﻿454: ON THE GENUS CROOMIA. 



nous, or dicotyledonous and exogenous. As the structure of the embryo could not be 

 made out, and the germination was unknown, we were obliged to depend mainly upon 

 the structure of the stem. A portion of the creeping rhizoma, attached to the base of 

 some of the original specimens, exhibited a closed zone of wood, surrounding a small 

 cellular pith, and itself surrounded by a thick and purely parenchymatous cortical 

 stratum. This we naturally took for an exogenous structure ; — a view we were the 

 more inclined to adopt, from knowing at the time no Smilaceous, Dioscoreaceous, or 

 other allied Monocotyledonous j)lant having a simple ovary and an arillus like that 

 which is so remarkable in Croomia. Among Dicotyledons the nearest approach to 

 our plant in structure and in habit was found in some of the herbaceous Berberidacece ; 

 and we accordingly referred it to that order, notwithstanding the simple and persistent 

 perigonium. 



No additional materials were at hand when the first volume of the Genera Ulustrata 

 was prepared. So the genus was left appended to the Berberidacece, but mentioned as 

 of wholly doubtful affinity. 



Within the last two years I have been supplied, by Dr. Cabell and the Eev. Mr. 

 Nevius of Alabama, with excellent dried specimens, including mature fruit and seeds. 

 Mr. Nevius has also communicated living rootstocks, from which I have now the plant 

 in blossom in the conservatory of the Cambridge Botanic Garden. I have therefore 

 had the opportunity of a complete examination of Croomia, except as to the genni- 

 nation, in respect to which our attempts have thus far been unsuccessful. 



From these observations an amended generic character is given above. One correc- 

 tion of considerable consequence is, that the stamens are not really hypogynous, but 

 are inserted upon the base of the calyx, which, moreover, shows some tendency to be 

 connate with the base of the ovary. In the nearly full-grown flower-bud, the cavity of 

 the ovary at its broad base sinks slightly below the line of junction with the perianth 

 (as in Fig. 3) ; in anthesis it deepens somewhat, so as to give a slight appearance of 

 adnation, as is represented in Fig. 4. As the fruit forms, however, this incipient 

 adnation disappears. The mature fruit is usually somewhat contracted or tapering at 

 the base, and above dilated where the large seeds are contained ; but the ceU extends 

 down to the very insertion of the perianth and stamens, so that there is no trace of 

 any intervening hj-pogjTious receptacle, as in Thalamiflora;. 



The singular arillus, if it may be so denominated, belongs Avholly to the funiculus, 

 in the first instance. In the young flower-bud it consists of short-cylindrical simple 

 cells, like forming root-hairs, growing from its whole surface. As these lengthen, 

 they are converted, by the formation of transverse partitions, into a simple series of 



