204 THE ROYAL SOCIETY OF CANADA 



have been made in liquid chlorine and hydrogen bromide. The results 

 of the first investigation form the subject of this paper. 



The apparatus for the measurement of the temperatures was a 

 Carey-Foster bridge of the kind used by Beckmann* and one of us. 

 It consisted of a metre bridge wire of 14-42 ohms resistance, in series 

 with two coils of 3,000 ohms. The thermometer was of the Heraeus 

 platinum type wound on quartz, with a resistance of 25.015 ohms at 

 0°. The balancing resistances were Post Office boxes, and these were 

 calibrated by a standard Wheatstone bridge. A Broca galvano- 

 meter proved most satisfactory, and at the temperature of freezing 

 chlorine ( — 101-5°) one millimetre on the bridge wire equalled 

 0-00143°C. The apparatus was sensitive to 0-0005° but the other 

 errors were naturally much greater. 



The cryoscopic apparatus is shown in the diagram. A is a 

 3-inch silvered Dewar flask into which is fitted a rubber stopper carry- 

 ing a test tube B, which serves as an air chamber for the freezing 

 point tube. T is the platinum thermometer with heavy copper 

 leads making contact with the bridge through two large mercury cups, 

 so that the tube could be easily removed. The stirrer S, is of platinum 

 and is moved by a platinum wire which passes over the pulley to an 

 excentric on the shaft of a small motor. Dry air is forced through 

 F to prevent the entrance of moisture. The evaporation of chlorine 

 at its F.P. (9-2 mm pressure) is very slight. The Dewar flask con- 

 tained solid carbon dioxide and ether boiling under reduced pressure. 

 A May-Nelson pump and a manometer with a system of taps enabled 

 the pressure, and consequently the temperature, to be maintained with 

 sufficient accuracy. (— 105°— 108°C). 



In making an experiment, a convenient quantity of chlorine, 

 usually about 50 grams, was condensed in the freezing point tube 

 and weighed, the weight of the boiling tube with stirrer, thermometer, 

 etc., having been previously determined under identical conditions. 

 (With a little practice this can be done to 0-1 percent.). The resist- 

 ances of the boxes were adjusted until the slider was at a convenient 

 point on the bridge and its position was noted at the freezing point of 

 the chlorine. The tube was removed, the chlorine melted and the 

 solute added from a pyknometer, through the side tube. The freezing 

 point was again determined and this procedure repeated till five or 

 six measurements had been made. The primary source of error and 

 one difficult to avoid was in the addition of the solute. Variations in 

 the results were generally traceable to a portion of the solute remaining 

 on the walls of the tube or on the stirrer, and precautions were taken 

 to avoid this error as far as possible. 

 1 Zeit. anorg. Chem. 67, p. 17, 1910. 



