256 NATURAL niSTORY OF PLANTS. 



From this group we shall first, simply from practical considerations, 

 and not losing sight of the artificial character of the proceeding,' take 

 those Anouacccp which appear to possess two calyces and one corolla. 

 The other series, on the contrary, contains those with two corollas 

 and one calyx.' 



Series of the M1LIUSE.E. — G3ma3ceum polycarpous. Perianth triple 

 hypogynous. Middle perianth more like the outer than the inner. 



Serien of (he Anonk.e. — Uyna^ceum and receptacle of the precediug ; 

 middle perianth (outer corolla) more like the inner than the outer 

 (calyx). We have seen that this series is then subdivided according 

 to the conformation of the corolla into five sub-series, as follows : — 

 1. UvariecB ; 2. Unoneee ; 2t. Xijlopiea ; ^. Rolliniea ; 5. OripultrcfP.^ 



The AnonacecB present great uniformity in the general characters 

 of their vegetative organs. We always find non-herbaceous plants 

 with alternate exstipulate leaves, varying greatly, it is true, in 

 the size, consistency, and duration of the parts. The stem is 

 almost always aerial ; in Eupomatia Bennettii alone there is a rhizome 

 creeping nearly horizontally below ground, and bearing the aerial 

 branches. The arborescent Anonacea are nowhere stated to attain to 

 any great size. The largest trunks in any collection are al)out as thick 

 as a man's thigh. There are in this order a large number of shrubs 

 or little bushes which divide into fascicled branches from the level 

 of the ground ; and in very many exotic species we find this arrange- 

 ment so well shown in our cultivated A-simina triloba. Often, aijain, 

 the stems or slender branches of the Anonacea twine round neijrh- 

 bouring objects, and many species are described as creepers or 

 climbers. Has this peculiarity any influence on the anatomical 

 structure of the stem? We must answer in the negative, but only 

 as regards those species Ave have had the opportunity of studying. 

 Except a rarefication of the cortical parenchyma, to form spaces in 

 the centre of the masses of cellular tissue between the rows of liber 

 bundles, we have found the same structure in the branches of the 

 climbing and in the nun-sarmentose species oi A/ioiki, U/tono, and 



' Sec Adavnonin, viii. .3(1'.). botwd nro, with otio cxrpptioii, those uwhI by 



* Kxci-pl wlicii Olio corolla tliwippcarH. It w Bkntiiam A IIoukkk, the only ililVorfiico Wm^ 

 nnually the inner one; in Enuntia jilone ih the the rehitivo iinportHnee iissi^jne*! to these jfrouj)* ; 

 outer corolhi iihtienl. Unl in nil tliCHu pluntB there for we term tubtei-iei what they have culled 

 are only three nepiiloid leavcB. tribtt. 



* The ihaructent on which theso divihionK are 



