1886.] Thermopile and Galvanometer combined. 



217 



II. "On a Thermopile and Galvanometer combined." By 

 Professor George Forbes, M.A. Communicated by Lord 

 Rayleigh, M.A., D.C.L., Sec. R.S. Received February 4, 

 1886. 



The author has lately made use of a special form of thermopile and 

 galvanometer combined, which is very sensitive for the measurement 

 of radiation. The apparatus is especially suitable for use as a line- 

 thermopile. 



The first experiments were made with two half tubes, one of anti- 

 mony the other of bismuth, soldered together so as to make a short 

 tube about 2 cm. external diameter, the walls being 2 mm. thick, and 

 the length of tube about 2 or 2-J- cm. The sides of the tube where 

 the junctions of the metals occur were then filed flat, so as to present 

 a thin wall to receive the radiations and to enable it to rise in tempe- 

 rature more rapidly, and also more uniformly throughout the thick- 

 ness of this wall. This tube was lamp-blacked. A Thomson's mirror 

 and magnets (by J. White, of Glasgow), in its usual brass cell, but 

 with an insulated coating, was then inserted in the tube, and the 

 whole apparatus mounted inside a brass cube with a brass cone at one 

 side to throw the radiations upon one junction, and with a circular 

 hole facing the mirror. This apparatus when properly adjusted with 

 a lamp and scale was found to be very sensitive. It had been the 

 intention of the author to use a telescope in place of a lamp, but the 

 radiations of the lamp were found to give rise to no inconvenience. 



Let us compare the probable sensitiveness of this instrument, say, 

 with a line-thermopile of the ordinary construction of 20 pairs, 

 forming a line of the same length as the tube, the double length of a 

 pair of antimony and bismuth in the line-thermopile being equal to 

 the circumference of the tube. Let E be the E.M.F. of one junction, 

 and let R be the resistance of one pair in the line- thermopile, and let 

 R' be the resistance of the galvanometer used with the line-thermopile. 

 The total resistance of the line-thermopile is 20R, and that ot the 



R. 



tubular one is ^q, and the currents in the line and tubular thermo- 

 20E 20E 



piles respectively are — — — — and — — . If the galvanometer were 

 20x1 ht It 



specially constructed to match the line-thermopile, R' would be made 

 equal to 20R, and the current would be — , or one-fortieth of the cur- 

 rent in the tubular thermopile and galvanometer combined. Thus it 

 would require forty turns of wire in the low-resistance galvanometer, 

 if these coils occupied the same space as the metal of the tubular 



