386 PROCEEDINGS OP THE AMERICAN ACADEMY 



For the mean Q^ — 40.3// = Aq 



Collecting results, we have : 



AtO° 



w+mi/i = Ao 



Qi - 8.6 = ^0 



Q2+I7.8 =Aa 



Q,-U.7 =Aa 



Q^— 40.3 =Aq Q4— 318.0 =Aq 16.71 



It will be seen that every value of the coeflicient is less than the 

 value communicated by the Society, viz. 19.155//. It seems im- 

 possible to resist the evidence given by this determination, that this 

 value is certainly too large. It is possible that the solution of the 

 discrepancy may be found in the fact shown b}^ these observations, that 

 when the graduations are upon silver inlaid in brass, the coefficient of 

 the brass controls that of the silver somewhat in proportion to the 

 relative masses of the two metals. In every one of the five cathetome- 

 ters stfindardized by the writer, the brass has carried the silver with 

 it, that is, the coeificient has been found to be that of brass, and not 

 that of silver. 



It will be seen that the cathctometer has a less coefficient than the 

 bronze bar i?^, which is identical in composition with the Imperial 

 Yard. 



It is probable, also, that the resultant coefficient of two metals which 

 have a mechanical junction depends somewhat upon the perfection 

 with which the junction is made. The writer lia.=! not found so large 

 differences as here given in the values of the coefficient for different 

 specimens of the same metal, but in this case the diflerences may be 



