58 Salisbury . — On the Structure and Relationships of 
out in the sarcotesta. What was doubtless the second half of this strand 
is seen at the edge of the bundle cavity, and clearly shows that the 
course of the two halves was divergent. 
In the next section, viz. S. 34, b, which is nearly parallel to the former, 
five sarcotestal bundles can be distinctly recognized, whilst the position 
of a sixth is marked by a lacuna. 
Where the bundles are cut on the lower side of this section, the level is 
only about one-third of a millimetre above that at which the central bundle 
is cut in the previous slide, so that the distance is scarcely sufficient for 
other bundles to have been given off. 
It would appear then, from these two preparations, that three bands of 
xylem in continuity with the peripheral part of the chalazal strand were 
given off simultaneously, each of which then almost immediately underwent 
bifurcation into two divergent halves. 
The soft parenchymatous tissue which surrounded the bundle consisted 
of cells which, when cut transversely, appear more or less polygonal and 
isodiametric : where cut longitudinally they measure about 0-12 mm. x 0-022 
mm., and are separated by slightly oblique transverse walls. This tissue 
may perhaps have served the purpose of phloem, though apparently quite 
undifferentiated ; that it had some definite function is supported by its 
sharp delimitation from the surrounding tissue, even where this too was 
parenchymatous in nature. A similar sheath accompanied each of the six 
sarcotestal strands in this, the basal region of the seed. 
As seen in S. 34, b (PI. V, Fig. 15) the central part of each bundle was 
occupied by a strand of elongated scalariform tracheides ( n.tr .), and around 
this was a complete ring of much shorter and stouter spiral or slightly 
reticulate elements ( s.tr .). Separating the outer zone from the central core, 
was a narrow space from which the bulk of the tissue has perished, but here 
and there can be seen patches of elongated parenchymatous elements similar 
to those forming the bundle-sheath, and no doubt they originally constituted 
a complete ring. 
Each sarcotestal bundle was then essentially similar to that supplying 
the chalaza. An examination of the bundles of Dr. Scott’s section 626 
of T. Pcirkinsoni through the same region as S. 34, b, though exhibiting 
far less perfect preservation, seems to agree with the description here 
given. 
Before leaving these sections it should be said that, though the type of 
preservation is the same as that of the bulk of the specimens of T. Shorensis , 
in none of the secretory elements present was I able to find the longitudinal 
striation characteristic of the wall in that species, though this feature is 
exhibited by well-preserved ducts in the stems of Medidlosa anglica. 
