i^o] WRIGHT— PIT-CLOSING MEMBRANE 241 



formed ones figured here. The scalariform bars in the former 

 are fine and rather far apart, in the latter broader and joined in 

 such a way as to produce the reticulate effect of fig. 10. 



In both Botrychium and Helminthostachys the tracheids of the 

 root wood, although slightly smaller and more regular than those 

 of the stem, resemble these very closely. There is, perhaps, a 

 greater amount of open scalariform tertiary thickening than in the 

 stem and less of the broad, close formation. The petiole wood 

 of both forms is also a likeness in miniature of that of the stem, 

 particularly of the first-formed elements of the primary metaxylem 

 of the latter. Frequently, however, the pit pores in Helmintho- 

 stachys petiole are long and oblique rather than round. 



The presence or absence of a pit-closing membrane in the Ophio- 

 glossaceae, as in all the vascular cryptogams, has been a matter 

 of dispute. Russow (7), in illustrating his article of 1872, ex- 

 pressed the prevailing view of the anatomists of his time with 

 regard to the vascular cryptogams in general, when he showed no 

 membrane in the pits of either the side or the end walls of Bo- 

 trychium. It was in the following year that Sanio, working with 

 Pinus sylvestris, demonstrated beyond a doubt the presence, in 

 the mature condition in that form, not only of a membrane but 

 also of a torus. From that time the pendulum of opinion began 

 to swing in the opposite direction. In response to the stimulus of 

 Sanio's discovery, evidence has steadily accumulated that the 

 membrane in the vascular cryptogams remains in the pits of the 

 mature wood, not only in the side walls of the elements but, with 

 few exceptions, in the end walls as well. In 1908, however, this 

 view was challenged by Gwynxe-Vaughan (5). In returning to 

 the idea of the earliest investigators, that the membrane disappears 

 through resorption in the mature wood, the author distinguishes 

 two types of ferns, represented by Pteris and Osmunda respectively. 

 Ferns of the Pteris type, he claims, lose their limiting membrane 

 only from the pit cavities, while those of the Osmunda type lose 

 it also from between the walls of adjacent tracheids in the region 

 between the pits. Gwynne-Vaughan describes a further modi- 

 fication of this type which, however, need not be discussed 

 here, as he classes the Ophioglossaceae with ferns of the Pteris 



