564 TEST ACE A ATLANTIC A. 



single islands, and not having colonized even their respective 

 Groups. In allusion to this subject, I mentioned at p. 58, 

 that, out of the 176 Pulmoniferous Gastropods which have been 

 ascertained to inhabit the Madeiras, 5 only are found on the 

 whole five islands of the assemblage ; and I may add that out 

 of a somewhat greater number at the Canaries, only two 

 (namely the European H, lenticula, Fer., and the H. lan- 

 cerottensis, W. et B., which latter occurs likewise on the oppo- 

 site coast of Morocco) have been shewn as yet to be universal.' 



There may doubtless be many explanations, perhaps equally 

 plausible, of these phenomena ; but I must confess that none 

 commends itself so thoroughly to my mind as the possible 

 breaking-up of a land which was once more or less continuous, 

 and which had been intercolonized along ridges and tracts (now 

 lost beneath the ocean y which brought into comparatively inti- 

 mate connection many of its parts, — even whilst others, though 

 topographically near at hand, were separated by channels which 

 served practically to keep them very decidedly asunder. It is 

 on some such principle as this that I woidd account for the 

 Canaries appearing to be not only as widely removed from the 

 Madeiras as perhaps even the Cape Verdes are, but (while 



' Although I believe the same principle of segregation to be indicated in a 

 scaroely less degree at the Azores, yet since I have not myself collected in 

 that Group, and am bound therefore by the published statements of MM, 

 Morelet and Drouet, I have had no choice but to register as unwersal every 

 species to which they append the observation, ' Habite tout Varchijjel.^ Con- 

 sidering, however, that they never visited the island of S. Jorge at all, and 

 there is internal evidence that this exjjression is employed in the loosest 

 possible manner, I must be excused if I should fail to be convinced, in every 

 single instance, of its absolute truth. For whilst, out of the 176 species 

 (aboriginal and naturalized) which have been met with at the Madeiras, fire 

 only are found on the whole five islands of the Group ; and whilst out of the 

 189 at the Canaries (which are composed of seven islands), only two have as 

 yet been proved to be universal, it is preposterous to suppose that the 

 seventy-one species to which the Azorean fauna was brought up by Morelet 

 and Drouet should include no less than twenty-three which were detected by 

 them (and that too in the course of a single expedition, occupying but five 

 months) on the whole nine islands of an archipelago which is far more 

 widely scattered than either of those to which I have just called attention. 

 And j'et this is what we are compelled to acknowledge if their oft-repeated 

 assertion, embodied in the expression 'Habite tout rarchipel,' is to be looked 

 upon as undeniably true. To my own mind it is almost certain, not that 

 MM. Morelet and Drouet had unmistakeable evidence, in each individual case, 

 that the particular species which is thus reported had been ascertained 

 positively to exist on the whole, nine detachments of an assemblage the parts 

 of which are so difficult of access, and so remotely dispersed, as the Azores ; 

 but rather that, having met with it on an appreciable number of the islands, 

 they merely tlioiKjlit that it must be found upon the rest, and so did not 

 scruple to register it as occurring 'dans toutes les iles.' But wliether this is 

 to be received as conchisive, and as necessarily in accordance with facts, the 

 much more carefully compiled statistics of the neighbouring archipelagos 

 (the Madeiran data having been arrived at from researches which extend 

 over a period of nearly tifty years) may perhaps ser^'e to teach us. 



