462 OPHIOGLOSSALES 



The facts thus stated relating to the vascular structure of the stock in 

 the three genera are all consistent with a theory of origin from a type 

 with primarily a solid protostele, and subsequently a medullated monostele : 

 for the structure of the vascular system in the mature shoots of them all 

 is in point of fact a hollow cylinder perforated by the leaf-gaps: in 

 Botrychium and Helminthostachys, where these are less closely grouped 

 than in Opkioglossum, the fundamental structure as a vascular cylinder is 

 plainly seen. The opening of the cylinder to give exit to the leaf-trace 

 is a characteristic of that type designated by Jeffrey " phyllosiphonic," and 

 he distinguishes it from the " cladosiphonic type," in which the leaf-trace 

 passes off from the axial system without any opening. It has already 

 been pointed out that these two types are the anatomical expression of 

 the relative prevalence in the whole shoot of the axis in the cladosiphonic, 

 and of the leaf in the phyllosiphonic type. Supposing in any phyletic 

 series there should be an increasing dominance of the leaf, it would be 

 reasonable to expect evidence in the individual of a transition from the 

 one vascular type to the other. In the young plants of the Ophioglossaceae 

 themselves there is no indication of any such transition, for the young 

 plants are phyllosiphonic from the first. It will, however, be shown 

 later that on comparative grounds there is reason to think the origin 

 of the phyllosiphonic state in the Ophioglossaceae was from the clado- 

 siphonic, following upon an increase of proportion and importance of the 

 leaf. 



The leaf-trace itself is typically a single strand of the collateral type. 

 This is seen in Botrychium and Helminthostachys, and in most species of 

 Ophioglossum. The collateral strand may have its margins curved together 

 on the adaxial side, so that in the petiole of large leaves it may approximate 

 to a concentric structure, as in B. virginianum ; but this is merely a 

 modification of the collateral structure. Even in the large-leaved Helmin- 

 thostachys the leaf-trace comes off as a single strand, though it branches 

 very soon, in fact before the cortex is traversed, to form the numerous 

 strands of the petiole (Fig. 2571;). The condition seen in some species 

 of Ophioglossum is interesting for comparison with this, forming as it does 

 an exception to the rest of the family. In the section Euophioglossum 

 the leaf-trace comes off, as in other Ophioglossaceae, as a single strand, 

 which soon branches into three ; and this fact is embodied in Prantl's 

 diagnosis as amended by myself. 1 But in the section Ophioderma the 

 numerous strands of the petiole arc not united into a single strand at 

 the base : they are inserted as separate strands upon the vascular system 

 of the stock. It is still uncertain whether or not ^CJuiroglossa shares this 

 character. A comparison with other forms of Ophioglossum shows this con- 

 dition to be exceptional, and it is probably a derivative state, the separation 

 of the strands shown in other species only in the upper leaf having been 

 continued in ^Ophiodcrnia down to their actual base of insertion on tin 



1 Ann. />/'/>'(>/., xvili., p 215. 



