﻿Prof. R. Bunsen's Calorimetric Researches. 167 



I have constructed this ice-producing apparatus in the form 

 shown at fig. 3. The two semicylindrical tin-plate vessels a and b 

 correspond to the single vessel A shown in fig. 2, and are con- 

 nected by tubes with each other and with the tube a ; while the 

 similar tin-plate vessel, the outside of which is marked c, corre- 

 sponds to the vessel B, fig. 2. These two vessels, forming two 

 concentric chambers round the tube a v possess a great cooling- 

 surface, and are both immersed in the same freezing-mixture. 

 The arrangement of tubes in fig. 3, by which the circulation of 

 the cooled alcohol is accomplished, is easily understood, as the 

 corresponding caoutchouc tubes are denoted by the same letters 

 as in fig. 2. The alternate suction of the alcohol is managed 

 by means of the stopcock H, which is connected at w with the 

 water-pump. When this stopcock is placed in one position, 

 the tube q communicates with the exhausting- tube w } and the 

 tube p with the outer air ; when placed in the other position, 

 the reverse takes place, and w communicates with p and q with 

 the outer air. The production of the cylinder of ice becomes a 

 very simple operation by this arrangement. The cooling-appa- 

 ratus with its caoutchouc tubes is placed in the freezing-mixture, 

 p and q are connected with the stopcock H, and w with the 

 water-pump ; the caoutchouc stopper with the tubes m } n at C 

 is sunk in the inner vessel of the instrument, and, lastly, the 

 tubes m and n are connected with the corresponding glass tubes 

 of the cooling-apparatus. If, after the stopcock of the water- 

 pump has been opened, the reversing stopcock is turned alter- 

 nately, the stream of cooled alcohol can be kept at work as long as 

 necessary for the production of the cylinder of ice. The formation 

 of this last is easily observed with the naked eye or with a tele- 

 scope, and presents some not uninteresting peculiarities. The 

 temperature of the air-freed water in the outer vessel (b, fig. 1) 

 sinks by degrees (without freezing taking place) far below 0° C, 

 while the outside of the vessel becomes covered with a coating of 

 ice formed from atmospheric moisture, which even vigorous 

 shaking cannot remove. At last, when the temperature has 

 sunk very low, the formation of ice begins suddenly, and spreads 

 in a few seconds from X to p,. The whole vessel down to these 

 limits is filled with thin plates and needles of ice; but the 

 water from //, to the level of the mercury at /3 is not frozen. 

 By continued cooling the formation of the ice-cylinder now 

 begins, and is allowed to go on until it has attained a thickness 

 of from 6 to 10 millims. This shell of ice below \x appears com- 

 pletely amorphous, as clear and transparent as the purest crys- 

 tal ; the upper part between fi and \, however, is not transpa- 

 rent, and has a coarsely fibrous texture. When, however, the 

 instrument has stood in snow for some days at 0° C, this coarse 



