414 NATURAL HISTORY OF PLANTS. 



thought to be unrepresented in Australia. 1 The series Liquid- 

 ambarecs is represented by Biicklandia in India and Sumatra, by 

 Liquidambar in America, Asia, and Malaysia. Brexia and Roussea 

 are restricted to the Mascarene Islands and Madagascar ; but the 

 other genera of Brexieee represent it by some species in Australia 

 and Tasmania. All the Francoece are Chilian ; the Myomrandrea 

 are confined to the Cape and the islands off the east coast of Tropical 

 and South Africa; and the Codiece are exclusively Oceanian. The 

 Platanece are only found in North America and Mediterranean Asia. 

 Of the four known Datiscete one alone is North American ; the 

 others belong to South-eastern Europe, West and Central Asia, 

 India, and the Indian Archipelago. The Pittosporece are all 

 Australian, excepting the genus Piitosporum, which is also found in 

 all hot countries of the Old World, extending obliquely from the 

 south of Africa to the north-east of China. The Ribesiece inhabit 

 the temperate regions of Europe, Asia, and North America, Mediter- 

 ranean Europe, and the Andes of South America. 



An order of the size and constitution of this must necessarily 

 present numerous affinities. 2 Some of these are so close that it is 

 extremely difficult to distinguish Saxifragacece by any absolute cha- 

 racters from several other natural orders, such as Rosacea* and Crassu- 

 lacece? as we have shown in treating thereof. Asfilbe, on the one hand, 

 Bauer a, Eucryphia, and Pterostemon 5 on the other, link Saxifragacece 

 and Rosacea, the latter being distinguished from the former by its 

 general (though not constant) lack of albumen. By Philadelphu8 y 

 Argophyllum l and Carpodetus, on the one hand, and Cassipourea, 

 Anisop/iyllea, and Crossostyles, on the other, the order Saxifragaceee is 

 linked to Myrtacea, RMzophorece, and Melastomacea respectively ; by 

 Thamnea, to Memecylece ; 6 by Fendlera, Raleighia, Abatia, and Crypte- 



1 The existence of the plant we have pro- : For details on these points we refer to onr 



visionally named Osbrearia australiana (in Observations siir les Saxifragees (in Adamonia, 



Adansonia, x. fasc. 4) was tlien unknown. v. 282 ; vi. ]). 



Though still hut very imperfectly known, it 3 See above, i. 414. 



seems to us a Hamamelid with a partially * See ahove, iii. 317. 



inferior ovary, a capsule dehiscing into two 5 See Adansonia, ix. 245. 



emarginate valves, and numerous flowers in long 6 See Adansonia, iii. 321). 

 spikes, terminating branches covered with alter- 

 nate simple entire glabrous lanceolate leaves. 



