Australian Flowering Plants. 181 



development of herbs among the old luxuriant tree forms in 

 the temperate regions ; second, by the great development of 

 herbs and undershrubs in the warm arid and subarid regions, 

 and third, by the development of vines and herbs and epiphytes 

 in the mild and moist regions. 



An alternative explanation may be found by admitting the 

 existence of deserts of limited area even in the Cretaceous and 

 Eocene and in the presence of large areas of rocky barren soils 

 such as those now found in South Africa and Australia. 



A much later period of climatic zoning throughout the world 

 is evidenced by the development of deciduous trees in the 

 northern lands and the great development of xerophytic tribes, 

 subtribes, genera and species in both the northern and southern 

 hemispheres. 



In other words the present classification of the flowering 

 plants suggests that the great alliances of orders are of long 

 standing, that the families are due to an early world-wide dif- 

 ferentiation of climate, and that the present xerophytic and 

 deciduous development is due to the later and local modification 

 of old and well-established families and tribes. 



III. The Difficulty attendant on the Determination of Fossil 

 Flowering Plants by Leaf Remains only. 



It is necessary for true progress in paleobotany that the 

 proper determinations of the angiosperms should be agreed 

 upon both by botanists and paleobotanists. In this connection 

 it seems peculiar that the leaders of the botanical world, past 

 and present, both were and are accustomed, respectively, to 

 demand full and abundant material before referring a plant, 

 previously undescribed, to its proper genus and species.* Paleo- 

 botanists, on the contrary, seem content with fossil leaves 

 alone for generic determinations of flowering plants although 

 those fossil remains may date back to the Cretaceous. It is 

 not that the plants so named may be referred to new genera 

 for stratigraphic purposes, or that they are referred to genera 

 whose names suggest a general likeness in leaf characteristics 

 to certain types of modern genera, which causes the confusion, 

 indeed upon first thought this may seem a procedure unlikely 

 to lead to serious error, but the application of this method has 

 led paleobotanists on the evidence of leaves alone to refer many 



* See also a paper by A. A. Hamilton on "The Instability of Leaf Mor- 

 phology in its relation to Taxonomic Botany," Proc. Linn. Soc. N. S. W., 

 vol. xli, pp. 152-179, 1916, and references therein. — [This note was received 

 after the article was put into type — Eds.] 



