AKATOMY OF CEETAIN SPECIES OF ENCEPHALAETOS. 45g 



formed vascular elements arose by tangential division of the large, rounded or ani,ndar, 

 cortical or pericyclic cells, resulting in the large, isodianietric, irrcgularly-sliaped 

 tracheides of precisely similar shape to those cells. After one or more of such elements 

 have been cut off, the parenchyma-cells begin to divide radially as well as tangentially, 

 in this way forming smaller and smaller elements with each centrifugal division, until at 

 length the majority of the tracheides are of the same size and shape as those of the 

 central cylinder or stele. The large first-formed tracheides very soon become displaced 

 owing to the pressure of surrounding tissues. This fact, together with that of the great 

 diffei-ence in size between these and the later-formed tracheides, renders the determination 

 of their mode of origin, at an older stage of the same vascular tissues, utterly obscure, 

 Tlie similar reticulate tracheides in the cortical concentric strands, more distant from the 

 central cylinder in the stem of Cycas, in all probability arise in precisely tlie same wa\ , 

 but, in that case, I did not succeed in definitely determining their mode of origin. 



5. The cambium of several of the vascular zones is, judging from the appearance 

 presented by its cells, simultaneously active. 



6. All the secondary tracheides of the leaf -trace bundles possess dense spiral or, more 

 probably, scalariform thickenings, thus differing somewhat from the leaf-traces of 

 Macrozamia Fraaeri, Miq., w^here only a certain number of the secondary tracheides are 

 spirally thickened. 



7. A medullary system of bundles occurs in all four species, similar in every respect to 

 that of Macrozamia Fraseri, Miq. This system is intimately connected with a similarly- 

 anastomosino- network of mucilage-canals which is continuous with that of the cortex. 

 The medullary vascular system is, however, quite independent of the leaf- trace system of 

 the cortex. The bundles are primary in origin. 



Although the presence of a medullary vascular system would appear, so far as one can 

 judo-e from the examination of but four species, to be universal in the genus Encei^halartos, 

 the same cannot be said of Macrozamia, for no trace of such a system could be detected 

 in the stems of either M. Denisomi, F. Muell., or M. spiralis, Miq., although it is 

 possible that the latter plant was at rather too young a stage of growth to be able to 

 exliibit the structure. 



8. When the plant, growing in its native habitat, has reached a certain age, the 

 primary tap-root dies aw^ay as the dry season approaches. The mucilage-canal system 

 then probably acts as a water-reservoir. Next season adventitious roots arise. These 

 exliibit in their upper and older portion a structure similar in many respects to that of 

 the primary root of F. horridus, Lehm., viz., either one or two vascular rings and one or 

 more semi-concentric cortical strands. 



Finally, after the above observations on the extremely interesting structures m the 

 vegetative organs of Fncephalartos, I should like, before concluding the paper, lo bring 

 forward some considerations with regard to the phylogenetic relationships of the 

 Cycadacese, suggested by the contemplation of the various structures in this genus. 



In the first place, I wish to restate the generalization made or inferred in my account 

 of the anatomy of Macrozamia Fraseri, Miq., viz., that the stem-structure (as also, of 

 course, that of the upper portion of the root) is in aU probability derived from the 



SECOND SERIES. — BOTANY, VOL. V. *^ ^ 



