[ 476 ] 

 LXV. Intelligence and Miscellaneous Articles. 



NOTE ON THE USE OP THE DIFFERENTIAL GALVANOMETER IN 

 EXPERIMENTS ON RADIANT HEAT. BY M. P. DESAINS. 



nnHE sources of heat ordinarily used in investigating thermal ra- 

 -*- diation are never perfectly constant. Hence special precautions 

 must be taken to exhibit with certainty the slight actions exerted on 

 rays from these sources by causes not acting with any great energy. 

 The difficulty is usually resolved by repeating a great number of 

 times pairs of crossed observations — that is, those made alternately 

 on the direct and on the modified pencil. This mode has the dis- 

 advantage of greatly prolonging the duration of the operations. In 

 order to avoid this, which is frequently objectionable, De la Pro- 

 vostaye and I have used a totally different method, and one which 

 gives the means of directly calculating the influence of variations of 

 the source. But this method (Comptes Eendus, vol. xxxviii. p. 440) 

 presupposes simultaneous observations, and hence the cooperation of 

 two observers ; it is, moreover, ill adapted to lecture-experiments. 

 On the other hand, I have found that great certainty can be ra- 

 pidly obtained in the most delicate thermoscopic observations, or 

 manifestations, by using a kind of differential apparatus consisting 

 essentially of a single source of heat, of two thermopiles, of a differ- 

 ential galvanometer, and, lastly, of a rheostat. 



Each pile is connected with one of the two wires of the galvano- 

 meter, and the rheostat is interposed in one of the circuits. By 

 this apparatus the actions of the two currents on the needle can be 

 equalized, and thus kept at zero of the scale if they are in opposite 

 directions. 



Once obtained, equilibrium continues in spite of variations in the 

 source ; but if from any cause whatever the intensity of one of the 

 radiations alters, the needle quits zero and sets in some other posi- 

 tion. Both piles must be of the same construction if they are to be 

 equally rapid in following the action of the heat. The difference in. 

 dicated is finally the greater the more energetic are each of the cur- 

 rents which traverse the apparatus. 



The first phenomenon I have investigated by the method I have 

 described is that of the absorption of heat by transparent gases. To 

 demonstrate in an easy and certain manner the very interesting re- 

 sults obtained by Magnus and Tyndall, I worked in the following 

 manner : — 



The source of heat is a lamp. The piles are at a distance of about 

 1*8 metre from the flame, to the centre of which their axes point, 

 making with each other a considerable angle. Between the flame 

 and each pile is placed a glass tube, about a metre in length by a 

 decimetre in diameter, which can either be exhausted, or in which 

 the air can be compressed to two or three atmospheres. Each of 

 these tubes is closed at the two ends by thick transparent glass 

 plates ; its axis is in the prolongation of that of the corresponding pile. 



The apparatus being thus arranged, one of the tubes is exhausted, 



