1869.] Sir W. Thomson on a New Astronomical Clock. 469 



clock, be held back every time it touches the escapement during the interval 

 required to set it right again from having gone too fast during the preceding 

 interval of motion. But in the ordinary clock the interval of rest is con- 

 siderable, generally greater than the interval of motion. In the new clock 

 it is equal to a small fraction of the interval of motion : 3^ in the clock 

 as now working, but to be reduced probably to something much smaller 

 yet. The simplest appliance to count the turns of this escapement-wheel 

 (a worm, for instance, working upon a wheel with thirty teeth, carrying a 

 hand round, which will correspond to the seconds' hand of the clock) com- 

 pletes the instrument ; for minute and hour-hands are a superfluity in an 

 astronomical clock. 



In various trials which I have made since the year 1865, when this plan 

 of escapement first occurred to me, I have used several different forms, all 

 answering to the preceding description, although differing widely in their 

 geometrical and mechanical characters. In all of them the escapement- 

 wheel is reduced to a single tooth or arm, to diminish as much as possible 

 the moment of inertia of the mass stopped by the pendulum. This arm 

 revolves in the period of the pendulum (two seconds for a one second's 

 pendulum), or some multiple of it. Thus the pendulum may execute one or 

 more complete periods of vibration without being touched by the escapement. 



I look forward to carrying the principle of the governed motion for the 

 escapement-shaft much further than hitherto, and adjusting it to gain only 

 about per cent, on the pendulum ; and then I shall probably arrange 

 that each pallet of the escapement be touched only once a minute (and the 

 counter may be dispensed with) . The only other point of detail which I 

 need mention at present is that the pallets have been, in all my trials, 

 attached to the bottom of the pendulum, projecting below it, in order that 

 satisfactory action with a very small arc of vibration (not more on each side 

 than of the radius, or 1 centimetre for the seconds' pendulum) may be 

 secured. 



My trials were rendered practically abortive from 1865 until a few months 

 ago by the difficulty of obtaining a satisfactory governor for the uniform 

 motion of the escapement-shaft ; this difficulty is quite overcome in the 

 pendulum governor, which I now proceed to describe. 



Imagine a pendulum with single-tooth escapement mounted on a collar 

 loose on the escapement-shaft just as described above — the shaft, however, 

 being vertical in this case. A square-threaded screwis cut on the upper quarter 

 of the length of the shaft, this being the part of it on which the collar works, 

 and a pin fixed to the collar projects inwards to the furrow of the screw, so 

 that, if the collar is turned relatively to the shaft, it will be carried along, as 

 the nut of a screw, but with less friction than an ordinary nut. The main 

 escapement-shaft just described is mounted vertically. The lower screw and 

 long nut collar, three-quarters of the length of the escapement-shaft, are 

 surrounded by a tube which, by wheel-work, is carried round about five per 



