Intelligence and Miscellaneous Articles. 79 



angle, a group of rays of almost the same refrangibilities, and forming 

 a band of feeble but constant magnitude, to investigate how the ca- 

 lorific action of this band varies with its mean refrangibility on the 

 one hand, and with the nature of the source of heat on the other. 



(2) To investigate further how the transmissibility of such rays 

 through a screen of given thickness changes when either their mean 

 refrangibility is varied, or else the nature of the source or that of 

 the absorbent is altered. 



The difficulties experienced in these researches are those always 

 met with in attempting to form, with rays other than the solar rays, 

 pure spectra of an intensity sufficient for calorimetric experiments. 

 I do not dare to affirm that I have completely solved these difficul- 

 ties ; but, at any rate, I think I have succeeded in finding the condi- 

 tions in which the mixture of the rays is so feeble as not to exert an 

 appreciable influence on the result of my experiments. 



To produce these spectra I concentrated the rays from the source 

 of heat on a narrow slit. A lens with a focus of about 1 6 centims. 

 was placed about 30 centims. from the slit, and formed a defined 

 image of it in the conjugate focus. The prism placed behind this 

 lens deflected the rays, and transformed the colourless image into one 

 whose luminous part extended over from 0015 to 0*025 metre, ac- 

 cording to the nature of the prisms used. The thermoscopic pile 

 was linear and very narrow, its aperture being scarcely broader than 

 0-001 metre. 



Under these circumstances the purity of the spectra, and therefore 

 the certainty of the results furnished by analysis, must obviously 

 depend on the breadth of the slit which served as the source of heat. 



The ideal case would be that in which this slit was infinitely nar- 

 row. This cannot be realized ; but in all the experiments whose 

 results I am about to indicate, I found that I could vary the breadth 

 of the slit from - 0005 to 0*001 5 metre (that is, in the proportion 

 of 1 : 3) without at all changing the conclusions to which I was led 

 concerning the distribution of heat in the various parts of the spec- 

 trum, or regarding the absorptions which the consecutive parts of 

 these layers experience in different media. I think I am thence 

 justified in assuming that in my experiments any injurious influence 

 of the mixture of the rays was eliminated. 



I worked with four different sources : — 



(1) A thick platinum wire kept at a red heat in the flame of a 

 Bunsen's burner. 



(2) A bat's-wing burner with the section turned towards the slit . 



(3) An ordinary moderator lamp. 



(4) A Bourbouze lamp. The flame of this lamp is a kind of 

 thimble of very close platinum-wire gauze, kept at a red heat by 

 means of a gas-flame fed by compressed air. 



With the first two sources I used lenses and prisms of rock-salt ; 

 with the two others glass lenses, and prisms of flint glass or of rock- 

 salt. In the experiments in which liourbouze's lamp was used, I 

 modified the radiation by making it pass through a glass trough full 

 of water interposed between the source and the slit. 



It would be impossible to detail all the results of my experiments j 



