372 E. C. ANDREWS. 
absence of similar types and the monotypic nature in New 
Zealand of these genera. Canavaliais evidently a common 
maritime type. 
The genera in New Zealand which may have arrived 
from the north west by a land connection between New 
Zealand and the tropical continents, are Hdwardsia and 
Oarmicheelia, with the monotypic Notospartium and Coral- 
lospartium. The two tribes represented are Sophores and 
Galegeze. It seems impossible from the evidence available. 
to avoid the conclusion that New Zealand was isolated 
from the great tropical lands before the differentiation of 
Leguminosze into Mimosez, Cesalpinieze, Dalbergieze,, 
Trifoliese and similar tribes, but not necessarily before the 
development of Sophorez and Galegez. 
Another point needing consideration is the possibility of 
the existence of soil or climate barriers. Thus neither 
Oastanospermum nor EHntada could cross a subarid sandy 
waste. 
The next point to be considered is the factor of marine 
transportation. Throughout the tropics are many legumi- 
nous species, which are either identical within the various 
countries considered, or are so much alike that they are: 
only separated systematically by their geographical station. 
In this connection the more striking examples of the African 
legumes have been cited in an earlier chapter. An analysis. 
of Australian, Asiatic and Brazilian Leguminose reveals 
features equally startling in nature. 
Guppy has made a long and careful study of the histories 
of strand plants and sea currents, and an analysis of his. 
observations’ suggests that these plants are best con- 
sidered, not as examples of arrested development predating 
the separation of the great tropical land masses, but as. 
* Guppy (35 and 36). 
