475 



He is also conscious that the phenomena themselves are the chief 

 ground on which he can rest a claim for originality, and that the 

 explanation of them may be better treated by those who are more 

 accustomed to deal with similar researches. 



In the course of these experiments the author has been especially 

 indebted for many valuable suggestions to the Rev. Dr. Booth, 

 Dr. Roget, Professor Sharpey, and Mr. Bishop ; and he is also under 

 obligations to Professor Stokes and to Mr. Brooke. 



XII. "On the Constitution and Properties of Ozone." By 

 THOMAS ANDREWS, M.D., F.R.S., Professor of Chemistry 

 in Queen's College, Belfast. Received May 16, 1855. 



The conflicting views which have so long existed as to the true 

 constitution of ozone, induced the author to undertake a careful in- 

 vestigation of the subject, particularly as he had reason to doubt the 

 accuracy of the only quantitative experiments which have yet been 

 made to elucidate this difficult question. According to the experi- 

 ments referred to, two substances have been confounded under the 

 name of ozone, one a compound body having the formula HO 3 , the 

 other an allotropic variety of oxygen. To ascertain whether, in con- 

 formity with this statement, ozone obtained in the electrolysis of 

 water contains hydrogen as a constituent, the author made two 

 series of experiments. In the first series, he followed nearly the 

 same method of investigation by which its compound nature is sup- 

 posed to have been established, but modified so as to avoid a source 

 of error, which, if neglected, vitiates altogether the results. Electro- 

 lytic oxygen, unless very great precautions be taken, is always ac- 

 companied by a small but appreciable quantity of carbonic acid, 

 which is liable to be partially absorbed by the potassa set free when 

 a neutral solution of iodide of potassium is decomposed by ozone. 

 By adding a little hydrochloric acid to the solution of iodide of 

 potassium before the commencement of each experiment, this error 

 may be avoided. 



The method of performing the experiment was to conduct a stream 

 of electrolytic oxygen through a compound apparatus previously 



VOL. VII. 2 X 



