KOFOID AND SWEZY: UNAKMORED DINOPLAGELLATA 19 



massive size and great niimber of the chromosomes. They furthermore show 

 indications of a high degree of specialization along several lines. One of these 

 is the development of a paradesmose, which reaches its maximmn in Noctihoca 

 with its "sphere," consisting of a mass of archoplasm containing the centro- 

 some, which forms the axis of the karyokinetic figure. The I'ole of the para- 

 desmose (centrodesmose) in the dividing nucleus of these forms is one which 

 still requires explanation and confirmation. The conflicting figures of Jollos 

 (1910) and Borgert (1910) on this point in Coiitium show clearly the need 

 of further investigation. 



Another line of development is indicated in the large, distinctly marked 

 chromosomes which, in some cases at least, do not entirely lose their individ- 

 uality from one mitosis to another. In the living organism these are remark- 

 ably clear, appearing as moniliform threads of a clear, homogenous substance, 

 and are evident in most of the iudiA'iduals observed. 



Figures of l)inary fission in the dinoflagellates are frequent in the literature, 

 l)ut a critical analysis of the various steps in the process has been made in only 

 a few cases. Outside of the classical examj^le of Noctiluca, only one member 

 of the GATunodinioidae, Gijrofliiiiuni fueoriim (Gijiintodiiiiton fuconini) Jollos 

 (1910), has thus far received such attention. The mitotic process has been out- 

 lined most fvdly in the thecate forms by Lauterborn (1895) in Ceratium liirun- 

 (JiiK'Uii and in ('cniH/ini tripos by Borgert (1910). The work of Jollos on 

 Gyrodinium fucorum, though incomplete, combined with the scattered refer- 

 ences to this subject that may be foimd in the literature, and our own obser- 

 vations on these forms, clearly indicates that the process is essentiall,y the same 

 for the majority, at least, of the Gymnodiniidae as that which has been described 

 for the thecate forms. This conclusion is further strengthened by the identity 

 in nuclear structure between the thecate and non-thecate dinoflagellates. As 

 illustrative of the process of mitosis in this grouj) we have therefore selected 

 Borgert's work (1910) on Ceratium tripos var. snhsala. 



At the onset of division the ordinary spheroidal or ellipsoidal nucleus be- 

 comes elongated and, in some cases at least, the surrounding membrane is lost 

 (fig. D, 1). The chromosomes lose their earlier parallel or subparallel arrange- 

 ment and are found in a tangled skein or spireme, grouped in pairs (fig. D, 2). 

 According to Borgert, this is the result of a longitudinal splitting of the indi- 

 vidual threads. Lauterborn (1895) did not figure this stage in Ceratium 

 hirundineUa, nor did Jollos (1910) for Gijrodinium fucorum, Ceratium tripos 

 or C. fusiis. The work of both of these investigators left this stage .still unde- 

 termined. 



Following the spireme phase the chromosomes become arranged in an 

 equatorial plate (fig. D, 3). Tn all cases the jxihii- axis of the mitotic figure or 

 spindle at this stage, as shown by the position of the chrouiosomes, lies in the 

 short axis of the nucleus, necessitated doubtless by the great nunil)er of chro- 

 mosomes. A definite sj)indle has not been figured by Borgert, Lauterborn, or 



