439° BOTANICAL GAZETTE | DECEMBER 
While in this species there is an early quadrant division of the cover- 
cell (see fig. 5), the four resulting cover-cells are larger than is usually 
the case, and there may apparently be a limited number of the outer 
neck-cells which are cut off from these cells. Such a case is shown in 
jig. 46, E, of my Mosses and Ferns. That the cover-cells of the liver- 
wort archegonium may undergo one or two divisions subsequent to the 
original quadrant-divisions, has been long known, but I have been 
unable to convince myself that any apical growth, in the sense in 
which it is understood among the true mosses, can be demonstrated 
in any of the liverworts examined by me. 
Six rows of peripheral neck-cells are regularly found in the arche- 
gonium of the Marchantiacez while the normal number is five in the 
Jungermanniaceez. I have found that in Sphaerocarpus terrestris 
var. Californicus and the allied Geothallus, there are six rows of 
peripheral neck-cells, in which respect, as well as others they are 
intermediate between the Ricciacee and thallose Jungermanniacez. 
Gayet disputes the accuracy of my statement in regard to Sphzrocarpus, 
and it is possible that the European form of the species may show but 
five rows of cells. Neither of the two figures of Sphzrocarpus shown 
by Gayet is a cross-section, nor does he say whether he actually 
examined such sections. In the few cross-sections of the arche- 
gonium neck which I have made, the number of cells was six (see 
fig. 4), although it is possible that this number may not always be 
constant. 
It is to be regretted that M. Gayet has not given a more detailed 
account, as well as additional figures, of the archegonium of the 
thallose Jungermanniacee. He finds that in Pellia, as well as other 
Anacrogyne, there may be six rows of peripheral cells, instead of the 
usual five rows hitherto supposed to be constant in this group, aside 
from Spherocarpus and Geothallus. It is not strange that these 
primitive forms should show this approach in their structure to the 
Ricciace with which they are closely connected by Sphzrocarpus. 
It is to be hoped that we may soon have further information on t is 
interesting point. 
In regard to the statement that in the Musci the neck canal-cells 
are not cut off from the base of the terminal cell, as has been hitherto 
supposed, it cannot be said that Gayet’s figures are very convincing. 
This very difficult point can only be settled by means of very thin 
axial sections of young archegonia. Here, too, a proper staining of 
eS 
