42 



J. F. Tcniiant — On Standard Weiglits. 



[No. 1, 



the English grain, because, to me, those units were more accessible (as they 

 will be to most readers of the English language) and not because I prefer 

 them. I have thought that it was more important to avoid conversions 

 of the data before using them than to adhere to general considerations ; 

 just as (with the late Warden of the Standards) I have preferred uniformity 

 of data for reduction ; rather than a possible scientific accuracy, which is, 

 after all, not demonstrably gained. 



Section I. — On Weights. 



In May 1879, I received from England a set of Bullion Weights of 

 gilt bronze, with their errors on the Commercial Standard of England 

 roughly given, and a Troy Ounce of Platinum-iridium, with its error in va- 

 cuo in terras of the Parliamentary Standard Pound PS. I at the same 

 time received a set of Metric Weights of Platinum-iridium from 100 grammes 

 to one milligram, with their errors in tei'ms of the Kilogramme des Archives, 

 which is the Normal Standard weight of France. My paper here will be 

 confined to dealing with some of the Bullion Weights : and it will be neces- 

 sary in order to understand the procedure I follow, and also the scientific 

 principles of weighing, that I should give an account of the English 

 system of weights. 



Ordinary weights are made of brass, iron, or some other cheap metal, 

 but all these are liable to oxidation, and thus none of these metals is suitable 

 for a Standard. The metal chosen for the English Standard was platinum, 

 which is nearly indestructible. Since then it has been found that, whereas 

 platinum is soft, an alloy with iridium is hard, has the other advantages 

 of platinum, and can be made with sufBcient readiness for the purpose 

 required : this alloy is used in my Primary Standards as it is in the European 

 Standards now being made in Paris. The use of such substances for 

 Standard Weights, however, leads to some complication : these metals are 

 heavy ; while the metals and alloys ordinarily used are comparatively light. 

 Now the weight of a body in air is different from its weight in vacuo by 

 the weight of the air displaced, and this varies with the state of the atmos- 

 phere : consequently the relative weight of a pound of brass and one 

 of platinum, which are alike in vacuo, will, in air, be found to vary con- 

 tinually relatively to each other. In order to avoid the inconvenience of 

 this, it has been found desirable that the Commercial Standard should be 

 of brass or bronze ; both of which, having nearly the same density as the 

 metals used in ordinary weights, will show the same differences at all times 

 and places, with sufficient accuracy for commercial purposes ; and which, 

 moreover, are cheap enough to allow of the weights of all sizes being made 

 of them. For general Standard purposes, weights are now made of gilt 

 bronze, the gilding preserving them to a great extent from changing by 

 oxidation. 



