352 MR R. C. DAVIK ON 



marginal set of tracheides. The method of pinna-supply in Blechnum orientals, L. 

 (PL XXXIV. fig. 11), is the most nearly marginal among the extramarginal types. 

 Only one or two of the marginal tracheides remain when the pinna-trace departs, but 

 the method of separation of the pinna-trace is certainly extramarginal. The typical 

 lengthening of the hook, the bridging of the narrow space between the adaxial curve 

 and the margin of the hook, and the retention of the marginal tracheides are all 

 found in the process in this Fern. 



' Thin partitions " divide the bounds of marginal and extramarginal types. Gymno- 

 gramma Pearcei, Moore, var. robusta (PL XXXIV. fig. 1 2), Ceropteris calomelanos 

 (L.) Unci., and C. calomelanos, var. chrysophylla, Klf., have the same type of leaf-trace 

 strand as Blechnum orientate. The pinna-trace goes off from a lengthened leaf-trace 

 strand, which, however, re-forms its hooked extremity from a swelling of tracheides 

 back from the margin. This swelling never catches up the marginal set of tracheides 

 of the original leaf-trace. The Gymnogramma type is but a step from the type 

 described in Blechnum orientale. 



Histiopteris incisa (Thbg.) J. Sm. (PL XXXIV. fig. 15 and PL XXXV. fig. 16) illus- 

 trates a curious combination of the two types in its process of pinna-supply. The 

 leaf-trace has the outline of that of Balantium culcita — a flat-topped arch with in- 

 curved sides. Small hooks are present at its extremities. The hooked extremity 

 lengthens in the usual way and gives off its tip in marginal fashion. Simultaneously 

 the projecting corner of the arch beside it also lengthens, and from it is nipped off a 

 ring of vascular tissue in the extramarginal manner (PL XXXIV. fig. 15). The two 

 strands pass towards the base of the pinna, come together, and by the opening out 

 of the adaxial side of the extramarginally-derived portion form a pinna-trace 

 exactly like the leaf-trace (PL XXXV. fig. 16). 



In one or two broken leaf-traces a reminiscence of this combined process seems to 

 appear. All three strands are unhooked in Aspidium Moorei (Hk.) Diels. Part of 

 the pinna-trace comes from the tip of the adaxial strand, part from the median strand 

 — those parts which correspond to tip and arch-corner in Histiopteris incisa. The 

 pinna-trace of Leptocliilus cuspidatus (Pr.) C. Chr. also comes from portions of the 

 leaf-trace corresponding to those supplying the pinna in Histiopteris incisa. The 

 leaf-trace is made up of six or seven strands (text-fig. 2). The adaxial pair have 

 incurved extremities and give off" the backs of their hooks to supply part of the 

 pinna-trace. The rest of the pinna-trace comes from the two subsidiary strands next 

 in order to the adaxial strand as we go towards the abaxial side of the petiole. These 

 subsidiary strands have simple plates of xylem, from which the strands going to the 

 pinnae are simply nipped off. 



The pinna-traces of most of the Cyatheacese bear a resemblance to these. 

 Gwynne-Vatjghan ('03) has called attention to the distinction made by Bertrand 

 and Cornaille ('02) between the two regions of the petiolar trace, (l) the abaxial 

 curve, and (2) the adaxial arcs. The pinna-trace of Cyathea Brunonis comes partly 



