kofoid: mutations in ceratium. 225 



indeed G. tripos (&. str) itself. It will suffice for present purposes to 

 designate it as C. tripos. 



The other three members of the chain are of a different type. The 

 differences lie in the following characters : (1) The antapicals are un- 

 equal, (2) posteriorly directed with some lateral curvature, and (3) the 

 apices except in II 3 taper to sharp tips. (4) There results from the 

 posterior direction of the horns a deep postindentation with sharply 

 defined rectilinear postmargin. In C. tripos (cf. Fig. A), on the other 

 hand, antapicals are curved anteriorly to a direction subparallel to the 

 apical and their tips are abruptly pointed. There is no postindentation, 

 and the postmargin is broadly curved with no sharp limits. The differ- 

 ences between the two types are profound and coincide with the distinc- 

 tions between the subgenera Tripoceratium and Biceratium. The three 

 forward members of the chain I 3 , II 3 , and III 3 belong to the species 

 G. californiense. In cell II 3 the antapicals are distally curved outwardly 

 a little more than in I 3 and III 3 , and the tip of the right is bluntly 

 rounded and that of the left abruptly pointed, but not in the C. tripos 

 fashion with a median symmetrical point, but with a sharp point at the 

 outer margin, a feature found in many other species of the subgenera 

 Biceratium and Amphiceratium, but not in those of Tripoceratium. 



The species Ceratium californiense was originally described by me 

 (1907) from the waters of the Pacific off San Diego. It occurs through- 

 out the eastern tropical Pacific in the material of the " Albatross " 

 Expedition. It has also been recorded from the Indian Ocean by 

 Karsten (1907). It is in my own experience a relatively rare species, 

 occurring sparingly in the plankton. There are many such species in 

 the different subgenera of Ceratium. The species C . tripos and C. osten- 

 feldi are, however, common species of wide distribution in all seas save 

 polar waters. 



Progressive Perfection of the Mutant. 



The form of individual II 3 is of great interest and possibly of consid- 

 erable significance. The wider spread of the horns and the absence of 

 the tapering tip bring this individual somewhat nearer than I 3 and II I 3 

 to the ancestral type, though neither curvature nor tip is like that of 

 C. tripos. It looks as though there were some subtile ancestral influence 



which at the first division (I x to ^T") tended, though unsuccess- 



fully on the whole, to prevent the complete saltation from C. tripos to 

 G. californiense. Two generations (asexual !) are here required to per- 

 fect the mutation. 



