114 Prof. Magnus on the Influence of the 



and that of the source of heat by which it is irradiated. If the 

 latter is constant, the pile takes the less heat the warmer it is 

 itself; but if its temperature alters, the indications are no longer 

 comparable. The use of the method of compensation presup- 

 poses that the changes of temperature which the pile undergoes 

 in the course of time, are small as compared with the actions to 

 be measured. 



The atmospheric air whose radiation was to be measured was 

 pressed into the brass tube by means of a large bellows. If it 

 was to be dry, it passed previously through a spacious vessel 

 filled with pieces of fused chloride of calcium ; and if it was to be 

 saturated with aqueous vapour, through a flask in which was 

 water that could be heated at pleasure. Arrangements were so 

 made by means of stopcocks that the air could be passed at will 

 either through the chloride of calcium or through water. 



In order to be quite certain that in both cases the tempera- 

 ture of the emerging air, if not exactly, was approximately the 

 same, and also in order to ascertain how high was the tempera- 

 ture of the radiating air, a thermometer was so introduced that 

 its bulb was in the middle of the ascending current and in front 

 of the middle of the aperture through which the heat radiated 

 into the box and against the pile. This position was at a height 

 of 60 millims. above the end of the brass tube. It indicated 

 there a temperature of 220° to 230° C. It was then placed at 

 a height of 180 millims. above the aperture, where it still indi- 

 cated 120° to 130° C. During the experiments it was always 

 in this latter position ; in which it could neither radiate against 

 the pile, nor hinder the ascending current. In both positions 

 the temperature was the same whether dry or moist air passed 

 through the tube, provided the pressure was the same under 

 which the gas issued. In the case of the other gases, of which 

 mention will immediately be made, the pressure in the heated 

 tube was so regulated that the thermometer remained at the 

 temperature 120° to 130° at a height of 180 millims. 



When dry atmospheric air was blown through the heated tube, 

 it produced an extremely slight action upon the column. The 

 deflection of the very delicate galvanometer was about 3 millims. 

 or parts of the scale*. If the air had passed through the flask 

 containing water, the deflection remained almost unchanged ; it 

 only increased by three to five divisions of the scale. 



If instead of atmospheric air dry carbonic acid was passed 

 through the ignited metal tube, the deflection of the galvano- 

 meter amounted to 100 to 120 divisions. 



Ordinary coal-gas furnished almost the same deflection. 



* The galvanometer was the same as that described in Poggendorff's 

 Annalen, vol. cxxiv. p. 479. 



