nezv standards of weights and measures, 43 



lutely free from error. But a national standard should be 

 accurately that which it professes to be. It is not enough to 

 determine its error, as the record of this may in process of 

 time be lost ; it therefore became necessary to devise a 

 method by which any perceptible error in those standards 

 which are the foundation of all the others, might ultimately 

 be annihilated. 



The four standard yards which I am about to describe are 

 of brass, one inch and a quarter wide, and half an inch thick. 

 This thickness is the same as that of Sir G. Shuckburgh's 

 scale, and was chosen in order that both might be affected 

 with equal readiness by any change of temperature ; for as 

 the imperial standard yard of 1760 is one inch square, I 

 thought it preferable to adjust the new standards by means 

 of Sir G. Shuckburgh's scale, which, as I have before re- 

 marked, does not sensibly differ from it. 



A disk of gold being let into the surface near one extre- 

 mity, a hole was drilled through the bar at the distance of 

 thirty-six inches from the centre of the disk, and being made 

 slightly conical, a plug of brass was ground in the hole so 

 as to fit it perfectly. A gold disk was let into the top of the 

 plug, and reduced to a level with the surface of the scale. 

 The other end of the plug projected beneath the scale, and 

 had a small hole through it to admit a wire, by means of 

 which it might be turned round. A very fine deep dot was 

 then made by Mr. Dollond upon each of the gold disks, as 

 nearly as it could be done at the distance of thirty-six inches 

 from each other, the dot upon the moveable disk not being 

 exactly in its centre. 



