CONCLUSIONS 



The fern flora of Madeira contains 42 known species of leptosporangiate ferns, 

 38 of which have been studied cytologically at the time of writing (Manton, unpubHshed) . 

 Madeira itself is a mountainous island off the coast of Africa placed in a subtropical 

 latitude quite outside the area of glaciation and which has therefore enjoyed a vegeta- 

 tion-cover without interruption since it emerged from the sea in Tertiary times. Of 

 its 42 species of ferns 22 have been attributed to species also present in Britain, the 

 remainder being species characteristic of warmer, in some cases of tropical, climates. 

 A few are endemic. 



If now the British species are examined in Madeira, Love and Love's generalization 

 is borne out since out of 22 species only about 7 are polyploid, i.e. one-third 

 instead of half the relevant part of the flora. On the other hand, if we include the 

 non-British species as well, the total incidence of polyploidy though slightly less than 

 in Britain (Table 9) has another, perhaps even more significant, diff'erence, which is 

 not shown in the Table. In Britain (the Ophioglossaceae apart) the grade of polyploidy 

 is in most cases that of tetraploid with only two cases of hexaploids in Cystopteris and 

 Polypodium respectively. In Madeira on the other hand, among the non-British species 

 so far analysed there are two octoploids and a decaploid clearly recognizable as such 

 by the fact that they belong to well-known British genera: the species in question 

 are Asplenium aethiopicum (Burm.) Bech. {=A. furcatum Thunb.) n=i^/^ (octoploid), 

 Polystichum falcinellum (Sw.) Pr. ^=164 (octoploid) and Adiantum reniforme L. ^=150 

 (decaploid). Further, while in Britain the polyploids are in most cases still in geo- 

 graphical contact with lower-numbered relatives (from which circumstance they are 

 thought to be recent), in Madeira these high chromosome numbers all belong to 

 undoubtedly ancient and isolated types totally devoid of local relatives and in 

 some cases, notably Polystichum falcinellum and Adiantum reniforme, either endemics 

 or nearly so. These must therefore be ancient species with a long past history 

 which, for this reason, is no longer spread before us as in so many of our own native 



species. 



My personal conclusion from this is that polyploidy as such is not in itself either 

 ancient or modern or an adaptation to cold or any other single climatic or ecological 

 factor but that it is correlated rather with climatic or geographical upheavals however 

 caused. Under stable conditions, the natural spread of species is probably accompanied 

 by some, though perhaps infrequent, polyploidy as new species come into contact with 

 old ones and hybridize with them. In a relatively undisturbed flora the incidence of 

 polyploidy might therefore be expected to be low. Under changing climates or topo- 

 graphy, on the other hand, the opportunities for hybridization and therefore for 

 allopolyploidy can hardly fail to be increased. The nature of the significant climatic or 

 topographical changes may be very varied and include perhaps cold, heat, drought, 

 inundation, mountain building, volcanic action, changes in the distribution of land and 

 sea or any other vicissitude which may affect the whole earth or portions of it. All of 

 these may be expected to leave their mark on the evolution of vegetation. 



If this is so, the high polyploids of Madeira may perhaps record world events of a 

 different type and older than the Ice Age whereas the polyploids of Britain and of 

 north-western Europe in general may still bear in their distribution and physiological 



283 



r 



