TRANSACTIONS OP SECTION K. 783 



was probably represented in the earlier terms of this series, and that the upright 

 position of the more advanced Cyatheacese was a secondary condition. 



Thus, by the help of the two synthetic genera above named, it becomes 

 possible to trace the origin of two great sequences of living ferns along lines more 

 or less parallel but distinct : the one leading to a multiplicity of ' mixed ' types 

 represented by the Pterideae, the other to some of the most prominent ' gradate ' 

 types of the Cyatheaceous affinity. And it appears that these may both have 

 arisen from types of the ' Simplices ' having affinities with the Schizseaceae and 

 Gleicheniacea?. 



3. On the Fossil genus Tempskya. 

 By R. Kidston, LL.D., F.R.S., and D. T. Gwynne-Vaughan, M.A. 



From the examination of some specially well-preserved Russian petrifactions 

 belonging to this genus it may be safely stated that these fossils consist of aggre- 

 gates of branching fern stems imbedded in a compact mass of their own inter- 

 woven adventitious roots. 



The ' false stems ' thus produced no doubt grew erect as columnar or conical 

 structures, attaining a height of nine feet or more. The individual stems are 

 only about the thickness of a slate pencil, and they possess a solenostelic vascular 

 system. Strange to say, although growing erect, they exhibit a well-marked 

 dorsiventral symmetry, with the leaves all on one side of the stem and adventi- 

 tious roots on the other. 



The existence of a fern posse3sing this peculiar habit of growth opens the way 

 to the suggestion that the strong, erect, arboreal stems of the modern tree-ferns 

 may not be a primitive feature of the order. These plants may have been 

 derived from forms with a Tempskya-like habit, in which the original axis has 

 so greatly surpassed its branches in size and importance that the latter are now 

 almost entirely suppressed. As a result we have now a single main axis strong 

 enough to grow erect without the additional support of a root coating. 



4. Further Observations on the Fossil Flower. 

 By Miss M. C. Stopes, D.Sc, Ph.D. 



The ovary and perianth of a petrified flower (Cretorarium japonicum, Stopes 

 and Fujii) from the Cretaceous rocks of Japan were recently discovered by 

 Professor Fujii and the author. The specimens were petrified in concretionary 

 nodules, and show the cell structure, but did not contain ovules. 



Another specimen has now been discovered, in which the ovary wall is much 

 better preserved, showing an outer fleshy layer as well as the sclerosed layer, and 

 also containing several ovules. 



5. The Morphology of the Ovule of Gnetum africanum. 

 By Mrs. M. G. Thoday. 



In the young ovule the three coverings arise together at the base of the 

 nucellus, and the greater portion of the nucellus stands free in the centre of 

 the ovule. During later development the region between the two inner coverings 

 becomes more and more stretched, so that the innermost covering arises near the 

 apex of the nucellus, and the free portion of the nucellus in the mature seed is 

 very .small in proportion to the part developed by intercalary growth. 



The outer covering resembles the bracts in structure, and is here regarded as 

 an extra covering, while the two inner are called integuments, and are compared 

 with the two integuments in Welwitschia. The middle covering or outer integu- 

 ment has a well-developed vascular supply and is traversed by numerous strands 

 of fibres reduced higher up to four or five. At the tip the outer integument 

 firmly clasps the micropylar tube, and a transverse section at this level shows the 

 four or five strands forming a star of strongly lignified tissue. 



The inner integument is formed of thin walled tissue, and in the young ovule 

 strands of meristem run up into it for some distance above the base; when the 



3 K 2 



