MOEPHOLOGT AND ANATOMY OF THE NYMPH^ACE.E. 280 



not unlike that figured by Trecul ; and in tlie absence of any sections of the prickle may 

 have led liim to look upon it as bearing an actual pore. 



Van Tieghem considers the prickles to be homologous with those on tlie surface of the 

 leaves of certain varieties of Ilex Aqiiifolium, or with the tentacles of Drosera. 



The method of growth of the leaves of Victoria regia has been fully described by 

 Planch on in his monograph on this plant, although he gives no figures attached. 



At the base of each mature leaf on its ventral side there lies a membranous scale, 

 curving aAvay from the leaf to Avhich it l)elongs, and sheathing all the younger structures 

 in the bud (PI. XXI. fig. 4). It is formed by tlie fusion of two stipules along their inner 

 margins on the adaxial side of the leaf. The first or tirst two leaves of the seedling 

 have no such axillary scale, but their bases are siiglitly winged by two small lateral ap- 

 pendages, which probably represent the free lateral stipules found in certain Nympha^as, 

 N. zanzibariensis &c. (fig. 5). The intrafoliar scale of Nclumhium is probably derived in 

 the same way as that of Victoria, 



The embryonic leaves of the seedling are quite different in appearance from the adult 

 leaves of the mature plant, and they show, by an interesting series of gradations, a 

 progressive change from an acicular primordial leaf (or petiole only, according to 

 Trecul) to the peltate form of the mature plant. The exact form of corresponding 

 leaves in different seedlings varies considerably. However, the first leaf is always found 

 to be acicular. The second possesses a lamina, usually elongate-lanceolate, sometimes 

 wdth two small hastate lobes at the base. The third (fig. 6) varies from elongate-hastate 

 to deltoid-hastate, the auricles are widely divergent, and at the base of the lamina, just 

 above the insertion of the petiole, there is invariably a little pocket or pouch on its 

 adaxial side, which appears to have been formed by the fusion of the auricles at their 

 bases across the adaxial surface of the leaf. The fourth (figs. 7 & 8) is the first that 

 bears spines, and that shows itself to be distinctly peltate. It is the first swimming leaf, 

 and has a lamina oval in outline with a subacute apex ; there are two auricles at the 

 base which do not diverge, but lie more or less close together, thus making the leaf 

 sagittate. The fusion between them at their bases is here carried much further than in 

 the previous leaf ; in fact, they are fused together for about a third of their whole length 

 from the point of insertion of the petiole, the line of fusion being followed by a small 

 vein. So the final form attained by this leaf closely resembles that of the adult leaves 

 of many NympliBeas, N. Lotus, gracilis, delicatissima, &c., in which the fusion of the 

 auricles is never carried any further. In Victoria regia, on the other hand, the 

 succeeding leaves become more and more orbicular in outline, and the auricles become 

 fused along a successively greater part of their length until the final form of the adult 

 leaf is attained. 



As previously stated, the leaf of the mature plant owes its peltation to the formation 

 of a zone of growth across the summit of the petiole on the adaxial surface of the leaf, 

 which by its activity forms a mass of tissue joining together two parts of the lamina 

 previously separate and distinct. Hence it would appear that the leaf of the mature 

 plant passes in its youth through stages which are quite parallel to those permanently 

 retained by the embryonic leaves. Thus, the tliird leaf of the embryo may be regarded 



2t2 



