262 SIDNEY F. HARMER. 



which hcas been described by Loraas ^ inCellaria fistulosa. 

 The young branches in this species, as in Crisia and doubt- 

 less in other jointed Polyzoa, arise as outgrowths of the older 

 branches, formed of zocecia which are at first rigidly connected 

 with their parents. By the " breaking " (no doubt accom- 

 panied by some absorption of the calcareous walls) of the 

 base of these zooecia a joint is formed, the broken ring dis- 

 playing chitinous tubes beneath, which connect the young 

 branch with the older one. 



The rootlets of S. neozelauica are probably compara.ble 

 with the connecting tubes of other jointed Polyzoa, although 

 they are continuous with the external epitheca. Assuming 

 the correctness of Nitsche^s account already referred to,- the 

 connecting tubes of Cellaria may be regarded as formed 

 from the chitin which is left on the inner side of the calcareous 

 wall by the calcification of the middle lamella only of the 

 original chitinous ectocyst. S. neozelanica is thus to be 

 regarded as a jointed species -, and if what has been said 

 above with regard to the arrangement of the zooecia is correct, 

 the group of tubes may be derived from an originally Escharan 

 (bilaminar) colony, with narrow branches given off in a pal- 

 mate manner, a longitudinal division of the colony taking 

 place with the formation of each fresh branch. 



The edge of the zocecia is very slightly raised, and slopes 

 continuously into the post-oral shelf. This is finely tuber- 

 cular, and often slopes continuously into the cryptocyst. The 

 oral shelf is represented by a mere line. The condyles are 

 small. The oral arch is thick, and practically not raised 

 above the level of the epitheca. The cryptocyst is thick and 

 tubercular, not much thickened at the commencement of its 

 descent, which is angulated and at right angles to the proximal 

 portion. The cavity of the median process is very deep, its 

 upper opening usually in the form of a longitudinal slit, the 

 cavity widening regularly as it descends until it reaches the 

 same transverse diameter as the tube where it joins this 



' ' Proc. Liverpool Biol. See.,' iii, 1889, p. 219. 

 2 See p. 227. 



