the Anatomy of the Cone and Fertile Stem of Equisetum. 443 
the annulus, between this level and the forking of the metaxylem strands. 
We may thus conclude that the forking of the protoxylem, whether it 
occurs somewhat below or a little above the insertion of the annulus, is 
probably, like the forking of the metaxylem, a reminiscence of a former 
node. Another explanation of the early division of the protoxylem 
would be that if the annulus is a reduced leaf-sheath only the free part 
is developed, the basal part being presumably concrescent with the axis. 
In Cones A and B of E. sylvaticum the supra-annular anastomoses are 
relatively numerous. 1 Among the species studied the only one which 
shows a proportion of such anastomoses at all comparable to that of 
E. sylvaticum is E. hyemale (Browne ( 3 ), pp. 242-3 and Text-figs. 2-4). 
In Cone B of E. sylvaticum eleven out of thirteen parenchymatous meshes are 
closed slightly above the insertion of the annulus and eleven fresh meshes 
arise, one of the first, one of the second, three of the third, two of the fourth, 
three of the fifth, and one of the ninth orders. In Cone A of this species 
nine out of twelve parenchymatous meshes are closed slightly above the 
insertion of the annulus. As already indicated (p. 432), I interpret the two 
parenchymatous meshes arising slightly below and to one side of the 
first and second traces of the lowest whorl of this cone as supra-annular 
in origin. This interpretation gives five meshes as originating above the 
annulus, two of the first and three of the second order. 
As the basal whorl of Cone A of E. sylvaticum consists of fifteen 
sporangiophores, and there are only twelve bundles in the lower part of the 
internode below the annulus, the forkings of the protoxylem and metaxylem 
in the neighbourhood of the insertion of the annulus of this specimen do not 
afford so good an argument in favour of the presence of a vestigial node as 
do the corresponding anastomoses in Cone B, where the lowest whorl of the 
cone consists of thirteen sporangiophores, and there are thirteen bundles in 
the internode below the annulus. Moreover, in Cone B the greater elonga- 
tion of the region between the insertion of the annulus and the lowest 
sporangiophores makes the relations of the strands to one another particu- 
larly easy to follow. Though there are, of course, slight differences in size 
between the bundles and their carinal canals, of which each bundle 
possesses one, it seems clear that the thirteen bundles of the internode 
below the annulus are equivalent, and arose in the usual way above and 
between an equal number of leaf-traces given off at the whorl below. 
Nevertheless, but one strand of Cone B passes unbranched through the 
1 It is interesting to note that according to Milde the branches, which are produced after the 
maturity of the cone by the originally unbranched fertile stems of E. sylvaticum, often develop 
immediately below the annulus (Milde, p. 288). Duval-Jouve writes (p. 147) that this anomaly is 
rather rare in France, but is presumably much more frequent in Switzerland, since Bernoulli, 
always so accurate, mentions the presence of branches below the annulus as a distinctive character of 
E. sylvaticum. This phenomenon, whether rare or common, appears to be a further indication 
of the less great reduction of the annulus in this species compared with most others. 
