68 PLANT MORPHOLOGY 



phytes. But objection may be taken to all these lines of evidence. 

 We should hardly look to either the embryos of seed-plants nor to 

 their inflorescences for safe guidance as to the origin of the funda- 

 mental characters of shoot-construction, for both are probably 

 highly specialized forms of shoot. Particularly would this seem to be 

 the case in the embryo, which is nursed with a supply of endosperm 

 within the seed, a condition far removed from what can possibly 

 be conceived as that of a primitive leafy shoot. Moreover, the fact 

 that certain monocotyledon embryos conform externally to such a 

 theoretical description as is given is not sufficiently cogent in the 

 absence of internal limits of demarkation of the constituent shoot- 

 segments. 



The details of comparison of the moss-plant and protonema are 

 quite beside our question, which relates to vascular plants: how- 

 ever interesting the analogies may be between the alternating gen- 

 erations, they cannot rank as evidence in such a question as this: 

 for it is quite conceivable that a perfect system of shoot-segmenta- 

 tion might rule in the one generation, while the leafy development 

 in the other, having originated by a distinct evolutionary sequence, 

 might show a quite distinct relation of leaf to axis. 



The last line of evidence is from segmentation at the apex of the 

 shoot in pteridophytes: if one cell-segment regularly produced one 

 leaf-bearing shoot-segment, this might be held to be valid evidence 

 of Celakovsky 's view. But this argument does not apply consistently, 

 as indeed Celakovsky himself admits. It is true that a leaf may be 

 produced from each apical segment in some ferns: but in dorsi- 

 ventral ferns, and hydropterids, leaves are not produced from the 

 ventral rows (Klein, Bot. Zeit., 1884). In Azolla and Salvinia leafless 

 internodes intervene between successive nodes: thus there is no 

 constant relation in ferns between apical segmentation and 'leaf- 

 production. Treub's investigations resulted in his statement that 

 " there cannot be any constant relation between the leaves and the 

 segments of the apical cells in Selaginella Martensii." Lastly, in 

 Equisetum, notwithstanding the regular segmentation of the tetra- 

 hedral apical cell, the leaves show no regular relation to the segments 

 in number or position, varying in number from three to about forty. 

 Thus the argument from apical segmentation even in pteridophytes 

 does not give consistent support to a theory of shoot-segmentation, 

 while such evidence is entirely wanting in the vast majority of 

 phanerogamic plants. Notwithstanding the ingenuity of the theory 

 as put forward by Celakovsky, in the absence of any structural indi- 

 cation of the limits of the shoot-segments in the vast majority of 

 cases, the theory does not appear to me to be sufficiently upheld by 

 the facts. 



An extreme, and indeed, a paradoxical position has been taken 



