292 Mr. E. W. B. Gill on the Electrical 



quickly to a second maximum higher than the first at 300 fi/i r 

 and remained at that value without any decrease till 210 fifi, 

 when a slight fall occurred to 200 jjl/jl, where the observations 

 terminated. Professor Dreyer*, however, made the inter- 

 esting observation that by sensitizing the bacteria by erythrosin 

 they could be made as sensitive to long as to short waves, in 

 much the same way as a photographic plate can be made 

 responsive to different rays, 



A little consideration v^ill show that there is not of 

 necessity any intimate connexion between the photographic 

 effect, the electrical effect, and the energy of any given 

 portion of the spectrum. The photographic effect may be 

 roughly estimated by the time of exposure required to photo- 

 graph any part of the spectrum. The visible spectrum has 

 a large effect upon a photographic plate, but as the wave- 

 length gets shorter the effect gets less and less, and even if 

 special plates (without gelatine, which Schumann found 

 absorbed the shorter waves) be used very great difficulty is 

 found in photographing as far as A =185 /xyit, using a quartz 

 prism and lenses. The electrical effect, on the other hand, 

 is practically zero in the visible spectrum, the i^s which 

 set free negative ions being almost entirely cut off by glass, 

 and, as will be shown later, the maximum electrical effect in 

 the spectrum of a spark- discharge does not coincide with the 

 rays which give the maximum photographic effect. In fact 

 the electrical method for detecting the light is much more 

 sensitive than the photographic for short waves. 



The distribution of the energy also is independent of the 

 intensities of either the photographic or the electrical effects; 

 thus the energy extends over the whole spectrum, and the 

 electrical effect is confined to the ultra-violet; but it is a 

 remarkable fact that Pfluger's energy curve reaches a 

 maximum for the same rays as the curve giving the intensity 

 of the electrical effect. The electrical curve, as will be seen 

 later, has a second maximum, but unfortunately Pfliiger's 

 results do not reach as far as this. 



In the electrical research to be described below the spectrum 

 was formed by a quartz prism, and it may be useful to sum- 

 marize the chief results obtained so far for the spectrum 

 produced in this manner. Rubens' curve (fig. 1) show r s that 

 for a small range of spectrum dn (n here and afterwards 

 denoting refractive index) the corresponding dX in the 

 visible spectrum is large, but gets less as n increases, and at 



* " Sensibilisering af Mikroorganisnier og dyriske Yoeo," Meddd. fra 

 Finsens medic. Lysinstitut, vii. 1903, p. 110. 



