i66 



BOTANICAL GAZETTE 



[SEPTEMBER 



Fig. 3 



further fused so as to form three more or less clearly separated areas, 

 as IS shown in the photograph of P. crispus (fig. 28) and in fig. 3, 

 ^ij ^2J ^2? h of the preceding diagrams have united to form the central 

 area, while the remaining bundles form the lateral regions. T^ may 



be distinguished from T^ by the double 

 lacuna of the central area. As usual, three 

 (or sometimes five) leaf trace bundles enter 

 the central cylinder, the large median one 

 joining the central area, and the laterals, 

 each composed as usual of bundles from the 

 petiole and one from the stipule of that side, 

 join the lateral areas of the central cylinder, DeBary (7, p. 274) 

 gives a singularly misleading description and inaccurate diagrams 

 of this species, owing probably to the crudity of his methods of 

 investigation. Schenck (28, p. 41) has given a correct interpretation 

 of the appearances presented by a cross-section, and I have con- 

 firmed his statements by series of sections through the nodes. P- 

 crispus is no more an ''anomalous monocotvledon" (DeBary) than 

 are the other species of Potamogeton, since, as shown above, its vas- 

 cular system can be easily derived from that of P. pulcher. 



A further reduction or fusion of the vascular elements is shown in 

 the species of which P. pectinatus {fig. 18) may be taken as the type. 

 In the internodes the xylem is represented only by a central cavity 

 surrounded by parenchyma and phloem, but at the nodes a division 

 of the stele into three areas is more or less plain. The last fact sug- 

 gests that the condition found in this species has been derived from 

 that seen in P. crispus, since ancestral characters are apt to persist at 



the nodes. 



from 



stipules show^s great variation in the different species, (i) They 

 may descend into the cortex of the succeeding internode, constituting 

 the usually numerous cortical bundles of a number of species, e.'g., P. 

 natans, eventually joining others of the same kind, or finding their 

 way into the central cylinder at a lower node. (2) They may descend 

 a short distance into the cortex, then disappear, while a short distance 

 below the node other strands make their appearance in the cortex^ 

 to disappear before the next node is reached, e. g., P. RohhinsiL 





! 



