8 PROF. MARGARET J. BENSON ON 



.ire fairly frequently preserved in the canopy, but are rarely seen in the periphery 

 of the embryo sac (slide 292, l). 



The eight regions respectively supplied by the eight vascular bundles correspond 

 with those of the crests. Lower down on the same radii on which the vascular bundles 

 lie, but at the very base of the sinus, between the canopy and the plinth, the epidermis 

 shows some curious tongue-shaped cells which have pitted thick walls and generally 

 dark brown contents (PI. I. figs. 1 and G, h,). They differ in form and appearance from 

 the papillate mucilage cells, which are thin-walled and of a pale yellow colour. 



The eight vascular strands approximate closely to these cells, often obviously 

 bending towards them and then retreating again above. There are probably eight 

 groups rather than a continuous ring of these cells, since occasionally the radial plane 

 of section does not pass through any. Such sections escape also the bundles. 



The form of these cells and their close approximation to the vascular strands 

 suggest an excretory function, but their different mode of preservation suggests that 

 the excretum was non-mucilaginous and hence possibly merely water. 



Let us turn now to the hypodermal structure of the exposed surface of the inner 

 integument. One sees in a well-preserved specimen a uniform single layer of fibres 

 approximately square in transverse section (text-fig. 3, b) and several times longer 

 than broad. 



The layer appears to be less regular in some ovules — some growth in length and 

 displacement probably taking place among the fibres. 



This sheath forms the only special mechanical tissue the ovule possesses, and must 

 have given toughness to the coat although it was unable to afford it any great degree 

 of rigidity. In several cases the ovule has been indented, or even folded like a 

 collapsed bladder, before fossilisation. The sheath is not continued beneath the 

 epidermis of the pedicel, but is completed by deeper-lying sclerotic cells forming a 

 transverse plate, which is only perforated to allow of the entry of the single vascular 

 strand into the ovule (text-fig. 1, s.). This plate forms the base of the great majority 

 of the ovules examined (PI. II. figs. 8 and 11). 



3. The Abscission Layer. 



Immediately outside the sheath, as it leaves the epidermis, the tissue in the pedicel 

 shows signs of breaking down. Eemains of this layer have been demonstrated in a 

 transverse section of an ovule which had probably been naturally severed from the 

 parent plant (PI. II. fig. 11, ab.). The tissue is seen also in the radial section of the 

 ovule in slide 387 (PI. II. fig. 9, ab.) which shows the layer of separation still in situ. 

 In this case the degeneration of the tissue has not yet encroached on the epidermis, 

 which is still regular and continuous. 



If this be the true interpretation of this region, it indicates that the ovule normally 

 fell out of the cupule, leaving that upon the parent plant. Thus it would seem probable 

 that such of the ovules in Dr Scott's slide 387 as are still surrounded by a cupule had 



