286 BOTANICAL GAZETTE [APRIL 
the moment of the development of an antheridium, when the 
wall is delimited from the potential spermatogenous cells, the 
latter are few in number and their walls become so much thicker 
than those which appear in their subsequent divisions, that the 
mature organ in section displays conspicuous plots, which fre- 
quently mature at different periods. 
Not rarely one or more of these 
cells fails to produce the same 
number of generations as the 
others, and so certain large spheri- 
cal cells (for they separate from 
one another) may sometimes be 
found intermingled with mature 
sperms (fig. 70). I am _ inclined 
to regard these as eggs. Campbell 
reports budding of “secondary” 
antheridia from the stalk of an 
older Anthoceros antheridium, and 
the same thing has been seen in 
ae oe ree seiaa this laboratory. This variation 1S 
ae ese tee sithahdin ae. analo gous to the formation of 
veloped from neck cells (2), an antheridia from the neck cells of 
apogsmons embryo(ed,/sud anormal: the fern archegonia. 
two-celled embryo in the venter (e). The antheridia of the apogamous 
fern prothalli described above were also sz generis. Those 
developed in young prothalli were typical. But those formed 
among the archegonia were frequently double, two antheridia on 
a single stalk (fig. rr). Upon investigation, this proved to be : 
case of budding of a secondary antheridium from another wel 
advanced. Both antheridia produced perfect sperms that were 
voided. A frequent variant on this was a rhizoid-like branch 
from the stalk of the antheridium. A most interesting case of 
apogamy was observed twice. In these cases the embryo was 
formed from the stalk cell of an antheridium! This has not been 
observed, I think, by Lang (10) or any other investigator who : 
has worked on apogamy. : 
In this connection and a propos of this unstable and plastic con- 
