CORRELATIONS 327 



each other, inasmuch as the performance of a definite function in one organ is 

 essential to the carrying out of other functions in other organs, although these 

 organs may be quite capable in themselves of performing such functions. In 

 normal ontogeny each functional member assumes a definite shape, and we 

 might readily suppose that it could not assume any other form. As a matter 

 of fact, however, we should be more correct in asserting that every organ 

 arising at the growing point may develop in a variety of ways ; the fact that 

 it is forced to assume a definite direction of development depends only on its 

 relations to other parts. Did no such regulation of the development of 

 parts exist and if every cell or every mass of embryonic tissue gave rise to such 

 structure as it was inherently capable of producing, the resulting plant would 

 be no longer an organism but merely a chaotic mass of living substance. 

 ' Harmonious development ' is possible only if correlations exist, and to gain 

 some acquaintance with examples of these correlations must be our first task, 

 thus giving us the opportunity of considering a whole series of phenomena 

 which we have hitherto only imperfectly studied. In only a few cases is it 

 possible to draw conclusions as to correlation from observation only. Berthold 

 (1882) has found that the lateral branches of many Algae may give rise to new 

 outgrowths at their bases on the convex side turned away from the chief axis, • 

 the effect of the chief shoot being to cause the normally radial branches to become 

 locally dorsiventral. Similar observations have been made on twigs of Cu- 

 pressiis ; the lateral branches bear more numerous and larger proliferations on 

 the side towards the apex, although their disposition should be basiscopic. 

 Occasional anomalies may also be observed. Thus De Vries (1891) records 

 the case of a flower-stalk of Pelargonium which formed a leaf-bud, and which, 

 instead of dying off after the bud had opened, remained in existence for several 

 years, and by vigorous secondary growth formed a woody cylinder like an 

 ordinary stem. As a general rule the flower-stalk forms no cambium nor has it 

 any capacity for forming merismatic tissue ; it remains bare and no foliage- 

 leaves arise on it ; in other words, a correlation exists between the development 

 of foliage-leaves and the power of secondary growth in the stem. Similar 

 examples worthy of notice often occur among malformations. Experimental 

 studies on this subject are, however, of more importance and more pertinent. 



From among such experimental researches we may select those which aim 

 at observing the effect on the rest of a plant of the removal of one of its organs. 

 This experiment necessitates, in the first instance, the formation of a wound 

 which the plant endeavours at once to cicatrize [Massart, 1898]. In the second 

 place, the plant exhibits a capacity for replacing the organ which it has lost. We 

 have here to deal with a reaction on the part of the plant to external mechanical 

 stimuli, such as we have described in the last lecture, but these processes 

 become intelligible here for the first time in view of the light shed on them by 

 a study of correlations. 



First of all a few words may be said as to the healing of wounds, and we 

 shall take our examples from the higher plants exclusively. The healing takes 

 place very differently according to the age and specific character of the tissue 

 affected. The injured cells and also the other cells adjacent to them die off, but 

 the layer of cells next below — if they still contain protoplasm — begin to react 

 so as to develop tissue over the wound. Many of the parenchymatous cells, 

 without actually growing, divide parallel to the surface of the wound, and their 

 walls become cutinized and thus interpolate a corky layer between the dead 

 and the living tissue. This reaction teaches us that fully-grown cells may still 

 retain the power of division, but we must not regard the destruction of the 

 connexion with neighbouring cells as the only factor inducing this reaction, for 

 materials from the injured cells, or other changes taking place during the inflic- 

 tion of the injury, might act as a stimulus. 



