of the Arctic and the Antarctic Faunas. 321 



is ill the tropical zone a considerable discontinuity in the 

 distribution-region of several liundre Is of genera of surface- 

 water animals occurring in the higher northern and southern 

 latitudes, and also in the distribution of very many — in any 

 case far more than a hundred — species of higher latitudes. 

 Either the discontinuity is really ))resent in the troj)ical zone, 

 or it appears to be so because of the present incomplete state 

 of our knowledge. The reasons for the first supposition we 

 have already discussed. Reasons in favour of the second 

 case undoubtedly exist also, and I believe that, with the 

 further progress of our knowledge, some regions of distri- 

 bution which are now believed to be discontinuous will be 

 found to be continuous. But that this may prove so in all 

 cases is an assumption which receives no justification either 

 from general considerations of probability or from the present 

 state of our knowledge. And it is not only our right, but 

 our duty to base our theoretical assumptions on the state of 

 knowledge at the time. But even if we go the length of 

 assuming that all bipolar genera and species may disa))pear 

 from science, will that affect a single point in our theories of 

 the relation of the arctic and antarctic faunas ? The Early 

 Tertiary roots of our present fauna remain the same whether 

 bipolarity exists or not ; and the similarity between the 

 faunas of higher latitudes also remains the same, whether the 

 distribution-regions of the genera or species be continuous or 

 not. 



The possibility, however, that a continuous distribution 

 through the deeper waters of the tropics may be demonstrated 

 for all the coeval surface-water inhabitants of the higher 

 northern and southern latitudes is simply not to be thought 

 of. The marine jilant-world of the higher latitudes exhibits 

 quite a pronounced bipolarity ; and the idea that this should 

 extend through the lightless layers of the tropical subsurface- 

 water beyond the equator cannot be entertained. The same, 

 of course, holds true of all animals which are directly depen- 

 dent on the plant-world. So that, as a general jjrinciple of 

 explanation, we may dismiss the hypothesis that the surface- 

 water animals of higher latitudes liave a continuous distri- 

 bution through the deeper layers of the subsurface- water of 

 the tropics. This, of course, does not prevent us from sup- 

 posing that the assumed connexion, which must have existed 

 at one time, does actually occur in some cases at the present 

 day, both in the deeper strata and in the surface-water. 



There is still another possible way in which the examples 

 of bipolar genera and species, which have hitherto been 

 present in literature, may disappear from science ; that is, by 



