12 Transactions. 



stages of the tree-fern the stem f.v. elements are essentially con- 

 nected with leaves, though later there is some differentiation be- 

 tween the leaf-traces to provide a complete ring and to prepare 

 for the insertion of leaves higher up the stem. But probablv the 

 mode of growth in Gleichenia and Loxsoma is the more primitive, 

 and that in the ferns with crowded leaves is a later development, 

 leading to the reduction of the cauline strand. In the primitive 

 types we may assume that the first bundle system was differ- 

 entiated to minister to the needs of a rapidly elongating spike 

 or strobilus. Then, as the vegetative appendicular organs 

 became larger, strands would be differentiated in them, and 

 fit on to the central strand. Then later, as the leaves outnum- 

 bered the sporophylls and the leaves were crowded on the stem, 

 the cauline strand was reduced, and on some ferns practically 

 gives way to leaf-traces. 



(2.) How did the transition to the solenostele take place ? 

 Of course, we can see that the ring is a better arrangement 

 of the f.v. elements than the solid strand. If the stem is to be 

 upright and bear a crown of leaves, only a few xylem elements 

 will be needed, and these will strengthen the stem more and be 

 better placed for leaf-insertion if they are in a ring ; and the 

 large undifferentiated pith may serve as a starch and water 

 reservoir. 



But how did the ring develop from the solid strand ? Now, 

 in Gleichenia we have a solid strand in the stem and a curved 

 strand in the leaves ; and when a part of the stem stele is cut 

 off for the leaf, the segment remains attached to the main stele 

 while it is assuming a curved form ; and especially in G. flabellata 

 the meristele remains attached at its edges to the stem stele 

 for some time. Thus are formed " nodal islands." 



Tansley and Lulham suggest that by the continuation of 

 the nodal islands through the internode above and below a 

 structure like Lindsaya would be reached — Lindsaya being, then, 

 a phytogenetic link between the protostelic and solenostelic 

 types. 



But it seems probable that the transition has taken place 

 quite independently in several groups, and the process need 

 not be similar in all. In the Schizceacece the protostele is pro- 

 bably primitive, but siphonostely and even polystely has been 

 reached in Anumia (Boodle). Similarly hi the Glekheniacea 

 the protostele persists in many forms, but a solenostele has arisen 

 in G. pectinata (Boodle). 



In the Marattiacecp, from the life-history of Angicphris (Fanner 

 and Hill, 1902), the change from protostele to siphon., stele is 

 due to parenchyma cells appearing in the centre of the xylem 

 and the leaf removing a segment stretching to this pith. The 



