1884.] for determining the density of Ozone. 209 



the jars brought into communication by sliding the dampers until 

 the holes were concentric. 



The diffusion thus established was allowed in all our ex- 

 periments to proceed for 45 minutes, after which the dampers were 

 closed. 



The quantity of chlorine or ozone present in each jar was 

 then determined by absorption with KI, and subsequent titration 

 by a solution of Na 2 S 2 3 . Whatever be the chemical reactions, 

 the amount of Iodine liberated, and therefore the number of 

 cc. of JNa 2 S 2 3 used is in ever}' case proportional to the amount 

 of chlorine or ozone present in the upper and lower jar. (For 

 convenience, we used a standard solution of Na 2 S 2 3 to obtain an 

 idea of the quantity of gas used in each experiment.) 



Let v and V be the number of cc.'s of Na 2 S.,0 3 required for 



the upper and lower jars respectively when chlorine was the gas 



v 



diffused, then ^ = ratio of amount of chlorine diffused to the 



K+ v 



whole amount of chlorine taken. 



This ratio is independent of the quantity of diffusing gas 



originally placed in the lower jar but is a function of the time, 



the temperature, the size of the smaller hole, and the density of 



the gases. If the time of diffusion were long enough this ratio 



would become |, but when the time is comparatively short the 



v 

 ratio ^ may be taken to be proportional to the velocity of 



diffusion of the particular gas, that is to —.-= where D — density. 



The mean of six determinations with chlorine gave 



V + v 



The mean of fourteen determinations with ozone gave 



V + v 



jr (117) 2 _ density of ozone 



(147)' 2 density of chlorine ' 



Assuming the density of chlorine to be 35"5 we get from this 

 equation the density of ozone to be 22"5, a result which sufficiently 

 justifies the formula 3 for the molecule of ozone. 



The following table gives the actual numbers recorded in our 

 experiments. 



