716 TRANSACTIONS OF SECTION K. 



whole. There is the same likelihood of effective correlations as in other portions 

 of the individual plant in which a new relation is set up. Cases in which a 

 new influence affects the meristematic structure of the plant — as, for instance, 

 in galls or the root-tubercles of the LeguminosEe — seem to be parallel. They 

 exhibit a differentiation of the investing tissue in relation to the enclosed por- 

 tion and a suggestive similarity between the vascular arrangements in the root- 

 tubercles and those in the more bulky ovules. The new construction in the gaWs 

 cannot be regarded as in any sense adaptive or purposive on the part of the 

 plant. We are forced to look at it causally, and perhaps such a mode of 

 regarding the ovule and seed may prove to be the most helpful. On such a 

 view the ovule would be a megaspore gall, later containing an embryo plant, 

 and the problem would concern the system of relations existing between the 

 parts as development proceeds. Corresponding causes might independently 

 result in corresponding constructions and the homologies between various seeds 

 or the contents of various embryo-sacs, though real, be homologies of organisa- 

 tion. This is consistent with the ideas we were led to entertain in the case of 

 the shoot and of alternation of generations. 



As a last example the embryo of the seed-plants may be referred to. <l-'nncrh 

 I must not venture far into the facts. While exhibiting differences, the embryos 

 of various seed-plants present certain common features which contrast with the 

 embryos of all spore plants, even the heterosporous torms. now lin an. ^ -..-.e 

 peculiarities causally connected with the seed-habit and to what extent are they 

 marks of phyletic unity ? We have seen grounds for suspecting that seeds are 

 parallel developments. What follows as to the construction of the embryo if we 

 contemplate such independent origins from ancient Filicales ? The su&pensor 

 has already been considered, and we can regard it as a persistent construction 

 finding a use in some cases. In the embryos of the Filicales we see a very 

 clearly marked type with a single relatively large cotyledon and a hypocotyledo- 

 nary region between the shoot and root. Without entering into the question 

 of whether or not the Monocotyledons were derived from Dicotyledons, it seems 

 clear on broad phyletic grounds that a single cotyledon condition lies behind the 

 dicotyledonous or polycotyledonous condition. There is thus a presumption 

 that there has been a change in the primary members of the embryo in this 

 direction. It seems worth while to emphasise that, from a phyletic point 

 of view, there is a real inconsistency between the origin of seed-plants from 

 the Filicales and the relative primitiveness of dicotyly. Is it not possible that 

 the conditions of early embryonic growth in the more or less cylindrical seed 

 may have led to a more symmetrical construction of the embryo, and that the 

 dicotyledonous or polycotyledonous condition, on the one hand, and the mono- 

 cotyledonous condition with an apparently terminal cotyledon on the other, are 

 two alternative expressions of this?*' For reasons already given, the anatomy 

 would follow the morphological change, and would have to be considered in the 

 light of this, and not as affording safe evidence by itself. 



Conclusion. 



I have touched on a nmnber of large questions, any one of which demanded 

 separate treatment. My concern has not, however, been with them indivi- 

 dually but as cognate problems justifying the deliberate adoption of a causal 

 explanation as the aim of morphological work. I have confined myself to 

 problems bearing on the development and self-construction of the individual 

 and tried to treat them so as to illustrate the causal attitude and possible 

 lines of attack. Preliminary speculations on the questions considered can at 

 best contain a germ of truth, and must be subsequently adjusted in the light 

 of further facts. I have discussed these questions rather than the smaller 

 modifications in individual development shown in metamorphosis, partly because 

 the latter have of late years been treated from a causal point of view -' and 



"■ Recent observations of Coulter and Land on the embryo of Agapanthus 

 appear to afford direct support to this view arrived at by quite a different line of 

 couniarison. — liat. Guz. vol. Ivii. p. 500. 



°^ Of. especially, ' The Fundamental Problems of Present-day Plant Mor- 

 phology,' Science, N.S. vol. xxii. p. 33, and other works by Goebel. 



