On the Acceleration of Gravity for Tokio, Japan. 293 



earth's surface can be calculated by a formula developed by 

 Clairault, who from pendulum-experiments made at a variety 

 of latitudes on the earth's surface has shown that, approxi- 

 mately, for any latitude X and any height h centimetres above 

 the level of the sea, 



#=980-6056-2-5028 cos X-O000003 L 



As the latitude of the Imperial College of Engineering, 

 Tokio, is about 35° 39', Clairault's formula gives for g the 

 value 979*7 centimetres per second per second. But the value 

 obtained by the series of experiments made with the Kater's 

 pendulums differed too much from this and from one another 

 to allow of our trusting them. 



We therefore decided on employing a totally different me- 

 thod and a much less laborious one : — A brass ball 2352*2 

 grammes in weight was suspended by a long steel wire 0*45 

 millimetre in thickness, and in the earlier experiments 978*7 

 centimetres in length. The wire was supported from a steel 

 knife-edge resting on a brass plate. Both the brass ball and 

 the bob of the seconds-pendulum of the standard clock were 

 fitted with fine pieces of platinum wire, either of which dipped 

 into a small cup of mercury when the pendulum to which it 

 was attached was vertical. The mercury-cups &c. were then 

 joined up with a battery and resistance-coils to a quick-run- 

 ning Morse instrument, as seen in the figure, in which M is 



