EvANs: TAXONOMIC STUDY OF DUMORTIERA 175 
describes further certain ridges or rays on the dorsal surface 
alternating with the lobes; these ridges branch by forking, the 
branches extending along the margins of the lobes. In D. tricho- 
cephala the branches are not connected in any way and thus leave 
sharp sinuses between the lobes; in D. hirsuta the branches are 
connected by thallus-substance, and the sinuses thus appear 
lunulate. These distinctions would be very helpful if they were 
at all constant but, in the writer’s experience, this is not the case. 
All gradations occur between strongly bristly receptacles and those 
with marginal bristles only, while the ridges with their branches, 
although sometimes fairly distinct, are often vague and evanes- 
cent. Neither is there any correspondence between differences 
in the thallus and differences in the female receptacle. A smooth 
or nearly smooth form, for example, may bear receptacles with 
many bristles scattered over the surface or with only a few bristles 
restricted to the margin. 
The characters assigned to the female receptacle of D. velutina 
seem at first sight to be more trustworthy. According to Schiffner 
(12, p. 26) the disc is depressed-conical with broad lobes and very 
shallow sinuses, the upper surface bearing few or no bristles. 
The figures published by Ernst (4, pl. 18, f: 1) and by Campbell 
(2, text-f. 3) represent the upper surface as perfectly smooth, but 
otherwise agree with Schiffner’s description. In material of 
D. velutina from Java, determined by Campbell, the receptacles 
agree closely with his figures and show no bristles. This is not 
the case, unfortunately, with the specimens distributed by Schiff- 
ner in his Iter Indicum, one of which came from Java (No. 27) 
and the other from Sumatra (No. 32). In these the immature 
receptacles are more or less bristly and some of the bristles are 
scattered over the upper surface. The older receptacles tend to 
be less bristly, but their marginal sinuses are often deeper than 
-Schiffner’s description and the published figures indicate. It 
would thus appear that the characters assigned to the female 
receptacle by Schiffner were subject to variation on therefore 
untrustworthy. 
Although the antheridia and archegonia are usually borne on 
separate receptacles in Dumortiera, bisexual receptacles have long 
been known. They were first described by Taylor (26, p. 391) 
