RHODOPHYCE^E, OR FLORIDE^ 219 



the fruit, at the place where the procarp originally 

 was, the cells of the wall part asunder, and there 

 arises a cylindrical canal through the whole thick- 

 ness of the wall. The cells at the sides of this fruit- 

 pore develop dense masses of hair-like cells at right 

 angles to the pore, and nearly filling the whole of the 

 canal. The pore is often formed very early, before 

 the formation of the spore-glomerules, and before 

 the spore-clusters have developed in size and have 

 become fully matured spores. Such cell-clusters 

 gradually develop into spores, the single cells become 

 larger and more intensely coloured ; and when they 

 are mature, the connections between the individual 

 spores are gradually loosened, the whole glomerule 

 becomes disintegrated, and finally, the fully de- 

 veloped round spores are entirely free." Their 

 escape through the pore is gradual, and in fact it 

 often happens that some do not escape at all, and 

 actually begin to germinate in situ. 



In contrast with the well-developed, vigorous 

 fronds of such genera as Chondrus, Gigartina, Phyllo- 

 phora, Stenogramme, and Gymnogongrus, Callophyllis 

 and Callymenia, the genus Adinococcus is of singular 

 interest. A. roseus grows parasitically on Phyllo- 

 phora Brodicei forming externally small cushions 

 which have been mistaken by some observers for the 

 tetraspore fruits of the host plant. Its rhizoid 

 tissue penetrates the inter-cellular spaces of the 

 host, but its parasitism is obviously not so complete 

 as that of Choreocolax (Harveyella ?) albus. The 

 antheridia and cystocarps are unknown, and the 

 tetraspores only have so far been recorded. A 



