08 MENISPERMACE.E. Diclinous Exogens. 



more or less houiotropal or campylotropal by the greater increment of the ovary 

 on the dorsal face. Drupes usually fleshy, containing a single nut obsoletely 2-valved, 

 greatly various in form and development, always 1 -celled, with the cell more or less 

 curved about a central process (condyle), which is a peculiar development of the 

 placenta. Seed single, partaking of the form of the cell, enveloped in a membranous 

 integument, attached by its ventral face to the condyle, albuminous or exalbuminous ; 

 albumen, when present,, abundant or small in quantity, either homogeneous in tex- 

 ture or partially divided into lamellar plates or convolutions, in which case the 

 integumental covering partly enters into its numerous interstitial spaces. Embryo 

 homotropous or campylotropous ; cotyledons either flat and foliaceous, and either 

 incumbently parallel or laterally widely divaricated, and placed in distinct cells of 

 the albumen, or with cotyledons narrow, flattened, accumbent, and coiled in a peri- 

 spherical form, or very long, slender, terete, accumbent or incumbent, and coiled in 

 a hippocrepical or somewhat annular form ; in the exalbuminous tribe the cotyledons 

 are large, thick, fleshy, and hippocrepically or reniformly bent, and incumbent ; the 

 radicle in all cases is superior, short, terete, curved more or less, and pointed to the 

 style or original apex of the fruit, which, by its inflexion, is often curved downwards 

 to near its base. — Miers. 



Our knowledge of this curious family has hitherto been extremely imperfect, and 

 I am indebted to Mr. Miers for the above outline and for the facts upon which the 

 following observations are founded ; he has prepared an extensive monograph of the 

 whole order, with numerous drawings and analyses, which are not yet published, but 

 he has given a slight outline of these results (Ann. Nat. Hist., 2nd. Ser. vii. 33). In 

 the " Introduction to Botany," this family was placed in the class Imperfectfe, on 

 account of its undeveloped flowers and curved embryo; but in my subsequent 

 arrangement in the former edition of this work (p. 307), it was arranged among 

 diclinous exogens, in a distinct alliance, under views in which Mr. Miers does not 

 concur. He thinks that Mouimiacea?, Atherospermacepe and Myristicacere form a good 

 alliance in the same position, but that Menispermacese, Schizandracese and Lardiza- 

 balacese constitute another valid group which should occupy a different place in the 

 system ; the former being monochlamydeous and essentially unisexual, might retain 

 its place among diclinous exogens, but the latter are truly dichlamydeous, for although 

 the petals are generally reduced to the size of scales, they are ever regular in their form 

 and number, and with rare exceptions, constantly present ; the diclinous character of 

 their flowers is due only to abortion, for they are sometimes hermaphrodite or poly- 

 gamous, and in the <J flower rudiments of the ovaria are commonly seen, and in the 

 ? the sterile stamens are of frequent occurrence. Mr. Miers, therefore, argues that 

 according to the rule laid down in this work (p. 240), such flowers should not be held 

 to be truly diclinous ; and he would say that on account of their many seried floral 

 envelopes and numerous 1-ovulate carpels, this alliance should find its place in the 

 system between Ranales and Berberales. I have already pointed out the affinity of the 

 Kadsurads to the Ma<moliads (p. 41S). The Menispermads at the same time approach 

 the Anonads (especially the tribes Bocageas and Xylopiese) in their bisexual flowers, 

 the frequently valvate aestivation of their flowers, their numerous unilocular carpels 

 with ovules attached to the ventral sutui'e, and in their seeds often with albumen 

 divided into lamellar plates. This is nearly the position long ago assigned to the family 

 by botanists ; De Candolle, however, suggested a resemblance to the Sterculiads. on 

 account of their mouadelphous stamens and peltate leaves, but they differ in all other 

 most essential respects. Accordiug to St. Hilaire they are related to Spurgeworts, 

 because Phyllanthus sometimes has its anthers born on a monadelphous column, as 

 in Cissampelos, a rare occurrence in both families, and there is little else to support so 

 distant a conclusion. Mallowworts have also been suggested, with as little foundation. 



Mr. Miers remarks, that there is probably no family so completely hetero- 

 morphous as the Menispermads, or that presents such extreme and aberrant features, 

 at variance with its normal structure ; these extremes are found in the habit of the 

 plants, in the texture and form of the leaves, in the various modes of inflorescence, 

 in the number, arrangement, and manner of aestivation of the floral envelopes, in the 

 form and position of the stamens, as well as in the structure of the anthers and their 

 mode of dehiscence, in the presence or absence of a distinct gynophore, in the 

 variable character of the style and stigma, in the extent of development of the ovules, 

 in the form of the nut, in the seed, sometimes exalbuminous, at others with albumen 

 highly developed, which is often fleshy and homogeneous, copious or "sparse "in 

 quantity, and in other cases singularly constructed of lamellar plates: and finally in 

 the variation of the form and development of the embryo, whose cotyledons are 

 sometimes large, fleshy, and adpressed, or they are slender and terete, or long and 

 ribbon-formed, foliaceous, thin in texture, divaricate, and placed in separate cells in 



