260 THE ROYAL SOCIETY OF CANADA 



The ether was a pure sample carefully freed from alcohol by- 

 repeated washing with water, dehydrated by distillation over calcium 

 chloride and finally allowed to stand for some days over sodium wire. 

 This latter step removes even the last traces of water which will 

 otherwise show themselves as a cloud in the ether when the latter is 

 cooled. 



The oxonium compound of ether and hydrobromic acid in mono- 

 molecular proportions was prepared by addition of the correct amount 

 of ether to a weighed quantity of the acid. This process required 

 some care in manipulation owing to the heat of combination which 

 tended to cause some of the acid to evaporate. Into a weighed dry 

 test-tube, dry hydrobromic acid was distilled and the whole reweighed 

 and thus the weight of acid obtained. To this ether was added drop 

 by drop. At first, when the heat evolution was greatest, a long interval 

 between the addition of drops was necessary for cooling, but as the 

 correct proportions were approached the speed of addition could be 

 increased. When the tube contained the required amount of ether 

 it was reweighed to determine just how great had been the loss of acid 

 in the process. The ether additions were made by means of a pykno- 

 meter as also were the subsequent additions of water to the compound. 



The freezing-point of the mixture was then determined and found 

 to be — 40-5°C., agreeing well with the value — 40°C. found by 

 Mcintosh.^ A slight discrepancy was to be expected due to the 

 difficulty in mixing the two components in exactly molecular pro- 

 portions. Definite amounts of water were added and the freezing 

 point of the system determined each time. The temperatures were 

 found by means of a platinum thermometer which gave readings to 

 0- 1°C. with ease. After the series the total contents were dissolved in 

 a large excess of water and the hydrobromic acid estimated by stand- 

 ard alkali, the loss of acid during the determinations being distributed 

 proportionately when calculating the percentages. After the first 

 few points were determined the viscosity of the solution began to 

 increase very rapidly, so much that the recognition of the exact 

 freezing point became a matter of much difficulty. Finally at a com- 

 position of about 16% water, on attempting to freeze out solid by 

 cooling, the mass became so viscous as to be correctly characterized 

 as a glass and the curve could be continued no further. The values 

 obtained are shown in Table I. 



iJ.A.C.S. 33, 70 (1911). 



