NOTICES OF BOOKS AND MEMOIES. 311 



male examples, and some only by female ones. By the liel^D of 

 many different herbaria I have partially filled up the deficiencies, 

 but there still remain far too many species of which the female 

 flower is miknoYm." 



In opposition to Mr. Bentham, but in accordance with many 

 other botanists, SmilacecB here ranks as an order distinct from 

 Liliacew, a position to which it appears fully entitled. M. De 

 CandoUe gives an interesting account of the morphology of 

 the leaves, inflorescence and flowers in the Order ; he recognizes 

 three genera, Heterosmilax, Smilcw, and Fihipoi/omuu, the first and 

 .last with but 5 species each, whilst Sunla.r possesses no less 

 than 209. (Hhjosmilax, Seem., is reduced to Heterosmilax, and 

 Nemesia, Pleiosmilax and Coprosmanthus to Smilax. This last great 

 genus is divided into four sections defined by the number of 

 stamens (six or more than six) the existence of two or one ovule 

 in each cell of the ovary, and the concave or reflexed perianth- 

 leaves of the male flower ; in the further divisions the ^nature of 

 the inflorescence affords characters. As to species of Smila.v, 

 Kunth's monograph, published in 1850, contained 193, of which 52 

 were doubtful and 12 more only known by name ; of the 209 in the 

 present memoir only 22 are doubtful or unknown. The species 

 here described for the first time are 55 in number, whilst 30 of 

 Kunth's have been reduced. The author says that he has abstained 

 from publishing or even naming a number of new species, of which 

 he has seen merely fragments in herbaria, as to do so would be to 

 throw into science mere puzzles and difficulties. Would that all 

 descriptive botanists had an equal restraint ! 



Restiacew is another very troublesome family, which Dr. Masters 

 has already much helped to elucidate by previous memoirs. In 

 this complete monograph the sum of the genera is 20, and of the 

 species 234. They are distributed between the Cape and Australia. 

 South Africa has 13 genera and 156 species, Australia 10 genera 

 and 77 species, only 3 of the genera and no species being common 

 to both areas. A curious fact in geographical botany is the occur- 

 rence of a single species, Leptocarpus chilensis, at Arique, on the 

 Rio Negro, in Chili. The Australian species have been quite 

 recently monographed by Mr. Bentham m the last volume of his 

 ' Flora Australiensis,' published this year; and Dr. Masters differs 

 but slightly from that standard work, which is quoted thoughout 

 as " ined," though it has turned out to be the first published. 



The Meliacecc occupy half the volume, and are monographed by 

 Casimir DeCandolle. Since the original account in the ' Pro- 

 dromus' this order has been the subject of a fine memoir, in 1830, 

 by A. de Jussieu ; and of a revision, so far as genera are concerned, 

 in 18(32 by Hooker in the ' Genera Plantarum.' From this latter 

 arrangement that here adopted does not greatly differ, but the genera 

 Cipadessa, TurrmantJius and Azadirachta are kept uj), whilst 

 Calodryinn is reduced to Quivisia, Epicharis to Bysoxylum, Milnea 

 to Af/Iaia, and MoscJwxijlum to TrichiUa. Vavcca is transferred 

 from the MeliccE to the TricliUece. The number of new sj)ecies is 

 large, the standard of specific rank being considerably lower than 



