and Vegetable Colouring -matters by the Spectrum Microscope. 159 



by a variable amount of absorption extending downwards from the 

 extreme blue. Fig. 2 will illustrate my meaning. In the case of 

 one group of colours, the addition of sulphite of soda almost imme- 

 diately removes the detached broad absorption in the lower part of 

 the spectrum, but leaves that at the blue end quite unchanged, or 

 only slightly modified by the solution being made more alkaline. If, 

 then, as in the case of magenta, there is no absorption at the blue 

 end, sulphite of soda makes the solution quite colourless ; whereas if 

 the blues are absorbed, as in the case of the ammoniacal solutions of 

 the colour of red roses and of some species of Dianthus, it changes the 

 colour from green to yellow. If the absorption extends continuously 

 down from the extreme blue to the orange, as often happens when 

 ammonia is added to yellow colours, sulphite of soda produces no 

 change. It is only when there is a more or less perfect division be- 

 tween the upper and lower absorption, that it has any effect ; and 

 then it simply and entirely removes the lower absorption. Some co- 

 lours are changed immediately, even when a very small quantity of 

 sulphite is added ; but others require more, and change gradually, 

 though still very soon. 



16. Groups A, B, and C. 



Colours which are thus altered when the solution is ammoniacal 

 constitute my group A. Frequently, however, sulphite of soda does 

 not remove the detached absorption when excess of ammonia is 

 present, but does so when there is an excess of citric acid. These 

 constitute my group B. As in the other group, any absorption 

 which extends continuously from the extreme blue end is not altered, 

 but the detached absorption in the green is almost immediately re- 

 moved ; and therefore a deep pink or red solution may at once become 

 quite colourless, or only a very pale yellow ; and in many cases this 

 residual colour is due to some yellow colouring-matter mixed with 

 the other. I have never seen a colour which was changed by sul- 

 phite when alkaline, and not when acid ; and thus citric acid never 

 restores the colour when it is added to the changed ammoniacal so- 

 lution. Excess of ammonia usually restores the faded acid solution 

 to nearly the original colour, and it is therefore not a case of actual 

 decomposition, but merely the result of some remarkable molecular 

 change. A third group of colours consists of those which are not 

 almost immediately changed by sulphite of soda, either when alkaline 

 or acid ; and these I call group C. Some of them may fade on keep- 

 ing several hours, and some do not fade even in several days ; but they 

 cannot thus be divided into two definite groups. When thus faded, 

 ammonia does not restore the colour ; and therefore it is evidently 

 the effect of decomposition, and not like the mere molecular change 

 met with in group B. 



On the whole, the groups A, B, and C are remarkably distinct. 

 There are, indeed, a few cases where the change takes place somer 

 what slowly ; and a few scarlet colours do not show very distinctly 

 the characteristic peculiarities of either B or C ; but there are other 

 very strong reasons for believing that some of these are really mix- 



