SUPPLEMENT 101 



tions. Since in the majority of cases we cannot demonstrate a physical effect 

 of these organs on each other, and since further it has been already shown that 

 NAGELI'S idea of the transference of ' vibrations ' has proved valueless, we can 

 scarcely do otherwise than assume that chemical processes play a great part in 

 correlation. What the nature of the active material is, however, how it arises 

 in the plant, how it moves from place to place and how it disappears, has yet 

 to be settled. Still, note should be made of the antitoxins met with in the study 

 of immunity, substances which are often characteristic of definite species. 

 Certain phenomena in this aspect of plant life, especially self-sterility, have 

 led us already to the belief in special materials of this kind (Josr, 1907). It is 

 probable that we must look for the explanation of correlation in investigations 

 of substances of this nature, though these are at present quite unknown. Just 

 as species may possess special materials, so also may different kinds of cells or 

 tissues. In reality, we have always assumed this, since one has, in the end, 

 attempted to refer all differences between organisms to differences in the proto- 

 plasm. But if substances such as those which are required in correlations 

 come out of individual cells and operate in some way or other at a greater or 

 less distance, they must be diffusible, and cannot be part of the plasma which 

 we know is confined within the cell itself. 



Without following out any further these intentionally wide generalizations, 

 and returning to the mere facts i. e. the cases of correlation empirically deter- 

 mined we have still to note in conclusion that these are demonstrated preferably 

 by what are in reality experimentally induced variations in the typical mode 

 of development of the plant malformations, we might term them but that 

 they also doubtless play an important part in normal ontogenesis. A few 

 examples will make this clear. 



Looking first of all at organogenesis at the growing point, it must be said 

 at once that the nature of this growing point (whether of stem or root) is 

 already fixed in the egg-cell, and it must develop on these lines. At the apex 

 of the stem-leaves appear with free spaces between them. No one can doubt 

 that any part of the merismatic cone is capable of forming a leaf in the same 

 way, and hence if one region grows out into a leaf, that part must obviously 

 inhibit leaf-formation in neighbouring parts ; in other words, we would explain 

 leaf arrangements to a certain extent by assuming correlations in different 

 regions of the growing point. Thus we are led to a subject as to which another 

 theory, very widely accepted, has long held the field, a theory which has a 

 mathematical basis, and hence, apparently, much more perfect than any other 

 botanical conception viz. the mechanical theory of phyllotaxis. 



We are far from suggesting that the relations existing between plant organs 

 are merely chemical in their nature, for reciprocal mechanical influences of 

 organs touching each other are quite possible. Hence we must distinguish 

 between a purely mechanical pressure and a stimulus effect, such as we saw in 

 a previous lecture might result from pressure, contact, &c. 



337, 1. 52, for Since read Thus 



338, 1. 19, for for read from 



1. 38, for do we gain much read do we obtain an exact explanation of the 

 facts 



1. 41, for to correlation, such as the read to correlation. We shall speak of 

 these presently. 



It is quite clear that correlations play an important part in the extension 

 of organs, for otherwise all lateral buds must shoot out, since they are certainly 

 all constructed on the same principle, nodes must develop in length to as great 

 an extent as internodes, or one individual node must alone elongate, &c., &c. 



We reach the same conclusion from a consideration of anatomical features. 



