592 Annals of the South African Museum. 



This part of the preparatory stage lasts a considerable time. 

 Eventually the lip of the phialopore begins to rise, and straightens 

 out, more or less completely closing the phialopore (fig. 6, B ; Plate 

 XLIII, C). At this stage it is often almost impossible to distinguish 

 the position of the phialopore, the young colony appearing broadly 

 elliptical or round in optical section, with an uninterrupted wall. 

 The change in shape of the cells is accomplished or immediately 

 followed by the commencement of cilia formation. 



(b) Inversion. There is no perceptible pause separating the two 

 phases perhaps the moment when the lip begins to straighten out 

 should really be regarded as the end of preparation and the beginning 

 of inversion proper. After the lip has straightened out completely 

 the pore begins to widen, and continues widening until it almost 

 reaches the diameter of the whole bundle (fig. 5, C, D ; Plate XLIII, 

 D to F) the " Thimble " stage. 



Meanwhile the change in the shape of the component cells con- 

 tinues, the cells having become very much elongated, the outer ends 

 sharply pointed, especially in the posterior part, often long drawn 

 out, the clear inner ends also sharply pointed (fig. 6, E). At this 

 stage the incipient cilia were first detected in favourable material 

 the very short cilia may be seen moving in the hollow, while in sections 

 they show distinctly. 



As the pore reaches its maximum extension the edges begin to bend 

 outwards and gradually fold back, making a very pretty series of hat- 

 shaped figures first a cloche hat, then a bowler, then wider brimmed 

 hats (fig. 6, F to H ; Plates XLIII, G to K, and XLIV, A, E), the 

 brim constantly enlarging at the expense of the crown. During the 

 ' hat " stages the protoplasmic strands connecting the young male to 

 the parent finally part, and the hat sometimes rolls over slightly, but 

 the displacement is much less marked than in the case of the daughter 

 colony. 



As inversion continues, the crown of the hat gradually flattens out, 

 the brim closes over the top, and usually the pore is completely 

 obliterated (fig. 6, I to K ; Plate XLIV, B, C, F). Sometimes, how- 

 ever, mature globoids were seen which still had a pore, usually round 

 in outline but sometimes either triangular or quadrilateral, in the 

 centre of the upper surface. This was, however, not seen in young 

 vigorous material, and is probably a sign of advancing senility in the 

 strain or colony, or of untoward conditions. 



