September 9, 1898.] 



SCIENCE. 



327 



Schuster and Gannon exceed Eowland's at 

 19. °1 on the same scale by about one pai't 

 in 500 (Johns Hopkins University Circulars, 

 June, 1898). These differences exist after 

 the final elaborate comparison of thermome- 

 ters and the reductions to the same abso- 

 lute scale of temperature. 



Since the electrical methods employed to 

 determine the mechanical equivalent of heat 

 involve either the current and the E. M. F. 

 of the Clark cell or the square of this E. M. 

 F., and since the E. M. F. of the Clark cell 

 is determined by means of the silver vol- 

 tameter it is evident that the current enters 

 the final result as the square. If the dis- 

 crepancy is due entirely to an error in the 

 value of the ampere, assuming the ohm to 

 be correct, then the ampere should be one 

 part in 1,000 to one part in 800 larger than 

 the present accepted value. That is, the 

 electro-chemical equivalent of silver should 

 be increased from Lord Eayleigh's value of 

 0.001118 to 0.0011191 or 0.0011194. Lord 

 Rayleigh does not claim for his result an 

 accuracy greater than one part in 1,000. 



The method used by Patterson and Guthe 

 was that of a specially constructed electro- 

 djmamometer of large dimensions, and the 

 employment of the torque of a phosphor- 

 bronze wire to equilibrate the countertorque 

 due to the effort between the magnetic 

 fields of the stationary and movable coils. 

 This method eliminates entirely the value 

 of gravity g. The torque of the wire was 

 measured by observing the period of vibra- 

 tion of a cylindrical brass weight of known 

 mass and dimensions when suspended by 

 the phosphorbronze wire. The entire success 

 of this part of the investigation was due to 

 the fact that the observations were made 

 with the whole apparatus enclosed in a 

 fairly good vacuum. Under these condi- 

 tions the vibrations could be followed for 

 hours at various intervals ; the logarithmic 

 decrement was almost entirely constant, 

 and it was easy to obtain a curve connect- 



ing temperatures and periods of vibration 

 as a torsional pendulum. The wire was so 

 connected to the support and to the brass 

 cylinder that it could be transferred from 

 the vacuum apparatus to the electrodyna- 

 mometer and back again without discon- 

 necting it from the terminal pins. From 

 personal inspection at the several stages of 

 the investigation assurance can be given 

 that the work has been most carefully exe- 

 cuted at every point, and all known sources 

 of error have been as completely eliminated 

 as possible. The weights employed were 

 compared with the standards at the U. S. 

 Bureau of Weights and Measures in Wash- 

 ington ; the standard of length was a half- 

 meter bar of speculum metal made for the 

 University of Michigan by the late Professor 

 AV. A. Eogers. The time was taken from a 

 standard Eiefler clock checked by compari- 

 son with the observatory time. The result 

 of the investigation is that the electro- chem- 

 ical equivalent of a used solution of neutral 

 silver nitrate, fifteen parts by weight of the 

 silver salt to eighty-five parts of distilled 

 water, is 0.0011192 gm. per ampere per sec. 

 This exceeds Lord Eayleigh's value by 

 by about ^ of one per cent, and causes the 

 discrepancy in the mechanical equivalent 

 of heat to disappear. 



The corresponding change in the E. M. F. 

 of the Clark cell will be from 1.4342 to 

 1.4327 at 15° C. A direct determination 

 has not yet been made and this redeter- 

 mination is reserved for the coming year. 



Dr. Kahle has obtained for the electro- 

 chemical equivalent of silver the value 

 0.0011182 ( Wied. Annal., Vol. 59, p. 532) 

 by the use of an electro-dynamometer de- 

 signed by von Helmholtz and a fresh solu- 

 tion of the salt. Pellat and Potier found 

 the same value as that of Patterson and 

 Guthe (Journ. de Phys. 9, p. 381, 1890). 

 Henry S. Carhaet, 



Secretary of Committee. 



Ann Aeboe, Mich., July 25, 1898. 



