XII LEPTOSPORANGIATyi: HETEROSPORE^E 409 



primary octant is reduced to the much narrower form found 

 in the okier sporophyte. 



The Root 



The first wall in the root quadrant strikes the basal wall 

 at an angle of about 60°, so that the octants are of very 

 unequal size (Fig. 212, C), and the larger one, as in other 

 similar cases, becomes at once the initial cell of the root, which 

 in both genera shows the same regular divisions that char- 

 acterise the Polypodiaceas. The segments of the root-cap do 

 not form any periclinal walls, and remain single-layered. The 

 root, like the cotyledon, is traversed by regular air-chambers, 

 and its transverse section resembles very closely that of the 

 leaf These air-chambers appear while the root is very young, 

 and at a point between the endodermis and the cortex. The 

 latter is at this stage divided into but two cells, the outermost 

 of which by a further tangential division becomes two-layered, 

 the outer forming the epidermis, and the inner by similar 

 divisions three-layered. The two outer layers divide by radial 

 walls, but the inner ones divide only by periclinal walls, and 

 form one-layered lamella separating the air-spaces and connect- 

 ing the endodermis with the outer cortex. 



TJie Foot 



The first divisions in the foot quadrant follow closely those 

 in the root, but this regularity soon ceases, and after the first 

 divisions no definite succession in the walls can be distinguished. 

 The foot remains small, but, as we have seen, the first segments 

 of the lower epibasal octants practically form part of it, and 

 doubtless all the lower cells are concerned in the absorption of 

 food from the spore. The volume of the protoplasm in the 

 spore increases as the prothallium grows, but loses more and 

 more its coarsely granular structure. In both Marsilia and 

 Pilularia the nucleus of the spore cavity soon becomes indis- 

 tinguishable, and in the former is from the first very small. In 

 Pilularia it is larger, and in the later stages bodies were 

 observed that looked as if they might be secondary " endosperm- 

 nuclei," like those of Azolla, but their nature was doubtful. 



The leaves are at first alike in both genera, and the earliest 

 ones do not show any trace of the circinate vernation of the 



