250 



THE ZOOLOGIST. 



imagine, and this, judging from their great dissimilarities, may 

 well be the case. 



There appears to be no reason why evolution may not, even 

 does not, act in a similar way on types closely allied or even 

 somewhat divergent, and if the process is repeated (and it would 

 be presumptuous to altogether deny the possibility of such a 

 proceeding), one would get parallelism (and even perhaps con- 

 vergence) carried out to a far greater extent than appears to be 

 credited or even taken into consideration. In fact, the possibi- 

 lities of this style of evolution are somewhat staggering, that is, 

 if we can admit that there is perhaps an inherent potentiality 

 in more or less similar forms to be affected by the causes of 

 evolution to similar and somewhat similar degrees. 



In this way kinship, or even germ plasm resemblance and 

 fertility, need not depend on recent near blood relationship, and 

 members of the same species might be unrelated in origin ; to 

 what degree of remoteness, it would be impossible to surmise. 



L. multitceniata sterile with 



L. rubicunda (A). 



L. m. sterile with 



(B) + L. r. (A) = fertile. 



In this way we have seen how one strain (B) of L. rubicunda 

 is produced from L. multitceniata and sterile (partly) with this 

 form, and on account of this, and also because the former is 

 fertile with another independently evoluted strain (A) of L. rubi- 

 cunda, it may be assumed this "A " strain is also sterile (partly) 

 with L. multitceniata (this should be tested). 



We have thus a case of convergence in L. multitceniata pro- 

 ducing a form of L. rubicunda indistinguishable from and fertile 

 with the original L. rubicunda strain. If this breeding had not 

 taken place under observation in partial confinement, who would 

 have felt inclined to admit the possibility of a genealogical tree 

 similar to the above ? 



I have mentioned above that the supposed fertility between 



