October 28, 1915] 



NATURE 



245 



forms of various grades of complexity the unicellular 

 condition is also our usual starting-point. What is 

 not so generally recognised or emphasised is the 

 importance of the filament as the primitive construc- 

 tion-form of most plants. I do not use the word 

 primitive in a phyletic sense, nor in the sense of an 

 ideal form, but to indicate a real stage in independent 

 [)rogressions underlying many homologies of organ- 

 isation. I cannot develop this fully here, but wide 



omparison of independent lines of advance suggests 

 ihat the main types of progress in complexity of the 

 plant-body have involved the elaboration of the single 

 tilament with apical growth and with subordinated 

 "branches." It is generally recognised that various 

 -groups of algae show how a solid multicellular axis 

 may come about, not only by the further partition of 

 the segments of the apical cell, but by the congenital 

 rortication of a central filament or the congenital 

 condensation of the subordinated "branches" on to 

 the central axis. The algae further show the change 

 from the dome-shaped apical cell of a filament to the 

 sunken initial cell with two, three, or four sides. The 

 central filament then only appears, if at all, as a sub- 

 sequent differentiation in the tissue, and the segments 

 seriallv cut off from the apical cell may or may not 

 bear projecting hair-shoots or "leaves." The algae 

 thus attain in independent lines a construction corre- 

 sponding to that of the plant in liverworts and 

 mosses. In the various parallel series of Bryophyta 

 the filament is not only more or less evident in the 

 ontogeny, but may be regarded as the form under- 

 lying both thallus and shoot, between which on this 

 view there is no fundamental distinction. The sporo- 

 gonium also can be readily regarded as an elaborated 

 tilament. While the same interpretation of the fern- 

 prothallus will readily be granted, to think of the 

 tern-plant as the equivalent of an elaborated filament 

 may appear far-fetched. So far from this being the 

 case, I believe that it will be found helpful in under- 

 standing the essential morphology of the shoot. In 

 a number of vascular 'cryptogams and seed-plants, 

 there is actually a filamentous juvenile stage, the 

 suspensor, while the growth by a single apical cell 

 is essentially the same in the fern as in the moss and 

 some algae. 



There follows from this a natural explanation of 

 the growth by a single initial cell so commonly found 

 in plants. The apical cell appears to be the one part 

 of the massive plant-body (for instance, of Laurencia, 

 a moss, or a fern) that persists as a filament ; it is a 

 filament one cell long. It may be replaced by a group 

 of initial cells, as we see in some algee, liverworts, 

 and Pteridophyta, and this leads naturally to the small- 

 celled meristems found in most Gymnosperms and 

 Angiosperms. The filamentous condition is then 

 wholly lost, though the system of relations and espe- 

 cially the polarity is maintained throughout all the 

 changes in the apical meristem. 



I feel confirmed in regarding the construction of the 

 sporophyte in this fashion by the fact that it fits 

 naturally with the conclusions resulting from the 

 masterly comparative treatment of the embryology of 

 the vascular cryptogams by Prof. Bower. These are 

 (i) the primary importance of the longitudinal axis of 

 the shoot, the position of the first root and the foot 

 being variable ; (2) the constancy of the position of 

 the stem-apex near the centre of the epibasal half of 

 the embryo ; (3) the probability that embryos without 

 suspensors have been derived from forms with 

 suspensors, without any example of the converse 

 change. These and other related facts seem to find 

 their morphological explanation in the shoot of the 



sporophyte being the result of the elaboration of a 

 filament. 



NO. 2400, VOL. 96] 



The Coiisiruction of the Shoot. 

 The view to which we arc thus led is that the uni- 

 axial shoot is a complex whole, equivalent to the axial 

 filament together with its congenitally associated sub- 

 ordinated "branches." This applies to the multi- 

 cellular plant-bodies found in various independent lines 

 of algai and Bryophyta, whether they have definite 

 projiecting appendages of the nature of leaves or not. 

 The discarding of the distinction between thallus and 

 shoot, which in practice has proved an unsatisfactory 

 one, is no great loss. Even taking the word in the 

 narrower sense of a stem with distinct leaves, the 

 shoots in algae, liverworts, and mosses, though ad- 

 mittedly independent developments, exhibit an essential 

 correspondence amounting to a homology of organisa- 

 tion. The resemblances are not analogies, for it is 

 doubtful whether the "leaves" in the different cases 

 correspond in function. The comparison of the shoot 

 of the sporophyte of a vascular cryptogam with, for 

 example, the shoot of the moss seems equally justi- 

 fiable. It is only forbidden by strict phyletic morpho- 

 logy, which for our purpose has no jurisdiction. The 

 general agreement as regards the leaf-arrangement 

 between the ferns and the Bryophyta suggests that 

 similar laws will be found to hold in the shoot of 

 both gametophyte and sporophyte. Apart from plagio- 

 tropic shoots, there is a constructionally dorsiventral 

 type of fern-rhizome. The leaves of this alternate as 

 in the leafy liverworts, while the radial type of fern 

 corresponds to the moss-shoot. It is significant that 

 the earlv leaves of radially constructed ferns usually 

 exhibit a divergence of | in the seedling, passing 

 higher up the stem into more complicated arrange- 

 ments, and the same is the case in mosses. I must 

 not enter into questions of phyllotaxy, but may remark 

 on the hopefulness of attacking it from the study of 

 the simpler shoots of algae and Bryophyta rather 

 than, as has usually been done, beginning with the 

 flowering plants. 



In some ferns (the striking example being Cerato- 

 pteris) the relation between the segmentation of the 

 apical cell and leaf-production is as definite as in the 

 moss, each segment giving rise to a leaf. This may 

 hold more widely for ferns than is at present 

 demonstrated, and the question deserves thorough re- 

 investigation to ascertain the facts independently of 

 any theoretical views. That the coincidence of the 

 segmentation of the shoot expressed by the leaf- 

 arrangement and the segmentation of an apical cell 

 is not a necessary one is, however, clearly shown in 

 other ferns, and is obvious in the case of shoots with 

 a small-celled meristem. The two segmentations 

 appear to be determined by some deeper system of 

 relations, which may also be manifested in a coeno- 

 cytic plant-body. 



In the complication of the uniaxial shoot introduced 

 by branching also there seems to be an advantage in 

 a' wide area of comparison. The question most often 

 discussed concerns dichotomous and monopodial 

 branching. If the details of development are to be 

 taken into consideration, the term " dichotomy " has 

 usually been very loosely applied. Apparent dicho- 

 tomy, the continuation of one shoot by two equally 

 strong ones, is fairly common. But in most cases 

 investigated in detail the branching seems to be really 

 monopodial and the forking due to the equally strong 

 development of a lateral branch close to the main 

 apex, not to the division of the latter. In plants 

 growing by a single initial cell almost the only case 

 of strict dichotomv known is the classic one of Dic- 

 tyota. The branching of the ferns has been ^ the 

 subject of numerous investigations, but there is a 

 great lack of developmental data. Usually the 

 branches stand in some definite relation to the leaves 



