xxvii] SPECIFIC DIFFERENCES 331 



The idea that the specific distinctions among the Potamogetons 

 are somewhat fluid, may be partly attributed to a too exclusive 

 use of external features in systematic work; there is no logical 

 reason for the exclusion of anatomical characters from taxo- 

 nomic study and their importance is fortunately now becoming 

 recognised^. It has been demonstrated, for instance, that, 

 though the flower and fruit characters of the Potamogetons 

 show very small difl-erences in the different species, and the 

 external characters of the vegetative parts which can be used 

 in diagnosis are few and variable, the anatomical characters of 

 the vegetative organs prove to be much more constant^. That 

 the majority of specific differences observed among the Pond- 

 weeds could be of any survival-value, seems almost incompre- 

 hensible, and the lack of any apparent utility in certain specific 

 characters is seen almost more clearly when we turn to the 

 marine Potamogetonaceae. In the case of these plants, the 

 anatomy of the leaves, taken by itself, furnishes data for exact 

 specific determination^. Dealing with Cymodocea and Halodule^ 

 Sauvageau* remarks, " It is an interesting fact that plants which 

 in general are of relatively simple structure, present such a 

 variation from one species to another, and, at the same time, such 

 constancy in specific anatomical characters." It can scarcely 

 be imagined that the majority of the specific differences, ob- 

 served in the anatomy of the vegetative organs of the marine 

 Potamogetonaceae, can serve any purpose in connexion with 

 the relatively uniform conditions of their submerged life, and, 

 unless these differences are advantageous, it is impossible to 

 suppose that they are due to Natural Selection. It is most re- 

 markable that in so simple a genus as Naias^ in which some, at 

 least, of the external specific differences can hardly, by any 

 stretch of imagination, be supposed to fit their possessors in 



1 The excellent method advocated by R. C. McLean (^Netv Phyt. 

 Vol. XV, 191 6, p. 103) for rendering herbarium material available for 

 anatomical work, makes the use of internal characters in systematic study 

 more practicable than hitherto. 2 Raunkiaer, C. (1903). 



^ Sauvageau, C. (1891I); see also p. 131. * Sauvageau, C. (iSgi^j. 



