369 
of Edinburgh, Session 1873 - 74 . 
David Brewster increased the limits of the visible spectrum by 
dilating his pupil with Belladonna, so as to increase the amount of 
light. 
From all these considerations, I believe that the diagram here 
given* is not a bad approximation to such a radiation diagram as I 
have described, data for its accurate determination being at present 
unobtainable. The principal Frauenhofer lines are marked below, 
and the numbers along the axis of abscissae represent thirteenth- 
metres .f 
I have drawn particular attention to what I have called the curve 
of limiting visibility, because a consideration of it affords an explan- 
ation of some curious facts which have, from time to time, been 
brought before the Royal Astronomical Society. 
Many fellows of that Society were puzzled by the varying colours 
of stars, and of Jupiter particularly, when observed with different 
telescopes. Mr Huggins suggested that the amount of light, as 
depending upon the magnifying power and aperture of the object 
glass, might be the explanation of it. Mr Browning tested this, 
and found that it afforded a complete explanation. Colonel Strange 
corroborated these views by an independent observation. 
I have tried, in a variety of ways, to produce this result experi- 
mentally, and believe that I have at length succeeded by employing 
gas light, and viewing it through a number of plates of the com- 
mon blue glass coloured with cobalt. This thickness of the glass 
allows only blue and red rays to pass; the boundaries of these bright 
bands in the spectrum being sharply defined. When a piece of 
white paper, illuminated by the gas-flame, is examined with this 
glass, it appears to be blue, but the gas-flame itself appears to be 
red. This is due to no effect of fluorescence. Now, let I r be the 
intensity of the red rays of the flame as seen through the glass, and 
I 6 the intensity of the blue rays. Also letL r be the limiting inten- 
sity for visibility in the red rays, and L 6 in the blue rays. Then, 
I,-L- ( 1 ) 
and 
I* -I* (2) 
* The diagram here referred to is not reproduced. It will probably appear 
in a text-book on Physics now in preparation. — G. F., 1874, April 20. 
+ A thirteenth-metre = 10 — 13 x 1 metre. 
