1899] ORIGIN OF AUSTRALIAN FLORA 279 
towards the elimination of xerophilous forms. But it is not certain 
that the xerophilous vegetation completely disappeared from South 
Europe and Asia Minor, for it may well be—to cite a few instances 
only—that some Chenopodiaceae, and species of Helichyswm (and a fair 
number of these still survive in countries bordering on the Mediterranean), 
may actually be descendants from herbaceous members of the xero- 
philous flora, and when the present distribution of these genera is borne 
in mind, there is, it is submitted, at least some probability for this 
view. 
An objector will, of course, ask why it is, if the theory above 
sketched be true, that we do not now find species of Hucalyptus and 
Banksia and Dryandsa flourishing in deserts north of the equator. 
These, he will remark, are precisely the places to which a xerophilous 
flora would retire for shelter when driven by stress of climate from its 
former homes. Undoubtedly it would do so, if the desert then existed, 
and if no stretch of sea interposed to cut off the retreat of xerophilous 
species. The available desert country reaches from the Atlas Moun- 
tains across Arabia into Baluchistan, but from this the Sahara must 
be deducted, since it was, till quite recent times, submerged beneath 
the sea, and until the nummulitic limestone emerged from the waves, 
the ocean in which that extensive formation was laid down would be 
an effectual barrier to migration. Since Eocene times, however, this 
barrier has not existed; but it is not clear that Arabia and the drought- 
stricken regions bordering on it were deserts at the time when the two 
floras, xerophilous and hygrophilous, were engaged in their life-and- 
death struggle. If Perim can be taken as a guide—and there is no 
reason why it should not be—there is every reason to believe that 
Arabia enjoyed, in Eocene times, a climate much like that of Europe; 
and all we have to suppose is that the same change went on there as 
in Europe, namely that the climate became more favourable to hygro- 
philous forms, which were thus enabled to eliminate their xerophilous 
competitors, and that desert conditions subsequently prevailed, and the 
absence of Australian genera from the great northern deserts is 
explained. That this explanation presents difficulties is not to be 
denied, for the elimination, at least of arborescent forms, has been so 
complete, we should have expected that at least some few forms would 
have been able to adapt themselves to the altered conditions.’ Still 
the disappearance of these forms is scarcely more remarkable than is 
the disappearance of “ Northern” genera such as Quercus and Alnus 
and Salix from a country which must have afforded them, one would 
imagine, many eligible stations where they should have been able to 
survive. 
Let us now turn to Australia. And first, one must express a 
1 The Indo-Malayan and East Asian species of the Proteaceous genus Helicia may 
perhaps be cases of adaptation in the sense used above ; so, too, East Asian species of such 
genera as Drimys, Baeckia, Leptospermum, Lewcopogon, etc. 
