SPH^EROSTOMA OVALE FROM PETTYCUR, FIFESHIRE. 11 



this body must be deferred until another occasion. So also must the description of 

 some curious bodies which often accompany the ungerminated grains in the pollen 

 chamber. 



No satisfactory specimen has yet come to hand in which the pollen chamber is 

 open. One of Williamson's slides, preserved in his collection at the British Museum 

 at South Kensington, and figured in his Eighth Memoir,* suggests this condition, but 

 the appearance may be due to injury before fossilisation, as only one side is open. 



7. The Processes assumed to occur during the Opening and Closing of the 



Pollen Chamber. 



That the important function of securing and nursing the pollen was correlated with 

 a series of progressive changes cannot be doubted. When the excavation of the pollen 

 chamber was complete, the epidermis of the nucellar cap underwent circumscissile 

 dehiscence along a line just within the periphery of the persistent column of tissue in 

 the centre. The roof of the chamber was thus made to overlap the column when it 

 returned, after dehiscence, to its original position (cf. figs. L = state before dehiscence 

 and 6 = state after closure). In fact, the relation of the margin of the column to the 

 roof of the pollen chamber resembles that of the rebate of a box to its lid. Hence the 

 downward curvature of the roof is prevented from being so excessive as to obliterate 

 the pollen chamber. 



The growth in length of the roof-cells and the deposition of thickening on their 

 vertical longitudinal walls would necessarily have set up a centrifugal strain tending 

 to bring about the circumscissile dehiscence. This was further aided by the mucilagin- 

 ous degeneration of the subjacent tissue. 



After dehiscence the roof would straighten elastically and thus an annular stomium 

 would be formed. Passage to the lower part of the sinus would be at least partially 

 blocked because the liberated margin of the roof of the pollen chamber would nearly 

 abut on the basiscopic papillae around the upper part of the sinus. Pollen would thus 

 be prevented from straying into the sinus. 



The mucilage glands would meanwhile have come into play, thus completing the 

 preparation for pollen. If grains were now to enter the micropyle they would be 

 caught in the freshly exuded watery mucilage and be drawn with it into the pollen 

 chamber. The entry must have been aided by the subsequent downward curvature of 

 the roof of the chamber. 



This return to the former position may have been due to an increase in their 

 turgescence on the part of the roof-cells. This is suggested by the domed form of the 

 peripheral cell-walls of the epidermis covering the older pollen chambers. These 

 peripheral walls offer a marked contrast not only to the thick vertical cell- walls, but 

 even in this respect to the basal walls, which, although thin and transparent, 

 are flat. 



* Williamson, he. cit., pi. xii. fig. 83. 



