Australian Flowering Plants. 173 



the Distribution and Development of the plants in the groups 

 Legummosse 1 and Myrtacese. 2 The distribution of these groups 

 alone, considered without reference to the distribution of other 

 families of animals and plants, suggested that the great tropical 

 lands had been connected directly with each other as from 

 Tropical America to Tropical West Africa, thence through 

 Madagascar and Asia through the Malay Archipelago to Aus- 

 tralia. But the study of the Amentales, the Composite, the 

 Ericaceae and many other families, and orders, suggested that 

 the discussion of the legumes and myrtles alone had shown 

 only one aspect, out of several possible, of the story of angio- 

 spermous distribution. The distribution of animals suggested 

 another line of evidence and it was decided finally to obtain 

 the independent testimony of various witnesses as to the pos- 

 sibility of land connections since the early Cretaceous and to 

 coordinate the evidence so secured. Thus the student either 

 of isostasy or of insular floras and faunas is driven to accept 

 the doctrine of the permanence of the larger features of the 

 ocean basins since the dawn of the angiosperms. For example, 

 it is known that the rock structures underlying the ocean basins 

 are heavier than those of the continental areas considered with 

 respect to unit columns lying above a certain plane about 100 

 miles below the sea-level and such a peculiar adjustment of 

 structures has not been attained in a hurry. Again Wallace 

 has observed 3 that the sedimentary deposits of the land areas 

 are suggestive of shallow-water conditions, while there is an 

 absence of continental areas of deposits such as are found 

 to-day on the bottom of the deep sea. 



Throughout these pages the idea of a great two-period dif- 

 ferentiation of climate has been made prominent, nevertheless 

 it must not be forgotten that the distribution and development 

 of almost all of the plant types considered in the present report 

 could be explained satisfactorily on the assumption of the 

 co-existence of waste places, of barren sandy tracts and of 

 subarid to arid regions with the widespread mild and moist 

 conditions of the Cretaceous and Eocene, such waste, barren, 

 and desert areas increasing in size at various periods, notably 

 since the Eocene or Miocene. Indeed it is necessary to postu- 

 late the uninterrupted existence of such desert conditions or of 

 barren soils throughout the later history of the angiosperms, 

 say, since the Lower Cretaceous, otherwise it is difficult to 



'E. C. Andrews, Proc. Eoy. Soc. N. S. Wales, vol. xlviii, pp. 333-407, 

 1914. 



*Tdem, Prop. Linn. Soc. N. S. Wales, vol. xxxviii, pp. 529-568, 1913. 

 3 Island Life, 1892, pp. 103-105. 



