and on the Motion accompanying Heat -conduction. 1G3 



inclination required towards the axis. In the following experi- 

 ments the two always made an angle of 35° with one another. 

 The thermo-electric pile with its cone was so fastened to the 

 mirror by means of an arm, that the rays falling upon the mirror 

 in the direction of the axis of the tube were reflected into the 

 cone of the pile. The entire arrangement could be turned with 

 the tube and its axis so that the mirror and thermo-pile could 

 be brought into all azimuths. To protect them from external 

 influences, they were surrounded by a large pasteboard box, 

 which was further protected by thick hanging covers. Only a 

 small part of the tube extended beyond the box, and by this 

 means the mirror and thermo-pile could be turned from the out- 

 side without opening the box. An index fastened to the tube 

 showed the position in the interior. Between the box and the 

 source of heat another large screen was erected, in which there 

 was an opening of the same size as those at the two ends of the 

 tube. As the rays passed through these three openings, they 

 reached the mirror in almost parallel directions. For the source 

 of heat a tin-plate vessel was employed, which could be main- 

 tained at 100° C. by means of a current of steam. This steam- 

 box was so fastened upon a stout wooden slab that it could 

 be turned upon a horizontal axis whose continuation passed 

 through the centre of the mirror. The face of this vessel, which 

 radiated upon the mirror, made an angle of 35° with the hori- 

 zontal. Other plates could be fastened upon the steam-box, so 

 as to serve as radiating surfaces. The steam entered the vessel 

 out of a large glass bolt-head, and left it by means of a tube 

 which led into a large flask containing many litres of cold water. 

 In order to remove the water condensed in the vessel which 

 served as the source of heat, the tube which led off the steam 

 was so bent in the inside of the box, that its opening always 

 reached down to the lowest part. When, therefore, the water 

 had reached this point, it was forced out of the box by the 

 freshly arriving steam. A moveable screen of double metal 

 plates stood between the radiating vessel and the fixed screen, in 

 which the first diaphragm was situated. It was removed when- 

 ever an observation had to be made, and replaced as soon as the 

 needle had completed its deviation. 



The observation was made by observing the first displacement 

 of the needle. It may, it is true, appear that it would be prefer- 

 able to wait until the needle had assumed a constant position. 

 But the results are then less sharp and exact. During the long 

 continuance of the radiation the pile becomes, in fact, warmed, 

 and consequently the position of the needle never remains quite 

 constant. At the same time the space in the inside of the box 

 becomes warmed, and the same value is not always obtained on 

 repeating the same experiment. Accordingly the method of 



M2 



