ON THE SEXUAL ORGANS OF THE CYCADACEZX. 95 
common point of attachment. In their earliest stages they appear as 
slightly elevated protuberances or papillz, green in colour, and covered 
by the epidermis, which they do not rupture. They are, in fact, ex- 
crescences of the parenchyma which are formed at particular points, 
and their internal tissue is consequently wholly cellular. Little by 
little they assume their elongated rounded form, and the entire mass 
of cellular tissue becomes pollen-generating tissue, because in the 
cavity, when matured, nothing but pollen ean be found. Each cell 
[‘ parent cell'] produces four other cells, and each of these forms a 
pollen-cell They are comparable to the regions where, in the loculi 
(^ loges’) of ordinary anthers, the production of pollen takes place; 
and should receive, therefore, the name of loculi. 
The formation of pollen does not take place over the whole organ as 
in angiosperms and most gymnosperms, but only at a considerable 
number of points on either side of the median line. The wall of the 
loculi is very firm. Its colour is brown at a period a little more ad- 
vanced, and its exterior is marked with short linear impressions. It 
opens from the top to the bottom on the inner side or that which 
is turned towards the other loculi of the group, and sometimes the 
slit is prolonged beyond the summit on to the opposite side. Pur- 
kiuje (‘De cellulis antherarum fibrosis”) was not wrong when he 
termed the wall ** mere epidermidalis," since the loculi are nothing more 
than erupted portions of the tissue of the androphyll, covered with the 
same epidermis as the rest. I may remark, however, that two cellular 
layers may be distinguished in this wall which may be also recognized 
in Purkinje’s figures (Plate I. and Plate XVIII., belonging to Zamia 
media and Encephalartos longifolius). The external layer is the epi- 
dermis, the inner one is a parenchymatous layer of peculiar appearance, 
composed of porous cells. 
The cells of the epidermis have a very narrow lumen. This gives 
rise to the superficial stripes mentioned above. ‘The pollen grains ex- 
hibit a great uniformity throughout the whole family: they are more or 
less elliptical, with a deep longitudinal fold which does not entirely disap- 
pear in water. Thus, as Schacht first pointed out (Pringsheim, Jahrb. 
ii. p. 145, plate xvii. fig. 26-28), two secondary cells are also formed 
in the intine among the Cyeads, so that the structure of the pollen is 
comparable in all respects to that of the Conifere. 
If the views which have just been explained on the subject of the 
H2 
