an archaic type of Seed from the Palaeozoic Rocks. 89 
jecture : analogy suggests that it might have been of a mucilaginous 
character, but at present the investigation of fossilized secretions has been 
too slight for any confident expression of opinion. It seems probable, 
however, that we have to deal in this case with a genuine secretion of the 
protoplasm, rather than with a degradation product of the cell-membrane. 
3. The Ground-tissue of the Integument. 
The Ground-tissue of the Integument consists of closely-fitting, thin- 
walled parenchyma cells of elongated, prismatic form, the longer dimension 
being parallel to the axis of the seed. The approximate dimensions of 
these cells are 200 /x x 25 /x. 
In the case of the tentacles, this tissue forms the whole of the filling 
substance. It was traversed in the median plane, just below the small- 
celled inner epidermis, by a vascular strand, but in the great majority of 
specimens the strand has broken down, and is represented by an intercellular 
space (PI. V, Fig. 7, IX 
In the body of the seed, where integument and nucellus are confluent, 
there are from five to six layers of these cells, forming an undulating belt 
(PI. V, Figs. 2 and 5). The outer layer (upon which the hairs are inserted) 
and the layer beneath it consist of somewhat larger cells than the deeper- 
lying regions ; they often show a slight radial extension. 
The cell-contents. The whole of these cells — alike in the tentacles and 
on the seed-body — show an interior contracted tube, light brown in colour, 
which on cursory inspection might be taken for a slightly plasmolysed and 
petrified protoplasmic body (Pl.V, Fig. 5, and PI. VII, Fig. 26,/./. and Fig. 28). 
These inner tubes are well seen in most specimens, though, as we shall 
see, the appearance is not invariable. However striking the resemblance to 
protoplasmic bodies, it is impossible to accept this interpretation without 
reserve in the absence of much more critical histological investigations 
than have yet been carried out upon fossilized cells and tissues. In the 
present instance this reserve is justified because a certain number of 
specimens have come under observation which show no inner tubes ; 
moreover, such specimens have the further point in common that the walls 
of the cells are unmistakably thicker than in the case of the more usual 
type of preservation (PI. V, Fig. 9). Hence the possibility is by t?o means 
excluded that the inner tubes or vesicles may not be protoplasmic vesicles 
at all, but rather the lining layers of the cell-wall which have become 
separated from the main reticulum of cell-walls. In view of these facts it 
would seem probable that the vesicles in question are really derived from the 
membrane and not from the protoplasm. 
Throughout the body of the seed the interior border of the parenchyma- 
zone of the integument abuts upon the very delicate zone in which the 
secretory sacs are situated — the region usually least well preserved of any 
