668 



CONCLUSION 



This general principle may be illustrated by comparison of certain of the 

 figures quoted in Part II. Thus in Fig. 185 A, i; (p. 348) of Lycopodinm 

 Phlegmaria the apical point T coincides very nearly with the intersection of 

 octants, though it appears unsymmetrical owing to unequal growth caused by 

 the precocity of the cotyledon (c), but it is righted by the appearance later of 

 the second leaf (Figs. 185 c, D). It cannot be doubted that the case of L. 



Selago is very similar, though the 

 detailed study of cleavages is not 

 yet to hand (Fig. 183, p. 346). 

 Fig. 1 86 (p. 349) of L. annoiinnm 

 shows the coincidence of the 

 apex (s) with the cleavage-wall 

 (n) very plainly indeed. As 

 the cleavages have not yet been 

 traced in the more aberrant 

 L. cerniium, it is impossible to 

 say more than that the ob- 

 served facts do not preclude a 

 similar origin of the axis, which 

 comparison with Phylloglossum 

 makes probable (p. 353). In 

 Selaginella spinulosa (Fig. 190, 

 P- 35?) a comparison of the 

 stages A, c, D clearly shows 

 that the small-celled tissue of 

 the apex of the axis includes 

 the intersecting octant walls. A 

 similar origin of the axis to that 

 in Lycopodium and Selaginella 

 s/i//ii//o$a may be traced for 

 fsoe/es, notwithstanding the ab- 

 sence of a suspensor and the 

 small size and late definition of 

 the apex (Fig. 191, pp. 359- 

 360). The case of S. Martcnsii 

 is interesting for comparison, 



Q 



FIG. 357. 



Diagrams to show in view from above and in section how 

 growth with a single three-sided initial cell may lie rstali- 

 'lishcd in an cpihasal hemisphere divided into octants. The 

 quadrant wall, O. Q. and the ,.. taut wall, o, (', are the first 

 of the series of cleavages, continued l>y the walls /, /, //. //, 

 etc. The result is that the initial eell (.r) is formed al lh< 

 nearest possible point to the centre, consistent with ilu 

 sequence of its segmentation. 



interesting 

 since there is a single initial 



cell, a condition which is probably derivative as compared with that of 6". 

 spiimlosn, with its small-celled mcristrm. Pfeffer's drawings 1 demonstrate 

 how this originates with the octant wall forming one of its lateral faces ; 

 in fact, at the nearest point to a central position compatible with its 

 existence as a single initial cell. The embryo of Eqidsetum shows this 

 even more plainly : if a single initial cell is to be carved out of an 

 epibasal hemisphere of four octants so as to be as near to the centre of 



1 Hanstein's Ablnnidl., i., Taf. iii. iv. 



