332 NATURAL SELECTION [ch. 



any special way for their environment, these specific or varietal 

 characters are exceedingly constant. Naias graminea^ var. Delilei^ 

 for example, has been known in Egypt to have the same cha- 

 racters for about a century, and when introduced into England 

 these characters remain wholly unchanged^. These considera- 

 tions seem to the present writer to confirm the conclusion 

 drawn by Willis from his study of the Podostemaceae a con- 

 clusion which has also been arrived at by various workers in 

 other fields that Natural Selection is incompetent to explain 

 the origin of the sharply defined entities which we call 

 species. 



But when we turn to Natural Selection in its second aspect 

 as one of the various factors to which adaptation may be due 

 Willis's conclusions seem to need some revision. Accepting the 

 view that we have, among the Podostemaceae, a case of evolu- 

 tion untrammelled by the limiting influence of Natural Selec- 

 tion, we find associated with this freedom, the development of a 

 large number of well-defined species, remarkable for their lack 

 of definite adaptation to the conditions of their life. The view 

 may well be taken that the lack of adaptation which Willis 

 finds so striking, is actually in part attributable to the absence of 

 competition and hence to the elimination of Natural Selection. 

 From this point of view, the Podostemaceae furnish evidence 

 negative but forcible for the importance of Natural Selection 

 in the development of adaptation, since here we have a case of 

 the absence of Natural Selection correlated with the absence of 

 special adaptations. Among the Podostemads, presumably, all 

 variations good, bad, or indifferent have had an almost equal 

 chance of perpetuation, provided they did not interfere with 

 those general features which gave the group its special capacity 

 for growth in the rapidly running water, which is so inimical to 

 most forms of plant life. Perhaps the present condition of the 

 Podostemaceae may be broadly compared with that of certain 

 of our domestic animals, consisting at the present day of many 

 sharply defined breeds, which could not have survived the 



1 Magnus, P. (1883). 



