and Vegetable Colouring -matters by the Spectrum Microscope. 1 63 



neutral aqueous solutions exhibit 0, 1, 2, 3, &c. decided absorption- 

 bands. Sometimes one of them may be so obscure that we may hesi- 

 tate whether it should be counted or not ; but practically this is no 

 very serious objection, if we decide to reckon only distinct bands, 

 and to look on the fainter as important merely in identifying indivi- 

 dual colours. If no absorption-band can be seen in the neutral 

 solution, we may take into account those seen when more or less 

 ammonia is added ; and if none occur in either case, we may make 

 use of those seen in the alcoholic solution when neutral, and after the 

 addition of ammonia. Whenever in this order of experiments the 

 solution gives any decided absorption-band, the subgroup is deter- 

 mined ; and it is only when none has been produced that the process 

 must be carried further. 



The general connexion of the subgroups will be best seen from 

 the following Table : — r am 



( al i am, 



am~ '< aL am,, &c. 



2' 



aq < am 1 [ al 2 , &c. 



1, A <j aq x [ am 2 , &c- 



( aq 2 , &c. 



The same system is applicable to each division, I, 2, and 3, and to 

 each group, A, B, and C. We can easily express the subgroups by 

 using one or more of the signs aq, am, al, am, with a figure to indi- 

 cate the number of bands in the first term that contains any, those 

 before it being given to show the facts more clearly. 



Each colour can be indicated by writing after the subgroup the 

 characteristic spectrum, or, for the sake of simplicity, merely the 

 position of the centres of the bands when they are seen as inde- 

 pendent as possible of general absorption. If the centres of the bands 

 are in different positions, the colours cannot be the same ; but if they 

 agree, it does not necessarily follow that they are the same : it is 

 probable, hut must be further proved by the correspondence of other 

 spectra. As examples I give a few well-marked cases. 



Purple Pansy 1, A, aq am x (4). 



Crimson Rose 1, A, aq a?n al am x (2|). 



Blue Lobelia (L. speciosa) 1, B, aq 2 (2|, A\). 



Pink Stock (Matthiola annua) 1, B, aq 2 (3§, 5|). 



Several blue species of Campanula 1, B, aq± (2§, 4, 5§, 7|). 



Brazil-wood (Ccesa/pinia crista) 1, C, aq x (5^). 



Logwood (Hcematoxylum campechianum) 1, C, aq x (4|). 

 Sandalwood {Pterocarpus santalinus) . . 3, C, aq 2 (6, 7%)- 

 Alkanet-root {Anchusa tinctoria) 3, C, aq 3 (4 J, 5f, 7|). 



21. Individual Colows. 

 Having then ascertained to which subgroup any particular colour 

 belongs, it is in the next place requisite to determine what particular 

 substance it is. When it gives rise to well-marked absorption-bands, 

 this may be more or less definitely decided by their exact position 

 and character — since they may of course occur in different situations, 

 or vary much in absolute and relative breadth and in intensity. 

 Thus, choosing closely related spectra, we have, for example : — 



M2 



