572 Annals of the South African Museum. 



of which has as yet been described for any species. The first, the 

 Preparatory Stage, lasts for some time, probably at least 2 to 4 hours ; 

 the second, the actual process of Inversion, when once it has com 

 menced, is more rapid, being completed in from 30 minutes to an hour, 

 the average time taken being about 40 minutes. 



A. Preparatory Stage. (1) At the end of cell-division the ostiole 

 or phialopore is very small, sometimes almost obliterated, with rather 

 ill-defined outline (Plate XL, A). The first sign of change preparatory 

 to inversion is seen in the lip ; the outline becomes more definite 

 (Plate XL, B ; Plate XLI, A), the cells forming the lip more regular 

 in outline and darker in colour, while the ratio between the axes of 

 the cells alters, the radial axis elongating at the expense of the 

 equatorial. Meanwhile the free edge of the lip curls gradually 

 inward and downward into the hollow of the daughter, the position 

 of the connecting strands often showing very clearly at this stage 

 (fig. 3, A, and Plate XL, C). 



(2) Almost coincident with this change in the lip of the pore, 

 the daughter colony, which at the end of cell-division presented a 

 firm rounded outline flattened on the outer side, begins to dent 

 irregularly. Dents appear first at one point then at another, becoming 

 progressively deeper and more extensive as the lip curls further in- 

 ward. During this denting period the young colony has a distinctly 

 battered appearance and is in a state of highly unstable equilibrium 

 (fig. 3, A-C; Plate XL, C, D). In preserved material, unless special 

 care has been taken in fixing, colonies at this stage nearly always 

 appear completely collapsed. 



(3) After some time the dents begin to disappear, the lip to straighten 

 out, and the colony has once more a firm, rounded outline (fig. 3, D ; 

 Plate XL, E, F). Now, however, the size of the colony is slightly 

 smaller ; the radial diameter is no longer the shorter, it is either equal 

 to or a little longer than the equatorial. The cells are longer and nar- 

 rower and more regularly polygonal, as seen in surface view. Both 

 the movement of the lip and the denting of the young colony are caused 

 at least in part by this change in the shape of the cells, which con- 

 tinues during the next stage of the process. The change does not take 

 place simultaneously in all the cells of the colony, but first in one 

 part, then in another ; hence the irregular appearance and disappear- 

 ance of the dents. What role exactly the free lip plays in the process 

 is not clear. It is evidently important, and apparently the change 

 in shape of the cells is initiated there, and yet later in the process it 

 is these cells which are most distorted and pulled out tangentially. 



