LINITEAN SOCIETY OF LONDON. XV 



And yet we must believe this if we admit Strasbui^er's pedigrees. 

 We must suppose that races, after having once secured the advan- 

 tages of a total separation of the two sexes and undergone modifica- 

 tions suited to their separate requirements, have again returned to 

 their primitive state of sexual proximity, and commenced a totally 

 different series of modifications destined to counteract the evil effects 

 of that proximity. A much more simple hypothesis would be that 

 Conifers separated from the parent stock before the development of 

 floral envolopes, the higher Dicotyledons before the separation of the 

 sexes. The arrangement of the vegetative organs, or phyllotaxy, had 

 probably acquired considerable perfection before the separation of 

 either of these primary classes of Dicotyledons ; for we have the ver- 

 ticiUate arrangement in alternating whorls in Frenelu, Ephedra, Ca- 

 suarina, Calycopephis, Hippuris, and many others belonging to the 

 most widely separated natural orders — the opposite and decussate 

 leaves in various genera of Conifers and Gnetaceae, as well as in nume- 

 rous orders, whether of Monochlamydese, Gamopetalae, or Polype- 

 talae ; and in Conifers, as in the higher Dicotyledons, the whorled or 

 decussate arrangement is variously broken up into the spiral, the al- 

 ternate, or the scattered. But the reproductive organs having at that 

 early stage taken the two directions of total separation of the sexes in 

 the one and their union in the other within a set of floral envelopes, 

 their progress was thenceforth in dift'erent directions, and homology 

 in a great measure disappeared. In Coniferae this complete separa- 

 tion of the sexes and fertilization through the agency of wind being 

 established, natural selection would only promote the development of 

 such floral envelopes as might be required for protection and would 

 not interfere with the fertilizing process and would necessarily be 

 very different in the male and in the female flowers. Accordingly 

 one great point established by Strasburger and others is that in Coni- 

 ferae and Gnetaceae there is no homology between the male and the 

 female flowers. In the higher Dicotyledons the male elements took 

 their place around the females, and axial appendages would be early 

 established or modified for the various purposes of assisting, protect- 

 ing, or controlling fertilization or maturation, all of which arrange- 

 ments would become more and more complicated as the plants came 

 to be benefited by cross fertilization through insect and other ex- 



Bcended from perfect hermaphrodite ones by the gradual abortion of one of the 

 sexual elements, in which the abortive element is occasionally again perfected, 

 is no real retrogression. An occasional perfect stamen in a female Euphorbia- 

 ceous flower cannot be said to be a real return to hermaphroditism. 



