Evans: NOTES ON GENUS HERBERTA 203 
antheridia occur not only in the axils of the bracts but also in the 
axils of the bracteoles.* In order to determine whether this phe- 
nomenon was widespread among the Ptilidioideae, to which 
group Herberta is usually assigned, he examined species of several 
other genera. The closely related Mastigophora, however, was the 
only one in which he was able to observe antheridial bracteoles. 
In the case of Herberta the bracteoles differ from the bracts in 
about the same way that the ordinary underleaves differ from the 
leaves. Both bracts (TEXT FIG. 20; PLATE 8, FIGs. 8, 9) and 
bracteoles are more or less imbricated and have a broad inflated 
pocket at the base enclosing a cluster of two or more antheridia. 
The pocket does not involve the entire width of the basal region 
but leaves a narrow flattish or revolute strip on each side. In the 
inflated portion the cells are broader and much paler than ordinary 
cells, and the thickenings of the walls are more irregular in outline 
and less strongly developed. The divisions, sinus, and margins 
are much the same as on ordinary leaves. 
The perichaetial bracts (PLATE 8, FIGs. 10, 11) and bracteoles 
are essentially alike. They occur in three or four closely crowded 
series and increase somewhat if size toward the perianth, only 
the apical portion of which projects beyond them. Those of the 
innermost series, when viewed from the outer surface, show two 
rounded ridges in the basal portion, separated by a narrow median 
groove. A marginal band on each side is flat or slightly concave. 
The divisions are much the same as on ordinary leaves but the 
margins are much more toothed, the teeth being irregular and 
extending higher up, sometimes almost to the apices of the divisions. 
Occasionally minute and irregular paraphyllia (PLATE 8, FIG. 12) 
with marginal papillae are found among the bracts. According 
to Spruce the bracts closely embrace the perianth but ar é free 
from it throughout their entire length; accor ding to Stephani they 
are coalescent with the perianth up to a considerable height. A 
longitudinal section through a young sporophyte and surround- 
ing parts (TEXT FIG. 8) shows that Spruce was essentially correct. 
Any coalescence which may be present is so slight as to be prac- 
tically negligible. Between the innermost bracts and bracteole 
and the leaves below the involucre there ie a.gredual ransom. 
* Untersuchungen iiber. Amphigastrial-Antheridien und tiber den Bau der 
Andr&cien der Ptilidioideen. Hedwigia 50: 146-162. f 1-39 1910. 
