xm LYCOPODINEJE 509 



of the plant is very variable in bulk. ... In the large as well as 

 the small specimens a single initial is usually present, but its seg- 

 mentation does not appear to be strictly regular, and it is diffi- 

 cult to refer the whole meristem to the activity of one parent 

 cell. . . . When a leaf or sporangiophore is about to be formed, 

 certain of the superficial cells increase in size, and undergo both 

 periclinal and anticlinal divisions so as to form a massive out- 

 growth, the summit of which is occupied, as seen in radial sec- 

 tion, by a single larger cell of a wedge-like or prismatic form. . . . 

 In these early stages I find it impossible to say whether the part 

 in question will be a vegetative leaf or a sporangiophore, and 

 even when older it is still a matter of uncertainty. . . . Those 

 which are to develop as sporangiophores soon show an increase 

 in thickness, while they grow less in length ; an excrescence of 

 the adaxial surface soon becomes apparent (Fig. 294, A, sy), 

 in which the superficial cells are chiefly involved. . . . The super- 

 ficial cells at first form a rather regular series, which may be 

 compared with the cells which give rise to the sporangia in Lyco- 

 podium clavatum, or in Isoetes: they undergo more or less regu- 

 lar divisions, which, however, I have been unable to. follow in 

 detail : a band of tissue some four or more layers in depth is thus 

 produced. About this period certain masses of cells assume 

 the characters' of a sporogenous tissue: but though they can be 

 recognised as such by the character of the cells, it is extremely 

 difficult to define the actual limits of these sporogenous masses." 



In Tmesipteris there are normally two masses of sporog- 

 enous tissue corresponding to the two loculi in the mature synan- 

 gium; in Psilotum, which correspond closely with Tmesipteris 

 in other respects, there are three. Whether additions are made 

 to the sporogenous tissue from cells outside the original arch- 

 esporium was not determined with certainty, but Professor 

 Bower thinks it not improbable. In Psilotum the young arch- 

 esporium is more clearly defined than in Tmesipteris, and it 

 seems not unlikely that each sporogenous mass is referable to the 

 division of a single primary archesporial cell. In both genera 

 some of the sporogenous cells do not develop spores, but simply 

 serve for the nourishment of the others, as in Equisetum. 



The fully-developed synangium has the outer walls of the 

 loculi composed of a single superficial layer of large cells, be- 

 neath which are several layers of smaller ones (Fig. 294, D). 

 The cells composing the septa are narrow tabular ones, with 



