100 



origin. They must have origiuated before, or at the same time 

 with, the Gymnosperms. The following diagram may show the 

 way of their possible origin. 



HETEROSPOROUS PTERIDOPHYTES. 



Gymnosperms [ Ephedroideae 



Gnetaccae / Welwitschoideae 



j I ^^Gnetum Gnemon 



Angiosperms f Gnetoideae 



^Gneta of Karsten. 



I am therefore inclined to consider the Gnetaceae as the 

 aequivalent of the Gymnosperms and Angiosperms combined 

 and running paralell to these, having originated entirely in- 

 dependent of them. 



In these two paralell rows the Gymnosperms can be com- 

 pared to the Ephedroideae the Angiosperms to the Gnetoideae. 



The sexual apparatus of the Gnetoideae is much more re- 

 duced than that of the Angiosperms. 



While no intermediate forms between Angiosperms and 

 Gymnosperms are known, the Welwitschoideae form a link 

 between the Ephedroideae and the Gnetoideae as has allready 

 been pointed out by Karsten '). It is as he says easy to derive 

 the Gneta described by him from a form like Welwitschia. To 

 derive Gnetum Gnemon from Welwitschia seems at first very 

 difficult, yet I think that the reader will agree with me in 

 considering Gn. Gnemon as the more primitive form. 



Of the groupe of the Gnetaceae Ephedra is doubtless the 



1) Karsten (see p. 18 of the historical review) mentions the fact that occasio- 

 nally three embryosacs may arise from the division of one embryosacmothercell. 

 If any of such triplets ever attain adult size it would be exceedingly interesting 

 to see if each perhaps formed but one, be it a steril or a fertil prothallium (see 

 here after). I have one abnormal embryosac of Gnetum Ula Brgn. divided by a 

 faint crosswall, each half of this embryosac is filled with a steril prothallium, yet 

 in normal embryosacs of Gnetum Ula, one can distinguish a fertil and a steril 

 part, though not so plainly as in Gnetum Gnemon L. This case will be discussed 

 more fully in a future paper on Gnetum Ula Brgnr. 



