380 



Notice of Booh : 



[April 



with our investigation. Let us suppose the plant is dicotyledonous, and 

 that we still follow DeCandoUe as our guide. 



He divides Dicotyledonous or Exogenous plants into four sub- 

 classes, the characters of which are taken from the fructification- These 

 four are respectively named Thalamiiiorcs, Calicijlorce, CorolliJlorcB and 

 MQnoc1ilamyde(s. 



I. " ThalamiJIorcB are distinguished by having several distinct petals 

 and stamens inserted on the Thalamus or receptacle of the flower, and 

 not cohering with the calyx. — (flowers hypogynous). 



" The petals and stamens do, however, occasionally cohere ; as, for ex- 

 ample, in MalvacecB, so as in some degree to resemble those of the third 

 class, CorolUfiorae, from which, however, in all such cases, their numer- 

 ous, nearly distinct, carpels distinguish them. 



II. " Calyciforae have also several petals, either distinct or united 

 to each other, but inserted along with the stamens on the calyx, usually 

 through the medium of a disk lining ils tube or concave portion. — 

 (flowers pyrigynous or epigynous). 



III. " Corolli/lorce have the petals united into one, bearing the 

 stamens, and inserted on the thalamus, or receptacle (hypogynous). 



IV. " MonochlamydecB have either no corolla, or the petals, if present, 

 are united to the calyx, forming together a single, not double, perigoni- 

 um or verticel of flower leaves round the pistil. 



Before going further it may be proper to observe, that this is a mere 

 artificial classification, intended simply to facilitate th^^ arrangement of 

 the orders in a linear series, and to aid in enabling b-'<.to determine to 

 what part of that series we ought to turn to find the order of any plant 

 under investigation. While it pretty well fulfills these objects, it must 

 be acknov»7l edged, that, like all such artificial combinations, it is liable, 

 on the one hand, to produce unnatural associations ; and, on the other, to 

 separate orders most nearly allied; besides presenting examples of mo- 

 nochlamydeous and dichlamydeous genera in the same order. These 

 objections, however, apply to the arrangement, only, of the orders, not 

 to the orders themselves, which may, though individually perfectly na- 

 tural, be most unnaturally placed in relation to each other ; a defect, to 

 which all systems yet proposed are more or less liable, and, we may 

 almost assume, ever will be, when we take into consideration that fa- 

 milies of plants, like provinces of a kingdom, touch each other, not by 

 one point only, but on all sides, and that we might as well expect to 

 arrange the latter in a correct, linear series, as the equally irregularly 

 formed and closely surrounded provinces, that is the natural orders, of 



