290 Proceedings of the Royal Physical Society. 



many names ; and so the insertion of a third name, which 

 might not be insisted on as a matter of instruction, was 

 convenient. Having had a little experience both of being 

 examined and of examining in turn, I can cordially sym- 

 pathise with the student, preparing for his trials, in any 

 complaint of " overpressure." But the question of how far 

 he is to master the details of nomenclature, or how many 

 scientific names he is to cram up, seems to me to be one for 

 the teachers and examiners rather than for the working 

 scientists. From my own experience, I rather fear that the 

 ordinary student is more oppressed by the recent advances 

 in comparative anatomy and embryology than by the 

 number of generic names which he has to learn. 



But the trinomial system chiefly discussed at the meeting 

 referred to is not connected with " subgenera," but with 

 " subspecies." Bearing in mind the only practicable defini- 

 tion of a species, namely, a form which has not yet been 

 positively shown to intergraduate with any other — (if we are 

 evolutionists, we believe that the connecting links have been 

 lost ; if we are not, we believe they have never existed) — it 

 has long been a recognised fact that under these species 

 subordinate forms or " subspecies " occur, often distinct 

 enough when extreme forms are compared, but passing into 

 each other when a well-selected series is put under observa- 

 tion. These " subspecies " belong to quite a different 

 category from accidental varieties or sports, being char- 

 acteristic of geographical regions where they breed true, 

 and in very many instances have been named, described, 

 and long considered as true species. They are, in fact, in 

 the eye of the evolutionist, species between which the links 

 have not been lost. 



That these are facts, and that their expression is not 

 provided for in a strictly binomial nomenclature, is acknow- 

 ledged by every one ; the question comes to be, how best to 

 give them expression ? Hitherto the usual solution of the 

 difficulty, without interfering with the integrity of the 

 binomial system, has been to give subspecific or varietal 

 names, and to append them to the proper name of the 

 species, the word " variety," contracted " var.," being inter- 



