105 



(iilVer from the coii-es|)onding thickenings or bands in ]i. niliciihild in several 

 ways. Tluis, their thinner middle portion passes evenly over into two thicker, 

 rounded marginal portions (fig. 2 k), and further they are much broader, occu- 

 pying more than half the periphery of the whole segment. They divide this into 

 lour, unequally large areas, of which the largest lies on the basal surface of the 

 colony and the narrowest on the frontal surface. This last area which does not 

 seem to have any membranous jjortion at all or to be jirovided with parietal 

 muscles is not rarely divided into two or more areas behind one another, as the 

 two calcareous thickenings may be connected by one or several transverse bridges. 

 The distal wall (figs. 2 1 — n, 2 j), which is saddle-shaped fiom side to side, is 

 over its whole surface provided with extremely numerous, small uniporous rosette- 

 plates, so that the whole di.stal wall migiit really be regarded as a large multi- 

 porous rosette-plate. In its whole extent it is attached internally to the wall of 

 the segment by ascending, often branched chitinous rods, which are apparent 

 from the outside anil produce digitate and lobate figures. 



The colony consists of a nunil)er of jointed stems which have two alternate 

 rows of llahellate branches with lour to five bifurcations and up to 40 linear 

 segments with biserial zooecia. Each of these branches is in connection with the 

 stem through a multiporous rosette-plate, which is placed a little proximally to 

 the end of a stem-segment (tigs. 2 n, 2 i, 2 j), and such a rosette-plate is only 

 found on that side of the segment where a branch issues. I have only in a few 

 cases found two successive branches arising from the same side but never two 

 branches placed at the same height. From each stem again 2 — 5 new stems 

 arise, in most cases just opposite a branch, more rarely alternately with these, 

 and in a few cases I have found a new stem arising just proximally to a branch. 

 New stems which are in connection with the main-stems through multiporous 

 rosette-plates seem to arise in a doul)le way, partly by transformation of branches 

 and partly independently. While the common branches are fixed by a zocecium, 

 the proximal part of which is provided with two rings, I have seen a number 

 of branches only ditferent from the others therein that 1 — 2 slender segments 

 are interposed between the zoo-cium and the stem, and most likely these branches 

 are about to be transformed into new stems. A great number, however, of the 

 young stems, which issue fiom the main-stems and for instance those, which are 

 seen in fig. 2 a, cannot well have arisen in this way, which in the first instance 

 may be inferred from their being generally placed opposite the branches, as two 

 branches are never [)laced at the same heiglh. In the next place these stems are 

 characterized by their being provided with branches only at a very late period 

 though there is a rather large ditlerence in that respect. Only in a single stem 



