopher Hill Ltd. The short answer to the appellants' claim under section 13 is that there was no breach of the implied condition as to description since what was included in "King Size" was to Mr. Udall's formula. This is the description relied on by the appellants in their own pleading, namely "'King Size' meaning thereby 'King Size' made according to Mr. Udall's formula." The formula included herring meal. *450 In considering poisoned food one has to ascertain whether there has been a loss of identity. Description is identity. The question of quality is dealt with in sections 14 and 15. It is emphasised that section 13 is concerned merely with description; section 14 contains an implied condition as to quality or fitness.
The following are examples of articles which, although suffering from contamination or adulteration, have not lost their identity: butter containing a very small amount of arsenic; margarine that has become rancid; a rusty iron bar. In all these cases the butter still remains butter, the rancid margarine still remains margarine and the rusty iron bar an iron bar.
Milmo J. relied on British oil Cake Co. Ltd. v. J. Burstall & Co. Ltd. (1923) 15 Ll.L.Rep. 46; Pinnock Brothers v. Lewis & Peat Ltd. [1923] 1 K.B. 690 and Robert A. Munro & Co. Ltd. v. Meyer [1930] 2 K.B. 312 but these cases are distinguishable, for in each of them there was the introduction of a substance into the article in question which made the goods sold lose their identity.
A case more analogous to the present is Ward v. Hobbs (1878) 4 App.Cas. 13. A pig suffering from typhoid fever does not cease to be a Pig.
As to section 14 (1) a purchaser of herring meal expects to get such substances as are reasonably and necessarily included in it for the purpose of making or preserving that which he is buying and so the inclusion of minute quantities of foreign substances would not import a misdescription. Some amount of preservation is inevitable. A purchaser getting herring meal by description would also get some sodium nitrite but it would still be herring meal by description.
The reliance on Christopher Hill by the representative of the purchasers did not go to the extent of relying on them to discriminate against a type of ingredient suitable for all normal purposes but, without any means of knowledge on their part, unsuitable for mink alone: see the finding of the Court of Appeal [1969] 3 All E.R. 1496, 1514-1515; [1969] 2 Lloyd's Rep. 425, 461. The admitted reliance of the purchasers on Chistopher Hill was to compound the food from such materials or ingredients as they normally used for their trade; they were given some discretion as to the source of those materials and ingredients according to availability and what they had in stock for their own trade.
The Cammell Laird case [1934] A.C. 402 is to be distinguished, since there the reliance placed by the purchasers on the skill and judgment of the manufacturers was that in the exercise of that skill and judgment they should produce a propeller suitable for ship No. 972. There was no equivalent reliance in the present case. The reliance was for the vendors to compound the food efficiently and to use the materials they had in stock. In the Cammell Laird case the defect was within the scope of the reliance placed on the manufacturers' skills and judgment. But in the present case the purchasers considered that they were in the position that if they had heard that there had been trouble over herring meal in Norway they would have been within their rights in telling the vendors that they wanted no ingredients from there. Accordingly this case did not come within section 14 at all. The goods in question were "King Size" mink food. There *451 was some relation between the description of these goods and whatever reliance was placed by the purchasers on the vendors. That relation was in any view, more limited than the Court of Appeal found.
On the evidence there are two possible views: (1) that the representative of the purchasers placed no reliance on the skill and judgment of the vendors or (2) that there was a partial reliance but one insufficient to support this claim. Either view is a matter of inference.
There was no reliance. The representative of the purchasers considered himself entitled to object to ingredients from any suspect source on the ground of inferior quality, so as to keep the safety of the mink under his own control. Though the vendors were left with a discretion as to the sources of the materials, this was subject to the right of veto by the purchasers' representative. He went to them because they had the "know how" and the facilities for compounding animal food but no warranty could be based on that. In the Cammell Laird case [1934] A.C. 402 no corresponding objection could have been taken to the manganese bronze.
The indications are that the representative of the purchasers was the expert in mink food, while the vendors had no knowledge of it. He produced the formula and would allow no change in it or in the ingredients without his approval. He took nothing on trust and there is no evidence that he relied on the vendors to do anything more than compound the food effectively from such stock as they had in hand or might buy for feeds for the animals they knew about. If there had been a partial reliance on the vendors it could only have been in respect of the stock which they normally used for themselves.
There was no breach of any warranty that the herring meal was reasonably fit for the purchasers' purpose, because in order to impose an obligation of reasonable fitness the purchasers would have had to make known the purpose for which the goods were required: see Griffiths v. Peter Conway Ltd. [1919] 1 All E.R. 685, 691. One thing never made known to the vendors by the representative of the purchasers was that mink have a particular sensitivity in the matter of herring meal. Herring meal normally fed to animals causes them no harm, but it does cause harm to mink. There was evidence that both herring meal and fish meal were normal foods for animals both in England and Norway.
In section 14 (1) of the Sale of Goods Act the words "the goods are of a description," etc., have a function in the making of the contract. The words which go before show the buyer's reliance on the seller's skill and judgment in making known the purpose for which the goods are required and the subsection concludes with the implication that the goods will be reasonably fit for that purpose. In looking in detail at the seller's business to see whether it is in the course of his business to sell goods of that "description," it is a question of fact in each case. One does not answer the question by finding that he is doing it for money or trade; one must look at the whole course of the business. One cannot simply take a broad genus of goods, the goods the seller deals in. The goods at which one is looking here are "King Size" animal food and not just herring meal. Thus if a trader dealt in soft furnishings that might include linen and carpets, but if he only sold carpets it would be wrong to treat that as including soft *452 furnishings. Cricket goods and golf clubs can be called sports gear, but if a trader dealt only in cricket goods and a customer persuaded him to get him some golf clubs, that would be a thing outside his business and there could be no implication under section 14 (1). If someone who made pottery figures of dogs were persuaded by someone to make him a large pottery jar outside his usual range and quite different in size and thickness, that could also be something which it was not in the course of his business to supply. Even within the range of animal figures a man who made horses and poodles might not be able to make a giraffe with a long neck. So in the present case one does not solve the problem by saying that the seller deals in animal foodstuffs.
The seller's course of business must bear some relation to the reliance by the buyer which he is being asked to accept. One must not imply reliance on the skill and judgment of a seller who does not profess to have expertise. The reliance on the skill and judgment must properly be personal. The implication of a condition under section 14 (1) must not go beyond the buyer's reliance on the seller's warranty.
The Spencer Trading case [1947] 1 All E.R. 284, relied on by the appellants does not help in the present case. The decision was right on its facts, but lays down no general rule. The arguments of the parties on this point were correctly summarised by the Court of Appeal: [1969] 3 All E.R. 1496, 1516; [1969] 2 Lloyd's Rep. 425, 462. The court was right in concluding that it had not been proved that any part of the shipment was unfit for animals other than mink.
The "description" must be suitable to the individual bargain and find what is the real course of the business for the purposes of that bargain. Here the bargain for the feeding stuff must be related to the animals the supplier knows about. Only in the case of them can there be the implication under section 14 (1). In the case of these suppliers their course of business was confined to the animals on their list. The feeding-stuffs for them were the goods of the description within the scope of the implication, and the reliance should go no further. One must rule out animal foodstuffs for animals of which the suppliers knew nothing, including wild animals in captivity.
As to section 14 (2), if one is dealing with a one-purpose product like "King Size" and the buyer makes known its purpose under section 14 (1) and yet fails to recover judgment under the subsection, even though the goods are not suitable, it would be illogical for him to recover under subsection (2), because the goods are not merchantable as mink food. Again the question is whether the goods were bought from a seller who dealt "in goods of that description." The Court of Appeal have held that "description" in subsection (2) referred to the contract description: [1969] 3 All E.R. 1496 , 1517- 1518; [1969] 2 Lloyd's Rep. 425, 463-464. These respondents adopt the Court of Appeal's finding on this point and submit that the word "description" as first used in subsection (2) must he read with the same word later in the subsection and accordingly does not have the same meaning as "description" in subsection (1): see the Hardwick Game Farm case [1969] 2 A.C. 31, 76-77. Subsection (2) is dealing with merchantability and not with the question whether the goods *453 are fit for the buyer's purpose. As to the meaning of "merchantable quality" see Sumner Permain & Co. v. Webb & Co. [1922] 1 K.B. 55, 63. This construction makes sense of the relationship of the two subsections. Subsection (1) relates to the buyer's special needs and purpose, but subsection (2) does not because the goods may still be merchantable, even though they do not fit the buyer's needs; its concern is with the quality of the goods. Subsection (1) relates to the buyer's reliance on the seller's expertise in the field of what the buyer wants in this particular instance, while subsection (2) relates only to the general merchantability of the goods referred to under the contract description, so that the buyer's special needs are irrelevant: see Jones v. Just (1868) L.R. 3 Q.B. 197, 203 cited by the appellants. Shepherd v. Pybus (1842) 3 M. & G. 868 did not go so far.
In subsection (2) "goods of that description," viz., in which the seller deals, refer to the contract description and in the present case goods of that description were something in which the seller did not deal so that the subsection was not brought into play. Even if on the construction of subsection (2) "description" has the same meaning as in subsection (1) the description in the present case does not go beyond mink food and the submissions under subsection (1) are repeated. What one is looking at is the actual dealing by the seller. For the present purpose that means asking: What goods did the sellers deal in before 1960? That takes one back to the seller's list of the goods in which they were expressly stated to deal and held themselves out as dealing, foodstuffs for certain known animals mentioned and limited to them. To make up food for an unknown animal is something quite different. On the facts the sellers in their business did not do the sort of thing they were required to do under the contract. Burnby v. Bollett (1847) 16 M. & W. 644, 648 (Alderson B.). 654 (Parke B.) suggests that the basis of the rule in section 14 (2) is the public good in maintaining the standard of merchandise warranted by the person who deals in these particular goods. What was said in the first paragraph of the introduction to the 1st edition of Chalmers on Sale of Goods (1894) is adopted. Section 14 (2) must be construed according to the ordinary canons of construction.
Andrew Bateson following. The trial judge held that the sellers had warranted the herring meal as fit for feeding all animals, including mink. The Court of Appeal held that the warranty was for the purpose of feeding all animals other than mink, and the judge held that the sellers expressly stated that they knew nothing of the feeding requirements of mink. If the warranty was as found by the Court of Appeal, it was for a totally different purpose and there was not in any sense a partial reliance by the buyers on the sellers. There was no reliance to provide feeding-stuff for feeding animals generally. If there were such a reliance a mink farmer claiming for breach of contract on the basis of an implied term would have to say that he was entitled to recover damages for the death of his mink because the food supplied would have killed cows or poultry. He could recover for breach of contract in respect of food which killed his mink because it would have killed cows or poultry, but not simply because it killed mink. The death of the mink would be irrelevant in deciding whether the food *454 was deleterious. That is not what is meant by a partial reliance: see the Cammell Laird case [1934] A.C. 402 , 422.
This food was required, not for animals generally, but for the particular purpose of feeding mink. If a farmer invites these sellers to compound a foodstuff, part of the ingredients of which are supplied by him, there would be a partial reliance on the sellers in respect of the ingredients which they put in that would be the true analogy to the Cammell Laird case. There may be a division of reliance. But it would be a considerable departure from the principles of section 14 (1) to relate a specific purpose to a general purpose and call that a partial reliance. The purpose is the general one of feeding cows and other animals. On the basis that it is fit for feeding those other animals the sellers cannot have broken a warranty (namely, fitness for feeding to mink) which they did not give because the goods are unfit for feeding to some other animal.
But here one is dealing with mink. A food might injure other animals and yet be harmless to mink. On that basis the question would be, not whether someone else's animals would be killed, but whether the buyer's animals would be killed.
G. R. F. Morris Q.C. in reply. There is no evidence that any herring meal but the Norwegian was poisoned. This Norwegian herring meal is in a separate category.
Section 2 (3) of the Fertilisers and Feeding Stuffs Act 1926 and the definition of white fish meal in Part II of Schedule 4 thereto indicates that if any part of the meal in the present case had been tainted with nitrate it could not properly be called fish meal. The statutory definition required that the goods should be called herring meal with a qualification, since other matter had been added. The herring meal with the addition was toxic. Fish meal was the genus and herring meal the species. The nitrite was meant to have a bacteriological effect, in preserving the freshness of the herring before it was processed. But the effect of the heating process created an additional ingredient in the meal.
There was no evidence that mink had a special sensitivity to DMNA. The effect of the evidence was that all animals were sensitive to it but that mink were most sensitive. There was no evidence on which the court could hold that the food was suitable.
There was no evidence that DMNA was not toxic to domestic or captive animals, That was not disputed at the trial which was fought on the issue whether or not the disease the mink contracted was attributable to it. The Court of Appeal sought to distinguish mink and wrongly attribute to them sensitivities not shared by other animals.
It is submitted: (1) No issue was defined on the pleadings to suggest that mink fall into any special category qua animals; (2) No evidence was given to suggest that they were in such a category; (3) The unchallenged evidence was that DMNA was toxic to all animals; (4) The scientific and field evidence showed that a number of factors, some known, some imponderable, affected the manifestation of damage within the animal but that damage could occur in the animal over a period of years and only be detected on a post mortem; (5) The animals in respect of which evidence was given at the trial included cows, calves, goats, sheep, chickens, mink, ferrets, ducks and foxes. These included experimental and free animals *455 to which toxic herring meal had been fed both before and after DMNA had been fed. By 1968 the cumulative experience was that DMNA was toxic to all animals without qualification and hazardous to man.
Any positive finding that the herring meal in "King Size" could have been fed to other animals without damage would be speculative having regard to the scientific evidence, it would be permissible to find that it was poisonous.
In section 14 (2) of the Act of 1893 the word "description" may be used in a general sense or the word "deals" in the ordinary general sense so as to include dealing in goods of that description. Taking the words "where goods are bought by description" the fact that the seller gets the goods to dispose of makes him a dealer before he disposes of them. The manufacturer is also a dealer. Dealing must involve the concept of preparing for the sale by acquiring the goods and, having done that, one is a dealer. The next step in sequence is when the goods are bought. The status of dealer is as necessary to give effect to section 14 (2) as "goods of that description." The most sensible approach is to treat "description" in section 14 (1) and (2) as referring to goods of that kind in a general sense. A restrictive approach has never been taken historically. In relation to "description" under section 14 (1) the expression covers the nature of the goods with which we are concerned, animal feeding- stuffs and and compounds. Under section 14 (2) "goods of that description " refer to the formula and the ingredients compounded. As the Court of Appeal said [1969] 3 All E.R. 1496, 1511: [1969] 2 Lloyd's Rep. 425, 457:
"The essence of the description, then, was that the goods were to consist of the ingredients specified in the formula, in the proportions stated therein."
David Croom-Johnson Q.C. replying on the Fertilisers and Feeding Stuffs Act 1926. The definitions in the Schedule indicate that one must not have added anything to the finished manufactured product. The Schedule is not saying that in no circumstances must the materials have had anything added to them. This is supported by the wording of the definition of "white fish meal" in Part II of Schedule 4 to the Fertilisers and Feeding Stuffs Regulations 1960 (S.I. 1960 No. 1165). The word "and" is left out, thereby bringing it into line with the other definitions. The evidence was that at the date of these regulations it was within the knowledge of the Ministry that the use of preservatives in fish meal was a common practice.

[SECOND APPEAL]

David Croom-Johnson Q.C. and Andrew Bateson for the appellants. The questions arise: (1) whether the goods supplied were "Norwegian herring meal, " corresponding to the contract description, within section 13 of the Sale of Goods Act 1893; (2) whether "fair average quality of the season" was part of the contract description, as these appellants contend; (3) whether, since, on the evidence, DMNA was not to be found in far the greater part of Norwegian herring meal in that season, its presence prevented this shipment from being "fair average quality of the season"; (4) whether latent defects are a breach of the requirement *456 of "fair average quality," so that if anything, whether patent or latent, results in the goods not being up to that quality, there is a breach; (5) whether, if there is a breach of the condition, the third parties are protected by general condition 3 of the contract; (6) under section 14 (1) of the Sale of Goods Act 1893 there arise the questions (a) whether "the particular purpose" for which the goods were required was made known; (b) whether, if there was no express making known of that purpose, the appellants made it known by implication, so as to show that they relied on the respondents' skill and judgment; (c) whether, the particular purpose being that the herring meal was required as animal foodstuff to be fed to all animals to which it might properly and reasonably be fed in reasonable quantities, mink were within that class of animals, a major factor in the determination of that question being whether these respondents could and should have had in mind the fact that Norwegian herring meal had been so used long before 1961; (d) whether, if the particular purpose was as assumed, the herring meal supplied was reasonably fit for it; (e) whether the damage was too remote.
On question (1) Christopher Hill adopt the arguments raised against them on section 13 by Ashington Piggeries. The next point of inquiry is; What was the description in the contract? The contract is on a printed form put out by these respondents. They therefore had the choice what should and should not go into it. It is submitted that the description goes down to the end of the clause. Quality may form part of the description.
Accordingly, on question (2), "fair average quality of the season" is part of the contract description.
On question (3), this shipment was not fair average quality of the season. The test is stricter if "fair average quality" is part of the description than if it were merely a warranty,
On question (4), if there is a breach of description it does not matter whether it was patent or latent; see Randall v. Newson (1877) 2 Q.B.D. 102, 109. The point was dealt with in the Court of Appeal [1969] 3 All E.R. 1496, 1521; [1969] 2 Lloyd's Rep. 425, 466, 467. But it is wholly inappropriate to apply the test of the distinction to defects of quality in section 13 where quality is part of the description. The Court of Appeal quoted from James Drummond & Sons v. E. H. Van Ingen & Co. (1887) 12 App.Cas. 284, 297, but the quotation was taken out of context: see pp. 296-297. The cases cited by the Court of Appeal do not support a proposition of the width which it lays down. The failure to comply with the description would give a right of rejection. The quality which one looks for in herring meal is its nutritional value. The quality of processing would determine whether it was of "fair average quality. " It might contain all sorts of deleterious things which might not be found by the trade analysis stipulated for in the contract. W. N. Lindsay & Co. Ltd. v. European Grain & Shipping Agency Ltd. [1963] 1 Lloyd's Rep. 437, 441, 445 gives some guide on "fair average quality," though it is not directly on the present problem. Steels & Busks Ltd. v. Bleecker Bik & Co. Ltd. [1956] 1 Lloyd's Rep. 228, 229, 235, 236, 237 bears on the point of the compliance with the contract description. It is authority only for the propositions (1) that if the arbitrator in a particular trade dispute found as a fact that the goods comply *457 with the description, that the description included quality and that the custom of the trade was that for all purposes quality is judged by visual examination only, the finding will not be upset and (2) that if the sale was also by sample, then, despite visual examination, the other provisions of the Sale of Goods Act would apply. In the present case the evidence did not support existence of any special trade custom. There is no reason for giving "quality" any other meaning than its full meaning with all its attributes: see also F. E. Hookway & Co. Ltd. v. Alfred Isaacs & Sons [1954] 1 Lloyd's Rep. 491, 492, 494, 498, 503, 509-510, 511, which is really a section 15 case. Once a breach is established, one does not look to see how small is the quantity of the deleterious substances. If the description is qualitative, the breach is qualitative. It is the effect of the departure from the quality which matters. The statistical approach inherent in the de minimis test is inappropriate. A typhoid germ is minute but its effect is catastrophic. If "fair average quality" did not cover latent defects, the buyer would be left without a remedy.
On question (5) these appellants have findings in their favour in both the courts below as to the meaning and effect of general condition 3. In the Court of Appeal see [1969] 3 All E.R. 1496, 1522; [1969] 2 Lloyd's Rep. 425, 468. The condition is an exclusion clause put in by the seller for his own protection and as such must be construed narrowly. All it amounts to is that where the buyer has a right to look at the goods to see whether they comply with the contract, that is the moment when he must take or reject them. It does not exclude his right to recover in respect of loss if they turn out to be defective.
As to the first of the four points arising under section 14, whether the particular purpose for which the goods were required was made known to the sellers, see Hardwick Game Farm v. Suffolk Agricultural Poultry Producers Association [1969] 2 A.C. 31, 93, 105, 125. On the evidence the purpose was made known to the sellers.
As to the second point, whether, if it be held that there was no express making known of the purpose, it was made known by implication, see Mash & Murrell Ltd. v. Joseph I. Emanuel Ltd. [1961] 1 W.L.R. 862, 867. The same inference should be raised in the present case, because it would make sense in the circumstances. In view of the Norwegian regulations, it would be sensible that the manufacturers should carry the risk rather than the consumers. The inference of reliance on the skill and judgment of the sellers should be drawn because: (a) it is natural to draw it in the circumstances; (b) the trial judge found on the evidence that herring meal forms part of the diet of mink and was properly so used; (c) there was no peculiarity about herring meal which rendered it proper food in one country but improper in another and these appellants bought it as a foodstuff for any animal in the diet of which fish meal was a material for compounding: (d) the sale from the sellers' agents to these appellants in England was without reservations, just as the sale in Norway was without reservations. This was a relevant factor: see the Hardwick case [1969] 2 A.C. 31, 95, 125; (e) the agents had not sold Norwegian herring meal to these appellants for some time; (f) the sale was by the co-operative of manufacturers themselves, and so there was an even stronger case for making the implication than in Hardwick.
*458 As to the third point, whether mink were within the class of animals to which herring meal might properly be fed, mink were included in that particular purpose. Using herring meal as a foodstuff for animals was an ordinary and reasonable use. At the time of this contract it was being used as a food for mink in England. The particular purpose for which the herring meal was required was to feed animals generally and not only the animals within the Fertilisers and Feeding Stuffs Act 1926. Feeding herring meal to pheasants would be an ordinary use, though they are not within the statutory definition of poultry. For section 14 (1) to apply the seller must be apprised of any abnormality in the object to which the goods are to be applied: see Griffiths v. Peter Conway Ltd. [1939] 1 All E.R. 685, 691. These respondents could have foreseen that the herring meal would find its way to mink. They knew that mink fur farming was an important Norwegian industry and that a product of another important Norwegian industry, herring meal, was being fed to them. The trial judge found this, after seeing the witnesses and the Court of Appeal was wrong to reverse this finding. Even if the respondents did not know this, they ought to have known it. If they had stopped to think, they would have known that this herring meal might find its way into the stomach of an animal.
As to the fourth point, this herring meal was not reasonably fit for the particular purpose: see the finding of the trial judge [1968] 1 Lloyd's Rep. 458 , 487. Even if only a few tons were contaminated, that was a large enough proportion to render the whole cargo unfit.
Fitness for one purpose coupled with unfitness for another is not sufficient to establish suitability. The degree of harmfulness is relevant. Food would be unsuitable even if only on very rare occasions it killed a consumer: see the Hardwick case [1969] 2 A.C. 31, 114-118.
As to the fifth point of remoteness of damage, if feeding to mink was a reasonable use of herring meal and Norwegian herring meal plus DMNA was harmful to animals, then the results produced in this case were not too remote: see the approach adopted in the Hardwick case at pp. 91, 103, 111, 117, 126. If feeding herring meal to mink was a normal use of it and this herring meal was poisonous to them, though it was only potentially poisonous to other animals, the damage was not too remote. The sellers should reasonably have contemplated that there was a serious possibility of that result: see C. Czarnikow Ltd. v. Koufos [1969] 1 A.C. 350, 385. If a cook served porridge with arsenic in it to a family of eight adults and a child, and only the child died, though the food was deleterious to all, the damage would not be too remote. But if the child, not liking the taste of the arsenic threw his shoe through the television set, that damage would be too remote. Similarly, if these mink, instead of dying after eating this herring meal, had got excited and broken out, doing damage, that would have been too remote.
Anthony Lloyd Q.C. and A. B. R. Hallgarten for Norsildmel, the respondents in the second appeal. The issues between these parties are (1) whether these respondents were in breach of section 13 of the Sale of Goods Act; (2) whether they were in breach of section 14 (1) of the Act; (3) whether they were protected by the exceptions in the contract; (4) whether the damage was too remote. It is no longer alleged that these respondents *459 were in breach of section 14 (2). It is conceded that the herring meal was reasonably fit for purposes other than that for which it was used.
As to the first question, whether the respondents are in breach of section 13, one must first ask: was this Norwegian herring meal and, if it was not, what was it? One must not confuse sections 13 and 14. This is not a question of degree but of kind. An admixture of 0.0008 does not make a thing different from wallet it is without that admixture. The proper description of the substance would be "contaminated herring meal." Unless there is something in the contract against contamination, the respondents have complied with their obligations under section 13 and escape liability.
"Fair average quality" was not part of the description. The case has been run on the line that the goods did not comply with the description because they were not "fair average quality." But the description is simply "Norwegian herring meal." One might say of every warranty in a contract that it is descriptive of the goods but "description" in section 13 refers to the language which the parties have used to identify the goods which are the subject-matter of the contract. The definition of "warranty" in section 62 of the Act assumes that the goods have been identified and agreed upon: see Chalmers' Sale of Goods, 15th ed. (1967), pp. 202-203. The function of section 13 is to do for unspecified or future goods what the parties do for themselves in the case of specified goods. If a seller delivers a car which is not the car agreed at the time of the contract, the exceptions clause would not apply. Similarly in the case of unspecified goods, if the goods delivered are not those agreed upon: see Chalmers' Sale of Goods, p. 201.
The goods to which the present contract relates are Norwegian herring meal. If these appellants had been asked what was in the herring meal bin in their store, they would have answered: "Norwegian herring meal" and not "fair average quality Norwegian herring meal." If the respondents had delivered fish meal, clause 3 of the exceptions would not have applied, not because of a breach of an implied term under section 13, but because the contract was not about fish meal. So too if they had delivered German herring meal: see W. N. Lindsay & Co. Ltd. v. European Grain & Shipping Agency Ltd. [1963] 1 Lloyd's Rep. 437, 438, 443.
Prima facie "fair average quality" indicates the quality, not the description. of the goods: Pacific Trading Co. Ltd. v. Wiener (1923) 14 Ll.L. Rep. 51, 54. The "quantity and description" part of the contract contains words which it would be hard to say were part of the description of the goods. The protein content is not part of the description. If it were, it would leave no scope for analysis: see Milmo J. [1968] I Lloyd's Rep. 457, 488-491. Precisely the right test is expressed in Varley v. Whipp [1900] 1 Q.B. 513, 517. That case is relied on for what is said as to identification of the goods as opposed to mere collateral warranty: see also MacPherson Train & Co. Ltd. v. Howard Ross & Co. Ltd. [1955] 1 W.L.R. 640.
The question has been raised whether the presence of a latent defect necessarily means that the goods are not "fair average quality." It begs the question to describe the I presence of DMNA as a latent defect. It is a defect vis-a-vis mink, but not so far as cattle, pigs and poultry are concerned. It is conceded that it did not harm some animals. The appellants' *460 claim that this herring meal was unsuitable for animals in general has wholly failed on the evidence. Admittedly all animals are sensitive to DMNA if they are given enough of it, but this amount was harmless to them. Substances are not to be held toxic if they are harmful only in large quantities. Cigarettes are not ordinarily regarded as poisonous, though nicotine taken in large quantities is. So too with alcohol. On this point there are concurrent findings of fact in favour of the respondents. It begs the question to say that I the proscenia of DMNA was a latent defect, because it was only a defect so far as mink were concerned, its inclusion in the herring meal being harmless to other animals.
Quality is not an abstract thing; it must be judged by ordinary usage. If the respondents had not waived the arbitration clause the question of fair average quality would have been determined by an arbitrator, who would have seen a sample and had it tested. If while it was being so tested he had been told that there was something deleterious in it he might have asked whether it would harm cattle or poultry and, when told it would not, he would have passed it on the basis of normal user as "fair average quality." In deciding that, one is not concerned with the undetectable but with the ordinary qualities which would be present to the eye or could be detected on a reasonable examination: Pinnock's case [1923] 1 K.B. 690, 697. The question is not what "quality " means on its own but what it means in the context of this contract. Hookway's case [1954] 1 Lloyd's Rep. 491 was not laying down what quality could mean in a general way but what it meant in the particular contract: see p. 511. "Quality" in other contexts might include latent defects. "Fair average quality" is not a very precise expression. In Drummond's case (1887) 12 App.Cas. 284, 287-288, Lord Selborne approached the question as one of construction of the contract. Even if Hookway's case were, as the appellants contend, a section 15 case, that would make no difference. If there were a clause "quality as per sample," quality would mean apparent quality. The meaning of the word cannot differ according as it is used separately or as part of the description. If "fair average quality" is held not to cover latent defects there is no hardship on the buyer since if those defects make the goods unmerchantable, the buyer would have a remedy under section 14 (2); and if the special purpose for which they were required was made known to the seller, the buyer would have a remedy under section 14 (1).
Groupement National d'Achat des Tourteaux v. Sociedad Industrial Financiera Argentina (The Milton B. Medary) [1962] 2 Lloyd's Rep. 192, 197 is also relevant.
Summarising the submissions under section 13:
(1) The description of the goods sold was "Norwegian herring meal." The description of the goods delivered was "Norwegian herring meal," though the herring meal was contaminated with DMNA.
(2) "Fair average quality" was not part of the description, but even if it was, this herring meal was "fair average quality," (a) because the presence of DMNA was harmless for the ordinary purposes for which herring meal was used; (b) because it is only concerned with apparent quality and not with latent defects; (c) because there was no evidence that *461 it was not "fair average quality; and (d) because there was no real evidence that it was different in any way from any other herring meal of the season.
As to the second question, whether the respondents were in breach of section 14 (1):
(i) This is not a section 14 (1) case at all, because no particular purpose was made known to the sellers. There was no mention of mink from start to finish. Only animal feeding-stuff was referred to. That was not sufficient to bring section 14 (1) into operation.
(ii) This is not a section 14 (1) case at all, because, if the purpose of the feeding-stuff is to be treated as feeding mink that purpose was not made known so as to show reliance on the sellers' skill and judgment for that purpose.
(iii) If a particular purpose was made known it was the purpose of use as an animal feeding-stuff and the herring meal was reasonably fit for that purpose. The wider the purpose made known, the easier it is to comply with it.
(iv) If normal use be relevant at all under section 14 (1), then compounding into mink food was not a normal use for herring meal within the United Kingdom or elsewhere. The question is not whether it was normal food for mink, but whether it was a normal use for herring meal.
As to submission (i), these goods were sold generally on the market and no particular purpose was made known. The obligations on the sellers were (a) that the goods comply with the description and be fit for their usual purpose and (b) that they should be merchantable, which is conceded. If the buyer wants to make sure that goods are fit for a particular purpose, he must state what it is. It was never stated that the purpose was to feed mink, which was the only purpose for which it was unfit. There may be some things for which there is only one possible use. Section 14 (1) may come into operation by implication, as where the use for which the buyer wanted the goods is obvious from previous dealings between the parties or from the nature of the buyer's business.
If it were the case that goods sold generally must be fit for all purposes for which they are normally used, section 14 (2) would be otiose, since under section 14 (2), goods sold generally have to be merchantable, i.e., fit for one purpose only. It would make nonsense of the relationship between subsection (1) and subsection (2). If as a matter of law the goods really did have to be fit for all normal purposes no buyer would make any particular purpose known to the seller, because he could only thereby narrow the seller's liability. Jones v. Padgett (1890) 24 Q.B.D. 650, 651, 654 illustrates the fallacy behind the argument for the appellants: see also Brown v. Edgington (1841) 2 Man. & G. 279. As to what is required to bring a particular purpose in, see the Hardwick Game Farm case [1969] 2 A.C. 31, 80 (Lord Reid). See also p. 123 (Lord Wilberforce) as to whether the buyer has told the seller enough to enable him to make up his mind what is to be supplied. The purpose must be stated with sufficient particularity in order to bring section 14 (1) into operation at all, because otherwise there would be nothing on which the skill and judgment of the seller could operate. See also p. 93 (Lord Morris of Borth-y-Gest). Before the Hardwick case Manchester Liners Ltd. v. Rea Ltd. [1922] 2 A.C. 74 was the leading case. In the present case unless the sellers were told *462 that the herring meal was intended for feeding mink, they were entitled to deliver herring meal suitable for animals generally: see also Griffiths v. Peter Conway Ltd. [1939] 1 All E.R. 685, 691.
Suppose a buyer went to a retailer to buy some carpet, without saying what he wanted it for, and the seller sold him carpet suitable for light domestic use, which he put on the floor of a restaurant, where it wore out in a week. If the appellants were right the buyer could succeed in a claim if he had said that he needed it for floor covering. "Floor covering" has as much particularity as "animal feeding-stuff" but it would not be sufficient particularity in that it did not specify the vital thing, i.e. that it was wanted for a restaurant. This was not a section 14 (1) case because there was not sufficient particularity.
As to submission (ii), that, even if there was sufficient particularity, there was no realliance on the seller's skill and judgment for that purpose. The following points are relied on: (a) nothing was said about mink at the time of the sale; (b) no evidence was called to the effect that the buyers were in fact relying on the sellers to ensure that the herring meal was fit for mink; (c) mink food was right outside the sellers' normal business: this was the first compound food for mink manufactured in the United Kingdom; (d) the sellers had no idea that this would go into mink food and they knew nothing of the buyers save that they were compounders in a general way of business; (e) if the sellers' representative had looked at the list of products manufactured by the buyers he would have found no reference to mink food; (f) Norwegian herring meal was not a new line of product which the sellers were supplying to the buyers, who had bought it before; (g) if either the buyers or the sellers had known about the mink (which neither of them did) they would have known that herring meal was regarded as unsuitable for mink food, not only because of DMNA, but also because it had too high a fat content. Reliance is placed on Teheran-Europe Co. Ltd. v. S. T. Belton (Tractors) Ltd. [1968] 2 Q.B. 545, 554, 560, 563.
As to submission (iii), if the particular purpose was made known and that purpose was use as an animal feeding-stuff, it is accepted that there would be reliance on the sellers' skill and judgment and section 14 (1) would operate. But there was no breach because the herring meal was reasonably fit for that purpose. The wider the purpose made known, the easier it is to show that the goods were fit for it.
"Toxic" means nothing in the abstract, but only in relation to a particular quantity and purpose. For example, salt or aspirins are toxic if taken in excess, and arsenic may be beneficial for certain purposes if taken in minute quantity.
The herring meal was reasonably fit for use as animal feeding stuff, because it was fit for all animals except mink.
As to submission (iv) if it is necessary to decide whether feeding to mink was a normal use of herring meal, it is accepted that the Court of Appeal upset the findings of the trial judge. In that they were justified. This is not a case in which the Court of Appeal were at a disadvantage in not seeing the witnesses: see [1969] 2 Lloyd's Rep. 425, 471, 475.
In order to succeed on this point the appellants must show at least that *463 herring meal was normally used for feeding mink (a) in England and (b) in Norway. The appellants did not succeed in establishing either point.
As to the third main question, whether the respondents are protected by the exceptions in the contract, the provision in the general conditions that the goods are to be taken "with all faults and defects" casts the risk of those defects on the buyers. The clause gives the buyers some remedy in a reduction in price if the goods are damaged or inferior: see Lloyd Del Pacifico v. Board of Trade (1929) 35 Ll.L.Rep. 217, 220, 221, 222, 225 and Couchman v. Hill [1947] K.B. 554, 557, 559. The court should not invent a meaning for words to avoid the well-established interpretation of the "faults and defects " clause.
On the final question, the claim for damage under section 14 (1) and (2) is too remote. For such a claim to succeed there must be some likelihood of what happened. It is not enough that it should be possible or reasonably foreseeable since the test is not the same as in tort: see Czarnikow's case [1969] 1 A.C. 350, 383, 387. Even if the respondents were wrong on every other point, they must succeed on this.
David Croom-Johnson Q.C. in reply. What was the description in the contract? What was the effect of "fair average quality" in it? The simplest statement is in Halsbury's Laws of England, 3rd ed., Vol. 34 (1960), para. 74, pp. 48-50. Goods sold may at one and the same time infringe section 13 (description), section 14 (1) (particular purpose) and section 14 (2) (merchantability): see Grant v. Australian Knitting Mills Ltd. [1936] A.C. 85, 100 and Arcos Ltd. v. E. A. Ronaasen & Son [1933] A.C. 470. As to the right to reject the goods, see Meyer Ltd. v. Osakeyhtio Carelia Timber Co. Ltd. (1930) 37 Ll.L.Rep. 76, 78, 79, and in re An Arbitration between North Western Rubber Co. Ltd. and Hüttenbach & Co. [1908] 2 K.B. 907. As to the "fair average quality clause," see Pacific Trading Co. Ltd. v. Weiner (1923) 14 Ll.L.Rep. 51 . As to breach of a condition that the goods should correspond to description, see Wimble, Sons & Co. v. Lillico & Son (London) (1922) 38 T.L.R. 296 .
In looking at the description clause in this case one should not carve the words relating to quality out of the description on the ground of the relationship between sections 13 and 14. This is not an L.C.F.T.A. form as in Lindsay's case [1963] 1 Lloyd's Rep. 437 . The form was sent to the appellants before the contract was made. The side notes which correspond to each clause must be taken into account. Part of the description clause is the expectation of analysis. The "fair average quality" clause was necessary because, if the protein was less than 70 per cent., it would not give a right of rejection. It would be grammatical nonsense to say that "fair average quality" was not part of the description: see Macpherson Train [1955] 1 W.L.R. 640, 641. As to the words "expected to analyse" in the description, see Sanday & Co. v. Keighley, Maxted & Co. (1929) 91 L.J.K.B. 624 and The Mihalis Angelos [1971] 1 Q.B. 164, 192, 193, 194, 197, 205. "Expected to analyse" is a condition. It is part of the description.
It would be strange if "fair average quality" were not part of the description. What was being bought was Norwegian herring meal expected to be not less than 70 per cent. protein. A lower percentage would be no use to the appellants for simple compounding reasons. They would assume that the protein would be of a certain percentage, which could be put with *464 their normal productions without recalculation. They wanted to be as near certainty as they could get. That was what mattered to them. "Fair average quality of the season " was part of the description identifying the herring meal. If "fair average quality" was not part of the description but was merely a warranty, the sellers could deliver any herring meal which was no use to the buyers because of its low protein value, on a reduction of the price. But the appellants could not be required to accept an inferior herring meal at a valuation, because they did not want to recalculate their formula. The words in the clause should be given their ordinary meaning as part of the description of the herring meal.
On the question whether this herring meal has been shown as a fact not to be "fair average quality" both the trial judge and the Court of Appeal have found that it was not.
The Milton B. Medary [1962] 2 Lloyd's Rep. 192, 196, 197, was not an authority for the proposition for which it was cited by the respondents. Drummond's case (1887) 12 App.Cas. 284, 287-288; Hookway's case [1954] 1 Lloyd's Rep. 491 and Steels & Busks' case [1956] 1 Lloyd's Rep. 228, 229 were sample cases.
The difference in the present case is that here under the appellants' contract with the Norwegian sellers there was nothing with which any comparison could be made. In construing this contract according to its ordinary meaning one does not get any help by looking at sample cases in which "fair average quality" was not mentioned.
As to condition 3, on the basis that "fair average quality" was part of the description, Munro's case [1930] 2 K.B. 312 could apply. But the probability is that a good proportion of the bags contained DMNA and if the presence of DMNA involves a breach of "fair average quality" condition 3 will not bite. See Hardwick Game Farm in the Court of Appeal [1966] 1 W.L.R. 287, 344.
The clause is an exception clause and must be read restrictively against the person relying on it, not giving it a meaning beyond that which it is necessary to give it.
In the construction of a condition like this it would be wrong to let an exceptions clause control the description. If the appellants are right, it need not be otiose. It has a function in an application to the implied condition of merchantable quality under section 14 (2). Lloyd Del Pacifico (1929) 35 Ll.L.Rep. 217 was decided on very special facts and does not detract from the scope of this clause.
As to the question of remoteness, see Czarnikow [1969] 1 A.C. 350, 384, 385, 405.
As to section 14 (1) see the Hardwick case [1969] 2 A.C. 31, 83-84, 93-94, 105-106, 114-115, 124.
Section 14 (1) is dependent on the knowledge of the seller; section 14 (2) is not. Each must be regarded in its own light and different tests must be applied. The Act is only concerned with the facts at the making of the contract. The goods in question may be single purpose goods, when section 14 (1) and (2) stand and fall together. But if the goods have more than one purpose, the result, when subsection (1) is invoked, depends on the seller's knowledge, and the existence of subsection (2) is no reason for limiting subsection (1) when it applies.
*465 Teheran-Europe [1968] 2 Q.B. 545, cited for the respondents, was a very different case from the present: see Lord Diplock at p. 560. The Hardwick case [1969] 2 A.C. 31 did not obliterate the Manchester Liners case [1922] 2 A.C. 74, 79, 81, 84, which is still good law: see Lord Morris of Borth-y-Gest at p. 94 and Lord Pearce at p. 115; and Mash & Murrell [1961] 1 W.L.R. 862, 867. The present case is not comparable on its facts to Manchester Liners but it is comparable to Hardwick. See also Griffiths' case [1939] 1 All E.R. 685, 691 and Grants' case [1936] A.C. 85, 92-93.
The respondents knew that feeding mink was a normal use of herring meal in Norway. This was an international market and it did not make any difference that the use was to be in England, the use being the same as in Norway. It was right to imply that the herring meal was fit for compounding for animal foodstuffs, including mink.
Their Lordships took time for consideration.

February 24. LORD HODSON.
My Lords, in July 1961 the first case was noticed of a new and hitherto unknown disease afflicting mink which are bred in a large number of farms in this country. A male kit was found with a grossly enlarged abdomen and died within a few days. Similar cases were reported from that time onwards in various parts of the country. The common factor was that all the afflicted mink had been fed a fortified cereal mink food marketed under the brand name "King Size" and made up according to a formula. This formula emanated from a Mr. Udall who, since the early nineteen-fifties, had been concerned with the breeding of mink in the Wimborne area and was recognised as an expert on mink farming. The company, Ashington Piggeries Ltd., the defendants in the action, was controlled by Mr. Udall. In 1960 he approached a Mr. Granger, who was the personal assistant to the managing director of the plaintiff company Christopher Hill Ltd.) with a view to the latter company compounding for him a mink food to be called "King Size" in accordance with a formula prepared by himself. This last-named company is an old-established and well-known animal feeding stuff compounder carrying on business at Poole. At this time the company was compounding 167 varieties of feeding stuffs principally for poultry, pheasants, calves and pigs but, until the events giving rise to this case, had had no experience or knowledge of mink.
To put the matter shortly, the ingredients were to be supplied by the plaintiffs and were to be of the best quality available. These were commodities which the plaintiffs were in the habit of handling in the course of their business and the manufacture of compounds for animal feeding to customers' formula was something which the trial judge found the plaintiffs habitually undertook. The contract of sale was entered into in May 1960 and deliveries of "King Size" commenced immediately either to the defendants or to their customers. Between May 1960 and the end of March 1961 "King Size" had been supplied to about 100 farms, but no real trouble arose until the end of July 1961. Mr. Udall's herd of mink was affected and he began to suffer increasing losses. "King Size" came under suspicion as being the cause of the outbreak of the severe liver disease from which the animals were suffering.
These proceedings were started by the plaintiffs claiming the price of *466 goods sold and delivered, namely, the "King Size." To this the defendants answered that the goods were worthless and relied in the first instance on a change made in the formula of the goods without their consent, making their attack on the use of an anti-oxidant called Santoquin No. 6 as being the cause of the liver disease in the mink. After several days this attack was abandoned and in its place the attack was directed against Norwegian herring meal claimed to have been included in "King Size" and to have been the cause of the toxin which killed the mink. It was said that the meal was manufactured from herring preserved with sodium nitrite in circumstances which rendered the meal toxic to animals and in particular mink. The substance said to be toxic was dimethylnitrasomine (DMNA) which was not a constituent of the formula. There was at the trial a conflict as to causation which no longer subsists. The findings of fact made at the trial and no longer disputed are that the cause of the liver disease in the mink was "King Size" and that the toxic element was DMNA which was in the herring meal because of the use of sodium nitrite for preservation purposes.
A great deal turns in this case on the use of the word "toxic" which is an emotive word equivalent to poisonous - and I must return to it, in view of the difference of opinion between the trial judge and the Court of Appeal as to the application of the Sale of Goods Act 1893, in particular section 14 (1), to the facts of this case.
The defendants have established that the mink were poisoned by ingestion of "King Size" because it contained Norwegian herring meal contaminated by DMNA used in the processing of herring to which sodium nitrite had been applied as a preservative.
The Norwegian herring meal had been obtained by the plaintiffs from Norsildmel, the third parties in these proceedings, under a written contract which must be treated separately from the contract between the plaintiffs and the defendants. It differs materially from that made between the original parties but involves consideration of the same sections of the Sale of Goods Act.
The herring meal which caused the damage was found to be 8 1/2 tons out of 333 1/2 tons of herring meal delivered to the plaintiffs at Poole from the motor vessel Hansa which had come from Factory 72 at Honningsvaag in Norway. Of the balance of 325 tons, 21 tons were sold to ordinary (nonmink) farms as feedings stuff and the rest was used by the plaintiffs in their various compounds. No animal which consumed any of this balance suffered any injury so far as is known.
The question of causation being out of the way, it is necessary to consider the legal issues which arise between the plaintiffs and defendants. Section 13 of the Sale of Goods Act 1893, so far as material, reads:
"Where there is a contract for the sale of goods by description, there is an implied condition that the goods shall correspond with the description; ..."
The language used is directed to the identification of goods. The point is made by Channell J. in Varley v. Whipp [1900] 1 Q.B. 513, 516:
"The term 'sale of goods by description' must apply to all cases where the purchaser has not seen the goods, but is relying on the description alone."
*467 This is not to say that it may not apply even where the purchaser has seen the goods if the deviation from the description is not apparent, but this has no bearing on the facts of the instant case. The essential point is that identification of the goods is that with which the section is concerned. The defendants rely on the oft quoted decision of Roche J. in Pinnock Brothers v. Lewis & Peat Ltd. [1923] 1 K.B. 690. This related to the sale of East African coppery cake. The goods were so adulterated with castor seed as to be poisonous to cattle. It was held that the goods were not properly described as coppery cake at all and that the sellers were not protected by a clause in the contract that:
"The goods are not warranted free from defect rendering same unmerchantable which would not be apparent on reasonable examination, any statute or rule of law to the contrary notwithstanding."
The relevant breach of description relates only to the herring meal element in the formula. The defendants say this was not herring meal, it was herring meal plus DMNA which is not an authorised ingredient of the formula. I agree with the Court of Appeal in arriving at the conclusion that the evidence does not support the view that DMNA was something added to the herring meal, as the castor seed was to the coppery cake. One of the defendants' own witnesses (Dr. Pearson) was disposed to agree that DMNA was something in the herring meal which had gone wrong. As the Court of Appeal pointed out, sodium nitrite was used as a preservative just as salt had been previously used. It produced a chemical reaction in some cases which produced DMNA. In my opinion, it is working the word "description" too hard to say that "herring meal" was a misdescription. The herring meal was contaminated but no poisonous substance was added to it so as to make the description "herring meal" erroneous.
The distinction between "description" of goods and their quality is made by Lord Dunedin in Manchester Liners Ltd. v. Rea Ltd. [1922] 2 A.C. 74, 80:
"The tender of anything that does not tally with the specified description is not compliance with the contract. But when the article tendered does comply with this specific description, and the objection on the buyer's part is an objection to quality alone, then I think section 14 (1) settles the standard, and the only standard by which the matter is to be judged."
The claim based on section 13 accordingly must, in my opinion, fail.
This brings me to consider section 14 (1) which, so far as material, reads:
"Where the buyer, expressly or by implication, makes known to the seller the particular purpose for which the goods are required, so as to show that the buyer relies on the seller's skill or judgment, and the goods are of a description which it is in the course of the seller's business to supply ... there is an implied condition that the goods shall be reasonably fit for such purpose, ..."
There is no dispute but that the defendants expressly made known to the plaintiffs the particular purpose for which the goods were required, namely, as food for mink, nor is there any dispute but that the goods were unfit for that purpose.
*468 Did the defendants rely on the skill and judgment of the plaintiffs to any and, if so, to what extent?
The judge's finding [1968] 1 Lloyd's Rep. 457, 480 reads:
"While Mr. Udall made it clear to the plaintiffs that the compound which was to be called 'King Size' was wanted for one purpose and one only, namely, as a feed for mink, I find that he made it no less clear to them that he was only relying on their skill and judgment to a strictly limited extent, namely, for the selection and acquisition of the various ingredients set out in the formula and for the actual mixing of their ingredients. As to whether the compound so made up would prove a suitable and good food for mink, he relied entirely upon his own judgment and experience. I am satisfied not only that this was in fact the position but that both parties were aware of it and accepted it from the start. On the one hand Mr. Udall was relying on his own judgment as to what his formula should contain and the levels at which the various ingredients in it should be included. On the other, he was relying, and had no alternative but to rely, upon the plaintiffs to obtain the ingredients, to see they were of good quality and not to use ingredients which, as a result of contamination, were toxic."
In using the word "toxic" the learned judge was clearly not limiting the adjective to the effect on mink alone. He expressly found, and there was evidence to support his finding, that all animals are sensitive to DMNA poisoning though mink are more sensitive than most. The level of DMNA was sufficiently high to be lethal to mink which are animals to which herring meal can properly be fed. There was no evidence as to the lethal dosage for other animals but, in disagreement with the Court of Appeal, it appears to me that the toxicity covers the range of animals for which the plaintiffs normally supply feeding stuff. As a result of contamination one of the ingredients, namely, the Norwegian herring meal, was toxic and the defendants were entitled to rely on the plaintiffs not to produce food poisonous to mink, leaving out altogether any question of the idiosyncrasy of mink.
The defendants have proved a general defect and that their animals were poisoned thereby. The expert called by the third parties, Nils Koppang, an expert from the Department of Pathology, Veterinary College of Oslo, Norway, described the disease which had existed as early as 1957. He himself referred to toxic doses in connection with DMNA in such a way that it appears that the toxic condition was not a peculiar one such as is illustrated by the case of Griffiths v. Peter Conway Ltd. [1939] 1 All E.R. 685, a case relied upon as a decision in favour of the seller. That was a case concerning the purchase of a Harris tweed coat by a woman with an abnormally sensitive skin who did not disclose the fact to the seller. She failed in her action because the unsuitability of the article arose from the special state of affairs relating to the buyer, of which the seller was not aware. It is otherwise here, where DMNA is shown to have been toxic to all animals, not only to mink.
It is well settled that the reliance on the seller's skill and judgment need not be total or exclusive: see Cammell Laird & Co. Ltd. v. Manganese Bronze and Brass Co. Ltd. [1934] A.C. 402, 427 per Lord Wright, and the earlier case of Medway Oil & Storage Co. Ltd. v. Silica Gel Corporation(1928) 33 Com.Cas. 195 *469 , 196, per Lord Sumner. Before the Sale of Goods Act in James Drummond & Sans v. E. H. Van Ingen & Co. (1887) 12 App.Cas. 284, 297 Lord Macnaghten pointed out that in matters exclusively within the province of the manufacturer the merchant relies on the manufacturer's skill.
I think the learned judge correctly applied these authorities. He pointed out that the propellers in the Cammell Laird case [1934] A.C. 402 had to be made to the buyer's detailed specification; nevertheless, it was held that the suppliers gave an implied warranty under section 14 (1) that they would be reasonably fit to be used in the ships for which they were required. The suppliers were not warranting the buyer's design but warranting the fitness of the propellers within the field in which they were called upon to exercise their own discretion. It may be thought that the Cammell Laird case went a long way against the vendors but it seems to be a close parallel to the instant case. When the buyers, that is the defendants, made a detailed specification of the ingredients to be included in the "King Size" food, the plaintiffs, acting within the field in which they were called upon to exercise their discretion, warranted that the herring which they would incorporate in the food would not contain poison. The herring meal which killed the mink did contain a poison at a level sufficiently high to be lethal to mink which are animals to which herring meal can properly be fed. Although mink are more sensitive to DMNA than most other animals all animals are sensitive to a greater or less degree to this form of poisoning.
I also agree with the learned judge that the plaintiffs cannot escape by arguing that "King Size" were not goods which it was not in the course of their business to supply. Their business was to make up compounds for animal feeding and they were only using raw materials which they regularly handled, cf. Spencer Trading Co. Ltd. v. Devon (Factual and Stickphast Ltd. Third Parties) [1947] 1 All E.R. 284.
In my judgment, the defendants succeed in establishing under section 14 (1) first that a condition of fitness was to be implied between the parties as to the suitability of the herring meal and that this condition was breached and, secondly, that the goods were of a description which it was in the course of the plaintiffs' business to supply.
So far as section 14 (2) is concerned, I agree with the Court of Appeal in the construction of the word "description" where it appears and, accordingly, I agree that the defendants fail to prove a breach of the section, because they did not deal in goods of that description, namely, the mink food called "King Size" made pursuant to the formula.
I would, accordingly, allow the first appeal, that is to say the appeal of the defendants against the plaintiffs, restore the judgment of Milmo J. and discharge the order of the Court of Appeal. Costs will be dealt with after counsel have had a further opportunity of addressing your Lordships upon the matter.
I come now to the second appeal, that is to say the appeal of the plaintiffs against the third parties, who were exonerated in the Court of Appeal. The third parties, called Norsildmel, are the successors of a Norwegian body being the Export Committee for Herringmeal and Herringoil from whom they took over in 1964. It is not necessary to draw any distinction between these two bodies for the purposes of this case. C. T. Bowring & *470 Co. Ltd., called "Bowrings, " were the exclusive selling agents of the third parties in the United Kingdom.
The contract of sale between the plaintiffs and the third parties was made in writing in the English language on the third parties' printed form with typed insertions and is set out in some detail in the judgment of Milmo J., but it will be sufficient to refer to a few portions in order to appreciate the issues which arise under section 13 of the Sale of Goods Act. Against the marginal words in block capitals QUANTITY & DESCRIPTION appear the words:
"About 300/350 ... tons at sellers' option of 2240 lb./1016 kilos of NORWEGIAN HERRING MEAL fair average quality of the season, expected to analyse not less than 70% protein, not more than 12% fat and not more than 4% salt."
The price clause concluded with a paragraph reading:
"In the event of any surpluses of fat and/or salt, sellers will make an allowance to buyers at the rate of 1% surplus = 1% of the contract price, fractions pro rata. If the fat contents exceed 13% or the salt contents exceed 5%, the buyers will have the right of rejection."
Shipment was to be from Norway during March 1961. There was provision for sampling and analysis and general conditions, of which No. 3 reads:
"The goods to be taken with all faults and defects, damaged or inferior, if any, at valuation to be arranged mutually or by arbitration."
The plaintiffs base their claim against the third parties under section 13 solely on a breach of the terms to be implied under section 13, that the goods shall correspond with the description. The contest on description has been as to whether "fair average quality of the season" is part of the description of the goods or whether, as the third parties contend, it is a warranty of quality. From the layout of the document it seems clear that the third parties are right on this matter. Looking at the photograph of the original contract and bearing in mind that a term ought not to be regarded as part of the description unless it identifies the goods sold, one sees that the words "Norwegian Herring Meal" appear in the printed form in capital letters and are sufficient to satisfy the marginal requirement of description. The words following, f.a.q. as it is called in brief. beginning "expected to analyse " etc., clearly do not form part of the description and cannot therefore be used to embrace, so to speak, the f.a.q. phrase and make it part of the description in that way. It is also to be noticed that general condition No. 3 has nothing to bite on unless f.a.q. falls for consideration under this rejection clause dealing with breach of warranty.
Although quality could be used, no doubt, as part of a description it is, I think, not so used in this case; there is a warranty of quality but no more. It is natural so to read it where it may be expected that there will perhaps be divergences in the goods supplied from the samples submitted. The reference to faults and defects in general condition 3 leads to this conclusion. I have used the word "sample" although this is not a case of sale by sample, agreeing with your Lordships that the use of the f.a.q. term performs the same function as sale by sample, cf. per Devlin J. in F. E. Hookway & Co. Ltd. v. Alfred Isaacs & Sons [1954] 1 Lloyd's Rep. 491, and per Sellers J. *471 in Steels & Busks Ltd. v. Bleecker Bik & Co. Ltd. [1956] 1 Lloyd's Rep. 228.
I am content to accept the formulation of the third parties accepted by the Court of Appeal [1969] 3 All E.R. 1496, 1520-1521:
"In essence, it relates, and relates only, to such qualities as are apparent on an ordinary examination or analysis of the goods, such as is usually done in the trade in relation to such goods."
Before leaving the construction of the contract I should add that I agree with the Court of Appeal that general condition 3 is to be read as purporting to exclude the buyer's right to reject goods for faults and defects but not as purporting to exclude his right to recover from the sellers compensation for any consequential damage which he may sustain by reason of acceptance of goods which thereafter turn out to be defective and cause loss or damage by reason of that defect. The plaintiffs' claim against the third parties based on section 13 must accordingly fail.
There remains for consideration the claim under section 14 (1). This claim stands in rather an unusual position for it depends on a finding of fact by the trial judge in favour of the plaintiffs and reversed by the Court of Appeal. The Court of Appeal came to the conclusion that in all the circumstances of the case the third party had no reason to contemplate that this consignment of herring meal would be used as food for mink and that it was impossible to draw the inference that the plaintiffs made known to the third party that particular purpose so as to show that they relied on the skill and judgment of the third party. The court held, therefore, that the condition under section 14 (1) of the Act is not to be implied and in any event the damage was too remote.
A number of documents were put in evidence which shows beyond doubt that a serious progressive liver disease had been observed in fur farms. In a treatise on Toxic Hepatosis in Fur Animals by Dr. Koppang and Dr. Helgebostad published in 1966 in Nordisk Veterinaermedicin, there is a reference, at p. 215, to "a disease which has ravaged in several fur farms since 1957. All fur farms afflicted by the liver disease have used herring meal." In an earlier treatise by Dr. Koppang, A Severe Progressive Liver Disease in Fur Animals (1966), p. 205, he referred to this disease as having been diagnosed by Bohler in 1957 and stated that in the following years the disease was observed in mink and foxes at more than 70 fur animal farms all over Norway. It is not possible to review in detail the evidence, oral and documentary, which the learned judge considered but for myself I do not find that I am persuaded by the arguments contained in the judgment of the Court of Appeal that the learned judge came to a wrong conclusion of fact. I would accordingly uphold his finding under section 14 (1) and allow the appeal of the defendants in the second appeal, namely, that to which the third parties are respondents; that is to say, I would allow both appeals.
Section 14 (2) does not arise.
What I have said earlier about costs refers also to the second appeal.

LORD GUEST.

My Lords the facts out of which these appeals arise have been fully stated in the courts below. It is only necessary to outline the essential matters for the purpose of this judgment.
*472 In May 1960 Ashington Piggeries, the appellants in the first appeal; were supplied by Christopher Hill Ltd., the respondents, with a mink food known as "King Size" compounded by Hill according to a formula supplied by Mr. Udall of Ashington Piggeries and agreed between them. It was expressly made known to Hill that the food was required for mink. The food was duly fed to mink and during July and August 1961 heavy losses occurred to mink which had been fed with "King Size." Although it was contested below, it is now conceded that the losses of mink were due to the presence in herring meal, which was one of the ingredients of "King Size," of a substance called dimethylnitrosamine (DMNA). This was proved to be highly toxic to mink in certain quantities.
The herring meal which was compounded into "King Size" to a total of 8 1/2 tons was part of a larger consignment of 333 1/2 tons of Norwegian herring meal supplied to Hill, the appellants in the second appeal, by Norsildmel a Norwegian firm, the respondents in the second appeal. It was conceded that the DMNA present in the herring meal resulted from the treatment of herring with a preservative, sodium nitrite. The herring were subsequently subjected to heat in the process of producing the meal and this was responsible for the DMNA.
Ashington Piggeries were sued by Hill for the price of the "King Size" and Ashington Piggeries counterclaimed for damages in respect of the losses occasioned by the death and injury to the mink. Hill in turn sued Norsildmel for indemnity.
Milmo J. ([1968] 1 Lloyd's Rep. 457) sustained Ashington's claim against Hill and found for Hill in their claim for indemnity against Norsildmel. The Court of Appeal ([1969] 2 Lloyd's Rep. 425) unanimously reversed the learned judge and rejected Ashington's claim and also Hill's claim for indemnity.

ASHINGTON PIGGERIES v. HILL

Ashington's case against Hill was for breach of contract based on section 13 and section 14 (1) and (2) of the Sale of Goods Act 1893 .
Section 13 of the Sale of Goods Act 1893: This section provides that, where goods are sold by description, there is an implied warranty that the goods shall correspond with the description. The order by Ashington to Hill was a verbal order for "King Size" mink food compounded according to a formula. The argument for Ashington was that as the herring meal, which was an ingredient of "King Size," contained DMNA in sufficient quantities to injure mink, it did not correspond with the description of "herring meal." In my view, this argument is not well founded. Herring meal is still herring meal notwithstanding that it may have been contaminated by DMNA. If there was a substantial addition to the commodity described, then it might be that the goods plus this addition would not correspond with the description. The case of Pinnock Brothers v. Lewis & Peat Ltd. [1923] 1 K.B. 690 is an example of this type of case, where the goods delivered were an admixture of coppery cake and castor beans, the latter being poisonous, and this was held to be a different substance from the coppery cake contracted for by description. A similar position arose in Robert A. Munro & Co. Ltd. v. Meyer [1930] 2 K.B. 312 where the meat and bone meal contracted for *473 contained an admixture of meat and bone meal and cocoa husks. It was, accordingly, not of the contract description of meat and bone meal. The quantitative approach was applied. In my view, the fact that the herring meal was contaminated by DMNA did not result in a different substance from the herring meal in the description. There was no loss of identity. In my opinion, no breach of section 13 by Hill occurred:
Section 14 (1) of the Sale of Goods Act 1893: Upon this matter I have had the advantage of reading the speech of my noble and learned friend, Lord Wilberforce. I have not found this matter easy, but I am prepared to concur in the result.
Section 14 (2) of the Sale of Goods Act 1893: This section provides that where goods are bought by description from a seller who deals in goods of that description there is an implied condition that the goods shall be of merchantable quality. There is an admission by Hill that the goods were not of merchantable quality. So that the only issue which arises is on the first limb of this section.
The argument for Hill was that as the goods bought were "King Size" mink food and as Hill had not previously dealt in goods of that contractual description the section did not apply. Such was the approach of the Court of Appeal. This is a plausible and possible construction of the subsection. But, in my view, such a strict construction would lead to absurd results. Suppose a customer goes to a tobacconist's shop and orders a box of Lara<<currency>>aga cigars in which the tobacconist had not previously dealt. If the cigars were not fit for smoking, there would be no liability on the tobacconist as he had not previously dealt in goods of that particular description, namely, lara<< currency>>aga cigars. I cannot believe that the section bears such a restricted meaning. Some support for the view which I have expressed is to be found in observations of my noble and learned friend, Lord Reid, in B. S. Brown & Son Ltd. v. Craiks Ltd. [1970] 1 W.L.R. 752, 755:
"The appellants mainly relied on the contention that, whereas cloth of this description had been commonly used for making dresses, there was no evidence that such cloth had ever been put to any industrial use. There is, I think, some ambiguity in saying that goods are of the same description where the contract description is a precise and detailed specification for their manufacture. One may mean of the same precise and detailed description, and that may be novel: or one may mean of the same general description, and that may be common. In most of the authorities the latter meaning seems to have been adopted."
Mr. Guest for Ashington Piggeries gave the House an elaborate examination of the common law before the passing of the Sale of Goods Act 1893. He argued that as the Sale of Goods Act was a codifying statute, it should in large measure represent the law on the subject before the passing of the Act. As Sir Mackenzie Chalmers put it in the introduction to his first edition of The Sale of Goods Act, p. iv: "The Bill is almost entirely a reproduction of common law, " and the changes in the Act were very slight. Mr. Guest referred to a number of cases in the 19th century which he said made it clear that there was no distinction between the word "description" at common law and as it subsequently appeared in section 14 (1) and (2). The word "description" in section 14 (1) clearly had a wide import. The principles *474 he argued, embodied in these sections applied where goods were supplied by a manufacturer or dealer in way of his business as opposed to where the seller sold goods in a private capacity. It is true that the distinction can be drawn between the case of Brown v. Edgington (1841) 2 M. & G. 279 where the seller of a piece of rope put himself forward as the manufacturer, although the rope was manufactured by an independent contractor, and was held liable, and the case of Burnby v. Bollett (1847) 16 M. & W. 644, where trading was done in an individual capacity and there was no liability for defective goods. These may be but straws in the wind. But there is the classic passage in Jones v. Lust (1868) L.R. 3 Q.B. 197, 202-203 of Mellor J. which is supposed to be the basis of the enactment of section 14 of the Sale of Goods Act 1893, which I quote:
"Fourthly, where a manufacturer or a dealer contracts to supply an article which he manufactures or produces, or in which he deals, to be applied to a particular purpose, so that the buyer necessarily trusts to the judgment or skill of the manufacturer or dealer, there is in that case an implied term or warranty that it shall be reasonably fit for the purpose to which it is to be applied: Brown v. Edgington (1841) 2 Man. & G. 279 , Jones v. Bright (1829) 15 Bing. 533. In such a case the buyer trusts to the manufacturer or dealer, and relies upon his judgment and not upon his own. Fifthly, where a manufacturer undertakes to supply goods, manufactured by himself, or in which he deals, but which the vendee has not had the opportunity of inspecting, it is an implied term in the contract that he shall supply a merchantable article: ..."
This case is referred to in the note to section 14 (1) in Chalmers (1st edition). If this be the principle lying behind section 14, I see no reason why the scope of the dealer's business in section 14 (2) should be restricted to the contract description of the goods.
If the purpose of the qualifying words "deals in goods of that description " was to confine the section to a dealer in the way of business as opposed to a private capacity, then I think a fair interpretation of the words would be "who deals in goods of that kind." If this is the proper interpretation of section 14 (2) then the conditions are satisfied because Hill had dealt before in goods of that kind, namely, animal feeding stuffs. In my view, Hill are in breach of the implied condition as the goods were not of merchantable quality.

HILL v. NORSILDMEL

The contract for the supply of Norwegian herring meal by Norsildmel to Hill is dated February 14, 1961. The buyers are Hill and the sellers Export Committee for Herringmeal and Herringoil, Bergen, Norway. The provisions of the contract which are relevant for consideration are:
"QUANTITY & DESCRIPTION:
About 300/350 (three hundred to three hundred and fifty) tons at sellers' option of 2240lb./1016 kilos of NORWEGIAN HERRING MEAL fair average quality of the season, expected to analyse not less than 70% protein, not more than 12% fat and not more than 4% salt." *475 GENERAL CONDITIONS: ...
"3. The goods to be taken with all faults and defects, damaged or inferior, if any, at valuation to be arranged mutually or by arbitration."
Section 13 of the Sale of Goods Act 1893: The primary argument for Hill was that the whole of the clause "Quantity and Description" was "the description" of the goods within the meaning of section 13, the terms of which have already been quoted. It was said that "fair average quality of the season " (f.a.q.) must be part of the description because on what has been conveniently described as the "sandwich principle" that part of the clause dealing with expected analysis of the meal was part of the description. It therefore followed that the intervening words f.a.q. must also be part of the description. In my view, the fallacy of this argument lies in the fact that the "expected analysis" is not part of the description. Where goods are unascertained, "description" implies a specification whereby the goods can be identified by the buyer. Such a case was Arcos Ltd. v. E. A. Ronaasen & Son [1933] A.C. 470 where the timber contracted for was precisely specified as to length, breadth and thickness. Neither f.a.q. nor the expected analysis provision identifies the goods. They prima facie indicate the quality of the goods: see Pacific Trading Co. Ltd. v. Wiener (1923) 14 Ll.L.Rep. 51, 54, Roche J. There is a case where the contract was for goods "afloat per s.s. Morton Bay due London approximately June 8"; these words were held to be part of the description (Macpherson Train & Co. Ltd. v. Howard Ross & Co. Ltd. [1955] 1 W.L.R. 640). But that is a different case from the present. It enabled the goods to be identified. I do not dispute that there may be cases where a qualitative description of the goods may come within the section. The case of Varley v. Whipp [1900] 1 Q.B. 513 is an example of such a case. It concerned the sale of a reaping machine stated to have been new the previous year and to have been used to cut only fifty or sixty acres. This was held to be a sale by description. But in that case the description would have identified the goods as a nearly new machine.
I have reached the conclusion without much difficulty that f.a.q. is not part of the description of "Norwegian herring meal" contained in the contract, nor is the expected analysis part of the description. Apart from the side note "Quantity & Description" in the sales contract, I can find no justification whatever for importing f.a.q. into the description of the goods. The side note by itself cannot control the clause where the rest of the clause is clear and unambiguous.
If f.a.q. is not part of the description of the goods, then it becomes unnecessary to consider the question whether the herring meal was f.a.q. and whether, if it was, it covers latent defects. These questions were extensively debated before the House. On the assumption that only the words "Norwegian herring meal" were part of the description, then the question whether there was a breach of section 13 is the same as that raised in the principal appeal and I would, accordingly, hold that there was no breach of section 13.
General Condition 3: I agree with the rest of your Lordships in thinking that this clause is no bar to Hill's action for indemnity. This was the view of both courts below and I see no reason to differ from them.
*476 Section 14 (1) of the Sale of Goods Act 1893: In view of the difficulties attaching to the construction of this clause, I quote it in extenso:
"Subject to the provisions of this Act and of any statute in that behalf, there is no implied warranty or condition as to the quality or fitness for any particular purpose of goods supplied under a contract of sale, except as follows:- (1) Where the buyer, expressly or by implication, makes known to the seller the particular purpose for which the goods are required, so as to show that the buyer relies on the seller's skill or judgment, and the goods are of a description which it is in the course of the seller's business to supply (whether he be the manufacturer or not), there is an implied condition that the goods shall be reasonably fit for such purpose, provided that in the case of a contract for the sale of a specified article under its patent or other trade name, there is no implied condition as to its fitness for any particular purpose: ..."
The courts below differed as to the application of this section and I have found this the most difficult part of the case.
If I set out the argument for Norsildmel it will focus the issue. Mr. Lloyd argued that the purpose which was made known to the seller was the purpose of compounding for animal feeding stuffs and that this was not a particular purpose within the meaning of section 14 (1). It was not so particular as to show that the buyer relied on the seller's skill or judgment. The expressed purpose of compounding for animal feeding stuffs did not give the seller an opportunity of deciding whether he would contract with the buyer. It was further contended that the feeding of herring meal to mink had not been shown to be a normal use of herring meal in Norway or Great Britain. Reference was made to certain passages in the speeches in Hardwick Game Farm v. Suffolk Agricultural Poultry Producers Association [1969] 2 A.C. 31. If I select one typical passage from the speech of my noble and learned friend, Lord Reid, it must not be thought that I have not had in my mind other similar passages from the other judgments. I quote from p. 80:
"It was argued that I whenever any purpose is stated so as to bring this subsection into operation, the seller must supply goods reasonably fit to enable the buyer to carry out his purpose in any normal way. But that can only be right if the purpose is stated with sufficient particularity to enable the seller to exercise his skill or judgment in making or selecting appropriate goods."
But before I proceed I must refer to another passage in the speech of my noble and learned friend. After referring to Manchester Liners Ltd. v. Rea Ltd. [1922] 2 A.C. 74, he says [1969] 2 A.C. 31, 81:
"I do not think that this case is any authority for the view which has sometimes been expressed that if the seller knows the purpose for which the buyer wants the goods it will be presumed that the buyer relied on his skill and judgment."
In Teheran-Europe Co. Ltd. v. S. T. Belton (Tractors) Ltd. [1968] 2 Q.B. 545, 554 Lord Denning MR. said that this passage in Lord Reid's speech had dealt a "knock-out blow" to the idea that where a particular purpose *477 was made known to the seller there is a presumption that the buyer relies on his skill or judgment. None of the others of their Lordships in Hardwick [1969] 2 A.C. 31 referred to this point. In fact Lord Pearce at p. 115 says that: "The whole trend of authority has inclined towards an assumption of reliance wherever the seller knows the particular purpose." I do not understand my noble and learned friend, Lord Reid, to be saying that the presumption can never be drawn from the mere fact that a particular purpose is made known to the seller. He emphasises that the question is whether in the whole circumstances the inference can properly be drawn that a reasonable man in the shoes of the seller would realise that he was being relied upon.
The whole of section 14 (1) so far as relevant must be read together:
"Where the buyer, expressly or by implication, makes known to the seller the particular purpose for which the goods are required, so as to show that the buyer relies on the seller's skill or judgment, ..."
The purpose must be a "particular" purpose. It must be a definite purpose. In Hardwick Game Farm the purpose of compounding into feeding stuffs for pigs and poultry was held sufficiently particular (see my noble and learned friend, Lord Morris of Borth-y-Gest, at p. 93). "A communicated purpose, if stated with reasonably sufficient precision, will be a particular purpose. It will be the given purpose." (See also Lord Pearce at p. 114.)
In the present case the purpose was for compounding into animal feeding stuffs, which would seem sufficiently definite. It distinguishes it from use as fertiliser. The knowledge of the seller need not be expressly communicated: it may be by implication, as the section provides. If the seller knows the purpose for which the buyer requires the goods, then no express intimation by the buyer is necessary. It will be implied (see Mash & Murrell Ltd. v. Joseph I. Emanuel Ltd. [1961] 1 W.L.R. 862, 866 and Manchester Liners Ltd. v. Rea Ltd. [1922] 2 A.C. 74).
The question in the present case therefore resolves itself into this: Whether in all the circumstances it is proper to draw the inference that there was reliance by the buyer on the seller's skill or judgment - in other words, whether the particular purpose of the herring meal for compounding into animal feeding stuffs having been made known, the sellers knew that it was likely that it would be fed to mink. Whether the test is "likely," "not unlikely," or "liable to" is a matter of taste, but all these tests are comprised in the speeches in C. Czarnikow Ltd. v. Koufos [1969] 1 A.C. 350. If the proper inference from all the evidence is that Norsildmel knew that herring meal was used as food for mink, then, in my view, it is sufficient to show the reliance required by the section. If the particular purpose is shown, then it is an easy step to draw the inference of reliance.
Hill start off with the undoubted advantage that they have a finding in their favour by the trial judge (Milmo J. [1968] 1 Lloyd's Rep. 457, 484, 490):
"Fishmeal, including herring meal, has been used for many years as a food for mink and I am satisfied that it is a perfectly safe and proper food to give mink assuming, of course, it is pure meal, that is to say, has not been contaminated by a deleterious substance and has not been produced in a manner which results in the presence of a potential toxin. Fishmeal, prior to the introduction of 'King Size' in the middle of *478 1960, was not used to any appreciable extent as a feeding stuff by mink farmers in the United Kingdom, though on the recommendation of Mr. Udall one farmer certainly did use it. On the other hand, I am satisfied that it was used extensively in other parts of the globe and had been used for many years ...
The argument addressed to me on behalf of the third parties on this point relating to damages was largely, if not entirely, based upon the hypothesis that the feeding of herring meal to mink was not a normal or well-known user of herring meal and further it was a user of which both Mr. Arnesen and Mr. Validness at the relevant time were unaware. I have already found that herring meal was at the time a perfectly normal and well-known ingredient of the diet of mink kept in captivity both in Europe, on the American continent, in Scandinavia and in Norway. I have also found that Mr. Validness was prior to 1961 aware that herring meal was being fed to mink in Norway. On the footing that Mr. Arnesen saw at the time the correspondence that passed between Sildemelutvalget and the Norwegian Fur Farmers Marketing Organisation and the Norwegian College of Agriculture to which I have already referred, I have found that he also was aware. The real question, however, is whether the sellers, Sildemelutvalget, of which Mr. Arnesen was the assistant director and who are now represented by the third parties, knew of the practice of feeding herring meal to mink and I am satisfied that they must have done."
The Court of Appeal, for reasons which they gave, disturbed this finding of fact and held that there was not sufficient evidence to establish that Norsildmel knew before the contract was entered into that herring meal was being fed to mink.
In my view, the Court of Appeal were not entitled to disturb what was essentially a finding of fact by the trial judge. It depended not only on documentary evidence but also upon the evidence of witnesses. Milmo J. did not accept the evidence of Norsildmel's two witnesses, Arnesen and Voldnes, who said they did not know of this fact before 1957. I find that there was sufficient evidence upon which the trial judge could make a finding as to the requisite knowledge on the part of Norsildmel.
Herring meal is an international commodity dealt with throughout the world as animal feeding stuff. The most significant fact is that from 1957-61 there were recurrent outbreaks of a liver disease in mink in Norway where herring meal was being fed to mink. These were investigated. It was not known then what was the cause of the disease or that it came from herring meal. But the fact that herring meal was being fed to mink must have been known to Norsildmel who were so heavily involved in the sale of that commodity. Mr. Validness admitted that they knew that from 1957 Norwegian mink farmers were feeding herring meal to mink.
It is apparent from a correspondence dated in November 1960 and produced by Norsildmel that herring meal was being pushed in Norway as a suitable food for mink. Mr. Validness said there was no reason why herring meal should not be fed to mink.
In the Nordic Handbook on Minkuppfodning (Mink Rearing) (1961), p. 125 there occurs this passage: *479 
"Large quantities of fish meal are produced from herrings, sold usually under the name of herring meal. The herring meal often has a large fat content, more than 6 per cent. By means of a long extraction process it is, however, possible to produce herring meal with a low fat content, similar to that of, for instance, cod meal. In Norway part of the herring meal is used for the feeding of mink. If the fat content is low - 5 per cent. and less - one can use a quantity of up to 4 per cent. in the mixture."
In the article of the Fur Trade Journal of Canada,"Feeding Spring Herring " (April 1952), p. 165 it is stated that herring ought to be a valuable food nutritionally for fox and mink.
If Norsildmel had knowledge that herring meal was being fed to mink in Norway and elsewhere I see no reason why it was necessary for Hill to prove use in Great Britain.
Mr. Lloyd for Norsildmel relied very strongly on Griffiths v. Peter Conway Ltd. [1939] 1 All E.R. 685. In that case a lady purchased a Harris tweed coat. After wearing it she developed dermatitis. She made a claim under section 14 (1) of the Sale of Goods Act. The evidence revealed that the plaintiff's skin was abnormally sensitive and that there was nothing in the cloth which would have affected the skin of a normal person. The abnormality of the plaintiff's skin was not made known to the retailer. The plaintiff failed on the ground of the abnormality of her skin, which no seller would assume to exist. I regard this as a highly special case which has no application to the present. All animals are sensitive to DMNA. Whether it is lethal depends on the animal and the quantity of meal supplied in the food. Mink may be more sensitive than other animals but the tweed coat in Griffiths v. Peter Conway Ltd. would not have harmed a normal person.
Remoteness: I do not desire to add anything to what has been said on this subject.
I would allow the appeal of Ashington Piggeries against Hill and the appeal of Hill against Norsildmel.

VISCOUNT DILHORNE.

My Lords, in 1960 Mr. Udall, who controlled the appellant companies, was minded to market a food for mink which would constitute part of their daily ration. He got in touch with the respondents who carry on business as compounders of food for animals and who produce animal feeding stuffs. They knew nothing about the nutritional requirements of mink and had not previously produced a food for mink.
Mr. Udall had bred mink since 1951 and was a well-known mink farmer. Various formulae were considered and the formula finally agreed on between Mr. Udall and the respondents included 3 cwt. of herring meal in every ton of the food for the mink. Herring meal and all the bulk ingredients of the mink food were in normal use by the respondents in the compounding of animal feeding stuffs and, as Milmo J. said [1968] 1 Lloyd's Rep. 457, 462, "commodities which the plaintiffs" (now the respondents) "were handling from day to day in the course of their business."
It is clear that Mr. Udall did not rely on the respondents to produce a food suitable for mink. He relied on them to produce a food in accordance *480 with the agreed formula and, if they did that, then its suitability for mink was a matter for which he was responsible.
The mink food compounded in accordance with the agreed formula was given the name "King Size." Marketing of the food began in May or June 1960. The appellants ordered it from the respondents and it was delivered either to the appellants or to their customers. By the end of the year sales had reached the level of about 25 tons a month and about 100 mink farms were being supplied.
The deaths of mink which gave rise to this litigation did not commence until the end of July 1961. Prior to that date considerable quantities of "King Size " had been supplied and fed to mink. Each batch of "King Size" contained 15 per cent. of herring meal and it was not until after Norwegian herring meal began to be used by the respondents that trouble occurred. It was suggested that the fat content of herring meal made it unfit for consumption by mink. Whether the fat content of Norwegian herring differs from that of other herring was not established but it is, in my opinion, an important fact in relation to this case that for many months "King Size" containing herring meal was fed to mink without causing their death or disease of the liver.
From the end of July 1961 until the end of 1963 considerable numbers of mink died in farms all over the country. Milmo J. was satisfied (p. 463) that "there was one disease, and one only, involved and that this disease was not only unknown in this country but no scientific literature was to be found at that time regarding it." Examination of dead mink did not show any bacterial infection. The only common factor was that the mink that died had been fed with "King Size."
Bearing in mind that "King Size" had been fed to mink for many months without any such in effects, Mr. Udall sought to find out whether without his knowledge any of the ingredients of "King Size" had been altered. He learnt that from March 1961 an anti-oxidant "Santoquin No. 6" had been used instead of the anti- oxidant prescribed in the formula.
On December 7, 1961, the respondents issued a writ against Ashington Piggeries Ltd., claiming £6,693 for the price of goods sold and delivered, and on January 9, 1962, they issued a writ against Fur Farm Supplies Ltd., claiming £634 3s. 4d. as the price of goods sold and delivered. The actions were consolidated.
By the defence and counterclaim the appellants alleged that the sales of "King Size" were sales by description; alternatively, that they had made known to the respondents the particular purpose for which "King Size" was required so as to show that they relied on the respondents' skill and judgment and that the mink food was of a description which it was in the course of the respondents' business to supply; and in the further alternative, that the food was bought by description from the respondents who at the material times dealt in goods of that description.
They thus relied on sections 13, 14 (1) and 14 (2) of the Sale of Goods Act 1893. They alleged breaches of the respondents' obligations under those sections in that the mink food did not contain the prescribed antioxidant but contained "Santoquin No. 6." They also alleged deficiencies in Vitamins A and E, but their main allegation was in respect of the inclusion of "Santoquin No. 6."
*481 The hearing of the action began on June 24, 1965. On the 12th day of the hearing, in the course of the cross-examination of Dr. Pearson, an expert called by the appellants, an extract from a work published in the United States was introduced. This extract referred to a disease in mink of unknown origin causing moderate to high mortality in Norway.
The appellants then examined the respondents' "mix sheets" which recorded the ingredients used in making each batch of "King Size." They had been disclosed just prior to the hearing and, until the production of this extract, appeared to have little relevance to the issues in the case. The appellants then discovered that the word "NORGR" had been written above the word "herring meal " on the mix sheet for March 30, 1961.
That week-end Dr. Pearson went to Norway. He learnt that prior to 1961 and since, scientific research had been going on in Norway into a new liver disease which affected mink and other animals and that it had been found that some herring meal produced in Norway had led to the deaths of ruminants and mink with symptoms closely resembling those found in the dead mink in the United Kingdom. On his return to England, the appellants obtained leave to amend their defence and counterclaim and the hearing was adjourned.
The respondents revealed that Norwegian herring meal was first used in making "King Size" on March 20, 1961, and that it had been used from March 20 Until July 20, 1961. In all 56 tons of "King Size" containing 8 1/2 tons of Norwegian herring meal had been compounded.
For some years sodium nitrite had been used in Norway for preserving fish. Nitrite so used gradually disappears, but if any is left when heat is applied in the course of making the meal, there can be interaction with the herring and the production of dimethylnitrosamine (hereafter referred to as DMNA).
In their amended defence and counterclaim the appellants abandoned the allegations of breaches which they had originally made and alleged that there had been breaches of the Sale of Goods Act 1893 by the inclusion in the compounding of "King Size" of herring meal containing DMNA which they alleged was toxic to animals and, in particular, to mink, and which was not a constituent of the agreed formula.
That DMNA was the cause of the deaths of the mink in the United Kingdom was strenuously disputed at the trial. Milmo J. [1968] 1 Lloyd's Rep. 457, 474 said that the appellants had satisfied him
"that the disease which attacked the English field cases was a progressive liver intoxication caused by a single toxic agent. On the balance of probabilities I find that this toxic agent was DMNA, that it was contained in 'King Size' and that it was not present in 'King Size' prior to mid-March 1961."
The Court of Appeal agreed with this conclusion and this finding was not challenged in this House.
The respondents admitted that the sales of the "King Size" were sales by description. The only question, therefore, to be determined in relation to section 13 of the Sale of Goods Act is, did the fact that the "King Size " supplied between March and July 1961 included herring meal containing *482 DMNA constitute a breach of the implied condition that the goods supplied would correspond with the description?
Milmo J. found that "King Size" which contained DMNA in quantities which rendered it toxic to mink, did not comply with the description contained in Mr. Udall's formula, and that it was goods so described that the plaintiffs had undertaken to manufacture, sell and deliver.
The Court of Appeal disagreed. They held that the appellants had failed to show that the goods delivered did not correspond with the contractual description. They came to the conclusion that the 8 1/2 tons or so of herring meal contaminated with DMNA used for compounding "King Size" was toxic to mink but [1969] 2 Lloyd's Rep. 425, 456 "was not so contaminated as to make it toxic to any other kind of animal, whether the herring meal was fed directly or as a part of a compound foodstuff." They also said (p. 458) that there was "no justification in the evidence for regarding the presence of DMNA as an addition to, rather than a condition of, herring meal" and they cited the following passage from Dr. Pearson's evidence:
"... there is no external extraneous substance which has been added to the herring meal; it is just something in the herring meal itself which has gone wrong.
It has been held that where a substance has been added, for example castor seed to coppery cake (Pinnock Bros. v. Lewis & Peat Ltd. [1923] 1 K.B. 690; British Oil & Cake Co. Ltd. v. J. Burstall & Co. Ltd. (1923) 15 Ll.L.Rep. 46), and cocoa husks to meat and bone meal (Robert A. Munro & Co. Ltd. v. Meyer [1930] 2 K.B. 312), the goods delivered did not comply with the description but it has not been decided that a failure to comply with the description can only be established on proof of the addition of a substance.
In my opinion, failure to comply may be brought about by a variety of causes. What has to be established is not that there is a particular cause but that the goods do not comply with the description. The addition of a preservative could by interaction with the meal and the application of heat cause the formation of a substance quite as deleterious as the addition of castor beans to coppery or cocoa husks to meat and bone meal and could lead to the product delivered not complying with the description of the goods ordered.
In this case the Court of Appeal appear to me to have decided that the goods did comply with the description, mainly on account of their conclusion that the meal was not harmful to other animals than mink. They based this conclusion on the fact that the 8 1/2 tons formed part of a shipment of 333 1/2 tons by the third party to the respondents. While it may well be that not all the bags in that shipment and not all the bags used for compounding "King Size" were contaminated, it is to the highest degree improbable that only the contaminated bags were used for making food for mink. The probability is that the balance of the 333 1/2 tons was contaminated to a like extent as the 8 1/2 tons. That balance was used for feeding animals but there was no evidence that any other animals fed had suffered any in effects. If the case rested there, then the conclusion that the meal containing DMNA was only toxic to mink would be warranted.
*483 It was not suggested in the pleadings by the respondents that mink had special idiosyncrasies which distinguished them from other animals in relation to DMNA, nor, I understand, was any evidence to the effect given at the trial. That mink were more sensitive to DMNA than other animals was said, but not that DMNA was harmful only to mink.
Dr. Roe gave evidence about experiments he had conducted using very small quantities of DMNA. He had received it as a liquid and he had instructed his assistant to keep the room in which the experiments were conducted well ventilated with the windows open and to wear gloves for, he said, "this material is particularly hazardous to man in concentrated form."
Dr. Koppang, an expert called by the third party, said that all animals were sensitive to DMNA poisoning but that the sensitivity varies from species to species.
In December 1964 Dr. Koppang had written a booklet An Outbreak of Toxic Liver Injury in Ruminants which was published by the Department of Pathology, Veterinary College of Norway.
It began with the following paragraph:
"In 1961, cases of a pernicious liver disease with a distinct clinical picture and characteristic pathological lesions began to appear in cattle and sheep. Several animals in the same stock were affected and many died. The disease had not previously been observed in Norway, nor have we seen it reported in the literature."
He said that the first cases occurred in February 1961 in sheep, and that all the animals affected had been fed with herring meal. He had conducted a number of experiments and wrote (p. 318):
"The feeding experiments have borne out what has been observed in practice, namely, that some herring meal may be toxic. There are close to 100 herring meal factories in Norway, but sacks with toxic meal have practically only been manufactured by one factory in Northern Norway."
Later another booklet written by him and Dr. Helgebostad was published by the Veterinary College of Norway, entitled Toxic Hepatosis in Fur Animals. This said among other things at p. 210 that "it was decided that it was after all necessary to determine if the herring meal which was found to be toxic for ruminants, could also cause the severe progressive liver disease in fur animals."
One of the conclusions they reached, at p. 215 was:
"The severe progressive liver disease in fur animals and the toxic hepatosis in ruminants are diseases which are both caused by toxic herring meal."
They also wrote, at p. 215:
"The feeding experiments on mink, blue foxes and one silver fox with herring meal which had proved toxic to sheep, resulted in the production of pathological changes similar to those found in severe progressive liver disease, a disease which has ravaged in several fur farms since 1957. All fur farms afflicted by the liver disease have used herring meal."
*484 And, at p. 223:
"The conclusions drawn from the investigations from the beginning of the ' fifties have given the laboratories and manufacturers a false feeling of security, assuming that the preservation of the raw material with nitrite could not cause poisoning. The danger of spoliage to large of raw material has led certain manufacturers to use nitrite in amounts which lie far above the regulation limits. The formation of unsaturated fatty acids in the fish oil were reduced and thereby a better price could be attained."
In the light of this scientific evidence the conclusion that the herring meal used by the respondents for the making of "King Size" was not toxic to any other kind of animal was not, in my opinion, warranted. I regard it as contrary to the balance of probabilities that herring meal, contaminated to such a degree as to cause the deaths of many mink, even when it formed only a small proportion of their food, was not also harmful to other animals. The fact that at the trial no evidence was adduced that other animals had died or suffered in consequence of consuming some part of the balance of the 333 1/2 tons, when the deaths of mink occurred in 1961 and herring meal was not suspected until 1965 is, in my opinion, wholly insufficient to counter the effect of the scientific evidence.
If, as a result of the consumption of herring meal contaminated by DMNA supplied by the respondents, other animals died or were made ill in 1961, it may well be that the deaths and illness were not attributed to the herring meal. It was not until long after the deaths of the mink that that was alleged to be the cause. However that may be, bearing in mind that there was no issue at the trial or raised in the pleadings as to such meal being only toxic to mink, I think it is clear that Milmo J. [1968] 1 Lloyd's Rep. 457, 480 when he said that Mr. Udall was relying on the respondents "to obtain the ingredients, to see that they were of good quality and not to use ingredients which, as a result of contamination, were toxic" he meant toxic to all animals to which the meal would normally be fed.
In Chalmers' Sale of Goods Act, 3rd ed. (1893), p. 27, the author, who drafted the Act, cites the following passage from the judgment of Lord Blackburn in Bowes v. Shand (1877) 2 App.Cas. 455, 480: "If the description of the article tendered is different in any respect it is not the article bargained for, and the other party is not bound to take it."
In Pinnock Brothers v. Lewis & Peat Ltd. [1923] 1 K.B. 690 it was only after the cake had been fed to animals and they had been made in that it was analysed and found to contain castor beans in so large a proportion as to make it poisonous. Roche J. held that there was a failure of the goods to comply with the description.
"Herring meal does not normally contain a poison" Milmo J. [1968] 1 Lloyd's Rep. 457, 481. Did the presence of DMNA merely affect the quality of the herring meal or did it make a difference in kind? If the former, then there was no failure to deliver in accordance with the description. If the latter, there was. In view of the fact that herring meal is not normally poisonous to animals, I think that, as in the Pinnock case, there was a difference in kind. Whether such a difference is brought about by the admixture of other substances or by the effect of manufacture when a particular *485 preservative has been used, seems to me immaterial. True it is that the presence of DMNA could not have been detected on a visual examination of the meal and that the presence of castor beans in the coppery cake was only discovered on analysis, but the fact that the adulteration or contamination is not discoverable on inspection does not, in my opinion, prevent a buyer from establishing that goods bought by description do not comply with the description.
The line between a difference in quality and a difference in kind may in many instances be difficult to draw. Here, where the distinction is between poisonous and non-poisonous herring meal, that was, in my opinion, more than a difference in quality, and I agree with Milmo J. in thinking that there was difference in kind and so a breach of section 13.
I also think that there were breaches of section 14 (1) and (2). The appellants made known to the respondents the particular purpose for which "King Size" was required, namely, as a food for mink. They did not, however, rely on the respondents' skill and judgment to provide food suitable for mink. They made known the purpose so as to rely on the respondents' skill and judgment in seeing that the ingredients used were of good quality. It was in the course of their business to supply herring meal. There was consequently an implied condition that the herring meal would be reasonably fit for use as a feeding stuff for animals.
I agree with Milmo J. that the meal containing DMNA was not reasonably fit for that purpose.
The Court of Appeal came to a different conclusion, as they held [1969] 2 Lloyd's Rep. 425, 462 - in my opinion, as I have said, erroneously - that the herring meal was "perfectly suitable as a feeding-stuff for all animals for which the plaintiffs" (the respondents) "normally compounded feedingstuffs, but which was toxic to mink alone." Presumably but for this conclusion, if they had held that the meal was toxic to other animals, they would have held that there had been a breach of section 14 (1).
Section 14 (2) applies when goods are bought by description from a seller who deals in goods of that description. It was not disputed that the goods were bought by description. Did the respondents deal in goods of that description, that is to say, of that kind? They had not made food for mink before but they dealt in all the ingredients which when compounded were labelled "King Size, " and they dealt in herring meal in large quantities.
Consequently, in my opinion, the case comes within section 14 (2) and there was an implied condition that the goods should be of merchantable quality. It was conceded by the respondents that they were not.
For these reasons, in my opinion, the appeal by the appellants succeeds.
I now turn to the respondents' claim against the third party.
On February 14, 1961 the respondents entered into a contract with Sildemelutvalget, the export committee for Herring Meal and Herringoil, to whom the third party is the successor, through C. T. Bowring & Co. of Leadenhall Street, for the purchase of herring meal. The contract contained the following paragraph:
"QUANTITY & DESCRIPTION. About 300/350 (three hundred to three hundred and fifty) tons at sellers' option of 2240 lb./1016 kilos of NORWEGIAN HERRING MEAL fair average quality of the *486 season, expected to analyse not less than 70% protein, not more than 12% fat and not more than 4% salt."
It contained a provision for the reduction of price if the fat or salt content exceeded the expected proportion, and gave the respondents a right of rejection if the fat content exceeded 13 per cent. or the salt content 5 per cent.
The sale was thus of goods by description. There was considerable controversy as to what was the description, the respondents contending that it included the words "fair average quality of the season"; the third party that it did not.
Where a sale is by description, the description identifies the goods. In my opinion, the words "fair average quality of the season" do not do so. They do not assist in the identification of the goods and merely relate to the quality of the herring meal. I do not think that the fact that this paragraph of the contract has the words "QUANTITY & DESCRIPTION" beside it suffices to justify the conclusion that the words were intended to be part of the description of the goods, though I appreciate that there may be cases where a reference to quality may be part of the description identifying the goods the subject of the sale.
In my opinion, the description of the goods the subject of the sale was Norwegian herring meal, and nothing more. The expectation that on analysis it would not have less than 70 per cent. protein or more than 12 per cent. fat and 4 per cent. salt cannot be part of the description identifying the goods.
The respondents alleged that there was a breach of section 13 of the Sale of Goods Act by the third party. Whether or not there was depends on whether the meal supplied could properly be described as Norwegian herring meal. Norwegian herring meal does not normally contain a poison. The meal delivered contained DMNA in sufficient quantites to kill mink and, as I have said, in my opinion, to be harmful to other animals.
The same question arises for decision in relation to the respondents' claim under section 13 as arose on the appellants' claim against the respondents. For the reasons I have already stated, and which I need not repeat, in my opinion the difference between Norwegian herring meal, which is normally harmless to animals, and Norwegian herring meal containing DMNA in sufficient quantities as to be harmful to animals is not just a difference of quality but one of kind and I would, therefore, find in favour of the respondents on this issue.
The respondents also alleged that there was a breach by the third party of its obligations under section 14 (1) of the Act. In relation to this claim the first question to be considered is: did the respondents, expressly or by implication, make known to the third party the particular purpose for which the herring meal was required so as to show that they relied on the third party's skill and judgment?
There are two main uses for herring meal, as a food for animals and as a fertiliser. In this case the quality and price stated in the contract indicated that it was required for feeding animals. I think that the purpose for which the meal was required was made known to the third parties so as to show that the respondents relied on their skip and judgment, and also that it was "the particular purpose" within the meaning of those words in the Act. *487 I regard it as almost unarguable that a person who goes into a shop and asks for a food for feeding to animals has not made known the particular purpose for which he is requiring the food and that he has only made known the particular purpose if he specifies the variety or varieties of animals ho wants to feed.
The Court of Appeal rightly said that for the respondents to succeed under section 14 (1) they had to show that the third party should reasonably have contemplated when the contract was made that mink was a type of animal to which it was not unlikely that herring meal would be fed; in other words, should the third party have realised, when it was made known to them that the meal was wanted to feed to animals, that one kind of animal to which it might be given was mink.
The third party called two witnesses, one an assistant director of theirs since 1960 and the other their chairman from 1940 to 1963. The assistant director said that they were not aware that herring meal was fed to mink until late in 1962 and the chairman said that he had always seen to it that herring meal should not be sold as mink food.
It was, however, admitted by the third party that herring meal had in fact been fed in Norway to mink since 1957. This admission was based on the statements in Dr. Koppang's booklets to which I have referred, namely (A Severe Progressive Liver Disease in Fur Animals, p. 205), that the first case of the disease had been diagnosed in 1957 and: "In the following years the disease was observed in mink and foxes at more than 70 fur animal farms all over Norway. " and (Toxic Hepatosis in Fur Animals, p. 215) that "All fur farms afflicted by the liver disease have used herring meal."
In October and November 1960 there were conferences between Sildemelutvalget and the Norwegian Fur Farmers Marketing Association and the Institute of Poultry and Fur Bearing Animals with a view to securing the sale of herring meal for consumption by mink in Norway.
Milmo J. said [1968] 1 Lloyd's Rep. 457, 484 that he was unable to believe that in the course of these conferences Sildemelutvalget "did not learn, if they did not already know, that herring meal was being fed to mink by Norwegian farmers." He found that the two witnesses to whom I have referred were aware of this, and he was satisfied that Sildemelutvalget must have known of this practice.
The Court of Appeal [1969] 2 Lloyd's Rep. 425, 475 rejected his conclusions and held that the respondents
"have failed to establish that, at the date of the contract, the third party should reasonably have contemplated that it was not unlikely that the herring meal, the subject matter of this contract, or part of it, might be fed to mink. On the contrary, we think on the evidence that the result of such contemplation would have been that it was highly unlikely that the herring meal would be fed to mink, to which alone it would be harmful."
I have already dealt with the conclusion that the meal was harmful only to mink and, while I accept the validity of the criticisms made by the Court of Appeal of some of the reasons advanced by Milmo J., nevertheless bearing in mind the information provided by Dr. Koppang and the fact that Milmo *488 J. had the advantage of seeing the two witnesses in the witness box, I do not think that those criticisms sufficed to justify the rejection of his conclusions and to entitle the Court of Appeal to substitute a finding of their own.
Milmo J. held that [1968] 1 Lloyd's Rep. 457, 485: "The feeding of herring meal to mink in and prior to 1961 was ... an ordinary and reasonable user of that commodity." If Norwegian herring meal was fed to mink in Norway and the third party was aware of this, then the third party should have contemplated that its use for food for mink in the United Kingdom was not unlikely.
In my view, the answer to the question posed by the Court of Appeal is in the affirmative. The third party in the circumstances should have realised that if the purpose of the purchase was to feed the meal to animals, it might be fed to mink. It-was in the course of their business to supply herring meal and thus, by virtue of section 14 (1), there was an implied condition that the herring meal would be reasonably fit for use as food for animals including mink.
The meal was not reasonably fit for such use as it contained DMNA and was poisonous and so, in my opinion, the respondents succeed on their claim under section 14 (1) against the third party.
The third party sought to rely upon condition 3 of the general conditions of the contract, which was in the following terms: "The goods to be taken with all faults and defects, damaged or inferior, if any, at valuation to be arranged mutually or by arbitration." I agree with the Court of Appeal that this clause does not exclude their liability for damages for breach of the implied condition that the meal should be reasonably fit for use as food for mink and other animals.
For these reasons, in my opinion, the respondents' appeal against the third party succeeds.

LORD WILBERFORCE.

My Lords, what caused the death of the appellants' mink in 1961? This was the substantial, and very difficult, issue at the trial. The appellants succeeded in proving after some fifty days that the heavy losses which occurred had been caused by the presence of dimethylnitrosamine (DMNA) in herring meal supplied to them by the respondents. This herring meal was an ingredient in a compound mink food ("King Size") ordered, on the appellants' behalf, by their controlling director Mr. W. T. Udall, a leading mink farmer and an expert in mink nutrition. The judge's finding as to the cause of the mortality was upheld by the Court of Appeal and has not been challenged in this House. We are only concerned with the appellants' rights under their contract of sale and under the Sale of Goods Act 1893 and consequentially with the respondents' rights against third parties from whom in turn the respondents acquired the meal. It is not, and cannot be, contended that because the presence of this chemical in the meal was unsuspected, and latent, at the date of the contract, and for some time after, that of itself affords a defence (other than a special defence under the fair average quality provisions) either to the intermediate sellers or to the manufacturers.
The judge decided that the respondents were in breach of each of the conditions implied by section 13, section 14 (1) and section 14 (2) of the Act *489 but on each point his decision was reversed by the Court of Appeal. I have no doubt that under two of these headings his judgment should be restored. Because of the way in which the Sale of Goods Act has slotted the pre- existent common law remedy into compartments, it is necessary to consider separately the three relevant provisions. It is well known that there is a good deal of overlapping between them, so that this subdivision is artificial and gives rise to difficulty. But there is no avoiding this procedure.
1. Section 13 of the Act: The question is whether the compound mink food sold by the respondents (under the name "King Size") corresponded with the description. The appellants' case was that the food was to be made up according to a formula which identified generically the ingredients and specifically the chemical additives, quantifying precisely the proportions of each ingredient. One of these ingredients was herring meal. The food delivered in certain relevant months, it was claimed, did not correspond with the description because it contained a significant quantity of DMNA. The proposition is that "King Size" made partly of herring meal which contains DMNA does not correspond with the description "King Size." This can be reduced to the proposition that the herring meal ingredient did not correspond with the description because it contained DMNA. The analogy was invoked, inevitably, by the appellants of coppery cake with castor seed; the respondents invoked that of oxidised iron. The learned judge accepted the former, the Court of Appeal the latter.
Whether in a given case a substance in or upon which there has been produced by chemical interaction some additional substance can properly be described or, if one prefers the word, identified, as the original substance qualified by the addition of a past participle such as contaminated or oxidised, or as the original substance plus, or intermixed with, an additional substance, may, if pressed to analysis, be a question of an Aristotelian character. Where does a substance with a quality pass into an aggregate of substances? I do not think that it can be solved by asking whether the chemical interaction came about by some natural or normal process, e.g., preservation by the addition of salt (sodium chloride), or by some alien intrusion by the production of DMNA from sodium nitrite through a heating effect. I cannot see any distinction in principle in this difference. Further I do not believe that the Sale of Goods Act was designed to provoke metaphysical discussions as to the nature of what is delivered, in comparison with what is sold. The test of description, at least where commodities are concerned, is intended to be a broader, more common sense, test of a mercantile character. The question whether that is what the buyer bargained for has to be answered according to such tests as men in the market would apply, leaving more delicate questions of condition, or quality, to be determined under other clauses of the contract or sections of the Act. Perhaps this is to admit an element of impression into the decision, but I think it is more than impression which leads me to prefer the answer, if not all of the reasoning, of the Court of Appeal that the defect in the meal was a matter of quality or condition rather than of description. I think that buyers and sellers and arbitrators in the market, asked what this was, could only have said that the relevant ingredient was herring meal and, therefore, that there was no failure to correspond with description. In my opinion, the appellants do not succeed under section 13.
*490 2. Section 14 (1) of the Act: I do not think it is disputed, or in any case disputable, that a particular purpose was made known by the buyers so as to show that they relied on the sellers' skill and judgment. The particular purpose for which "King Size" was required was as food for mink.
Equally I think it is clear (as both courts have found) that there was reliance on the respondents' skill and judgment. Although the Act makes no reference to partial reliance, it was settled, well before the Cammell Laird case [1934] A.C. 402 was decided in this House, that there may be cases where the buyer relies on his own skill or judgment for some purposes and on that of the seller for others. This House gave to that principle emphatic endorsement.
The present is certainly such a case. In the words of Milmo J. [1968] 1 Lloyd's Rep. 457, 480:
"On the one hand Mr. Udall was relying on his own judgment as to what his formula should contain and the levels at which the various ingredients in it should be included. On the other, he was relying, and had no alternative but to rely, upon the [respondents] to obtain the ingredients, to see they were of good quality and not to use ingredients which, as a result of contamination, were toxic."
The word "toxic" will require some examination but, subject to this, I consider that this passage correctly states the position as regards reliance.
The field thus left to the sellers can be described in terms of their responsibility as merchants, to obtain and deliver ingredients, and relevantly herring meal, not unfit by reason of contamination, to be fed to animals, including mink. The field reserved to the buyers, on the other hand, was that of particular or specific suitability for mink. There was no doubt that herring meal, as such, was suitable for mink; on the other hand, the particular consignments supplied in 1961 were unsuitable because of the presence of DMNA. What, then, was the nature of this unsuitability?
If mink possessed an idiosyncrasy, which made the food as supplied unsuitable for them though it was perfectly suitable for other animals, this would be the buyers' responsibility, unless, as is not the case here, they had made this idiosyncrasy known to the sellers so as to show reliance on them to provide for it. But any general unsuitability would be the sellers' responsibility. Although the evidence was not very complete, it is sufficiently shown, in my opinion, that mink are more sensitive to DMNA than most other animals to whom compound foods would be sold. Chicken and pigs are among the least sensitive, next cattle and then sheep, with mink at the top of the scale. So the question arises, what does the buyer, alleging unfitness, have to prove? If the fact were that the herring meal supplied, while damaging to mink, was perfectly harmless to all other animals to whom it might be fed, it would be unjust to hold the sellers liable. If, on the other hand, the herring meal was not only lethal to mink but also deleterious, though not lethal, to other animals, the sellers' responsibility could be fairly engaged. A man can hardly claim that the product he sells is suitable, especially if that is a foodstuff, merely because it fails to kill more than one species to which it is fed.
In this case, because of the difficulty of tracing the lethal element, the *491 evidence as to its presence, its strength and its effect was not scientifically complete. It was not until 1964 that DMNA was identified. By that time all the infected herring meal had been disposed of, and all other animals to which it had been fed had died. The critical question in this part of the appeal is whether the buyers proved enough to show that their mink died because of some general, that is, non-specific, unsuitability of the herring meal through contamination. The burden was upon the buyers to show that this was so.
The Court of Appeal, who decided against the buyers on this point, based their decision on a conclusion regarding the balance of the contaminated herring meal which was not used for mink food. The consignment which killed the appellants' mink was part (8 1/2 tons) of a total of 333 1/2 tons, the rest of which was sold by the respondents for inclusion in feeding stuffs in the normal way. The Court of Appeal [1969] 2 Lloyd's Rep. 425, 462 found that this balance was "perfectly suitable" as a feeding stuff for animals for whom the respondents normally compounded foods, and from this drew the conclusion that the deaths of the mink were due to some specific idiosyncrasy. I cannot accept that this conclusion was justified. The evidence certainly showed that no complaints were received from any person who received meal from the balance of the consignment, but that was all. There was no evidence in what quantities, or over what period, any of this meal was fed, nor even, accepting that no deaths resulted (and premature deaths may have occurred), was there any evidence that the animals in question were not adversely affected in weight, fertility or in damage to their livers. The absence of complaint is insufficient by itself to establish perfect suitability.
But the matter does not end with rejection of the Court of Appeal's finding: the basic difficulty remains. Given that the buyers had to show general unsuitability and not merely specific unsuitability for mink, did they do so? It is here that the concept of toxicity becomes relevant.
Toxicity is, to some extent, a relative quality. There are few products which are lethal or even deleterious to all species, in all conditions, in whatever doses. Some "poisons" may be quite harmless to certain species if they do not exceed a threshold quantity and if they can be eliminated before a fresh dose is taken: some may even be beneficial if taken in small doses but lethal if taken in large. Other products, though not immediately damaging, become cumulatively so through build-up in the tissues or cells. From this point of view the Court of Appeal was no doubt right in its reservations as to the use of the word "toxic" where quoted above: the word may involve a petitio principii.
But when this is said, it also remains true that, in certain contexts, toxicity may be an absolute quality: particularly is this so where foodstuffs are concerned. Where an element in feeding stuffs is shown to be (i) lethal in some quantities to one or more species (ii) damaging in other quantities to one or more species and in more than one respect and (iii) when it is not suggested that in any circumstances the chemical is beneficial; when moreover the expert evidence shows that the full implication and effects of feeding it have yet to be scientifically established, then there is every justification for describing it as toxic, and, which is the relevant consideration, for placing responsibility for its exclusion firmly upon the seller.
*492 I am satisfied that DMNA was shown to be toxic in this sense. That, in a general sense of the word, it was toxic was never disputed at the trial. The issue, fought over so many days, was whether it was shown that DMNA was present in the 8 1/2 tons used for making food for the appellants' mink, and whether it was DMNA that caused their death: secondarily also whether it was shown to be present in the rest of the 333 1/2 tons consignment or in the product of the season 1960-61.
It was not an issue, on the pleading, or, as I understood it, in evidence, whether DMNA was toxic generally or whether the disputed herring meal was toxic only quoad the mink. The description of DMNA as toxic was that of Dr. Koppang, the expert witness called for the third parties - the effective antagonist of the appellants. It was he who in the witness box said that all animals are sensitive to DMNA and that mink were more sensitive than others. There was evidence of the specific damage inflicted by DMNA, namely, damage to the liver: not only mink have livers, and damaged livers of other species were instanced and photographed: this is generic damage, not idiosyncratic. There was evidence of disastrous losses to Norwegian fox and mink over six years from 1957 with similar symptoms to those of a toxic hepatosis in 1961 among ruminants to whom toxic herring meal had been fed. Dr. Roe, a pathologist, showed that DMNA was particularly hazardous to man in concentrated form and that he had instructed persons carrying out examinations of poisoned mink to be extremely careful. My noble and learned friend, Viscount Dilhorne, has fully extracted the essence of this evidence.
Against this the respondents sought to rely on certain passages in the evidence to the effect that as regards other animals than mink, including pigs, chickens and cows, DMNA below a certain quantity is harmless. But such evidence would have to be of a very convincing character before it satisfied (as it clearly did not satisfy the judge) that foodstuffs containing DMNA were suitable for their purpose, even if the concentration was low. In fact, the evidence was of a very imprecise kind. It was agreed that one species differs from another, one sub-species from another, one individual from another. As regards sensitivity to DMNA the evidence as regards the effect of feeding it was partial and far from comprehensive. The Norwegian expert admitted that scientific knowledge was imperfect. The serious ravages among ruminants and furbearers had taken place seven years or more before the trial, the opportunity for accurate research in the field did not exist. The equation is, therefore, this: that DMNA is inherently and generally toxic: that as included in the relevant herring meal it poisoned the appellants' mink: that this poisoning was not due to any specific idiosyncrasy among mink: that its effect on other species or sub-species is not accurately known beyond that it is potentially deleterious at least to the liver and that most normal species are more tolerant of it than mink.
In my opinion, the appellants made good their case: they proved the cause of their losses to lie in the inclusion of a generally (viz. non-specific as regards mink) toxic ingredient in the food. It was not for them to show that this same food killed, or poisoned, other species. So to require would place far too high a burden on a buyer. The buyer may have no means of ascertaining what the effect on other species may be. The whole of the contaminated consignment may have been fed to the buyer's animals: is the *493 buyer to fail because he cannot show that this particular consignment killed, or at least injured, other animals? He must, I think, carry his proof to the point of showing that the guilty ingredient has some generally (as opposed to specifically) toxic quality. But once he has done this, has he not shown, at least with strong prima facie force, that a feeding stuff which contained it was unsuitable? Is he not entitled to throw on to the seller the burden of showing, if he can, that the damage to the buyer's animals was due to some factor within the field of responsibility reserved to the buyer? I would answer yes to these questions. In the end, it is for the judge to decide whether, on the evidence, the buyers have proved their case. Milmo J.'s conclusions are expressed in three passages, one in the main action, the others in the third party proceedings (the whole case was heard together):
"Herring meal does not normally contain a poison. The herring meal which killed the English mink contained DMNA which is a poison, and it contained it at a level sufficiently high to be lethal to mink, which are animals to which herring meal can properly be fed. All animals are sensitive to DMNA poisoning, though mink are more sensitive than most." ([1968] 1 Lloyd's Rep. 457, 481.)
"I find that the meal which poisoned the English mink was not reasonably fit for use as an ingredient in animal foodstuffs because of the fact that it contained in substantial and significant quantities DMNA which is a toxic substance to which all animals are sensitive." (p. 487.) "While I accept that there was no evidence that the meal had a deleterious effect upon any animal or other type of livestock other than mink, I do not consider that it was proved affirmatively that the meal which killed the mink could have been fed with impunity to all other types of livestock." (p. 486 - emphasis supplied.)
This is precisely the position: coupled with the general finding as to toxicity (something to which all animals are sensitive, i.e., liable to suffer liver damage) it amounts to a rejection of the only line of defence open to the respondents - namely, that the relevant consignment was fit to be fed to all normal animals and only unfit to be fed to mink.
In my opinion, these findings were justified and correct.
So much for the facts, but there remains one legal argument on this part of the case. Section 14 (1) contains the words "and the goods are of a description which it is in the course of the seller's business to supply." The respondents relied on these words and persuaded the Court of Appeal to decide that the requirement was not satisfied because, briefly, the respondents were not dealers in mink food. A similar argument was put forward on the words in section 14 (2) "where goods are bought by description from a seller who deals in goods of that description." The Court of Appeal decided this point, too, in the respondents' favour. The respondents, they held, did not deal in mink food, or "King Size," before Mr. Udall placed with them the orders which produced the defective goods. I have some doubt whether this argument is even correct on the facts, because Mr. Udall had been ordering "King Size" for several months before he ordered the fatal consignment. But we must deal with the legal argument because it is clearly of general importance. It appears never previously to have been *494 accepted and it substantially narrows the scope of both subsections. It rests, in the first place, upon a linguistic comparison of the meaning of the word "description" in the three places where it appears and on the argument that it must mean the same in each place.
I do not accept that, taken in its most linguistic strictness, either subsection bears the meaning contended for. I would hold that (as to subsection (1)) it is in the course of the seller's business to supply goods if he agrees, either generally, or in a particular case, to supply the goods when ordered, and (as to subsection (2)) that a seller deals in goods of that description if his business is such that he is willing to accept orders for them. I cannot comprehend the rationale of holding that the subsections do not apply if the seller is dealing in the particular goods for the first time or the sense of distinguishing between the first and the second order for the goods or for goods of the description. The Court of Appeal offered the analogy of a doctor sending a novel prescription to a pharmacist, which turns out to be deleterious. But as often happens to arguments of this kind, the analogy is faulty: if the prescription is wrong, of course the doctor is responsible. The fitness of the prescription is within his field of responsibility. The relevant question is whether the pharmacist is responsible for the purity of his ingredients and one does not see why not.
But, moreover, consideration of the preceding common law shows that what the Act had in mind was something quite simple and rational: to limit the implied conditions of fitness or quality to persons in the way of business, as distinct from private persons. Whether this should be the law was a problem which had emerged, and been resolved, well before 1893. The first indication of the point arose in Jones v. Bright (1829) 15 Bing. 533 (copper sheathing). Two of the judges regarded it as an essential allegation that the defendant should have been the manufacturer of the defective copper. Park J. in fact, at p. 546, used the words "distinguishing, as I do, between the manufacturer of an article and the mere seller." In Brown v. Edgington (1841) 2 Man. & G. 279 , 291 (the crane rope) we find a description of the defendant by Bosanquet J. as "a dealer in articles of that description," clearly a reason for holding him liable though he was not the manufacturer. The distinction between the dealer and the private seller is clearly brought out in Burnby v. Bollet (1847) 16 M. & W. 644 , where a man bought a carcase in the market but later sold it to another farmer. His exemption from liability for defects in the carcase was explicitly based on his private character; he was "not clothed with any character of general dealer in provisions" (p. 649), he was "not dealing in the way of a common trade" (p. 655). and finally in the forerunner case of Jones v. Just (1868) L.R. 3 Q.B. 197 we find Mellor J. in his fourth and fifth categories, which anticipate respectively section 14 (1) and 14 (2) of the Sale of Goods Act 1893, referring to a manufacturer or dealer contracting to supply an article which he manufactures or produces, or in which he deals, and to a manufacturer undertaking to supply foods manufactured by himself or in which he deals, so clearly following and adopting the prior accepted division between sales by way of trade and private sales.
One asks, therefore, what difference the insertion in the Sale of Goods Act of the word "description" made to these well-accepted rules. It seems at least clear that the words now appearing in section 14 (1) "and the goods *495 are of a description which it is the seller's business to supply "cannot mean more than" the goods are of a kind. ..." "Description" here cannot be used in the sense in which the word is used when the Act speaks of "sales by description," for section 14 (1) is not dealing with sales by description at all. If this is so, I find no obstacle against reading "goods of that description" in a similar way in section 14 (2). In both cases the word means "goods of that kind" and nothing more. Moreover, even if this is wrong, and "description" is to be understood in a technical sense, I would have no difficulty in holding that a seller deals in goods "of that description" if he accepts orders to supply them in the way of business; and this whether or not he has previously accepted orders for goods of that description.
So, all other elements being present as I have tried to show, I would hold that section 14 (1) applies to the present case. I would agree with the judge that section 14 (2) equally applies and disagree with the reasons abased on the "description" argument) which led the Court of Appeal to a contrary opinion.
That the goods were unmerchantable was conceded in both courts - in my opinion, rightly so. Goods may quite well be unmerchantable even if "purpose built." Lord Wright made this quite clear in the Cammell Laird case [1934] A.C. 402; so equally with "King Size" mink food.
I would therefore allow the appeal.
The appeal of Christopher Hill Ltd. (the respondents) against Norsildmel ("the third parties") raises different, and, in one respect at least, more difficult issues. The goods supplied were in this case Norwegian herring meal and they were supplied under the terms of a commodity market contract in writing. A number of points arise under it. On the following I express my concurrence with others of your Lordships, and do not think it necessary to add reasons of my own.
1. The respondents were not in breach of a term in the contract implied by virtue of section 13 of the Sale of Goods Act 1893. The goods supplied were, in my opinion, Norwegian herring meal. The words "fair average quality of the season" were not in this contract part of the description. I do not find it necessary to consider whether, if they were, there was a breach of any implied condition that the goods should correspond with this description. They were not relied upon as themselves importing a warranty; but, if the contention is open, I am in agreement with my noble and learned friend, Lord Diplock, for the reasons which he gives, that they do not cover the particular defect which existed.
2. The exemption clause contained in general condition 3 does not exclude a claim for breach of any warranty implied under section 14 (1) of the Act.
This leaves the substantial question whether a term as to reasonable fitness ought to be implied under section 14 (1) of the Act. There was also raised a question as to remoteness of damage but, in the view which I take, this depends on the same considerations as those necessary for determination of liability under section 14 (1). I now consider this question.
In so doing I should make it clear that, although I refer to Norsildmel as the third parties, the actual contract for sale was made with a committee called Sildemelutvalget to whom Norsildmel succeeded in 1964, but no distinction has been made between these organisations. What is necessary *496 to determine is whether any particular purpose for which the goods were required was made known by the buyers to the sellers so as to show that the buyers relied on the sellers' skill and judgment: what the particular purpose was: finally, whether the particular purpose included feeding to mink. The particular purpose relied upon by the respondents was that the meal was required for inclusion in animal feeding stuffs to be compounded by them. They do not contend that feeding to mink was explicitly stated as a purpose; but they say that feeding to mink was known to both parties as a normal user for herring meal, and that it was sold without any reservation or restriction as to the use to which it might be put. The sale was negotiated through an agent in England - C. T. Bowring & Co. Ltd. on behalf of Sildmelutvalget, but no point has been taken as to any limitation upon their knowledge as compared with that of their principals.
The scope and application of section 14 (1) of the Sale of Goods Act 1893 was fully considered by this House in Hardwick Game Farm v. Suffolk Agricultural Poultry Producers Association Ltd. [1969] 2 A.C. 31. The opinion expressed in that case endorsed a tendency which other cases (such as Manchester Liners Ltd. v. Rea Ltd. [1922] 2 A.C. 74) had shown, to expand the scope of section 14 (1) so as to cover territory which might otherwise, on a first reading, have been thought to belong to section 14 (2). I think that this tendency essentially reflects a reversion to the more general approach to questions of the seller's liability under implied warranty adopted by the common law, as contrasted with the compartmentalisation into separate, but inevitably overlapping, provisions adopted by the Sale of Goods Act. Naturam expels furca is a maxim which tends to apply to codifications. At any rate it is dear that this House in the Hardwick case [1969] 2 A.C. 31 accepted that the "making known" so as to show reliance which the section requires is easily deduced from the nature and circumstances of the sale, and that the word "particular" means little more than "stated" or "defined." As Lord Pearce said in Hardwick, at p. 115: "There is no need for a buyer formally to 'make known' that which is already known": and here there is no doubt that the third parties, through their selling agents C. T. Bowring & Co. Ltd., and also directly, knew what the herring meal was required for, namely, for inclusion in animal feeding stuffs to be compounded by the buyers, and no special purpose in relation to mink was relied on. The third parties were, moreover, a committee, or co-operative, of manufacturers of herring meal: in this case, whether one speaks of implication or presumption, the conclusion can hardly be otherwise than that of reliance by the buyers to produce a product reasonably fit for the purpose. I observe, indeed, that my noble and learned friend, Lord Guest, who felt difficulty in Hardwick as to the application of section 14 (1) against persons who were dealers in the market, said that he could well understand, where the sale is by a manufacturer to a customer, that the inference (sc. of reliance) can easily be drawn (p. 106). I agree with Milmo J. that it ought to be drawn in this case.
Then was the purpose, to be used for inclusion in animal feeding stuffs to be compounded by the buyers, a particular purpose? In my opinion, certainly yes. It is true that the purpose was wide, wider even than the purpose accepted as particular in Hardwick (for compounding into food for cattle and poultry), and, if one leaves aside a possible alternative use as *497 fertiliser, on which there was some indefinite evidence, the purpose so made known covers a large part of the area which would be within section 14 (2). But I do not think, as the law has developed, that this can be regarded as an objection or that in accepting a purpose so defined, as a "particular purpose," the court is crossing any forbidden line. There remains a distinction between a statement (express or implied) of a particular purpose, though a wide one, with the implied condition (or warranty) which this attracts, and a purchase by description with no purpose stated and the different condition (or warranty) which that attracts. Moreover, width of the purpose is compensated, from the seller's point of view, by the dilution of his responsibility: and to hold him liable under an implied warranty of fitness for the purpose of which he has been made aware, wide though this may be, appears as fair as to leave him exposed to the vaguer and less defined standard of merchantability. After all, the seller's liability is, if I may borrow the expression of my noble and learned friend, Lord Morris of Borth-y-Gest, no more than to meet the requirement of a buyer who is saying to him "that is what I want it for, but I only want to buy if you sell me something that will do." I think that well expresses the situation here.
The next point is whether, when the meal turned out to be unsuitable for feeding to mink, this was a matter to be treated as within the seller's responsibility. There are two distinct points here: the first is whether feeding to mink was a normal use, within the general purpose of inclusion in animal feeding stuffs: the second is whether, assuming that the seller's implied warranty did not extend beyond that of general suitability for animals, including possibly mink, the buyers were able to show a breach of that warranty.
The first point involves an issue of fact which received lengthy examination in the courts below. The decision on it depended to a great extent upon the view taken of two Norwegian witnesses called by the third parties, who were the assistant director of the third parties and the chief executive of a Norwegian herring oil factory at the relevant time. These witnesses were called to show that the third parties did not know in 1961 that herring meal might be fed to mink.
Unfortunately the courts below reached different conclusions. Milmo J. did not accept the disclaimer of the Norwegian witnesses. He found that both were aware in or before 1961 that herring meal was being fed to mink in Norway and that herring meal was a normal and well-known ingredient of the diet of mink kept in captivity in Norway and (he added) in other countries. On this basis he found that Sildmelutvalget knew of the practice of feeding herring meal to mink.
The Court of Appeal reached an opposite conclusion. They held that the respondents had failed to establish that at the date of the contract the third party should reasonably have contemplated that it was not unlikely that the herring meal, the subject-matter of the contract, might be fed to mink. On the contrary, they found on the evidence that the result of such contemplation would have been that it was highly unlikely that the herring meal would be fed to mink, and they added (though this would appear to be both an irrelevant matter and, since the meal was not known to be toxic, also factually incorrect) to which alone it would be harmful.
On this issue, the careful re-examination of the evidence which took place in this House, convinced me that the Court of Appeal was not justified in *498 reversing in this matter the findings of fact of the trial judge. The latter were supported by the impression made on him by the two Norwegian witnesses in the witness-box, by some important letters written by the third parties in late 1960 on the subject of herring meal and its potentiality as mink food, and by the general probabilities of the case, the fact being that there were numerous mink and fox farms in Norway to which herring meal had been fed. In my opinion, we must reinstate the judge's conclusion, that feeding to mink was a normal user in 1961 and known as such to the third parties. That the relevant consignment or part of it might be fed to mink was also known to the respondents.
I should add that some, but slight, reliance was placed on the fat content of the meal supplied. The argument was that herring meal with a high fat content is unsuitable for mink, that the herring meal in question had a high fat content and that this showed that the buyers could not be relying on the seller's judgment. There are several answers to this. In the first place, the actual fat content (9.6 per cent.) was below that fixed by the contract (12-13 per cent.) which seems to show that the buyer regarded the herring meal as in this respect suitable. And, secondly, even if it were excessive, this could be a matter within the buyer's responsibility quite consistently with the purity of the meal being within the seller's responsibility. The argument is, in my opinion, conclusive.
If I am right so far on the question of suitability and reliance, similar considerations arise on the question whether the consignment was in fact unsuitable, so as to involve a breach of warranty, to those already discussed as between the appellants and the respondents, and for the same reasons the conclusion follows, in my opinion, that a breach of warranty under section 14 (1) was proved. The respondents did not, in this part of the appeal, pursue a claim under section 14 (2).
Finally, any question as to remoteness of damage is disposed of by the finding that feeding to mink was a normal user and contemplated as such by both parties to the contract.
I would allow both appeals.

LORD DIPLOCK.

My Lords, it was no one's fault that the Norwegian herring meal which was an ingredient of the compound feeding-stuff for mink made the food poisonous to those animals. At the relevant time the possibility of the chemical reaction which produced the toxic substance DMNA in the course of manufacture of the meal was unthought of. In the then state of knowledge, scientific and commercial, no deliberate exercise of human skill or judgment could have prevented the meal from having its toxic effect upon mink. It was sheer bad luck.
The question in each of these appeals is: Who is to bear the loss occasioned by that bad luck? Is it to be the buyer of the compound feeding-stuff, who fed it to his mink? Or the seller, who compounded the feeding-stuff using as an ingredient herring meal which he himself had bought for this and other purposes? That is the issue in the first appeal between the appellants as buyers and the respondents as sellers whom, for ease of reference I will personalise as "Udall" and "Hill." and if it is to be Hill as seller of the compound feeding-stuff, can he as buyer of the herring meal recover the loss *499 from his own seller, the third party whom I will call "the Norwegians "? That is the question in the second appeal.
Initially the facts were complex and obscure. But they have been sifted and analysed by Milmo J. and the Court of Appeal and, save upon one matter, they are clear. What is not proved one way or the other is whether or not the herring meal sold by the Norwegians to Hill would have had any deleterious effect at all upon any animals, other than mink, to which it might reasonably be expected to be fed, if it were fed to them as an ingredient of their total diet in any proportion that herring meal might reasonably be expected to be used.
Such chemical and biological facts as are now known are simple - and scientific. "Poison" and "toxic" are emotive words in ordinary speech. Revulsion is one's instinctive reaction to any suggestion that a poison or toxin may be present in a food. But it is now known that there are many substances which are harmless, or in some cases even beneficial, when ingested in small quantities yet are damaging to health, or even fatal, if consumed in higher concentrations. Up to a certain concentration in the diet the natural processes of the body will eliminate the toxic substance before it can do any detectable harm. Above that concentration it is poisonous. Below that concentration it is harmless - it has no deleterious effect. The critical cocentration or "threshold" for any toxic substance may vary widely between one species of animal and another, and within a single species there may be variations according to age groups or between individual animals.
It is in this scientific, non-emotive sense that DMNA (dimethylnitrosamine) is a "poison" or "toxic." It is derived from amino-acids by chemical reaction in which enzymes may play a part. Amino-acids are present in all proteins and, as the evidence in the instant case disclosed, traces of DMNA are commonly found in many human food-stuffs such as instant coffee and cereal breakfast foods. But it is present in proportions which are well below the critical concentration at which the food-stuff is harmful to the consumer or the nutritive value of its other ingredients is affected.
The critical concentration or threshold of DMNA for the different species of animals and birds has not yet been fully investigated by scientists. So far as it goes the undisputed scientific evidence in the instant case shows that there is a wide variation in the threshold for different species and that the threshold for mink is considerably lower than that of any other species of animal or bird which has been the subject of scientific investigation. In particular the evidence was that it was considerably lower than for cattle, sheep, pigs and poultry. This does not, of course, mean that DMNA is not poisonous or at least deleterious to these other animals if consumed in sufficiently high concentrations. It is in this sense that "all animals are sensitive to DMNA" as Dr. Koppang, the acknowledged expert on this subject, stated; for it is only in this sense that this answer is reconcilable with all the rest of his lengthy and detailed evidence of the results of his scientific investigations.
There is thus nothing inherently improbable in a consignment of herring meal containing DMNA in a concentration which, on the one hand, was high enough to be fatal to mink when fed to them as 3 per cent. of their diet, yet, on the other hand, was low enough to have no deleterious effect upon the wholesomeness and nutritive value of the herring meal as a feeding-stuff for *500 all domestic livestock and birds, other than mink, to whom it might reasonably be expected to be fed, if it were fed to them as an ingredient of their total diet in any proportion that herring meal might reasonably be expected to be used.
It was not possible, seven years after the event, to determine by chemical analysis what was the concentration of DMNA in any part of the Norwegian herring meal with which these two appeals are concerned. It had all been consumed: 8 1/2 tons by mink in "King Size" mink food and the remaining 325 tons by other domestic livestock and birds, namely, pigs, calves, poultry, pheasants and rabbits. One can only infer what was the concentration of DMNA in the herring meal from its effect upon the animals to which it was fed. Both Milmo J. and the Court of Appeal were satisfied upon the evidence that the composition of the herring meal which was fed to mink and that which was fed to other animals was substantially the same. It may have varied slightly from one bag to another, but the 2 1/2 per cent. of the consignment which was fed to mink was fairly representative of the whole.
Nearly all the mink who consumed "King Size" mink food of which the Norwegian herring meal was an ingredient either died or suffered in their condition. One can infer, therefore, that the concentration of DMNA in the herring meal was above the "threshold" for mink when fed as about 3 per cent. of their diet. All that one knows about the effect of the Norwegian herring meal upon the well- being of other animals to whom it was fed, either as an ingredient of other compound feeding-stuffs sold by Hill or as mixed with other feeding-stuffs by customers of Hill to whom the herring meal was sold straight, is that whatever may have been the percentage of their diet constituted by the herring meal no complaints were received by Hill from any of his customers. The identity of the customers was known to all the parties to the appeals. None was called to give evidence at the trial. This may well have been because at that stage of the proceedings it was not appreciated that this would be the crucial issue of fact upon which both these appeals, in my opinion, turn.
Milmo J.'s finding upon this issue was stated thus [1968] 1 Lloyd's Rep. 457, 486:
"While I accept that there was no evidence that the meal had a deleterious effect upon any animal or other type of livestock other than mink, I do not consider that it was proved affirmatively that the meal which killed the mink could have been fed with impunity to all other types of livestock."
I am content to accept both parts of this finding as to the effect of the evidence. The first part was expressly concurred in by the Court of Appeal. Some of the language in their judgment is consistent with its being their opinion that the innocuous character of the herring meal as an ingredient in the diet of the other animals and birds to which it was fed had been established affirmatively. But it does not seem that their minds were directed to the question as to the party upon whom the onus of proving or disproving this lay in either of the appeals. I do not think that the evidence as recorded justifies an appellate court departing from the finding of the judge who heard the witnesses at the trial.
*501 My Lords, the claim in each of these appeals is for damages for breach of a contract for the sale of goods. It will, therefore, be necessary to relate them to the relevant statutory provisions of the Sale of Goods Act 1893. In the form in which the Bill was originally drafted by Sir MacKenzie Chalmers that Act was intended to state the common law rules relating to the sale of goods as they had been developed by judicial decision up to 1889. Although a number of amendments were made in committee during the passage of the Bill through Parliament, they did not alter the essential character of the Act as expository of the common law of England at the date at which it was passed. But the exposition contained in the Act is only partial. It does not seek to codify the general law of contract of England or of Scotland. It assumes the existence as a basic principle of the English law of contract that, subject to any limitations imposed by statute or by common law rules of public policy, parties to contracts have freedom of choice not only as to what each will mutually promise to do but also as to what each is willing to accept as the consequences of the performance or non-performance of those promises so far as those consequences affect any other party to the contract. The paramountcy of this freedom of choice as to promises made in contracts for the sale of goods is acknowledged by section 55 of the Act. The provisions of the Act are in the main confined to statements of what promises are to be implied on the part of the buyer and the seller in respect of matters upon which the contract is silent, and to statements of the consequences of performance or non-performance of promises, whether expressed or implied, where the contract does not state what those consequences are to be. Even a code whose content is so limited must proceed by classifying promises, both those which are expressed and those to be implied; the circumstances which give rise to implied promises, and how they are to be performed and the consequences of performing each class of promise or of failing to perform it. Because of the source of the rules stated in the Sale of Goods Act 1893 the classification adopted is by reference to the promises made in relatively simple types of contracts for the sale of goods which were commonly made in the nineteenth century and had been the subject of judicial decision before 1893. But although the language in which the rules are expressed is appropriate to these simple types of contracts, it has to be applied today to promises made in much more complicated contracts which cannot be readily allotted to any single class of contract which appears to be primarily envisaged by a particular section or subsection of the code. Unless the Sale of Goods Act 1893 is to be allowed to fossilise the law and to restrict the freedom of choice of parties to contracts for the sale of goods to make agreements which take account of advances in technology and changes in the way in which business is carried on today, the provisions set out in the various sections and subsections of the code ought not to be construed so narrowly as to force upon parties to contracts for the sale of goods promises and consequences different from what they must reasonably have intended. They should be treated rather as illustrations of the application to simple types of contract of general principles for ascertaining the common intention of the parties as to their mutual promises and their consequences, which ought to be applied by analogy in cases arising out of contracts which do not appear to have been within the immediate contemplation of the draftsman of the Act in 1893.
*502 In each of the instant appeals the dispute is as to what the seller promised to the buyer by the words which he used in the contract itself and by his conduct in the course of the negotiations which led up to the contract. What he promised is determined by ascertaining what his words and conduct would have led the buyer reasonably to believe that he was promising. That is what is meant in the English law of contract by the common intention of the parties. The test is impersonal. It does not depend upon what the seller himself thought he was promising, if the words and conduct by which he communicated his intention to the seller would have led a reasonable man in the position of the buyer to a different belief as to the promise; nor does it depend upon the actual belief of the buyer himself as to what the seller's promise was, unless that belief would have been shared by a reasonable man in the position of the buyer. The result of the application of this test to the words themselves used in the contract is "the construction of the contract." So far as the reasonable belief of the buyer as to what the seller's promise was, would have been influenced by any conduct of the seller before the contract was made, any implication as to the nature of his promise falls to be determined by applying to his conduct the general principles for ascertaining the common intention of parties to a contract for the sale of goods which underlie the relevant provisions of the Sale of Goods Act 1893.
My Lords, since I believe that the basic principle of the English common law of contract, including that part of it which is codified in the Sale of Goods Act 1893, is to give effect to the common intention of the parties as to their mutual promises in the sense that I have just described, I prefer to deal with each appeal by considering first the transaction between the buyer and the seller in the light of common sense and good faith in business, before examining the particular provisions of the code upon which the parties rely.
In the first appeal the goods which were the subject-matter of the contract of sale between Hill and Udall were a compound feeding-stuff, "King Size" mink food, which Hill, the seller, knew were intended by Udall, the buyer, to be fed to mink. The feeding-stuff was to be compounded by the seller, according to a formula prepared by the buyer, of ingredients specified by the buyer and to be provided by the seller. Most of the specified ingredients, including the herring meal, were of a kind which the seller had previously used in the course of his business of compounding feeding-stuffs to be fed to domestic animals and birds other than mink.
The buyer was an expert on the nutritional requirements of mink. The seller disclaimed all knowledge of this subject and the buyer knew that he had never before compounded or sold a feeding-stuff to be fed to mink. The seller, however, professed a general expertise in the compounding of feedingseller for domestic animals and poultry (including pheasants) and in the selection and purchase in the market of the various ingredients needed in the preparation of compound feeding-stuffs intended for that purpose. Nothing was said in the contract itself about the suitability as food for mink of the compounded feeding-stuff or any of its ingredients.
If the law of sale of goods is as sensible and simple as it ought to be, the question of law for the court should be: What, if any, responsibility as to the characteristics of the goods to be supplied under the contract would the seller reasonably understand that the buyer believed that he, the seller, was *503 accepting? Or, put in the obverse form: What responsibility as to the characteristics of the goods to be supplied would the buyer reasonably believe that the seller was accepting by entering into the contract? and in the circumstances of the present case that question of law can be reduced to this: Would Hill's reasonable understanding and Udall's reasonable belief be that Hill was accepting responsibility to deliver feeding-stuffs characterised by their being compounded in accordance with the formula provided by the appellants and composed of ingredients of the kinds specified and of a quality suitable for use in compound feeding-stuffs for domestic animals and poultry? Or were their respective reasonable understanding and belief that Hill was accepting an additional responsibility that feeding-stuffs so compounded and composed would be suitable for feeding to mink?
So stated, as it seems to me, the question answers itself. The additional responsibility was not one which, in the circumstances of this case, either party would understand or believe that the seller was accepting. So, if the law of sale of goods is in accord with common sense and good faith in business, Udall, upon the findings of fact to which I have referred, has failed to prove that the goods delivered by Hill did not conform with the only characteristics for which Hill had accepted responsibility, and Hill has failed to prove that the goods did so conform.
I turn, then, to the particular provisions of the Sale of Goods Act 1893. Sections 13 and 14 (2) apply to contracts for the sale of goods "by description." This expression is also used in section 18, rule 5, but the Act nowhere defines its meaning. Not all statements about the characteristics of goods which are the subject-matter of a contract of sale form part of the "description" by which they are sold. Sections 13 and 14 draw a distinction between the description by which goods are sold, on the one hand, and their fitness for the particular purpose for which they are required and their quality, on the other. Section 14 (4) recognises that a contract for the sale of goods may contain an express statement about the fitness of the goods for a particular purpose or their quality which does not form part of the description by which they are sold but constitutes a separate stipulation in the contract which may be either a condition or a warranty, depending upon the construction of the contract (see section 11 (1) (b)).
A contract for the sale of goods is one whereby the property in goods which have been physically identified is transferred from the seller to the buyer (see: sections 1 (1) and 16). But a contract may be made for the sale of unascertained goods before the actual goods in which the property is to be a transferred are physically identified and agreed upon. At the time of making such a contract the kind of goods which are its subject-matter can only be identified verbally and/or by reference to a sample. The "description" by which unascertained goods are sold is, in my view, confined to those words in the contract which were intended by the parties to identify the kind of goods which were to be supplied. It is open to the parties to use a description as broad or narrow as they choose. But ultimately the test is whether the buyer could fairly and reasonably refuse to accept the physical goods proffered to him on the ground that their failure to correspond with that part of what was said about them in the contract makes them *504 goods of a different kind from those he had agreed to buy. The key to section 13 is identification.
Udall bases his claim against Hill in the first instance on section 13 of the Act. The goods, he submits, did not correspond with the description by which they were sold. The contract was oral. The subject-matter was unascertained goods, and it is common ground that the description by which they were sold was contained in Udall's formula which set out in detail the ingredients of the feeding-stuff to be compounded by Hill. One of the described ingredients was "herring meal." Udall contended that the description also included the expression "mink food" - but for the purposes of the claim under section 13 I do not think this matters.
DMNA was not an ingredient referred to in the formula. Milmo J., following the cases about the sale of "copra cake" which was contaminated with castor seed, held that a feeding-stuff which contained DMNA in quantities which rendered it toxic to mink did not correspond with the description by which the goods were sold.
I agree with your Lordships and the Court of Appeal that this is not so. Udall's formula was commercial not chemical. The ingredient described as "herring meal" did not cease to comply with that description because it was manufactured from herrings to which a preservative had been added to prevent them from deteriorating. The most usual preservative is common salt (sodium chloride) but the evidence showed that another salt of sodium, sodium nitrite, had been used in Norway for several years before 1961. In certain conditions which can occur during the normal process of manufacture of herring meal the amino-acids naturally present in the herring break down into an organic chemical, dimethylamine, which can react with sodium nitrite to form DMNA (dimethylnitrosamine). The occurrence of this reaction may affect the quality of the meal. It does not alter its identity as "herring meal."
Alternatively Udall sought to bring his claim under section 14 (1) or 14 (2). These two subsections are interrelated. Unlike section 13 they are not concerned with the identity of the goods which are the subject-matter of the contract, but with their "quality" or their "fitness for any particular purpose."
All goods which are sold finish up by being used for some purpose or another by the ultimate user, consumer or converter. This is not necessarily the actual buyer under the contract of sale. He may intend to pass them on by sale or otherwise to the person who actually puts them to that use. But the purpose for which they will be used by the ultimate consumer or converter is the "purpose for which the goods are required" within the meaning of section 14 (1).
The only condition as to "quality" which is to be implied under section 14 (2) is that the goods shall be of "merchantable quality." This is likewise concerned with use. Goods are of merchantable quality if they are fit for use for any purpose for which goods which correspond with the description by which they were sold would normally be used.
The dichotomy between the two sections is thus between fitness for use for a "particular purpose" and fitness for use for one of several purposes for which goods which correspond with the description by which they were sold would normally be used. But if there is only one purpose for which *505 goods of that description would normally be used, the condition to be implied under each of the subsections is the same. In such a case one would expect to find some common principle which links the circumstances which give rise to the implied condition under subsection (1) and the circumstances which give rise to the implied condition under subsection (2).
So far as concerns the conduct of the buyer, the circumstances which give rise to the implied condition under subsection (1) are, first, that he should make known expressly or by implication to the seller what is the particular purpose for which the goods are required and, secondly, that he should do so in such a way as to make the seller reasonably understand that he is relying upon the seller to exercise sufficient skill or judgment to ensure that the goods are fit for that particular purpose. This he generally does by selecting a seller who makes it his business to supply goods which are used for purposes of that kind. It does not matter that the seller does not possess the necessary skill or judgment nor does it matter that in the then state of knowledge no one could by exercise of skill or judgment detect the particular characteristic of the goods which rendered them unfit for that purpose. This may seem harsh upon the seller, but its harshness is mitigated by the requirement that the goods must be of a description which it is in the course of the seller's business to supply. By holding himself out to the buyer as a manufacturer or dealer in goods of that kind he leads the buyer reasonably to understand that he is capable of exercising sufficient skill or judgment to make or to select goods which will be fit for the particular purpose for which he knows the buyer wants them.
I have used the word "kind" as meaning the same as the word "description" in the subsection. The subsection applies to all types of contracts for the sale of goods: to contracts for the sale of specific goods or for sale by sample, which are not contracts of sale of goods "by description," as well as to contracts which are. Where the contract is not one for the sale of goods by description the expression in the subsection "goods of a description which it is in the course of the seller's business to supply" can only mean that the seller does deal in goods of a kind that can be verbally identified by a description that is wide enough to include goods which are intended for use for the particular purpose for which the buyer requires the goods which are the subject-matter of the contract. I do not think that this expression bears any different meaning when the contract is one for sale of goods by description. I agree with the Court of Appeal that it is not limited to the "description" by which the actual goods that are the subject-matter of the contract are sold. It would have been no defence to the sellers in Cammell Laird & Co. Ltd. v. Manganese Bronze and Brass Co. Ltd. [1934] A.C. 402 to say that they had never previously supplied a ship's propeller of the particular dimensions ordered although these did form part of the description by which the goods which were the subject-matter of that contract were sold. It was sufficient that it was in the course of their business to supply ships' propellers.
In contrast to subsection (1) the corresponding conduct of the buyer which gives rise to the implied condition under subsection (2) is that he should have bought the goods "by description." The subsection does not apply to any other type of contract for the sale of goods. By describing the goods that are to be the subject-matter of the contract the buyer makes it known to the seller by implication that the goods are required for, at any rate, one *506 of the purposes for which the goods of that description are normally required - or, if there is only one such purpose, that they are required for that purpose. But, as in the case of subsection (1), the mere ordering of goods by description does not suffice to show the buyer's reliance upon the seller's knowledge of the purposes for which goods of that description are normally used or of the characteristics needed to make them fit for any of those purposes. It is only by selecting a seller who makes it his business to supply goods of the same description as that by which the goods which are the sub-ject-matter of the contract are bought, that the buyer shows his reliance upon the skill and judgment of the seller to supply goods which are reasonably fit for one of the purposes disclosed by the description by which the buyer has bought the goods. For this description constitutes the only information which the buyer has given the seller about the purpose for which the goods are required.
It follows that I agree with the Court of Appeal that in the phrase in subsection (2): "Where goods are bought by description from a seller who deals in goods of that description," the words "that description" refer to and mean the actual description by which the goods which are the sub-ject-matter of the contract were bought. Not only is it impossible to ascribe any other meaning to it grammatically but also, as I have endeavoured to explain, it makes good commercial sense. The expression "that description" in subsection (2) as referring to goods in which the seller deals, is thus narrower in meaning than the expression "a description" in subsection (1) as referring to goods which it is in the course of the seller's business to supply.
The key to both subsections is reliance - the reasonable reliance of the buyer upon the seller's ability to make or select goods which are reasonably fit for the buyer's purpose coupled with the seller's acceptance of responsibility to do so. The seller has a choice whether or not to accept that responsibility. To enable him to exercise it he must be supplied by the buyer with sufficient information to acquaint him with what he is being relied upon to do and to enable him to appreciate what exercise of skill or judgment is called for in order to make or select goods which will be fit for the purpose for which the buyer requires them.
This consideration, in my view, throws light upon two matters arising under section 14. The first is the meaning of "particular purpose" in subsection (1). The second is the application of the doctrine of "partial reliance" under both subsection (1) and subsection (2).
To attract the condition to be implied by subsection (1) the buyer must make known the purpose for which he requires the goods with sufficient particularity to enable a reasonable seller, engaged in the business of supplying goods of the kind ordered, to identify the characteristics which the goods need to possess to fit them for that purpose. If all that the buyer does make known to the seller is a range of purposes which do not all call for goods possessing identical characteristics and he does not identify the particular purpose or purposes within that range for which he in fact requires the goods, he does not give the seller sufficient information to enable him to make or to select goods possessing a characteristic which is needed to make them fit for any one of those purposes in particular, if the same characteristic either is not needed to make them fit, or makes them unfit, for other purposes within the range.
*507 A "range of purposes" case thus poses a stark question of legal policy as to whether the seller's responsibility ought to be to supply goods which are fit for at least one of the purposes within the range or to supply goods which are fit for all of those purposes unless he expressly disclaims responsibility for their fitness for any one or more of them. The answer to this question of policy has, in my view, been pre-empted by subsection (2) of section 14 of the Sale of Goods Act 1893.
The commonest way in which a buyer makes known to the seller a range of purposes for which the goods are required is by the description by which he buys them and by nothing more. This is the case that is contemplated by subsection (2). This, as it has been authoritatively construed by the courts, provides that the only condition to be implied as to the responsibility of the seller is that the goods should be reasonably fit for one of the purposes within the range.
To supplement the description by which the goods are bought, or to replace it if they are not bought by description, the buyer may identify with greater precision the purpose for which the goods are required, by making it known to the seller in some other way. This is the case contemplated by subsection (1). He may do this expressly or by implication. At any rate, if he does so expressly he can make it known to the seller that he relies upon the seller to supply goods that are fit for more than one purpose or, indeed, for all possible purposes which lie within a range. But the mere fact that the seller knows that the buyer is engaged in a business in which goods of the description by which they are bought may be needed for any one of a number of purposes within the range of those for which goods of that description are normally used, adds nothing to what he might reasonably infer from the fact that the buyer ordered the goods by a description which covers goods fit for a range of purposes, without particularising which of those purposes he requires goods for. It might be otherwise if the seller knew that the buyer was engaged in a business in which goods of the description by which they were bought were needed for one or more only of the purposes within the whole range.
It would, in my view, conflict with the principle of reliance which underlies section 14 (1) and (2) and would be a misuse of a statutory code of this kind, to treat a range of purposes for any one of which the buyer may require the goods, on the one hand, as constituting "the particular purpose for which the goods are required" and so giving rise to an implied condition under subsection (1) that they shall be reasonably fit for all purposes within the range, if the seller's knowledge of the range is derived in whole or in part from some circumstance other than the description by which the goods are ordered; and, on the other hand, as giving rise to an implied condition under subsection (2) that the goods need only be fit for one of the purposes within the range, if the seller's knowledge of the range is derived solely from the description by which the goods are ordered. So to construe the code would for practical purposes deprive subsection (2) of any effect.
I turn next to "partial reliance." The actual words of subsection (1) appear to contemplate two classes of contracts only; one, where the buyer does not rely at all upon the skill or judgment of the seller to see to it that the goods supplied are reasonably fit for a particular purpose; the other where the buyer does so rely and the other requirements of the subsection are *508 satisfied. As a matter of linguistics it is possible to construe the expression "so as to show that the buyer relies" as referring to a reliance which was only partial, in the sense that the reliance was not the only or even the determinative factor which induced the buyer to enter into the contract. But it is not possible to extract from the language of the subsection any qualification upon the implied undertaking by the seller, if there is such reliance, that the goods supplied by him shall be reasonably fit for the particular purpose for which they are required by the buyer. Yet as a result of technological advances since 1893 there are an increasing number of cases where the preparation of goods fit for a particular purpose calls for the exercise of more than one kind of expertise. The buyer may himself possess one of the kinds of expertise needed but lack another and may choose a seller who has led him to believe that he, the seller, possesses it. The only reliance by the buyer upon the skill or judgment of the seller is that in the preparation or selection of the goods he will exercise that kind of expertise which he has led the buyer reasonably to believe that he possesses. The goods supplied may then be unfit for the particular purpose for which both parties knew they were required, either because of a defect which lay within the sphere of expertise of the seller or because of a defect which lay within the sphere of expertise of the buyer himself.
The way in which the principle of reliance which underlies subsections (1) and (2) should be applied to a more complex contract of this kind, which was not in the immediate contemplation of the draftsman of the code, poses another stark question of legal policy. In large part this decision was made by your Lordships' House in 1934 in the Cammell Laird case [1934] A.C. 402. It was there laid down that if the defect in the goods which rendered them unfit for their purpose was due to a characteristic which it lay within the sphere of expertise of the seller to detect and avoid, the responsibility for their unfitness lay with the seller. The ratio decidendi leads ineluctably to the corollary that if the defect was due to a characteristic which it lay within the sphere of the expertise of the buyer to detect and avoid, the seller was not contractually responsible for it. It did not attract the implied condition under subsection (1). The field of the seller's undertaking as to the fitness of the goods for the purpose corresponded with the field of the buyer's reliance upon the skill and judgment of the seller.
My Lords, this seems to me to be consistent with common sense and business honesty. It was accepted as the correct principle by both courts below and by all parties to the appeals in this House. But the Cammell Laird case leaves open for decision an ancillary question of legal policy which your Lordships are now called upon to decide for the first time. That question is whether, in a case of partial reliance of this kind, once the goods have been proved to be unfit for the purpose for which they were required the onus lies upon the buyer to prove that the defect was due to a characteristic which it lay within the field of expertise of the seller to detect and avoid; or does it lie upon the seller to prove that the defect was due to a characteristic which lay within the sphere of expertise of the buyer.
I do not think that there is anything in the Sale of Goods Act 1893 or in hitherto accepted law which inhibits your Lordships from deciding this question whichever way commends itself to the majority. The choice depends largely upon one's personal view as to whether the swing of the pendulum *509 since 1893 from caveat emptor to caveat venditor has now gone far enough and ought to be arrested, or whether it should be given a further impetus, albeit a minor one, upon its current course. For my part I would have been in favour of arresting it. But I recognise that a decision to the contrary is simply one of policy and, as it commends itself to the majority of your Lordships, I accept it with good grace as now forming part of the law of contracts for the sale of goods.
Once it is accepted, it is decisive of the first appeal. Hill's sphere of expertise upon which Udall relied lay in the selection of ingredients for "King Size" mink food which would be of a quality suitable for use in compound feeding-stuffs for domestic animals and poultry other than mink. The particular purpose, known to both parties, for which Udall required the "King Size" mink food was as a feeding-stuff for mink. Udall proved that it was unsuitable for that purpose and that its unsuitability was due to the quality of an ingredient, herring meal, selected by Hill. Upon the findings of fact of Milmo J., Hill failed to prove affirmatively that the herring meal was of a quality suitable for use in compound feeding-stuffs for domestic animals and poultry other than mink. Therefore, Udall is entitled to succeed upon his claim for damages for breach of the condition to be implied under section 14 (1) of the Sale of Goods Act 1893.
On the other hand, I do not think that the contract between Udall and Hill was one which attracted any implied condition as to merchantable quality under section 14 (2) of the Act. The expression "merchantable quality" in relation to goods implies that at the time of the contract of sale there already exists a market in goods of the kind described from which it is possible to identify a standard of comparison or norm with which the goods should correspond in order to be acceptable to a reasonable buyer of such goods in that market. There was no existing market in "King Size" mink food. It was an entirely new product in which Hill made plain to Udall he had never dealt before. The contract was for the sale of goods by description; but whether the description was limited to what was stated in Udall's formula as to the ingredients and their proportions or whether it also included a reference to the goods as a feeding-stuff for mink the description was not one of goods in which Hill dealt within the meaning of section 14 (2).
To hold the contrary would, in my view, conflict with the principle of reliance which underlies both subsections (1) and (2) of section 14. If the "King Size" compound had been unsuitable as a feeding-stuff for mink solely because any food compounded in accordance with Udall's formula would have been unsuitable for mink, it would have been unmerchantable for there was no other purpose for which it could be used in commerce. Yet it would be contrary to commonsense and business honesty if Hill, who had done precisely that which Udall had reasonably relied upon him to do, were liable for a defect in the goods in a respect in which no reliance had been placed upon Hill.
I turn next to the second appeal in which Hill seeks to recover from the Norwegians, as damages for breach of their contract of sale with him, the amount which he is liable to pay to Udall.
This contract was in writing. It was for Norwegian herring meal, a commodity in which the Norwegians dealt. Commercially it is classified as a fishmeal, but since characteristics of fish meals differ it is generally sold *510 under the name herring meal with or without a reference to the country of origin. It was common knowledge in the trade that its normal use in the United Kingdom was as a protein rich constituent in the diet of a variety of domestic animals and poultry, including pheasants. But prior to its inclusion in "King Size" mink food it had not been used in the United Kingdom as a feeding-stuff for mink. At the date of Hill's contract with the Norwegians no one but Hill and Udall knew that herring meal, as distinct from fishmeal, was ever fed to mink in this country. The goods were ordered through the Norwegians' agent in London, Bowring. He knew that Hill was a compounder of animal feeding-stuffs and that the Norwegian herring meal was required by Hill for the purpose of inclusion in compound feeding-stuffs for domestic animals and poultry, including pheasants. He did not know, and had no reason to suspect, that Hill intended to include the meal as an ingredient of a feeding-stuff for mink. If he had known he would have inquired of his principals, the Norwegians, whether Norwegian herring meal was suitable for that purpose.
Milmo J. found as a fact that the Norwegians knew that by the date of the contract herring meal was being fed to mink by Norwegian farmers. The Court of Appeal considered that this finding was contrary to the weight of the evidence. I do not, however, think that an appellate court would be justified in setting aside a finding of primary fact which was based in part at least upon Milmo J.'s view of the reliability of the oral evidence given at the hearing. On the other hand, it is clear upon the evidence that it was known to the Norwegians that in order to be suitable as an ingredient in diet of mink in Norway herring meal needed to possess special characteristics which were not present in most herring meal produced in Norway nor needed in order to make it suitable as an ingredient in the diet of domestic animals and poultry. One such characteristic was a low fat content of less than 6 per cent., whereas the normal fat content of herring meal sold by the Norwegians, at any rate in the winter season, was about 12 per cent. The herring meal ordered by Hill was stated to have a fat content of this order.
My Lords, before coming to the specific provisions of the Sale of Goods Act 1893, I again prefer to start by posing the question: What, if any, responsibility as to the characteristics of the goods to be supplied under the contract would the seller reasonably understand that the buyer believed that he, the seller, was accepting? and in the circumstances of the present appeal that question can be reduced to this: Would the Norwegians' reasonable understanding and Hill's reasonable belief be that the Norwegians were accepting responsibility to deliver Norwegian herring meal which possessed characteristics which made it suitable for inclusion as an ingredient in the diet of mink, as distinguished from other animals?
So stated, as it seems to me, the question answers itself. Hill never let the Norwegians know that he required the herring meal for feeding to mink. They had no reason to suspect that it was required for a use to which it had never previously been put in the country in which it was to be consumed. The fat content provided for in the contract made it unsuitable for the only use of herring meal as a food for mink of which they had any knowledge, namely, in Norway. They were given no opportunity to exercise any skill or judgment in selecting herring meal which would have the special *511 characteristics which would render it suitable as a food for mink as distinguished from other animals.
So I come once more to section 13 of the Sale of Goods Act 1893, upon which Hill in turn founds his claim against the Norwegians.
Since the contract was in writing the description by which the goods were sold must be determined by construing the words used by the parties in the contract. What the contract said about the goods was
"Norwegian herring meal, fair average quality of the season, expected to analyse not less than 70% protein, not more than 12% fat and not more than 4% salt."
I agree with your Lordships that the description by which the goods were sold is limited to the words "Norwegian herring meal." That is what identifies the subject-matter of the contract. Where a contract contains an express statement about the quality of the goods to be supplied the prima facie inference is that this was intended by the parties not as an identification of the kind of goods that are alone the subject-matter of the contract, but as an express stipulation as to the standard of quality to which goods of that kind supplied under the contract shall conform. Such an express stipulation may be intended as a condition or as a warranty. Which it is, depends upon the construction of the contract. This, in turn, depends upon what can be inferred, from the whole of the terms of the contract and the circumstances in which it was negotiated, as to the importance that the parties attached to it. Did they regard exact conformity with the stipulated standard as so vital to the contract that no distinction was to be drawn between a minor divergence or one which was great? Or did they intend to differentiate according to the magnitude of the divergence? If the former, the stipulation is a condition and the consequence of a breach of it would be the same as if the stipulation did form part of the description by which the goods were sold, though it would be classified under section 11 as an express condition and not as an implied condition under section 13. If the latter, the stipulation is an express warranty, but does not exclude reliance upon the implied condition of merchantable quality under section 14 (2) if the divergence is so great as to make the goods not reasonably fit for any purpose for which they are normally used.
In a contract, such as that between Hill and the Norwegians, for the sale of a commodity in which there is an established market and for which there is more than one normal use, a divergence from the stipulated standard of quality which falls short of rendering them useless for every normal use, would be reasonably expected by both parties to affect the value of the goods according to the magnitude of the divergence, and prima facie to amount only to a warranty, a breach of which would not entitle the buyer to reject the goods as being dehors the contract. That such was the intended effect of the stipulation "fair average quality of the season" is, in my view, confirmed by general condition 3: "The goods to be taken with all faults and defects damaged or inferior, if any, at valuation to be arranged mutually or by arbitration." I construe this clause as excluding any right of the buyer to reject the goods for any non- conformity with the stipulation as to quality which falls short of rendering them unmerchantable.
Hill did not allege any express warranty or quality in his pleading. He *512 relied upon the words "fair average quality of the season" as being part of the description by which the goods were sold. I will, however, come back to their effect as an express warranty of quality later.
I turn next to section 14 (1) upon which Hill also relies. The most that Hill made known to the Norwegians about the purposes for which the herring meal was required was what I have previously termed a "range of purposes." The extent of that range was limited to what their agent in London had learnt from Hill in the course of previous dealings as to the nature of Hill's business. The range so made known included use as an ingredient in feeding-stuffs for many kinds of domestic animals and poultry. What it did not include was use as an ingredient in feeding-stuffs for mink. This seems to me to be conclusive that even if the Norwegians knew that Norwegian herring meal was a commodity which might be used as an ingredient in the diet of mink, use for that purpose can neither be nor form any part of the particular purpose for which the goods were required which was made known by the buyer to the seller, so as to give rise to the implied condition under section 14 (1) that they should be reasonably fit for feeding mink.
My Lords, it will already be apparent that, for the reasons which I have already advanced in discussing "range of purposes," the decision of this House in Hardwick Game Farm v. Suffolk Agricultural Poultry Producers Association [1969] 2 A.C. 31 that the fourth parties were liable to the third parties for breach of the condition implied under section 14 (1) in my view goes to the utmost limit of what can be held to be a "particular purpose" within the meaning of that section without amending the Act itself. However desirable it may be to make such an amendment, to do so lies beyond the competence of this House of Parliament acting alone - even in its judicial capacity. I myself would distinguish that part of the decision in the Hardwick Game Farm case whenever I can. I do not think that it is open to your Lordships to extend it.
There are at least two relevant distinctions between the facts in the second appeal and those in the appeal between the third and fourth parties in the Hardwick Game Farm case. The first I have already mentioned. Neither expressly nor by implication had Hill ever made known to the Norwegians that the range of purposes for which he required the herring meal included use as an ingredient in the diet of mink, whereas in the Hardwick Game Farm case the third party had at least by implication made known to the fourth party that his range of purposes did include use as an ingredient in the diet of pheasants, for which it proved to be unsuitable.
The second distinction is that at the time of the relevant contract between the third and fourth parties in the Hardwick Game Farm case it was not known that there could be any characteristics of the commodity sold, Brazilian ground-nut extractions, which might render them unsuitable for feeding to pheasants that would not also render them unsuitable for feeding to all other species of animals and birds. As was pointed out by Lord Reid, at p. 83, the information that the goods were required for feeding to pheasants as well as to other animals and birds would not in fact have enabled the fourth parties to exercise any skill or judgment which they then possessed other than that which was called for in selecting goods fit for the purpose of feeding to all kinds of animals and birds. In the instant appeal it was known *513 that there were some characteristics of Norwegian herring meal, in particular the fat content, which might render it unsuitable for feeding to mink though not unsuitable for feeding to other animals and birds. The information that the herring meal was required for feeding to mink would in fact have enabled the Norwegians to exercise skill and judgment in selecting goods fit for this particular purpose and had they been given this information it is unlikely that, because of its high fat content, they would have selected the particular consignment of herring meal for delivery under the contract. It is true that, at the time, the reason for their not doing so would not have been the presence of DMNA in the consignment, of which they knew nothing, but its fat content which was much higher than was then thought to be suitable in a food for mink. But the fact that to their knowledge there was at least one characteristic which distinguished herring meal which was suitable for feeding to mink from herring meal which was suitable for feeding to some other animals was, in my view, sufficient to displace any inference that Hill, who did not make it known to them that he required the herring meal for feeding to mink, showed that he relied upon their skill and judgment to supply a meal which was suitable for that purpose.
I recognise that, unlike the first, the second distinction is fortuitous. Fat content bears no necessary relationship to DMNA content. But if, as I have suggested earlier, the key to section 14 (1) is reliance, mere knowledge by the seller that the goods may be required for use for feeding to mink is not enough. Unless they know that the goods are required for that purpose and the source of their knowledge is the buyer himself, there is no ground for any reasonable inference that the buyer was relying on the skill or judgment of the seller to select herring meal which is fit for feeding to mink, if there is any characteristic which distinguishes herring meal which is fit for that purpose from herring meal which is fit for feeding to other animals or birds.
Section 14 (2) has not been relied upon by Hill in your Lordships' House. The description by which the goods were bought was "Norwegian herring meal." Goods of that description are normally used for a range of purposes, i.e., as feeding-stuffs for a number of different species of domestic animals and birds. Milmo J. held [1968] 1 Lloyd's Rep. 457, 487 that: "There was no evidence to prove that any part of the shipment was unfit for other animals such as pigs or poultry" and accordingly found that "there was no evidence which would have supported a finding that the herring meal was unmerchantable." This finding Hill does not seek to disturb.
Finally, I turn to the effect of the inclusion in the contract of the express warranty that the herring meal should be "fair average quality of the season. " A term to this effect, often referred to as "f.a.q.," is common in contracts made on the markets in commodities which consist of raw or processed animal or vegetable products. The business purpose of an "f.a.q." term is to enable the buyer and the seller to ascertain what their respective rights are under the contract at the time when the goods are delivered. It fulfils the same function as a physical sample in a contract of sale by sample. It provides a standard of comparison which enables the buyer to ascertain whether it has been departed from, by subjecting the goods to reasonable examination at the time of delivery.
Sales by sample are dealt with in section 15 of the Sale of Goods Act 1893 which is based upon the common law as it had been recently expounded *514 by Lord Macnaghten in James Drummond & Sons v. E. H. Van Ingen & Co. (1887) 12 App.Cas. 284 , 297. The implied conditions under section 15 (2) (a) that "the bulk shall correspond with the sample in quality" and under section 15 (2) (c) that "the goods shall be free from any defect, rendering Íhem unmerchantable, which would not be apparent on reasonable examination of the sample" are, in my view, related to one another. If the sample contains a latent defect not apparent on reasonable examination which would render it unmerchantable but the goods do not, it cannot have been intended that the buyer should be entitled to reject the goods under (a on the ground that the goods did not correspond with the sample in quality. The reconciliation of the two implied conditions is that "quality" in section 15 (2) (a) is restricted to characteristics of the goods which a reasonable examination of the sample ought to have revealed, that is, characteristics which would be apparent upon physical inspection of the sample and the goods and the application to them of such other tests as are customarily applied in the trade.
The same principle ought, in my view, to be applied by analogy to "f.a.q. " terms in contracts for the sale of commodities and to other contracts in which some object that can be subjected to physical examination is adopted as the criterion of quality to which goods are to conform. In my experience, this is how "f.a.q." is understood by dealers in commodities and, as the Court of Appeal pointed out, the same principle has been applied to other contracts which were not strictly sales by sample, namely, by Devlin J. in F. E. Hookway & Co. Ltd. v. Alfred Isaacs & Sons [1954] 1 Lloyd's Rep. 491 by Sellers J. in Steels & Busks Ltd. v. Bleecker Bik & Co. Ltd. [1956] 1 Lloyd's Rep. 228.
I accordingly agree with the Court of Appeal that "fair average quality of the season "relates only to such qualities as are apparent on an ordinary trade examination or analysis of the goods. The presence of DMNA in the herring meal was not such a quality and no breach of the express warranty of quality has been established by Hill.
I may add that, if a breach had been established, general condition 3 would not, in my view, exclude the right to recover the consequential damage (in section 54 of the Sale of Goods Act 1893 called "special damages") claimed in the second appeal. Under section 55 of the Act that right could only be negatived or varied by express words. There are no such words in general condition 3.
For my part I have reached the conclusion that it is not open to this House to allow the second appeal upon the ground which commends itself to the majority of your Lordships or upon any alternative ground.
I would accordingly dismiss the second appeal.

Representation

Solicitors: Sharpe, Pritchard & Co.; Metson, Cross & Co.; William A. Crump & Son .

Appeals allowed. (F. C. )

(c) Incorporated Council of Law Reporting For England & Wales
[1972] A.C. 441

END OF DOCUMENT