120 BOTANICAL GAZETTE [FEBRUARY 
gonium from some form of primitive compound gametangium. 
In such a case, by sterilization, a protective wall was secured and 
the number of gametes greatly reduced, perhaps to a single row 
as the form _took on terrestrial habits. Barnes (16, p. 289) 
regards each member of such a central row as the homologue of 
a simple ovary. These undergo sterilization from the apex 
downward, a natural consequence of conditions of nutrition when 
the eggs are retainedin the ovary. Sucha process of progressive 
sterilization would not occur in the antheridium since the forma- 
tion of a large number of simple sperms is the function, and with 
this effected the work is done. The egg, on the other hand, 
must provide for the development of the young sporophyte and 
requires an abundant store of food for this purpose. 
Recently Davis (23) has proposed a theory for the origin 0 
the archegonium that has very much in its favor. Because there 
seems to be no organ among the heterogamous algae from which 
the archegonium could have arisen, Davis goes to the lower 
Phaeophyceae for a type to illustrate his views: He takes the 
asexual plurilocular sporangium of certain Ectocarpaceae as a 
starting-point. From the asexual zoospores of such an organ 
sexual gametes arise through differentiation. With sexuality 
established, the gametangia differentiate until organs are reached 
from which archegonia and antheridia can be produced by 
processes already indicated. The aperture at the apex of each 
organ is significant, and the sterilization of all the gamete mother- 
cells but one a great advantage to the egg. Davis goes a step 
further and records that under certain conditions, Scezomerls 
Leibleintt and Draparnaldia have developed zoospores in organs 
that strikingly resemble plurilocular sporangia; and he is dis- 
posed to look upon these structures as forerunners of well 
differentiated plurilocular sporangia in the Chlorophyceae. 
These views of Barnes and Davis seem to be in the right 
direction; though it is understood that in any case it is 
necessary to assume the existence of heterogamous Chloro- 
phyceae of which we have no knowledge at present. Since these 
views have been advanced considerable evidence in their suppott 
has been accumulating, and several cases in Mnium, described in 
