Thermostat for Controlling Temperature of a Room. 211 



If the same kind of argument were used here as has been 

 applied to the two kinds of positive ions, the experiments 

 would have been said to show that there are more negatives 

 than positives. It does not therefore appear to me that the 

 evidence Professor Millikan has brought forward to show 

 that all the positive ions in the gas have single atomic charges, 

 is conclusive. 



Yours truly, 



John S. Townsexd. 



XIX. A Thermostat for Controlling the Temperature of a 

 Room. By The Earl of Berkeley, F.R.S., and C. V. 

 Burton, D.Se* 



THE thermostat here described was designed for regulating 

 the temperature of an interferometer room ; a some- 

 what small cellar, which owing to its situation had to be 

 heated by electricity. The current dealt with was 7 amperes 

 at 100 volts, but the method to be described is capable of 

 handling considerably larger currents, provided they are 

 split up in parallel. 



For our purpose it was essential that the temperature of 

 the room should not vary by more than about half a degree 

 for several days and nights on end. 



The sensitive member was a flat compound strip of brass 

 and invar about a foot longf (we are indebted to Mr. A. 

 Campbell for this suggestion), made by hard-soldering face 

 to face a strip of brass and one of invar, and then rolling the 

 compound strip to thin and harden it. 



In the figure (p. 212), A is the brass-invar strip, whose upper 

 end Ai is fixed. B x and B 2 are platinum-tipped screws, one on 

 each side of a platinum contact-piece rivetted into the lower 

 end of the strip A. These screws are adjusted so that the 

 freedom of motion of the strip is restricted to a small fraction 

 of a millimetre. C L and C 2 are coreless solenoids, adjustably 

 fixed to the board E, which carries the parts already named. 



F is a wooden beam which can turn about a pivot by an 

 amount limited by stops (not shown) ; G 1 and G 2 are short 

 rods of soft iron, suspended from the ends of the beam F, 

 co-axially with the solenoids H x and H 2 . K is a sealed glass 

 tube fixed to the beam ; it contains some mercury and is 

 exhausted to a high vacuum. Two platinum wires are sealed 



* Communicated by the Authors. 

 t Later increased to 18 inches in length. 

 P 2 



