158 Royal Society : — Mr. H. C. Sorby on Analysis of Animal 



known substance gave rise to similar spectra, with bands at unequal 

 intervals, we ought strongly to suspect either that it was naturally 

 composite, or that some new compound had been formed by decom- 

 position. As a very good illustration I may refer to the product of 

 the action of acids on chlorophyll. The band in the red is not at 

 an equal interval ; but, on careful examination, it is seen to be made 

 up of two bands, the upper of which is at an equal interval. I was 

 not aware that these were due to two different substances, but was led 

 to think it very probable ; and Professor Stokes informs me that he has 

 proved it to be the case. As an illustration of another kind of excep- 

 tion, I refer to the colouring-matter of the pink Stock (Matthiola 

 annua). The aqueous solution shows two bands, whose centres are 

 at about 3f and 5^ ; and on adding ammonia the upper is removed, 

 and the lower depressed to 3J. In spirit of wine they are at 4 and 

 5^, and ammonia developes a third at 3, which are not equal inter- 

 vals. However, if absolute alcohol be used, the bands are at 2f , 4g, 

 and 5^, which are equal intervals ; and thus we see that the abnormal 

 inequality is due to the presence of water, which causes the spec- 

 trum to be as if due to a mixture of two colours, when in reality 

 it is the same colour dissolved in two solvents. 



In spectra showing one absorption-band, there is very commonly a 

 general absorption extending from it towards the blue end ; whereas 

 it so seldom extends towards the red end that it is doubtful if it 

 ever occurs in substances undoubtedly not a mixture of two colours. 

 It can, however, so easily occur in mixed colours, that any substance 

 giving rise to such a spectrum is probably a mixture. Many illus- 

 trations might be given ,* but I will select Brazil-wood, and the same 

 artificially mixed with the colour of beet-root. Adding bicarbonate 

 of ammonia to both, we have — 



* 

 Brazil-wood alone 4 j — 5| ... 7 



Brazil-wood and beet-root 3|...4f — 5| 8 



Here, then, the shading below the absorption-band from 3 J to 4| is 

 evidence of the second colour ; and if such a mixture had occurred 

 naturally its mixed character might easily have been overlooked. I 

 have found many cases similar to this, and had proved that they were 

 mixtures before I was aware that the spectra indicated it. If these 

 very common facts turn out to be general laws, we might thus detect 

 at once the presence of as many as three different substances, or at 

 all events might learn what further examination was desirable. 



1 5 . Sulphite of Soda. 



Sulphite of soda is a most valuable reagent, and its action very 

 remarkable. It enables us to divide colours into three groups, ac- 

 cording as it produces a change in an ammoniacal, in an acid solution, 

 or in neither. The action is related in a very simple manner to the 

 spectra. Having added an excess of ammonia, there may be a well- 

 marked broad absorption over more or less of the red, orange, yellow, 

 and upper green, and above this a clear transparent space, limited 



