HEATING EFFECTS PRODUCED BY KATHODE RAYS 315 



in this way should have been about the same for each plate. There 

 was practically no loss of heat by conduction through the vulcanite. The 

 loss by radiation was equalized by coating both plates with thin aluminum 

 leaf. The heat losses from all causes were slight, for the temperature in 

 the tube was never raised more than a few degrees. The plates were held 

 in position by small drops of water-glass placed at each end and by the 

 pressure of the thermoelectric couples placed at the centers of their back 

 faces. Any asymmetry in the bundle of kathode rays could be eliminated 

 by rotating the vulcanite ring to which the plates were attached. 



The thermoelectric couples were made by joining No. 24 Ni and Cu 

 wires at an angle of about 30 with silver solder. The junctions were 

 then filed away until they presented points about o. 25 mm. square in area. 

 The wires were insulated by small glass tubes and sealing-wax where 

 they passed through the the brass cap. The second Ni and Cu junc- 

 tions, outside the tube, were protected from air currents and sudden 

 changes in the room temperature by lumps of soft wax molded around 

 them, a device which served its purpose admirably by eliminating 

 entirely the previous erratic behavior of the galvanometer. The resist- 

 ances and the thermoelectric properties of the couples were tested in 

 various ways and found to be practically identical. 



With the above arrangement, which was exceedingly sensitive to 

 changes in temperature at the junctions within the tube, many experi- 

 ments were made, the results of which show no appreciable difference 

 in the heating effects produced in lead and zinc. With other arrange- 

 ments the results were erratic, sometimes showing a heating effect 

 greater in the lead and again greater in the zinc. In almost all such 

 cases the real heating effects were found to be masked by static charges 

 which accumulated because of imperfect grounding. My assistant, 

 Mr. F. A. Giacomini, repeated some of the experiments in 1907 with 

 a slightly modified form of apparatus and his results confirm those 

 obtained before. All the results obtained have been reduced to the 

 form of curves but as they present nothing particularly new it does not 

 seem necessary to publish them. The main object of this note is to 

 make known the general results obtained by the method of attacking 

 the problem described above. 



