62 



REPORT — 1856. 



these bodies are not self-luminous. The light seems to be owing to the me- 

 teor, instead of the light of the meteor ; probably the great speed causes a pe- 

 culiar property of the upper regions to ignite, at the instant of ignition being 

 an intense blaze, and then subsiding into a phosphorescent flame, which may 

 linger for a length of time and be wafted along by currents of air, as was the 

 case in several instances. In the case of the meteor of Dec. 19, 1855, it moved 

 over 1 8^° in less than a second of time ; it cannot therefore be supposed that 

 the meteor itself could be within 5° of this path 10 minutes afterwards. Now 

 if we suppose the meteor burst at this point (which to me seems improbable), 

 it must have burst in a medium where light could shine, and if so it is as easy 

 to suppose some substance should be ignited, as the meteor itseli' should blaze. 

 The intense brightness is too great fur reflected light. „ p j j „ » 



Fig. f^. 



Fig. 5. 



Photochemical Researches. By Professor Bunsen, of Heidelberg, 

 and Dr. Henry E. Roscoe, of London. 



We had the honour of laying before the Chemical Section of the British 

 Association at the Glasgow Meeting, a short account of a series of experi- 

 ments which we had undertaken with the view of becoming more nearly 

 acquainted with the laws which regulate the chemical action of light, and of 

 obtaining, if possible, a measure for this action. 



These experiments, the continuation of which has been assisted by a 

 grant from the Association, have been extended during the present summer 

 months, and we beg to lay before the meeting, in a short report, the chief 

 results as yet obtained. 



The method employed by us for measuring the chemical action of light is 

 founded upon the well-known fact that chlorine and hydrogen combine when 

 exposed to light. The employment of this reaction as a measure of the 

 chemical action of light was proposed and practically carried out by Dr. 

 Draper of New York in 184'4', to whom belongs the great credit of first 

 having attempted to obtain a measure for this action. A number of experi- 

 ments instituted for the purpose of testing the accuracy of the instrument 

 proposed by Draper, assured us, however, that not only for observations 

 extending over a considerable period of time, but even for those of short 

 duration the indications of the instrument were not reliable. The possibility 

 of obtaining exact photometrical results with a mixture of chlorine and hy- 

 drogen, depends upon the fulfilment of various conditions which in Draper's 

 tithonometer have not been regarded. Of these conditions the two most 

 essential are — 



1. The constant composition and purity of the gaseous mixture. 



2. Constant pressure exerted upon the gas. 



It is easy to show from the laws of gas absorption that the method em- 



