for Systematic Biology. 273 



seemed to be no need to alter our methods. But is it true — that 

 we still only see apparently fixed species ? Is it not a fact that, 

 as our collections grow and as our work becomes more exact and 

 our comparisons closer, variability is being revealed to us all along 

 the line until it is folly to continue to work as if we could ignore 

 it. Only continue your collection — say of Lepidoptera — and you 

 will find, as the collection of my friend Mr Elwes has amply 

 demonstrated, that the ' species ' melt away into apparently 

 endless ' varieties.' In the stony Corals, the variation is so great 

 that any attempt at genetic grouping into species can only be the 

 purest guesswork. No two workers would do it alike. Our 

 types have become ridiculous. They are not the types of any- 

 thing in nature : their special value is purely historical ; they 

 were the forms accidentally first described. In the great majority 

 of animal groups, then, variability is being revealed in some cases 

 but slowly, but in others the moment any competent person under- 

 takes to describe a really large collection. On all sides, indeed, 

 we are hearing demands for some more elastic system of work than 

 that supplied us by the Linnean binominal species name. The com- 

 plexities of the organic world due to the procession of life through 

 time are clearly too great to be investigated by so clumsy and 

 indefinable a unit. 



It is I think, then, obvious that we must cease to use a purely 

 ideal genetic group such as the Linnean ' species ' as a unit for 

 work. We must have one which more nearly fulfils the require- 

 ments already laid down, it must at least be an ascertainable 

 fact of Nature. One such unit we have and as far as I can see 

 only one, viz. that supplied us by the form, each form being 

 an aggregate of structural characters regarded in the abstract. 

 The different forms assumed by living matter are the units with 

 which we must work. 



Now this conclusion that the form is the only possible unit 

 for accurate scientific work is not only what my practical work 

 with the stony Corals resulted in but it is what we might have 

 arrived at theoretically. For only the form can be the unit in 

 evolutionary classification. Organic evolution means nothing 

 more than the gradual modification of relatively simpler forms 

 in various directions resulting in the production of relatively 

 more complex and specialised forms. The forms as such, that is, 

 in the abstract, are the only important things for the evolutionist 

 and morphologist. It is not easy to keep this clear and to 

 convince others of it. The fact that the individual concrete 

 forms of life possess the power of almost exactly reproducing 

 themselves, sometimes through many generations and over great 

 areas, somewhat dazzles us. These great armies of living beings, 

 reproducing themselves so far as we can see exactly, have natur- 



