54 Salisbury. — On the Structure and Relationships of 
bands, the width of the more clearly delimited varying between i mm. 
and i*3 mm. 
The internal diameter of the megaspore cavity, as seen in section S. 33, c 3 
from which the above data were obtained, is 4.4 mm., which represents 
a periphery of approximately 13-8 mm. If we assume each band to have 
been of equal width, this gives us 1-15 mm. as the measurement of each if 
their number were twelve. This figure is sufficiently close to the actual 
measurements to support the belief that this was the actual total of the 
bands. Unfortunately, none of our other preparations furnish any corro- 
borative data, but it is worthy of note that the repeated hexamerous 
character of the seed-structure again recurs in this figure. 
The next section (S. 33, d), though ill preserved, shows the tracheides 
were completely segregated into lenticular strands which were much narrower 
(0*25 mm. -0.5 mm.) and more numerous than at the preceding level (probably 
over twenty). 
These strands were probably mesarch in structure, of from four to five 
elements in the radial plane, diminishing both in number and size on either 
side. In some cases, tracheides at the edge of the xylem bands can be seen 
passing horizontally, as is the case in Trigonocarpus Parkinsoni } but actual 
connexions were not observed. 
Tangential sections show that the individual tracheides were often 
oblique or curved, and the appearance of horizontally directed elements, in 
the slightly oblique transverse sections, might easily be accounted for in this 
way. In spite of the excellent preservation, no indication of phloem has 
been observed in relation to the nucellar strands. 
The final section of the transverse series only contributes confirmatory 
evidence that the bands of xylem were disconnected, and that their smallest 
elements were centrally placed. 
For the further course of the bundles only longitudinal and oblique 
sections are available. In these, the bands of tracheides can be traced to 
what must have been a level close to the floor of the pollen chamber. Only 
two sections, viz. S. 32, c and Section I of Dr. Scott’s series, actually pass 
through this structure ; neither of these shows any tracheides within, but the 
internal preservation of the latter section is very poor, whilst the former has 
a broken and ill-preserved floor. 
The vascular elements of the nucellar system appear to be very uniform 
in structure (0-022-0.03 mm.), and exhibit scalariform thickenings which 
occasionally show anastomoses between adjacent horizontal bars. (Interval 
between bars, 0-0015-0-003 mm.) 
From the above description it will be seen that the internal vascular 
system is almost identical with that described by Scott and Maslen for 
Trigonocarpus Parkinsoni (loc. cit., pp. 120-1), and it may be of some 
1 Scott and Maslen, loc. cit., PI. XIV, Fig. 13. 
