5i6 MOSSES AND FERNS chap. 



much thickened at -a later stage, and thus is formed the dia- 

 phragm which is so conspicuous in most species, and which led 

 Pfeffer to suppose that the first division in the young prothal- 

 lium proper from the lower part of the spore, in which later the 

 "secondary endosperm" is formed. 



Scattered through the protoplasm of the spore-cavity below 

 the diaphragm are numerous nuclei. The protoplasmic layer 

 becomes rapidly thicker (Fig. 297, A), and finally completely 

 fills the cavity of the spore. The thickenings upon the outer 

 spore-coat are very evident even before the primary nucleus 

 divides, and they increase rapidh^ in size, as the spore develops. 

 A very casual examination suffices to show that the tapetal cells 

 of the sporangium here play a most important part, not only 

 in the development of the spore-coat, but also in the growth 

 of the prothallium. The rapid increase in the amount of pro- 

 toplasm in the spore during the growth of the prothallium, as 

 well as the growth of the spore itself, can only be accounted for 

 by the activity of these cells, which are in close contact with 

 the spore, and show every evidence of being active cells, through 

 whose agency the materials are conveyed to the spore for its 

 further development. 



The first archegonia begin to form shortly before the spores 

 are shed, and soon after, the exospore splits along the three ven- 

 tral ridges and exposes the central part of the prothallium. 

 This, like that of Isoetes, is quite destitute of chlorophyll, and 

 is entirely dependent upon the food materials in the spore for 

 its further development. About this time also begins the cell- 

 formation in the part of the spore below the diaphragm (Fig. 

 298). This is simply a continuation of the same process by 

 which the apical tissue was developed, but the cells are larger 

 and more irregular. 



The archegonia are produced in considerable numbers, and 

 apparently in no definite order. Their development corre- 

 sponds with that of Lycopodinm, but the neck is very short, 

 like that of the Marsiliacese, each row of neck cells having but 

 two cells. No basal cell is formed, and the central cell is sepa- 

 rated from the diaphragm onl}^ by a single layer of cells. The 

 neck canal cell (Fig. 298) is broad, like that of Isoetes, but the 

 nucleus does not, apparently, divide again. The Qgg (Fig. 298, 

 E) shows a distinct receptive spot, and the nucleus is clearly de- 

 fined. At this stage the diaphragm is very evident and much 



