145 



A B 

 Fig. fi8. 



Ncnialioii niultifidnm. Young plants from thi' mole 

 at Gilleleje, Novrmlier lOll'i. r.,:!:i:1. 



composed of long, narrow cells with less protoplasmic contents are developed in 

 the continuity of the primary ones. They form later at their upper end fasciculated 

 branches reminding one of the peripheral assimilative filaments in the older frond. 

 The author named supposes that such thin 

 erect filaments may meet and twist together, 

 thus giving rise to an erect frond. I have not 

 observed the species in this stage of develop- 

 ment. In February at Gilleleje I found crusts 

 apparently formed by densely united creeping 

 filaments like those described by Miss Chester, 

 but almost no erect filaments were observed. 

 Young plants with the normal structure are 

 shown in fig. 68; the assimilative filaments were 

 only less numerous in the lower part of the 

 plants than later. 



The assimilative filaments terminate in hya- 

 line hairs of various length, generally rather 

 short. The shortest are almost entirely filled 

 with protoplasm, while in the longer the pro- 

 toplasm with the nucleus is concentrated in 



the upper end, the rest of the cell containing only a thin parietal layer. When a 

 hair dies a new one is often formed at the same place from the subjacent cell; 



the lower part of the old membrane remains however 

 surrounding the base of the new hair as a sheath. A 

 new hair may also be formed beside and below the terr 

 minal one, and this may also be renewed (fig. 69). 



The antheridial branches form clusters at the ends 

 of the assimilative filaments as described by Bornet 

 and Thuret (1867) and Wolfe. Each cell in the an- 

 theridial branch gives rise to four antheridia or fewer 

 (fig. 70). The spermatogenesis has been worked out by 

 Wolfe. 



The carpogenic filaments are terminal on the assi- 

 milative filaments and usually 3-celled , (according to 

 Wolfe 2- to 5-celled). Concerning the fertilization and 

 the development of the cystocarp reference may be 

 made to the quoted papers of Bornet and Thuret, 



Nenmlion multiridum. End-cells of JanCZEWSKI, ScHMITZ, WiLLE and WoLFE. The fertilised 

 assimilutive filaments with hairs; in _ 



D the nucleus is visible. (i80:i. carpogouium divides by a transverse wall into a basal 



or placental cell, remaining undivided, and an upper 

 cell dividing by vertical walls into a number of cells giving rise to branched spo- 

 rogenous filaments, the end-cells of which produce carpospores. 



A 



B 



Fig. fi9. 



1). K. I). ViilensU. Selsli. Ski 



. MiukUe, n:itin'\i(leiisli (ij; HKitlu'iii. AlVi. I. 



19 



