CELESTIAL CHEMISTRY. 235 



retain the image of a star for any length of time exactly withia 

 the jaws of a slit only the l-300th of an inch apart. By patient 

 perseverance these ditiiculties have been ovei'come, and satis- 

 factory results obtained. We considered that the trustworthi- 

 ness of our results must rest chiefly upon direct and simultaneous 

 comparison of terrestrial spectra with those of celestial objects. 

 For this purpose we contrived the apparatus which is represented 

 in the diagram. 



" By this outer tube the instrument is adapted to the eye-end of 

 the telescope, and is carried round with it by the clock motion. 

 Within this outer tube a second tube slides, carrying a cylindrical 

 lens. This lens is for the purpose of elongating the I'ound point- 

 like image of the star into a short line of light, which is made \o 

 fall exactly within the jaws of a nearly-closed slit. Behind the 

 slit, an achromatic lens (and at the distance of its own focal 

 length) causes the pencils to emerge parallel. They then pass 

 into two prisms of dense flint glass. The spectrum which results 

 from the decomposition of the light by the prisms is Anewed 

 through a small achromatic telescope. This telescope is provided 

 with a micrometer screw, by tvhich the lines of the spectra may 

 be measured. 



" The light of the terresti'ial substances, which are to be com- 

 pared with the stellar spectra, is admitted into the instrument ia 

 the following manner : — 



" Over one-half of the slit is fixed a small prism, which receives 

 the light reflected into it by the movable mirror placed over the 

 tube. The mirror faces a clamp of ebonite, provided with for- 

 ceps to contain fragments of the metals employed. These metals 

 are rendered luminous in the state of gas by the intense heat of 

 the sparks from a powerful induction coil. The light from the 

 spark, reflected into the instrument by means of the min-or and 

 the little prism, passes on to the prisms in comi3any with that from 

 the star. In the small telescope, the two spectra are viewed in 

 juxtaposition, so, that the coincidence and relative positions of 

 the bright lines in the spectrum of the spark with dark lines in 

 the spectrum of the star, can be accurately determined. 



" Moon and Planets. — I now pass to the results of our observa- 

 tions. 



" I i-efer, in a few words only, to the nioon and planets. These 

 objects, unlike the stars and nebula3, are not original sources of 

 light. Since they shine by reflecting the sun's light, their spectra 

 resemble the solar specti'um ; and the only indications in their 

 spectra which may become sources of knowledge to us ai-e con- 

 fined to any modifications which the solar light may have sufl'ered, 

 either in the atmospheres of the planets, or by reflection at their 

 surfiices. 



"Moon. — On the moon, the results of our observations have 

 been negative. The spectra of the various parts of the moon's 

 surface, when examined under different conditions of illumina- 

 tion, showed no indication of an atmosphere about the moon. I 

 also watched the spectrum of a star, as the dark edge of the moon 

 advanced towards the star, and then occulted it. No signs of a 

 lunar atmosphere presented themselves. 



