COMPARATIVE MOEPHOLOGY OP GYMNOSPERMS 167 



In Tumhoa and Gnetum, free nuclear division, which in 

 Ephedra sometimes results in but two nuclei, is abandoned en- 

 tirely, and the whole oospore behaves as do one of the free cells 

 in Ephedra, elongating in suspensor fashion, and cutting off at 

 the tip an embryonic cell. 



The case of Ginkgo is a very interesting departure from the 

 ordinary sequence. The germination of the oospore begins with 

 free nuclear division, but suspensor development is entirely 

 omitted, and as a consequence the oospore is filled with a com- 

 pact tissue which organizes directly into the embryo, and this 

 encroaches upon the endosperm by the growth of the whole 

 mass. It is important to note that the method of Ginkgo was 

 also that of the Bennettitales, and this fact raises the question 

 whether this does not represent the type of embryogeny among 

 the primitive Gymnosperms, in which the three embryonic re- 

 gions of modern Grymnosperms have not been differentiated. 

 No evidence as to the embryo of the Cordaitales has been ob- 

 tained, so that the very important testimony which they might 

 give is not available. If this be the more primitive type of 

 embryogeny, it would follow that the conspicuous suspensor of 

 most living Gymnosperms is a structure of secondary origin in 

 the group. However, these two types of embryogeny may not 

 hold any definite relation to each other. 



The development of the embryo proper seems to follow no 

 rigid sequence. The first division may be transverse or vertical, 

 and the subsequent divisions may be in almost any order or 

 direction. Perhaps too much emphasis has been laid upon the 

 sequence of events in the earliest stage of the embryo ; at least, 

 the data now at hand seem to indicate the entire lack of a definite 

 sequence. A study of the organization of the growing points of 

 the great body regions, however, should lead to important re- 

 sults, but unfortunately the knowledge of Gymnosperms in this 

 regard is too meager to justify any definite statement. 



It is evident that polyembryony is natural in the group, and 

 numerous embryos in various stages of development are com- 

 monly found imbedded in the endosperm, but for more than one 

 embryo to come to maturity seems to be a rare phenomenon. 



12 



