[sHAVv] A SENSITIVE BATH-THERMOSTAT 135 



It will be seen that as Hi — Ho became greater, p2 became less. 

 In the above case the automatic regulation covered a possible external 

 change in temperature of nearly five degrees. In almost every 

 building a room can be found which will not change so much in tem- 

 perature during a season. If necessary however, R2 and R3 can be 

 altered to suit. In the above case R2 was a 32 c.p. carbon filament 

 lamp, and R3 was an 8 c.p. carbon filament lamp. An assortment 

 of carbon filament lamps of 4, 8, 16 and 32 candle power respectively, 

 with a four socket lamp-board will be sufficient under all ordinary 

 external temperatures, to obtain regulations at temperatures below 

 30°C. With more lamps this range could be extended upwards and 

 with cooling coils it can be taken below room temperature, if desired, 

 to within a few degrees of the temperature attainable in the cooling 

 coils. In every case the same constancy in regulation should be 

 apparent. 



In adjusting the thermoregulator for use at a given temperature, 

 the bath is first heated by the coils approximately to the desired 

 temperature and the level of the mercury is adjusted by means of the 

 tap D through which mercury may be added or removed easily. The 

 final adjustment should be made by moving the wire in the collar at 

 A (see the diagram), and checking by consulting a Beckmann or an 

 electrical thermometer immersed in the bath. It will be seen that 

 a considerable movement of the mercury per hundredth of a degree 

 change is necessary if sticking is to be avoided and if close adjustment 

 of the wire is to be possible. In our case with about 1,100 cc. of 

 toluene and 50 cc. of mercury, the expansion up the tube above B 

 was about 6 mm. per hundredth of a degree change in temperature in 

 the bath. 



In order to increase the sensibility further (i.e. lessen the range 

 of variation) it is necessary to make H2 nearer Hi in magnitude. 

 Hi being as small as possible. This requires greater constancy in 

 external temperature, or more frequent readjustment of the electrical 

 circuit. It will be seen what an important part the double sides play 

 in making further refinement possible. 



In conclusion the writer desires to express again his indebtedness 

 to Dr. H. L. Bronson for his direction and association in the con- 

 struction of the apparatus described. Many thanks are due also, to 

 Dr. H. T. Barnes who suggested the investigation which required 

 its use; and full acknowledgment to Messrs. Wolff and Waters for 

 the general design. 

 Macdonald College, 



McGill University. 

 Mav, 1917 



