BY E. C. ANDREWS. 549 



for that family to preserve, therein, the purity of the types already 

 firmly established.. 



A slight difficulty arises in attempting to establish the order of 

 the appearance of the Eugenias and the Myrtles. Both contain 

 single flowers in certain species; the Eugenias are plants typically 

 with embryos possessing large cotyledons and small radicles, with 

 flowers of four petals, and with inflorescences racemose or clus- 

 tered; while the Myrtles are plants in which there is a typical 

 development of large radicles and small cotyledons, of flowers with 

 five petals, and of rather simple inflorescences. 



The evidence suggests that Eugenia was more nearly related to 

 the earlier type, by reason of its embryo, and the extratropical 

 range of Myrtus. On the other hand, the 4-petalled-flower of 

 Eugenia is a departure from type, and it would appear that there 

 were still eariier types, from which both these important genera 

 deployed. Moreover, in Rhodamnia and Rhodomyrtus among the 

 Myrtese, and in Leptospermum, Melaleuca and Callistemon among 

 the Leptospermeae, the leaves are generally strongly nerved, and 

 this interesting survival likewise suggests that nerved leaves existed 

 among some of the earlier, but now extinct, forms of the family. 



Apparently a land-bridge existed in the Cretaceous between 

 some portions of tropical America and tropical Africa. The land- 

 bridge between Asia and Australia appears to have been destroyed 

 in the Upper Cretaceous. 



This led to two important modifications of the Myrtaceae. In 

 Asia, fertile tropical conditions still continued, but there arose a 

 severe floral competition during later and Post-Cretaceous times in 

 Asia, owing to the deployment there of other vigorous and aggres- 

 sive families. In America, fertile tropical conditions continued; 

 while, in Australia, the vigorous outside competition was not ex- 

 perienced, but the soil there was neither so fertile, nor was the 

 climate so genial, as in the other regions. These conditions were 

 most noticeable in the south-west of Australia, where great sandy 

 expanses of land existed* ; while the northern portion of Australia 

 probably was very similar to the other portions of the tropics. 



* See also A. R. Wallace, " Island Life," (1892), pp. 487-508. 



