368 ARTHUR J. MASLEN [may 



It should be pointed out, however, in this connection, that Dr. 

 Scott, in his recent paper on " The anatomical characters presented 

 by the peduncle of Cycadaceae," already quoted, has shown that in the 

 peduncle of Stangeria paradoxa, T. Moore — which shows the mesarch 

 structure and consequent development of both centripetal and 

 centrifugal primary xylem — -the small cauline cortical bundles which 

 also occur in the peduncle sometimes exhibit a typically concentric 

 structure. These concentric bundles are of considerable interest, not 

 only as the only recorded cycadean bundles with a primarily concentric 

 structure, but also as probably indicating a recurrence to a condition 

 which appears to be a relic of an ancestral fern-character. To 

 quote Dr. Scott : " The evidence for the origin of cycads from ferns 

 is now overwhelming, Professor Ikeno's discovery of the multiciliate 

 spermatozoids of Cycas completing the proof in the most striking- 

 manner." 1 



Still more recently Mr. Worsdell has described " The vascular 

 structure of the sporophylls of the Cycadaceae," 2 and has shown that, 

 exactly as in the case of the peduncle, the sporophylls borne on it 

 retain certain primitive characters which are not found in the foliage 

 leaf. He shows that this primitive character is exhibited here also by 

 the development of concentric bundles which occur in both fertile and 

 barren sporophylls in several genera. " These concentric bundles are 

 absent from the foliage leaves, the structure, number, and orientation 

 of whose bundles are extremely regular, constant and well defined, 

 whereas in the sporophylls the reverse is the case, a fact which 

 probably points to their possessing a more primitive structure, viz. one 

 not so perfectly adapted and stereotyped to subserve a special 

 physiological function, as is the case with the foliage leaves." 3 



On the outer surface of the cortex in both petiole and stem of 

 Lyginodcndron are seated some curious emergences, the structure of 

 which suggests a gland of some kind. It was the similarity of these 

 structures on stem and leaf-stalk which first led Williamson to surmise 

 that the petioles (Rachiopteris asperci) were borne on Lyginodcndron stems. 

 Similar appendages are found on some recent tree-ferns, more especially 

 on Alsopliila australis. 



That the fossils previously described by Williamson as Kaloxylon 

 Hookcri are the adventitious roots of Lyginodcndron appears to have 

 been proved almost simultaneously and independently by Williamson 

 and Scott and by the late Mr. Hick. 4 In the memoir by the two first 

 authors the evidence on which this conclusion is based, and the some- 

 what peculiar structure of these appendages is fully described. That 

 they are really roots is proved not only by their primary structure, but 



1 Loc. cit. p. 409, and Annals of Botany, June 1897. 



- Annals of Botany, June 1898. s Loc. cit. p. 239. 



4 "On Kaloxylon Hookcri, Will., and Lyginodendron Oldhamium, Will.," Mem. and 

 1'roc. Manchester Lit. and Phil. Soc. (1895), vol. ix. p. 109-11 0. 



