OF ARTS AND SCIENCES. 295 



the yard with the Imperial Yard. This comparison was made with 

 great care by Mr. Chaney, and, at his request, by the writer also. It 

 was then sent to Breteuil for comparison of the meter with the Inter- 

 national Meter. The bar was received from Paris, February 4, 1883. 

 According to the report of Dr. Fernet, this meter is 310 /a shorter 

 than the Metre des Archives at 0° C. Inasmuch as the comparisons 

 with the prototype were made near 1°, 7°, and 12° Centigrade, with the 

 greatest exactness, the data are at hand for the determination of the 

 length of this metre at 16°.67 C, or 62° Fahr., but they have not yet 

 been received. It is, however, quite as well that the present investi- 

 gation should have been made independently, and without any knowl- 

 ediie of the results obtained at Breteuil. 



The graduations are traced upon plugs of silver inserted at the 

 bottom of wells sunk to the line of the neutral axis. The defining 

 lines of the meter present sharply defined edges, but one of the 

 defining lines of the yard has a somewhat broken appearance, being 

 broader at some points than at others. Two lines about 3 mm. apart 

 are drawn at right angles to the defining lines. It appears, from 

 repeated examinations, that the defining lines, both for the meter and 

 the yard, are sensibly parallel. The comparisons made by the writer 

 have been made at points about midway between the horizontal lines. 



V. The Glass Yards and Meters designated G^ and G^. 



These bars were made by Chance and Sons in 1870, for the Stan- 

 dards Department, London. G^j was presented to the writer by Mr. 

 Chaney, under the authority of the Board of Trade, in 1880. G^ is 

 the property of the Standards Department. It was intrusted to the 

 writer for the purpose of graduation. 



If one may judge from the character of the lines ruled with the 

 diamond, these bars are far from homogeneous throughout their whole 

 extent. With the same pressure upon the diamond, some of the lines 

 are slightly broader than others, and while nearly all of the lines are 

 of good quality, a few of them present a decided granular structure. 



In the transfer of the graduations from one standard to another 

 having a different composition, I have found it advisable to first make 

 a provisional transfer, with an assumed relative coefficient of expan- 

 sion between the two bars, and with an assumed series of corrections 

 to the prototype on account of the horizontal curvature of the ways. 



The bars are then placed upon the universal comparator, and the 

 relations between the prototype and the standard constructed from it 



