16 UNIVERSITY OF CQLOKADO STUDIES 



ing the mercury completely corrected this fault, we surmise 

 that the leakage must have taken place along the contact surface 

 between the mercury and the iron tubing. No appearance of 

 alcohol at /' was noticed, but it could easily have evaporated 

 too rapidly. 



The binding posts at g and h were connected to a circuit con- 

 taining a dry cell and a relay. The relay opened and closed a second 

 circuit containincr an electromagnet, which in turn controlled the 

 main heating circuit. 



For temperatures only a few degrees above that of the room the 

 makes and breaks at /' may open and close the heating circuit com- 

 pletely ; but for temperature 30 ° or 40 ° C. above that of the room 

 it is better to arrange the circuits so that the makes and breaks aty 

 throw an extra resistance into and out of the main circuit. In 

 this case, of course, the larger of the two main currents must 

 be sufficient to heat the bath up to a higher temperature than 

 the required one, and the smaller insufficient to maintain it at 

 that temperature. 



In our apparatus the glass tube e was open at the top. A large 

 variation in the atmospheric pressure might produce a change in the 

 temperature at which the makes and breaks take place on account of 

 the slight compressibility of the alcohol. Xo such effect has been 

 noticed, however. If it occurred the fault could easily be rem- 

 edied by fastening a tube with a large bulb on its end to the 

 tube e. The bulb, of course, would have to be below the surface 

 of the bath, so that the temperature of the air within it would not 

 change. 



In practice it is easy to set the regulating device to working at 

 any desired temperature between that of the room and one a few 

 degrees below the boiling point of the liquid in the tubes, as follows. 

 With the cap A unscrewed, allow the bath to heat up slowly, and 

 when the desired temperature has nearly been reached, screw 

 the cap down. A little practice will enable one to set the ther- 

 mometer at a temperature within a small fraction of a degree of 

 the desired one. 



