1894.] on the Astronomical Telescope. 309 



Tlie lecturer here exhibited in action an equatorial clock controlled 

 from an independent pendulum as above described, with an arrangement 

 of bells added by which the audience were enabled to judge of the perfect 

 synchronism of the controlled clock and the controlling pendulum. The 

 lecturer purposely introduced errors into the clock train to illustrate the 

 power of the controlling apparatus to erase these errors. 



Suppose now we have our clock as perfect as is possible, it is 

 further necessary to see that that perfectly uniform motion is trans- 

 mitted to the instrument ; in other words, that any gearing between 

 the controlled clock and the polar axis be as far as possible without 

 error. This gearing consists mainly of the endless screw, called 

 the right ascension screw, and the toothed sector. The precautions 

 taken for the ensuring of this perfection have been elsewhere de- 

 scribed, and are of too technical a character to deal with here, but one 

 observation only I would desire to make. 



In a recent paper by Professor Pickering, commenting on instru- 

 ments he saw during a recent visit over here, he is kind enough to 

 make complimentary allusions to some of these arrangements, but he 

 takes exception to the use by us in this country of long radius sectors 

 for driving the polar axis instead of entire circles. Perhaps it may 

 be well, therefore, if I take this opportunity of saying why we 

 prefer the sector. 



Bear in mind that the greatest possible perfection of clock driving 

 is what we are aiming at, and you will easily see our reason for adopt- 

 ing the sector. When a sector or portion 'of a circle only is used it 

 is possible to get a radius much greater than in the case of an entire 

 circle. 



No mechanism ever made is absolutely free from error. Call the 

 residual error of this screw anything you like, one 10-thousandth or 

 one 20-thousandth of an inch ; whatsoever that error be its effect on 

 the accuracy of the driving of the telescope will be exactly in the 

 inverse ratio of the radius at which it acts ; therefore any given error 

 will only have one-third the effect on the driving of the telescope if 

 working (as it may in a sector) at three times the radius. One 

 10-thousandth of an inch at say 10-inch radius will produce an 

 angular error of about 2 seconds of arc ; at 30 inches radius it will only 

 produce two-thirds of a second error. This may seem a small advan- 

 tage, but the nearer we approach to perfection the more difficult it is 

 to obtain any given increment. 



It does not take much coal to increase the speed of a locomotive 

 from 10 to 11 miles an hour, but it is very different if we want to 

 increase it the same 10 per cent, from say 60 to 66. 



Another fact that has been brought to light by the experiments 

 of the last few years is that atmospheric disturbance, the bete noire 

 of the astronomer, has not so much effect on most of these photo- 

 graphic results as in the case of visual observations. 



In the number of the 'Observatory' published in December 1889, 

 Dr. Gill makes the following remarks in a note accompanying a 



