Anatomy of Solenostelic Ferns : L Loxsoma . 93 
recognize indications of a horseshoe design in the more or 
less curved form of the xylem-strand. This possibility had 
already suggested itself to me before I had the opportunity 
of reading the interesting paper recently published by Boodle 
upon the anatomy of this order 1 . He there describes a type 
of petiole, previously unknown to me, which goes far to 
bridge over the gap that existed between the petiolar strands 
of the Hymenophyllaceae and the hippocrepiform meristele of 
the petiole of Loxsoma . This type is to be found in Tricho- 
manes apiifolium , Presl, where not only the xylem-strand, 
but also the outline of the meristele itself, has the form of 
a horseshoe or arch, and further, the ends of the arms of the 
xylem are prolonged into incurved ‘ hooks ’ precisely as in 
Loxsoma. It also appears that there is a protoxylem-group 
in the bay of each hook, and probably one in the middle 
of the curved region. Trichomanes Prieurii , Kz., is also a 
most instructive type ; for here, although the xylem-strand 
exhibits the same form as in T. apiifolium , sometimes with 
two protoxylem-groups on the curved region, the outline 
of the meristele no longer follows that of the xylem, but is 
approximately reniform. This state of affairs will permit 
of a fairly close comparison with the meristele in Loxsoma at 
a point near the top of the rachis, where it becomes diminished 
in size and simplified in form (cf. Fig. 7, c), and especially 
so since this particular Trichomanes also presents cavity- 
parenchyma and fibrous elements, some of which may re- 
present sclerosed sieve-tubes. Granted these two examples, 
the petiolar strands of many other species fall into line as 
a series of progressive simplifications of the horseshoe design 
in accordance with the diminution in their size. Thus, in 
those of the type of Trichomanes scandens , L., the hooks at 
the ends of the arms of the xylem have practically disappeared, 
leaving a A or crescent-shaped strand with a protoxylem at 
the end of each arm and one median dorsal. The outline 
of the meristele itself is now circular. By the disappearance 
of the median protoxylem a still simpler type is produced, 
1 Ann. Bot., vol. xiv, p. 455, 1900. 
