PROGRESS or MICROSCOPICAL SCIENCE. 19 



with one, and in another instance with another colouring-matter, 

 whilst in other cases he attributes certain variable spectra to the 

 same single substance, modified by some unknowoi and, as I believe, 

 altogether imaginary cause, whereas it can easily be shown that all 

 these variations are due to variable mixtures of substances well known 

 in an approximately pure state, having perfectly definite and constant 

 properties. For example, he describes and figures the spectra of the 

 blue-green colouring-matters from Deiitzia scahra, and from Oscilla- 

 toria, and shows that they diifer in constant and important particulars. 

 He hence concludes that they are two distinct substances, without any 

 simple and definite connection, whereas by employing the methods I 

 have alluded to it may be most conclusively proved that what I have 

 called blue chlorophyll is the principal constituent of both, but that 

 it is mixed in the one case with a substance found in all green leaves, 

 though not in Oscillatorm, and in the other case with another, occur- 

 ring in large quantity in OsciUaforice grown in bright light, though 

 in relatively very small quantity in green leaves, and even in Oscilla- 

 torice when grown in a very shady jjlace. It is chiefly in the case of 

 the colouring-matters belonging to what I have called the xanthophyll 

 group that the author attributes the variation in the specti'a to some 

 unknown cause, and I look upon it as very important to be able to 

 show that this variation is simply due to a variable mixture of two or 

 more substances, for it would lead to loose and inaccurate observations 

 if we were to suppose that the optical characters of any separate com- 

 pound could vary when dissolved in the same liquid. On the contrary, 

 now that it can be shown by various methods that the solutions giving 

 these variable sj^ectra are mixtures of substances that can often be 

 sei)arated, and that the results can be easily imitated by artificial 

 mixtures, there is no kind of reason for supposing that in like circum- 

 stances the optical characters of any of the separate constituents are 

 variable. Not only is it imj)ortant to establish this fact, but by dis- 

 tinguishing the different constituents, and determining their relative 

 quantity in different cases, we have a perfectly simple and intelligible 

 method of comparison, which otherwise would not be the case. The 

 value of such principles in studying comjiarative vegetable chromato- 

 logy will be seen at once, since it enables us to understand the exact 

 connection and difference between the coloured constituents of dif- 

 ferent classes of plants. With such exceptions as these, which are to 

 be attributed in great measure to the application of new methods of 

 research, I must express my high opinion of the merits of the work, 

 and I trust that its publication will be the means of leading to the 

 more comi^lete and accurate study of a branch of research which will 

 probably yield most valuable results ; only, as I believe, these will be 

 derived, not from the discovery of rare and exceptional colouring- 

 matters, but from the careful and accurate qualitative and compara- 

 tive quantitative analysis of complicated mixtures of the most common 

 and fundamental, which may not have attractive properties, but yet 

 probably play an important part in the economy of particular classes 

 of plants. When we thus study the subject, and do not ignore what 

 might be looked upon as insignificant details, it seems possible to 



