56 Salisbury . — On the Structure and Relationships of 
four tracheides (0*0075 mm - x from 0*0075 mm.-o-oi 5 mm.) with thicker walls 
than those around. From the mesarch protoxylem there extends in the 
centrifugal direction an irregular fan-shaped group (PL V, Fig. 17, cf.x .), the 
elements of which, though larger than those of the protoxylem, are dis- 
tinctly smaller than both those which lie on their flanks and those on 
the inner side of the bundle. 
The centripetal xylem (PI. V, Fig. 17, cp.x.) exhibits a further dis- 
tinction in that the elements here are mostly cut more or less obliquely, 
so that the scalariform or reticulate sculpturing of their walls can be seen. 
A single layer of xylem elements on the centripetal side of the bundle 
immediately next the protoxylem do not show this character, although they 
exhibit the increased dimension. 
The bulk of the centripetal wood evidently then consisted of short 
tracheides, such as are frequently present in centripetal xylem that is 
becoming obsolete, as is exemplified in Lepidodendron v as cid are 1 or Osmun - 
dites KolbeiS 
(c) Additional Facts as to the Vasctdar System of Trigonocarpns 
P arkinsoni. 
The course of the sarcotestal bundles at the base of the seed of 
Trigonocarpns Shorensis cannot be traced, as the peripheral portion is lack- 
ing in the sections which pass through that region. 
Two sections probably belonging to Trigonocarpus P arkinsoni , viz. 
S. 34, a and S. 34, b of the University College collection, have however 
recently come to hand, which Prof. Oliver has also placed at my disposal 
for description, and which, owing to their excellent preservation, add con- 
siderably to our knowledge of the vascular structure in this region. The 
absence of similar data regarding our own seed, and the essential simi- 
larity between the vascular systems of the two species, are sufficient 
justification for inclusion here. 
The first section, viz. S. 34, #, just passes through the lower limit of 
the sclerotesta, and the second, parallel to the first, obliquely through 
the seed base, but without traversing the megaspore cavity. In this 
latter section eight ribs can be distinguished, of which three are doubt- 
less the primary ribs, and from the distribution of the remaining five it 
seems likely that the seed was twelve-angled, the missing ribs having 
become obliterated either by obliquity or in consequence of the lower 
level at which the seed is cut on one side. A further point of impor- 
tance is that this seed was evidently blunt at the base and not tapering, 
1 Hovelacque, M. : Recherches sur le Lepidodendron selaginoides , Sternb. Mem. Soc. Linn. 
Normandie, vol. xvii. 
2 Kidston and Gwynne-Vaughan : On the Fossil Osmundaceae, Pt. IV. Trans. Roy. Soc. 
Edinburgh, 1910. 
