ON THE CHEMICAL REACTIONS Or LIGHT. 69 



small quantity of the substance is put into a very fine thin cylindrical txibc, 

 which is closed with the lamp at both ends after the introduction of the 

 body, and can be, when necessary, easily broken to pieces by a slight shock 

 against the tiibc. The mercury gas-holders which receive and conduct the 

 gases are glass, carefully divided, so as to measure the volume of the gases 

 experimented on. In lieu of an air-pump I have always made use of a 

 mercury exhauster, the only apparatus which can be absolutely relied on, and 

 which enables the operator, when the reaction is completed, to withdraw 

 the gases from the tube for analysis; the mercury exhauster likewise 

 enables us to measiure the elastic force of the gases before and after the reac- 

 tion, and thus indicates the variations of volume of the gas employed. 



What therefore gives these experiments a peculiar character is the com- 

 plete elimination of^aU gaseous vehicles employed to convey the vapour. The 

 conditions of my experiments differ in this respect from Mr, Tyndall's. 



The tubes, of smaller dimensions, which I have employed do not give 

 indeed the splendid results which Mr. Tyndall presented to his delighted 

 audience ; but these tubes are easier to set up and to clean : the mercury 

 exhauster possesses moreover another advantage ; it allowed me to ascertain 

 whether the gases were perfectly dry — a most essential point. 



As to the means of conveying the solar light, the process was the following. 

 A broad mirror receives and reflects horizontally a voluminous pencil of rays 

 which is refracted by a lens 22 centimetres in diameter and of 40 centi- 

 metres focal length. The luminous cone is enclosed in a metallic box, 

 whence it issues to penetrate into the tube through the glass plate. The 

 summits of the two cones, the converging and the diverging, are situated 

 pretty nearly in the centre of the tube (A) ; the two cones are therefore easily 

 visible when the modifications of the bodies contained in the tube are pro- 

 duced. It must not be forgotten that the periphery of the two cones is 

 coloiired, as we said before, in consequence of the non-achromatism of the 

 large condensing lens. 



Under these circumstances I was very much surprised to see that a mix- 

 ture of hydrogen and nitrogen, perfectly pure and dry, produced the reaction 

 cloud. They had been dried by a very slow passage through pounded glass 

 which had been calcined and moistened with sulphiu-ic acid, pure and highly 

 concentrated. I changed the desiccating substance, and successively made 

 use of potash, chloride of calcium, phosphoric acid, all recently melted. In 

 the three latter cases the cloud did not appear ; the action was null, and the 

 solar light passed unperceived. What coiild the s?(?^)7fHr/c acid, then, convey ? 

 evidently some little sulplmrous acid ; for sulphuric acid is in fact a real 

 emitter of it, and always adds a certain quantity of it to any pure gas 

 Avhich passes through it in minute and successive quantities ; and this acts 

 in fact as an absorbent of the dissolved gas. Therefore hydrogen and nitro- 

 gen cannot be united by solar influence wlicn perfectly freed from other 

 gases. Eut, in truth, what could the action of the sulphurous gas be ? 



I applied myself to a special study of sulphurous acid, and have ascertained 

 how easy the decomposition of this gas is. As soon as the light passes through 

 it, the white cloud appears ; and if its manifestation is followed with care, it 

 will easily be seen that it is produced not only at the summit of the cones, 

 but likewise on the blue periphery of the first part of the diverging cone, 

 and in the interior of the converging cone. I know of no body more 

 sensible to luminous action ; and the use of condensed light seems hardly 

 necessary for the purpose, since the cloud is formed at other points than the 

 summit of the cones. With this body it is certainly both easy and admirable to 



