1904] DAVIS—SEXUAL ORGANS IN PLANTS 251 
gonium” (DAvIs ’03¢) and the reader must be referred to that for 
a detailed treatment of the subject. 
Miss Lyon (’o4) has discussed the interesting problem of the rela- 
tion of the sunken gametangia, characteristic of certain pteridophytes 
(especially Lycopodium) and such liverworts as Anthoceros and 
Aneura, to the stalked archegonium and to my theory associating 
these structures with plurilocular sporangia. She is inclined to derive 
the sunken structure from an indeterminate region of gametogenous 
cells which later might develop into an emergence with the general 
characters of a gametangium (plurilocular sporangium). This view 
carries the origin of the archegonium still further back, and allows the 
organ to develop through an emergent gametangium into the stalked 
Structure, or to remain partially or wholly imbedded in the tissues of 
the gametophyte. In the first group the archegonia would become 
definite gametangia, comparable to plurilocular sporangia; while in 
the second they would remain as less defined or indeterminate regions 
of §ametogenous tissue. The chief difficulties in this view, in the 
author’s opinion, lie in the remarkable unity of structure displayed 
by the archegonium, in the presence of a single terminal opening, 
and the situation of the egg at the bottom of an axial row of gametog- 
fnous cells, which conditions imply an origin from some definite 
type of gametang'um whose fertile tissue was limited to a central 
on. The rarity among the known thallophytes of indeterminate 
regions of §ametogenous tissues present further important difficulties 
in Miss Lyon’s theory. 
hee 98 ('o4, Pp. 289), however, is inclined to pass lightly over 
readih § ficulties, believing that transitory conditions may “be 
br eo a. among the algae.” She discusses several types and 
consid “5 diagram of Ulva indicating a gametogenous tissue of 
ka thickness at the period of reproduction. This is a very 
a am, for not only are there no walls formed between 
latter are 84 segments of protoplasm in the mother-cells, but the 
its neighbo tmarkably well-defined sporocysts, each independent of 
4 thallus several € membrane of Ulva is very far from constituting 
In Phyllitis * layers of cells thick, or even a differentiated tissue. 
Within the ere appears a successive segmentation of the protoplasm 
mother-cells with the formation of walls by which the 
