38 Captaifi Kater's account of the 



the steel plates to which the stirrups are attached coincide 

 with the ends of the knife edges, otherwise some error 

 might arise from a possible want of parallelism. 



The bushel measure being placed in one of the scales, 

 together with 80 lbs. * the whole was counterpoised with 

 brass ; the 80 lbs. was then removed, and the bushel filled 

 with water by means of a syphon. The temperature of the 

 water was then very carefully taken, and the height of the 

 barometer registered. A circular piece of plate glass, as 

 plane as could be procured, having a hole about a quarter of 

 an inch diameter in its centre, was slid over the bushel, and 

 the air-bubbles which appeared on the under surface of the 

 glass were removed by touching them with a^ small bent 

 tube made of flexible metal, a method thought of by 

 Mr. Bate. Water was added through the hole in the glass ; 

 and by carefully observing the curvature of the fluid surface, 

 it was soon evident that the filling of the bushel might be 

 repeated without a greater difference than three or four 

 grains. 



It is scarcely requisite to remark that many experiments 

 were made, not here detailed, before the adjustment of the 

 bushel was perfected ; the very great accuracy and masterly 

 workmanship however of Mr. Keir, who had been furnished 

 with gauges for the diameter and depth of the bushel, ren- 

 dered little alteration necessary. 



* This was obtained by taking twice the weight of the four weights of the 

 gallon of water. 



