r 63 
We have taken pains to see all the available American ma- 
terial of Physcomitrella patens in the Sullivant herbarium as well 
as in that of the Department of Agriculture and others. In Drum- 
mond’s Southern Mosses, No. 5, collected near St. Louis, the 
specimens are still in perfect condition for examination. Many 
of the capsules have not split, but several of those which have 
show with a low magnification a perfectly circular rim. Under a 
high magnification, after boiling and mounting, the walls are torn 
irregularly, but a distinct median line of separation, where the 
cells are elongated with longer transverse walls, was seen, even on 
the irregularly torn capsules. 
No. 30, Sull. & Lesq. Musci Bor. Am. Ed. 2 (1865), collected 
in Ohio, also are in fine condition. Few of the capsules have 
split, but after mounting in glycerine jelly, three capsules did split 
regularly in half, and although we have seen no differentiation of 
the cells along the suture, yet from its regularity, we suspect it is 
there. It is almost impossible to determine this after the dehis- 
cence of the capsules, as the walls are so thin, the cells so tender, 
that in boiling to get rid of the spores, the walls are injured be- 
yond possibility of examination. The stomata are basal, and 
Seem partly immersed. | 
In Sullivant’s own herbarium we found specimens from Colum- 
bus, Ohio, collected in 1852, several capsules of which had split 
regularly in half as seen by reflected light with a low power, 
Quite recently we have received from Fort Snelling, Minne- 
sota, specimens collected by Grace E. Sheldon on October I, 
1894, which show a differentiation of the cells around the middle 
of the capsule. Fig. 8 of our-plate was drawn from these speci- 
mens. These capsules are very fresh and young, the cells of the 
walls still contain chlorophyll, and are more distinct than any we 
have seen. Starting from the stomatose base, the longest diame- 
ter of the cells is vertical, they measure .013%.027 mm., gradually 
the cells become larger, still keeping their longest diameter verti- 
cal, and twice the length of the shortest, measuring .027.054. 
At the line of dehiscence the cells are transversely elongated, 
three times longer than wide, and measures .o16xby .054 mm. 
Taking the above observations only for what they are worth, 
it seems to me that we may at least give the question the benefit 
