STRUCTURE OF Fi 
the two envelopes are altogether united, the papilliform 
process of the primine is open; from its mouth a canal 
proceeds, which widens as it descends into a large cavity 
in which the cone of the nucleus is included, the apex 
of this is not exserted but terminates a little above the 
widening of the canal. The cone is solid, the base of 
the nucleus is occupied by alarge globular cavity in 
which the sac of the quartine is contained. This sac is 
perfectly shut, and has no communication with the wall 
of the cavity of the nucleus. It is easily detached, 
is of a globular form and is filled with pulp, consisting 
of very lax, slightly angular cellular tissue of a perfectly 
homogeneous nature. | 
8. Ovule further advanced, the papilla of the primine hae 
fallen off, and the canal is obliterated. No other change 
has taken place except in size. 
It does not seem to me to come within the bounds of 
what we know of general structure, to reduce the male or- 
gans of Cycas to the type of an ordinary anther.* 
Either each scale is to be considered as an anther, or each 
polliniferous bag is to be so considered. 
All analogy is against the former, for no instance is 
known of the cells being distinct organs projecting beyond 
the surface. Though pluri-locular anthers are known; 
yet the cells are contained in the general substance of the 
anther, and at one period at least contained within its com- 
mon covering or surface of that organ. 
The ordinary type of an anther is 4-locular, the pollen is a 
Separation or modification of the cellular tissue equally on 
both sides of the mesial line of the leaf: no instance is 
known in which the whole substance of the anther is so mo- 
dified, except perhaps in Sarcophyte, but a part remains solid 
and unchanged as connectivum. Now in Cycas each pollini- 
erous bag is a distinct organ, and the pollen is the entire 
modified cellular internal substance. 
* No analogy between anther cells of Cycas and Phenogs. 
