358 
changed. The spectrum of transmitted light shows two ab- 
sorption-bands situated very, considerably further from the, 
blue end than before, and the light of fluorescence is yellow- 
green, giving the spectrum No. 5 with a bright band nearly 
in the centre of the green and a fainter between the green 
and yellow. This change takes place much more slowly in 
the case of the solution in ether, but much more rapidly. 
when crushed insects are exposed to the air, and a third, 
compound is formed, which may be obtained in a very satis- 
factory manner by digesting dead insects, kept dry for some 
weeks, in a solution of bisulphide of carbon in alcohol, and. 
after it has remained for a few days agitating the clear solu- 
tion with excess of the bisulphide. This sinks to the bottom 
with the greater part of the required substance, and leaves 
various impurities dissolved in the alcohol. After washing 
with more alcohol, the solution in bisulphide when evapo- 
rated leaves an oily or waxy substance coloured brown orange. 
When dissolved in bisulphide of carbon this gives most 
remarkable spectra. The transmitted light is of an orange-red 
colour, giving five well-marked absorption-bands, one in the, 
orange, dark, narrow, and well defined; one at the yellow 
end of the green, very dark and well defined, with some 
general shading on the green side; a third and a fourth, less, 
dark than the above two, one nearly in the centre of the 
green and the other at the green end of the blue, whilst the 
fifth is nearly in its centre. This spectrum is not only re- 
markable for the number of bands thus spread over so large’ 
a space, but also for the manner in which they are related to 
one another. This is much like what might be due to a 
mixture of two substances, and yet there is no further eyi- 
dence of its being so.1 The solution is strongly fluorescent, 
the light of fluorescence is orange-coloured, and its spectrum 
is as shown by No. 6. The yellow, gréen, and blue are 
entirely absent ; there is a red band, but it is comparatively 
so faint that the light may be said to be nearly monochro- 
matic, being almost entirely due to the well-defined orange 
band shown by the figure, which is so narrow that it is only 
about =th part of the whole visible spectrum of daylight. 
As will be seen, it is quite on the red side of the sodium 
line D, but when the substance is dissolved in ether instead 
of bisulphide of carbon, the centre of the bright band almost 
exactly coincides with D, and all the various bands in the; 
other spectra already described are raised to about the same: 
1 See my late paper, “ On the Examination of Mixed Colouring Matters,” 
‘Monthly Micros. Journal,’ vol. vi, pp. 124—134. 
