138 



tudes, at all events) be a mere conjecture or romance. 

 Such a notion, however, I would not for a moment 

 entertain, for there is too much direct evidence in 

 support of distinct epochs of diffusion, to allow of any 

 hypothesis, when endeavouring to account for the 

 phenomena which we now behold, to supersede the 

 assumption of a once continuous tract. No matter if 

 we be compelled to suppose, whilst attempting to in- 

 terpret what we see, that the disseminating current has 

 flowed in exactly opposite courses, at different and 

 remote periods, over the surface of that ancient land, 

 seeing that the fact (if such in reality it be) remains 

 untouched, that the land itself is at any rate there. I 

 am not, however, prepared to assert that the opinion at 

 which I had independently arrived, from the insect 

 statistics, does positively require a northerly prolonga- 

 tion of that area beyond the line of the central Mediter- 

 ranean districts ; yet, after making every possible allow- 

 ance for accidental introductions since the subsidences 

 have taken place, there is still left a large residuum 

 which I am convinced can never be explained (unless 

 the doctrine of specific centres be a myth) except 

 through the means of ordinary and regular migration 

 over an unbroken continent. Nevertheless, though I 

 would not presume, from insufficient material, to insist 

 upon an extension of this Atlantic region into higher 

 latitudes than those which I have just referred to, I 

 must express my individual belief that, the more the 

 subject is examined, with reference to the distribution of 



