1041 
Affinities of Sutcliffia. 
secondary wood, and in XXI it has become entirely free from it. The 
secondary xylem closes up immediately after the exit of this bundle, and 
none of it accompanies the strand during its outward passage. Imme- 
diately it becomes free from the c meristele 5 it turns vertically upwards and 
can be traced as far as XXXI (Text-fig. 4) ; after this level it is no longer 
present in the series. The strand consists of a small group of primary 
tracheides mixed with parenchyma, and only one protoxylem group 
appears to be present. Its preservation is far from satisfactory, but it 
is certainly concentric in structure, though how the phloem was gained 
could not be ascertained. In the upper sections in which it appears the 
phloem has associated with it the ring of the large elements which are 
so characteristic a feature of the phloem of the leaf-trace bundles ; bundles 
of fibres also occur immediately beyond the phloem. There can be no 
doubt that this strand is a leaf-trace bundle of the unilateral type, passing 
out directly and prematurely from a { meristele ’. 
In tracing the course of the ‘ meristele * y the central region of it shows 
signs of injury quite early in the series, and in XIX there is a complete 
break at the middle part. A division of the ‘ meristele ’ into two portions 
in all probability occurred in life, but the appearance in the fossil is con- 
siderably emphasized by crushing and injury (Text-fig. 9, Figs, XII, XXVII, 
and XXIX). A small portion, y 3 , is soon cut off from y 2 , and subsequently 
y 1 and y 2 appear to fuse. The evidence for this re-fusion is however doubtful, 
for considerable crushinghas taken place in the upper portions of the stem, and 
added to this there is further injury in preservation at the critical region. In 
any case, by the time XLVIIIis reached (Text-fig. 9, Figs. XXXIII-XLVIII) 
y has divided into two larger and one small portion, y 1 , y 2 , and y 3 . Above 
this level the primary wood of y 1 is dividing up into smaller portions, which 
do not leave the c meristele 5 in the course of our series ; so far as one can judge 
from this stage, the ‘ meristele 5 y 1 is dividing into radially symmetric leaf- 
trace strands. Division also begins in y 3 , and three small strands of 
primary xylem — each with a single protoxylem group — are produced ; they 
begin to pass out through the secondary wood, but they cannot be traced 
to the stage where they gain the phloem zone. On this account the 
unqualified statement that the { meristele 5 y 3 is used up entirely in the pro- 
duction of unilateral leaf-trace strands cannot be made, but comparison with 
the similar bundle given off by the ‘meristele 5 y in Section XVIII leaves 
little room to doubt that it is correct. 
In summarizing the evidence provided by the course of the three 
‘ meristeles ’ a, ( 3 , and y, it appears evident that large masses of vascular 
tissue are cut off from the central axis, that these run parallel to the latter 
for some distance, giving off leaf-traces and ultimately dividing up into 
smaller strands, often unequal in size, and that the primary wood of these 
strands was ultimately entirely used up in the production of leaf-trace 
