201 



in the same colony. Such a Iransformation of the cyUndrical iiileiiiodes is found 

 in most of the more developed colonies in one or several secondary hranches, 

 and a growth in length and a division of the younger inlernodcs take place at 

 the same time. 



We have already mentioned that in a colony at a certain stage of develop- 

 ment the outenno.st or sometimes the two outermost secondary hranches on each 

 side are only composed of cylindrical internodes, and if we were to examine a 

 numher of colonies at dilTerent stages of development we should lind that these 

 whip-lilie secondary branches i.ssue nearer the top stem-internode in the younger 

 colonies tlian in the older ones. Thus, if we were to designate the internode of 

 a main l)ranch, issuing directly from the hifurcate, distal stem-internode, no. 1, 

 tlie next no. 2 and so on, we should lind that in colonies with 6 — 8 secondary 

 branches the whip-like secondary branches issue from internode 3 — 4, in colonies 

 with 9 — 12 from internode 4-6 and in colonies with 13 — 16 from internode 5 — 7. 

 This fact can only mean that all the zocecia in a number of outer secondary 

 branches arise by a transformation of cylindrical internodes. There is however 

 no certain rule for the time of the appearance of the fust whip-like secondary 

 branch, as in some colonies it may appear later than in others. The outermost 

 secondary branch on each side generally remains untransformed, and I have only 

 in very few cases found 1 — 3j^of the proximal internodes transformed into zoa-cia 

 on one side of an older colony. A consequence of the conception that a number 

 of the older whip-like secondary branches are transformed into zooecia-bearing 

 ones and that new ones are formed outside these is, that the inner cylindrical 

 internodes of the main hranches must at the same time be transformed into 

 bifurcate internodes. 



Whilst all the colonies of this species examined by me have arisen by gem- 

 mation from a branched stolon connected with other colonies.a fact explaining 

 the possibility that the development of a colony may begin with the formation 

 of a numher of individuals (.slem-internodes and branch-internodes) without or- 

 gans of nutrition, 1 have no doubt that a colony, proceeding directly from a 

 larva, must begin with the formation of a zocecium. Of this species I have been 

 able to examine numerous colonies obtained from a jointed calcareous alga 

 taken at Ajaccio by Dr. Borgesen. 



Family Alysidn'dae. 



The jointed colonies, springing from a stolonate network, consist of zo(Ecia 

 and gonozott'cia, the latter borne by stem-like kenozooecia. The zoivcia, the distal 

 half of which has a dei)ressed cryptocyst, are furnished with a simple opercular 



