i6 



finds and those of others, and although this record also 

 embraces a number of varieties subsequently improved in 

 type by selection of the progeny of the wild finds, those 

 raised are always so described, and cannot be confused 

 with the wild finds themselves. Moreover, and this as 

 another material point, the plants have invariably been 

 collected as living plants, and not merely as herbarium 

 materia], and having been carefully preserved and multi- 

 plied by off-sets and bulbils, the bulk of them are still 

 existent in British Fern collections. It is also worthy of 

 note that although all our native species, without exception, 

 are also indigenous to other temperate countries, and some 

 of them almost ubiquitous, this particular study has almost 

 exclusively been confined to the British Isles, and the 

 specimens extant are, therefore, with but a few exceptions, 

 truly British wild plants. We have, therefore; an immense 

 mass of purely unsophisticated material to deal with in 

 considering the comparative variability of this section of 

 plant life under wild and cultural conditions, while the 

 records are incontestible as regards their wild origin, and 

 in this particular connection are of the greater value, as 

 they were not compiled with any view to establish a theory 

 of variation, and are, therefore, quite unbiassed. The latest 

 and most complete record of this kind is found in the 

 late ^Ir. E. J. Lowes' "British Ferns," 1891, published 

 by Swan, Sonnenschein & Co. This is a descriptive list, 

 giving the date and locality of the discoveries, and the 

 names of the finders, together with enhanced types raised 

 by selection, these being marked " raised," with the raiser's 

 name. It is noteworthy that these latter are, in every case, 

 of similar type to the wild parent, though this may be as it 

 were emphasized, i.e. more marked. For a really new type 

 we have invariably to look to the wild fern. This fact, 

 ripely considered, goes far to prove greater variability 

 under wild conditions than under culture, instead of the 

 generally accepted converse case. We may now come to 

 the consideration of how far this fifty year record demon- 

 strates the proportion of wild sports to those which may be 



