FAMILY CERATIACEAE 



57 



thin, hyaline plates. Most of area anterior to girdle 

 covered by one large plate, the largest of the area (fig. 

 671). This plate is bordered along most of its margin 

 with a line of closely spaced minute tubercles. The rest 

 of the plate is covered with fairly regularly spaced tu- 

 bercles of larger size. At left side of this plate there 

 is a long, narrow, smooth plate. Ventral area posterior 

 to girdle composed for the most part of two smooth 

 plates. On the right side these are shorter than the an- 

 terior plate because of the projection of the horn trough 

 into the ventral area. The above four plates were each 

 dissected from the others, so that their occurrence as 

 distinct integral parts is well established. They include 

 all the plates of the right side of the ventral area. 

 There are other plates, however, on the left side. The 

 number and position of these as well as the exact loca- 

 tion of the flagellar pore or sulcus could not be deter- 

 mined. The plates in this region are extremely mem- 

 branaceous and nothing regarding their true shape or 

 orientation could be ascertained. 



Comparisons . Our analysis of the ventral area of 

 C. pavillardii agrees in part with that of Entz (1927) for 

 a number of fresh-water species. Entz also found that 

 most of the area is made up of three large plates. In 

 his species, however, the shapes and relative sizes of 

 these plates varied considerably in different species. 

 It is probable, however, that the presence of these three 

 plates is characteristic of all species of Ceratium , both 

 fresh- water and marine. 



In regard to the other features of the ventral area, 

 no analysis has been completely satisfactory. Entz 

 shows for most species two small plates at the left side 

 of the area alongside the longitudinal furrow. The na- 

 ture of this furrow or the position of the flagellar pore, 

 however, must still be considered uncertain. 



Entz's analysis of the right side of the ventral area 

 and distal end of girdle is also unsatisfactory, so that no 

 direct comparison can be made with our analysis of that 

 region in C. pavillardii . 



Our own analysis of C. arctlcum indicates that there 

 is probably considerable variation among the species in 

 respect to this region, so that the condition of C. pavil - 

 lardii cannot be considered representative of the entire 

 genus. In C. arcticum the fourth girdle plate is molded 

 into a cap for the horn trough as in C. pavillardii, but 



the shape of the distal end of the third girdle plate is 

 regular, not hammer-shaped. Furthermore, there is a 

 continuation of girdle lists, or at least ridges, to the 

 ventral area, so that the distal end of the girdle proper 

 may be said to extend entirely to the ventral area, in 

 marked contrast with the condition in C. pavillardii . 



Chain formation . Ceratium pavillardii is often 

 found in chains of several individuals. For this reason 

 the structure of the end of the apical horn and the horn 

 trough of the body is of particular interest. The end of 

 the apical horn is covered by a small platelet, the api- 

 cal platelet. It is pierced in the center by a large pore. 

 In chain specimens (except in the anterior cell) this 

 platelet, along with the ends of the apical plates, is very 

 much thickened and convoluted, so that a knobby process 

 is formed which holds the end of the horn fast in the 

 socket of the top of the horn trough of the anterior spec- 

 imen. This socket in the anterior specimen is formed 

 mostly by the fourth girdle plate. Specimens not in 

 chains have neither the knobby apical horn nor the deep 

 socket at the end of the horn trough. Specimens in 

 chains, with the exception of the apical cell, have much 

 shorter apical horns. Perhaps a time factor operates 

 here. It is suggested that, after division, if the cells 

 separate immediately, there is a growth at the end of 

 the apical horn which results in an increase in the length 

 of the horn, whereas, if the specimens remain connected 

 for a short length of time, this growth results, not in 

 increase in the length of the horn, but in formation of a 

 knob at the end of the horn which acts to fasten the two 

 cells together so that by the time cell growth is com- 

 pleted the two cells are firmly secured to each other. 

 Chain formation, on this basis, then, would depend on 

 the formation of the knob during a short critical period 

 while the mother cell undergoes division. 



Distribution . The genus Ceratium is distributed 

 over all the oceans of the world and is one of the most 

 valuable genera of the peridinians for distributional 

 studies. Not only are there cold- and warm-water spe- 

 cies, but many species show minor phenotypic variations 

 which are useful in tracing dynamic conditions. 



The discussion of the distribution of aU the species 

 of Ceratium of the Carnegie collection constitutes an- 

 other report. Since the distribution of C. pavillardii is 

 included in that work, it is not reported here. 



