216 bulletin: museum of comparative zoology. 



tendency, more or less pronounced, to adhere in chain for some time after 

 schizogony. This tendency is moi-e marked in some species than in 

 others. For example, in the subgenus Euceratium Gran in the species 

 C. vultur, C. schrariki, C. carriense, and G. pahnatum, chains of 4-20 

 individuals are not unusual, while in C.furca, C.fusus, C. biceps, and other 

 representatives of the subgenera Biceratium and Amphiceratium rarely 

 more than two individuals (except in C candelabrum) are seen in chain 

 and chains are relatively infrequent. Chain formation is most readily 

 found in collections made in the night or in the early morning. In my 

 own experience at San Uiego (latitude 32° 40') they ai-e most abundant 

 in collections made between 3 and 7 o'clock in the morning, cell division 

 seeming to be favored by the conditions of illumination and possibly by 

 the falling temperature prevailing during the night. The schizonts part 

 company shortly after the completion of schizogony and skeletal forma- 

 tion, and chains are absent or relatively rare in collections made later in 

 the day. 



The morphology of chain formation is correlated with the presence of 

 an apical pore at the end of an apical horn. As the new skeletal moie- 

 ties are formed respectively on the posterior and anterior regions of the 

 diverging schizonts, the plasma of the posterior member is drawn out in a 

 long strand which becomes the apical horn. Its tip rests inmiediately 

 upon the distal end of the newly forming girdle (plate l), at which 

 point the plasma of the two individuals remains in continuity without 

 interference by the forming skeleton. As the newly forming skeletons 

 are completed, the apical pore of the posterior schizont is set under the 

 anterior shelf or list of the distal end of the girdle at the margin of the 

 ventral plate (plate l) of the anterior schizont. The posterior list of 

 the girdle is not formed at this point, and the apical horn as it passes 

 posteriorly lies in a channel or depression on the ventral face of the 

 midbody along the right margin of the ventral plate. The place of 

 junction on the anterior schizont I designate as the attachment area 

 (att. a.) and the depression as the chain channel (ch. ch.). The anterior 

 end of the apical horn is also modified, its ventral side being prolonged in 

 a short lobe, giving to the apical pore an irregularly oblique opening 

 (Entz, 1905), a condition found in individuals in chain and also in those 

 but recently released from chain formation. In most free individuals 

 the apical pore is transverse, as in the anterior parental skeleton of the 

 schizont I3 in the chain shown in plate 1. 



In the course of my investigations upon the dinoflagellates of the 

 Pacific conducted for the past eight years at the San Diego station and 



