Pteridophyten. 311 



petiole of Dariaea; but the present writer adduces very numerous 

 cases observed by himself and by other botanists in which the leaf 

 trace is mesarch at its very base: i. e. in the cortex of the stem. 

 This is held to be a primitive condition and evidenee from fossils 

 is adduced to support this theory. From P. Bertrand's researches 

 it would seem that a very primitive condition of the trace was a 

 mesarch one with two protoxylems; the monarch State was probably 

 still earlier for it is found at the attachment of the trace of Clepsi- 

 dropsis and Asterochlaena to the stele and in the Lycopods, Psilo- 

 tales and Equisetales. The Ophioglossaceous and Osmundaceous 

 traces are monarch — the latter clearly primitively so, as shown by 

 Kidston and Gwynne Vaughan's researches on their geological 

 history; it seems not unlikely that the Osmundaceae are related to 

 the latter order and that these two families branched off from the 

 primitive fern stock, before the leaf bündle had become complicated. 

 It is also held that the primitively diarch, double Marattiaceous 

 bündle may have arisen from the diarch mesarch Strand already 

 constricted of certain Zygopterideae. Psaronius would be an early 

 side branch from these diarch Zygopterideae. The origin suggested 

 for the great body of Ferns included under the forms with the base 

 of the leaf primitively triarch is that they are derived from diarch 

 forms by the division of the two protoxylems at the end of the 

 more or less elliptical trace and the fusion of the two resulting 

 abaxial protoxylem groups in a median and mesarch position, such 

 a process being, perhaps, foreshadowed in Asterochlaena. Moreover 

 from the readiness of the median protoxylem group to divide in 

 the course of the higher development of the trace in living Ferns 

 it may perhaps be inferred that fusion is not yet complete in all 

 cases. Owing to the retention of centripetal woody elements in the 

 petiole Lygodium seems to be the most primitive member of the 

 Schisaeaceae ; its primitiveness seems also to be shown in the 

 isodiametric form of the petiolar bündle; its exarchy is not, how- 

 ever, considered primitive and the species with centripetal xylem 

 only are regarded as derived from Lygodium palmaturn with a 

 mesarch trace. Among the Gieicheniaceae the petiolar structure of 

 Eugleichenia is regarded as more primitive than that of Mertensia, 

 while Platysoma, owing to reduction has but two small protoxylems; 

 it is. further, collateral and endarch in the cortex, concentric in 

 the petiole and apparently mesarch in the rachis. The trace of 

 Matonia, especially of M. sarmentosa, is little modified from the 

 hypothecated primitive type. The Dipteridineae appear to have been 

 derived from the Maionineae and their trace is of the same type, 

 though the protoxylems are never so few as three. In the Hymeno- 

 phyllaceae the trace is often triangulär with three or more proto- 

 xylems; but in numerous cases it is by reduction a monarch colla- 

 teral bündle. The similarity of the stouter forms to the Schisaeaceae 

 and to the Gieicheniaceae is opposed to Boodle's, Bower's and 

 Tansley's views, according to which the stouter forms have arisen 

 by amplification of such a type as Hymenophyllum dilatatum. They 

 seem to represent a reduction series starting frorn Trichomanes 

 scandens and ending in the collateral Hymenophyllums. The Loxsoma, 

 Dennstaedtia , Dicksonia alliance is not far removed as regards the 

 trace from the hypothetical form with a triarch triangulär trace, 

 though the protoxylems tend to be more numerous. Bertrand and 

 Cornaille's Orwclea trace is found in most of the smaller and 

 simpler Mixtae; it consists of two curved diarch bundless; in every 



