MOEPHOLOGY AND ANATOMY OF THE NYJlPHJiACE^. 293 



The seedling of Nymplicea zanzibariensis also is at first monostelic, and for about the 

 same distance upwards as in Victoria regia, but the transition from the stele of the 

 epicotyl into that of the internode immediately above takes place in a different manner. 

 The epicotvledonary stele is exactly similar to that of Victoria, and the two bundles 

 w^hich it contains as they pass upwards increase in size and subdivide ; however, the 

 several bundles thus formed do not separate from one another, but remain with their 

 xylems confluent. Hence, although there is for a time a single central cylinder with a 

 common endodermis, the bundles do not become distinct and arranged in a ring around 

 a medullary region as they do in Victoria, Therefore the central cylinder must always 

 be very slender, and any further extension to meet the increase in diameter of the stem 

 has to be provided for by a complete separation of the bundles. This will also account 

 for the absence of a peripheral zone around the central vascular mass in the mature 

 rhizome of this genus. 



The structure of the seedling of Nelumhium speciosum is widely different from those of 

 Victoria and Nymphcca, and is particularly remarkable for the large number of vascular 

 bundles exhibited in the ej^icotyledonary internode. There are some 20-30 of them 

 arranged in concentric circles, much as in the internodes of the mature rhizome, 

 from which the chief differences are, that in the epicotyl the bundles are fewer, and 

 that the two circles of inverse bundles are absent. The second internodal region has a 

 structure almost identical with that of the mature rhizome. 



It is a remarkable fact that in the seedling of ^elmnhium an almost complete absence 

 of primitive features is met with both in the leaves and in the stsm. Moreover, the 

 primary root is completely abortive and never escajies from the seed-coat. The 

 precocity with which this plant assumes its definitive form stands in strong contrast 

 to the gradually i)rogressive stages exhibited by the other members of the order. 



Apical Meristem. 



I have not been able to obtain preparations of the apex of a mature plant oi Victoria 

 regia, but microtome series of that of Nymphaca tuherosa, prepared by Dr. Scott, show 

 clearly that the apical cone is entirely composed of a number of homogeneous meri- 

 stematic cells, no desmogen strands appearing therein until considerably lower down, 

 Avhere they are developed in an entirely indiscriminate manner. Nothing resembling a 

 plerome is at any time to be discovered. I have also found a very similar state of 

 affairs in Cabomba aquatica. The apex of the floating shoot has the frSm of a small 

 rounded cone composed of rather large meristematic cells, among which no sort of 

 differentiation is exhibited, until, at a point lower down, the procambial rudiments of 

 the two vascular strands appear on opposite sides of the stem. 



On the other hand, in the young seedlings of Victoria regia and of Nympha-a zauzi- 

 bariensis, taken while yet barely out of their monostelic stage (PL XXI. tig. Oj, the 

 meristematic tissue of the apical cone itself is continued down into the stem as a central 

 column or cylinder of very similar meristematic cells, in which, later on, desmogen 

 strands make their appearance. In the case of the Nynipluca they are irregularly scattered 

 throughout it ; in Victoria those situated at the periphery seem to be the first developed. 



