THE BRANCHING-SYSTEMS OE THE ALCYOKACEA. 531 



nombreuses branches que les polypes sont agglomeres en panicules ovoides. Chaqiie 

 individu a un pedicule distinct, court, d'un In-un rougeatre, ou couleur de bistre, de 

 meme que les luiit tentacules, obtus et arrondis, qui le termiaent." 



Anatomi/ and Ilislologij. 



The description given, in the introductory part of this paper, of the mode of aggrega- 

 tion of the Nephthyidie applies in every jjarticular to Lemnalia. I made several series ot 

 sections through branches and a few sections through the stem of L. nitida, and I was 

 able to make a series of sections through detached branches of L. rhahdota and L. peri- 

 styla. I was not al)le to examine the other species microscopically. 



Fig. 19 (PI. 42), which is a section of a tolerably large branch of L. nitida, shows that its 

 branches (and the same is the case with the stem) are composed of fused elongate anthosteles. 

 It can be seen that the middle of tlie branch is occupied by some ten or twelve large 

 anthosteles, whose cavities measure some -75 mm. in diameter, wliile the cortical portion is 

 occupied by smaller anthosteles whose cavities vary from "!') to "1 mm. in diameter. 

 Nearly all the cavities display ridges or folds which are the prolongations of the 

 mesenteries. The tissues in the centre of the branches and stem were badly preserved, 

 so that I could not make out the details of the mesenterial tilaments in all cases ; but in 

 some sections, both of the stem and branches, I could distinguish tne characteristic 

 grooved and ciliated asulcar or "dorsal" filaments, and the remaining mesenteries seem 

 to bear tilaments down to the very base of the colony. The larger central anthosteles 

 are the chief zooids, the more peripheral anthosteles the secondary and tertiary zooids 

 of the syndete. The former may be called the chief members, the latter the lateral 

 members of the branch. It can be seen that the chief members do not communicate 

 one with another, but that, as at m. m., they communicate by short and wide solenia with 

 the lateral members, and also that, as at n. i/., some of the lateral members of larger size 

 communicate by similar solenia with lateral members of smaller size. It is obvious that 

 there are very few solenia in Lemnalia nitida, and that those Avhich exist are almost 

 exclusively confined to the cortical part of the branch. What is true of the branch is 

 also true of the stem. 



If lateral members, such as those marked m in fig. 1!), be traced downward from their 

 solcnial connection with a chief member, they are observed to gradually thin out and 

 end in a cul-de-sac. If they are traced upward they can l)e followed through a large 

 number of sections, gradually increasing in diameter, till eventually they emerge on the 

 surface as anthocodias. In other words, the lateral members do not rise directly from the 

 chief members, but indirectly, through the intervention of solenia, and they grow both 

 acropetally and basipetally, bnt their basipetal growth is inconsiderable and may be 

 neglected. Although the lateral members originate indirectly from the chief members 

 in this manner, they may be regarded, for present purposes, as springing directly from 

 them, for by so regarding them we may form a clearer idea of the branching-system of 

 the anthodete. 



The branch of which a section is represented in tig. 19 broke up into a number of 

 branchlets l)earinu: anthocodice. The group of anthosteles marked A, B, C (chief members) 



73 * 



