91 
an archaic type of Seed from the Palaeozoic Rocks. 
wide, owing to the projection into it from below of the apical prolongation 
of the megaspore and surrounding tissue of the nucellus (PI. V, Fig. 8, p.c.). 
In most specimens the chink widens in its lower half (PI. VI, Fig. 17), 
but this condition is probably not the natural one. In Williamson’s 
original section, 1 and in a very few other specimens, 2 the central projecting 
column is buttressed below by a horizontal belt or cushion of tissue — which 
forms the actual lining of the chamber on its inner side — in such a manner 
that the width of the chink or lumen remains appreciably uniform through- 
out (PL VI, Fig. 18, cu). 
As a rule both the inner and outer walls of the pollen-chamber are 
preserved as black, structureless crusts, so that reliance must be placed on 
occasional specimens for anatomical details. The outer wall seems to have 
consisted of cubical cells with rather thick walls (PI. V, Fig. 3,7), the lower 
side), which doubtless represent the epidermis ; whilst in the inner wall, 
which was several cell-layers in depth, traces of the tapetum overlaid by 
secretory sacs may occasionally be recognized. The projecting cushion of 
tissue of the Williamson specimen (W. C. No. 1439) consisted of parenchyma, 
and that is all that can be said of it. No doubt in life it must have had 
assigned to it the double function of secreting a collecting drop and of 
providing for the nutrition of the pollen-grains during their development in 
the pollen-chamber. Morphologically, this tissue would be the equivalent 
of the prominent central cone of the Lagenostomas. If the megaspore- 
chamber of a Lagenostoma be conceived as invading the central cone tissue, 
so as to occupy almost the entire space except a residue on the flanks, the 
resulting relations would closely correspond with the Physostoma-condit\or\. 
The orifice of the pollen-chamber is situated on a low papilla at the 
apex. It is about 150 fx across (PI. VI, Figs. 17 and 18, o.p.c). In none of 
the specimens does the papilla project further than here shown, nor do 
horizontal sections cut above the pollen- chamber afford any reason for 
supposing that it is merely the persistent base of a longer beak. If this 
seed stood alone, it would be quite superfluous even to suspect that a 
caducous beak had been present ; but when regard is had for the structure 
of the pollen-chamber in the allied seeds such a possibility must be 
recognized (cf. p. 92). 
5. P ollination. 
To judge from the quantity of pollen found in the pollen-chamber, the 
arrangements for pollination in Physostoma must have been unusually 
efficient. A section across an empty pollen-chamber is the exception. 
The specimen showing the largest number of pollen-grains that has yet 
1 Williamson Collection, No. 1439 : figured in his eighth Memoir, PI. XI, Fig. 77, and PI. XII 
Fig. 79, h. The position occupied by the cushion here seems to correspond more closely with the 
natural condition than in our Fig. 18. 
2 e. g. Scott Collection, No. 1798. 
