806 Captain Ahney [Feb. 25, 



case by bare glass. Bebind tbe " red " pbotograpb we place a red 

 medium, sucb as red glass, wbicb occupies but a small part of the 

 spectrum and is equivalent to the red sensation, and behind the 

 green photograph a green medium, also taking in but a small part of 

 the spectrum, and behind the blue photograph a blue medium, and if 

 the luminosity of the mixed light coming through the red, the green, 

 and the blue when unshaded hy the positives forms white, we shall have 

 a representation of the spectrum. 



Suppose that we were going to reproduce the image of the elec- 

 tric light carbons by combining three distinct photographs together 

 backed by proper media, and that we wished to know what each trans- 

 parency would look like when illuminated with its proper colour, we 

 can show this in a fairly simple manner. Close in front of the slit of 

 the spectroscope is a lens of such a focus that a sharp image of the car- 

 bon points is thrown on the surface of the j^risin. The prism analyses 

 the colours, and a lens in front of the spectrum collects the coloured rays 

 again and gives us an image on the screen of the carbon points. Placing 

 three slits in the spectrum, we alter their width until the image again 

 appears white at the brightest part. We may substitute three lenses 

 of equal foci for the single lens, and we have three images side by 

 side, which, as just seen when combined together, will give the white 

 image of the crater and the redder image of those parts where 

 the heat is less intense. We can vary this experiment. If we place 

 against the prism a small square made up of circular glasses of dif- 

 ferent colours, we have the image of the glasses on the screen when 

 the whole spectrum is used. With the slits inserted as before, we 

 also get white light and the colours of the glasses (Fig. 4). The 

 three lenses, also placed before the slits, give the separate images 

 such as we wish to obtain by photographic means. 



But how about securing these photographs ? Can we find three 

 different photographic plates which will be exactly sensitive to the 

 required parts of the spectrum, excluding all other parts ? 



It will be seen that the parts overlap (see Fig. 1). Thus the 

 green and red curves overlap, as do also the green and the blue. 

 It may at once be stated that there are no such different kinds of 

 plates to be found. But if we can find one plate which is sensitive 

 to the whole spectrum, we can, by using absorbing media, cutoff those 

 portions which are required. Now the ordinary plate, with short 

 exposure, is not sensitive much beyond the blue (see No. 4, Fig. 5), 

 but if we give it a slightly longer exposure it will tJfe found sensitive 

 to the green and yellow as well, and, with a still further exposure, to 

 the extreme red ; so that we can use an ordinary plate for the purj)Ose, 

 as it is sensitive, but in vastly different degrees, to the whole spectrum, 

 but we have to cut off all the parts we do not want. 



In the spectrum of the light transmitted by an orange glass in the 

 spectrum, we see that the red, yellow and green alone penetrate, and 

 this is the region of the spectrum that the red sensation curve occu- 

 pies. A blue-green glass cuts off most of the red and the violet, and 



