1903.] The Cosmical Function of the Green Plant. 435 



starch is only the resultant of two conflicting processes, of its pro- 

 duction and its dissolution, so that at a certain point when the 

 assimilation is not sufficiently intense the former may be counter- 

 balanced by the latter, no surplus starch being stored in the chloroplast. 



This leads us to the discussion of the very important question con- 

 cerning the relative effect of the rays of the more refracted part of the 

 spectrum. It might have been remarked that my researches were 

 restricted to its less refracted half. 



Concerning the other half, the current opinion was that these 

 rays play a comparatively small part in our photo-chemical process. 

 This opinion was considered the result of Draper's classical experi- 

 ments and of a great many experiments with yellow and blue 

 coloured screens. But all these latter experiments are little to be 

 relied upon, since it may be proved by photo-metrical measurements 

 that the intensity of the transmitted light is diminished in very 

 different proportions by these media, and that in consequence the 

 results obtained are not directly comparable. And they are likewise 

 unreliable when the prismatic spectrum is used, because of the great 

 dispersion of these rays, and that was the reason why I confined 

 myself to the less refrangible half of the spectrum and did not push 

 the experiment further into the blue and violet rays. It may be 

 shown, by-the-bye, how capricious and unreliable in this respect are 

 the results obtained by Engelmann's bacterium method. It will suffice 

 to put en regard the results obtained by Engelmann and those of one of 

 his most faithful adherents Professor Pfeffer. In the former case we 

 have a distinct second maximum corresponding to the blue rays in 

 the latter no such maximum could be observed. 



It might seem that the defect inherent in the spectrum method 

 might be easily remedied by a simple calculation, assuming that the 

 effect is proportional to the intensity. But the law which regulates 

 the relation of the process to the intensity of light not being known, 

 this assumption would be gratuitous, and in fadt we shall see later that 

 this law is much more complicated. 



The simplest experimental way of solving the question would, of 

 course, seem to be to make the experiment in a diffraction spectrum. 

 But a series of experiments I made with a beautiful Rowland grating 

 gave only a negative result just as in Mtiller's first experiments the 

 light intensity not being sufficient, no reduction of the carbon dioxide 

 could be directly revealed. In consequence, I was obliged to fall back 

 on the prismatic spectrum. I may, perhaps, be allowed to refer to the 

 fact that even Mr. Langley, notwithstanding the wonderful sensitive- 

 ness of his bolometer, remarks in one of his latest papers, " The prism 

 is, on the whole, far more convenient than the grating." 



To work with the prism, care being taken to avoid the error arising 

 from the differences of dispersion, was in this case simply to adopt the 



