306 GEOGRAPHICAL DISTRIBUTION [CH. 



to enter upon any discussion of the grounds upon which Willis 

 bases his view, or of the criticisms to which it has been sub- 

 jected. It must suffice to see whether it can be applied to any 

 aquatic plants, and, if so, with what result. The only hydrophytes 

 with which Willis himself deals are the Tristichaceae and Podo- 

 stemaceae. He points out that, owing to the peculiar morphology 

 of these plants, it seems possible to say with some degree of 

 certainty which are the older forms. Tristicha and Podostemon 

 are almost radially symmetrical, and do not diverge greatly 

 from the ordinary type of submerged plant. Lawia and Cas- 

 telnavia, on the other hand, show the most extreme dorsi- 

 ventrality of structure and have highly modified flowers ; most 

 botanists would probably agree that Tristicha and Podostemon are 

 the older types, while Lawia and Castelnavia represent a more 

 recent evolutionary development. If this view be accepted, the 

 families in question form a striking illustration of the principle 

 of Age and Area, for Tristicha and Podostemon cover the whole 

 range of distribution of the families, while Lawia and Castel- 

 navia are both limited to comparatively small regions 1 . 



The difficulty of applying a morphological test to Willis's 

 or to Guppy's theory lies in the fact that botanists seldom agree 

 as to which members of any given family or genus are to be 

 considered primitive and which are more specialised. The Water 

 Starworts (Callitriche\ however, seem to the present writer to 

 present a case which is, in this regard, less problematical than 

 most. Within Callitriche we have two sub-genera, one of which, 

 Eu-callitriche, has the upper leaves floating and is characterised 

 by aerial pollination, while the other, Pseudo-callitriche^ is com- 

 pletely submersed throughout life. Most botanists would pro- 

 bably admit that the genus is descended from terrestrial 

 ancestors, and that the submerged Pseudo-callitriche is hence a 

 more highly specialised and recent type. In distribution, Eu- 

 callitriche is almost cosmopolitan, while Pseudo-callitriche is 

 confined to the North Temperate regions. The distinction holds 

 even within our own country, where C. verna and its sub-species 

 i Willis, J.C. (1917). 



