296 DISTRIBUTION [CH. 



is by no means what one would expect at first glance, since it 

 might reasonably be supposed that salt-water areas, mountain 

 ranges, and wide tracts of arid country would prove insuper- 

 able barriers to the migration of plants of fresh water 1 . This 

 difficulty was so keenly felt by Alphonse de Candolle 2 that he 

 was forced to the conclusion that the facts of the distribution 

 of aquatic species were scarcely explicable except on the theory 

 that there had been multiple centres of creation. 



For the sake of simplicity we may first consider the distri- 

 bution of hydrophytes within a single country such as our own, 

 which, on a small scale, presents the same difficulties. A partial 

 solution of the problem might be reached, if former con- 

 nexions between the existing river basins could be postulated, 

 in order to account for the uniformity of their floras. But the 

 history of the land surfaces at once disposes of this possibility. 

 In the words of Clement Reid 3 , whose labours disinterred so 

 much of the geological history of our present flora, "Each 

 year's work at the subject makes it more clear, that ever since 

 our climate became sufficiently mild to allow of the existence 

 of our present fauna and flora, many of the river-basins of 

 Britain have formed isolated areas." It is no doubt possible 

 that floods may, in some cases, give a species the opportunity 

 of introducing itself into fresh situations; an extension on a 

 small scale, of the area of distribution of certain aquatic plants 

 was induced by the great floods in East Anglia in 1912. 

 Furthermore, floods may even, as Guppy 4 has suggested, oc- 

 casionally bring about an exchange between plants belonging 

 to different rivers traversing extensive level regions. But such 

 effects can never be more than partial and they will not explain 

 the passage of any species over a well-defined watershed 5 . 



1 This paradox was noted by Darwin, C. (1859). 



2 Candolle, A. P. de(i855). 



3 Reid, C. (1892). 4 Guppy, H. B. (1906). 



5 Dr Guppy has suggested to the writer " that the permanent head- 

 springs of rivers in elevated regions where the sources of rivers may lie 

 in proximity would serve as centres of dispersion for the same plants in 



