40 6 West.— A Contribution to the Study of the Maratiiaceae. 
archegonium, grows straight upwards and eventually pierces the prothallium, 
emerging on its upper surface (cf. Campbell, 20 , p. 138, Fig. 108, A). The 
root, which in this family is a strictly endogenous structure, as soon as 
its apex is established, rapidly increases in length, and with the cotyledon, 
which in the meantime has been growing actively upward, forms almost 
a straight line, so that the young sporophyte may be described as bipolar 
(cf. Campbell, 20 , p. 145). Thus the evidence derived from a consideration 
of the Marattiacean type of embryogeny points clearly to a radially 
organized ancestry for this group of Ferns. 
Strong evidence in favour of this view can also be derived from 
a comparative study of the structure and development of the skeletal 
framework in this family. In every genus and species, with the exception 
of Kaulfussia , the shoot of the young sporophyte, for a time at least, is 
radially organized, the earliest leaves forming an irregular spiral around the 
stem, and in the case of Marattia , Macroglossum , certain species of Danaea 
(e. g. D. nodosa , D. elliptic a , and D. simplicifolia ) , and probably Archangio- 
pteris , this radial organization of the shoot is permanently retained. How- 
ever, in other species of Danaea , as for example D. alata , D. Jenmani , &c., 
the apex of the shoot sooner or later bends over and growth proceeds 
horizontally, the leaves being for the most part confined to the dorsal and 
lateral surfaces of the rhizome whilst the roots arise from all sides of 
the stem. 
Lastly, we have in Katdfussia a plant in which a dorsiventral structure 
is indicated at a very early stage, the second leaf arising quite close to 
the cotyledon and on the same side of the stem-apex (cf. Campbell, 20, 
p. 156 ; 19 , p. 79). The distichous arrangement of the leaves, thus early 
initiated, is retained permanently on the slender creeping rhizome of this 
genus. Campbell (l.c., p. 218) maintains that Kaulfussia is probably the 
most primitive of the living representatives of the Marattiaceae ; in the 
opinion of the present writer, however, this view is untenable, since not only 
is this genus the most aberrant type externally, but the comparatively 
simple type of vascular anatomy found in the adult rhizome is to be 
regarded as a derived or specialized condition correlated with the habit 
of the plant. 
I have omitted Angiopteris from the series traced above because there 
has been considerable divergence of opinion with regard to the symmetry 
of the axis of this plant. According to the observations of Campbell ( 20 , 
p. 201), plants of Angiopteris growing upon level ground are always strictly 
radial in structure. 
Farmer and Hill ( 29 , p. 380) remark that ‘both Marattia and 
Angiopteris also exhibit a tendency to dorsiventrality, but it is not very 
marked in young plants ’, while Shove ( 58 , p. 531) states that the stem of 
Angiopteris examined by her presented definite dorsiventrality. Charles 
