298 . MK. D. T. GWTNNE-YAUGHAN ON THE 



Summary. 



We see, therefore, that the plants in this Order, although generally astelic, sometimes 

 aggregate all the vascular bundles present in certain of their members into so many 

 steles, or, it may be, that only those bundles wliieh are set apart for a certain function are 

 thus dealt with, although they may be lying in a region of the plant otherwise astelic. 



These steles vary greatly in compactness, size, and particularly in the number of 

 bundles which assist in their formation. Prom the large many-bundled root-bearing 

 steles of Victoria regia, through the smaller ones of about a dozen bundles of NijmphcBa 

 alba and iV". tuberosa, we pass to the steles containing four to three only in the primary 

 stolon of N.flava, and finally to those containing two bundles in the secondary stolons 

 of the same plant and throughout the whole structure of Cabomha and Brasenia — if, 

 indeed, we are entitled to call the latter steles at all. In relation to this point the close 

 similarity that these paired bundles show in all points of structure to the pair of bundles 

 found in the cylinder of the epicotyl of Victo^Ha and Wymphcea should be borne in mind, 

 and the latter, in virtue of its position, must be acknowledged as a diarcli stele. 

 Moreover, there are many other admitted steles which consist of two bundles only in the 

 stems and petioles of many Perns and SelagincUas. If this question be answered in the 

 afiirmative, Cabomba and Brasenia must no longer be described as astelic, but as 

 essentially polystelic. However, the exact manner in which the two pairs arise from 

 the central cylinder of the epicotyl should first be determined before any final decision 

 is arrived at. 



Finally it may be remarked that the simplicity of the anatomical structure in 

 Cahornha and Brasenia is completely in accordance with the structure of their flowers, 

 the want of complexity in which also distinguishes these plants from the rest of the 

 Order — the parts of the flower being arranged in whorls of three only, and the 

 gynsecium being superior and apocarpous. 



EXPLANATION OP THE PLATES. 



Plate XXL 



Figs. 1-4. Young stages of mature leaves of Victoria regia. 1. Very young, front view (xlO). 



2. Older, back view (x8). 3. Still older, side view ( x8). 4. Young leaf, in situ, seen from 



behind, and showing axillary scale, which curves away from the leaf (X about 2). 

 Fig. 5. Young stage of mature leaf of Njimphcea sanzibariensis, showing the free lateral stipules (x4). 

 Figs. 6-8. Embryonic leaves of Victoria regia seedling (nat. size). 6. 3rd leaf showing the " pocket'' 



of the lamina at the insertion of the petiole. 7. 4th leaf, front view. 8. 4th leaf, back view. 

 Fig. 9. Median longitudinal section of apical region of young seedling of Nymphaa zanzibaritnsis : 



It., leaF-trace j rt., root-beariug stele; vv., desmogen strands in central meristematic cylinder 



(X240). 



