24 



correct, have equalled what the former distance lacked of 

 three feet. The distance was measured, and from the mean of 

 six observations was found to be 9182 divisions of the micro- 

 meter, which Captain Kater says " differs so little from the 

 919 that this standard yard may be considered as perfectly cor- 

 rect." The amount of error was that the standard official yard 

 was "00003424 inch less than the Shuckburg scale, which error 

 was of course not important enough to justify rejecting the 

 measure. It was, however, considered by the Astronomer 

 [Royal that little importance was to be attached to the Shuck- 

 burg scale itself on account of its imperfections and the dis- 

 cordancies of its comparisons, and Mr. Sheepshanks (the 

 designer of the present microscopic comparing instrument 

 described below) did not place much trust in Eater's experi- 

 ments, on account of the objectionable mode of resting the 

 microscopical apparatus upon the scale. The idea of dividing 

 a bar into approximate halves, as in this process, originated 

 with Prof. Airy, the Astronomer Eoyal, who points out that 

 " by placing two such end measures of a yard in a straight line 

 end to end a line measure of a yard is formed, which may be 

 compared with a line standard, and experience proves that line 

 measures can be much more readily and accurately observed 

 than end measures. 



The microscopical comparing apparatus now used by the 

 Standards Department consist of two principal parts — a travel- 

 ling rest or stand for carrying the bars to be compared, and a 

 fixed shelf of slate carrying four microscopes with micrometers, 

 as per drawing (Plate IX.). The whole of the top part of the 

 table, namely, the three plates of gun-metal, marked 1' 1" 1'", and 

 their connections, are enclosed after adjustment in a mahogany 

 box, together with the measures to be compared attached to 

 them, to prevent any variations of temperature during testing. 

 As a further precaution a thick woollen rug is wrapped round 

 this with holes cut in it for observation. 



The micrometers of the microscopes were carefully rated, 

 and it was determined provisionally that one division of the 

 micrometer equalled '00003173 inch — eventually settled at 

 "0000318 inch. This is the instrument used for the verification 

 of line measures, or measures a traits, as they are termed ; but for 

 end measures, or measures a bouts, a specially-devised contact ap- 

 paratus is employed. The instrument is made of brass, as per il- 

 lustration (Plate IX.). By means of small contact pieces of steel, 

 with fine lines engraved on gold studs upon them, the measures 

 are converted into measures a traits. The contact pieces must 

 first be rated by ascertaining the exact distance between the 

 fine lines of the two pieces when in contact, and this constant 

 always deducted from the observed measure gives the actual 



