290 ME. D. T. GWYNNE-VATJGHAN ON THE 



as the first in which the transverse zone appears at all; its activity being as yet but 

 little marked, the structure arising from it is correspondingly minute. Whereas in the 

 fourth and succeeding leaves the duration of its activity and the results obtained thereby 

 become more and more considerable. So, in this case at least, it may be said that the 

 ontogeny, if it may be so termed, of a single leaf of the mature plant repeats, in some 

 degree at any rate, the true ontogeny of the leaves of the plant as a whole — that is to 

 say, it repeats the developmental stages of the embryo. 



The embryonic leaves of NymplKBa and Nuphar pass through a series of stages some- 

 what similar to those of Victoria until a more or less sagittate form, is reached. In many 

 species no further advance is made, the auricles remaining practically free throughout 

 (Nymphcea pygmcea, N. stellata). In others (those mentioned above) they fuse together 

 to about the same extent as they do in the fourth leaf of Victoria. 



The adult leaves of Barclaya show a still simpler form, being elongate, strap-shaped, 

 and slightly auricled. After consideration of the above gradations they might perhaps 

 be held as approximating to the ancestral leaves of the NymphcBocecc in their form. 



With regard to its embryonic leaves, Kelumbiiim signalizes its distinction from the 

 other NymphceacecB by the fact that the first leaf is not acicular, but it and the succeeding 

 leaves closely resemble in miniature those of the mature plant. 



Anato7ny of the Ilature Uliizome. 



The chief feature in the structure of the rhizomes of the NympliccacecB is the astelic 

 arrangement of their vascular bundles, and in most cases a second very prominent 

 characteristic is occasioned by the almost indescribable confusion and complexity 

 presented by these ])undles in consequence of the exceedingly irregular courses they 

 pursue and the intricate anastomoses they undergo. This complexity is found at its 

 maximum in the rhizomes of Victoria and Nym'phcea. Tlie vascular bundles are for the 

 most part massed together in the central region of the rliizome, where they run indis- 

 criminately in all directions and without any attempt at order, the utmost diversity also 

 being exhibited in their orientation. On the outside of this central region there lies a 

 belt of continuous parenchyma comparatively free from vascular bundles, traversed only 

 by those passing outwards from the central mass towards the exterior. Another narrow 

 zone of tissue of exactly the same nature lies immediately below the surface of the 

 rhizome. Finally, between these two there is a broad belt of "cavernous tissue" 

 [cf. Henfrey, Phil. Trans., 1852, pp. 289-294). The latter is formed by a number of 

 large lacunae filled up with spongy masses of very loosely packed cells, and separated 

 from each other by supporting plates or laminae of continuous tissue connected on 

 either side with the previously mentioned zones of firm parenchyma. 



Those species in which the rhizomes remain throughout comparatively small, snch as 

 ^ymplidea Jlava, show, in accordance with the diminution in the number of bundles 

 present, a corresponding decrease in the complexity of their arrangement. Further, in 

 this species and in Is. tuherosa it is noticeable that almost all the inner bundles are 

 oriented in the same direction, i. e. inversely; many of them having their xylcms 

 confluent with those of the outer normally oriented bundles. Differences are in 



