BOTANICAL SUBJECTS. 291 



without a specialized \asculai- system ; aud in those we have 

 stems aud leaves and i'eprodueti\e or«5aiis all the same. It does 

 not appear to me that, because triehomes have no connexion with 

 the fibro-vascular system, therefore they are not homologous witli 

 the tentacles of Drosera. 



The reason why triehomes are not usually connected with the 

 vascular system may well be that cit its termination this system 

 usually aborts^ and leaves the epidermis and trichomas to be 

 nourished b}- imbibition of juice poured out by the terminal fibrils 

 of the vascular branches, in the same way that the veinlets of the 

 leaves terminate in the midst of a group of parenchyma cells, 

 which are nourished by the fluid poured out. 



We see the epidermal hairs of Begonia phyllomaniaca turning 

 into branches, and it is presumable that the need of specialized 

 vascular branches gives rise to their development. Similarly, in 

 Figs. 121-123, we see that the tentacles of Drosera intermedia 

 can now be ovules, now glandular hairs, now branches, with sub- 

 divisional tentacles as leaves, and now simple leaves, with tentacles 

 or ovules on their margin, while other tentacles have aborted into 

 ordinary hairs. Can there be anything more like a branch, with 

 verticillate suli-branches, than the hair of Alternanthera ? * 



We seem to be afraid of accepting the revelations of these 

 reversions for what they really mean, because we are biassed by 

 the notion which we have set up that disconnection with the fibro- 

 vascular system also means disconnexion with the homological and 

 morphologicitl continuity of the whole plant. This break of 

 continuity may be due, as I said, to abortion of the terminal 

 vascular veinlets, because the cells in their vicinity can l)e 

 nourished by cell-to-cell imbibition. 



We seem to see striking examples of this homologv at every 

 turn. Examine for instance, the young pink leaf of Cheuopodium 

 petiolare. Its whole susface is closely packed with minute 

 globular cells, giving it a pollen-like surface. At the tips, where 

 the leaves are turning green, these cells decay and are shed, and 

 when the leaf turns wholly green, they have all disappeared. They 

 are evidently what are called hairs, and possibly may be either 

 protective or excretory. Examine then the placenta of Lobelia 

 Tupa, and you will find the surface closely packed with similar, 



* Le Ma out and Decaisne, " Syst. of Bot.," p. 139. 



T 2 



