168 



THE INDIA RUBBER WORLD 



[February I, 1911. 



Specifications for the Purchase of Materials— II. 



/)'.V l-'icdeiic Vatnicrth, Ph.D.* 



THE manufacturers' criticism of a Specification, before the 

 same is issued, is to-day regarded as a most important 

 part of the proceedings. The old idea that the consumer 

 was to dictate to the manufacturer what the former wanted has 

 given way to the broader view — namely, that the specification 

 should contain or should represent the best that is known en th« 

 subject, no matter where the inforniaiion has been obtained. Tc 

 leave out the other party to the contract is unwise, unfair, short 

 sighted and foolish. The specification should be so drawn that 

 the manufacturer can actually see that he has had a part in 

 its making. 



After criticisms from all those qualified to criticize have beer. 

 received, the provisional specification is remodeled so as to in 

 elude all suggestions which may be of actual value. In those 

 cases where criticisms conflict, it has been observed tliat this is 

 frequently due to local conditions. There is one pre-eminent 

 thought that appears worthy of careful consideration in the wholt 

 mode of procedure — namely, the greater the care, the larger the 

 amount of study and the more well directed time and effort thai 

 are put upon the specification before it is issued, the less will 

 probably be the difficulty connected with it after it has once be 

 come a part of the contract. It might, in fact, be said that the 

 most valuable specification represents the fruition of the studies 

 of those who make investigation into the properties of useful 

 materials, and of those who use them. 



VALUE OF CHEMICAL DATA. 



When specifications for steel were first issued there wa.' 

 opposition on the part of manufacturers to the insertion of 

 chemical data. They claimed that the consumer should only 

 specify the physical properties of the metal and exclude or limit 

 by chemical data only the objectionable constituents, leaving the 

 steel maker free to vary those constituents (upon which the mos( 

 valuable properties of the steel depend) according to his own 

 ideas. Now it is clear that a certain set of physical propertie? 

 produced by high carbon and low manganese in steel may yield 

 a steel more valuable to the consumer than approximately the 

 same physical properties produced by lower carbon and higher 

 manganese, or the interchange of the constituents commonly 

 affecting the physical properties of steel. To repeat what has 

 been previously said : A good specification is the result of the 

 joint effort (1) of those who know steel from its behavior while 

 it is being manufactured, and (2) of those who know steel from 

 its behavior while in service. 



SIZE OF SAMPLE FOR TESTING. 



It has been found difficult to make a hard and fast rule as tc 

 the weight or volume of material which should be represented 

 by one sample. Where the material is made in batches, and the 

 sample is selected at random, the sample obviously represents 

 that batch. In the case of oils and paints the question of 

 sampling is more difficult. Such shipments are made up of 

 material resulting from a number of like operations, without 

 any certainty as to uniformity in the output of each complete 

 operation. In such cases the sampling must needs be rather 

 arbitrary, but if there should be any indication of lack of uni- 

 formity in the shipment, an amendment to the specification is 

 probably called for. 



NUMBER or SAMPLES TAKEN. 



Let us now consider the question : How many individual parts 

 shall make up the average sample? If the shipment consists of 

 20.000 pounds of soap, shall we take one pound for examination 

 or shall we take several pounds from different parts of the ship- 



''roi:snIting Clicmical Engineer, PIiilarle)i)hi,n. 



ment? If 50 barrels of linseed oil have been received, shall we 

 sample one barrel or every barrel? Some consumers contend 

 that the shipment is presumed to be uniform, and that the number 

 of samples should therefore be limited, small unimportant varia- 

 tions being allowed for in the specifications. 



RE-TESTS. 



When n shipment is received, sampled in the prescribed way, 

 tested and found wanting — what then? The producer frequently 

 asks for another test, in the hope that this may shew more 

 favorable results and allow the material to slip in. If the 

 second test is favorable, the consumer is naturally anxious to 

 make a third or "decision" test. Specifications are not drawn 

 for the purpose of making it easy ' for irregular and possibly 

 carelessly made material to be accepted. It is far better to make 

 the limits of the specification wide enough, when they are first 

 drawn to cover all the uncertainties in manufacture and 

 eliminate carelessness, bad judgment, or any attempt to sell an 

 inferior product at the price of a good one. 



BRANDING REJECTED MATERIAL. 



Another difficult point is covered by the question : "Is it 

 possible to so draw a specification that material which has once 

 been rejected will not be offered a second time?" The Govern- 

 ment has adopted the system of placing marks on rejected ship- 

 ments, a practice which lowers the value of the so rejected 

 material and makes it necessary for the manufacturer to raise the 

 price in order to cover himself. But for the larger number of 

 products which are covered by specifications (oils and soaps) 

 identification marks are unavailable. One way in which to 

 obviate any diflSculty is to insert in the contract a clause stating 

 that the manufacturer must pay return freight on rejected 

 material. One purchasing agent settles doubtful cases by refus- 

 ing to accept future bids from firms who make it their practice to 

 return rejected goods. 



HOW SPECIFIC.\TIONS AFFECT THE BUSINESS OF THE PRODUCERS. 



Many manufacturers object to specifications on the ground that 

 they are annoying and really serve no good purpose; others have 

 distinctly requested them. Some consider a difficult specification 

 a direct advantage as it eliminates the competition of inferior 

 products. It must ever be borne in mind that low prices must 

 be the result of some unusual manufacturing facilities or, as is 

 more frequently the case, the indicator of poor quality. A good 

 specification is the best protection that can be offered the 

 manufacturer of honest goods. 



HOW SPECIFICATIONS AFFECT PRICES. 



In many cases the consumer is afraid to adopt specifications, 

 thinking that the price of the product will be raised by the pro- 

 ducer. Experience has in fact shown that after the producers 

 have become accustomed to the new specifications, their prices 

 invariably drop to a greater or less extent. This observation has 

 been variously explained: (1) as being due to the fact that all 

 bidders are bidding on the same quality of goods ; (2) the ma- 

 terial defined in a specification represents what might be called 

 "standard material" so that manufacturers can, \yithout great 

 risk, fill in idle time with its manufacture; (3) as the material 

 is "standard" the manufacturer can without fear of loss purchase 

 the ciiiistituents in a favorable market. 



RULES GOVERNING SPECIFICATIONS. 



In addition to the points previously emphasized, there are 

 given below certain rules (by Dudley) which may be followed 

 to advantage by those who draft specifications. 



