82 BANGIACEAE 
suborbicular or suborbicular-pulvinate and pseudoparenchyma- 
tous, the discs attaining a diameter of 180 u, mostly monostro- 
matic, but often irregularly 2-4-stratose in the commonly elevated 
and sometimes bullate central portion; ramification lateral or sub- 
dichotomous, occurring mostly in subterminal cells; cells in surface 
view ovoid, suborbicular, oblong, pyriform, deltoid, rhomboid, 
cuneate, or lunate, 4-14 и in maximum diameter, those of the 
mature disc, except the marginal, commonly elongate vertically 
and appearing ovoid, rhomboid, or subclavate in a vertical section; 
spores 4-5 u in diameter. [PLATE 30, FIGURES 1-7.) 
Growing in the outer rather firmly gelatinous walls of 
Leptocladia peruviana, “dredged in five fathoms,” in the Bay of 
Sechura, Coker 157 p.p. It is commonly associated with Нуейа 
infestans and Chlorogloea endophytica and is sometimes more ог 
less intermingled with them. Well-developed thalli bear some 
superficial resemblance to the young nemathecia of the host 
plant, but may be distinguished by their darker color, repent 
Margins, etc. 
Erythrocladia irregularis Rosenv. and possibly also E. subintegra 
Rosenv. (loc. cit. 72—75. f. 11-14), authentic specimens of which, 
on Polysiphonia urceolata, we owe to the kindness of Dr. Rosen- 
vinge, appear to us to be ordinarily immersed in the walls of their 
host and to be endophytic rather than epiphytic. However, it 
would, we think, be wrong to attempt to identify E. endophloea with 
either of them. The younger stages bear some resemblance to Е. 
irregularis, but the older are more like E. subintegra, differing, how- 
ever, in the relatively less elongate form of the cells in the surface 
view, in the absence of dichotomy in the marginal cells, in the 
finally convex and often irregularly 2-4-stratose character of the 
older portions of the thallus, etc. The cell walls of the Erythrocladia 
are so blended with those of its host that they are scarcely visible 
as arule and our measurements and most of our figures are drawn 
from the protoplasts alone. 
Erythrocladia endophloea is primarily and essentially an endo- 
phyte, but is often, especially in its thicker central parts, so close 
to the surface of its host that it may present superficially much the od 
habit of an epiphyte. : 
Rosenvinge, in establishing the genus Erythrocladia (loc. cit. 
71), discusses its possible relationship to the endophytic Colaco- - 
