58 Sir John Conroy. On the Amount of Luminous and 



1 cm., 5 cm., 10 cm., and 15 cm., long; each of the washers was about 

 1 mm. thick, so that the radiations which passed through the cells 

 had to traverse 3 mm. of glass and 12 cm., 5'2 cm., 102 cm., or 

 15 2 cm. of water in Cells I, II, III, and IV. 



The determinations were made by placing one of the cells in front 

 of the thermopile, and about 9 cm. from it, a screen with a rect- 

 angular aperture, 5 cm. by 3 cm., being fixed at about 2 cm. from the 

 end of the cell nearest the pile, in order to prevent any radiations 

 reflected from the surfaces of the water in the cell reaching the pile ; 

 the glass plates were always well cleaned immediately before being 

 used, and the cell filled with freshly filtered distilled water. 



An Argand burner, 1*5 cm. in diameter, with a glass chimney 4*5 cm. 

 in diameter and 15 cm. high, was placed beyond the cell, a tin-plate 

 screen being interposed. 



In the experiments recorded in this paper the axis of the burner 

 was 31 '8 cm. from the face of the pile. 



The index of the galvanometer was then, if necessary, brought to 

 zero and the circuit closed ; closing the circuit almost always caused 

 the line of light to move in one or other direction. By adjusting the 

 height of the flame of the compensating burner, which was invariably 

 small, from 0'3 cm. to 0'8 cm., and its distance from the thermopile, 

 the line of light was brought back to the zero, or nearly so. 



A horizontal wire was clamped to the stand of the burner whose 

 radiation was to be observed, 10 cm. above the plate of the burner, 

 and by means of a tap the height of the flame so regulated that its 

 tip just appeared above the wire. 



The metal screen was then removed, the amplitude of the first 

 swing of the galvanometer needle observed, and the screen replaced. 



The oscillations of the needle were extremely slow, and, therefore, 

 instead of always waiting till it had completely come to rest again, as 

 soon as the oscillations had become small, about half a division of the 

 scale or less, the screen was again removed and another reading 

 made, care being taken to always remove the screen whilst the index 

 was passing the point taken as zero, and in the opposite direction to 

 that in which the deflection due to the heat would occur. 



In this way twelve readings were made, and then the cell was 

 removed and the lamp placed at a greater distance from the pile, 

 which was shielded by a tin-plate screen which could be moved to and 

 fro by strings. 



It was usually necessary to readjust the compensating lamp, and 

 when the index had again been brought to zero the screen was 

 removed and the deflection noted. After a few trials a position was 

 found for the lamp in which the deflection was about the same as that 

 produced by the radiation which had passed through the water ; a 

 reading of the deflection was made with the lamp in this position, and 



