122 DR. J. STEPHENSON ON THE MORPHOLOGY, CLASSIFICATION, 



be reached by different paths ; in other words, certain groups 

 with the same anatomical characters may have a polyphyletic 

 origin. > 



Must it then be an article of faith that each genus has arisen 

 once and once only ? Variations may be innumerable — no doubt 

 every organ and part may vary and does vary independently ; 

 but the variations that mean anything, that come to anything 

 from the point of view of evolution, are not innumerable — they 

 are limited both in their seat and in the direction they take. 

 And it would seem that similar steps are being taken in many 

 parts of the range of a group; with the consequence that the end 

 condition is similar also. In other words, we have a polyphyletic 

 origin of certain groups. 



Objections may be raised to the above line of argument. It 

 may be said, for example, that what I have been discussing are 

 cases of convergence, which nobody has ever denied. Or it may 

 be said that if, as I have claimed, Megascolex or any other genus 

 has a multiple origin, then it is not a true genus, and that the 

 group we know as Megascolex really consists of several genera 

 with different lines of descent. 



The term " convergence " is applicable to the case of Monogaster 

 and Octochcetus, just discussed; it is applicable to that of the 

 group of species of Megascolex descended from Perionyoc and the 

 group descended from JVotoscolex, as well as that descended from 

 &pence?*iella, if this origin should be confirmed. But it is not 

 applicable to the different groups of species of Megascolex 

 descended from different Notoscolex forms, nor, generally, to the 

 multiple origin of one genus from another single genus : 

 there is no convergence here — the developments are parallel. 



Further, along with the use of the term convergence there 

 seems to go an idea that a careful morphological examination, or 

 a consideration of distribution, will ultimately suffice to distinguish 

 groups of different origin, and that a different descent will always 

 betray itself to sufficiently careful and minute investigation. My 

 point is that it may not do so. Naturally, in giving specific 

 instances to support the a priori argument, I have had to give 

 cases where som.e features of certain species of a genus seem to 

 countenance a separate origin for these species ; otherwise if there 

 had been no anatomical features at all to support the hypothesis 

 of a double origin, the argument could have been deductive only. 

 I have tried to show that in the genus Megascolex we can with 

 some probability separate off small groups here and there which 

 have originated at different times, from different ancestors — 

 these ancestors, too, belonging to more than one genus. But, 

 even if these groups were separated off, are we prepared to say 

 that the \arge number of species which remain (the great 

 majority of the genus) own a single origin ? — that we have been 

 able to discriminate all such independent groups ? Such a claim 

 would be, to my thinking, extremely rash. 



