250 Mr Scott Elliot on the 



The conditions are so much more sharply defined and so 

 varied in every way that one is able to get a sort of glimpse 

 of the way in which to work. I have tried to give some 

 hints in this paper, which are pretty obvious and seem 

 probable, but it would require years of residence and unin- 

 terrupted ol)servation to obtain a really satisfactory notion 

 of how evolution has progressed. 



Taking the order Scrophularinete, it was most interesting 

 to me to watch the manner in which it took on the pre- 

 valent form of the vegetation round it. Thus in the Perie 

 forest one finds a scrophularineous evergreen tree, Halleria 

 clliptica, some 40 feet high, or even more. It has leaves just 

 like most of those around it and berries, which the birds eat, 

 and so distribute the seeds. In the smaller brushwood on the 

 Boschberg, Somerset, I found Pliygdius capensis, a shrub in 

 foliage and appearance quite like its neighbours. On the 

 flanks of Table Mountain and the dry hill-slopes in the south- 

 western region there are several genera which take on the 

 cricoid appearance. These are small shrubby perennials, with 

 small hard leaves and small flowers, such as Lyperia, Chcenos- 

 toma and others, or with sticky leaves which apparently answer 

 the same purpose, e.g. 8phcnandra. Although not in this 

 order, the nearly allied Selagineae consist wholly of this kind 

 of plant. On the tops of the mountains, again, amongst the 

 long grass and plants with chiefly radical leaves and long 

 peduncles, one finds Zahizianshya, Buchncra, &c., which have 

 rosettes of long, slightly cut radical leaves exactly like their 

 neighbours. In the Karoo one finds the depressed, moss-like 

 tufts of Aptosimum and Pcliostomum. The extraordinary 

 parasitic Hyohanclie has even become fleshy, so that in this 

 one order the typically South African forms have taken on 

 what seems to me the general type of the districts to which 

 they belong. 



Instances of this kind could be multiplied indefinitely. 

 The Leguminosse show the same modifications remarkably 

 well. We have the typically cricoid forms in Aspalathus 

 and the sub-tribe Crotalariea?. Some Karroid Aspcdathi even 

 have succulent leaves which, I think, is very rare in Legumi- 

 nosse. They also grow into trees such as Erythrina 

 Calpurnia and Virgilia. I have even noticed what seems to 

 me similar tendencies in places where one would least expect it. 



