1888.] Evolution as illustrated hy the Geogr. Distrib. of Plants. 279 



The theory of simultaneous variation accounts for the appearance 

 of the same species both over a continuous floral area and in widely 

 separated areas as follows : 



From an ancestral species probably equally widely spread, having 

 everywhere a capability of varying in a limited number of directions 

 and a tendency to vary in still fewer, similar circumstances have over 

 some parts of the area favoured the production and survival here of 

 one variety and there of another. The localities in which the same 

 varieties have been favoured, are not necessarily conterminous or 

 even in the same area. In other places again, circumstances not being 

 favourable to the new varieties, the species has disappeared. 



The difference of the flora of regions similarly situated as far as 

 physical conditions are concerned would then be explained as follows : 

 Two areas, say Mexico and South Africa, have formerly been subject 

 to different physical conditions and have thus developed different 

 specific and generic forms. Owing to change of circumstances the 

 physical conditions of these two regions have subsequently become 

 assimilated to each other. The conditions requisite for the same flora 

 are not then here present. Similar conditions operate on dissimilar 

 forms. In this way not identical but parallel species are produced, 

 resembling each other outwardly but not in the structure of the part 

 .on which classification depends. Thus in Mexico we find the succulent 

 Cactaceae, in South Africa the succulent Euphorbeaceas. 



The theory of the individual or dual origin of species on which 

 the migration theory rests, or rather, perhaps, on which the denial 

 .of the plurality of seats for the same species depends, seems to 

 be surrounded with difficulties. 



According to this theory either the new species is produced per 

 saltum or it is produced from a single line of ancestry which has 

 left no collateral issue. 



Either hypothesis is inconsistent with the admitted prevalence of 

 . cross-fertilization. 



If variations are all-round variations, following no perceptible law 

 .and selected by environment only, then where cross-fertilization is 

 the rule every attempt at va-riation would be swamped, or only those 

 new species Avould be produced which are self-fertilized, so that cross- 

 fertilization would either become unknown or very rare. 



The production of new species ]per saltum is hardly worth serious 

 consideration. In the case of all animals and dioecious plants it 

 would involve the accidental and simultaneous production of two 



