212 MINNESOTA BOTANICAL STUDIES. 



connected tissue similar to that which forms the inner portion of 

 the cystocarp wall. These filaments are not present in all cys- 

 tocarps. They appear to result from the tearing of the sub- 

 cortical tissue in the formation of the cavity of the cystocarp. 



In the young cystocarp (Fig. 12) the thickened wall is al- 

 read}^ present and shows its permanent division into two layers. 

 The cells of the outer layer are arranged in filaments perpen- 

 dicular to the surface, those of the subcortical layer are ar- 

 ranged in oblique rows converging towards the apex of the 

 pericarp. The placental area and spore mass are represented 

 by a few connected cells with very dense contents. The cavity 

 of the cystocarp is, at this stage, comparatively small. The 

 carpostome is already developed even in very young cystocarps. 

 It appears to be formed by the tearing apart of the cells to- 

 gether with the destruction of some of the cells. As was men- 

 tioned in the account of the gross anatomy of the lamina, in 

 two specimens the apical portions of the frond present a pecu- 

 liar mottled appearance. Cross sections of these areas show 

 that in places there are slight protuberances from the surface of 

 the thallus associated with an unusual development of cortical 

 and, in some cases, also of the subcortical cells. The cortical 

 cells are rather narrower than elsewhere, and more elongated 

 perpendicularly to the surface of the frond. The cortical layer 

 is also a greater number of cells deep than elsewhere, and in 

 some cases there is also increase in the number of the smaller 

 central cells situated immediately beneath the cortex. . These 

 areas in some cases involve only a few cells, in others they are 

 .5 mm. in diameter. The structure of the larger protuberances 

 agrees essentially with that of the young cystocarp described 

 above, except that in no cases could any cavity be discovered 

 in them. They appear to me to be very young stages of cysto- 

 carp development, but in no case could positive evidence of 

 their nature be discovered. 



No indubitable cases of tetraspore formation were seen, but 

 in some cross sections taken through the up.per part of the 

 lamina, some of the cortical cells are peculiarly divided pro- 

 ducing somewhat the appearance of tetraspore formation (Fig. 

 6). Except for their peculiar arrangement, these cells appear 

 in all respects similar to the ordinary cortical cells. Whether 

 they are tetraspores or not, could not be determined. Their 

 method of division is cruciate or somewhat irregular approach- 

 ing the tetrahedral arrangement. 



