211 



laiiied a certain size their presence is already sliowii on leganliiig liie IVoiital 

 surface of tlie respective zooecia, the latter then lieing half transparent. Pis. VII 

 and VIII show dilTerent stages in development of such ooecla in Cell, aiistralis. 

 Cell. ri<jiila and Cell, atlantica. 



The avicularia, only occurring in small numbers, vary much botli in size 

 and form, and the largest of them, the dimensions of which are similar to those 

 of the zocccia, reach right to the axis of the colonj', while this is not the case 

 with the smaller of them, the latter being only wedged in between the zooecia. 

 Judging from the figures given by Busk in his account of the Bnjozon of the 

 Challenger Expedition we should imagine that these avicularia had constantly a 

 complete subopercular cryptocyst. But although the latter may be unusually 

 strongly developed, it is only in exce[)tional cases and in older zooecia, e. g. in 

 Cell, fistiilosa, that it reaches right up to the operculum. As a rule it has either 

 one median or two symmetrical incisions dislally, and in Cell. iiKilviiiensi.s, which 

 is in fact one of the species figured by Busk, the median incision is separated 

 from the opercular area by a tiny cryplocj'st arch, which unites the two inner 

 ends of the suspensory facets of the mandible. 



The very peculiar fact, that the areas perceptible on the surface of the colony 

 are by no means ecpud in size and extent to the zooecia, has hitherto escajied 

 notice, and Busk's description of the separate superficial divisions as »areas is 

 in accordance with his incorrect conce[)tion of the above mentioned liliform chi- 

 tinous thickenings as a system of hollow lilamenls shared in common by the 

 whole colony, imbedded in and effecting the growth and the calcification of the 

 separating walls, which he imagines to exist between the separate areas. There 

 can however be no doubt that Busk thinks every area to correspond with a 

 zoa'cium (»Zoa^cia completely immersed, each corresponding to an area '). That 

 the areas and the zooecia do not correspond in this family is most easily seen 

 on isolating a single zoo3cial layer of Cell. (ilUuitiai and grinding away the basal 

 wall (fig. 2 c), as the narrow eloiigated zocrcia and the much shorter and broader 

 rhombic areas may then be seen at the same time. In regarding a longitudinal 

 grinding it will also be very obvious that the orecia, the bottom of which is a 

 part of the separating wall between two zoircia lying in the same longitudinal 

 row, open in the distal part of an area but in the proximal part of a zocxcium 

 (PI. VIII, figs. 1 a, 1 I), PI. VII, fig. 4 f). 



We may now by means of longitudinal and transverse sections make a closer 

 inspection of the way in which the separate zotrcia are mutually connected in 



' 8, p. 83. 



14* 



