106 



with two segments I have found a little terminal hranch showing a double bi- 

 furcation while a common branch on a main-stem as already said shows 4 — 5 

 bifurcations. In all the other young stems the tip is devoid of a branch, which 

 no doubt has fallen of. While a terminal branch is a direct continuation of a 

 stem and does not arise from a rosette-plate, the formation of a rosette-plate 

 always precedes the formation of a branch, and even if all the lateral branches 

 have dropped of, as is the case in many of the stems examined, their position 

 and number is indicated bj' the respective rosette-plates. The examination of a 

 number of stems shows that the first rosette-plate appears proximally to the 

 partition-wall between the end-segment and the one next to it, and that the 

 development of these formations goes steadily downwards. I shall here give a 

 few instances showing the difference in the appearing of the rosette-plates. 



a stem with: 

 6 joints 



Tlie fifth segment with a beginning rosette-plate. 



7 — 



Rosette-plates on the six segments. 



A Rosette-plate on the seventh segment. The partition wall between the eighth and 

 the ninth segments is not whoUj' developed. 



11 



Rosetplates on the ninth and the tenth segments. 



13 



RosetiJlates on the eleventh and the twelfth segments. 



The number of segments in the trunk gradually increases by division of the 

 older segments, and when two short segments follow one another this is a sign 

 that a division has taken place lately. The proximal segment of the two has 

 then not yet got the lateral rosette-plate, and the proximal end of the distal seg- 

 ment does not yet show the rounded swelling defined by a more or less distinct 

 constriction, which is seen on the completed segment. A division like this is 

 always introduced by the two strongly calcified lateral belts on each side sending 

 prolongations towards one another, which at last join together to form a bridge. 



The radical fibres, which issue from the proximal end of the stems serve 

 exclusively to attach these to the surroundings. They are much branched, furnished 

 with irregular expansions and swellings, and like the stem divided into sections 

 (kenozooecia), which internally are separated by multiporous rosette-plates, 

 but externally have no distinct constrictions. We further find a multiporous 

 rosette-plate everywhere where a new branch arises from an older one. 

 While the radical fibres have in the beginning on arising from the stem a struc- 

 ture similar to the latter their wall soon becomes evenly calcified over the whole 

 of its surface. 



