FAMILY CERATOCORYACEAE 



39 



discrepancy in the appearance of the two daughter cells 

 immediately after fission (fig. 48A-I). These new spec- 

 imens have frequently been mistaken for taxonomic u- 

 nits (see under "Historical," below). 



Reproduction may, possibly, take place by the pro- 

 duction of autospores. One specimen was found in sam- 

 ple 284, station 50 (fig. 50), which indicated that the en- 

 tire shell was newly formed, and similar specimens 

 were found in sample 563, station 95. The walls of the 

 first specimen were very thin, the pores were small 

 without pits, the lists were embryonic, and the spines 

 were absent. A slight indication of the formation of two 

 of the antapical spines was found in the usual position. 

 The size was normal for the adult. This specimen 

 strongly suggests that ecdysis of the shell occurs in this 

 species and that one or more new individuals are formed 

 from the old protoplast with the formation of entirely 

 new shells. 



Variation . Although the size of the specimens of the 

 Carnegie material varied, these specimens represented 

 a homogeneous group and no separations of a taxonomic 

 nature could be made on this basis. The same held true 

 for the length of spines. Figure 51 shows the distribu- 

 tion of the frequency of various horn lengths in Carne - 

 gie material. It is evident from this that, so far as horn 

 length is concerned, only one species can be detected. 

 Lemmermann (1899) described var. lor^icornis as hav- 

 ing a horn length of 60 microns and Pavillard (1931) de- 

 scribed var. extensa with horns 120 to 130 microns long. 

 Sixty microns is near the average for this species in 

 Carnegie material (fig. 51). There are not, however, e- 

 nough data in the 120- micron range to test the validity 

 of Pavillard's var. extensa . Short-horned forms are 

 shown in figures 49D and F; long-horned forms in fig- 

 ures 49A, C, and G. 



There is considerable variation in the thickness of 

 the thecal walls, in the thickness of the lists and spines, 

 and in the ornamentation of the body generally. One fre- 

 quently finds strongly sculptured individuals like that 

 shown in figure 49B. There is every gradation from this 

 to the more deUcate forms represented by figure 47. 

 These variants are not taxonomically distinct. The heav- 

 ily sculptured individuals apparently do not represent a 

 response to changed environmental conditions, as no cor- 

 relation could be found between their occurrence and 

 the hydrographic conditions. In four cases specimens 

 with thin and specimens with thick thecae were found in 

 the same sample; in five cases such specimens were 

 found at the same station. It is probable that the strong 

 ornamentation indicates age, and thus that there is a 

 continuous accretion of skeletal material in the course 

 of the life of the individual. 



Occasionally supernumerary spines are developed. 

 These always appear in already existing lists, usually in 

 the spine lists as double spines (fig. 49C, F). Some- 

 times a spine falls to develop (fig. 49D). Lindemann 

 (1925, fig. 18) reports a specimen with only three antap- 

 ical spines. The absence of one spine in his figure 19, 

 however, may be caused by the particular view of the 

 specimen. Karsten's Ceratocorys horrida var. airicana 

 is a many-spined form. 



Historical . Ceratocorys horrida was first estab- 

 lished by Stein (1883, p. 159). Stein's figxires are re- 

 markable for their clarity and accuracy and far excel 

 any illustrations of the genus since published. 



Lindemann (1925, p. 101) apparently was not aware 

 of Kofoid's (1910) revision of the genus. Lindemann 



stated that the tabulation in this species is uncertain and 

 variable, and that there are always four precingular 

 plates (there are in fact five) and one intermediate plate 

 which breaks up into a number of plates, the pattern be- 

 ing variable as indicated by the lists crossing this plate. 

 As shown by the present investigations, careful dissec- 

 tion reveals that the pattern in constant. The number 

 of plates in the apical region, as well as in all other 

 parts of the theca, does not vary, although the pattern is 

 not always indicated by the development of lists across 

 the apical region. 



Ceratocorys tridentata Daday (1888) was accepted 

 by Entz (1905) and Kofoid (1910) as a short-horned form 

 of C. horrida . In the writer's opinion, howevfer, this 

 form cannot possibly belong to C. horrida because of its 

 elongated shape, high epitheca, and absence of ventral 

 and dorsal spines. Indeed, in the figure there is nothing 

 to indicate that it even belongs to the genus Ceratocorys . 



Ceratocorys horrida var. longicornis Lemmermann 

 (1899) and probably C. horrida var. extensa Pavillard 

 (1931) are long-horned variants of no systematic impor- 

 tance. Karsten's (1907) C. horrida var. africana, fig- 

 ured with eight feathered and two simple spines, is also 

 a variant of no systematic importance (see above). 



Dinophysis j ourdanii Gourret (1883) must be includ- 

 ed in the synonymy of C.- horrida (see p. 43 under C. 

 gourretii). 



Ceratocorys hirsuta Matzenauer (1933) must be 

 placed in the list of doubtful species. In his descriptioa 

 Matzenauer states that the species is similar to C. ar- 

 mata but has longer, very tufted spines. His figure (p. 

 453, fig. 23), however, shows the body shape and girdle 

 lists of C. horrida . There are four antapical spines but 

 no dorsal or ventral ones. This specimen is possibly 

 the right daughter cell of C. horrida (see p. 36 and a- 

 bove). 



Entz, as early as 1905, correctly described fission 

 in this species, opinions of later workers notwithstand- 

 ing. His indication of the skeletal fission line was cor- 

 rect except for details of the sulcal area. 



The form to which Mangin (1926, p. 71) referred 

 was probably the left (two-spined) daughter cell, as 

 Dangeard (1927c) stated. Mangin's statement reads, 

 "La plupart des individus sont caracterises pardepuis- 

 santes cornes later ales aux angles de la valve inferi- 

 eure, mais il existe des formes divergentes dans les- 

 quelles deux Opines trfes developpees sont opposees I'une 

 i. I'autre, les autres Opines restant trfes minces." 



Dangeard (1927c) understood the fission line of this 

 species and the significance of the two-spined speci- 

 mens. He presented iclear drawings of the left (two- 

 spined) moiety although he indicated the fifth precingu- 

 lar plate incorrectly as going to the left moiety. 



Pavillard (1931) curiously refused to accept the two- 

 spined form as a division form in spite of the clear 

 drawings of the fission line by Entz (1905) and Dan- 

 geard (1927c) and the statement by Kofoid (1910) that 

 "short horns appear at fission in the younger parts of 

 the skeleton of otherwise long-horned forms." Pavil- 

 ard suggested, on the other hand, that these forms 

 might be identical with his var. extensa, which is char- 

 acterized by the length and straightness of the horns 

 rather than by any difference in their number. 



Matzenauer (1933) apparently followed Pavillard's 

 interpretation of Mangin, but considered var. extensa 

 separate from Mangin's "forme divergente." He gives 

 figures under both titles. 



