464 M. J. Habermann on a Modification of Dumas's 



The experiments are carried out in precisely the mode de- 

 scribed by Dumas ; that is, the balloon filled with air is weighed, 

 the barometer and thermometer are noted, and all the other 

 well-known precautions which influence the accuracy of the re- 

 sults are taken. About a gram of the substance is introduced 

 into the balloon in the ordinary way ; it is then clamped between 

 two iron rings in the position indicated in the figure, and sunk 

 to a suitable depth in the bath. The bending, a, of the capil- 

 lary neck must be in proximity to the edge of the vessel and 

 yet above it. The open end of the capillary tube projects beyond 

 the edge of the vessel without touching it, and is inclined 

 upwards. One end of the bulb-tube is connected with the 

 pump, and the other end with the capillary neck of the balloon 

 by means of the caoutchouc tube, in such a manner that about 

 1 centim. of the caoutchouc tube is within the bulb-tube. 

 The position which the bulb-tube assumes will be evident from 

 the figure. It may be supported on a movable stand. 



If the caoutchouc connexion at b is suitably adjusted, which 

 is very easily effected, and the connexion with the pump made 

 with care, the apparatus will be air-tight, and the experiment 

 may be commenced. The small balloons are evacuated up to 

 a pressure of 500 millims., the larger to 600 millims. and up- 

 wards, as indicated on the manometer ; the cock d is then closed 

 and heat is applied. 



The liquid in the balloon is rapidly heated until it com- 

 mences to boil, which is readily known from the passing of 

 drops over into the receiver; and the flame is so regulated that 

 the temperature of the bath rises about 1° in from 1 to 2 

 minutes. When the thermometer in the oil-bath indicates a 

 temperature about 10° or 20° below the boiling-point of the 

 substance employed, the evacuation is reinstituted until the 

 mercury column in the manometer, which meanwhile will 

 have fallen a few centimetres in consequence of the higher 

 temperature and the vapour- tension, has risen to its ori- 

 ginal height, the portion of the capillary tube projecting out 

 of the bath up to the bend a being bathed uninterruptedly in 

 the liquid of the bath. 



When the thermometer in the oil-bath nearly indicates the 

 boiling-point of the substance, the neck at c is softened by 

 means of the blowpipe-flame and the balloon is sealed by 

 drawing away the bulb-tube, which under the given circum- 

 stances can be done with great ease and ■s'sdthout risk to the 

 experiment. Of course, the temperature of the oil-bath and the 

 heights of the manometer and barometer are at the same time 

 noted. I have never perceived in my experiments that, at 

 the temperature of sealing, liquid still distilled over out of 



