April 4, 19 18] 



NATURE 



97 



em; you must gain the confidence of the manufac- 



ter'and lead "him to realise you are out to help 



m. and that yoU really know something, probaibly 



ore than he does, of the strength and weakness of 



- goods. Nearly all Englishmen are anxious to 



lintain the reputation of their country, and welcome 



ir tests which show up bad work and make for its 



nprovement. Our statistics show the improvement 



ihat is produced by tests properly carried out. 



So much, then, for my first contention, that research 



mI standardisation must go hand in hand; the truth 



the second, that the testing authority should be inde- 



iident of the manufacturer, is, I think, obvious. It 



i? necessary to give confidence. 



A certificate has but little value, even if it states 

 ine truth and nothing but the truth, unless it comes 

 tiam an absolutely impartial source. If I bear wit- 

 n. ss of myself, niy witness is nothing. To the old 

 customers of a well-established firm the assurance of 

 :he firm is sufficient; a stranger looks for some inde- 

 pendent evidence before he accepts as true all the claims 

 ade ibv the man who desires "to sell his latest produc- 

 n as something far superior to all else on the market, 

 [mpartiality is the first attribute of justice, and the 

 ^picion that the judge may be swayed by something 

 -ides the strict merits of the case is fatal. Again, 

 r: is necessary for the good of the manufacturer. False 

 praise is dangerous to the recipient. The man who 

 relies on the verdict of a too friendly critic may easily 

 fail to maintain the high quality of his products and 

 find himself outstripped by one who has fceen spurred 

 to effort by fair and judicious criticism. 



A testing laboratory controlled by a,n association of 

 manufacturers for the advancement of their trade is of 

 much less value, both to them and to the country, than 

 one in which the ultimate decisions rest with an inde- 

 pendent authority. Of course, the standards to be 

 worked to must 'be determined in closest co-operation 

 with the trade. No specification is ever adopted by 

 the Engineering Standards Committee until it has been 

 fullv discussed at meetings at which the trade is fully 

 represented ; in no case is the decision as to whether 

 an article comes up to the standard left to such a meet- 

 ing, and this has had an important bearing on its 

 success. At the laboratory we have advisory com- 

 mittees on various matters. Executive powers rest with 

 the Executive Committee or with the director acting 

 under the instructions of that committee. He signs all 

 the certificates, and is responsible only to the com- 

 mittee, and this seems to me the proper plan. 



[The lecturer then proceeded to describe and illustrate 

 on the screen some of the principal tests now carried 

 out at the laboratory.] 



Having now dealt with the test work at the labora- 

 tory, and the method of procedure, let us turn to 

 the future. Is the work of value? If so, what steps 

 iiave 'been taken to make -it of more value still, to 

 increase its range, and to widen its influence? Are 

 further steps desirable, and, if so, what should they be? 

 Its value is, I think, recognised ; the recent growth 

 in many branches of our work, besides that of testing 

 gauges for engineers, is evidence of this ; the proposals 

 to establish standardising laboratories in various cen- 

 tres of industry point in the same direction. Engineers 

 are coming to recognise more and more the importance 

 of interchangeability, the ndvantage of working to 

 limits, the gain in producing power — 'Combined, I fear, 

 with deadly dullness in much of the work — secured 

 by the standardisation of parts. 



Here, I think, a word of caution is necessary. Local 

 standardising institutions are desirable in certain cases ; 

 local standards are most undesirable. I am not sure 

 how many wire gauges used in the sale and purchase 

 of wire and thin metal sheets there are In a recent 

 NO. 2527, VOL. lOl] 



list I saw enumerated' some six or eight, each with its 

 own tolerances, or in many cases with no tolerances 

 at all; each has been introduced to fill a need, but 

 with no thought for other needs. There is a risk, I 

 fear, that the establishment of local testing laboratories^ 

 unless care is taken to connect them with some central 

 institution responsible for maintaining their standards 

 and co-ordinating their methods, may tend to perpetuate 

 like anomalies. There is already, as many of us know, 

 a standard inch and an "Enfield" inch; we do not 

 want Manchester, Leeds, and Birmingham inches. 



Transit is easy, and the delay involved in sending 

 goods to a central institution need not be great; the 

 uniformity of results secured in this manner is worth 

 much. Where this cannot be done there should be 

 some organisation devised to keep the standards em- 

 ployed in all parts of the country alike within agreed 

 limits, and to maintain this connection with the results 

 of research. 



The increase in the number of clinical thermometers 

 ha^ already been mentioned. Tests on optical instru- 

 ments of all kinds are growing, and steps have been 

 taken to add to the staff and improve the facilities 

 for handling these. 



The quantity of glassware used in chemical labora- 

 tories throughout the country is enormous. In pre- 

 war days this was almost all of German manufacture, 

 and much came into the country with Reichsanstalt 

 certificates. English manufacturers have taken up the 

 quesition, and are now prepared to offer large supplies, 

 and a scheme has been arranged for its standardisation 

 and the issue of certificates. This is the outcome of 

 discussions of a committee on which were representa- 

 tives of the Department, the manufacturers, the users, 

 and the laboratory. The limits of error for the various 

 classes of articles have been provisionally fixed, and a 

 schedule of fees settled which the makers' think reason- 

 able, and it is hoped will in time enable the 

 work to be carried on without loss. For the present a 

 house has been secured at Teddington, and is being 

 equipped, in which the testing can for the time go on 

 — a certain amount of this class of ^york has always 

 been carried out at the laboratory. Additional build- 

 ings are to be erected, and the scheme put on a per- 

 manent footing. 



The quantity of the various articles is very large, 

 and it is not necessary that all should be tested to the 

 same limits of accuracy, nor would it be possible 

 to send them all to the laboratory. This difficulty will 

 be met by having two classes of goods treated differ- 

 ently. For work of the highest accuracy it is necessary 

 that the articles should be sent to Teddington and be 

 tested individually. Those that pass the tests will con- 

 stitute Class A, and receive the laboratory mark. 

 The vast majority will be dealt with at local centres 

 organised by the laboratory and manned, at least so 

 far as the more responsible positions go, by members of 

 the laboratory staff. These centres will, in some cases, 

 be at the large works ; in others it is hoped to interest 

 the local universities or technical colleges. At the head 

 of each will be the N.P.L. inspector, who will be free 

 to visit the works, inspect the methods of manufac- 

 ture, and select for test from each batch such articles 

 as he thinks fit. So long as the methods remain 

 satisfactory and the goods come up to standard, the 

 firm will be licensed to mark the articles in some 

 distinctive way. 



A fee will be charged for each article tested 

 at the laboratory. In the case of the arti- 

 cles inspected or tested in bulk, it is proposed to cover 

 expenses by a royaltv reckoned on the numbers pro- 

 duced, which would be charged for permission to use 

 the trade-mark. 



Such a scheme, it is clear, requires the cordial co- 

 operation of the makers and the inspecting authority. 



