228 BULLETIN : MUSEUM OF COMPARATIVE ZOOLOGY. 



C. intermedium, especially in the left horn, than it is in C. ostenfeldi. 

 Compare on this point the original figure of G. macroceros (Claparede et 

 Lachmanu (1858-1861, plate 19, fig. 1), Ostenfeld's (1903, p. 584, figs. 

 136-139) figures of G. intermedium, and Kofoid's (1907 c, plate 26, 

 figs. 22-25) of C. ostenfeldi. The three species here named form a group 

 of closely related but clearly separable species, much confused in the 

 earlier and some of the recent literature (see Paulsen, 1908, for the 

 synonymy of the first two, and Kofoid, 1907, for that of G. ostenfeldi). 



Evidences of Genuineness and Completeness. 



The internal evidence is conclusive that this is a genuine chain and 

 not a chance union. The connection is in every way typical. On the 

 other hand, it is not certain that the whole of the chain is here repre- 

 sented. The relatively short apical horn of the anterior member is 

 indicative alike of the recent formation of the chain and of this part of 

 the skeleton ; that is, it is not Aj. but A 1+x . A comparison of this chain 

 with the one previously discussed and that of C. vultur, shown in plate 4, 

 fig. 7, will tend to confirm this conclusion. Again, it is also uncertain 

 whether or not the posterior member of the original chain is here repre- 

 sented by the rear cell. I can find no internal evidence that is abso- 

 lutely conclusive in either direction. The chain may therefore be 



comparable with ^ or with — f of the chain shown on plates 1-3, or 

 1V 3 -ll^a 



in more general terms with ±=r= f, in which x represents the number 



& (II + x) y 



of these individuals in the chain beyond the first, and y the number of 

 the generation to which they belong. It is, however, to my mind more 



probable that the pair was not terminal (y^) but intermediate f ^j- J 



in position, in such a relation as exists in the case of j^T in the chain 



on plates 1-3 ; that is, the present cells were not separated in the last 

 schizogony, and consequently the anterior skeletal moiety of the fore- 

 most cell and the posterior of the rear one were not joined in the skeleton 

 of any parent cell. They are not sister cells, but cousins of indetermi- 

 nate degree of nearness or remoteness. The grounds upon which this 

 conjecture rests are : (1) the forward cell is 20 per cent wider at the 

 girdle than the rearward one ; (2) the lengths of the sides of the plates 

 along the fission lines differ so in the two cells that they cannot be 



