462 BOTANICAL GAZETTE [JUNE 
Examination of serial sections through nodes of a cone-bearing 
branch of E. palustre and of a fertile stem of E. sylvaticum showed 
that the continuity of the protoxylem strands through the node, 
their forkings and fusions, could very well be followed in these 
species. In both the course of the protoxylem at the node corre- 
sponds with that of the protoxylem in FE. gigantewm. The dis- 
tinction between nodal wood and protoxylem is very clear, although 
the innermost nodal tracheids are relatively small. In both species 
the persistent protoxylem is sometimes locally separated from the 
inner edge of the nodal wood by a parenchymatous cell or two. 
In E. sylvaticum the persistent tracheids of protoxylem are relatively 
numerous, and some of the medianly situated ones may persist for a 
little distance above the departure of the tracheids of the trace. 
In a young node of a fertile stem of EF. arvense that had not yet 
appeared above ground it was possible to observe the persistence 
of the protoxylem and indications of its forkings and fusions at 
the node. The phenomena, however, were not so clear. 
Serial sections through a vegetative node of the cone-bearing 
branch of E. debile showed that in this species also two small groups 
of protoxylem, consisting of a few tracheids only, persisted after 
the departure of the trace. On the breaking up of the nodal xylem 
these two small strands entered separate but adjacent bundles. 
Each of the newly constituted bundles, therefore, possessed two 
of these small strands of protoxylem. These, however, did not — 
always fuse to form the protoxylem strand of the internode. Fre- 
quently one, sometimes both, seemed to die out. In the latter 
case the protoxylem of the upper internode was unconnected 
(except by nodal tracheids) with that of the lower internode. 
In other cases a few elements of one or both branches of protoxylem 
linked up the protoxylem of one internode with that of the other. 
In the specimens of E. hiemale of which I have examined serial 
sections, the protoxylem disappears at the nodes; in other words, 
the whole of it passes out into the trace. In this species the inter- 
nodal bundles are relatively far apart, and the bundles of suc- 
cessive internodes are formed by the oblique course and fusion of 
adjacent halves of neighboring bundles. During the oblique 
course of the halves it is easy to see that all the tracheids com- 
