584 Annals of the South African Museum. 



that runs along near the Wetton Road vlei on 12th October, and 

 thus, when these observations were made, had been kept in culture, 

 although in vlei water, for four days. Further, even if this had not 

 affected it, when a colony of such a size, especially when so loaded 

 with daughters, is placed on the slide a great part of its weight is 

 no longer supported by water, and it is very liable to split merely 

 under its own weight. Hence, though interesting, it is an exception 

 on which much reliance cannot be placed, since conditions were 

 certainly far from natural. 



The parent colony may continue active for some time after the 

 escape of the daughters, even though it has several large gaps in its 

 posterior half. Eventually, however, it disintegrates. 



As the daughters escape their rate of rotation and progression 

 becomes very high, and they swim about exceedingly rapidly. At 

 first they are very deep green owing to the close juxtaposition of 

 the component cells, but as development proceeds the protoplasts 

 become further and further apart owing to the gelatinisation of the 

 intervening walls, the protoplasmic connections become more 

 apparent, the eyespots in the anterior half become well developed, 

 and soon the gonidia begin to divide. 



In vigorous strains young escaped daughters of all ages are seen 

 moving about among the mature colonies. The daughter colonies 

 are large at birth, on the whole larger in V. Rousseletii than in V. 

 capensis, but the actual size varies enormously ; this depends partly 

 on the age of the strain, partly on external conditions, but also to a 

 great extent on individual idiosyncrasy of the gonidium, cf. Plate 

 XIV, A, where all 14 daughters in the large colony are mature, but 

 show very great variation in size. 



In V. Rousseletii one of the smallest daughters seen free measured 

 182 x 204 /z, while the largest seen in situ was 630 x 672 /z. This 

 was, however, exceptional, although daughters over 500 fj, in dia- 

 meter (e.g. 525 x 567 /z) are not uncommon at times. The average 

 in vigorous normal strains lies between 300 /j, and 400 p,, falling 

 considerably as the strain ages. 



In V. capensis the largest daughter colony measured in situ was 

 467 x 482 fji ; colonies averaging 320 x 360 /JL are common, but here 

 again the size decreases with the age of the strain. In cold weather, 

 too, the daughters are noticeably smaller and composed of fewer cells. 



In general, the size of the daughter at birth is directly proportional 

 to the size at the beginning of inversion, and this, too, varies greatly. 

 Daughters were measured at the beginning of the preparatory stage, 



