iv] GREEN CELLS OF CONVOLUTA 109 



When just-hatched C. roscoffensis are examined with 

 the high power of the microscope, no green cells are 

 to be seen in their bodies, nor are there present any 

 colourless cells resembling in shape or structure the 

 green cells, nor do either eggs or larvae appear to 

 contain leucoplasts. 



If just-hatched animals are transferred to sea- 

 water filtered by means of a Pasteur-Chamberland 

 filter, though, in the course of time, some may become 

 green, many remain colourless. Therefore it is highly 

 probable that the green cells do not owe their origin 

 to^colouflesTantecedents (leucoplasts) present in the 

 eggs7~For7were such forerunners of the green chloro- 

 plasts present, they would develop into chloroplasts 

 in all, or at all events in the great majority, of the 

 larvae. On the other hand, if young animals are 

 hatched and allowed to remain in ordinary unfiltered 

 sea-water, green cells make their appearance with 

 certainty in the animals in the course of one or two 

 days. In the youngest larvae, there are to be seen 

 no more than two or four green cells, each of them 

 lying in a clear vacuole and occupying a fairly definite 

 situation in the body. Two such cells lie right and 

 left, a little behind the otocyst and two right and 

 left about the middle of the body. By their repeated 

 division is produced ultimately the whole contingent 

 of green cells of the adult body. 



At stages earlier than this, no green cells are to 



