158 



Capt. W. de W. Abney on 



[Mar. 7, 



be excessively cautious to take the readings under precisely similar cir- 

 cumstances, and also to check them from time to time by the extinc- 

 tion of some standard light. Thus I found that the maximum 

 sensitiveness of the eye was practically attained after it had been 

 excluded from the light for two minutes. Readings taken before that 

 interval had elapsed were invariably too high, while afterwards they 

 were fairly equable. 



The annexed series shows this result. The readings were commenced 

 after exposure of the eye to an ordinary reading lamp, and continued 

 for more than ten minutes* 



3 '16 at the commencement. 



3 '28 at the end of f minute's darkness. 



3-38 at the end of 1± „ 



3-42 at the end of 2& 



3-48 at the end of 3 



3"45 at the end of 5 ,, „ 



3-48 at the end of 10 



A similar form of instrument was made for Sir Charles Wheatstone 

 without his knowledge of the existence of mine ; but I am unaware of 

 the results he obtained. 



In the diagrams which accompany this paper I have given the curve 

 of intensity of the magneto-electric light generated by a machine, 

 similar to that to which reference will subsequently be made, and it 

 will be noticed how close is the agreement between it and the curve 

 which indicates the intensity of the blue light. 



Later it was my endeavour to compare the values of the different 

 coloured components radiated from any source of light, and my 

 original apparatus consisted of a double spectroscopic arrangement 

 by which the lights from the two sources entered by different slits into 

 the spectroscope, and to the eye the spectra appeared vertically one over 

 the other. After screening off all of the two spectra except a narrow 

 strip, and by applying a telescope with a divided lens, as will shortly 

 be described, the intensities of the two strips were equalised by widen- 

 ing or narrowing one or other of the slits. The ratio of the widths of 

 the slits gave approximately the values of the different intensities of 

 the light in the particular part of the spectrum. For light giving a 

 continuous spectrum this method answered admirably, and I believe it 

 was on the same principle that Lord Rayleigh measured the relative 

 values of sky and sunlight ; but for lights whose spectra were discon- 

 tinuous the method failed owing to the labour it involved. This 

 method was then abandoned, and another was sought which should 

 integrate two or three parts of the visible spectrum. 



In my hands Bunsen's method of photometry has always been un- 

 certain, perhaps through some personal obliquity, though I believe that 



