82 JOURNAL OF THE ROYAL HORTICULTURAL SOCIETY. 



of a variety is altogether independent of the botanical interest inherent in the 

 fact that the fern has suddenly departed from its ancestral habit of growth 

 and adopted an entirely new plan, whether merely eccentric or symmetrically 

 beautiful matters not in this connection. Some of the inconstant forms, 

 indeed, are in this respect more interesting than the constant, since in 

 them we see the evidence of a sort of conflict between the varietal and 

 normal forces, each of which occasionally obtains the mastery. In hybrid 

 plants ,we can to some extent understand this, since we know that two 

 specific forms are fused or rather mixed in the blood, and from Mendel's 

 and others' experiments it appears that the alliance is always subject to 

 dissolution at the critical reproductive period. 



Fig.. 34. — Pteris aquilina cristata. 



In fern varieties, however, we have to do with a pure species, an 

 individual of which at some period has formed, as it were, a new 

 architectural plan, which in constant sports is thenceforth strictly 

 adhered to, for the progeny, when they vary again, only vary in the 

 extent of development of the new type, or, in inconstant sports, is apt 

 to be replaced partially or wholly by the old one. The reason of these 

 sudden changes is utterly unknown to us ; none of the theories advanced 

 so far will hold water when confronted with actual facts, as regards 

 nature of habitats and environment generally, as indeed I have already 

 stated. Wherever ferns are abundant, the assiduous hunter finds these 

 "sports " mingled with hosts of the common types, and search for any of 

 the intermediate type which may suggest gradation is invariably vain. To 

 all intents and purposes they are " special creations," and considering the 



