S6 



STUDIES IN THE MORPHOLOGY, TAXONOMY, AND ECOLOGY OF THE PEMDINIALES 



mediately adjacent to the suture. 



Spines . Prominent hypothecal spines are a charac- 

 teristic of the genus. All species bear them on the antapi- 

 cal plate, and two species, C. horrida and C. gourretii, 

 have, in addition, a dorsal and a ventral spine. Dorsal spine 

 produced in dorsal body list and ventral spine in ventral 

 body list. Antapical plate normally bears two spines in 

 C. bipes , three in C. skogsbergil and C. gourretii , and 

 four in C. horrida . C. armata , C. reticulata , and C. aul - 

 til. When four spines are present, they occur at the four 

 edges of the antapical plate. When three are present, it 

 is the spine of the right corner that is absent; when two, 

 the spines of the right and left corners are absent. 

 Smaller spines, representing thickenings in various 

 body lists, may occur. There are "brushes" in the long 

 spines of C . horrida and C. gourretii ; in the other spe- 

 cies spines smaller and usually simple. 



The spines of Ceratocorys are formed on the plan 

 of the junction of four lists and are not hollow as has 

 frequently been reported in the past. Each spine, when 

 viewed on end, presents the form of a cross (fig. 491, K). 

 This effect is more pronounced at the base of the spine, 

 where the lists are wider, but is evident to a certain ex- 

 tent even at the tip. On the antapical spines the lists are 

 frequently quite short, not running onto the body to any 

 extent (figs. 47, 58); but in other cases they may extend 

 to adjacent spines to join with their lists (figs.49G, 59D). 

 Lists of dorsal and ventral spines (C. horrida and C. 

 gourretii ) are more extensive. The anterior member of 

 each runs up to, and joins with, the posterior girdle 

 list. The "brushlike" or "featherlike" effect present- 

 ed by the distal half of these spines is due to the forma- 

 tion of side ribs or riblets from the center of the spine 

 along each of the four lists. This structure gives a very 

 dense appearance to outer parts of spines. There is, in 

 addition, a thickening where the four lists converge, 

 forming a sort of solid shaft, and a similar thickening at 

 the base of the spines where these join the body. This 

 thickening is especially pronounced distally, usually 

 forming a club-shaped mass on which the riblets are 

 superimposed. 



The four elements of these spines cannot be sepa- 

 rated; each is an integral part of the whole. The spines 

 do, however, occasionally break off. Specimens have 

 been found with stubs of spines and with only the thick- 

 ening at the base remaining (figs. 49H, 59D). Spines 

 have been observed in the process of breaking off. In 

 such cases no separation of the four lists could be ob- 

 served; there was a shear transverse cleavage. Wheth- 

 er this is autotomy representing an adjustment to a 

 change in external environment, as Kofoid (1908) has 

 suggested in the case of Ceratium . is impossible to say. 



Although these spines occur approximately along 

 sutures, they do not in any case run at the very edge of 

 a plate. There is always a zone between the base of the 

 spine and the actual suture. This is true of most of the 

 nonspinulate lists of the body as well, except in very re- 

 cently divided individuals. In such individuals, however, 

 the spines on the new moiety have not yet developed. 



Reproduction 



Binary fission is the only type of reproduction defi- 

 nitely known in the genus (see p. 36). Daughter cells 

 with one moiety undeveloped were found in C. horrida . 

 C. armata . and C. gourretii . The line of fission is ir- 

 regularly oblique, to the right of the longitudinal axis in 



the epitheca and to the left of it in the hypotheca. To the 

 right moiety on the epitheca go precingulars 3, 4, and 5. 

 The fission line runs to the left of the flagellar pore, so 

 that the right moiety gets all the sulcal plates except the 

 anterior. It receives also postcingulars 5 and 6 and the 

 antapical plate (figs. 48, 53C). 



Methods of Study 



As an aid to defining the species of this genus, cer- 

 tain body dimensions were utilized. These are illustra- 

 ted in figure 45. All measurements of the body were 

 made with the specimen presenting the full ventral view. 

 The length of body ( 1_) was measured from the apex of 

 the body, excluding the lists, to the most posterior part 

 of the body, excluding the spines and lists, which, in this 

 view, is near the dorsal part of the specimen. The di- 

 ameter of the body at the girdle (d) is self-evident. The 

 relative length is expressed by thej/d ratio. The height 

 of the epitheca (e) was measured from the apex of the 

 body to the proximal end of the girdle at its anterior 

 edge. This dimension is always expressed as a fraction 

 of the body length (e/1. ratio). The angle ;6 is the angle 

 which the left contour of the body makes with the longi- 

 tudinal body axis, A-B. This angle gives a measure of 

 the amovint of narrowing of the body posteriorly. 



Historical Review 



The genus was first reported by Stein (1883), who 

 figured C. horrida in several views. The larger plates 

 of the theca were distinctly shown and the attachments of 

 all the six spines correctly indicated. The pattern of the 

 major plates is easily established from Stein's figures. 



In the same year this species was described and 

 figured by Gourret (1883) under the name Dinophysis 

 i ourdanii . which later became attached to another spe- 

 cies, C. gourretii Paulsen (see p. 43). Schutt (1895) re- 

 ported a third species under the name Goniodoma acumi - 

 natum var. armatum ; and Cleve (1903) reported another 

 species under the name Goniodoma bipes . These last 

 three species were later ascribed to their proper genus 

 by Kofoid (1910), who appreciated the resemblances of 

 their plate patterns to that of Ceratocorys horrida Stein. 



Murray and Whitting (1899) figured, under Cerato - 

 corys spinifera . a species of Gonyaulax and a species of 

 Ceratocorys . The latter species was renamed C. magna 

 in the revision of the genus by Kofoid. This revision, 

 thus, included five species. We owe to Kofoid the proper 

 allocation of all these forms to the genus Ceratocorys . 

 He did not figure any of the species, however, and did not 

 analyze the sulcal plate pattern. 



Following Kofoid's (1910) revision, only one possi- 

 bly valid species has been introduced, viz., Ceratocorys 

 kofoidii Paulsen (1931, p. 36, fig. 22A-C). This species, 

 which is related to C. gourretii . was not found in the 

 Carnegie collections. Paulsen's sketches do not furnish 

 much more than the general body outline. He gives the 

 length as 115 microns. This exceeds the length of any 

 specimen of the genus Ceratocorys which the author has 

 ever measured, the longest being a specimen of C. arma- 

 ta, which measured 99 microns. This size can best be 

 appreciated if we contrast it with that of C. gourretii , 

 which is the smallest species of the genus, and has a 

 length of 49 (38-62) microns. In Paulsen's description 

 of C. kofoidii it is stated that that species is related to 

 C. gourretii. that it differs in the less rounded form. 



