THE PALZEOZOIC SEED LAGENOSTOMA OVOIDES. 465 
median longitudinal section, which is known to be approximately correct, 
since it accords with all available sections (except К. 28) within the limits 
of individual variation. That such occurred, as in recent plants, would be 
difficult to show directly, since it would require at least two sections, known 
to be cut in precisely similar directions through two different seeds ; but it 
may be inferred from а comparison of the oblique sections В.48 (2) and 
R.51, which both pass through the base, the former cutting the bundle, 
while the latter does not. Now if the two seeds were identical in size, it is 
evident that the base of a section which passes through the bundle should be 
of greater dimensions than the base of any other section, assuming that the 
bundle enters symmetrically. This we are entitled to do both оп compara- 
tive grounds and from the entire organization of the seed, although we have 
as yet no transverse section of the chalazal region. The base of В. 51 is, 
however, fully 7 °/, longer than that of К. 48 (2), which thus indicates that 
individual variation to the extent of, say, 10°/) characterized these seeds. 
With this margin, all sections can be plotted on the median section except 
R. 28, which cuts the neck of the lagenostome, and yet from the configuration 
of the integument falls considerably below this point. As this appearance 
does not tally with any of the oblique longitudinal sections, we must conclude 
that it is abnormal, and that the whole lagenostome has been in some way 
dragged down into the body of the seed. 
X Terminology. 
Owing perhaps partly to the fact that the delicate tissue of the integument 
was never preserved in any of Williamson’s specimens, he did not recognize 
the latter as a definite unit of structure, but called the hard enveloping shell 
of the integument the “testa,” and the inner fluted layer of the free part 
the * canopy.” 
The possession by seeds of a vascular system was not then a current 
conception, and it is curious to note that though in a subsequent memoir 
(21, p. 518) Williamson described and figured the “ delicate prosench y matous, 
barred or spiral cells” of a longitudinal section, he failed to grasp their 
true nature, or that they were identical with the cells of the “dark spot” 
(20. p. 239) he had seen in transverse section within each of the concavities 
of his “canopy” when first describing the seed, but considered them part 
of an inner layer of the testa. 
We know now that vascular strands run in the integuments of many seeds, 
recent and fossil, so that their occurrence in Lagenostoma presents no 
difficulties ; and, in the light of recent knowledge, it is scarcely possible to 
see in the chambered upper part of the integument other than the equivalent 
[4 
of the fused “tentacles " of Physostoma. Williamson’s “ canopy ” is there- 
fore nothing but the inner, and of the same morphological value as the outer 
