12:5 



As new erecl filaments are con- 



times penetrate into one cell. The filaments often branch within the host ceil, and 

 some of the branches may again become free, growing outwards through the wall 

 of the host, and the same occurs with endophytic filaments without branching (in 

 fig. 51 the free endings of the hauslorial filaments are not shaded). As far as I 

 have observed, these filaments do not penetrate from one cell into another, and 

 therefore do not serve as propagating organs. The protoplasm of the host cell is 

 more or less shrunk and evidently yields nourishment to the Chantransia which is 

 thus a veritable parasite. 



A great number of erect filaments are given otl" from the creeping filaments, 

 from the peripheral part as well as the central, 

 stantly produced, a fully developed plant shows 

 numerous erect filaments of different sizes, giving 

 the plant a cushion-shaped appearance. Most of 

 these filaments attain only an inconsiderable 

 length, the greatest are about 200 n long; they 

 are either unbranched or bear one or a few 

 branches which are much shorter than the main 

 filament. Hyaline hairs frequently occur at the 

 ends of the filaments, becoming lateral by the 

 continued growth of these. In the lower part of 

 the filaments the cells are more or less swollen 

 at their upper ends or in the middle. The struc- 

 ture of the cell is the same as in Ch. Dumontiiv 

 (see p. 124) , the chromatophore being stellate 

 without pyrenoid, while a body staining inten- 

 sely by hsemalum and undoubtedly a nucleus is 

 to be seen under the chromatophore. 



The sporangia are always sessile on the sides 

 of the erect filaments, in their whole length. From 

 the first each cell bears one sporangium, but very 



soon one or two others appear, and each cell ments are shaded. Iheir tree emerging ends are 



white; p, protoplasm ot the liost cell. 550:1. 



bears thus usually two or three sporangia, the 



two being as a rule opposite. The latest formed sporangium is sometimes seated 

 at a lower level than the other, near the middle of the cell. Terminal sporangia 

 were not observed. Nearly all the sporangia were monosporous, very few tetra- 

 sporous; the latter were somewhat larger than the other. Possibly some of the 

 undivided sporangia were unripe tetrasporangia ; I imagine, however, that most of 

 them were really monosporangia. 



The structure of the cell and the mode of growth bring this species near to 

 Ch. Dumontice; it diflers from it in particular by the intracellular haustoria, by 

 shorter, less branched erect filaments, by shorter cells and by the want of terminal 



1(5- 



Fig. 51. 



Chantransia cytophaga. A , plant growing on 

 the margin of the frond of Porphyra um})ili- 

 ealis, to the left two haustoria penetrating into 

 the same cell. B. plant growing on the flat 

 side of the frond seen from above, showing 

 three haustorial filaments. The endophytic fila- 



