336 PROCEEDINGS OP THE AMERICAN ACADEMY 



I am indebted to the courtesy of Professor Hilgard for the oppor- 

 tunity of comparing R.^ with "Bronze 11" upon Comparator No. 1, 

 which was sent to Wasliington for this jDurpose. Since subsequent 

 comparisons were to be made with this comj^arator, it seemed impor- 

 tant that the relations between these standards should be determined 

 under the same conditions as those under which future investigations 

 would be made. 



For this work I was assigned to a room in the basement of the 

 building, in which a pretty steady high temperature could be main- 

 tained. Afterwards the comparator was removed to a small observa- 

 tory building in the rear, which was admirably adapted for the [tur- 

 pose. Here a nearly constant low temperature was maintained for 

 three days. The comparator was then mounted again in its former 

 location, and further observations were made at a nearly constant 

 temperature, which had now become reduced to about 62°. 



The following are the results of the comparisons. 



Mean....— .000027 

 We have, therefore, 



" Bronze 11 " — R.^- — —0.000027 inch. 

 " Bronze 11 " — F = —0.000088 '' 



Finally, 



R.^' — Y = — O.OOOOGl " 



„ a . 0.000036 in. + 0.000061 in. ^ 

 ^2 ' H o = ^• 



Or, R.^'- -f 0.000048 in. = Y. 



Since the metal in each bar has the same composition, it is assumed 

 that they have the same coefficient of expansion. 



Comparisons were also made between the Tresca meter and tlie 

 Coast Survey meter "No. 49," which bears a known relation to tl.e 

 Berlin meter, and thence with the Metre des Arcl'.ives. 



