TRANSACTIOKS OF SECTION A. 567 



Depaktment III. 



18. A Mi'thod for accurately DetcrmlniiKj the Freezinff -point of Arpieous 

 Solutions ivhich freeze at Temper at xires just behno 0° C. JUj tlie Late 

 P. B. Lewis. Communicated by Dr. Mejer AVildermann. 



The tbermometers used for measuring minute variations in temperature were 

 f^radnafed, the one to read to thousandths the other to hundredths of a degree. The 

 total vohime of the solution used in the various determinations was J, 250 c.c. The 

 beaker containiu"- tlie liquid for examination was placed on a felt cushion, wrapped 

 in thin gutta-percha, inside a zinc vessel which acted as an air-chamber. The 

 cover of this vessel was provided with a circular opening into whicli litted a cork, 

 holding the thermometers. The stirring of the liquid was efi'ected by a porcelain 

 stirrer which consisted of two parallel porcelain plates, in which there were corre- 

 sponding holes for the thermometers to pass through, and besides these eight round 

 holes so arranged that the holes in the upper plate did not correspond with those 

 in the lower plate. On moving the stirrer up and down currents are sent' in all 

 directions. The thermometers are kept uniformly at a temperature of 0° C, even 

 when not in use, by keeping them plunged in a glass beaker, tilled with ice and 

 water at 0° C. This second beaker is contained in a similar zinc vessel, which acts 

 as a cold-air chamber. The two zinc vessels, the one containing the distilled water 

 and the other the solution to be examined, are placed in a larger zinc vessel, which 

 acts as an ice-bath, and which is encased in a yet larger zinc vessel, so that there 

 should be a cushion of air of some centimetres' thickness between the two. The 

 outermost zinc vessel stands on a thick felt mat, and is wrapped in a thick felt 

 cover. Ilie ice-bath was covered over as far as possible with pieces of asbestos 

 cloth. 13y these means a fairly constant temperature is maintained in the bath. 

 The temperature of the ice-bath should only vary between 1°'8 and — 2° when the 

 It'mperature of the room is 12° to 18°, and onl}^ between — 2°"1 and 2°-2 at a tem- 

 perature of 25°. It is necessary for a successful experiment to overcool the solu- 

 tion by about 0°-7 to 1°. Under these conditions the most accui-ate determination 

 of the freezing-point can be obtained : the mercury in the large thermometer has 

 suHicient time to become exactly of the temperature of the solution, the ice is obtained 

 in the form of thin films, and the freezing-point remains fairly constant for from 

 ten to fifteen minutes, though slight variations of from O°'O0Ol to 0°'0002 occur, 

 and in the more concentrated solutions of 0°0003. As regards the action of the 

 theriUo:ueters it is important to know how far the capacity of their bulbs is affected 

 by the pressm-e caused by a rise, especially a rapid rise, of mercury in the stem, and 

 how quickly the bulb recovers from the expansion or contraction which variations of 

 pressure or temperature cause. It can only be here stated that if such variations exist 

 they lie within the limits of ordinary experimental error, namely, 0°'0001, 0°'0002, 

 and I'ai-ely 0°-0003. On the other hand, it was found that when the more delicate 

 thermometer had been allowed to remain for several hours at the ordinary tem- 

 perature, a rise in its indication of the freezing-point could be observed for many 

 days afterwards, and a long-continued exposure to a rise or fall of atmospheric 

 pressure causes a variation of about 0°-000.3 in the reading of the freezing-point 

 for 1 mm. rise of the barometer. A similar variation is produced in the reading of 

 the smaller thermometer. In the case of very dilute solutions only those determi- 

 nations of the freezing-point which have been made during almost constant condi- 

 tions of atmospheric pressure are to be depended on. While the column of mercury 

 is rising the stirring must be maintained together with a light tapping with the 

 fingers on the cork. The maximum is reached when further tapping causes no 

 rise of the mercurial column. AVhen this point has been reached readings are made 

 every minute for from eight to ten minutes. The readings are made with a small 

 lens. After the constant point has been observed for from eight to ten minutes the 

 form taken by the ice produced is examined. If the experiment has been success- 

 ful the ice will have formed in fine films and in considerable quantity throughout 

 the solution. Art account of this investigation will be given in the Journal of the 

 Chemical Society, and in the 'Zeitsch. f, physik. Chemie.' 



