346 Proceedings of the Royal Physical Society. 



The recommendation which I claim for this machine is 

 that of cheapness combined with an efficiency which enables 

 it to meet all ordinary demands. I do not mean to suggest 

 that it possesses any advantage over the other forms referred 

 to, other than that it is much cheaper than any with which 

 I am acquainted, and that I have never found it incapable 

 of fulfilling any demand I have had to make on it in the 

 course of an inquiry which has proved, I believe, a fair test 

 of its efficiency. 



The essential characters of a good recording apparatus 

 seem to me to be : — 



1. That it have a steady motion. 



2. That it possess a reasonable variety of speeds. 



3. That it can be readily adaptable for both smoked 



and continuous paper work. 

 These properties will, I believe, be found in the apparatus 1 

 am about to describe. 



1. Apparatus foe Smoked Papee. 

 The motive power employed is a strong clock driven by a 

 spring, and capable of running for 15 minutes without 

 re-winding. It can of course be wound without interfering 

 with the running. This part of the apparatus was originally 

 made from drawings by Mr John Shewan, at one time 

 laboratory assistant to Professor Rutherford. It is shown 

 with the drum attached in PL XII., Fig. 1. The clock has 

 a train of five wheels, and is supported between two plates 

 10 in. by 3 J in. The last wheel of the train is bevelled and 

 drives a horizontal spindle, carrying a regulator of very 

 simple construction, and which seems to me to serve all the 

 purposes of the more complex governors usually applied. 

 It is shown in PL XIL, Fig. 2. The spindle referred to 

 carries a crosspiece of brass having four square eyes, two in 

 each side of the spindle. Into these eyes a short brass wire 

 (square in section) passes on either side. This wire carries 

 at its further end a thin metal plate, 2 in. by 1 in., and 

 moves freely in the eyes. The plates are pulled towards each 

 other by a spiral spring of slender wire, the ends of whicli 

 are hooked into holes piercing the near sides of the plates. 



