180 Ethel N. Thomas. 
Instead of the horseshoe leaf-trace, so frequently correlated 
with the solenostele, we meet with a bewildering number of small 
bundles, at the first glance giving almost the impression of being 
arranged in three concentric series. The position of the proto- 
xylem elements in these bundles at once attracts attention 
(Text-fig. 36). 
The bundles forming a row along the flattened upper (adaxial), 
side' of the petiole, have their protoxylem groups uppermost in the 
normal manner (Text-fig. 36). The protoxylem groups of the semi¬ 
circle of bundles arranged around the periphery of the convex under 
(abaxial) side are also directed towards the upper surface but 
within them is a smaller semicircle of bundles with inversely 
orientated protoxylem groups. This series, however, appears to 
be continuous at its ends with the outer series, and gives rise to 
the suspicion that they represent a lateral indentation of bundles 
arranged on the horseshoe plan. 
The central internal system, and the upper row of bundles 
first alluded to, can similarly be explained as elaborations of the 
free ends of the horseshoe. 
From an examination of the petiole alone, Mr. Gwynne- 
Vaughan, in a letter to Professor Oliver, has expressed his belief 
that the vascular system of the petiole of AcrosticJiuin aurenm can 
be “ thus resolved.” I fortunately had in the material at my disposal, 
a young plant from which a microtome-series of the base of one 
of its youngest fronds was obtained. It revealed the petiolar 
vascular system described above, but in its simplest terms (Text-fig. 
34). The peripheral series of the lower side is composed of six 
bundles; within this is the inverted series consisting of only four 
bundles—two on each side and separated by a space. The central 
system is represented by one large bundle with several protoxylem 
groups. It is only slightly displaced from the middle of the row of 
bundles on the upper side. 
Node. 
A series of transverse sections was made through the stem, and 
they probably included the insertion of the petiole just described. 
As the node is approached from below, the first sign of the 
departure of the leaf-trace, is the bulging of the solenostele at one 
of its angles to form a horseshoe-like loop (Text-fig. 27). Very soon, 
however, the continuous horseshoe is broken up into bundles. This 
1 In Text-figs. 34 to 37 the upper (adaxial) side of the petiole is 
turned downwards to correspond with the position of the leaf- 
trace in Text-figs. 27 to 33. 
