49 
Trigonocarpus Shorensis , sp. nov. 
The individual sclerotic fibres were roughly rectangular, with a slight 
flattening in the radial direction, and usually about 2-5 x 4*5 yot. The wall 
was much thickened (about o*8 fi) and formed of an outer transparent, 
yellowish layer and an inner and slightly thicker brown layer, both of 
which show clearly defined lamellation. 
The longitudinal sections do not show the outer region of the sarcotesta 
preserved. The maximum longitudinal dimension in oblique sections is 
a quarter of a millimetre, and no doubt the real length was much greater. 
The width of the intervals between successive plates is usually fairly 
regular, but rarely they are seen in very close proximity even where there 
is no evidence of displacement, and these are, perhaps, like the larger 
aggregations, to be interpreted as due to anastomoses. 
Irregularly interspersed amongst the sclerotic cells are numerous 
secretory sacs, some 7*5 ' m diameter, which were no doubt situated in the 
soft parenchyma between the sclerotic plates, and become very numerous 
just beneath this zone. 
Here and there the secretory sacs occur two or three together, arranged 
radially, and it is possible that these too, as in the nucellus, formed rows 
alternating with those of the sclerized elements. 
Where cut obliquely, the secretory cells show fine longitudinal striation 
of the external surface of their walls ; in one section, which is especially 
well preserved, this wall is seen cut transversely as a series of dark beads 
separated by clear spaces of about equal width. Surrounding the carbonized 
contents is a clear space which probably represents a thick sclerotic wall : 
the elements, in fact, were almost identical in appearance with the thick- 
walled mucilage sacs found in the sporophylls and testa of present-day 
Cycads. 
In favourable cases, four or five connexions at more or less regular 
intervals can be seen between the dark central mass and the thin external 
wall. They probably represent radial pittings of the cell-wall, but must 
not be confused with the numerous and much finer and fainter radial 
striations, which are doubtless a matrix effect. 
Occasionally the secretory sacs were situated at one or other end of 
the sclerotic plates, or even in the middle, replacing the sclerotic elements 
themselves. 1 
The presence of the numerous hard plates in the peripheral zone of 
the sarcotesta must have given to that region considerable rigidity, and the 
dark layer often seen between it and the soft underlying tissue is no doubt 
to be interpreted as originating through the compression of the outer cells 
of this latter. 
(c) The Secretory System. The sarcotesta is not only distinguished 
from that of other seeds by its extreme development, but also by the 
1 Cf. Petiole of Medullosa. Scott’s Studies in Fossil Botany, Fig. 176. 
E 
