Anatomy of Solenostelic Ferns : I. L ox soma. 91 
are also present. In many cases also, the cells bounding the 
intercellular spaces of the mesophyll are beset with the small 
rod-like structures described in L ox soma. If it is necessary 
to select any particular species for still closer comparison, 
Dicksonia apiifolia , Hk. ( Dennstaedtia ), and D. cicutaria , Sw. 
may be referred to as perhaps most suitable for the purpose. 
In the first place, these are the only two Ferns that I am 
as yet aware of, besides Loxsoma, in which the protoxylem- 
elements of the stem are all scalariform, and not aggregated 
into definite groups, but evenly distributed around its external 
periphery. Further, the small islets of parenchyma scattered 
amongst the sclerenchymatous ground-tissue, that form so 
characteristic a feature of the stem in Loxsoma , are also to 
be found in these two Dicksonias, with the same little rodlets 
projecting into their intercellular spaces. Cavity-parenchyma 
is present in the hippocrepiform petiolar meristele of Dicksonia 
apiifolia , and the sieve-tubes situated in the bays of the hooks 
of the xylem-strands have especially thick walls although 
they remain unlignffied. Whether the petiole of Dicksonia 
cicutaria presents the same structure I did not ascertain, 
because my material was accidentally lost, and I have not 
been able to replace it. 
The affinity thus brought to light between Loxsoma and 
the Dennstaedtiinae is in no wise discountenanced by com- 
parisons which may be drawn from the morphology of the 
sorus and sporangia. For Professor Bower ( 1 . c.) has shown 
that in this respect also a number of important characters 
are held in common by them both ; such as the marginal 
position of the receptacle, the position and form of the 
indusium, the basipetal succession of the sporangia, the 
obliquity of the annulus, and the relatively low output of 
spores (64). It must be borne in mind, however, that the 
consideration of the sorus and the sporangium also indicates 
affinities as close, or even closer, to* the Gleicheniaceae and 
Hymenophyllaceae. It is true that suggestive parallels may 
be drawn between the anatomy of Loxsoma and that of 
certain members of these families, but it seems to me that 
