22 NATURAL HISTORY OF PLANTS. 



ledonous embryo. 1 These constant or nearly constant characters are 

 those of the order. To split it into tribes or series others have 

 been used, which, though no doubt not unexceptionable, are the 

 only ones at present available in the study of a group usually 

 represented in collections by more or less imperfect specimens. 2 

 We have hence adopted provisionally the classification proposed by 

 J. Hooker & Thomson, based mainly on the structure of the fruit 

 and seed. 3 In certain Menispermacea the embryo is exalbuminous as 

 in Tacliygone ; in others the albumen envelops it. In Cocculus and 

 Menispermum, the two cotyledons are applied to one another by the 

 whole extent of their inner faces, while in Chasmantfiera or Burasaia 

 they diverge at their very insertion, leaving an angular space of 

 variable breadth into which the albumen penetrates. Hence we get 

 the three series, Cocculece, Cliasmantherece, and Pacliygonece ; the fourth 

 series, Cissampelidea, rests on totally different characters, drawn from 

 the structure of the flower and the number of its parts. These are 

 not regularly trimerous. 4 At the same time the androceum consists of 

 stamens which cohere for a certain distance to form a column on top 

 of which the anthers, sessile or subsessile, are united into a sort of 

 terminal cap. The gynseceum is unsymmetrical, reduced to a single 

 unilateral unilocular ovary. 5 We may then sum up the general 

 character of the four series as follows. 



I. Coccule^. — Seed with a narrow embryo, whose cotyledons are 

 applied to one another, surrounded by albumen. Drupe with stylar 

 cicatrix subterminal or more frequently brought down near the base 

 of the incurved fruit. Stone with an internal projection of variable 

 form from the ventral angle on which the seed is moulded bowed or 

 inflexed. Carpels 3-6, more rarely 9-12 (8 genera). 



more or less intimately united. This view seems 2 Whether because only the male or female 



to be purely theoretical. In the other genera, plant of a given specimen is there ; or while the 



where the jierianth leaves are united, some ap- flowers are well known, the characters of the 



pear to show true gamosepaly, like Cyclea ; but embryo are unknown ; or again, because the only 



more commonly the sepals that are united for a examples are on the contrary fructiferous, 



certain distance, as in Synclhia, TUiacora, &c, 3 " Distrib. in trib. (forte minis artificiales) 



appear merely to stick together by their edges. imprim. ad fruct. v. sem. limitat." (B. H., 



1 Only one cotyledon is well developed in Gen., 30.) 



Rameya and Trlclisia. Mr. Mieks informs 4 " Flores non regulariter trimeri." (Hook. 



me that he has shown this in the latter genus, et Thoms.) 



and that the same thing occurs in Spirosper- 5 The top of the style is however usually di- 



mum. vided into three branches. 



