558 Annals of the South African Museum. 



particularly in their mode of inversion. The subject is an interest- 

 ing one and would probably repay further study. 



(vii) Contractile Vacuoles. In addition to the highly refractive 

 food vacuoles already referred to, each cell possesses a varying number 

 of contracting or pulsating vacuoles. They may occur in the colour- 

 less throat (i.e. upper ventral portion) of the cell, in the outer layers 

 of the broad basal part usually near the bases of the connecting 

 strands, or actually in the strands themselves (fig. 1, Plate XIX, 

 C-E). 



Those in the upper part of the cell are usually difficult to distin- 

 guish, particularly in the anterior part of the colony ; in the cells 

 near the posterior pole they are often clear, especially in older colonies. 

 Whether they are always present in this region is not clear ; more 

 than one in the throat of a somatic cell was rarely seen. In the basal 

 part of the cell, on the other hand, they are very easily seen in surface 

 view, particularly if the oil-immersion lens is used. In number they 

 vary considerably, one cell may show only one vacuole in this region 

 or as many as four or five. They appear to be within the outer layer 

 of the chloroplast, since in diastole a thin layer of green is sometimes 

 distinguishable outside the vacuole. When the vacuoles occur in 

 the connecting strands, usually in old colonies, they show very clearly. 



As the colony ages, the number and size of the contractile vacuoles 

 tend to increase. This also depends to some extent on external 

 conditions ; for instance, material which has been kept in culture for 

 some time often becomes strongly vacuolate, as noticed by Klein 

 (1890, p. 9) in the European species. 



The period of pulsation for any one vacuole remained fairly con- 

 stant during the time of observation, and varied from 15 to 55 seconds 

 in those which were timed. Most often the period was in the neigh- 

 bourhood of 30 seconds. No relations could be found between the 

 periods of the various vacuoles of any one cell. 



In the reproductive cells the number of vacuoles is fairly high, 

 but there, too, excessive vacuolation would appear to be a patho- 

 logical sign. 



The system of pulsating vacuoles here observed is obviously very 

 similar to that found in I 7 , globator and described by Klein (1890, 

 p. 9), Janet (1912, pp. 52-53), etc., the chief differences from their 

 descriptions being the presence apparently of only one, or possibly 

 not even one, vacuole in the throat of the cell, and the position of the 

 vacuole, apparently within the outer layer of the chloroplast, not 

 merely within the ectoplasm. 



