390 West . — A Contribution to the Study of the Marattiaceae. 
internal strands continue their course from the pulvinus far up into the 
region of the rachis, where, according to Bertrand and Cornaille (2, pp. 1 63 
and 212), the primary branches of the petiole are supplied direct with 
strands from the internal system, in addition to those they obtain from the 
outer ring. 
5. Angiopteris . 
The adult sporophyte of Angiopteris , which closely resembles in size 
and in habit that of most species of Marattia , also has a bulky tuberous 
caudex surmounted by a rosette of huge leaves, which are usually twice- 
pinnate. 
The first account of the vascular anatomy of the genus Angiopteris 
was published in 1853 by Harting (De Vriese and Harting, 27 ), who found 
that the vascular system of the stem consisted of a confused network of 
bundles which traversed the ground-tissue in all directions. This investi- 
gator, however, not only failed to distinguish the foliar traces from those 
of the intracortical roots, but made no attempt to trace out the actual 
course of individual vascular strands. 
In 1864 Mettenius ( 49 ) published an important monograph in which 
the course of the bundles in the stem and leaf-bases of an old plant of 
Angiopteris was described in great detail. He showed that the very 
numerous vascular strands of the tuberous stem form a number of concentric 
cylinders of anastomosing bundles each having the general form of an 
inverted cone. Meshed segments leave the outermost cylinder to supply 
the leaves ; the gaps in this cylinder formed by the departure of the leaf- 
traces are compensated by corresponding segments from the next inner 
zone. In a similar manner segments from the third cylinder compensate 
for the gaps produced in the second, and so with successive inner 
cylinders. Mettenius ( 1 . c., p. 590) also stated that strands from the second 
cylinder sometimes contribute directly to the leaf-trace. 
De Bary (1) pointed out that the stem-stele of a young plant of 
Angiopteris consisted of a c Biindelrohr 5 with foliar gaps, while Leclerc 
du Sablon ( 54 ) gave a brief account of the vascular anatomy of the young 
plant as it appeared in a series of transverse sections of the stem. 
Shove ( 58 ) described the vascular anatomy of a large specimen of Angio- 
pteris and in most of the important details confirmed the results obtained by 
Mettenius. This plant, however, exhibited a dorsiventral structure which 
was especially well marked towards its base, where the lower surface was 
quite destitute of leaves, but thickly covered with roots. This observer ( 58 , 
p. 505 ; PL XXVIII, Figs. 5 and 6) found that, correlated with the external 
dorsiventrality of the plant, the meshes of the vascular network were much 
longer on the ventral root-bearing side than on the dorsal leaf-bearing side. 
All the strands of the leaf-trace were derived solely from the outermost 
