306 



Royal Astronomical Society, 



" Not having been able yet to get my tourmalines applied to a 

 telescope, I cannot report on their actual sufficiency for measuring 

 the magnitudes of stars ; the principle, however, seems to answer, 

 inasmuch as a mode of observation is produced, in which the only 

 variable element is angular motion. 



" Numerical measurement of the periodical and secular changes 

 in the colour of stars would seem to be as important as that of their 

 lustre. My plan for doing it cannot be put into action without a 

 clock-moved telescope. The principle is merely this : — a prism is 

 to be introduced into the tube of a telescope so as to make the linear 

 spectrum of a star, instead of the round white image to be viewed by 

 the eye-piece. The mean yellow ray is to be brought by mechani- 

 cal adjustments to a fixed point in the field, and, the spectrum being 

 divided into a number of spaces of certain angular extent, the bright- 

 ness of each is to be measured separately, as in the case of the mag- 

 nitudes of the stars. 



" White stars of all sizes will, of course, show the same relative 

 intensities in the different divisions of the spectrum ; while coloured 

 stars will exhibit a preponderance in the part which answers to their 

 colour. To eliminate the differences of the effects of the atmosphere, 

 and to determine what is white light, a method must be pursued si- 

 milar to that in the discussion of the measurement of magnitudes ; 

 and the colour of the star will ultimately be expressed by numbers, 

 "without attempting to determine by what particular designation such 

 a tint would be conceived by a person of ordinary eyesight. 



2- 

 1- 

 



^1+ 



.2+ 



3+ 



" Let this be a spectrum divided into equal angular spaces of 2" 

 each, the one marked being the yellow ray ; by means of a perfo- 

 rated plate in the focus of the telescope, each of these portions is to 

 be brought singly and successively into view, and have its brightness 

 separately measured as a single star ; the several portions of colour 

 will then be represented thus : — 



" The method is undoubtedly rough, and inapplicable to the smaller 

 magnitudes of stars visible in any telescope, but if a measurement is 



