422 



together, one sporangium springing from every cell, but sometimes 

 vegetative cells intervene. The sporangia are divided not only by 

 transverse walls, but also by more or Jess oblique longitudinal 

 walls. As the sporangia gradually arise along the filament a pe- 

 culiar development takes plaee, viz. the cells whence the sporangia 

 spring divide and turn into sporangia, transforming the whole of 

 the sporangia-bearing portion of the erect filament into what may 

 be termed a single, large pluriloeular sporangium. The latter generally 

 ripens and is emptied successively from the apex downwards to 

 the base, though not always, for, as shown in lig. 78, c specimens 

 occur in which some of the sporangia are emptied here and there 

 along the filaments, and most commonly the cells whence these 

 sporangia spring are emptied simultaneously. The pluriloeular 

 sporangia are about 11 \i broad and 40 \a long. I have only found 

 a small quantity of unilocular sporangia (fig. 78, d); the latter 

 vary somewhat in shape being obovate or nearly so and they 

 sometimes form a sporangium in connection with the cell whence 

 they spring, and sometimes are separated from this cell by a wall. 

 Besides the sporangia, the erect filaments bear now and then true 

 Phæosporé-hairs which grow endogenously and have a distinct 

 sheath as indicated by Sauvageau, 1. c. p. 47. These hairs are 

 usually lateral and I have only rarely come across terminal ones 

 such as are shown in fig. 78, a. The hairs are about 6 — 7 /* thick; 

 as I said before they do not occur on all the erect filaments. 



Besides long branches, quite short sporangia-bearing branchlets 

 with only one vegetative cell are occasionally met with as, e. g. 

 shown in fig. 78, /'. 



The chromatophores consist of a parietal, irregularly branched 

 or perforated plate (lig. 78, e). 



This plant was found on the conceptacles of Himanthalia lorea, 

 where it occurred as a short, dense mat, often associated with 

 Myrionema globosum. 



A species which appears to me to be most closely allied to M. 

 speciosum is M. globosum though the latter dillers considerably from 

 the former, more particularly, e. g. in the different ramiiication of its 

 erect filaments as also in the form and position of its sporangia. 



Myrionema sj)eriosum somewhat resembles the Ectocarpus tomen- 

 tosoides var. norvegica Gran 1 , which is fully described and lignred 



1 Gran, H. H.: Ed norsk form at' Ectocarpus tomentosoides Farlow y\ Chri- 

 stiania Vidensk. Sc-lsk. Forhandlinger tor 1893. No. 17). 



