1884.] for determining the density of Ozone. 211 



Hence largest quantity of ozone diffused was about 2 cc. 



We hope shortly to apply this method to other gases and 

 vapours ; the chief difficulty however will be to find methods 

 sufficiently delicate for the estimation of small quantities. 



If the gas experimented upon, and the gas into which it is 

 diffused be of very different densities, a considerable source of error 

 will be introduced through mass motion. 



If we follow the motion of the chlorine molecules in the lower 

 jar and of an equal number of oxygen molecules in the upper 

 jar, we see that when diffusion has commenced the more rapidly 

 moving oxygen molecules will enter the lower jar by diffusion 

 in greater number than the chlorine molecules can enter the 

 upper jar. There will therefore be a tendency for molecules to 

 accumulate in the lower jar, in other words for the pressure to 

 rise, as is the case when a porous septum is used, and there- 

 fore there will be a continuous efflux of mixed gases from the 

 lower jar not due to diffusion proper. This will tend to make 

 v and therefore the ratio too great. 



Trial proved that the apparatus was gas tight for considerable 

 differences of pressure ; however, to gain additional security we 

 frequently lubricated our dampers with concentrated H 2 S0 4 . 



The absorption of the chlorine at the end of the diffusion 

 was effected by placing each jar closed by its damper mouth 

 downwards in a solution of KI, and the solution allowed to come 

 in contact with the gas. 



The oxygen in the jar was then finally washed from the last 

 traces of chlorine by blowing up a fine jet of KI solution from a 

 curved pipette. The jars containing the ozone were treated in 

 the same way. When the ozone came in contact with the KI 

 the jars became filled with a dense white nearly opaque fog, the 

 autozone of Schonbein. 



In these experiments the time allowed for each diffusion was 

 that adopted by Soret, and the size of the diffusion hole was rather 

 less than his. 



We think it will be worth while to enlarge the diffusion hole 

 considerably so as to be able to shorten the time. 



(3) On the effects of self-induction of the galvanometer in the 

 determination of the capacity of a condenser. By J. C. M'Connel, 

 B.A. 



In the calculation which Mr Thomson gives with reference to 

 the method of determining the capacity of a condenser, to which 

 your attention has been so lately drawn by Mr Wilberforce, there 

 is a point of some importance which he passes by without notice. 



