1880.] 



J. F. Tennant — On Standard Weights. 



63 



always been considered necessary to have the primary unit very indestructi- 

 ble, and no doubt this is a very important point : the lead was taken in 

 France, where the Normal Kilogramme was made of platinum ; platinum was 

 again used in England for the Standard Pound, and now standards of re- 

 ference are made of a Platinum-iridium alloy. The cost of the mere metal 

 is very heavy (a kilogramme is at present worth £60 for mere material), 

 and the use of such a metal for large weights is of course out of the ques- 

 tion. It seems to me doubtful whether eqi;al accuracy could not be 

 obtained by employing a large weight of gilt or nickelized bronze ; from 

 which copies could be made with far greater accuracy than they could be 

 separately deduced from the small primary. It is possibly too late to 

 change the material of Primary Standards now, but at all events the 

 standard of Commercial Weight should be a large mass of gilt bronze. 



Acting on these principles, I have neai-ly made a set of weights from 

 1000 tolahs to 0-OOL tolah from these bullion weights. There will be 

 several copies of the largest, carefully compared, some of which I trust 

 Government will allow me to distribute. The individual weights are on 

 what I have called the English Grain system : that is, there are — 

 1000 tolahs. 100 tolahs. 10 tolahs. 1- tolahs. 0-10 tolahs. O'OIO tolahs. 



600 



)j 



GO 



J? 



6 



„ 0-6 



)> 



006 



„ 0-006 „ 



300 





30 



); 



3 



„ 0-3 



)> 



003 



„ 003 „ 



200 



!) 



20 



)> 



2 



„ 0-2 



)! 



002 



„ 0-002 „ 



100 



>J 



10 





1 



„ 01 



;> 



001 



„ 0-001 „ 



The final adjustments and deductions have yet to be made ; but after what 

 I have said, there will be little new in this. I have been very greatly 

 assisted by Mr. Durham, Senior Assistant in the Assay Office, who has 

 superintended all of the gilding ; and to whom I owe devices which will 

 allow the gilt weights to be made true almost to the accuracy of a single 

 comparison by substitution. 



Table I. 



Logarithms for calculating tlie Weight of the Air adapted to Fahrenheif s 



Thermometer. 



This Table gives 10 + the logarithm of the ratio which the weight of 

 air at the temperature named and at Calcutta bears to that of the same 

 volume of water when at its maximum density, the logarithm of the height 

 of the barometer. 



If B be the reading of the barometer reduced to freezing point ; the 

 temperature and V the elasticity of the vapour in the air 



then log sq. of air = At + log (B — 0-238 V). 



