499- 



more concentrated state, the ordinary methods of determining the 

 density of a gas are plainly inapplicable. The difficulty of the pro- 

 blem was farther increased by the rapid action of Ozone on mercury, 

 which rendered it impossible to collect or measure the gas over that 

 metal ; and the tension of aqueous vapour, as well as the gradual 

 destruction of Ozone by water, prevented the use of the latter. After 

 numerous trials, the method finally employed was to measure the 

 change of volume which occurs in exposing a gaseous mixture con- 

 taining Ozone to a temperature of 230 C., or upwards. The volume 

 of the gas after this treatment was invariably found to have increased ; 

 and by eliminating the effects of alteration of temperature and pres- 

 sure during the course of the experiment, by the aid of a similar vessel 

 to that containing the Ozone, the authors succeeded in estimating the 

 change of volume which took place, to an extremely small fraction 

 of the entire amount. 



The vessels employed in different experiments varied in capacity 

 from 200 CC. to 600 CC., and terminated in tubes of about 2 milli- 

 metres in diameter, bent in a U-form and containing sulphuric acid. 

 The amount of Ozone was ascertained by passing a stream of the gas 

 through two other vessels, one placed on each side of the vessel to 

 be heated, and afterwards analysing their contents by the method 

 described in the communication already referred to. It was easy to 

 measure with certainty a change of pressure amounting to g^th 

 of the whole ; but on account of the ordinary fluctuations of atmo- 

 spheric pressure between two consecutive observations of the primary 

 and auxiliary vessels, it was rarely possible to work to this degree of 

 accuracy. 



The experimental data have not yet been completely reduced, and 

 some slight corrections have yet to be investigated ; but the general 

 result of the inquiry, which has been a very protracted one, gives 

 on the assumption that Ozone is oxygen in an allotropic condition 

 for its density as compared with that of oxygen, nearly the ratio of 

 4 to 1. 



The following approximate formulae were employed in the reduc- 

 tion of the experiments. They are sufficiently exact for the purpose 

 of calculation on account of the smallness of the quantities observed. 



I. To reduce the change of level observed in the auxiliary vessel 



