26 UNIVERSITY OF COLORADO STUDIES 



change in volume the system is placed in a spiral glass tube (Fig. 4) 

 on one end of which is sealed a capillary tube A B, and on the 

 other end a ground stopper and cup C. The reacting system may 

 fill the entire spiral or only a part of it, the rest being filled with a 

 liquid that does not combine with the given system. If the volume 

 increases during the reaction, the cup A and most of the tube A B are 

 filled with mercury. If the volume decreases, the cup and only a 

 small part of the tube are filled. Evidently during the change of 

 volume the mercury is drawn in or pushed out through the capillary 

 tube and the positions of the end of the mercury column measure 

 the amount of change. 



The particular form represented (Fig. 4) was adopted, firstly, in 

 order that the temperature of the system should remain that of a 

 bath in which the spiral is immersed owing to the large surface pre- 

 sented to the bath; and secondly, in order that the tube might be 

 filled, and the first reading of the position of the mercury column 

 taken as soon after the reaction started as possible. The liquid is 

 first poured in at C and allowed to fill the spiral, the tube A B and 

 the cup A, Mercury is then poured into A and finally the stopper 

 inserted at C. A little practice in manipulating the stopper and 

 tilting the whole tube will enable one to set the dividing surface 

 between the mercury and the other liquid at any desired point in the 

 capillary tube. The tube A B is graduated or a graduated scale is 

 fastened to it. 



In most cases the change in the volume of a system is small. 

 Hence, if the temperature of the apparatus varies during the reaction, 

 even by a small amount, the consequent change in volume will be a 

 large fraction, perhaps, of the change we wish to measure. It is, 

 therefore, necessary to immerse the spiral tube in a delicate ther- 

 mostat. In the following experiment on the inversion of a 25 per 

 cent, solution of cane sugar the electrical thermostat described by my 

 assistant, Mr. Lory, and myself in this Journal for March, 1900, was 

 used. The temperature of the bath did not vary as much as ^yVir^^ 

 of a degree except once, and then the variation was only about j-jjig-g-th 

 of a degree. 



