128 On the Examination of 



We might conclude that very probably two substances were 

 present, one giving the bands A and B, and the other C and D. 

 Similar principles would of course apply to cases where more or 

 fewer bands were present. Very often one colouring matter might 

 give well-marked bands, and the other only a general absorption. 

 We might then separate it with two products, and on making the 

 solution of these of the same general depth of colour, one might 

 show a spectrum with well-marked bands, and the other scarcely 

 any trace of them. The same general facts may be seen when 

 the characteristic bands are only developed by the addition of 

 different reagents, which completely alter the colouring matter, but 

 then more care is required to avoid any differences that may result 

 from the varying action of more or less of the reagents. To attempt 

 to describe all the necessary precautions would make this paper 

 most um-easonably long, and therefore I trust that the above rough 

 and general account will at all events suffice to point out the sort of 

 plan that may be followed, premising that in all cases the results 

 should be checked by other experiments, in order that no accidental 

 peculiarities may lead to error. As illustrations of the appHcation 

 of these principles I would refer to those colouring matters of Mno, 

 and of Gummi rubra, which are soluble both in ether and in water. 

 The kino contains two such, one more soluble in ether, of pink 

 colour, which in its natm'al state gives a well-marked absorption- 

 band between the yellow and green, and the other, more soluble in 

 water, of orange colour, which gives a broad absorption-band between 

 the green and blue on the addition of bicarbonate of ammonia. 

 Gummi riibra contains this same orange colour, without the 

 pink, mixed with another orange or yellow substance, relatively 

 more soluble in water, not yielding a spectrum with decided bands 

 in any state that I have yet seen. 



When a solution contains more than two colouring matters the 

 recognition of each becomes somewhat more difficult ; but still, 

 by following out this system, and dividing the material into more 

 than two portions, a very good opinion may be often formed of the 

 general optical properties of each substance. When some of them 

 give well-marked and characteristic absoiption-bands, and when the 

 absorption of others may be removed by the addition of sulphite 

 of soda, the study of a complex mixture is very greatly facili- 

 tated, and especially if the spectrum of one or more of the con- 

 stituents is of such a marked character that it can be at once recog- 

 nized as that of some substance already known in a pure state. A 

 tolerably good opinion may then be formed of the spectrum of the 

 rest, by, as it were, subtracting that of the known constituent. This 

 leads me to the description of the spectra of certain colouring matters 

 which are met with so far separated naturally that their compound 



