668 PROCEEDINGS OF THE AMERICAN ACADEMY. 



had been established, often for hours at a time no change could be de- 

 tected on a Beckmann thermometer which could easily be read to thou- 

 sandths of a degree. In the latter part of the work the mercury 

 occasionally caught in the annular space outside the capillary inlet 

 tube, causing sudden and vexatious fluctuations of temperature. This 

 would perhaps be obviated by a different ratio between the size of the 

 capillary entrance tube, the size of the annular space, and the head of 

 the water. It was found more convenient, however, to discard alto- 

 gether the method of cutting off the flow of hot water directly by means 

 of the mercury column, and to use the mercury as a means of making and 

 breaking a feeble galvanic current. When made, this current actuated a 

 relay which in turn sent a stronger current through an electro-magnetic * 

 arrangement for cutting off the supply of hot water. This apparatus 

 worked admirably, and was used in a few of the last experiments. 



The Cold Thermostat. 



It was endeavored, at first, to maintain a constant temperature below 

 that of the room, by passing a rapid stream of tap water around the 

 pipette containing the solution. This gave surprising constancy at 

 times, but was usually far too variable for the purpose. In the hope 

 of increasing its constancy, the tap water was then made to pass first 

 through a large mixing vessel, itself well insulated by an outside jacket 

 of running water. This was a slight, but inadequate improvement. A 

 special apparatus was therefore devised, but this apparatus was finally 

 supplanted by a much simpler one, and therefore need not be described 

 in detail. It utilized the tap water as a cooling agent, being available 

 whenever this is tolerably constant, and consisted essentially of two water 

 jackets, one inside the other, insulated from one another with cotton 

 wool. The inner one contained the pipette, wherein the solutions were 

 placed. The liquid within the pipette has been kept in this device for 

 over an hour without suffering the slightest change in the temperature 

 as observed by means of a very sensitive Beckmann thermometer. 



As has been said, this method also was discarded for the much simpler 

 one of packing the pipette in ice, — a proceeding which would have been 

 adopted in the first place had it not at first seemed advisable to maintain 

 a temperature somewhat above the freezing point. By using pure ice, 

 mixed with previously chilled, boiled, distilled water, we were able to do 

 away entirely with the measurement of the temperature of the cold solu- 

 tion, because it was exactly zero. This of course was a distinct gain. 



