68 MEMOIRS OF THE UNIVERSITY OF CALIFORNIA 



attained by the free-living forms. The motile individual is a typical Gymno- 

 dinium-VikQ flagellate of minute size usually and of simple organization (fig. 

 J, 5) . When it enters upon its parasitic career it loses its characteristic features. 

 The body becomes greatly enlarged, as in Blastodinium pruvoti (figs. J, 2— t) 

 and in this genus divides into macrocyte (mac.) and microc}i:e (diic). These 

 follow individual lines of development that are still somewhat obscure, with the 

 final result, the production of a number of motile Gy)nnodinhim-]ike flagellates 

 (fig. J, 5). 



A different type of life cycle is that described by Dogiel (1906) for 

 Chytriodinium roseum (= Gyninodimum 7-oseum). The parasitic form as orig- 

 inally described consisted of two cells attached to each other, one. minute, the 

 other huge (fig. K, 1) . Part of the contents from the large cell passed into the 

 small one, which increased in size Avith a corresponding reduction in the bulk 

 of the other (figs. K, 2, 3). This continued until but a small amount of cyto- 

 plasm remained of the larger cell in a somewhat collapsed cell membrane (fig. 

 K. 4). A later and probably more correct interpretation of these two cells 

 (Doflein, 1909; Chatton, 1912) is that the smaller one is the amoeboid dinofla- 

 gellate and that the larger one is a copepod egg to which the parasite has Ijecome 

 attached. It is thus a case of true parasitism, the dinoflagellate feeding upon 

 the contents of the egg and developing at the expense of its host. 



With the increase in size of the parasite the nucleus begins to divide (fig. 

 K, 2), followed by cytoplasmic division (figs. K, 3, 4) of the entire body within 

 a cyst wall which has been formed around it. This process continues until 

 eight, sixteen or more small uninucleated bodies are formed (fig. K, 5). These 

 break out from the primary cyst, form secondary cysts corresponding to the 

 crescent-shaped ones of Gymnodinimn lunula, and each daughter organism or 

 sx^oroblast, as Dogiel terms them, begins to divide (fig. K, 6). Division may 

 continue three times, the resulting minute bodies developing into a Gi/nniodi- 

 nuo^-like flagellate (fig. K, 8). 



In Oodinium parasiticum a still further complexity has been introduced in 

 that the parasite enters the eggs and part of its development takes place therein. 

 The entire process is more obscure in this species than in Cliijtriodinium roseum. 

 The products of the final division of the parasite are small flagellates with the 

 tjT^ical flagella and girdle of the Dinoflagellata. 



The striking similarity between the various stages of the life cycle of 

 Chytriodinium roseum and Gymnodinium lunula is probably due to the ecto- 

 parasitic mode of life of the former. Thus, while deriving its nourishment 

 from the body of the host, it is still exposed to the various influences of the 

 oceanic waters in about the same degree as Gymnodinium lunula. An appar- 

 ently much smaller gap is left in the life cycle of the parasitic form than in that 

 of the free-living one, since the small, amoelioid organism attaching itself to 

 the egg is similar in appearance to the small flagellate which has rounded up 

 and lost its flagella. It is possible, however, that an intervening stage may be 

 present here also. 



