an archaic type of Seed from the Palaeozoic Rocks. 99 
and that showing radially arranged black rods, are degradation products 
of a layer comparable, for descriptive purposes, to a zone of corky periderm 
that has undergone collapse. 
Transverse sections near the base of the seed, where the embryo-sac is 
tapering, show the layer as a broader zone (as in M. H. 369, PI. V, Fig. 1) 
than is the case higher up ; whilst in appropriate sections still nearer the 
chalaza, the tapetum appears to fill the whole cavity of the seed (PI. VI, 
Fig. 14, Ip). 
It is very rare indeed to find this layer with structure preserved in the 
region of the pollen-chamber ; the specimen figured in PI. VII, Fig. 25, is an 
exception to the general rule. It is a somewhat oblique transverse section 
across the apical cone of the nucellus and pollen-chamber ; and although 
the tapetum (which alone is figured) shows a long tangential rift, the layer 
shows substantial agreement in structure with that found in other parts of 
the seed. 
The compression and poor preservation of the tapetum afford some 
ground for the supposition that its full functional activity corresponded to 
an earlier stage in the development of the seed than the one to which the 
majority of specimens oi Physostoma belong. This supposition finds support 
in certain small-sized seeds or ovules of Physostoma , of which two specimens 
have come under observation. In these specimens the dimensions are about 
three-quarters the normal size, and the general preservation has something 
in common with the interesting little seed of Lagenostoma Lomaxii} 
A transverse section of one of these small specimens is represented in 
PL VII, Fig. 29, and the principal feature shown is the tapetum (//), which, 
partly separated from the outer wall and contracted in stellate manner, is 
an extraordinarily conspicuous object. It is reasonable to suppose that 
this specimen, which shows other peculiarities (cf. p. 100), represents 
a considerably younger stage in development than the generality of 
specimens. 
It seems probable that a tapetal layer may have been a common 
feature in the seeds of the Lagenostoma-gron p. For in addition to Physo- 
stoma , occasional specimens of another seed, the rare Conostoma oblongum 
of Williamson, [now undergoing reinvestigation, show a well- developed 
tapetal zone resembling that of Physostoma , whilst a reexamination of 
the young seed of Lagenostoma Lomaxii raises the question whether the 
black-walled £ pipes ’ (there regarded as the shrivelled nucellus and mega- 
spore wall) may not represent a tapetal layer in a state of collapse. 2 
The available information as to the occurrence of tapetal layers in the 
Lagenostoma-gxox) p of seeds may be shortly summarized as follows 
Physostoma and Conostoma oblongtim had a many-layered tapetal zone, 
1 Oliver and Scott, On Lagenostoma Lomaxii , Phil. Trans., B., vol. 197, p. 212. 
2 Oliver and Scott, loc. cit., PI. X, Fig. 34, i.s. 
H 2 
