NOTICES OF BOOKS. 187 



absent as an order in West Australia, tbongli the errors under this last 

 head are unimportant as compared with those under the other five. 



In fine, although a large number of additional facts are now to 

 hand, it may be doubted whether, in the discussion of the Austra- 

 lian, New Zealand, and Antarctic flora the general results of Dr. 

 Engler are at all an advance upon Sir J. D. Hooker's Essay 

 of 1853. 



Madagascar is very imperfectly known botanically ; nearly all 

 the collections have hitherto been made round Antananarivo and 

 on the routes thereto ; the plants received in England during the 

 present year prove that it is at present vain to found an argument 

 on the absence of any genus from Madagascar. 



The discussion of the African flora is one of the best of Prof. 

 Engler's interesting book ; he has wrought out the analogy in 

 structure, climate, and vegetation between Africa south of the 

 Sahara and hither-India south of the Gangetic Plain, the wet 

 West Malabar being paralleled in the wet West Cameroons. 



The tabulation of genera for South-East Asia occupy six large 

 quarto pages, but they are confined to a half-dozen isolated orders. 

 Prof. Engler says, in the text, that he has tabulated Cogniaux's 

 Cucurhitacea ; but in the table only four genera appear. The 

 selection of the orders and genera for tabulation is so arbitrary 

 that the only general inference Prof. Engler is enabled to draw 

 from them is that the genera overlap each other a good deal, and 

 that it is not easy to lay down limits for phytographic sabareas, an 

 inference which might perhaps have been arrived at without tabu- 

 lation. Attention is drawn to the fact that the oaks and ^Dines, 

 though continued from the Himalaya to low levels in the Malay 

 Archipelago, do not appear in Ceylon and South India, even in the 

 mountains, on which (and other parallel facts) the condensed 

 Indian geology of Blanford and Medlicott may supply Professor 

 Engler with food for further reflection. 



Any article treating on the geographic distribution of the 

 plants of the world must be built in chief on secondary evidence, 

 and the results will be invalidated in numberless cases by errors for 

 which the compiler is not responsible. The genus Dicellostyles is 

 cited by Engler as a' solitary instance of a genus of two species, 

 whereof one is endemic in Ceylon, the other endemic in the Hima- 

 laya ; but it ha^Dpens that the Himalayan Kijclia jujuhiflora, Griff., 

 is a true Kydia, and has been erroneously referred to Dicellostyles 

 by Bentham and Masters. Dr. Engler cannot be criticised for 

 accepting such authority; but many anomalies in distribution 

 which stand in our books disappear on studying the plants them- 

 selves. 



In estimating the richness of tropical floras. Prof. Engler has 

 perhaps made insufficient allowance for elevation, i.e., whether 

 mountains, affording great variety in level, are included under the 

 area treated of. Thus he speaks of the comparative poorness of 

 the flora of China, which is true of the great plain of cultivation : 

 this, like all tropical plains of one level and of one character of 

 soil, affords a very uniform flora, consequently not rich in species 



