172 TBANSACTIONS LIVEEPOOL BIOLOGICAL SOCIETY. 



the Mediterranean to the West Indies, and, most interest- 

 ing of all, the Eed Sea to the Mediterranean and vice- 

 versa through the Suez Canal. We have here an 

 interesting field for future work — and one that I may 

 appropriately recommend to a Liverpool audience. 



The great harriers interposed between the natural 

 regions of the marine flora are continental areas, the 

 depths of the sea, areas of different temperature in the 

 sea and deserts of sand and friable rock. We have an 

 excellent example of the continental barrier in the mass 

 of Africa separating the tropical Atlantic from the Indian 

 Ocean and stretching southwards into a colder region 

 which effectively blocks the way. Geologists tell us that 

 the continental areas are "permanent," that their antiquity 

 is at least so great that their beginnings are lost in " the 

 illimitable azure of the past." An examination of the 

 marine flora of the tropical Atlantic and that of the 

 tropical Indian Ocean discloses the fact that while the 

 genera are very largely common to both, the species are 

 in a high proportion different. Let us suppose an argu- 

 ment founded on this in favour of the high antiquity of 

 the genera of Algse— an argument that suggests itself with 

 fatal facility. It is true that these regions are as effectively 

 separated as if they were on different planets. But we 

 know that the present areas of temperature have been by 

 no means of such great duration as the continents, and 

 their changes have been far reaching. It is also admitted 

 as an important agent in determining such variations of 

 climate in the northern and southern hemispheres (or at 

 all events a constant accompaniment of such — which is 

 the same thing so far as our argument is concerned) has 

 been change of direction of the great ocean currents. Let 

 us suppose in the past — a past by no means so remote as 

 the birth of continents — a more genial climate in the 



