284 F. O, BOWER. 



divisions in the parenchyma of the petiole preparatory to the 

 formation of one of these glands may be seen in fig. 14. 

 These glandular structures are found at the base of the 

 petiole of the cotyledons, as well as of all foliage leaves. 



Polyemhryony. 



From what has been stated above it will be seen that 

 polyemhryony is the rule in Gnetum, in the same way as in 

 other members of the group, viz. a number of suspensors are 

 found in each seed (and probably several spring from each 

 corpusculum) ; on the apex of each of these an embryo may 

 be formed. 



But with Gnetum Gnemon the matter does not end here. 

 It has been above observed that branched suspensors occur 

 occasionally in the ripe seed ; and we have no reason to 

 suppose that embryos may not be formed on the apex of each 

 of these branches. I have, moreover, been able to observe 

 directly that one suspensor can produce more than one 

 embryo. In the case represented in fig. 20 a number of 

 bodies, which resemble normal embryos in their structure, are 

 borne on a single branched suspensor. Further, I can 

 suggest no other explanation for irregular bodies, such as 

 those represented in figs. 21, 22, which are not ofunfrequent 

 occurrence, than that they are instances of a like branching 

 of suspensors (or perhaps proliferation of embryos, as it 

 might be more correctly termed, in the case of fig. 22) seen 

 at an earlier stage. 



Besides such cases as these it very frequently happens 

 that embryos are united by their posterior part, without its 

 being possible to trace them back to the single suspensor 

 (or, possibly, the suspensors) from which they originated ; 

 and in considering these cases it must be remembered that 

 in G. Gnemon each suspensor, as a rule, pursues an individual 

 course, otherwise this observation would have no bearing 

 upon the point under consideration. It must remain an 

 open question whether these embryos are simply in juxta- 

 position or are genetically connected. However this may be, 

 the fact is demonstrated by instances, such as those in figs. 20 

 and 22, that there is to be found in some cases in G. 

 Gnemon a higher order of polyemhryony than that described 

 for others of the Gymnosperms. 



In each seed only one embryo becomes finally developed. 

 The others, though rich in protoplasm in their early stages, 

 lose their activity after a time : they become transparent, 

 and are often squeezed out of shape ; in fact they are aborted. 



