GERMINATION AND EMBRYOGENY OF GNETUM GNE.MON. 291 



ducted through these tissues. There is in the species of 

 Ephedra above named no lateral outgrowth of the hypocoty- 

 ledonary stem comparable to the feeder of Welvvitsohia or 

 Gnetum, but it ^viIl readily be seen that, were the store of 

 nutritive substances in the endosperm larger, as is the case 

 in the other two genera, the need of some special provision 

 for the transfer would be more pressing. Moreover, in the 

 seedling of Ephedra we have a parallel process to that 

 which is the chief function of the feeder. In both cases 

 nourishment is received by the seedling, and absorbed from 

 without at the surface of the base of the hypocotyledonary 

 stem. 



Conclusion. 



Though the development of the corpuscula and the pro- 

 cess of fertilisation have not yet been observed in Gnetum 

 Gnemon, it seems probable that in these respects it resem- 

 bles Ephedra. But in G. Gnemon it is found that in the 

 ripe seed the suspensors bear no embryo, appearing merely 

 as more or less branched tubes, which permeate the apical 

 part of the endosperm. On the germination of the seed 

 these suspensors are capable of further development and of 

 producing embryos, the seed itself being meanwhile inde- 

 pendent of the parent plant. 



A similar process has been described by Strasburger 

 ('Conif. und Gnet.,' p. 295) in Ginkgo hiloha, and by 

 Warming in Ceratozamia, and some other Cycads (' Bidrag 

 til Cycadeernes Naturhistorie,' French summary, 1879). It 

 would be a mistake to attach much importance to this 

 matter as a character in common between the groups in 

 question (compare Heer, ' Bot. Centralblatt,' 1882, p. 239). 

 The inconstancy of the phenomenon in each of the three 

 groups is sufficient argument against such a view. 



These cases may be compared with those x\ngiosperms in 

 which the embryo is still rudimentary in the ripe seed. 



The mode of development of the embryo in G. Gnemon 

 corresponds closely to the type for the Gymnosperms. At 

 an early stage an apical cell makes its appearance, and it 

 may be recognised in much older embryos than is possible 

 in Ephedra. In its appearance and the length of period of 

 activity of its apical cell the embryo of G. Gnemon com- 

 pares more closely with that of Juniperus or Thuja than 

 with that of Ephedra. In Welwitschia according to Stras- 

 burger (' Angiosp. and Gymnosp./ pi. xxii) there is no apical 

 cell. 



In the differentiation of the embryonic body G. Gnemon 



