GERMINATION AND EMBRYOGENY OF GNETUM GNEMON. 295 



and Gnetum the formation of the outgrowth begins at a still 

 later period, when the cotyledons, and apex of stem and root 

 are already clearly defined. 



This difference of time of first appearance of these out- 

 o'rowths should be considered in its relation to the physiolo- 

 gical position of the young embryo. In the Ferns the body 

 of the embryo does not remain long embedded in the tissues 

 of the prothallus ; and the foot is thus required at an early 

 period to transfer nutritive materials from the latter to the 

 embryo, which is still unable to sup^iort itself. In Sela- 

 ginella the embryo remains till a later period embedded in 

 the tissues filling the macrospore, and the foot is accord- 

 ingly required and developed at a later stage. In Welwit- 

 schia and Gnetum the embryo remains embedded in the 

 endosperm till a still later period, and accordingly the out- 

 growth makes its appearance at a still later stage. 



We have already seen that this lateral outgrowth of the 

 hypocotyledonary axis is not a constant organ throughout the 

 Gnetaceee, it being absent, at least in some species of 

 Ephedra ; also that in Welwitschia and Gnetum it is pro- 

 portionate in size and complexity to the size of the seed. 

 These facts lead to[the conclusion that in the Gnetaceae at all 

 events the outgrowth is formed only where it is required, 

 and that its size and complexity varies according to the 

 duties it has to perform. Referring to the vascular Crypto- 

 gams, we find a similar inconstancy of development of the 

 foot. Cases occur in which the foot is not fully developed, 

 and in which, were it not for the comparison of such em- 

 bryos with the Ferns, the very slight lateral swelling of the 

 axis would hardly be noticed. For instance, in Equisetum 

 the foot always remains small, and is never fully developed 

 (Sadebeck, in Schenk's ' Handbuch der Botanik,' p. 224). 

 In this genus, moreover, a distinct lateral outgrowth is not 

 required, since the root of the young embryo perforates the 

 prothallus, and the embryo remaining thus in close contact 

 with it can draw supplies directly from it. A similar absence 

 of a developed swelling of the hypocotyledonary axis is said 

 to occur in Azolla. 



In Marsilea we have a foot which is distinctly adapted in 

 form to the function of transfer of nutritive matters ; it 

 differs from that of the true ferns in having a concave apex, 

 which fits upon the convex surface of the prothallus. 



"We thus see that, in the vascular Cryptogams, as in the 

 Gnetaceae, these organs of nutrition of the young embryo 

 appear on the hypocotyledonary axis as outgrowths of the 

 tissue ; they are not universally develvoped, nor are they, 



