1544 PHANEROGAMS. 



few forms are known ; but when we come to the Dakota stage of 

 the United States, which is usually correlated with the lower part of 

 the Upper Cretaceous of Europe, Dicotyledons formed a large pro- 

 portion of the flora, more than three hundred species having been 

 described in 1885, which belong to both sections of the sub- 

 class. This abundance of forms pointing to the conclusion that 

 the origin of Dicotyledons must be looked for in considerably 

 earlier epochs. 



Very different views are held by authorities as to the classification 

 of Dicotyledons, some dividing them into several primary groups, 

 while others only admit the three divisions Apetalae, Polypetalae, and 

 Gamopetalse. Dr Endlicher, however, makes only two divisions — 

 Choripetalae and Sympetalse, — and since this arrangement is adopted 

 by Dr Schenk in his ' Palaeophytologie,' it will be followed in this 

 work. 



The larger proportion of the Cretaceous and Eocene Dicotyledons 

 belong to the Choripetalse, and especially to those orders in which 

 the flowers are fertilised merely by the agency of the wind ; the 

 more specialised Sympetalae, which require the aid of insects for 

 this purpose being in great part of later date. The labours of Baron 

 von Ettingshausen have shown that the Cretaceous and early Ter- 

 tiary Dicotyledonous (and also the Monocotyledonous and Gymno- 

 spermous) flora was almost similar throughout the world ; and that 

 its present division into provinces is in great part due to the influ- 

 ence of one or more glacial periods. Bearing in mind this world- 

 wide distribution of so many of the genera, it will be unnecessary in 

 most cases to do more than mention their earliest established occur- 

 rence ; and owing to the immense number of families only the more 

 important types can be even mentioned at all. Space will not 

 permit any reference to the gradual differentiation of the existing 

 floras, or to the interesting distribution of modern arctic types in 

 the Pleistocene, for which the reader must refer to the works of 

 Baron von Ettingshausen, Sir J. W. Dawson, and other writers. 

 It should be mentioned that in the case of trees known only by 

 the leaves the generic determination may be open to doubt in 

 some cases. 



Division A. Choripetal^e. — This large group is characterised by 

 the petals being either absent, or if present not united together. 



Order i. Amentace^e. — This order comprises trees which are 

 typically apetalous with the male flowers in the form of catkins. 

 It includes five families, and comprises some of the earliest known 

 representatives of the subclass. The first family, Casuarinidce, is 

 now represented only by the well-known Casuarina of the Australian 

 and the eastern Malayan regions : this genus apparently occurs in the 

 Lower Tertiary of Sumatra, but it is very improbable that the Euro- 



