466 R. KIDSTON AND D. T. GWYNNE-VAUGHAN ON 



General Remarks on the Ancestry of the Osmund^ce^. 



From the summaries appended to the several preceding parts of this memoir it is 

 sufficiently clear that we regard the Osmundacese, as a whole, as an ascending series of 

 forms whose vascular system is to be derived from a primitive protostele with a solid 

 homogeneous xylem. At the same time it is clearly recognised that the existing 

 Osmundacese do not represent the highest vascular complexity that has ever been 

 attained in the order. They were already far surpassed by the Cretaceous Osmundites 

 Skidegatensis, and it is quite possible that other fossil Osmundacese may yet be 

 discovered that have advanced still further. There is, indeed, no reason why some 

 forms should not have developed a vascular system which might be regarded as a 

 dictyostele. Such an Osmundaceous dictyostele, if it ever existed, would, of course, 

 have been arrived at by a series of steps entirely different from those traversed by the 

 Polypodiacew, etc. The fact that the complex Osmundites Skidegatensis already existed 

 in Cretaceous times forbids the summary rejection of the possibility that the more recent 

 Osmundacese are reduced from some such type. It by no means follows, however, that 

 such a reduction has actually taken place, and we prefer to regard both Osmunda 

 cinnamomea and Osmundites Skidegatensis as indicating a comparatively recent advance 

 towards amphiphloic structure along an original line of their own. 



In whatever manner, however, this question may be decided, it does not in any way 

 affect the stability of our main proposition : that the typical Osmundaceous stele has 

 been derived by the medulla tion of a protostele with a primitively solid xylem, and the 

 subsequent breaking up of the peripheral xylem ring thus formed into separate strands. 

 The type of stele that finally results might fittingly be termed a dictyoxylic monostele. 

 It follows, of course, that the first appearance of internal phloem in the more complex 

 forms must have taken place in a stele that already possessed a pith. 



In taking this view of the matter, we find ourselves in complete accordance with Dr 

 Paul Zenetti, who came to much the same opinion fourteen years ago.* Dr Zenetti 

 arrived at his conclusions from a careful and exact study of the anatomy of Osmunda 

 regalis alone, altogether without the support of the evidence that we have gained from 

 the fossil representatives of the order — evidence which we believe has raised this view 

 from an astonishingly acute speculation to the status of a sound working hypothesis. 



Since this theory is chiefly based upon the interpretation of the structure exhibited 

 by the steles of Osmundites Dimlopi,Zalesskya gracilis, Z. diploxyJon, and Thamnopteris 

 Schlechtendali, it might perhaps be objected that it is hazardous to refer the two latter 

 genera, at any rate to the Osmundacese at all. We only do so with a certain reservation ; 

 all that is known of the plants in question is the structure of the stern and of the lower 

 part of the leaf-trace, and it is, of course, impossible to say what might be done with 

 them were we acquainted with the whole plant, including the spore-bearing organs. 

 They might, indeed, possess unsuspected characters of such a nature as to necessitate 



* lias Leihmgsystem im Stamen von Osmunda regalis und dessen Ubergang in den Blattstiel, pp. 25-2C, 1895. 



