474 PROCEEDINGS OF THE AMERICAN ACADEMY 



telescope as a camera lens, results were obtained which compared 

 very favorably witli those reached with the larger instrument. The 

 prism was composed of dense yellow flint glass, and experiments 

 with specimens of ordinary Hint and crown glass 7 cm. in thickness 

 gave no more absorption in the infra-red spectrum, as far as observed, 

 tlian they would in the visible portion ; which is entirely contrary 

 to the general belief. Au absorptive medium was placed iu front of 

 the slit, in order to destroy all light save tiiat of the wave length 

 which it was desired to photograph. This precaution is necessary, 

 as, owing to the reflections from the surfaces of the lenses and prism, 

 a certain amount of diffused light finds its way to the plate, together 

 with the spectrum, and should tliis diffused light be of short wave- 

 length, it would fog the plate and the spectrum image would be de- 

 stroyed. It is also necessary to coat the back of the plate with black 

 varnish, iu order to prevent the formation of a halo, owing to reflec- 

 tions from the back of the plate. The absorptive medium consisted 

 either of two pieces of red copper glass, or of a piece of red and a 

 piece of blue glass, or of a thin layer of asphalt varnish on glass, of 

 such density as to be slightly lighter than the combined red and blue 

 glass. These all gave about equally good results, with possibly a 

 slight advantage iu favor of the as})lialt. Other media were experi- 

 mented upon, including red glass (single, triple, and quadruple), 

 iodme dissolved in carbon bisulphide, lampblack, and hard rubber. 



The iodine solution (see diagram) transmitted a large quantity of 

 light in the vicinity of the II lines, — so much, in fact, as to reverse 

 the spectrum in that region. The lampblack showed a slight broad 

 absorption baud between F and G, with a maximum at G. Otherwise 

 the spectrum was quite uniform between A and H, and faded away 

 at the two ends, disappearing at wave-lengths .37 and .94 micron. 

 If one wished for a photograph of the visible spectrum only, it woidd 

 seem as if it might be obtained very satisfactorily by merely inserting 

 a piece of smoked glass in front of the slit. The glass should be 

 smoked until it is about as dark as two pieces of ordinary red dnd 

 blue glass placed together appear when viewed by transmitted light. 



The bai-d rubber spectrum was obtained with a piece of rubber 

 about .02o cm. in thickness. One could readily see the sun through 

 it, and by close examination detect the window bars when no light 

 came over the shoulder. In structure it was not transparent, but 

 translucent like porcelain, and filled with little irregularities, con- 

 sisting of short, narrow, opaque lines, lying in the direction in which 

 the sheet had been rolled. Ou placing it in front of the slit, its spec- 



