179 



connecting links is the stumbling block which prevents the 

 three forms of the old Lastvea filix-mas (species of Wollaston) 

 from being universally accepted as species. The types 

 are distinct enough, and the distinctions can be precisely 

 defined ; each one is also capable of holding its own in the 

 struggle for existence, since wide districts can often be 

 found in which each is the prevailing fern, if not the sole 

 occupier, but, unfortunately, for precise botanical classifica- 

 tion, there are numerous individuals which do not belong 

 altogether to one type but present characters of two kinds, 

 or intermediate characters. In this and similar cases I think 

 we may fairly fall back upon the term sub-species. They 

 are, in fact, species in process of evolution, not yet com- 

 pletely separated. It is quite conceivable that in no very 

 long time the intermediate individual may die out (they 

 are even now much less numerous than the types), and we 

 should then have three fully developed species. In the 

 case of the Polystichums aculeatum and angulare, we have 

 two sub-species in which the process of separation has gone 

 a little further; that is to say, the intermediate individuals 

 are much less numerous, and in many localities are 

 scarcely to be found at all. In North Hampshire, how- 

 ever, they are frequent, and plants are often to be found 

 which are difficult to assign to one " species M rather than 

 to the other. To go back to L. filix-mas and its varieties 

 or sub-species, Mr. Rowlands, quoting Babington, refers 

 to affinis as a variety. This is the same thing as MoDre's 

 L.f. mas iucisa, and is by Newman given the same rank as his 

 Bovveril (pseudo-mas of Wollaston) and abbveviata (propiuqua 

 of Wollaston). It has also been described as a species 

 (Aspidium affiue and Polystichum affiue) by German botanists. 

 The figure of it in Moore's Nature-printed Ferns, how- 

 ever, appears to me to be much nearer to the typical filix- 

 mas than either pseudo-mas or propiuqua. I have certainly 

 seen wild individuals much more distinct than Moore's 



