EvANs: TAXONOMIC STUDY OF DUMORTIERA 171 
In its histological features the thallus differs strikingly from 
most of the other genera of the Marchantiaceae, since it lacks, 
either partially or completely, the usual dorsal epidermis and 
system of air-chambers. , The simplification in structure which is 
thus exhibited is regarded as a derived condition, associated in 
some way with the usual moist and shaded habitat of the plants. 
By the earlier writers air chambers were supposed to be invariably 
absent. Taylor, indeed, in his account of Hygropyla irrigua 
(26, p. 391), described a system of branched and anastomosing 
lines (rami) on the upper surface of the thallus but did not intimate 
that there was any connection between these lines and chambers. 
Many years later Leitgeb (7, p. 308), working mainly on Taylor’s 
species, confirmed his observations and showed that the lines 
were ridges, representing the boundary-walls of vestigial air 
chambers. In the vicinity of the apex he was able to demonstrate 
a short-lived epidermis with distorted pores and he noted that 
epidermal fragments or isolated cells sometimes persisted for a 
while on the boundary walls of older chambers. Close to the 
receptacles he observed the frequent occurrence of elongated 
papilliform cells in the spaces enclosed by the boundaries and 
pointed out the homology between these and the green cell-chains 
_ filling the air chambers of Marchantia. 
In the forms which Leitgeb studied these papilliform cells 
seem to have been restricted to the vicinity of the receptacles. 
At any rate he made no mention of them in other parts of the 
thallus, and the figures which he afterwards published (8, pi. 8, 
f. 8-11) show a smooth superficial layer of cells. A few years 
later Spruce (16, p. 566) gave a detailed account of “D. hirsuta,” 
based primarily on South American material and apparently 
found a very different state of affairs. According to his descrip- 
tion the superficial cells are “‘ papilloso-prominulae” and give the 
living plants a velvety appearance. If there were exceptions to 
this condition in any of his plants he did not allude to them. 
Goebel (5, p. 224, f. 63) described the surface of “D. hirsuta” in 
much the same way and figured the papilliform cells very clearly. 
Unfortunately he did not state the source of his material. He 
contrasted the species with an unnamed form from Ceylon, in 
which he found vestigial chambers but no surface papillae, this 
