COLOURING MATTEU ASSOCIATED WITH CHLOEOPHYLI,. 17 



of the layers ; nor does he state, that with an increase of the chro- 

 mule additional bands do not appear. I doubt not that the whole 

 perhaps of his yellow chromules — one only excepted — would suddenly 

 reveal also the chlorophyll bands of the hrst half, if my method wei-e 

 applied." 



Now I must say, with reference to nearly the whole of these state- 

 ments, that I cannot comprehend how anyone could have been led to 

 make them if he had read my various published papers, more especially 

 that on comparative vegetable chromatology (Proceed. R. S., 1873, 

 vol. xxi., p. 442). 



It will be seen, on referring to this at p. 444, that I never em- 

 ployed any reagents to separate the colouring matters but neutral 

 alcohol, water, and carbon bisulphide, and was always most careful 

 to check the results by the examination of the spectra of the objects 

 in their natural state. These are quite as definite and distinct as 

 those of their solutions, and, as I showed at pp. 452 to 464, the various 

 substances are in most cases quite as much distinguished by their 

 chemical characters as by their spectra. Thus, for example, taking 

 the two kinds of xanthophyll which have been so commonly con- 

 founded together — my xanthophyll and yellow xanthophyll — when 

 both are dissolved in carbon bisulphide, so that the difference in their 

 spectra cannot be due to diifereuce in physical state, the absorption 

 bands of xanthophyll are situated at about wave-lengths 507 and 

 475 millionths of a millimeter, whereas those of yellow xantho- 

 phyll are situated at about 498 and 467. On dissolving both in 

 absolute alcohol and adding a little potassii nitrite and hydro- 

 chloric acid to both, the xanthophyll rapidly becomes colourless, 

 whereas the yellow xanthophyll is changed into a new splendid blue 

 substance. 



As to my disregarding the influence of the solvents on the spectra, 

 I need only refer to what I said at p. 443. I there considered ** the 

 absorption-band-raising power of solvents " under that special heading, 

 and in the subsequent portions of my paper I treated the subject as 

 though this fact were so thoroughly well understood as not to need 

 any further consideration. 1 always described the difference between 

 the spectra of solutions of the various substances loheii dissolved in the 

 same liquid, and in every other particular in exactly the same physical 

 condition. Moreover, in describing the spectra I relied more especially 

 on the exact position of the bands when the solutions were of such a 

 strength as to show them to the greatest advantage. I am more and 

 more convinced that by far the most important fact is the exact wave- 

 lengths of the centre of the bands when they are seen well-defined and 

 symmetrical ou both sides of the centre. Since in many, if not in most, 

 cases we have to deal with mixtures, the description of the absorption in 

 other particulars is often of no value at all as a character of the indi- 

 vidual substances. Thus, for example, the xanthophylls are almost 

 always found mixed with lichaoxanthine, which gives rise to a very 

 variable intensity of general absorption at the blue end as compared 

 with that due to the bands of the xanthophyll itself, and any de- 

 scription of the entire absorption characters of the xanthophyll obtained 

 from leaves would most certainly be characteristic, not of any one kind 

 of xanthophyll, but of a variable mixture of xanthophyll and lichno- 



c 



