72 Sykes. — The Anatomy and Morphology of Tmesipteris. 
latter . 1 It can be shown that the processes involved are essentially similar. 
The division of the protoxylem group, which is the first step in the 
formation of the branch-trace, takes place about two centimetres below the 
departure of the latter from the stem. The strand of xylem which is to 
pass out as a branch-trace, is soon nipped off, and runs for a considerable 
distance as a separate constituent of the stem stele, seemingly forming 
an additional strand (Text-figs. V and VI, bt ly _ bt 2 ). It then passes 
out, becoming in the ordinary way enclosed by phloem, and leaves a gap 
(Text-fig. VI, g) behind it in the ring of xylem strands. Hence the 
apparent difference in the appearance presented at the point of departure 
of leaf- and branch-traces is that in the former case a strand of xylem is 
left in the stele exactly opposite the trace, while in the latter there is a gap 
in that position. Yet the only real cause of this difference is that in the 
one case the division of the strand to form the trace takes place earlier 
than in the other. 
The lower fertile branches seem to receive their traces in the same 
way as the leaves, or, at any rate, less time elapses between the formation 
and departure than is required by the later 
branch-traces. Almost immediately after 
a branch-trace has gone off, or sometimes 
even before, another of the xylem groups 
divides to form a new branch-trace which 
remains for some time in the stele and brings 
the number of strands up to the original 
number (Text-fig. VI, 7 c, bt 7 ). In one or 
two cases both products of the division of a 
group were seen to pass out as branch-traces 
fairly low down in the stem (Text-figs. V, 
VI, b, bt. 2 and bt 4 ) ; always, as the apex is 
approached and the xylem groups become 
confluent, obliterating the pith, one group 
after another behaves in this manner, the 
number thus being gradually reduced, until finally only one group is left. 
Here in one series of sections four leaves occurred, and their leaf-traces 
were formed in the ordinary manner by the splitting off of a strand of 
xylem and its enclosure by phloem and subsequent rapid departure 
from the stele. In this same series there was next formed a branch, its 
trace being of necessity also produced by a similar division of the only 
remaining group. This division took place longer before the departure of 
the trace than was the case in the formation of the four leaf-traces just 
mentioned. Finally a branch w^as formed which limited the growth of the 
main axis (Text-fig. VII). As described above , 2 a synangium is usually borne 
1 Dangeard, 1 . c., p. 179. 2 Ibid., pp. 65-6. 
Figure VII. Apex of an 
adult specimen, showing a fertile 
‘ branch ’, one limb of which is 
branched again. 
