176 TEANSACTIONS LIVEEPOOL BIOLOGICAL SOCIETY. 



showing that there is such a subject and that it deserves to 

 be pursued. For this purpose I have selected three regions 

 of the ocean which are well cut off from each other, which 

 have little or nothing in common except the fact that their 

 Algse have been well collected and studied. These regions 

 are the Arctic Sea (taking Kjellman as the basis of calcu- 

 lation) ; the West Indian region (based on my own 

 catalogue) ; and the Australian region (based on Sonder's 

 catalogue in Baron von Mueller's Fragmenta Phytographice 

 Australia, vol. XI., Supplement and also in A dditamenta) . 

 The first and the second are purely natural regions, but 

 the third is hardly so. Australia is both tropical and 

 south temperate, but the fact may be urged in favour of 

 our treating it as a whole that the portion of the coast 

 best known at present is south temperate. I wish for the 

 purpose of this table it had been possible to so delimit 

 Australia as to exclude its tropical forms, but it was hardly 

 possible to do this with accuracy. Nevertheless, the 

 whole region is well cut off from the others, and the fact 

 that the whole of its flora is included would concede 

 chances against my argument for the geographical diversity 

 of algal forms. So let it stand. Before entering upon an 

 examination of this table, I would point to a discrepancy 

 between the totals of forms common to the "West Indies 

 and Australia now given and those given in my catalogue 

 of West Indian Algae. The discrepancy arises from the 

 fact that the present enumeration is based principally on 

 Sonder's Catalogue and the former one on Agardh's 

 Species, Genera et Ordines Algarum, and on other 

 books. Sonder's Catalogue is more exhaustive than the 

 Agardhian enumeration for the reason that part of the 

 latter (the Phseophycese) was done long before Sonder's 

 catalogue. There is here a difference of four. Again under 

 the Floridese there is a difference in excess of ten, arising 



