Qi5 



NATURE 



[April 4, 19 18 



silver ranks high amongst the mineral productions of 

 Peru, the increase is less marked than it would other- 

 wise have ibeen. The leading products are copper, 

 34,727 metric tons; petroleum, 363,162 metric tons; 

 silver, 294,425 kilos.; vanadium ore, 3145 metric tons; 

 gold, 1690 kilos ; coal, 290,743 metric tons. These are 

 the only minerals the annual value of which exceeds 

 20o,oooi. ; all the others are far less important. The 

 production of copper, already very important, appears 

 to be likely to increase still further. It is also note- 

 worthy that of the total export of copper no less 

 than 93-85 per cent, was in the form of bars, so that 

 practically the whole of the copper ores produced in 

 Peru are' now smelted in that country. This effect is 

 largely due to the heavy rise in freights; before the 

 war these were about 30s. to -zl. per ton, whereas in 

 1915 they rose to 5Z. to 6Z. per ton without taking the 

 increased cost of insurance into account, so that for 

 any ore or matte containing under 40 per cent, of 

 copper the rise in freights would outweigh a rise of 

 loL in the price of the metal. This effect would be 

 even more marked in the case of ores of a cheaper 

 metal like lead, so that nowadays Peru exports few 

 ores except those of such metals as vanadium, tung- 

 sten, molybdenum, etc., which, on account of their 

 considerable intrinsic value, are proportionately less 

 affected by a rise in freights. It is worth noting that 

 the production of coal has only increased from 283,860 

 tons in 1914 to 290,743 tons in 1915, whilst the imports 

 have fallen from 139,312 tons to 55,662 tons, in spite 

 of the increased development of the metallurgical in- 

 dustry, as just pointed out, the reason being that the 

 use of petroleum to replace coal as a fuel is on the 

 increase, the output of oil having risen 437 per cent, 

 above that in 19 14. 



NATIONAL LABORATORIES AND 

 INDUSTRIAL DEVELOPMENT.^ 



II. 



A National Proving House and Standardising 

 Laboratory. 



CERTAIN general principles seem to me essential to 

 success, namely : — 



(i) Standardisation and testing must, if they are to 

 be of value, depend upon research, and be closely con- 

 nected with it. 



(2) While there must be the closest union between 

 the testing authority and the trade concerned with the 

 production of the goods to be certified, the authority 

 should not be dependent on the trade for financial sup- 

 port, and while the wishes of the trade as to the 

 standards to be attained must be fully considered, the 

 executive of the testing institution should be an inde- 

 pendent authority. 



Testing must go hand in hand with research. For, 

 in the first place, research is necessary in order to set 

 up the standards required. Take, for example, our 

 standards of length. The yard or the metre is the 

 distance between two marks on certain standard bars 

 very carefully preserved. They are both arbitrary 

 standards, it is true, and it is clearly of the greatest 

 importance that they should be invariable. Do we 

 know that this condition is secured, and, if so, how 

 do we know it? Materials certainly alter their dimen- 

 sions with changing temperatures, and possibly also 

 with time; for standard work we must know the tem- 

 perature at which we make our comparisons, and this 

 need leads at once to the investigation of the methods 

 of measuring temperature and of the amounts by which 



1 Abridged from two lectures delivered at the Royal Institution on 

 February 26 and March 5 by Sir R. T. Glazebrook, C.B., F.R.S. Continued 

 from p. 77. 



NO. 2527, VOL. lOl] 



various materials change in size with changes of tem- 

 perature. A wide field of investigation opens directly ; 

 temperatures are measured by thermometers. How 

 are the various kinds of thermometer connected? Da 

 a mercury thermometer and a gas thermometer give 

 the same results? Is the glass of which an ordinary 

 mercury thermometer is made of importance ? Or, again : 

 To what extent is the length of a yard measure of brass 

 or isteel dependent on the temperature? Can we find 

 a material less sensitive to tem^perature changes than 

 the platinum-iridium alloy of which the standard metre 

 is made ? And so on. The investigations necessary before 

 we can standardise our yard measure have called for 

 much research. But, again, what security have we 

 that even if we keep the standard with the greatest 

 care and make our comparisons under the most favour- 

 able conditions of temperature, its length is invariable? 

 Is the metre the same length now as when it was first 

 deposited at the Bureau des Poids et Mesures at 

 Sevres? To answer this question a research of great 

 difficulty was carried out at Sevres by Michelson when 

 he compared the length of the metre with the wave- 

 length of light under certain specified conditions. There 

 are cogent reasons for supposing that to be an invari- 

 able quantity. 



At the laboratory during the past two years we have 

 tested vast numbers of gauges and the improvement in 

 manufacture has been very marked; this has been 

 reached only by careful investigation into each cause 

 of error by attention to small details, and by research 

 into methods of measurement with a view to their 

 simplification so that they could ibe used in the work- 

 shop, and to improvement in accuracy so. that the 

 results obtained were not vitiated Iby errors in the 

 method of obtaining them. 



A visit to the gauge-testing-room of the National 

 Physical Laboratory will show anyone how closely 

 research and standardisation go together, how hopeless 

 it -would be to try to run a standardising laboratory 

 apart from research. Or, again, to take an example 

 from another department of science. Ohms and volts 

 and amperes are nowadays familiar words ; you 

 measure the one with a Wheatstone bridge, or more 

 probably with an ohmeter, you read off the others in a 

 voltmeter or an ammeter. But the definitions of these 

 quantities are highly technical and scientific. 



Do you realise what research has been required be- 

 fore our present practical system of making electrical 

 measurements was evolved, and how much you owe to 

 that (research? Compare the rate of advance of the 

 electric motor and the steam engine. 



The work of the Engineering Standards Committee 

 has been of untold advantage to the country. At every 

 step of that work the committee has kept in close 

 touch with scientific principles, and researches of the 

 most varied character have been carried out, and are 

 being carried out now, with the view of determining 

 what standards to set up and what tests to prescribe. 



Nor is it enough to say that much of this has been 

 done and need not be carried further; the principles 

 on which amneters and voltmeters are made have 

 been thoroughly investigated, the optical laws with 

 which telescopes and lenses must comply are well 

 known ; lay down your tests and specifications, and 

 train observers, analysts, and tesiters to enforce them, 

 and you have done all. 



Stagnation and death, not life and progress, lie that 

 way. It is not our Object merely to apply with rigid 

 fairness the laws laid down, and to be pleased rather 

 than otherwise, like the mythical examiner, when we 

 "plough" them every one. The standards set must be 

 reasonable, but they must tend to raise the quality of 

 the product tested. Recurring defects must be watched 

 and investigated, and the tests modified to prevent 



