92 OBJECTIONS TO THE HYPOTHESIS [pt. i 



Zealand, etc., coincide with that of the endemics, and both 

 decrease together from tliat point, the endemics much the more 

 rapidly? 



Other formidable arguments against this view are given below, 

 in Chapters xv, xvi of Part II. 



The hypothesis of youth (within a country) and area can only 

 be accepted if one be prepared to accept with it the numerous 

 absurdities to which it leads. In particular, it involves a most 

 remarkable amount of rising and falling in the scale of area of 

 distribution, for which we have no warrant. The distribution of 

 the plants of the outlying islands of New Zealand (p. 66) seems 

 to provide a very strong case against it, for how can youth 

 ensure that a species shall reach more of these little islands? 



"The families Tristichaccae and Podostemaceae also afford an 

 excellent test case for the question of age or youth, for owing to 

 their peculiar morphology one can say with reasonable approach 

 to certainty which are the older forms. He would be a bold man 

 who would say that such forms as Laxvia in the one family, or 

 Castelnavia in the other, with their violently dorsiventral struc- 

 ture, shown in the lichen-like vegetative body and the extra- 

 ordinarily modified flowers, were older than such forms as Tri- 

 sticha or Podostemon, Avhich are almost radially symmetrical, 

 and come near to the ordinary type of submerged water plant. 

 Yet the latter are widespread and almost universal, covering 

 the whole range of distribution of families, while the violently 

 dorsi^'ent^al forms are all endemic to comparati\'ely small areas, 

 Latvia, for example, occurring from Cej'lon to Bombay, Castel- 

 navia in the Araguaya and one other river in Brazil. It is im- 

 possible to talk of local adaptation in these plants, as I have 

 elsewhere pointed out (124); there is nothing to be adapted to. 

 The non-dorsiveutral forms are just as common as the dorsi- 

 ventral, whether in slowly or in swiftly moving water" (quota- 

 tion from 128). 



Mrs Arber (4, p. 306) has brought up a parallel case in the 

 genus CaUitriche. 



Some, while admitting that in general endemics are not relicts, 

 say (24) that the endemics of mountains, at any rate, are usually 

 such, especially as the wides, not infrequently, do not ascend as 

 high as they do. This latter fact is a strong argument against 

 the explanation often given of mountain endemics, that they 

 have retreated upwards to escape the competition of the wides 

 in the plains below, for it would be very remarkable if they 



