of the Acceleration of Gravity for Tokio, Japan. 301 



would be to use a rigid homogeneous pendulum, of such a 

 shape that its moment of inertia could be easily calculated with 

 accuracy from its linear dimensions measured at any tempe- 

 rature — for example, a sphere at the end of a cylindrical rod 

 with a knife-edge, all cast in one piece and turned true in the 

 lathe, or a cylindrical bob cast at the end of the cylindrical 

 rod. The only possible objection to this method would be the 

 possible want of homogeneity in the metal. This might be 

 allowed for in the following way : — Instead of casting the cy- 

 lindrical bob in one piece, make it consist of a number of con- 

 centric tightly fitting cylindrical shells accurately turned. 

 Experiments would then be made first with the cylinders all in 

 one position, then with some of them twisted slightly round, 

 and so on until in the mean result the errors of eccentricity 

 of mass would probably be eliminated. 



Another, and perhaps the best of all methods, would be for 

 rigid compound pendulums to be accurately timed experimen- 

 tally at Greenwich at a number of different temperatures, and 

 sold with a scale of temperature-corrections for the time of 

 vibration attached. 



One set of experiments, then, with one of these pendulums 

 anywhere would at once give the value of g; and such a pen- 

 dulum would undoubtedly be the most suitable for surveys 

 and expeditions in foreign countries. The fact is, the mathe- 

 matical beauty of the principle involved in the Kater's pendu- 

 lum has, in our opinion, caused far too much importance to 

 be attached to it as a practical instrument for determining 

 experimentally the value of the acceleration of gravity. 



We have to thank several of our late students, and especi- 

 ally Messrs. Honda, Kikkawa, A. Kasai, J. Nakahara, and 

 H. Nobechi for assistance rendered us during this investi- 

 gation. And it may here be mentioned that this investigation, 

 like the many others we have been enabled to carry out during 

 the last few years, has resulted from the plan we have followed 

 of teaching the laboratory students not, as is customary in 

 Colleges, to repeat well-known experiments, but to endeavour 

 in their investigations to advance, in some small degree at any 

 rate, the bounds of existing knowledge. And this system of 

 enlisting the assistance of even quite young students in ori- 

 ginal research we have found to create an enthusiasm in ex- 

 perimental work otherwise unproducible. 



Phil. Mag. S. 5. Vol. 9. Ho. 56. April 1880. 



