456 MR. W. C. WORSDELL ON THE COMPARATIVE 



in the stem ; the primary tap-roots of Stangeria and Bowenia, two genera with but a 

 single vascular ring, do not exhiljit these outer concentric strands, as would, in all 

 likelihood, also be the case with Ceratozamia, Zamia, and Dioon, the remaining three 

 o-enera possessing a single vascular ring in the stem. This is a fact of importance, for it 

 points, with a fair amount of certainty, to the conclusion that the concentric or partially 

 concentric strands, or the fragmentary portions of such, situated at various radial 

 distances outside the central root-stele, are the homologues of the second, third, or fourth 

 vascular icings in the stem. And it is probable that, at a later stage of growth, the place 

 of these strands would be occupied by a vascular ring of collateral structure, when the 

 inner portion of each strand would become displaced and isolated as small bundles Avith 

 inverted orieatation of their parts, such as were shown by me to exist in the lower 

 portion of the stem of 3facro::amia 'Fraseri, Miq. 



PL XLIII., fig. 9, 1 regard as of extreme importance in helping materially to substantiate 

 the views I have put forward as to the phylogenetic origin of the successive vascular 

 rings of Cycads, for it presents within small compass that which, if rightly interpreted, 

 is, to my own mind, at least, a key and clue to the whole question. I hold, then, the 

 significance of this structure to be as follows : — The strand cs^ represents and is homologous 

 with the second vascular ring of the higher regions of the axis, but it retains (and this 

 is the important point) almost the entire contour of the primitive concentric constituent 

 of the ring ; the two bundles cs- represent two concentric strands of the third vascular 

 ring (of which, in each case, all but one of the outer segments, exhibiting, of course, the 

 orientation of the first ring or central cylinder, have become obsolete) ; in the same way 

 the fourth vascular ring has for one of its representatives at this level of the axis the 

 bundle cs^, the sole remnant of a concentric strand whose innermost segment has alone 

 survived. It is highly illustrative of what I have been endeavouring to point out with 

 regard to the meaning and origin of these cortical strands, that cs^, at the same distance 

 in the cortex from the central cylinder as cs^, affords what I regard as an example of 

 one of these concentric strands which has, happily, retained in .more perfect form than 

 the rest its primitive structure. Its parts are, it is true, rather loosely connected together 

 and isolated, thus giving this strand the appearance of a group of bundles rather 

 than of a single one ; but it should be remembered that this loose aggregation of the 

 segments is one of the chief and peculiar characters of Cycadean vascular strands, 

 rig. cs^ also probably represents, along with cs"^, one of the primitive constituents of 

 what, in the ancestors of the plant, would have constituted a ring of small concentric 

 strands. 



The fact that, as in the case of Encephalartos horridus, Lehm., and Cycas revoluta, 

 Thunb., the concentric strands, or parts of such, in the root are local in their occurrence 

 —appearing sometimes, as in the first-named plant, at one point only of the periphery of 

 the central stele — the greater portion of the subsequent vascular ring being constructed 

 according to the ordinary collateral plan, may be explained by the fact that these ancient 

 types of structure are gradually becoming obsolete and extinct, that they represent, in 

 truth, the last sporadic rudiments of a once dominant concentric type of stinicture, which 



