KOPOID AND SWEZY: UNARMORED DINOFLAGELLATA 63 



thtit most, if not all, dinoflagellates pass through a "pyrocystis" stage or its 

 equivalent in their development. 



Our knowledge of the development of the so-ealled Pijrocystis lunula 

 (Gym nodi III u))i lunula) is still incomplete, but it serves to represent what we 

 already know of the life cycle of this group. In our somewhat diagrammatic 

 figure representing this cycle (fig. I) the beginning of the series is placed where 

 it very obviously does not belong. The beginning pro1)ably lies in the still 

 unbridged gap represented by the question mark. There is some evidence, very 

 slight to be sure, and needing further confirmation, that conjugation may occur 

 at this stage between the large globular organisms, and probably not in the 

 small flagellated forms, as might be expected, just escaped from the cyst 

 (fig. 1,8). 



This large spherical organism, without visil)le means of locomotion, first 

 attracted the attention of Murray on the voyage of the Challenger Expedition 

 of 1873-1876. His notes on it were published in his reports of the results of 

 the expedition in 1876 and 1885, with the name Pijrocijstis suggested as an 

 appropriate designation. It has since that time 1)een found to occur in all 

 oceanic waters where dinoflagellates have been studied, its distribution being 

 coextensive with that group. 



Pouchet (1885r() described a crescent-shaped cyst as "Peridinium voisin de 

 Gi/mnoduiium spirale," showing a single large Gijmuodinium together with 

 other t^-]iical lunula cysts, one of which he figured with five smaller Gymno- 

 diniunt-YikQ flagellates contained within it. In Schiitt's monograph on the 

 Peridiniales (1895) he figured as Gymnodiniu ni lunula the crescent-shaped 

 cysts, and later (1896) changed their generic designation to Pyrocystis, as P. 

 lunula, thus recognizing their relation to the forms earlier described by Murray, 

 but as independent organisms, not as two phases of the same cycle. He divided 

 the development of Pyrocystis into two stages, the first the tetraspore stage in 

 which the cells are rounded to spindle-shaped, and the second or Gymnodinmm- 

 like stage developing from the first. 



The development of the spindle or crescent-sha])ed cysts from the large 

 spherical ones was not established until Dogiel (1906) and Apstein (1906) 

 pu])lished the results of their studies on the same forms. They followed the 

 development from the ])yrocystis stage through the siiccessive mitoses to the 

 formation of the small Gym nodinium and its escape from the crescent or second- 

 ary cyst. The account of the life cycle presented lierewith is largely taken from 

 Uogiel's acbnirable work, supplemented Ijy the results of our own observations. 



The change of the small Gymnodinmm (fig. I, 11) to the large spherical 

 form has thus far escaped observation. This (fig. I, 1) is a naked form with- 

 out a cyst wall, though at first glance it might pass for an encysted individual. 

 The periplast is thick and closely adherent to the ])ody. Tlie nucleus is found 

 at one side close to the periphery, in a rather dense, granular mass of cyto])lasm. 

 From this mass long streamers of cytoplasm run out, following the periphery 



