me 
sporangia. Ch. polyblasta and Ch. humilis differ by the want of endophytic filaments 
and by the presence of a distinct pyrenoid. 
Locality. Su: Only found at Helsingor, growing on Porphyra umbilicalis on the outer side of 
the southern mole in September. 
17. Chantransia Dumontiæ sp. nov. 
Thallus cæspitulosus ad 0,5 mm. altus, e filis 1° horizontalibus epiphyticis 2° 
erectis ramosis sporangiferis et 3° endophyticis constructus. Spora germinans disse- 
pimento verticali diametrali in duas cellulas æquales divisa est quarum utraque 
filum horizontale procreat. E filis primariis lateraliter fila repentia et superne 
fila erecta egrediuntur. Fila endophytica intercellularia ex parte saltem e filis erectis 
egrediuntur. Fila erecta a basi ramosa, ramis numerosis sparsis plus minus ramosis; 
cellule diametro fere triplo longiores, superne 6,5—9 » crass, inferne tenuiores, 
chromatophorum stellare, ut videtur sine pyrenoide, in parte superiori cellule situm 
continentes. Pili hyalini terminales vel pseudolaterales adsunt. Sporangia tetraspora 
oblonga—ovata, latitudine fere duplo longiora, 15—19 » longa, 8—11 y lata, in filis 
lateralia et terminalia, plerumque sessilia, sparsa vel opposita, nonnunquam in ra- 
mulis unicellularibus singula vel bina. 
The species forms numerous small, dark-purple tufts or cushions on fronds 
of Dumontia filiformis, They consist of numerous erect branched filaments given 
off from the creeping filaments, partly also from the endophytic threads. The ger- 
mination takes place as in Ch. cytophaga and others of the above described species 
(fig.52 A, B). The epiphytic creeping filaments are often somewhat irregular, thick 
and short-celled, and, as shown in fig. E, they are not always densely attached to 
the surface of the host. I am not able to say if the first endophytic filaments are 
given off from the underside of the creeping filaments or not. At all events endo- 
phytic filaments are also given off from the base of the erect filaments (fig. D). The 
endophytic filaments are much branched growing intercellulary in the host, and 
free erect filaments may again be given off from them through the surface of the 
frond. I believe that this may take place also at a greater distance from the point 
of departure of the endophytic filaments, these thus serving to propagate the Chan- 
transia in the host. The free filament shown in fig. D has probably emerged from 
the endophytic one. It appears that relatively few endophytic filaments are given 
off in the same cushion. 
The erect filaments arise in great number from the creeping filaments, from 
their peripheral as well as their central parts, and the plant forms therefore tufts 
or cushions of '/2—1 mm. in diameter. These filaments are fairly strongly branched, 
as a rule from the base, and often a branch is given off from each cell in a great 
part of the primary filaments, and the branches may also be branched. The cells 
are usually broader at the upper end than below, depending on the abundant ram- 
ification. In the central part of the stellate chromatophore I was not able to 
detect any pyrenoid staining stronger with hemalum than the remaining substance 
