Laboratory at St. Johrfs College, Oxford. 183 



boxes, whose fly-wheels are driven by an arrangement of this 

 kind. In these the teeth of the driving-wheels are cut quite 

 roughly : they are mere slits, and are cut straight, not to the 

 angle of the screw. I first cut the required teeth in the shape 

 of square slots, cut to the angle of the screw, but not rounded 

 or finished in any way. The wheel so cut drove the disk 

 beautifully. I then gave some further cuts to the sharp edges, 

 and rounded them off so as to fit the surface of the screw rather 

 neatly. The wheel then refused to drive the disk at all. It 

 appeared as if the greater surface in contact increased the 

 friction so as to stop the action. Eventually I cut away the 

 greater part of the rounded edges of the teeth again, and the 

 wheel now drives the disk quite well. I believe that straight 

 square slots, in which the surface in contact would be reduced 

 to a minimum, would be the best form of tooth for this purpose. 



The differential-velocity machine, which forms the essential 

 part of the clock governor, has been constructed according to 

 the description given in my original paper. No points calling 

 for mention arose in connexion with it. 



A number of electro-pneumatic levers have been supplied 

 by Messrs. Bryceson, for the purposes indicated in my original 

 paper. Connected with the bellows, having wind of a pressure 

 equivalent to 7 inches of water, each lever lifts easily a 2-lb. 

 weight through nearly an inch, on passing a current through 

 its circuit. A pair of these levers is applied to drive one side 

 of the differential-velocity governor from the clock, while the 

 other side of the governor is driven by the disk-machine. 



When this was all ready, and the electric connexions were 

 tried, it appeared that there was something wrong with the 

 circuits, which consist of two underground cables connecting 

 my rooms with the laboratory. As much with a view to 

 general convenience as for this particular examination, I took 

 the opportunity of forming a sufficient set of resistance-coils 

 and other testing-apparatus. The resistance-coi]s were copied 

 from standards in my possession. One of the cables proved 

 to have been cut across, forming a fault with an earth-con- 

 nexion. The Leclanche battery, which had always given 

 trouble, proved also insufficient for maintaining the requisite 

 power for any length of time. A suitable battery has not yet 

 been obtained ; but that will not present any difficulty. A 

 clock with wooden seconds' pendulum has been prepared for 

 occasional use in the laboratory itself, that we may not be 

 interrupted by any further failure of the circuits. But it is 

 judged best to keep the standard clock in my rooms, where it 

 is unaffected by the variations of temperature that occur in 

 the laboratory. Some preparation has been made for other 



