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



difficulty of thinking of them as growing in any other way than rigidly erect has to be 

 faced by every theory that is brought forward to account for them. 



It is not easy to think of erect leaves of this nature as growing from an upright 

 stock, but such fronds are associated without difficulty with creeping or prostrate rhizomes. 

 There is very little doubt that one Zygopterid stem, at any rate, was a rhizome. We refer 

 to Zygopteris Romeri, where the small size of the stem relative to the leaves, the apparent 

 absence of supporting sclerenchyma, and the presence of numerous irregularly distributed 

 adventitious roots, all point to a subterranean habit. At the same time, other Zygopterid 

 stems seem to have grown erect, but even in this case the leaves may have been approxi- 

 mately erect. However this may be, our suggestion is that a leaf of the ordinary 

 fern type took to a more or less rigidly erect habit, and its whole structure accommodated 

 itself to the change. By what morphological steps this change took place is another 

 question which it is not proposed to discuss here, because we have as yet no tangible 

 evidence to bring forward in favour of any of the conjectures that might be made. 



This theory of a primitively two-rowed Zygopterid leaf is in opposition to that held 

 by Tansley and Bertrand. These authors maintain that the Zygopterid leaf was 

 primitively radially symmetric, with appendages all around — a theory which is bound 

 up with the supposed origination of the leaf as a subsidiary and outgrown branch-system 

 of the main axis or stem. As these displaced branch-systems began definitely to take 

 up the function of leaves, they gradually lost their radial symmetry by reducing their 

 lateral appendages — first of all, to two rows of opposite pairs, and finally to a single row of 

 appendages on each side. Tansley regards Stauropteris as the most primitive Zygo- 

 pterid petiole, because it appears to retain so much of the supposed original radial 

 symmetry. On the other hand, Bertrand shows that the radial symmetry is more 

 apparent than real, and regards Stauropteris as one of the most specialised types in the 

 order — a point upon which we are heartily in agreement. 



In conclusion, it would appear that, although the stems of the Zygopteridese, possess 

 a comparatively primitive type of stele, their leaves have advanced very far indeed along 

 special lines of their own — lines quite peculiar to the order, which have culminated in 

 the production of two rows of appendages on each side of the rachis. In Anachoropteris 

 they may even have returned again to a single row by the reduction of the adaxial 

 appendages — a suggestion first put forward by Tansley, and now supported by Bertrand. 

 In any case, the order as a whole is so far advanced beyond anything really primitive 

 in the Filicales that they certainly do not deserve to rank among the Primofilices. In 

 fact, no group of plants has yet been described whose general structure is such that the 

 application of this term to them appears to be justified. 



