586 Annals of the South African Museum. 



stages of cell-division) in different androgonidia, and the course of 

 inversion in single individuals. The difficulties in obtaining good 

 photographs are considerable. After the first few cell-divisions we 

 are dealing with a dense, very small individual (diameter about 

 30 ft when mature) suspended inside the parent. When young the 

 latter, too, is comparatively dense, hence the best results were 

 obtained from large, fairly mature colonies where many of the globoids 

 had already escaped, and the cells of the parent were wide apart 

 and consequently not very dense ; even late in life such colonies 

 often still contain young developing antheridium mother-cells, and 

 thus offer favourable subjects for observation. 



In the case of the earlier stages a fortunate chance helped consider- 

 ably. A young male colony with antheridia in all stages happened 

 to get split by pressure of the cover-slip, one half lying with the 

 hollow inside facing upwards. Along the free edge of this several 

 antheridium mother-cells and antheridia in the very early stages of 

 development, each with its vesicle intact, were exposed, and pre- 

 sented specially good subjects for photographing (cf. Plate XLII, 

 D-J), since they were not masked by over-lying cells. The best 

 results were obtained from living material, using the 1-9 mm. oil- 

 immersion objective and a Leitz-3 ocular. The stronger oculars 

 were less satisfactory, cutting off too much light, with some loss of 

 definition. It was found preferable to obtain the additional magnifi- 

 cation by using a greater extension of the camera attachment. 



In obtaining the series showing inversion, it proved exceedingly 

 difficult to obtain a good, complete series. To begin with, the pre- 

 paratory stage is prolonged, and it was extremely difficult to be sure 

 when a bundle was just beginning to invert. Then it involved 

 keeping the chosen bundle under observation with the oil-immersion 

 lens for several hours without any crushing or drying. In several 

 cases a good series was begun, only to be spoilt in the middle. Series 

 which proved excellent at some stages were spoilt by slight movement 

 of the object at crucial moments. In a series where the end was 

 good, the beginning was often unsatisfactory, and vice versa. The 

 one chosen to illustrate this paper (Plate XLIII) is in some ways 

 not the best obtained, nor is it complete, since after the twelfth 

 negative had been exposed, the projection lamp came to the end of 

 its life, and work was held up for some days before another could be 

 procured. As a consequence the last three photographs showing the 

 completion of the process were underexposed, and these stages have 

 had to be supplied from another series. But the early stages on the 



