4i8 MOSSES AND FERNS chap. 



Ferns, consists of two layers, in some places of three. The 

 number of sporogenous cells is usually eight, but some or all of 

 these may divide again, so that the whole number ranges from 

 eight to sixteen. The dissolution of the tapetum walls and 

 subsequent division of the spores follow precisely as in Azolla. 

 In stained sections the nucleated protoplasm of the tapetal cells 

 is very evident after the walls have disappeared. At this point 

 the difference in the two kinds of sporangia becomes manifest. 

 Those in the lower part of the sorus, i.e. the oldest ones, form 

 the macrosporangia, the upper ones microsporangia. In the 

 latter all the spores mature ; in the former, as in Asolla, one 

 spore grows at the expense of the other, and finally fills the 

 sporangium completely. As in Azolla no trace of an annulus 

 is seen, either in the young or fully-developed sporangium.^ 



As the sporocarp ripens, the outer cells become excessively 

 hard, especially the first layer of hypodermal cells (Fig. 2 1 7), 

 whose walls become so thick as to almost obliterate the cell 

 cavity. The second hypodermal layer is also thickened, but not 

 so strongly. At maturity the sporocarp of P. Americana forms 

 a globular body about 3 mm. in diameter, covered with 

 hairs, and attached to a long peduncle which bends downward 

 and buries the ripe sporocarp more or less completely in the 

 earth. The statement ^ that this species has but three 

 chambers is incorrect, and except for the longer pedicel of the 

 fruit, and a slightly thinner epispore in the upper part of the 

 macrospore, it corresponds exactly to P. globulifera. The 

 sporocarp splits into four parts, corresponding to the four lobes 

 of the young fruit, and the membranaceous margins of the leaf 

 form a tough indusium surrounding the sporangia. This 

 indusium is not, at least in P. globulifei'a^ readily pervious to 

 water, and germination does not begin for a long time after the 

 valves separate, unless the indusium is artificially opened. 

 Except for the number and position of the sori, and the relative 

 position of the two sorts of sporangia, Marsilia agrees exactly 

 with Pilularia. The sorus canals form two longitudinal rows 

 along the sides of the elongated fruit rudiment, which may be 

 compared to a pinnate leaf. In Marsilia^ occupying the middle 



^ For the details of the development of the macrospore, see Meunier (i), pp. 

 382-387. 



- Goebel (10), p. 240 ; Underwood (4), 2nd ed., p. 127 ; " Botany of California," 

 vol. ii. p. 352. 



