FERNS. 



269 



attaches the embi-yo to the protliallium, and transports nourisliment for the 

 embryo until it can become attached to the soil and lead an independent ex- 

 istence. Durint; this time tlie wall of the archegonium grows somewhat to 

 accommodate the increase in size of the embryo, as shown in tigs. 312, 313. 

 But soon the wail of the arcliegonium is ruptured and the embrvo emerges, 

 the root attaches itself to the soil, and soon the prothallium dies. 



The emljryo is first on the inider side of the prothallium, and the first leaf 



Fig. 311- 

 Two-celled embryo of Pteris serrulata. Remnant of arcliegonium neck below. 



and the stem curves upward between the lobes of the heart-shaped body, and 

 then grows upright as shown in fig. 314. Usually only one embryo is formed 

 on a single prothallium, but in one case I found a prothallium with two well- 

 formed embryos, which are figured in 315. 



559. Comparison of ferns with liverworts and mosses. — In the ferns then 

 we have reached a remarkable condition of things as compared with that 

 whicli we found in the mosses and liverworts. In the mosses and liverworts 



