i6 COENOPTERIDACEAE [CH. 



there are many features which indicate for them a relatively primitive 

 position in that Class, such as would fully accord with their early geological 

 history. This is shown by the protostelic structure of the stem, whether 

 short with crowded leaves or elongated with leaves inserted at long intervals. 

 They share also with primitive Ferns the radial construction of the shoot as 

 a whole, and the monodesmic leaf-trace, which shows at the base of the 

 adult leaf a structure such as is seen only in some of the most primitive of 

 living Ferns. The much-branched distal region of the leaf and its narrow 

 segments with dichotomous branching again afford features that are primi- 

 tive. The hairs though apparently of complex structure are actually of the 

 simple transversely septate type, characteristic of the earlier Ferns. The 

 sporangia of B. forensis are grouped in simultaneous sori after the type of 

 the Simplices. They are not of the largest size: but from their spore-output 

 and complex annulus they clearly belong to the Eusporangiate series. All 

 these characters, vegetative and propagative, taken together indicate, at 

 least for the genus Botryopteris, that it comprises Ferns of an extremely 

 primitive type. 



A further point of interest is the suggestive relation of axis and leaf 

 afforded by the shoot of B. cylindrica. The elongated axis dichotomises: 

 the shanks of the dichotomy may be equal or unequal. The latter is illus- 

 trated by Fig. 312, A. The double protoxylem of the weaker shank is 

 immersed in the metaxylem, as also are the three protoxylems of the 

 stronger; but it is nearer to the inner face than the outer of the slightly 

 oval xylem. The origin of the leaf-trace resembles the origin of the supply 

 to a weak branch very closely indeed (Fig. 312, B). While there may be 

 two or three internal protoxylems lying centrally in the stem-stele, in the 

 petiolar trace the protoxylem (represented by two almost confluent strands) 

 lies at the adaxial surface of the xylem. It may even be temporarily 

 enveloped higher up by a few tracheides of metaxylem on the adaxial side, 

 but later it emerges, being directed towards the upper surface of the leaf 

 (Compare Fig. 153, Vol. I, p. 161.) This envelopment of the protoxylem, 

 which is temporary and inconstant in B. cylindrica, appears to be a regular 

 feature of B. antiqiia, in which adaxial metaxylem is present at the level 

 where the trace separates from the stele of the stem (Fig. 314). Thus in this 

 most ancient species the leaf-trace at its base resembles very closely the 

 vascular supply passing to the weaker shank of a dichotomy of the axis. At 

 the same time its structure as seen in B. antiqua resembles that of the 

 extreme base of the leaf in Thamnopteris, a primitive member of the Osmun- 

 daceous series (Fig. 155, 2, Vol., I, p. 162). 



The similarity of anatomical structure thus seen in B. cylindrica between 

 the leaf-base and a weak branch points towards the conclusion that the 

 leaf of a Fern may actually be phyletically a shank of a dichotomy, as was 



