14 



UNIVERSITY OF COLOEADO STUDIES 



might seem at first thought, that, owing to the time required for the 

 heat to penetrate through the walls of the tubes to the expanding 

 liquid within, the making and breaking of the circuit would take 

 place rather slowly, and the temperature of the bath would be oscil- 

 latory instead of constant. This is undoubtedly true to a certain 

 extent. Practically, however, with our arrangement of apparatus 

 the variation of temperature is too small to be detected even by a 

 differential thermometer that would indicate a change of tempera- 

 ture of 2 o\ th of a degree Centigrade, Indeed the elBiciency of the 

 apparatus is due to the fact that the makes and breaks follow each 

 other so rapidly, that there is not time for the temperature to change 

 perceptibly between them. Often two, three or even more interrup- 

 tions of the current occur in one second. At first we used an ordi- 

 nary wash-boiler containing about 150 liters of water as a bath. 

 The boiler was placed in a 

 large wooden box and packed 

 in around the sides with wool. 

 A light wooden frame im- 



mersed in the water served as 

 a rack to hold the wires con- 

 veying the current. We en- 

 countered considerable difli- 

 culty, however, owing to the 

 wires becoming very brittle 

 and breaking after a few days' 

 use. This was the case with 

 wires of copper, iron and 

 German silver. A direct 

 current, too, seemed to pro- 

 duce a much greater effect 

 than an alternating one of 

 the same heating power. 



To obviate this difficulty 

 we constructed a wooden 

 trough (80*='"X40'='^X 50"="") and filled it with a weak solution of 



B 



Fig. 1. 



