h'J^ ' -s^ .0 sV HARVARD UNIVERSITY LIBRARY OF THE Museum of Comparative Zoology JAN 1 0 1929 MEMOIRS OF THE MUSEUM OF COMPARATIVE ZO(»LO(,'y AT HARVARD COLLEGE. VOL. LI. CAMBRIDGE, MASS., U. S. A. printeb for tbe /iDuseum. 1928. JAU 101923 MEMOIRS OF THE MUSEUM OF (JOMPARATIVE ZOOLOUY AT HARVARD COLLt:GE. VOL. LI. CAMBRIDGE, MASS., U. S. A. pvinteD fov the /IDuscum. 1928. printed by The Cosmos Press, Incorporated Cambridge, Mass. ADemotrs of tbe /IDuseum of Comparative Zoology AT HARVARD COLLEGE. Vol. LI. REPORTS ON THE SCIENTIFIC RESULTS OF THE EXPEDITION TO THE EAST- ERN TROPICAL PACIFIC, IN CHARGE OF ALEXANDER AGASSIZ, BY THE U. S. FISH COMMISSION STEAMER " ALBATROSS," FROM OCTOBER, 1904, TO MARCH, 1905, LIEUT.-COMMANDER L. M. GARRETT, U.S.N., COM- MANDING. XXXV. THE DINOFLAGELLATA: THE DINOPHYSOIDAE. By CHARLES ATWOOD KOFOID AND TAGE SKOGSBERG. WITH THIRTY-ONE PLATES. [Published by permission of Henry O'Mallet, U. S. Fish Commissioner]. CAMBRIDGE, U. S. A. IPrinteO for tbe flDuseum. December, 1928. CONTENTS. REPORTS on the scientific results of the expedition to the Eastern Tropical Pacific, in charge of Alexander Agassiz, by the U. S. Fish Commission Steamer "Albatross," from October, 1904, to March, 1905, Lieut. -Commander L. M. Garrett, U. S. N., commanding. XXXV. The Dinoflagellata; The Dinophysoidae. By Charles A. Kofoid and Tage Skogsberg. 766 pp., 31 plates. December, 1928. CONTENTS Page PART I. Introduction and Collections 13 Acknowledgments ............. 15 Methods of collecting ' .... 16 Distribution of collections 18 Examination of collections 19 Number of genera and species 21 Orthogenesis and convergence 23 Procedure used in the accounts of genera and species 26 PART II. Systematic Account 29 Classification 29 Dinophysoidae Kofoid, 1926 30 Diagnosis 30 Derivation. Subdivisions. Relationships among the families .... 30 1. Dinophysidae Stein, 1883 : 32 Diagnosis 32 Subdivisions. Relationships among the genera ...... 32 Distribution 35 Key to the genera 35 Heteroschisma, gen. nov. 36 Diagnosis 36 Distribution 36 H. aequale, sp. nov. 36 H. inaequale, sp. nov 38 Phalacroma Stein 40 Diagnosis 40 Organology 40 Reproduction . 51 Distribution ........... 53 Historical discussion and systematica 58 Adaptive and systematic value of the characters. Principles used in description of the species ........ 63 Subdivisions. Relationships among the species 66 Discussion of species groups 68 1. Contractum group 83 P. contractum, sp. nov. ........ 83 2. Rotundatum group 85 P. parvulum (Schiitt) Jorgensen 85 P. lativelatum, sp. nov. ........ 89 P. lens, sp. nov 91 P. porosum Kofoid & Michener 93 p. lenticula Kofoid 96 CONTENTS. Page 3. Argus group 98 P. porodictyum Stein .... 98 P. argus Stein ..... . 104 P. apicatum, .sp. uov. .... 111 P. circunicinctuin Kofoid & Michener 114 P. ovum Scliiitt lis P. pyriforme, sp. nov. . 122 4. Cuneus group . 124 P. cuneus Schiitt .... . 124 P. striatum Kofoid .... . i;ii P. gigantcuni Kofoid & Michener . 136 5. Rapa group . 139 P. rapa Stein . 139 P. favus Kofoid & ^Michener . 146 P. hindmarchi Murray & Wliitting . 150 6. E.xpulsum group .... . 1.5.-) P. protuherans, sp. nov. 155 P. e.xpulsum (Kofoid & Michener) . 157 7. Limbatum group .... . 161 P. Iinil)atum Kofoid & Michener . 162 P. bipartitum, sp. nov. . 166 P. pulchrum Kofoid & JNIichener . 168 S. Doryphorum group .... . 172 P. niucronatum, sp. now . 172 P. doryphorum Stein .... 175 P. circunisutum Karsten 182 P. cuneolus, sp. nov. .... 186 9. Praetextum group .... 188 P. praetextum Kofoid & Michener 189 10. Reticulum group .... 192 P. fimbriatum Kofoid & Michener 192 P. reticulatum Kofoid 195 P. turbineuin Kofoid & Michener 198 Dinofurcula, gen. no\- 201 Diagnosis 201 Distribution 202 D. ultima (Kofoid) 203 D. ventralis, sp. nov. 205 I )inophysis Ehrenberg . 207 Diagnosis .... 207 Organology 207 Reproduction 218 Distribution 219 Historical discussion and systematics 224 .\daptive and systematic value of the characters. Principles used n the description of the species .... 233 Sulxlivisions. Relationships among the species . 235 1. Acuta group 239 D. exigua, sp. nov. . . . . 239 CONTENTS. f. nov D. sphaerica Stein .... D. similis, sp. nov D. okamurai, sp. nov. D. fortii Pavillard .... D. norvegica Claparede & Lachmann [?] D. schrtideri Pavillard 2. Hastata group I), hastata Stein ..... D. uracantha Stein .... D. urceolus, sp. nov. .... D. monacantha, sp. nov. D. trapezium, sp. nov. D. swezyi, sp. nov. .... D. collaris Kofoid & Michener D. schiitti Murray & Whitting D. nias Karsten D. jnrgenseni, sp. nov. L). triacantha Kofoid .... 3. Caudata group D. caudata Saville-Kent D. caudata Saville-Kent f. acutiformis, noni Histiophysis, gen. nov Diagnosis ....... H. rugosa (Kofoid & Michener) 2. Amphisolenidae, fam. nov Diagnosis ........ Subdivisions. Relationships between the genera Distribution Key to the genera ...... Amphisolenia Stein Diagnosis Organology ...... Reproduction ...... Distribution Historical discussion ..... Adaptive and systematic value of the characters. P the description of the species . Subdivisions. Relationships among the species Key to the species of Amphisolenia . 1 . Species of uncertain or more or less isolated A. inflata Murray & Whitting A. laticincta Kofoid .... A. brevicauda Kofoid .... A. schauinslandi Lemmerniann A. rectangulata Kofoid A. astragalus Kofoid & Michener 2. Extensa group ..... A. extensa Kofoid .... A. elongata, sp. nov posi ncipli es US' ed ii Page 241 247 250 253 256 257 261 261 273 281 283 286 289 292 296 303 307 312 314 314 330 333 333 334 335 335 336 339 340 340 340 340 347 349 353 355 360 365 366 366 369 372 374 378 380 382 383 386 10 CONTENTS. Page 3. Globifera group 388 A. globifera Stein 388 A. quadrispina Kofoifl 393 4. Bispinosa group . . . 395 A. bispinosa Kofoid 395 A. curvata Kofoid 398 5. Schroderi group ...... 400 A. schroderi Kofoid 400 A. clavipes Kofoid 402 A. complanata, sp. nov 404 A. truncata Kofoid & Michener .... 406 6. Palmata group 409 A. bidentata Schroder 409 A. lemmermanni Kofoid ..... 419 A. pahnata Stein 422 .\. palaeotheroide.s Kofoid ..... 427 A. asymmetrica Kofoid 429 7. Bifurcata group 432 A. bifurcata Murray & \Vhitting 432 A. projecta Kofoid 436 A. thrinax Schiitt 438 A. quadricauda Kofoid & Miciieucr 444 A. quinquecauda Kofoid ..... 445 Triposolenia Kofoid ....... 446 Diagnosis ......... 446 Organology 447 Reproduction ........ 453 Distribution ........ 454 Historical discussion ....... 459 Adaptive and systematic value of the characters. Principles used in thf descriptions of the species ..... 45!) Subdivisions. Relationships between the subgenera and among th( species . ; 461 Key to the species of Triposolenia .... 462 Subgenus 1. Triposolenia, nom. subgen. nov. 463 T. truncata Kofoid 463 T. depressa Kofoid 468 T. bicornis Kofoid ...... 473 T. intermedia, sp. no\ . ..... 477 T. longicornis Kofoid ...... 479 T. fatula Kofoid 483 T. ambulatrix Kofoid 486 Subgenus 2. Ramiciformia Kofoid .... 48!) T. ramiciformis Kofoid 489 3. Ornithocercidae, fam. nov. ...... 494 Diagnosis 494 Subdivisions. Relationships among the genera 494 Distribution 495 Key to the genera 496 CONTENTS. 11 Ornithocercus Stein Diagnosis .... Organology .... Reproduction .... Distribvition .... Historical discussion and systematics Adaptive and systematic \-alue of the characters. Prir description of the species . . vSubdivisions. Relationships among the species Key to the species of Ornithocercus 1. Splendidus group .... 0. heteroporus Kofoid O. splendidus Schiitt .... 2. Magnificus group .... O. magnificus Stein, s. str. Schiitt O. thurni (Schmidt) .... 0. steini Schiitt, s. str. 0. orbiculatus Kofoid & Michener O. quadratus Schiitt .... O. carolinae Kofoid .... 3. Formosus group .... O. formosus Kofoid & Michener . Specimens of Ornithocercus of ([uestionable specific a Parahistioneis, gen. nov. .... Diagnosis Organology Distribution Historical discussion ..... Subdivisions. Relationships among the species Key to the species of Parahistioneis . 1 . Species of uncertain generic allocation P. rotundata (Kofoid & Michener) 2. Garretti group P. garretti (Kofoid) .... P. paraforniis, sp. nov. P. para (Murray & Whitting) 3. Reticulata group .... P. karsteni (Kofoid & Michener) P. reticulata (Kofoid) P. diomedeae (Kofoid & Michener) Histioneis Stein ...... Diagnosis ....... Organology Distribution Historical discussion and systematics Aflaptive and systematic value of the characters. Princ description of the species .... Subdivisions. Relationships among the species Key to the species of Histioneis . pies used : the I turn plesi in tht Page 496 496 496 .501 .503 .509 512 515 516 517 517 521 528 529 540 551 559 561 572 577 577 580 582 582 583 588 589 590 592 592 593 596 596 598 601 603 603 605 60S 611 611 611 625 629 631 634 645 12 CONTENTS. atefl speci es of soniewhiit 1. Primitive species and higlily clitt'erent isolated positions H. costata Kofoid & Michener H. pauiseni Kofoid H. inciinata Kofoid & Michener H. inornata Kofoid & Michener H. reginelhi Kofoid & Michener H. panaria, sp. nov. 2. Remora group H. elongata Kofoid & Michener H. carinata Kofoid H. navicula Kofdid 3. Biremis group H. biremis Stein . H. highleyi Murray & Whittiiig 4. Longicollis group H. longicollis Kofoid H. hyalina Kofoid & Michener H. pacifica, sp. nuv. 5. Pulchra group H. gubernans Schiitt H. striata Kofoid & Micliencr H. pulchra Kofoid H. niitchellana Murray & Wliitting H. panda Kofoid & Michener 6. Dolon group H. helenae Murray & Wliitting H. milneri Murray & Whitting H. dolon Murray & Whitting H. hippoperoides Kofoid & INIicl H. josephinae Kofoid 4. Citharistidae, fani. nov. . Diagnosis ...... Citharistes Stein .... Diagnosis ..... Distribution .... Historical discussion . C. regius Stein C. apsteini Schiitt . PART III. Distribution of I )inophvsoidae .\t the Stations of the Expeditio.v Bibliography Index. I. Names of systematic units higlier tluin species .... II. Specific and subspecific names ...... Page 647 647 650 652 654 656 659 661 6(il 663 667 669 669 673 676 677 679 681 683 684 684 686 690 694 696 696 697 698 701 704 707 707 707 707 708 708 709 712 717 739 759 760 I. INTRODUCTION AND COLLECTIONS This report deals with the pelagic Dinophysoidae, a tribe of the subclass Dinoflagellata, taken by the U. S. Fish Commission Steamer Albatross during an expedition to the Eastern Tropical Pacific, from October, 1904, to March, 1905, under the leadership of the late Alexander Agassiz. The Mastigophora or Flagellata are represented in the plankton of all seas, and especially in the tropics, by three main groups, called orders in Doflein's (1916) system, namely, the Chrysomonaduia, the Dinoflagellata, and the Cysto- flagellata. The first-named group is represented by two divergent families, the Silieoflagellidae and the Coccolithophoridae. Owmg to their minute size, the Silicoflagellidae are not taken in numbers in the net plankton even with the finest of silk bolting cloths. One must turn to the centrifuge or filter as an effective means of determining the relative abundance of these representatives of the nannoplankton. The natural filters of Salpa and other feeders upon the nanno- plankton also offer remarkalale resources for observing this group. The calca- reous internal skeletal elements of the Coccolithophoridae formed the coccoliths of the mythical "Bathybius" of the reputed primordial slime discovered by the Challenger Expedition. They are retained m calcareous oozes, but the flagel- lates which produce them, owing to their small sizes, must also be sought in the nannoplankton or in the digestive tracts of animals feeding thereon. The Dinoflagellata, on the other hand, are of much larger size and are to be found m great numbers in the plankton taken in silk nets, even as coarse as number 12. They are very numerous in most plankton collections, vying with the diatoms for preeminence in numbers and variety. In northern waters the species are less numerous, but the individuals are at times extraordinarily abundant, the total production massive, and their predominance extreme. In the tropics, on the other hand, the number of individuals of any one species is quite generally snmll, but the number of species is greatly increased and the processes of differentiation reach here their highest expressions. The Chrysomonadina and the Dinoflagellata are both included in the Phytomastigina, but the Dinoflagellata contain many holozoic genera and species, notably among the Gymnodinioidae. In this tribe also cell-organs such as 14 THE DINOPHYSOIDAE. ocelli, nematocysts, and tentacles have arisen. Because of these facts and also because of the size and differentiation within the group Kofoid and Swezy (1921) have treated the Dinoflagellata as a subclass coordinate with the Phytomastigina and the Zoomastigina. The third main group of Flagellata represented in the plankton of the ocean is the Cystoflagellata, containing the widely known genus Noctiluca. This, however, as Kofoid (1919) has shown, is only a highly specialized form of a gym- nodinioid dinoflagellate. Since the other representatives of the Cystoflagellata are probably either highly modified dinoflage Hates or radiolarians, this group should ultimately disappear from the scheme of classification of the Mastigophora. The Dinoflagellata are of special interest to the observer of marine life be- cause of their extraordinary power of luminescence. This appears upon stimu- lation by contact, shock, or chemical irritation, at night only, as shown by experiment. The extent to which luminescence is present among the various dinoflagellates is unknown. It occurs, however, in the Gymnodinioidae in Nocti- luca, in the Peridinioidae in Gonyaulax, Ceratium, Blepharocysta, and Peridin- ium. No critical records of this phenomenon in the Dinophysoidae are known to us, but its occurrence among them is highly probable. Much more remains to be done in relation to the occurrence, location, duration, cause, and nature of the light thus produced in Dinoflagellata generally. These flagellates are also the cause of the notorious outbreaks of "red water" in tropical and warm-tem- perate seas (Torrey, 1902, and Kofoid, 1911) m which the enormous numbers present give a reddish, brownish, or yellowish discoloration to the surface waters, often for long distances along shore, or in great areas of the off-shore tropical currents. The collections of the Expedition have been found to contain 132 species of Dinophysoidae, including 82 new species and four new genera. All but one of the known genera are represented or occur in the area traversed by the Expedition and 132 of the 198 known species, or 66.7%, occur in the Expedition material. The species not recorded are mainly those restricted to colder regions or they are \'ery rare tropical forms. The report upon the tribes Gymnodinioidae and the Peridinioidae of this Expedition is in the course of preparation. The Dinophysoidae are marine and widely distributed in the plankton of coastal waters and in the uppermost 300 fathoms of the illumined zone of the high seas in all latitudes. The species of this group are remarkable for their diversity, for their occasional brilliant coloring, and for the extraordinary evolu- tion of organs of flotation in the form of processes of the body or in the form of INTRODUCTION AND COLLECTIONS. 15 lists differentiated as parachutes, sails, wings, or rudders. The thecal wall or cuirass which forms the surface of the body retains throughout the whole group the structural simplicity in the matter of the numbers of its constituent plates found in the simplest and most primitive members. It is bivalved and each valve contains three elements, an epithecal, a cingular, and a hypothecal plate. In certain genera such as Dinofurcula, Amphisolenia, Triposolenia, and in a few of the more divergent tropical species of Dinophysis, the body, while retaining its bivalved simpUcity of component parts, undergoes a remarkable evolutionary flare of elongation, bifurcation, multiple branching, and distal twisting into asymmetrical form, all of which serve the function of flotation. In this assem- blage of species in which the body proper undergoes these extensive diversifica- tions in form, there is very little evolution of the lists of the girdle and the sulcus. Since the Dinophysoidae are predominantly, if not wholly, either photo- synthetic, or associated with commensal photosynthetic phaeosomes, they find their optimum conditions of life within the lighted zone in the upper levels of the sea. Adaptations to flotation and to hydrostatic adjustments thus loom large in their structural evolution. The structural expansions of the surface lend them- selves to regulatory control by resorption, especially at binary fission, and to extension by renewed outgrowth to meet diverse conditions of the environment. Acknowledgments No investigation involving so extensive collections, so prolonged search of the diversified constituents of the tropical marine plankton, so much drawing and sketching for purposes of record and comparisons, and the use and clerical analysis of so widelj' scattered a literature, can be brought to completion without the collaboration of several workers. During the Cruise the senior author made daily observations and notes on the plankton as collected, and throughout the work has continued in close collabo- ration in the microscopical, morphological, and systematic analysis of the mate- rial. He has resolved the structure of the individuals utilized in the elaborately finished drawings, and supervised the completion of the illustrations. He is responsible for the plan of the work, of the method of treatment of species, and jointly with Dr. Skogsberg for the morphological analysis and systematic arrange- ment of the text. The manuscript of Part II of the work has in the main been prepared by Dr. Skogsberg, with the continuous collaboration and joint analysis of all moot points with the senior author. 16 THE DINOPHYSOIDAE. The original drawings and records of occurrences based on the painstaking examination of the plankton collections are the work of Mrs. Michener from May 1905 to June 1908, and from June 1909 to July 1910. Her sketches have been utilized in the text figures and her detailed drawings, finished in pencil, have been transferred and prepared for reproduction on Ross board by Mr. A. B. Streedain. His skill in portraying contour and detail by this method has con- tributed much to the accuracy and beauty of the plates. We are indebted to the late Dr. Alice Robertson for assistance in organizing the multitudinous details of several of the genera. For grants in aid of the work we are also indebted primaril,y to Alexander Agassiz, who after the close of the Cruise continued to manifest a deep interest in the progress of the work; and to the Carnegie Institution of Washington, through the late Dr. Alfred G. Mayor, Director of the Department of Marine Biology, for a grant for assistance in the preparation of tlie manuscript. The work could not have been completed except for substantial grants made bj' the Board of Research of the University tif California for the past four years. Methods of Collecting The plankton collections were made, in the main, by the following methods : — Under normal conditions, except on the Manga Reva-Acapulco line, a collection was made each evening at 8 p.m. with number 12 and number 20 silk nets towed from the port boom on either side of the large 000 plankton net. The depth at which these nets fished was generally within the uppermost fathom. The nets were out for about twenty minutes. The same nets were used in the morning collections at 8 a.m., but at this time sufficient cable was payed out to lower them to a depth of 300 fathoms. They were towed for twenty minutes, steaming slowly, and were then hauled in. They were open and fishing during Ixith descent and ascent and the catch there- fore represented the plankton from all knels traversed by the nets, but mainly from the level at which they were towed for twenty minutes, since the total time of lowering and hoisting was not more than one third of the time of towing at the lower level. As a rule the number 12 and number 20 nets were attached to the cable when hj'drographic samples were taken at a depth of 800 fathoms. These catches are thus from a vertical column of water from this depth to the surface. On the line from San Francisco to Panama collections were made in vertical INTRODUCTIOX AND COLLECTIONS. 17 hauls with the number 12 and number 20 nets attached in series to a vertical bar lashed to the cable. This method was abandoned after leaving Panama. The plankton nets used in these collections were of the Kofoid (1898) type, of number 12 and number 20 Dufour silk bolting cloth respectively, fourteen inches in diameter at the top, five feet in length, with a brass bucket at the lower end. Two such nets may be cut readily from two yards of silk and are of a size convenient to handle. They have a surface for filtration of such proportions to the opening that the coefficient of filtration is low. These plankton collections made regularly in course of the routine of the Cruise were supplemented from time to time by wing-net catches taken in small nets (four inches in diameter) tied to the upper bars of the runners of the trawl when trawhng. These catches contain samples of plankton from bottom to surface, since they were so suspended as to fish during both descent and ascent as well as while trawling. An attempt was also made to utilize a silk filter connected with the ship's circulating system near the intake and running continuously during the Cruise. A number 20 silk net with a fourteen-inch opening was suspended in a vertical cylinder with suitable overflow for the filtered water. This net received a con- tinuous stream from a pipe tapped into the circulating system within twenty feet of the intake. It functioned satisfactorily for the first fortnight, but its efficiency was destroyed by the accumulation of attached plankton-feeding organisms, principally barnacles, as shown by their faeces and remnants, which grew along the pipe fine and so reduced the plankton in amount as to render the catches of little significance. Iron rust and the cjuick disintegration of the silk net in the warm waters of the tropical seas also added to the ineflficiency of this method. Catches were made by this method for over half of the Cruise, but their contents are meager and add nothing of value to the material. Local surface and a few short vertical hauls with the number 12 and number 20 nets were made in the harbors at Panama, Manga Reva, and Acapulco. Con- ditions at Callao, at Chatham Island in the Galapagos, and at Easter Island re- spectively were so similar to those of the environing ocean that local collections at these points were of no great significance. In the main these local collections added little to the material of the Dinophj-soidae, since they were remarkably poor in organisms of this group except for a few of the hardier types which were invariably found also in the coastal waters. As will be seen in the station lists our main resources were in the regular hauls from the surface and from 300-0 fathoms, especially in the latter, which, 18 THE DINOPHYSOIDAE. as a rule, have a wider range of species since they contain the contributions from practically all zones in which dinoflagellates normally live. The catches of the wing-nets from the greatest depth and those at the hydrographic stations from 800 fathoms contained no additional Dinophysoidae not found in hauls from 300 fathoms and no noticeable change in the proportionate representation of the species. Distribution of Collections The collections were taken at 127 stations on the Cruise. A list will be found in Part III, together with the pertinent data. A fuller accoimt appeared in the General Report of the Cruise by Mr. Agassiz (1906). Their distribution on the six lines of the Expedition is shown in Plate 24. Stations 4571 to 4627 are on the San Francisco-Panama line, 4631-4671 on the Panama-Callao line, 4673- 4692 on the Callao-Easter Island line, 4694-4716 on the Easter Island-Galapagos line, 4717-4739 on the Galapagos-Manga Reva line, and 4740-4746 on the Manga Reva-Acapulco line. The collections at regular stations number 225. Of these eighty-three are from the surface, twenty-four more from Salpa stomachs treated as from the surface, ninety-four from 300-0 fathoms, twenty from 800-0 fathoms, and four from various other depths to the surface. The surface hauls, those from Salpa, and those from deeper levels are distributed on the six lines of the Expedition as follows: — on the (first) San Francisco-Panama line, 19, 2, and 19; on the (second) Panama-Callao line, 21, 5, and 29; on the (third) Callao-Easter Island line, 11, 2, and 15; on the (fourth) Easter Island-Galapagos line, 13, 6, and 19; on the (fifth) Galapagos-Gambier fine, 9, 6, and 31 ; and on the (sixth) Gambier- Acapulco line, 10, 3, and 5, respectively. In addition to these from the surface, from Salpa, and from the deeper levels there were incidental plankton collections at the anchorages at Panama, off Easter Island, and in the harbor at Acapulco, Mexico. These are, however, utilized only incidentally and are not treated as regular stations of the Cruise. The few collections made with the Chun plankton closing net are also omitted as they add nothing of systematic significance to the Dinoflagellata from the other collections. The distribution of the pelagic stations in the several oceanic currents is shown in detail in the following table : — INTRODUCTION AND COLLECTIONS. 19 DISTRIBUTION OF PELAGIC STATIONS Total number Surface end Surface Total of Total of of stations vertical oniy vertical surface California Current 4 4 4 4 Mexican Current 13 1 8 5 9 Panamic Area 17 2 11 6 13 Peruvian Current 27' 15 5' 22 201 Easter Island Eddy 10 5 5 5 Galapagos Eddy 4 2 2 2 South Equatorial Drift 45 1 22 23 23 South Equatorial Current 3 2 1 2 Equatorial Counter Current 2 2 2 North Equatorial Current 2 2 2 127 23 59 68 82 ' One of these is from Salpa only. Examination of Collections The plankton catch, as soon as it arrived on board ship, was released from the bucket into glass sorting dishes, treated with chloretone for anaesthetizing the larger organisms, and promptly fixed in 10% formalin. The plankton con- sisted of the fine sedimented residues remaining after the larger organisms such as the Coelenterata, larger Copepoda, Amphipoda, pelagic MoUusca, Tunicata, and larval fishes were removed. Small amounts were occasionally preserved in various fixing agents. This Report is, however, based on the formalin material since formalin, far better than any other fluid, preserves and reveals the skeletal structures upon which in final analysis the systematic relationships are based, and in which individual variations are revealed. The collections were examined in the original sea-water and formalin without other treatment than replacing the loss by evaporation with distilled water in some cases. It is thus often possible to roll the organisms about under the cover- glass and to secure diverse views in many cases of the same individual, and some- times also to shatter it into its constituent plates when this is necessary. The method has the disadvantage of not making it possible to reexamine the same individual with certainty a second time. There is, unfortunately, no safe and certain method of sealing permanent microscopical mounts in formalin. When- ever it was desired to isolate some special individual organism, the material under the cover-slip was washed carefully into a vial instead of into the general catch. The type specimens and cotypes are thus preserved. After a sufficient knowledge of the more abundant species had been acquired, each catch was searched by use of the mechanical stage until no additional species 20 THE DIXOPHYSOIDAE. were detected. The lists of occurrences at the record stations were thus made up. The station hsts (Part III) incorporate these data as modified by all later more critical revision. In order to have at least a quasi-quantitative record of relative frequency in individuals of the component species of each catch, records were made during the search of each collection of the number of individuals seen of each species, up to the total of the first one hundred individuals seen. Thereafter each additional species detected at that station was merely recorded as "also present." The number of individuals recorded is thus the percentage frequency at that station. These numbers ha^■e been used in mapping the local distribution of the genera. The numbers are recorded in Roman numerals at the ends of the radii from the circles marking the locations of stations on the route (Plate 24). The records of relative frequency used in the discussions of the distribution of each species, thus refer solely to the relative numbers of the different species in the one catch and give no indication of relative numbers of the species in question in different catches. The data thus accumulated have involved certain difficulties and discrepan- cies, especially in those cases in which a species, originally conceived in the wider sense, was later in the preparation of the manuscript broken up into several species by the withdrawal of divergent groups. In all such cases only those individuals which had Ijeen drawn were perforce included in the groups thus segregated off. All others recorded in the original analysis wei'e left under the original specific name in the later restricted sense. The concept of each species in the inception of the work of necessity rested upon previously published figures or upon the characters of the first individual whose structures were analyzed and from which the first figures were made. As the allocating of the individuals to definite species progressed, our concept was widened or modified by the detection of variants from the figure of the type individual. To meet this condition and to make available for purposes of com- parison the structure of individuals of the same species from different localities, it became necessary to sketch the outline of the most apparent and easily de- termined parts of an ever-increasing number of individuals. This was especially true in most cases of the more abundant and generally more w idely varying com- plexes. These groups of sketches constituted the great mass of data on which the manuscript has been based, and they form the sources of the text figures which illustrate the range of \'ariability and the aberrant types which are included within our concept of the several species. INTRODUCTION AND COLLECTIONS. 21 The fact should be noted that these habitus sketches represent the range of variation observed and not the normal distribution within that range. The constant tendency quite naturally in the premises was to make graphic record of all the aberrant individuals e\'en though these constituted only a very small pro- portion of the total representatives of the species. It thus follows that the di- versity in the few individuals within the species is emphasized rather than the uniformity among the many. Number of Genera and Species The microplankton of the Expedition contains a total of 132 species of Dinophysoidae. No less than eighty-eight new species have been established in connection with this Report. Of these new species thirty-two were described by Kofoid (1907a), twenty-nine by Kofoid and Michener (1911), and twenty- seven are described in the present paper. Of these twenty-seven species, twenty- six were found in the material of the Expedition; one {Phalacroma paulscni, p. 60) was based on a figure published previously (Paulsen, 1911b, p. 305, fig. 2). The following is a list of the new species found in the material of the Expedition : Described in Kofoid {1907a) Phalacroma lemmermanni Ornithocercus lentirula palaeotheroides carolinae reticulatum projecta heteroporus striatum quadrispina Parahistioneis Dinophysis quinquecauda (: as Histioneis) triacantha rectangulata garret ti Amphisolenia sohroderi reticulata asymmetrica Dinofiircula Histioneis bispinosa (as Phalacroma) carinata brevicauda ultima josephinae clavipes Triposolenia longicollis curvata longicornis navicula extensa fatula paulseni laticincta ambulatrix pulchra In Kofoid and Michener {1911) Phalacroma Histiophysis karsteni circumcinctum (as Dinophysis) rotundata expulsum (as Dinophysis) rugosa Hii ?tioneis favus Am|5hisolenia costata fimbriatum astragalus elongata giganteum quadricauda hippoperoides limbatum truncata hyalina porosum Ornithocercus inclinata praetextum truncata inornata pulchrum orbiculatus panda turbineum Parahistioneis reginella Dinophysis (as Histioneis) striata coUaris diomedeae 22 THE DINOPHYSOIDAE. Described in this Report Heteroschisma aequale inaequale Phalacroma apicatum bipartitum contractum cuneolus lativelatum lens mucronatum protuberans pyriforme Dinofurcula ventralis Dinophysis exigiia jorgenseni monacantha okamurai similis swezyi trapezium urceolus Amphisolenia complanata elongata Triposolenia intermedia Parahistioneis paraformis Histioneis pacilica panaria Besides the new species mentioned above, the following new specific and subspecific names are introduced in this Report: — Dinophysis baltica (p. 229) for D. ovum Schiitt var. baltica Paulsen (1908) Dinophysis caudala f. acutijormis (p. 227) for D. homunculus var. ventricosa Pavillard (1916) Dinophysis reciirva (p. 228) for D. lenlicida Pavillard (1916) Dinophysis renijormis (p. 228) for D. pavillardi Schroder (1906a) As shown in the following table there are in all fourteen genera, of which five are new. Total iiiimber of Bpecific and sub- speciHr namefi estab- lished in litpralurp Number of valid species and subspecies Number of new 8perie.s and sub- species described in connection with this Report Total number of species and sub- species found by the Expedition Theradinium, gen. nov. 2 2 Pseudophalacroma 1 1 Oxyphysis 1 1 Heteroschisma, gen. nov. 2 2 2 2 Phalacroma 73 50 23 31 Dinofurcula, gen. nov. 2 2 2 2 Dinophysis 123 52 10 20 Histiophysis, gen. nov. 1 1 1 1 Amphisolenia 32 31 19 26 Triposolenia 9 9 4 8 Ornithocercus 16 8 4 9 Parahistioneis, gen. nov. 9 9 6 7 Histioneis 45 28 17 24 C'itharistes 2 2 2 Totals 31S 198 88 132 Of the five new genera, four — viz., Heteroschisma, Dinofurcula, Histiophy- sis, and Parahistioneis — are based on species found in the material of the Expedi- tion; one, Thccadinium (p. 32), was founded on two previously described species, Amphidinmm kofoidi var. pelasalwn Herdman (1922, p. 26) and Phalacroma ebriola Herdman (1924, p. 34). INTRODUCTION AND COLLECTIONS. 23 Up to April 1, 1926, there have been described 318 species and subspecies in the Dinophysoidae. Of these only 198 are regarded as valid. Of the 198 valid species 132 or 66.7% of the known species were found in the collections of the Expedition. The eighty-seven new species constitute 65.9% of the total number (132) found on the Expedition, and an increase of 80% to the 111 species previ- ously known. This has resulted from the large representation of the rich tropical fauna in the collections and from the intensive search of the richer plankton collections and of the contents of Salpa stomachs for the rare and often minuter species. The accompanying table records under each genus the number of species and subspecies in literature, the numbers of these which in our opinion, with some reservations, are to be considered valid, the numbers of new species established in connection with this Report, and the total numbers of species found in each of fourteen genera. Orthogenesis and Convergence The species are arranged within the genera according to their structural resemblances, upon the assumption that the degrees of structural resemblances are indications of commensurate degrees of genetic relationship. It is fully realized that such resemblances may, however, not be accurate quantitative measures of the degrees of genetic relationship, or unequivocal demonstrations of genetic . origins. Nevertheless they are the only evidences of the nature and results of the evolutionary process in this highly diversified group, and while their quantitative value may well be quite hypothetical, their qualitative significance is undoubtedly of high value in showing the method, direction, and trend of the evolutionary process in the group. The record is all the more definite and clear because of the fact that it is registered in the non-mobile exoskeleton or theca of the body and its outgrowths. While it is true that the more superficial structures such as lists, fins, and sails are subject to regulatory modifications and to considerable individual variation, we nevertheless find beneath such variation the more basic similarities and differ- ences, which may serve as guides in the establishment of genetically related groups. The groups of species thus established within the genera exhibit relationships in structure among themselves which make feasible their grouping in a dendritic, diverging fashion, suggestive of phylogenetic descent. Within the subgenera or other secondary groups of several of the genera, the known species, even though there be but a few, may usually be arranged in 24 THE DINOPHYSOIDAE. something of an orthogenetic fashion according to the degree of specialization of one or more structural features, in ascending series from the simpler to the more complex. These series, which are not always equally spaced, are ortho- genetic, since they appear to exhibit progressive (or regressive) movements in their evolution or speciation. The orthogenetic tendencies within these lines may be summarized as (1) increase in size; (2) increase in relative length, especially in Amphisolcnia and Triposolenia; (3) branching or extensions of the body, as in Amphisolcnia and Dinofurcula; (4) increase in the development of the reticulations on the surface of the theca, often by heavier ribbing and larger size of the reticulations; (5) ex- tension and ribbing of the anterior and posterior cLngular lists about the girdle culminating in Ornilhocercus splendidus and Histioneis josephinae; {6) extension and structural modifications of the right and left sulcal lists, as in Histioneis; and (7) the progressive elaboration of the phaeosome chamber in Ornithocercus, Histioneis, Parahistioneis, and Citharistes. The net result of our analysis of the process of speciation in the Dinophy- soidae has been a growing conviction that there is in this group a more or less orderly divergence from primitive species with a small, spheroidal body, with a smooth undifferentiated surface, low, simple cingular and sulcal lists, and no phaeosome chamber, toward the more elaborate types along a number of diver- gent lines. Furthermore the progress of this process of evolutionary divergence is marked to a surprising degree by an orthogenetic aspect. The evolutionary steps seem to be represented in living Dinophysoidae, at least to such a degree that the path presumably followed in the process is still indicated by them. It may be that they represent only laterals along this path, but these still-existing types have seemingly remamed near enough to their levels of divergence from the main paths to mark their courses. A striking feature is that the process of speciation has proceeded in this group to a marked degree of specialization in the seeming absence of sexual reproduc- tion. There is as yet no evidence of the occurrence of this mode of reproduction in the Dinophysoidae. Another striking feature of the evolution of the Dinoflagellata is the fact that there is no evidence of closely related species being isolated from each other either geographically or bathymctrically. Both the most primitive and the most diver- gent and highly specialized genera and species occur together in the same region, and very closely related forms are not infrequently recorded in the same surface catch or from the same Salpa stomach. Specific adaptations to zones of illumi- I IXTRODITTIOX AND COLLECTIONS. 25 nation and temperature undoubtedly exist, but di;u-nal migrations in heliotropic responses to the rhythm of solar energy and to other crsniic and meteorological factors may be expected to produce some vertical extension of the zones inhabited by each species, and to vary these extensions with changing conditions. Save for the larger differences due to latitude and depth, areal distribution affords little aid in the solution of evolutionary problems. Along broad lines the species are widely distributed in all seas in waters of comparable temperatures and illuminations. The ecologic niche which each fills is in reality a wide shelf girdling the tropical seas and extending to a considerable depth. Moreover, many of these areas of distribution appear to be to a large degree coincident, even within the species of a single orthogenetic line. The factor of geographical isolation, so significant among birds and mammals in which feeding range and sexual behavior so immediately affect the genetic processes and evolutionary progress, sinks into relative insignificance, if indeed it ever had any considerable part, in the process of speciation among the Dinophysoidae. It is also interesting to note that the most highly specialized genera such as Citharistes and Histioneis and the most highly specialized species within the genera, are, in the main, relatively rare in individuals. The development of adaptive structures of a highly specialized type is not in many instances accom- panied by a corresponding reproductive vigor or ec^ual enlargement of survival value. Attention may also be invited to another phenomenon of evolutionary significance which has emerged in the analysis of the Dinophysoidae, namely, the appearance in Heteroschisma and Pseudophalacroma of a structural character identical in location and morphological relations with a fundamental structural feature of the Peridinioidae. This character in Heteroschisma is the marking out by structural lines ofan area corresponding to plate 1 of the postcingular series of the latter tribe. The Peridinioidae differ from the Dinophysoidae in the facts that the skeletal elements of the former are {1} in five (or six) instead of three horizontal zones and (2) that the pre- and postcingular zones have five instead of two plates or elementary subdivisions. It is a striking fact that the postcingular zone of Heteroschisma has developed the structural features of the first plate of the post- cingular series next to the flagellar pore, sulcus, and blepharoplast (see Hall, 1925). This is illustrated in detail in the discussion of Heteroschisma acquale and inaequale (Plate 1). In the genus Pseudophalacroma there are evidences of a comparable emergence of a structural field homologous in morphological re la- 26 THE DINOPHYSOIDAE. tions with plate 1 of the precingular series and also located near the flagellar pore. Both of these genera are among the less specialized members of the tribe, although the species of Heteroschisma have a somewhat advanced development of the surface. The phenomenon here observed is of the same category as that described by Kofoid (1926) in Oxyphysis oxytoxoides, in which a rather highly specialized genus, Oxyphysis, of the Dinophysoidae, has the facies of Oxytoxum, of the Peridinioidae. In the cases of Heteroschisma ^and Pseudophalacroma the con- vergence of characters is found in a single one of the plates or unitary elements of the skeleton. In the case of Oxyphysis it appears in the external form of the body as determined by the total skeleton, the location of the girdle and sulcus, and the shaping of the two apices. These three instances of convergent features of structure between the Dinophysoidae and the Peridinioidae lead strongly to the inference that the structure of the species or larger systematic category, as well as that of the individual, may not wholly reveal its genetic potentialities. The Dinophysoidae appear to carry potentialities of producing in the course of their evolution, in these three genera only, so far as is known, certain structural features which belong primarily in the tribe Peridinioidae and are characteristic of it, in the case of the plate of Heteroschisma and Pseudophalacroma, or of a single genus in the tribe, as in the case of Oxyphj'sis. Pkocedure used in the Accounts of Genera and Species Nearly all the descriptions of genera, species, and subspecies published up to the present time are short and incomplete, with but few data concerning the nature and amplitude of variation. Ours is the first attempt at a more or less exhaustive treatment of the various systematic units. A consequence of this is that we have had to elaborate a new technique of description and treatment. Our basic principle is to keep everything that implies comparisons as stereotyped as possible in order to obtain the greatest possible perspicuity and ease of com- parison and reference. A usual shortcoming of current taxonomic papers is that the various systematic descriptions are not treated in a consistent arrangement and manner. Frequently it is necessary to traverse a mass of facts in order to find the one that is desired. In short papers this defect may be more or less inconsequential, in large monographs and in genera with numerous species it creates almost insuperable difficulties and an annoying loss of time. In the present paper each character has the same relative place in all of the diagnoses and descriptions and in so far as possible is discussed in the same phraseology. INTRODUCTION AND COLLECTIONS. 27 The families have been treated under the following headings: — (1) Diag- nosis; (^) Subdivisions. Relationships among the genera; (3) Distribution; (4) Key to the genera. In the case of the genera the following arrangement has been adopted: — (1) Synonyms; (2) Diagnosis; (3) Organology; (4) Reproduction; (5) Distribu- tion; (6) Historical discussion; (7) Adaptive and systematic value of the charac- ters. Principles used in the descriptions of the species; (5) Subdivisions. Rela- tionships among the species; {9) Key to the species. These are the subheadings under each species: — (/) Synonyms; {£) Diag- nosis; (S) Description; (4) Dimensions; (5) Variations; (6) Comparisons; (7) Synonymy; (8) Occurrence. On account of the fact that the various genera are quite decidedly different morphologically, it is impossible to apply the same procedure of description to all of them. The differences in descriptive technique are, however, not very great and inside each genus the treatment conforms to the type established. In the general account of most genera a diagrammatic figure elucidates our term- inology and methods of measuring proportions and angles. At the end of the description of each species information is given as to the number of specimens, the proportions of which were measured. With the ex- ception of the length of the body, all statements as to proportions and dimensions refer to these specimens only. The following random example : — "the posterior cingular list is 0.50 (0.49-0.53) the length of the body from the apex" should be understood as follows: — the figures inside the parenthesis indicate the range of variation; the figure in front of the parenthesis, the average. This average is not weighted and based on random sampling, but is the average of the drawings which are based on selected specimens, some of which were "typical," others more or less extreme and selected because of their departure from the norm. In describing curves: — gently, moderately, and strongly (or boldly) convex or concave indicates an increasing degree of convexity or concavity (quantita- tive) ; in order to express an increasing uniformity of curvature, the terms irregu- larly, somewhat irregularly, subuniformly, and uniformly convex or concave (qualitative) are used. In recording the species all the samples from the same depth are counted as a single record; for instance, if a species is found at one station in two or more samples from 300 fathoms to the surface, all these records are considered as one. In the accounts of the distribution of the species, according to previously pub- lished literature, the material is arranged geographically. Beginning with the 28 THE DIXOPHYSOIDAE. data from the northern Atlantic Ocean we proceed south to the Antarctic Ocean; then the data from the Mediterranean, proceeding from the west to the east; from the eastern Mediterranean we proceed through the Red Sea, the Gulf of Arabia, the Indian Ocean, Antarctic Ocean, Malay Archipelago, Australian waters, Chinese and Japanese waters, and finally in an easterly direction across the Pacific Ocean. J II. SYSTEMATIC ACCOUNT Classification Phylum 1. Protozoa Goldfuss Superclass 1. Plasmodroma Doflein Class 1. Mastigophora Diesing Subclass 1. Dinoflagellata Biitschli Order 1. Adiiiiferidea Kofoid and Swezy Tribe 1. Athecatoidae Kofoid and Swezy Tribe 2. Thecatoidae Kofoid and Swezy Order 2. Diniferidea Delage and Herouard emend. Tribe 3. Gymnodinioidae Poche emend. Tribe 4. Amphilothoidae Kofoid and Swezy Tribe 5. Peridinioidae Poche emend. Tribe 6. Dinophysoidae Kofoid Family 1 . Dinophysidae Stein emend. Genus 1. Thecadinium, gen. nov. 2. Pseudophalacroma Jorgensen 3. Oxyphysis Kofoid 4. Heteroschisma, gen. nov. 5. Phalacroma Stein 6. Dinofurcula, gen. nov. 7.^ Dinophysis Ehrenberg 8. Histiophysis, gen. nov. Family 2. Amphisolenidae, fam. nov. Genus 9. Amphisolenia Stein 10. Triposolenia Kofoid Family 3. Ornithocercidae, fam. nov. Genus 11. Ornithocercus Stein 12. Parahistioneis, gen. nov. 13. Histioneis Stein Family 4. Citharistidae, fam. nov. Genus 14. Citharistes Stein DINOPHYSOIDAE "der Dinophysiden" Stein, 1883, p. 23. Dinophysida Butschli, 1885, p. 1009. Dinophyseae Schutt, 1896, p. 20. Dinophysidae Doflein, 1909, p. 464; 1911, p. 530; 1910, p 436; Kofoid & Michener, 1911, p. 268; Lebouh, 1925, p. 75. Dinophysiaceae Pavill.^rd, 1916, p. 44; Jorgensen, 1923, p. 3. Dinophysaceae Oltmanns, 1922, p. 54. Dinophy.soidae Kofoid, 1926, p. 215. Diagnosis: — Dinoflagellata with well-developed girdle and sulcus, with permanent longitudinal and transverse flagella located respectively in sulcus and girdle, with a single flagellar pore located in sulcus, and with body surrounded by theca composed of two subequal halves joined approximately in sagittal plane and, as a rule, made up of but three plates, epithecal, cingular, and hypothecal. Girdle usually anterior in position. Exclusively marine. Derivation. Subdivisions. Relationships among the Families While several investigators have united the dinophysids and the peridiniids {e.g., Lebour, 1925), we have adopted Kofoid's (192G) decision that they form two separate tribes, viz., Dinophysoidae and Peridinioidae, and have done so on account of the fundamental structural differences and because of the fact that all available evidence appears to support the opinion expressed in the pedi- grees of the dinoflagellates constructed by Butschli (1885, p. 1016) and Kofoid and Swezy (1921, p. 84), viz., that these groups have evolved independently from the tribe Gymnodinioidae. In all the discussions of the hiterrelationships of the various taxonomic units of Dinophysoidae, we have (contrary to Kofoid and Swezy, 1921, p. 84) adopted the viewpoint that this tribe evolved from ancestral forms very closely resembling Gymnodiniuyn ovulum Kofoid and Swezy (1921, pi. 5, fig. 58). These ancestral forms might be characterized briefly in the following manner: — (1) body small and spheroidal, its length, depth, and width subequal; (:2) girdle subequatorial, its distal portion not displaced; (3) sulcus short on both epitheca and hypotheca; (4) cingular and sulcal lists small and structurally undifferen- tiated; (5) without accessory lists or sails. Small size and subrotund shape of body are features characteristic of the simply organized members of most of the genera of Dinophysoidae, e.g., Heteroschisma, Phalacroma, Dinophysis, Orni- thocercus, Parahistioneis, and Histioneis. It is an almost universal rule in this SYSTEMATIC ACCOUNT. 31 tribe that the increase in structural complexity is accompanied by an increase in size. The distal portion of the girdle is displaced only in one of the known species of Dinophysoidae, viz., in Dinofurcula ultima (Plate 5, fig. 4, 6). The shortness of the hypothecal sulcus is a characteristic of most genera, and its anterior prolongation occurs but seldom and only in genera (Pseudophalacroma, Oxyphysis) which show other indications of primitiveness. There is within most of the large genera a distinct tendency for the cingular and sulcal Usts to increase in size and structural complexity; and accessory lists and sails occur only in a few genera and very seldom in the simply organized members of these. The assumption that the girdle originally was equatorial in position is somewhat more uncertain than the others. In C5ymnodinioidae this structure has no fixed position but may be found anywhere between the apex and the antapex; in Amphidinium it is anterior, in Gymnodinium more or less equatorial, and in Torodinium posterior. In Dinophysoidae, on the other hand, it is either sub- equatorial or more or less anterior. The reasons for assuming the equatorial position as the most primitive in Dinophysoidae are as follows: — {1) the gu'dle is anterior in all the more or less highly differentiated families, viz., in Amphiso- lenidae, Ornithocercidae, and Citharistidae, and Histioneis, the most complex of all the genera of this tribe, is characterized by a smaller epitheca than any other genus; {2) in the Dinophysidae the epitheca is small in the two most differ- entiated genera, viz., Dinophysis and Histiophysis, and relatively large in most of the more or less primitive genera, Pseudophalacroma, Oxyphj-sis, Hetero- schisma, and Phalacroma. The onlj' strikmg exception to this rule is Theca- dinium, and this genus is of uncertain systematic position and probably did not develop from the same ancestral forms as the rest of the family, but from Axnphi- dinium-like forms. The genera of the tribe, Dinophysoidae are assigned to one of the four fami- lies, viz., Dinophysidae, Amphisolenidae, Ornithocercidae, and Citharistidae, all of which are either new or emended. As to the interrelationships of these four families but little can be said with any degree of certainty. The Dinophysidae comprises the structurally simplest members and presumably evolved directly from Gymnodinioidae. Amphisolenidae may have sprung from the same ancestral forms of Dinophysis which evolved into Dinophysis caudata and D. miles and which were characterized by the tendencies toward elongation, atten- uation, and posterior bifurcation. Ornithocercidae presumably evolved from some ancestral forms closely related to Phalacroma, Dinophysis, and Histiophysis. Citharistes approaches Dinophysis in the shape and size of the anterior cingular 32 THE DIXOPHYSOIDAE. and the sulcal lists. It differs strikingly from this genus in the differentiation of the dorsal portion of the girdle into the large phaeosome chamber. This chamber probably originated by a gradual increase in the dorsal width and concavity of the transverse furrow, such as can be seen in the genus Parahistioneis (cf. for instance, P. reticulata, Plate 19, fig. 7). 1. DINOPHYSIDAE Stein, 1883, emend. Diagnosis: — Body of diverse shapes; ratio between length and depth usually somewhere between 1.0:1 and 1.7:1. Epitheca small, disk-like to large, some- times even as large as hypotheca. Transverse furrow narrow and, as a rule, about as wide dorsally as ventrally; its dorsal width less than 0.30 the greatest depth of body. Cingular lists, when present, subhorizontal or inclined anteriorly, narrow or of moderate width, usually less than 0.30 the greatest depth of body. Phaeosomes occur only in one genus. Length of body, 21-148 fi. Marine, usually eupelagic, in all seas. Subdivisions. Relationships among the Genera Dinophysis Ehrenberg (1840a), the type genus of Dinophysidae, was the first genus of the family to be established. Later Phalacroma was introduced by Stein (1883), Pseudophalacroma by Jorgensen (1923), and Oxyphysis by Kofoid (192G); and in the present paper four new genera, viz., Heteroschisma, Dino- furcula, Histiophysis, anil Thecadinium, are described. Of these eight genera, Phalacroma and Dinophysis are by far the largest, the former comprising fortj'- seven, the latter forty-one presumably \alid species. Pseudophalacroma was founded on a single species previously assigned by Stein (1883) to Phalacroma. Oxyphj'sis also comprises but one species, 0. oxyioxoides Kofoid (192G). Of the four new genera, Heteroschisma is based on two new species; Dinofurcula on two species, one of which is new, the other previously described and figured by Kofoid (1907a) under the generic name of Phalacroma; and Histiophysis on one species previously (Kofoid and Michener, 1911) referred to Dinophysis. Theca- dinium comprises two species, viz., T. petasalum and T. ebriolum. Of these, the first was described and figured by Herdman (1922, p. 26) under the name of Amphidiinum kofoidi var. petasatu7n and later (Herdman, 1923, p. 34) trans- ferred to Phalacroma as P. kofoidi. The second species was introduced by Herdman (1924, p. 34) as Phalacroma ebriola. The members of Thecadiniinn are, on the whole, the most primitive represen- tatives of this family. Indeed, even their assignment to the tribe Dinophysoidae SYSTEMATIC ACCOUNT. 33 must be considered as uncertain. It is true that they are characterized by a distinct theca (Lebour, 1925, pi. 11, fig. Ic, d), but the thickness of the pelUcle in Gynmodinioidae is variable, and it appears probable that there are stages intermediate between unarmored forms with exceedingly thin and flexible pellicle and armored ones, i.e., forms in which the pellicle has become more resistant and inelastic. If the theca is left out of consideration, the two species of Thecadinium are structurally much closer to Amphidinium than to any of the genera of Dino- physoidae. They evidently lack cingular lists and the left sulcal list is very narrow or absent. Furthermore, their longitudinal furrow appears to be deep, while in all the other members of this tribe it is flat or but slightly impressed. The close structural similarity between Amphidinium and Thecadinium indicates that the latter genus evolved directly from the former. It may be noted in this connection that the anterior position of the girdle is considered as secondary; in other words, in this respect these two genera are more advanced than some of the highly differentiated members of Phalacroma (e.g., P. limhatum, Plate 3, fig. 5). Next to Thecadinium, Pseudophalacroma and Oxyphysis appear to be the most primitive. This assumption is based on the following structural features : — (1) the longitudinal furrow continues anteriorly beyond the anterior cingular list (Stein, 1883, pi. 18, fig. 1 ; Kofoid, 1926, pi. 18, fig. 1) ; this peculiarity is found in a great number of the representatives of Gymnodinioidae, but in no other mem- bers of Dinophysoidae (Jorgensen, 1923, p. 4); {2) the epitheca is comparatively high and deep; (3) the cingular lists are very narrow and lack structural differen- tiation; (4) the sulcal lists are small, and the left of them lacks the three ribs and the posterior angularity characteristic of most of the members of the higher genera, viz., Phalacroma and Dinophysis. Pseudophalacroma appears to be very close to the evolutionary line that led to the development of Phalacroma and Dinophysis. Oxyphysis, on the other hand, forms an aberrant and in many respects highly differentiated branch that split off at a very early stage. It embodies many of the tendencies inherent in the peculiar genus Oxytoxum uf the tribe Peridinioidae and furnishes one of the most striking examples of conver- gence known among the Protozoa (Kofoid, 1926). The genus Heteroschisma is characterized especially by two structural features: — {1) the entire left sulcal list belongs to the right \alve, its fission rib being located at the level of the posterior cingular list; (2) the left valve of the hypotheca has a triangular postcingular plate on the ventral side. The behavior of the left sulcal list in binary fission is not known in Thecadinium, Pseudophala- croma, and Oxyphysis. On the other hand, in the highly organized genera of 34 THE DINOPHYSOIDAE. Dinophysidae, viz., Phalacroma, Dinophysis, and Histiophysis, the fission rib of this Ust is displaced posteriorly; in other words, the anterior portion of this list belongs to the left valve and the posterior to the right. Although there are no proofs in support of this assumption, Heteroschisma is considered primitive in this respect. Except in Heteroschisma, a triangular postcingular plate is not known to occur in any of the members of this tribe (with the possible exception of Phalacroma fimhriatum, Plate 2, fig. 1). For these reasons and because of the rounded outline of the body, the large size of the epitheca, and the simple struc- ture of the left sulcal list, Heteroschisma is considered an independent evolu- tionary branch that split off at a very early stage. Furthermore, the entire organization of Heteroschisma, when compared with that of Thecadinium, Pseudophalacroma, Oxj'physis, and Phalacroma, indicates that this genus is much more closely related to Phalacroma and Pseudophalacroma than to The- cadinium and Oxyphysis. Dinofurcula recalls Phalacroma in the great depth of the epitheca and in the moderate width, subhorizontal position, and ribbed structure of the cingular lists. On the other hand, it differs very strikingly from this genus in having the hypo- theca bifurcate posteriorly. Another peculiarity of Dinofurcula ultima (Plate 5, fig. 4, 6) is the posterior displacement of the distal portion of the girdle, a feature not found in any other member of Dinophysoidae, but frequent in Gymno- dinioidae and Peridinioidae. Dinofurcula, like Pseudophalacroma and Hetero- schisma, represents an independent branch near the base of the evolutionary line that led to the differentiation of Phalacroma. Of the three remaining genera of Dinophysidae, viz., Phalacroma, Dinoph}'- sis, and Histiophysis, all of which are structurally very closely related, the first is, on the whole, the most primitive. This is indicated by the following features : — {1) the body is frequently subrotund in lateral outline; {2) the epitheca is very large and the girdle is subhorizontal not only in some of the primitive, but also in some of the highly advanced species; (S) the cingular lists are relatively narrow. Phalacroma and Dinophysis embody several common tendencies, e.g., the tend- ency to develop accessory sails and lists, and even their generic separation "seems ... to be somewhat arbitrary'" (Jdrgensen, 1923, p. 5). It is very difficult to allocate generically not only several of their primitive but also some of their advanced members. Histiophysis, which is structurally more closely related to Dinophysis than to Phalacroma, shows certain similarities to Parahistioneis. For instance, its epitheca is very small, its anterior cingular list is unusually high, has a pronounced funnel shape, and simplified ribbing, its posterior cingular list SYSTEMATIC ACCOUNT. 35 is reticulated instead of ribbed as in Phalacroma and Dinophysis, and it has phaeosomes in the girdle. The highly developed structure of the sulcal lists and the lack of a posterior main rib in the left sulcal list should also be taken into account when the systematic position of Histiophysis is under consideration. In short, of the eight genera, six, viz., Pseudophalacroma, Heteroschisma, Dinofurcula, Phalacroma, Dinophysis, and Histiophysis, are structurally quite closely related, while the two remaining, viz., Thecadinium and Oxyphj^sis, occupy rather isolated positions. Thecadinium is still on the borderland to Gymno- dinioidae, and Oxyphysis is a highly differentiated genus that branched off at a very early stage from the evolutionary line that led to the differentiation of the first six genera. Of Pseudophalacroma, Heteroschisma, and Dinofurcula, which also split off very early, the last is in some respects very aberrant, while the others are still, on the whole, quite primitive (see also Hensen, 1891, pi. 2). Distribution: — The Dinophysidae are very widely distributed, exclusively marine, and all the genera are eupelagic, except Thecadinium which is found on sandy beaches. Most of the species occur exclusivelj^ or almost so, in tropical, subtropical, and warm-temperate seas, but some of them are endogenetic in the Arctic and the Antarctic Oceans, where no representatives of the other families of this tribe are to be found. Most species appear to be rare, but some have been recorded as rather frequent. The optimum habitat is the lower levels of photo- synthesis. Key to the Genera 1. Cingular lists absent Thecadiniuin, gen. nov. 1. Cingular lists present 2. 2. Sulcus continues anteriorly beyond anterior cingular list 3. 2. Sulcus does not continue anteriorl}' beyond anterior cingular list 4. 3. Epitheca narrow, cone-shaped Oxyphysis Kofoid. 3. Epitheca rounded, dome-shaped Pseudophalacroma Jorgensen. 4. Body niolariforni in lateral outline Dinofurcula, gen. nov. 4. Body not molariform in lateral outline 5. 5. Entire left sulcal list belongs to right valve Heteroschisma, gen. nov. 5. Entire left sulcal list does not belong to right valve 6. 6. Posterior cingular list finely reticulated Histiophysis, gen. nov. 6. Posterior cingular hst not finely reticulated Dinophysis Ehrenberg and Phalacroma Stein. 36 THE DINOPHYSOIDAE. Heteroschisma, gen. nov. Diagnosis: — Body subcircular to broadly subobovate in lateral outline. Epitheca large, but slightly narrower than hypotheca. Transverse furrow nar- row, of subuniform width throughout, 0.43-0.70 the greatest height of epitheca. Posterior cingailar list 0.36-0.46 the length of body from apex. Left hypotheca characterized by a triangular postcingular plate, occupying its ventroanterior corner; this plate is somewhat deeper than high, and its ventral height is 1.5-3.0 the width of the transverse furrow. Cingular lists subhorizontal and subequal, slightly wider to slightly narrower than transverse furrow; without structural differentiation except dorsally and ventrally, where they may have a few ribs. Right sulcal list about 0.33-0.45 as long as body, with a maximum width 0.66- 0.75 the width of transverse furrow. The most important feature of the left sulcal hst is that it belongs to the left valve in its entire extension; i.e., that it crosses over from the left to the right valve at the le\'el of the posterior cingular list. Other characteristics of this list are: — it is 0.53-€.67 as long as body, of subuniform width throughout the greater portion of its length, its average width being subequal to width of transverse furrow, forms a round to ear-shaped lobe posteriorly, and either lacks structural differentiation or has but a single rib near the posterior end. Type. — Heteroschisma inaequale, sp. nov. Distribution: — Heteroschisma has never been found outside the area ui- vestigated by the Expedition; and by the Expedition it was taken at five stations only. Two of these five stations (4665, 4671) are in the Peruvian Current; one (4699) is in the Easter Island Eddy; and two (4701, 4721) are in the South Equatorial Drift. All records refer to hauls from 300-0 fathoms except one (Station 4699) which refers to a specimen from the stomach of a Salpa taken in surface waters. The frequency is less than 1 % at all the five record stations. Heteroschisma aequale, sp. nov. Plate 1, fig. 7, 8. P'igure 1 :1, 2 Diagnosis: — Body subcircular in lateral outline, deepest near the middle; in dorsoventral view obovate, widest at girdle. Posterior cing-ular list 0.45-0.46 the length of body from apex. Ventral margin of right sulcal list strongly sig- moid, concave anteriorly and convex posteriorly. Left sulcal list 0.53 the length of bod}', of subuniform width throughout the greater portion of its length; its I SYSTEMATIC ACCOUNT. 37 average width subequal to width of transverse furrow; postmargin of list pro- tracted into narrow, ear-shaped lobe ; without ribs or other structural differentia- tion.. Theca finely and faintly reticulate. Length, 43.2-51.0 m- Eastern tropical and subtropical Pacific. Description: — Body subcircular in lateral outline and deepest near the middle. The margins are confluent and subuniformly convex, and the epitheca, which is highest in the center, is about as broadly rounded as the posterior portion of the body. The longitudinal axis is about perpendicular to the girdle. The transverse furrow is flat or gently con\ex, and its width is about 0.43-0.44 the greatest height of the epitheca. The posterior cingular list is about 0.45-0.46 the length of the body from the apex. The ventral height of the triangular field is about 1.5 times the width of the transverse furrow. In dorsoventral view the body is obovate and widest at the girdle. Figure 1. — 1, 2, Heleroschifwa aequale, gen. et sp. nov. 1, right lateral view of type specimen, Station 4671 (300-0 fathoms); 2, ventral view, Station 4701 (300-0 fathoms); 3, Heteroschisma inaequale, gen. etsp. nov., right lateral view of type specimen, Station 4665 (300-0 fathoms), x 430. The cingular lists are subhorizontal and subequal, about as wide as the transverse furrow or slightly narrower or wider, and, as far as the records go, without structure. The right sulcal list is 0.33-0.40 as long as the body; its ventral margin is strongly sigmoid, concave anteriorlj', convex posteriorly; its maximum width is about 0.75 the width of the transverse furrow. The left sulcal list is about 0.53 the length of the body and of subuniform width throughout the greater portion of its length; its average width is subequal to the width of the transverse furrow ; its greatest width is located in or somewhat behind the middle ; its ventral margin is gently sigmoid, slightly concave or almost straight anteriorly, convex in the middle, and slightly concave posteriorly; its postmargin is pro- tracted into a narrow, ear-shaped lobe at an angle of about 50°; there are no ribs or other structural differentiations. The thecal wall is finely and faintly reticu- late; the meshes are of subuniform size and polygonal; pores were not seen. Only megacytic stages were recorded. Two specimens had exceedingly wide intercalary zones. 38 THE DINOPHYSOIDAE. Dimensions: — The dimensions of two specimens were : — Length of body, 43.2-51.0 /J (type, 51.0 yu). Total length of megacytic specimens, 54.0-55.4 m- Variations: — Judging by the few specimens available, this species appears to be quite constant in shape and structure and fairly variable in size. Co7nj)arisons: — The description of the shape of the body is somewhat un- certain, since all the specimens examined are megacytic with very broad inter- calary zones. The simplicity of the structure of the thecal wall seems to indicate that Heteroschisma aequale is more primiti\'e than H. inaequale. Occurrence: — Heteroschisma aequale is recorded at four of the 127 stations. There are 0, 1 , 0, 2, 1, and 0 stations on the six lines of the Expedition. Of these four stations, one (4671), the type locaUty, is in the Peruvian Current; one (4699) is in the Easter Island Eddy; and two (4701, 4721) are in the South Equatorial Drift. At one station (4699, Salpa stomach) the species was taken in surface waters. The remaining records refer to hauls from 300-0 fathoms. The temperature range of these four stations at the surface is 66°-75°; the average was 72°. At Station 4699 the surface temperature was 75°. The frequency in all the cases is less than 1 r/ Heteroschisma inaequale, sp. nov. Plate 1, fig. 1, 2. Figure 1 : 3 Diagnosis: — Body broadly subobovate in lateral outline, deepest at girdle. In dorsoventral view subbiconical, widest at girdle, side contours of epitheca and hypotheca almost straight or even slightly concave, except posteriorly where they are convex. Posterior cingiilar list 0.36 the length of body from apex. Ventral margin of right sulcal list almost straight anteriorly, gently convex posteriorly. Left sulcal list 0.67 the length of body, of subuniform width throughout; its average width subequal to width of transverse furrow; with rounded postero- ventral lobe; with short posterior rib that ends at middle of postmargin. Wall of left hypotheca, except triangular field, has fine and rather faint reticulation; remaining portion of theca with well-developed and rather wide-meshed reticu- lation. Length, 51.0 ;u. Eastern tropical Pacific. Description: — Body broadly subobovate in lateral outline and deepest at the girdle. The margins are subuniformly convex and confluent, and the epi- theca, A\lii(h is highest somewhat ventrally to the center, is somewhat more SYSTEMATIC ACCOUNT. 39 broadly rounded than the posterior portion of the body. The longitudinal axis is about perpendicular to the girdle. The transverse furrow is shghtly concave, and its width is about 0.70 the greatest height of the epitheca. The posterior cingular list is about 0.36 the length of the body from the apex. The ventral height of the triangular field is about three times the width of the transverse furrow. In dorsoventral view the body is subbiconical and widest at the girdle ; in the type specimen (megacytic) the ratio between the length and width is 1.20: 1 ; the side contours of the epitheca and hypotheca are almost straight or even shghtly concave, except at the posterior end of the body where they are more or less convex. The cingular lists are subhorizontal and subequal, about as wide as the trans- verse furrow or somewhat narrower, and without structural differentiation except dorsally and ventrally where a few (1-4) ribs are to be found. The right sulcal hst is about 0.45 as long as the body, of subuniform width throughout the greater portion of its length, and with a maximum width about 0.66 the width of the transverse furrow; its ventral margin is almost straight anteriorly and gently convex posteriorly. The left sulcal list is about 0.67 as long as the body, of subuniform width throughout its entire length, and its average width is subequal to the width of the transverse furrow ; it is characterized by a rounded postero- ventral lobe and has no structural differentiation, with the exception of a short posterior rib that ends at the middle of the postmargin of the list and has a pos- terior inclination of about 80°. The thecal wall of the right \'alve and of the epitheca, the transverse furrow, and the triangular field of the left valve have a well-developed reticulum of rather large polygonal meshes. The meshes of the epitheca and of the right hypotheca are subuniform in size; and about twenty meshes border the girdle posteriorly on the right valve. In the entire triangular field about sixteen meshes are to be found. The polygons of the transverse furrow are arranged in two rows; on each valve there are about sixteen or seventeen polygons in each row. Except in the triangular field the wall of the left hypotheca is characterized by a fine and rather faint reticulation resembling the one found in Heteroschisma aequale. Scattered pores are to be found. The only specimen examined was megacytic. Dimensions: — Length of body, 51.0 fi. Total length of megacytic specimen (including intercalary zone), 54.8 m- Comparisons: — The description of the shape of the body is somewhat uncertain since the only specimen examined is megacytic. 40 THE DIXOPHYSOIDAE. ' This species is easily distinguished from Heteroschisma nequale by the well- developed and comparatively wide-meshed reticulation of its theca. Occurrence: — Heleroschisma inaequale is recorded at only one (4665) of the 127 stations, on the second line of the Expedition, in the Peruvian Current, from a depth of 300-0 fathoms, and at a surface temperature of 68°. The frequency is less than 1% (one specimen). Phalacroma Stein Phalacroma Stein, 1883, p. 23. Butschli, 188.5, p. 940, 1009. DsL.'iGE & Herouard, 1896, p. 385. ScHtiTT, 1896, p. 26. Paulsen, 1908, p. 19. Lebodr, 1925, p. 75. Phalarocoma Daday, 1888, p. 99 {lapsus pennae). Phalacromo Nathansohn, 1908, p. 601 (typ. err.). Diagnosis: — Body usually subcircular, subellipsoidal, subobovate, subbi- conical, subcuneate, or fig-shaped in lateral outline, distinctly longer than deep (length: depth, 0.86-1.84: 1), and more or less compressed bilaterally. Epitheca usually large (ratio between its depth and depth of hypotheca in most species somewhere between 0.8:1 and 1.0:1, sometimes as low as 0.54: 1), either high, even as high as hypotheca, or more or less flattened. Transverse furrow narrow, of subuniform width throughout, flat or but slightly convex or concave; its width, 0.07-0.12 the length of body. Cingular lists usually subhorizontal and subequal, slightly wider to slightly narrower than transverse furrow, with or without ribs; sometimes somewhat inclined anteriorly. Right sulcal list usually somewhat narrower than the left and ends somewhere between Rj and R3 of left. Left sulcal list usually subtrapeziform, somewhat wider posteriorly than anteriorly, with pronounced posteroventral angle, and with three main ribs, one (Ri) in front, and one (R3) behind fission rib (R2) ; R2 somewhat behind girdle ; sometimes this list extends to antapex, sometimes its length is only 0.33 the length of hy- potheca; R . and R3 at most 0.22 and 0.55 the greatest depth of body, respectively ; margin may be rounded posteroventrally. Sometimes with accessory lists and sails. Type. — Phalacroma porodictyum Stein. Organology: — The body (theca) in Phalacroma is simple in structure, but diverse in shape. When seen laterally, it usually is somewhat asymmetrical, but symmetrical species also have been found {e.g., P. oinwi, Figure 11:4). In dorsal or ventral ^•iew it appears always to be syminetrical. In many species the longitudinal a.ris {a. I., Figure 29) has maintained its original position, i.e., it is perpendicular to the girdle; frequently, however, it is more or less deflected posteroventrally. In the two species with the most pronounced posteroventral SYSTEMATIC ACCOUNT. 41 deflection of this axis, viz., P. porodictyum and P. expulsum, an inclination of l°-20° and 5°-15°, respectively, was found. The only species in which this axis has a slight (2°) posterodorsal inclination is P. circumcinctum (Plate 1, fig. 5). WTien seen laterally, the hodij usually is distmctly longer than deep, only seldom is it deeper than long. In the most elongated species (P. iurbineum, Plate 2, fig. 3) the ratio between the length and the depth is 1.84: 1 ; in the species with the relatively deepest body [P. giganleum, Figure 14: 1) the corresponding ratio is 0.86: 1; in most species it is between 1.1: 1 and 1.3: 1. Sometimes the body is deepest in or near the middle, but usually the greatest depth has shifted to the anterior third of the body. As a rule, the body is deepest at or just behind the girdle. No species with the greatest depth located anterior to the girdle has been found. Usually the body, seen in lateral view, is circular or subcircular {Phalacroma lalivelalum, Figure 3:2, 3), ellipsoidal or subellipsoidal {P. parindiwi, Figure 3: 4-6), obovate or subobovate (P. circumsulum, P. doryphorum, Figure 23). Of these shapes the last two are by far the most common. Although the circular shape is the most primitive, it is found not only in some of the primitive but also in some of the highly differentiated members of the genus (Plate 3, fig. 5). The following shapes of body, seen laterally, have also been found : — subcircular, with a shght (P. lens, Figure 4:3) or striking (P. contractum, Figure 3: 1) con- striction at the girdle ; subcircular, with the anterior end subtruncate (P. porosiim, Plate 1, fig. 6; P. pulchrum, Plate 3, fig. 6); rounded subbiconical (P. praetexlum , Plate 4, fig. 6); subbiconical, with well-rounded apices (P. fimhriatum, Plate 4, fig. 4); biconical, top-shaped, narrowly to fairly broadly rounded anteriorly, subacute posteriorly (P. reticulatuni, Plate 4, fig. 3; P. turbineum, Plate 2, fig. 3); irregularly obovate, with strikingly conical epitheca, or sublozenge-shaped (P. apicatimi, Figure 10); sac-like, truncate anteriorly, with a dorsal shoulder-like constriction at the girdle (P. expulsum, Figure 20: 1) ; subobovate, with a broadly rounded angle at the posterior end of the left sulcal list (P. acutinn, Schiitt, 1895, pi. 3, fig. 17:1); cuneate, with a very broadly rounded epitheca (P. cuneus, Figure 12); subcuneate, with the posterior margin strikingly sigmoid, convex anteriorly, concave posteriorly (P. wrtim, Schiitt, 1895, pi. 4, fig. 18: 1) ; subcuneate, very broadly rounded to nearly flat anteriorly, narrowly rounded to subacute but not mammilliform posteriorly, and with the ventral margin strikingly angular at the posterior main rib of the left sulcal list (P. rapa, Figin-e 16); subcuneate, broadly rounded anteriorly, constricted, mammilliform posteriorly, and with the ventral margin more or less angular at the posterior main rib of the left sulcal list 42 THE DINOPHYSOIDAE. (P. favus, Plate 2, fig. 7); inverted fig-shaped (P. hindmarchi, Figure 18). A peculiarity worth mentioning is that species with ovoidal body have not been found as yet in Phalacroma, although this shape is fairly common in the closely related genus Dinophysis. Generally the body is more or less compressed bilaterally ; only a few species (e.g., Phalacroma globulus, Schiitt, 1895, pi. 2, fig. 12: 1; P.fimbriatum, Plate 4, fig. 4; P. giganieum, Plate 2, fig. 1) have the dorsoventral and transverse diameters subequal. It may also be mentioned that in the type specimen of P. turbineum (Plate 2, fig. 3; Plate 4, fig. 1) the transdiameter even exceeded the dorsoventral diameter, but this specimen possibly was megacytic (see p. 200). In the .species with the most pronounced bilateral compression (P. lens, Figure 4: 1-3) the ratio between the length and the transdiameter is 2.38-2.66: 1, while the ratio between the length and the dorsoventral diameter is 1.05-1.11: 1. When seen in dorsoventral view, the body generally is obovate (P. pulchrum, Plate 3, fig. 1), eUipsoidal (P. mucronatum, Figure 22:6), or lenticular (P. limbatum, Plate 3, fig. 3). In all the species with ellipsoidal or lenticular outline and in some of the species with obovate outline, the body is widest in or near the middle ; in the remaining species with obovate outline the greatest width is located in the anterior third of the body, often near the girdle. The following shapes of body, seen dorsoventrally, have also been found: — subcircular (P. globulus, 't^Qh.uii, 1895, pi. 2, fig. 12: 1, in which the body approaches the fundamental, spherical form); biconical, with more or less narrowly rounded apices (P. apicatum, Figure 10: 2; P. circumcincium , Plate 1, fig. 4); biconical, with acute or subacute apices (P. fimbriatum, Plate 2, fig, 1 ; P. reticulatum, Plate 4, fig. 5) ; cuneate, broadly rounded anteriorly, narrowly rounded or subacute posteriorly ; and with the side contours of the hypotheca straight, or evenly and gently convex or concave (P. cuneolus, Figure 22: 7), or more or less undulating (P. expulsum, Figiu-e 20: 2, 5; P. favus. Figure 14:2); inverted pyriform (P. pyrijorme, Figure 4:4); inverted fig-shaped (P. hindmarchi) . One of the most characteristic features of Phalacroma is the relatively large size of the epitheca (epi., Figure 29). In lateral as well as in dorsoventral view^ the epitheca usually is about as deep as or but slightly narrower than the hypo- theca. In most species the ratio between the depth of the epitheca and that of the hypotheca is somewhere between 0.8: 1 and 1.0: 1 ; in the species with the relatively smallest epitheca (P. expulsum, Figure 20: 1, 3) the corresponding ratio is 0.60 (0.54-0.66): 1. Only in exceptional cases (P. reticulatum, Plate 4, fig. 3) is the epitheca slightly deeper than the hypotheca. In the ancestral forms of SYSTEMATIC ACCOUNT. 43 this genus the epitheca and the hypotheca presumably were subequal in size, and this condition is still found in some of the recent species, e.g., in P. limhatum (Plate 3, fig. 5), P. praeiextum (Plate 4, fig. 6), and P. circumcindum (Plate 1, fig. 5). However, in most of the recent forms the height of the epitheca has been more or less reduced without a corresponding reduction in depth. In some species this reduction in height has proceeded so far that the epitheca has the shape of a large and but slightly vaulted disk {P. porosum, Plate 1, fig. 6; P. giganteum, Plate 2, fig. 2; Plate 3, fig. 2). Only in a few of the known species (P. expulsuw, Figure 20: 3) has this reduction in height been accompanied by a decided decrease in depth. Phalacroma expulsum, which in several respects is intermediate between Phalacroma and Dinophysis, is also the only representative of its genus in which the epitheca in lateral view is not at all or but slightly visible above the anterior cingular list. As a rule, the epitheca is highest in or near the center, and only in a few species has its greatest height shifted dorsally or ven- trally (Figiu-e 20). In some species {Phalacroma praetextum, Plate 4, fig. 6; P. limhatum, Plate 3, fig. 5) the transverse furrow {tr. /., Figure 29) has maintained its original equa- torial position, but usually it is distinctly nearer to the anterior than to the pos- terior end of the body. Sometimes (P. porosum, Plate 1, fig. 6) it is even so close to the apex that its width exceeds the greatest height of the epitheca. Its distal portion is not displaced posteriorly, i.e., it does not form a spiral about the body. It usually crosses the lateral faces of the body in an almost straight line, but in P. porosum (Plate 1, fig. 6) it forms a gentle, sigmoid ciu"ve. Its width is subject to but slight variations; in obovoidal and subcuneate species the width is about 0.07-0.09, in subrotund species about 0.11-0.12 the length of the body. Its floor usually is flat or but slightly concave or convex, and only in a few species (P. coutra.ctum, Figure 3:1) is the concavity pronounced; this character appears to be fairly variable even within the species. The longitudinal furrow usually is about half as long as the hypotheca but may be somewhat longer or shorter; e.g., in Phalacroma porosum (Plate 1, fig. 3, 6) it is about 0.63 and in P. turhineum (Plate 4, fig. 1) about 0.26 the length of the hypotheca. Anteriorly it does not extend beyond the girdle, i.e., into the epi- theca. It is at most but slightly impressed. The lists of the furroios are not excessively developed. They arise from low and narrow basal ridges (Plate 3, fig. 1, 6) and are hyaline and usually exceed- ingly delicate. In almost all the species the cingular lists {ant. cing. I., post. cing. I., Figure 44 THE DINOPHYSOIDAE. 29) are subhorizontal, subequal, of subuniform width throughout, and about as wide as or shghtly wider or narrower than the transverse furrow. The anterior appears usually to be closed, but may be open ventrally in some species (Plate 2, fig. 1; Plate 1, fig. 5, of Phalacroma fimbriatum and P. circumdnctuni) . The posterior is always open ventrally. Only in exceptional cases are these lists strikingly inclined anteriorlj^; in P. expulsum (Plate 6, fig. 1) and P. pulchrwn (Plate 3, fig. 6), which have this anterior incUnation better developed than any other members of the genus, these Usts are inclined at about 30°-45° and 20°-40°, respectively. It should be remembered, however, that the pressure of t he co\er- glass is likely to change the inclination of these lists. In the species with the widest cingular lists (P. protuherans, Figure 20:8), these are about 1.7-2.3 times wider than the transverse furrow; in P. canlraclum (Figure 3:1) and P. lenticula (Figure 3:7), which form the opposite extreme, they are only about half as wide as the transverse furrow. In the most primiti\e as well as in some of the highly developed species {P. circumcinclum , Plate 1, fig. 5; P. pnlchrum, Plate 3, fig. 6), the cingular lists appear to lack structural differentiation, except for one or two ribs dorsally and ventrally on each valve near the sagittal suture. In the remain- ing species both these lists are ribbed. The ribs usually are rather strong, straight, simple, and nearly equidistant; but they may be fairly irregular (P. striatum, Plate 2, fig. 5), or they may even anastomose into a reticulum (P. cuneiis, Schutt, 1895, pi. 3, fig. 14: 2; P. striatum, Jorgensen, 1923, fig. 12). The number of ribs in the anterior list is either subequal to or somewhat larger than that in the posterior. In some species (P. turhineum, Plate 2, fig. 3) these lists have only a few ribs, but in others the number is fairly large; for instance, in P. giganteum (Plate 2, fig. 2), each of these lists has, on each valve, as many as seventy-five ribs. The sidcal lists, which form a direct continuation of the posterior cingular list, run about parallel to each other on either side of the longitudinal furrow. While the entire right sulcal list belongs to the right valve, the anterior portion of the left list belongs to the left valve and the posterior portion to the right. The right sulcal list {r. sulc. I., Figure 29) is fairly \-ariable in size and shape even within the species. It always is smaller than the left. In some species {Phalacroma porosum, Plate 1, fig. 6; P. famts, Plate 2, fig. 7) it extends to or slightly beyond the posterior main rib of the left sulcal list; in others (P. striatum, Plate 2, fig. 8) it ends at a point somewhere between the fission rib and the poste- rior main rib of the last-mentioned Ust; only exceptionally (P. expulsum, Plate 5, fig. 1) does it end at or even in front of the fission rib. Usually, it either (P. limhatum, Plate 3, fig. 5) is subtriangular to rounded triangular, about as wide as SYSTEMATIC ACCOUNT. 45 the transverse furrow anteriorly, and decreasing gradually in width posteriorly; or [P. favus, Plate 2, fig. 7) of subuniform width throughout the greater portion of its length, fairly strongly rounded posteriorly, and about as wide as the trans- verse furrow anteriorly. The following shapes have also been found : — crescent shape, wider in the middle than anteriorly and posteriorly (P. porosum, Plate 1, fig. 6) ; the anterior half to two thirds of the free margin is more or less concave and forms a distinct angle with the posterior portion of the margin which is either straight iP. fimhriatum, Plate 4, fig. 4) or more or less convex (P. striatum, Plate 2, fig. 8) ; the free margin is sigmoid, concave anteriorly and con\'ex poste- riorly (P. giganteum, Plate 2, fig. 2). Usually this list lacks structural differentia- tion, but sometimes it has a marginal rib (P. striatum, Plate 2, fig. 8; P. argus, Figure 8: 1), a T-shaped rib (P. turhineum, Plate 4, fig. 1), or a more or less de- veloped reticulation (P. striaiurn, Plate 2, fig. 8; P. circurncinctum, Plate 1, fig. 5); the reticulation is always first developed along the base of the list. The ventral margin is always free. The left sulcal list (/. sulc. /., Figure 29) is also strikingly variable in size and shape. It usually is about half as long as the hypotheca or slightly longer, but sometimes {Plialacroma turbineum, Plate 2, fig. 3) it is only about 0.33-0.39 as long as the hypotheca, or (P. striatuin, Plate 2, fig. 8) it may extend as far as to the antapex. In most species it is subtrapeziform, somewhat wider posteriorly than anteriorly, with a pronounced postero ventral angle, and with the ventral and posterior margins straight, or gently concave, convex, or sigmoid; in some of the species of this category (P. porosum, Plate 1, fig. 6) the ventral margin may be very broadly angular near the middle. The posteroventral angle frequently is between 80° and 95°, but sometimes this portion of the Ust is more or less acuminate (P. bipartitum. Figure 21 : 2; P. circumsutum, Figure 23: 6), or it shows intermediate stages between an angular and a rounded outline. In some species (P. apicatum. Figure 10:4) the entire free margin of the list is almost evenly rounded, in others (P. argus. Figure 8:1) it is more or less sigmoid, concave an- teriorly and convex posteriorly. It also should be mentioned that in one species, viz., P. praetextum (Plate 4, fig. 6), this list is subtriangular, decreasing gradually in width posteriorlj'. In nearly all the species this list has three well-developed cross-ribs, which in the present paper have been named the anterior main rib, the fission rib, and the posterior main rib. The anterior main rib (a«/. m.r., Figure 29) is located anteriorly, at the junction of this Ust and the posterior cingular list; the fission rib (/. r., Figure 29), at the place where the list is divided in binary fission, i.e., at the sagittal suture and usually somewhat in front of the 46 THE DINOPHYSOIDAE. middle of the list; the posterior main rib {post. m. r., Figure 29), near the posterior end of the list. Usually the anterior main rib and the fission rib are subequal, or either of them is slightly longer than the other; and the posterior main rib, as a rule, is between 1.5 and 2.5 times longer than the fission rib. The relative length of the posterior main rib is more variable than that of either of the two anterior ribs. Sometimes {P. doryphonwi, Figure 23; P. ovum, Figure 11) this rib is as much as 2.8-3.5 times longer than the fission rib, sometimes (P. giganteum, Figure 14: 1 ; P. argus, Figure 8: 1, 2) it is more or less reduced in length or even absent (P. expulsum, Plate 6, fig. 1). When the posterior main rib is well de- veloped, the posteroventral margin of the list is angular, and the rib ends at or near the vertex of the angle ; when the length of this rib is more or less reduced or when this rib is absent, the posteroventral margin of the list is rounded (P. giganteum, Figure 14: 1) or almost straight (P. praetextum, Plate 4, fig. 6). When compared to the depth of the body, the maximum length of the fission rib and the posterior main rib is about 0.22: 1 and 0.55: 1 (P. circumsutum. Figure 23: 6). The minimum length of the fission rib, expressed in the same manner, is about 0.07: 1 (P. argus, Figure 8:1). The main ribs always are unbranched, straight or but slightly curved, and usually of moderate strength, and tapering distally. The fission rib frequently (always?) is double, i.e., there is one rib on either side of the suture (P. porosum, Plate 1, fig. 3, G); and it may be T-shaped, in which case it is divided distally in two branches, which form a marginal rib along the anterior half of the list (P. circumcinctum , Plate 1, fig. 5). The posterior main rib, which sometimes is more or less club-shaped distally (P. doryphorum, Figure 23: 1 ; P. hindmarchi. Figure 18:3), generally has a posterior inclination of about 45° and may be directed straight back (P. slriatmn, Plate 2, fig. 8). In all the primitive and in many of the more or less highly differentiated species this list appears to lack structural differentiation except for the main ribs; in others it may have a number of short, secondary ribs (P. f aims, Plate 2, fig. 7), or a more or less de- veloped reticulation (P. striatuvi, Plate 2, fig. 8; P. hindmarchi. Figure 18:3). It also should be mentioned that in P. Kmbaturn (Plate 3, fig. 5) and P. pulchrum (Jorgensen, 1923, fig. 18) this list has a fine rib just behind the posterior main rib that extends to the margin of the list. The left sulcal list of P. turhineum (Plate 2, fig. 3) is of an exceptional type; its length is, as previously mentioned, only 0.38-0.39 the length of the hypotheca; and its greatest width is 0.12-0.13 the greatest depth of the body; its free margin is gently and evenly convex; it has a weak cross-rib somewhat in front of its middle and a submarginal rib along its entire length; with the exception of these two ribs, this list appears to lack struc- SYSTEMATIC ACCOUNT. 47 tural differentiation. With regard to this list in P. fimbriatum, see p. 192. In most species this hst ends a very short distance behind the posterior rib, but in some {P. f aims, Figure 14: 4, 5) it is more or less decurrent. The last condition is, in a sense, an initial step in the development of a parasagittal list (see P. cuneolus, Figure 22:1-3). The primitive species have no lists except the cingular and sulcal lists. On the other hand, in many of the more or less highly evolved forms, accessory lists are developed. As an example of a highly differentiated and, if we may say so, almost schematical system of accessory lists, these lists in Phalacroma limbatum (Plate 3, fig. 3, 5) may first be described. In this species two lists, the para- sagittal lists {par. I., Figure 29), encircle the body, one on either side of the sagittal suture ; however, in the transverse furrow parasagittal lists never are developed, and the left list ends at or near the posterior end of the longitudinal furrow. The left parasagittal list is of subuniform width throughout its entu-e length, has a maximum width which is subequal to or somewhat less than the width of the transverse furrow, and is furnished with a moderate number of simple, free or anastomosing, incomplete riblets. The right parasagittal list, which directly continues the left sulcal list, resembles the left, but forms on the antapex an acute, wedge-shaped process, the posterior sail (post, s., Figure 29), directed posteriorly. When fully developed, the posterior sail is 0.15-0.27 the greatest depth of the body in length and has a central rib, the proximal half of which might form a more or less complex reticulum; the angle at its tip is 35°-50°. Of this fundamental system there are a great number of variants, the most important of which will now be described. In P. fimbriatum (Plate 4, fig. 4; Plate 2, fig. 1) the two parasagittal lists also encircle the whole body. On the epitheca both lists are of about the same width and structure; from the apex, where their width about equals 0.09 the greatest depth of the body, they slightly decrease in width toward the girdle ; each of them has about twenty almost equidistant ribs, a few of which are branched. On the hypotheca the left list is very narrow, at most about half as wide as the transverse furrow. The right is much wider than the left, lacks a posterior sail, and merges so completely in the left sulcal list that the boundary between these two lists cannot be estabUshed with certainty. From the posterior main rib of the left sulcal list to the antapex, the right parasagittal list is of nearly subuniform width, about 0.14-0.20 the greatest depth of the body; anterodorsally to the antapex this list gradually becomes narrower, and its average width on the dorsal side of the hypotheca is only about 0.07 the greatest depth of the body. The right parasagittal Ust of the hypotheca has about 48 THE DINOPHYSOIDAE. twenty-six fairly strong and almost equidistant ribs, most of which are simple and almost straight, a few irregulai- or iri-egular and branched. The structure of the left list is unknown. As a last example of a species with the two para- sagittal lists extending around the whole body, P. turhineum (Plate 2, fig. 3; Plate 4, fig. 1, 2) may be mentioned. In this species the right parasagittal list forms a direct continuation of the left sulcal list and lacks a posterior sail. On the average both the parasagittal lists are about twice as wide on the posterior half of the body as on the anterior half, and their maximum width is about 0.08- 0.09 the greatest depth of the body. Each of them has a moderate number of short, simple ribs, approximately six on the epitheca and eighteen on the hypo- theca. In most species the parasagittal hsts are less developed than in the examples mentioned above. Generally speaking, the phylogenetic development of these Usts appears to have proceeded from the posterior end of the body to the anterior, and the development of the right list seems to have preceded that of the left. In order to illustrate this statement, a few examples will be given. P. pulchrum (Plate 3, fig. 6) has two parasagittal lists on the hypotheca, but none on the epitheca. The left one of these two lists extends, on the dorsal side of the body, from the posterior cingular list to the antapex; it is very narrow, its maxi- mum width being less than the width of the transverse furrow ; and it is furnished with a few cross-ribs; in some specimens it may even be absent. The right para- sagittal list forms a direct continuation of the left sulcal list and either extends to the posterior cingular list or is absent from the dorsal side of the hypotheca; dorsally it is of about the same type as the left list; posteriorly and postero- ventrally of about the same type as the corresponding list in P. Umbatuin, onlj^ as a rule, somewhat wider and with a slightly different structural differentiation. In P. bipartitum (Figure 21:2) only the right parasagittal hst is present; it is restricted to the posterior half of the hypotheca, but still is connected with the left sulcal list. A peculiar feature of this species is that there are two posterior sails of about the same type as the corresponding structures in P. limbatum, one on each side of the antapex. In a fairly great number of species (P. doryphorum, Figiu'e 23) only the posterior sail of the right parasagittal list is developed. This sail is subtriangular or wedge-shaped and is either situated on the antapex and directed posteriorly or slightly displaced ventrally and inclined posteroventrall}'. When fully developed, its length usually is about 0.20-0.40 the greatest depth of the body, and it is about as wide at the base as it is long or more or less decidedly narrower {P. pugiunculus, Jorgensen, 1923, fig. 19). Sometimes this sail has a central rib, sometimes it is more or less reticulated, especially in its central SYSTEMATIC ACCOUNT. 49 portion; sometimes both the central rib and the reticulation are developed; sometimes no structure can be distinguished. Only in a few of these species (P. circumsutum, Figure 23:6; P. cuneolus, Figure 22: 1, 2) a connection is de- veloped between the posterior sail and the left sulcal list. The flagellar pore (/. p., Figure 29) is a fairly large opening^ located in the longitudinal furrow, i.e., on the right valve, and just behind or about a girdle width behind the posterior cingular list. Its shape usually is subcircular or slightly elongated, in exceptional cases subtriangular (Phakicroma porosum, Plate 1, fig. 3) or slightly irregular. It is variable in size, but its greatest diame- ter appears never to exceed the width of the transverse furrow. Other pores to be mentioned in this connection are as follows: — in P. praetextum (Plate 4, fig. 7) there is a fairly small pore on the ventral side of the left valve, near the sagittal suture, and just behind the anterior cingular list. In P. pulchrum, P. giganteum, and P. striatum (Plate 3, fig. 1, 2; Plate 2, fig. 5) there are one to three fairly small pores on the ventral side of the left \alve, near the sagittal suture, and just in front of the anterior cingular list. Although pores of the last-men- tioned kind have been observed only in a comparatively small number of species, it is not impossible that they are characteristic of the genus as a whole. Their significance is not apparent. "Boih flagella arise in the flagellar pore and are well developed. The trans- verse flagellum (Schiitt, 1895, pi. 2, fig. 11 : 2) lies in the girdle and encircles the body from the left around to the right. The longitudinal flagellum (Schiitt, 1895, pi. 2, fig. 13: 6), which is about as long as the body or slightly longer or shorter, passes posteriorly between the sulcal lists. The structure of the thecal mdl is very different in the various species and affords important systematic characteristics. In most species there is a varying number of pores and poroids (Schiitt, 1895, p. 21, 22) moi-e or less uniformly scattered all over the theca. Sometimes (Phalacroma fimhriatum, Plate 4, fig. 4) these structures are numerous, sometimes (P. ttirhineum, Plate 4, fig. 1, 2) they are few, sometimes (P. praetextum, Plate 4, fig. 0, 7) they appear even to be absent. In P. lurhineum they seem to be restricted to the transverse furrow and only about fifteen in number. In some species the theca appears to lack struc- tural differentiation except for the pores and poroids, but usually it is charac- terized either by areolation or by reticulation. The areoles, which are small, rounded pits and usually more or less closely set (P. circumcinctum, Plate 1, fig. 5), resemble the poroids, with which they are connected by transitional structures (Schiitt, 1895, p. 22). The meshes of the reticulation usually are 50 THE DINOPHYSOIDAE. fairly uniform in size within the species, but show a wide range of variation within the genus as a whole. In P. striatum (Plate 2, fig. 8), which is charac- terized by meshes of moderate size, twenty-five to thirty-five meshes on each valve border the posterior margin of the transverse furrow; in P. f aims (Plate 2, fig. 7) the corresponding value is twenty to twenty-five ; in P. praetextum (Plate 4, fig. 6), about fifteen; in P. fimbriatum (Plate 4, fig. 4), about ten; and in P. turhineurn (Plate 2, fig. 3), five. In the last species the meshes are relatively larger than in any other known species. The ridges between the meshes some- times are so fine that the reticulation hardly can be distinguished, sometimes (P. fimbriatum, Plate 4, fig. 4) they are rather heavy. It should be remembered in this connection that the degree of structural differentiation varies with age. The new valve of a recently divided specimen is thin and almost without struc- ture; in old specimens the thecal wall is more or less heavy and its structural peculiarities are well developed. In most species the areolation or reticulation covers the whole thecal surface, in others {P. limbatum, Plate 3, fig. 4, 5) the regions nearest to the sagittal suture remain more or less undifferentiated. In P. cuneus (Figure 12) and related species a spine-like process projects into the cytoplasm from a point near the middle of the right sulcal list. P. praetextum (Plate 4, fig. 6, 7) is especially characterized by a dumbbell-shaped area, located on the ventral side of the hypotheca, just to the left of the sulcus, and extending on both sides of the sagittal suture from the posterior cingular list to the antapex. This area, which has a maximum width of about 0.33 the dorsoventral diameter of the body, is characterized by a structure quite different from that of the remaining portion of the theca. It has a faint reticulum of very small meshes, and in the middle of each or at any rate of most of these meshes there is an ex- ceedingly fine pore. In other words, this area has the structure of a cribriform plate. Presumably it is a permanent structure; i.e., it appears not to correspond to the intercalary zones of megacytic specimens (see p. 189). With regard to the shape and structure of the intercalary zones of the mega- cytic specimens, see p. 51. The sagittal suture of the two valves appears always to be finely serrated (Plate 2, fig. 1, 4). The three plates, epithecal, cingular, and hypothecal, of each valve very seldom are found separated. In one specimen of Phalacroma giganteum, the epithecal was detached from the cingular by means of the pressure of the cover-glass. The protoplasmic contents are coarsely alveolar and hyaline, or may be of a pale rose or brownish color. The nucleus, which usually is located on the dorsal SYSTEMATIC ACCOUNT. 51 side of the hypotheca (Schiitt, 1895, pi. 2, fig. 11 : 1, 2; 13: 3, 6), is rather large, ellipsoidal, ovoidal, or spherical, and has a moniliform chromatin reticulum. Probably in all species there are one or two, usually two, large pusules, which open by short and slender canals into the flagellar pore (Schiitt, 1895, pi. 2, fig. 11:2). In many species chromatophores are absent ; when present, these structures appear usually to be small, rod-shaped, and yellowish (Schiitt, 1895, p. 64, pi. 3, fig. 16: 4; 17: 2). A characteristic feature of many species is the presence of rod-like or thread-like rhabdo&omes, which in most cases are arranged radially (Schiitt, 1895, pi. 2, fig. 11: 2; 13: 3). Metaplasmic inclusions of different kinds, e.g., globules of fats and oils, also have been found. See also Schiitt (1895) and Biitschli (1885). The cell sometimes {Phalacroma pidchdhi, Lebour, 1922, p. 817) is "full of large refractive bodies." The length of the body is very variable within the species as well as within the genus. For the genus as a whole the range of variation in length thus far estab- lished is from 21 ix {Phalacroma pulchellum) to 148 /j. {P. giganteum). Reproduction: — The only mode of reproduction known in Phalacrcma is binary fission of the free-swimming individuals. The line of fission or, in other words, the sagittal suture lies in the sagittal plane except in the region of the sulcus. Probably in most species the whole flagellar pore and the entire longi- tudinal furrow belong to the right valve (Schiitt, 1895, pi. 3, fig. 16:3). It is possible, however, that sometimes the portion of the longitudinal furrow that is in front of this pore belongs to the left valve, as in the specimen of Dinophysis jorgenseni (Plate 5, fig. 3). Since in all species the flagellar pore is located near the junction of the sulcus and the cingulum, by far the larger poi'ticn of the longitudinal furrow always belongs to the right valve. The whole right sulcal list belongs to the right valve. The left sulcal list is divided at the fission rib, which probably is always double (Plate 1, fig. G) although, when seen in lateral view, it usually appears to be single. The portion of this list which is in front of the fission rib belongs to the left valve, the portion behind this rib to the right valve. The right parasagittal list and the posterior sail or sails belong exclu- sively to the right valve, the left parasagittal list exclusively to the left valve. Neither the cingular and sulcal lists nor the parasagittal lists and the pos- terior sails are resorbed in binary fission (Schiitt, 1895, pi. 2, fig. 10). Whether or not the flagella (of the right valve) are shed is not known. The flagella of the left valve are fairly well developed before the detachment of the two schizonts (Schiitt, 1895, pi. 3, fig. 16: 4), and so the dividing specimens probably continue their active swimming during the fission processes. 52 THE DINOPHYSOIDAE. The growth of the cingular and sulcal lists and the development of the thecal wall in the new halves of the two schizonts probably take place fairly rapidly. This is indicated by the fact that one rather seldom finds specimens in which these structures are not completely developed. Furthermore, judging by Schiitt's figure (1895, pi. 3, fig. 16: 3), the cingular lists are fairly well developed even before the schizonts separate; see, however, Meunier (1910, p. 59). On the other hand, the development of the parasagittal lists and of the posterior sails, which pre- sumably are phylogenetically rather young structures, appears to be rather slow. In any case, the posterior sails of the specimens of P/io/ocrowa doryphorum, etc., examined were found to be very variable in size, and in species with parasagittal lists many specimens lacked these structures or had them \'erj' incompletely developed. For a better understanding of the value of the last-mentioned struc- tures in the differentiation of species, their post-fission development should be submitted to a careful study. Double specimens or groups of specimens resulting from fission or repeated fissions, such as have been found in Dinophysis caudata and D. rniles, are unknown in Phalacroma. A phenomenon that possibly may be characteristic of the genus as a whole (Pavillard, 1916, p. 47) is the pronounced increase in size that precedes binary fission and that is accompanied by the development of the so-called intercalary bands (Stein, 1883, p. 23) along the sagittal margins of the two valves. Speci- mens with intercalary zones or, according to Pavillard's (1915a) terminology, megacytic specunens, have been found in twenty-one of the forty-seven presum- ably valid species of this genus, viz., Phalacroma acutum (by Pavillard, 1916), P. argus (by Jorgensen, 1923), P. cuneolus (by us), P. cuneus (by Pavillard, 1915a, 1916; and by us), P. doryphoruvi (by Pavillard, 1915a, 1916; Jorgensen, 1913; and by us), P.j'avus (by Jorgensen, 1923), P.fimbriatum (by us), P. giganteurn (by us), P. globulus (by Schiitt, 1895), P. hindinarchi (by us), P. mitra (by Pavillard, 1915a, 1916), P. mucronatum (by us), P. otmm (by us), P. paulseni (by Paulsen, 1911b), P. porodictyum (by Stein, 1883; Schiitt, 1895), P. porosum (by us), P. pulchellum (by Lebour, 1922), P. rapa (by Stein, 1883; Pavillard, 1915a, 1916), P. rotundaium (by Meunier, 1910), P. rudgei (by Murray and ^\^litting, 1899), and P. vastuni (by Schiitt, 1895) ; see also p. 198. The intercalary zone, which appears to be of varying width in the different species, is relatively wide on the dorsal side of the body and decreases gradually in width toward the ventral side; in the region of the sulcus it appears usually to be very narrow or not developed at all. Sometimes, as in Phalacroma mitra SYSTEMATIC ACCOUNT. 53 (Pavillard, 1916, p. 49), it has about the same structure as the thecal valves. In most cases, however, its structure is strikinglj' faint. In some species, e.g., in P. doryphorum (Pavillard, 1916, fig. 12), this zone under low magnification ap- pears to have closely set fine transverse striation; under higher magnifications each of the striae proves to be a row of closelj' set minute areoles. The mtercalary bands are transitory structures. When the new vahe is formed in binary fission its sagittal margin is attached to the premegacytic sagittal margin of the old ^'ah'e, and the intercalary band soon disappears; pre- sumably it is resorbed. For further discussion of this interesting phenomenon, see Pa\'illard (1915a, 1916), who was the first to submit it to a careful study. Finally, it may be mentioned that Bergh (1881b, p. 226) recorded a specimen of Dinophysis laevis [= Phalacroma rotundatum (Clap, and Lach.) var. laevis (Claparede and Lachmann) Jorgensen] in a restmg condition ("im Ruhezustand, zu einer Kugel innerhalb der Schale zusammengezogen"). The significance of this observation must be regarded as uncertain (see Blitschli, 1885, p. 1020), since encystment has not been investigated or definitely known to occur in the Dinophysoidae. Distribution: — The rather extensive data on the distribution of Phalacroma published at the present writing show that this genus is marine, almost exclusively eupelagic, of general oceanic distribution, but that the great majority of its species are limited to waters of tropical, subtropical, or warm-temperate nature or origin. The great difficulties implied in the determination of several of the species of this genus, the broad concept of species held by many of the inves- tigators in this field, and the fact that but relatively few of the data are accom- pamed by figures or descriptions by means of which the determinations of species may be checked (see p. 58) make the general distribution of many of the species rather uncertain. The distribution of the genus as a whole, on the other hand, may be regarded as fairly well established. Phalacroma rotundatum, P. omtuni, P. paulseni, P. rotundatum var. laevis, P. minutum, and P. rudgei are known to occur in the cold waters of the far north. Phalacroma rotundatum has been found along the whole coast of Norway, off Greenland, and in the .4i"ctic Ocean; P. ovnium, on the west coast of Norway and off Greenland; P. paulseni, oE Greenland; P. rotundatum var. laevis, on the west coast of Norway; and P. minutum. and P. rudgei, in the Atlantic Ocean between lat. 58° N. and lat. 60° N. Of these species, P. rotundatum, P. ovatum, P. paulseni, and P. rotundatum var. laeins are endogenetic in the northern waters, while P. minutum and P. rudgei are probably southern forms, occasionally carried to the 54 THE DIXOPHYSOIDAE. far north by the Gulf Stream. In the Antarctic Ocean P. rotundatum var. laevis has been found by Karsten (1905) as far to the south as lat. 63° S. It is a striking fact that all these species are small, rotund, and simply organized. The more or less luxuriant species appear to be restricted to warm-water regions (Schiitt, 1893, p. 269). The only species of this genus known to be endogenetic in brackish waters is P. rotundatum, which has been recorded repeatedly from the less saline regions of the Baltic Sea. Phalacroma is found in surface waters, but its optimum habitat seems to be the deeper levels of photosynthesis (Karsten, 1907, p. 442, 444). Some species, possibly sapruzoic, ha\e been found almost if not quite below the photic zone. It should be remembered, however, that the vertical distribution of this genus is very incompletely known, since 1/ut few records based on catches made with closing nets are available. The following from Karsten's (1905, 1906, 1907) reports on the phytoplankton of the Valdivia Expedition are the only published records. Phalacroma rotundatum var. laevis was found at Station 132, 400- 300 m.; Station 151, 200-100 m.; P. operculatum at Station 170, 200-100 m.; Station 229, 200-20 m.; and Station 268, 63-46 m. Phalacroma doryphorum was taken at the following stations: — Station 227, 1000-800 m.; Station 228, 420- 350 m.; Station 229, 1600-1400 m.; Station 239, 120-105 m.; Station 268, 105- 88 m. and 63-46 m. All the specimens recorded were reported as "living." The specimens of P. doryphorum taken at great depths were found to be without chromatophores. Although many of the species are widely distributed, none of them has been found to be abundant or even common. Indeed, most of these species appear to be rare, some even very rare. Representatives of Phalacroma were found at 100 (78.7%) out of the 127 stations of the Expedition from which dinoflagellates were recorded. These 100 stations are distributed over the whole area covered by the Expedition as shown in the following table and Plate 24: — Number and Number of Stiitions and Occurrences % of Occurrences % Cabfornia Current 4571,4.574,4.580,4.583 4 4 100 Mexican Current 4.587, 4588, 4590, 4592, 4594, 4596, 4598, 4r>(IO, 41)04, 4605, 4607, 4.545, 4546 13 13 100 Pan.amic Area 4609, 4611, 4613, 4615, 4617, 4619, 4621, 4631, 46:34, 46.3.5,4637,46:38,4640,4644 /4 17 82.3 Peruvian Current 4646, 4647, 4648, 46.50, 4651, 4652, 4655, 4657, 46.59, 4660, 4661, 4662, 4663, 4664, 4665, 4666, 4667, 4668, 4669, 4670, 4671, 4673, 4675, 4676, 4678 Easter Island Eddy 46S9, 4691, 4(i92, 4695, 4697, 4699 Galapagos Eddy 4713, 4715 25 27 92.6 6 10 60 2 A 50 Number and % of Occurrences % 30 4.5 66.6 3 3 100 2 2 100 1 2 50 SYSTEMATIC ACCOUNT. 55 Number of Stations and Occurrences South Equatorial Drift 4679, 4680, 4681, 4683, 4685, 4687, 4701, 4705, 4706, 4707, 4709, 4711, 4717, 4719, 4720, 4721, 4722, 4724, 4725, 4728, 4730, 4731, 4732, 4733, 4734, 4736, 4737. 4739, 4740, 4741. South Equatorial Current 4540, 4742, 4743 Equatorial Counter Current 4541, 4.542 North Equatorial Current 4543 At sixty-seven of these 100 stations Phalacroma was taken in vertical hauls and at forty-five in surface hauls. Most of the records from vertical hauls are from 300-0 fathoms, some are from 800, 400, 150, 100-0 fathoms. Vertical hauls were made at sixty-eight out of the 127 stations. The genus thus was found at not less than 98.5% of the stations at which vertical hauls were made; it was absent from one station only, 4574 in the California Current, at which it was found in a surface haul. Surface catches and Salpa stomachs of specimens taken in surface waters (eighty-one surface catches and twenty-four samples of Salpa stomachs) were examined from eighty-two stations. The genus, as previously mentioned, was taken in surface hauls at forty-five stations, i.e., at 54.9% of all the surface sta- tions. Taking into consideration the surface stations only, these forty-five record stations are distributed in the following manner: — Number of Stations and Occurrences California Current 4574, 4583 Mexican Current 4588, 4590, 4592, 4596, 4600, 4604, 4607, 4545, 4546 Panamic Area 4611, 4615, 4617, 4619, 4624, 4631, 4635, 4640, 4644 Peruvian Current 4648, 46.50, 4657, 4660, 4661, 4664, 4666, 4669, 4675, 4676, 4678 Easter Island Eddy 4692 Galapagos Eddy South Equatorial Drift 4680, 4706, 4709, 4720, 4725, 4731, 4733, 4741 South Equatorial Current 4743, 4540 Equatorial Counter Current 4541, 4542 North Equatorial Current 4543 On account of the prevalence of Phalacroma in deeper waters, the distribu- tion and the frequency of the record stations at which vertical hauls were made should be used as an indicator of the horizontal distribution of this genus in the area investigated by the Expedition. These records show that the genus is almost evenly distributed throughout this area; indeed, it was found at not less than 98.5% of the stations at which vertical hauls were made. Furthermore, the even distribution of this genus is also evident when the distribution and fre- quency of the records of species are considered (Plate 24). The only region in Number and % of Occurrences % 2 4 50 9 9 100 9 13 69.2 n 20 55 1 5 20 0 2 0 8 23 34.7 2 2 100 2 2 100 1 2 50 56 THE DINOPHYSOID.\E. which Phalacroma was found tci be represented rather sparingly was the CaU- fornia Current ; north of Point San Lucas there were only four records of species, viz., three records of P. parvuluin and one record of P. rapa. This scarcity pos- sibly may be due to the fact that the California Current comes from the north, in other words, from a region where this genus is less abinidant than within the tropics and subtropics. Of the eight most frequent species, P. argus, P. cuneus, P. doryphorum, P. porodidyum, and P. rapa (Figure 9, 13, 24, 7, 17) were found to be almost evenly distributed throughout the area investigated; P. reticulaturn (Figure 26) was not fountl nortli of the equator; P. striatum (Figiu-e 15) was not found in the California, Mexican, and Peruvian Currents; and P. hindmarchi (Figure 19) was restricted to the southwestern portion of this area, in other words, it had the same pecuUar distribution as Amphisolenia schauinslandi (Figure 51) and A. thrinax (Figure 59). The surface record-stations of Phalacroma are very unevenly distributed. ^^^lile this genus was found at twenty-nine (69.0%) out of the fortj^-two surface stations in the Mexican Current, the Panamic Area, and the Peruvian Current, it did not occur at more than nine (30.0%) out of the thirty surface stations in the South Equatorial Drift, the Easter Island Eddy, and the Galapagos Eddy. The causes of this uneven distribution may lie in the greater influence of vertical circulation in the regions nearer to the western shores of the continents. It should also be mentioned that there is no surface species of this genus prevalent in the coastal regions as in the case of Dinophysis (D. caudata). There are 309 records of species of Phalacroma from vertical catches. Out of these 309 records 1(0.3%; Station 4664) showed a frequency of 40%; 1(0.3%; Station 4663) 10%; 2(0.6%; Stations 4613, 4666) 6%; 2(0.6%,; Stations 4663, 4709) 5%; 2(0.6%c); Stations 4590, 4662) 4%,; 8(2.5%; Stations 4598, 4664, 4667, 4671, 4695, 4713, 4715, 4742) 3%; 11(3.6%; Stations 4609, 4634, 4638, 4675, 4681, 4689 [2 records], 4697, 4699, 4713, 4740) 2%o; 62(20.1%; Stations 4580, 4590, 4594, 4598 [2 records], 4605, 4613 [2 records], 4617, 4634 [2 records], 4637, 4638 [2 records], 4647, 4648 [3 records], 4650, 4651 [2 records], 4655, 4657 [2 records], 4665, 4667, 4671, 4676 [2 records], 4679 [2 records], 4681 [3 records], 4689 [3 records], 4691 [2 records], 4695, 4697, 4699 [3 records], 4701 [3 records], 4705, 4719 [2 records], 4721 [2 records], 4722, 4730, 4732 [2 records], 4734, 4737, 4739 [3 records], 4742) 1 %; 220(71.2%) showed a frequency of less than 1 %. There are seventy-four records of species of this genus from surface catches. Out of these seventj'-four records 1(1.4%; Station 4546) showed a frequency of 7%; 1(1.4%; Station 4619) 5%; 1(1.4%; Station 4669) 4%; 3(4.1%; Stations SYSTEMATIC ACCOUNT. 57 4604, 4542, 4545) 3%; 6(8.1%; Stations 4615, 4624, 4650, 4669, 4675 [2 records]) 2%; 19(25.7%; Stations 4574, 4600 [2 records], 4604, 4617, 4619 [2 records], 4635, 4657, 4664, 4666, 4720, 4731, 4743, 4540, 4541, 4542, 4543, 4545) 1%; 43(58.2%) a frequency of less than 1% or did not have the frequency estabUshed. Coincident occurrence of difTerent species of Phalacroma in catches from 300 (800, 400, 150, 100)-0 fathoms is recorded at the following of the sixty-seven stations mentioned above: — 15 species occurred coincidently at 1 station (1.5%; Station 4730) ; 12 species at 1 station (1.5%,; Station 4681) ; 11 specie-s at 1 station (1.5%; Station 4734); 10 species at 2 stations (3.0%; Stations 4699, 4701); 9 species at 1 station (1.5%; Station 4740) ; 8 species at 3 stations (4.5%; Stations, 4713, 4732, 4737) ; 7 species at 6 stations (9.0%; Stations 4679, 4697, 4705, 4711, 4734, 4742); 6 species at 7 stations (10.4%,; Stations 4587, 4637, 4691, 4695, 4709, 4721, 4739) ; 5 species at 6 stations (9.0%; Stations 4590, 4613, 4634, 4689, 4719, 4722); 4 species at 12 stations (17.9%; Stations 4594, 4598, 4605, 4617, 4638, 4655, 4662, 4664, 4665, 4676, 4683, 4717); 3 species at 9 stations (13.4%; Stations 4583, 4609, 4648, 4650, 4667, 4671, 4707, 4715, 4736); 2 species at 11 stations (16.4%; Stations 4580, 4647, 4651, 4652, 4657, 4661, 4663, 4666, 4675, 4685, 4728). Coincident occurrence of different species in surface catches is recorded at the following of the forty-five surface stations mentioned above : — 3 species occurred coincidently at 7 stations (15.6%; Stations 4604, 4617, 4619, 4743, 4540, 4542, 4545); 2 species at 15 stations (33.3%; Stations 4592, 4596, 4600, 4635, 4644, 4660, 4664, 4669, 4675, 4676, 4678, 4720, 4731, 4741, 4541). At the forty-five surface stations just mentioned, there are records of only nine of the thirty-one species found in the material of the Expedition. The number of surface records of each of these nine species is as follows: Phalacroma doryphorum, 26 records, 4 of which are from Salpa stomachs; P. rapa, 22 records, 2 of which are from Salpa stomachs; P. cuneus, 13 records, 2 of which are from Salpa stomachs; P. f aims, 6 records, 1 of which is from a Salpa stomach; P. argus, 2 records, 1 of which is from a Salpa stomach; P. parvulum, 2 records; P. hind- marchi, 1 record; P. expulsum, 1 record (Salpa stomach); P. lalirelatum, 1 record (Salpa stomach). Thus not less than sixty-one (82.5%) out of the seventy-four surface records refer to three species only. These three species are evenly dis- tributed over the whole area investigated by the Expedition (see p. 55, on the distribution of the surface stations). It may also be mentioned that P. poro- dictyum was taken in a surface haul in Acapulco Harbor, a station not included in the 127 discussed above. 58 THE DINOPHYSOIDAE. The fact that Phalacroma was found to be evenly distributed throughout the area investigated makes it highly plausible that this genus is fairly evenly distributed throughout all the tropical and subtropical seas. Finally, it may be mentioned that not less than sixteen out of the forty-seven presumably valid species have not been found except in the material of the Expedition. Historical Discussion and Systematics Stein (1883, p. 23), who estabhshed Phalacroma, gave no diagnoses but confined himself to a few words of comparisons and to the figuring of six new species, viz., P. nasutum, P. operculatum, P. porodictyum, P. argns, P. doryphnrum, and P. rapa. Diagnoses of this genus have been published by Biitschli (1885, p. 1009), Delage and Hcrouard (1896, p. 385), Schiitt (1896, p. 26), Paulsen (1908, p. 19), and Lebour (1925, p. 75). Of these diagnoses the one by Schiitt (1896) is by far the most nearly complete and satisfactory. No extensive descrip- tion and discussion of this genus have been published. Besides the sLx specific names established by Stem (1883), the following specific and subspecific names are to be found in the literature: P. armatwn Hensen (1895) P. haslatum Hensen (1895) P. operculoides Schiitt (1895) P. glnbiilvs Schiitt (1895) P. porodictyum Stein var. parvula Schiitt (1895) P. cunem Schiitt (1895) P. vasltim Schiitt (1895) P. vastum var. aciila Schiitt (1895) P. milra Schiitt (1895) P. jourdani (Gourret) Schutt (1895) P. ovum Schiitt (1895) P. hhckmani Murray & Whitting (1899) P. hivdmarcbi Murray & Whitting (1899) P. doHchopleriigium Murray & Whitting (1899) P. rudgci Murray & Whitting (1899) P. ynirtulum Cleve (1900c) P. cerntncorys Entz (1902a) P. cernlocorys var. tride7itata{Dadny) Entz (1902a) P. acuminatum (Claparede & Lachmann) Zacharias (1906) P. spliaericum (Stein) Zacharias (1906) P. knlicuJn Kofoid (1907a) P. reticulntum Kofoid (1907a) P. Mrintum Kofoid (1907a) P. ultimum Kofoid (1907a) P. circumnutum Karsten (1907) P. hastatum Pavillard (1909) P. CU71C11S Stuwe (1909) P. circtimcinclum Kofoid & Michener (1911) P.favuis Kofoid & Michener (1911) P. fimbriatum Kofoid & Michener (1911) P. giganleum Kofoid & Michener (1911) limbatum Kofoid & Michener (1911) porosum Kofoid & Michener (1911) praetextum Kofoid & Michener (1911) pulchrum Kofoid & Michener (1911) rotundatiim (Claparede & Lachmann) Kofoid & Michener (1911) turbineum Kofoid & Michener (1911) arcuatum Hensen (1911) P. argn Hensen (1911) P. cuter Hensen (1911) nblrquum Hensen (1911) propuhans Hensen (1911) rotundntum Hensen (1911) acrilum (Schiitt) Pavillard (1916) pulchellum Lebour (1922) kofoiiU Herdman (192.3) rotundatum (Claparede & Lachmann) var. tncpis (Claparede & Lachmann) .lorgensen (1923) P. ovntum (Claparede & Lachmann) Jorgensen (1923) P. parrulum (Schiitt) Jorgensen (1923) P. ehngatum Jorgensen (1923) P. stetiopterygiiim Jorgensen (1923) P. simulans Jorgensen (1923) P. pugiuncului Jorgensen (1923) P. ebriola Herdman (1924) P. irregulare Lebour (1925) SYSTEMATIC ACCOUNT. 59 In the present paper the following new specific names have been estab- lished:— P. apicatum, P. hipartitum, P. contradum, P. cuneolus, P. lativelatum, P. lens, P. rnucronaium, P. pavlseni, P. protuberans, and P. pyriforme; and two previously described species, Dinnphysis galea Pouchet (1883) and D. expulsa Kofoid and INIichener (1911), have been transferred to this genus. Several of the species mentioned in the last two paragraphs ha\'e been trans- ferred to other genera. Schiitt's (1895) transfer of Dinophysis jourdani Gourret to Phalacroma and Entz's (1902a) allocation, under the name of Phalacroma ccra- tocorys, are incorrect. This species belongs to Ceratocorys of the tribe Peridioni- oidae, as has been shown by Kofoid (1910a, p. 183) and by Jorgensen (1911, p. 147). Ceratocorys tridentata Daday (1888), which was treated by Entz (1902a) under the name of Phalacroma ceratocorys var. tridentata (Daday), is so inade- quately described and figured that its generic allocation must be considered as uncertain. Zacharias's (1906, p. 536, 530) use of Phalacroma for Dinophysis in designating Dinophysis acuminata Claparede and Lachmann and D. sphaerica Stein was probably due to slips of the pen, since at the remaining several places in his paper he maintained the original generic allocation of these two species. In any case, it is erroneous; both these species are typical representatives of Dinophysis. Herdman's (1923, p. 34) transfer of Amphidinium kofoidi var. petasatum Herdman (1922) to Phalacroma, under the name of P. kofoidi, is incorrect, since this species has a very small epitheca and since its transverse and longitudinal furrows are very deep and without distinct lists (see the Dinc- physidae, p. 32). It may also be mentioned that if this variety is given the status of a species, then its name must be petasatum and not kofoidi. For the same reasons Phalacroma ebriola Herdman (1924) does not belong to this genus; (see the family Dinophysidae, p. 32). The transfer to the genus Phalacroma of Dinophysis rotundata Claparede and Lachmann (by Kofoid and Michener, 1911), of D. ovata Claparede and Lachmann (by Jorgensen, 1923), and of D. laevis Claparede and Lachmann (by Jorgensen, 1923) probably is justified. These primitive forms are on the borderland between Phalacroma and Dinophysis, but they appear to be somewhat more closely related to the former than to the latter. It should be remembered, however, that these two genera merge into each other, and that their separation is almost arbitrary. As to whether Dino- physis laevis should be treated as a distinct species or as a variety of Phalacroma rotundatum is a question of its own that cannot be settled as yet. Phalacroma hastatum Pavillard (1909), although fairly highly differentiated, is of uncertain generic assignment. In the present paper we have followed Jorgensen's (1923, 60 THE DINOPHYSOIDAE. p. 31) suggestion and have referred it to Dinophysis hastata Stein (see p. 269). Apparently with full justification, Jorgensen (1923, p. 3) has made Phalacroma nasutum Stein (1883) the type of a new genus, Pseudophalacroma. Phalacroma ullimum Kofoid (1907a) has in the present paper been made the type of another new genus, Dinofurcula. The elevation to the rank of species of Phalacroma porodictyum Stein var. parvula Schiitt (1895) (by Jorgensen, 1923) and of P. vasium var. acuta Schiitt (1895) (by Pavillard, 1916) probably is justified. Phalacroma arcuatum, P. argo, P. arnudum, P. enter, P. hastalum, P. obliquum P. propulsans, and P. roiundatum Hensen (1895, 1911) are nomina nuda. The identity in the arrangement of the species in the two tables given by Hensen (1911, p. 166, 167) and the great similarity in the names indicate that the new names P. arcuatum and P. argo were intended for P. armatum and P. argus. Phalacroma rotundatum Hensen (1911) does not refer to P. rotundatum (Clapa- rede and Lachmann), since Hensen (1895, 1911) records the latter as Dinophysis rotundata. P. cxmcus Stiiwe (1909), which also is a nomen nudum, evidently is due to a typographical error and stands for P. cxmeus. Phalacroma simulans Jorgensen (1923) is a synonym of P.famis Kofoid and Michener (1911); and P. stenopterygiu7n Jorgensen (1923) is a synonym of P. expulsum (Kofoid and Michener, 1911). Jorgensen (1923, p. 5) suggests that P. rudgei Murray and Whitting (1899) "is perhaps only a thick megacytical stage of Phalacroma rotundatum ^'ar. laevis." However, on account of our scant knowledge of the small and rotund species of this genus, it is advisable to treat P. rudgei, preliminarily, as a valid species. With regard to P. ovum Schiitt (1895) and P. operculoides Schiitt (1895) see p. 118. With regard to P. galea (Pouchet, 1883) see p. 121. Phalacroma irregulare Lebour (1925) is based exclu- sively on fission stages with incomplete lists and with the intercalary border not yet absorbed. It must be regarded as too insufficiently known for certainty of specific identification. Finally, Paulsen (1911b, p. 305, fig. 2) figured a specimen, under the name of Dinophysis rotundata, which is referable to Phalacroma but not to any of the species known at the present time. For this form we suggest the name Phalacroma paulsenl, sp. nov. Under the designation of Phalacroma sp., four specimens have been figured, viz., by van Breemen (1906, pi. 1, fig. 4a, b), by Okamura (1907, pi. 4, fig. 26 and pi. 5, fig. 42), and by Lindemann (1923, fig. 13). P. sp. van Breemen (1906) possibly may be referable to P. rudgei, as suggested by Paulsen (1908, p. 19). P. sp. Okamura (1907, pi. 4, fig. 26) has been allocated to P. apicatum (p. 111). SYSTEMATIC ACCOUxNT. 61 P. sp. Okamura (1907, pi. 5, fig. 42) probably belongs to P. mitra Schutt (see Okamura, 1912, p. 18), and not to P. elongatum as suggested by Jorgensen (1923, p. 11). P. sp. Lindemann (1923) cannot be identified at the present time. As will be seen from the previous paragraphs, there are at present forty-seven presumably valid species and one variety of Phalacroma, as follows : — P. acutum (Schutt) Pavillard (1916) P. apicatum, sp. nov. P. argus Stein (1883) P. bipartitum, sp. nov. P. blackniani Murray & VVhitting (1899) P. circumcinctum Kofoid & Michener (1911) P. circumsutum Karsten (1907) P. contractuni, sp. nov. P. cuneolus, sp. nov. P. cuneus Schutt (1895) P. dolichopterygiiun Murray & Whitting (1899) P. doryphorum Stein (1883) P. elongatum Jorgensen (1923) P. expulsum (Kofoid & Michener) nobis P. favus Kofoid & Michener (1911) P. fimbriatum Kofoid & Michener (1911) P. giganteum Kofoid & Michener (1911) P. globulus Schutt (1895) P. hindmarchi Murray & Whitting (1899) P. lativelatum, sp. nov. P. lens, sp. nov. P. lenticula Kofoid (1907a) P. limbatum Kofoid & Michener (1911) P. niinutum Cleve (1900c) P. mitra Schtitt (1895) P. mucronatum, sp, nov. P. operculatum Stein (1883) P. operculoides Schutt (1895) P. ovatum (Claparede & Lachmann) Jorgensen (1923) P. ovum Schutt (1895) P. parvulum (Schutt) Jorgensen (1923) P. paulseni, sp. nov. P. porodictyum Stein (1883) P. porosum Kofoid & Michener (1911) P. praetextum Kofoid & Michener (1911) P. protuberans, sp. nov. P. pugiunculus Jorgen.sen (1923) P. pulchellum Lebour (1922) P. pulchrum Kofoid & Michener (19il) P. pyriforme, sp. nov. P. rapa Stein (1883) P. retieulatum Kofoid (1907a) P. rotundatum (Claparede & Lachmann) Kofoid and Michener (1911) P. rotundatum (Claparede & Lachmann) var. laevis (Claparede & Lachmann) P. rudgei Murray & Whitting (1899) P. striatum Kofoid (1907a) P. turbineum Kofoid & Michener (1911) P. vastum Schutt (1895) Besides these forty-seven valid species, we have Phalacroma galea (Pouchet, 1883), which may be identical with either P. doryphorum or P. circumsutum, and P. irregulare. All the published descriptions of the species are short and incomplete and have no or but few data of variation for consideration. The descriptions of Kofoid (1907a) and of Kofoitl and Michener (1911) are preliminary and based on the same material as those of the present paper. Some of the forty-se\-en species mentioned above, e.g., Phalacroma globulus Schutt and P. minutum Cleve, are so incompletely known that their certain specific identification will cause considerable difficulty. Furthermore, the synonymies of some of these species, e.g., those of P. rolundatum (Claparede and Lachmann) and of the other species occurring in northern European waters, are so exceedingly complicated that it appears more than doubtful whether anybody will be capable of unravel- ling them in a satisfactory way. Finally, a few of these species, as conceived in the present paper, are so variable that there are reasons to beUeve that future 62 THE DIXOPHYSOIDAE. investigators with adequate material at their disposal will find it necessary and feasible to undertake a further systematic subdivision. The first investigator to attempt a subdivision of Phalacroma (Jorgensen, 1923) divides the genus into six sections, viz., Paradinophysis, Euphalacroma, Cuneus, Argus, Podophalacroma, and Urophalacroma. The species included in these sections are as follows: — Paradinophysis: — P. rotundatum, P. rolunda- tum var. laevis, P. rudgei, P. ovatum, and P. parvulum. Euphalacroma: — P. ovum, P. porodictyum , P. operculatum, P. acutum, and P. elongahwi. Cuneus: — P. expulsum, P. cuneus, P. hlackmani, and P. striatum. Argus: — P. argus. Podophalacroma: — P. rapa, P. dolichopterygium, P. mitra, P. favus, and P. hindmarchi. Urophalacroma: — P. doryphorurn, P. circumsutwn, P. pugiunculus, and P. pulchrum. Using Jorgensen's system as a basis, Pavillard (192.3a) suggested a division of the genus into two subgenera, Euphalacroma and Para- dinophysis. To the first of these he referred four of Jorgensen's sections, viz., Oblongata (nom. nov.), Cuneus, Argus, Podophalacroma. To the second sub- genus he referred the remaining two of Jorgensen's sections, viz., Rotundati (nom. nov.) and Urophalacroma. The situation of the sulcal lists relative to the thecal valves was unknown even to Stein (1883, p. 23). This problem was settled by studies on the closely related genus Dinophysis (see p. 224). Biitschli (1885) and Schutt (1895, 1899) have given fairly extensive accounts of the structure and organization of the thecal wall and of the cell contents. With regard to the morphological changes that take place in the theca pre- ceding binary fission, it may be mentioned that megacytic specimens were figured by several investigators, e.g., by Stein (1883), Schutt (1895), Murray and Whitting (1899), and Meunier (1910) before Pavillard (1915a, 191(i) dis- cussed at length and correctly interpreted this phenomenon. Schutt (1895, p. 148) misunderstood the significance of the intercalary zones, while Meunier (1910, p. 59) gave a short l)ut correct account of these structures. It may also be mentioned in this connection that Pouchet (1894) interpreted megacytic specimens in the genus Dinophysis as prefission stages. Contributions to our knowledge of the distribution of Phalacroma are to be found in the following papers, not specifying those previously mentioned in this section:— Cleve (1899c, 1900b, 1901a, c, 1902b, 1903b), Daday (1888), Entz (1902b, 1904, 1905), Forti (1922), Graf (1909), Gran (1912b, 1915), Karsten (1905, 1906, 1907), Lemmermann (1899a, 1901a, 1904, 1905a), Lindemann (1924, 1925), Lloyd (1925), Lohmann (1902, 1908a, 1920), Mangin (1912, 1915), Nathansohn SYSTEMATIC ACCOUNT. 63 (1908, 1909, 1910a), Ostenfeld (1898a, 1906, 1915, 1916b), Ostenfeld and Paulsen (1904), Ostenfeld and Schmidt (1901), Paulsen (1904), Pavillard (1905, 1912), Rauschenplat (1900), Schiller (1911a, b, c, 1912), Schmidt (1901), Schroder (1900a, 1906a, 1911), Stuwe (1909), and Whitelegge (1891). None of these papers contains original figures by means of which the correctness of the de- terminations may be checked. Forti (1922) gives reproductions of previously published figures of the species recorded. Due to the great difficulties implied in the determination of several of the species of Phalacroma and to the broad concept of species held by many of the investigators, much of the data on the distribution of this genus should be accepted tentatively. References to Phalacroma or minor contributions to our knowledge of this genus are found also in Balbiani (1884c), Bergh (1884), Calkins (1902), Chun (1903), Doflein (1909, 1911, 1916), Forti and Issel (1923, 1924), Hensen (1891), Kofoid (1906c), Lindemann (1923a), Nathansohn (1910b), Oltmanns (1922), Schutt (1893, 1900a), Steuer (1910, 1911), Walther (1893), and Willey and Hickson (1909). Pavillard (1923a) gives a critical review of Jorgensen's (1923) important paper on this group. Adaptive and Systematic Value of the Characters. Principles used IN the Description of the Species Although Phalacroma is limited largely to tropical, subtropical, and warm- temperate waters of relatively low buoyancy, it is characterized by the fact that the body (theca) is simple in shape, viz., subspheroidal (Schutt, 1895, pi. 2, fig. 12: 1), lenticular (Plate 3, fig. 3, 5), cuneate (Plate 2, fig. 5, 8), inversed fig-shaped (Plate 2, fig. 7), subbiconical (Plate 4, fig. 1-7), or bilaterally flattened ellipsoidal (Figure 3:6) and subobovoidal (Figure 11). In no species does the body show adaptations to flotation in the form of pronounced elongation or protuberances as in Amphisolenia, Triposolenia, and certain species of Dino- physis; they maintain themselves in their optimum habitat, i.e., in the upper strata of the sea mainly as follows: — (1) they are fairly active swimmers; (2) most of them are relatively small; (3) their protoplasm contains large pusules (Schutt, 1895, pi. 4, fig. 19:2) and inclusions of low specific gravity, such as fats and oils (Schutt, 1895, p. 82-86), and they thus have a specific gravity approaching that of the surrounding medium; (4) in many of them the surface friction is increased by structural differentiations of the outer surface of the theca (Plate 4, fig. 1-7) ; (5) some of them have relatively large cingular, sulcal, or para- 64 THE DINOPHYSOIDAE. sagittal lists (Plate 2, fig. 8; Plate 1, fig. 6; Plate 4, fig. 4; Plate 3, fig. 5; P'igure 23:6). The most important of these five points probably is the third one, i.e., the lowering of the specific gravity of the body. Without stressing the first point, it is known that the species of Phalacroma are fairly active swimmers, but, on the other hand, there are no data as yet on the question as to how their swimming capacity compares with that of the members of such genera as Amphisolenia and Triposolenia in which the body is highly adapted for flotation. As to the second point, i.e., the relatively small size of the body, it may be mentioned that the largest of all the known species of Phalacroma, P. giganteurn, is characterized by its thin and fragile theca, while its close relative, the comparatively small P. cuncus, has a rather heavy theca. The reduced thickness of the thecal wall in P. giganteurn may be interpreted as an adaptation to flotation; the decrease in the weight of the theca at least to some extent counterbalances the reduction in the relati^'e surface of resistance caused by the increase in the \'olume of the body. When the adaptive value of the lists is considered, it should be remem- bered that these structures are relatively small in all the species of the genus occurring in cold waters (viz., P. rolundatum and its variety laens, P. ovatuni, P. paulseni, P. minutum, and P. rudgei), and that they reach their highest de- velopment in waters of low buoyancy. A thorough understanding of the func- tions of each of the three different kinds of lists present in this genus, viz., the cingular, sulcal, and parasagittal lists, will require a careful analysis from the viewpoints of mechanics and hydrodynamics. Such an analysis is outside the scope of the present paper. The following summary statements must suffice. The cingular lists, although small, undoubtedly act as parachutes when the body is suspended with the apex uppermost. The nearly symmetrically de- veloped parasagittal lists of some species, e.g., of P. Umbatum and P. fimhrialum (Plate 3, fig. 5; Plate 4, fig. 4), act as parachutes when the organism is turned upon one of its lateral faces. The asymmetrical parasagittal lists of some species, e.g., of P. pulchnini (Plate 3, fig. 6), and the asymmetrical left sulcal hst increase the surface of resistance and cause the organism to sink in a descending zigzag or spiral path whenever it rests on one of its lateral faces. They also function as keels and rudders, thus stabilizing the progressive spiral swimming character- istic of Phalacroma {cf. Kofoid, 190Ge, p. 129). Noteworthy in this connection are the small size of the right sulcal list, when compared to the left, and the fact that the left sulcal list is somewhat obhque in positit)n, passing from the left valve to the right, and often somewhat spiral-shaped (Plate 2, fig. 1, of Phala- SYSTEMATIC ACCOUNT. 65 croma fimhriatum; Plate 5, fig. 3, of Dinophysis jorgenseni). Finally, it may be added that the inclination of the longitudinal axis of the body found in several species of Phalacroma may be connected with the spiral mode of swimming. With regard to the significance of the spiral swimming, see Jennings (1901). Although in the present paper all the external characters have been taken into account in the establishment and characterization of species, some of the specific separations in Phalacroma should be regarded as tentative. This is due partly to the simple organization of many of these species and partly to the fact that all the species characters are subject to variations of the fluctuating kind, in varying degree of amplitude. The characters which are most variable within the species are the size and shape of the bodj^ (theca), the relative height of the epitheca, the shape of the left sulcal list, and the development and structure of the accessory lists and sails. These are also the characters which have been most profoundly modified within the genus as a whole, and it is largely upon them that the subdivisions of the genus have been founded. The variations in the development of the parasagittal lists appear at least in part to be due to the comparative!}' slow formation of these structures following binarj^ fission. WTien not otherwise stated, the following principles have been apphed in the descriptions of species of this genus : — {1) All characteristics refer to specimens in lateral view. (2) Diagnoses and descriptions refer to premegacytic specimens only; when only megacytic specimens were available, descriptions and measurements were made from valves exclusive of the mtercalary zone. (S) The terms length and depth of bod}', and symmetrical and asymmetrical species refer to body (theca) exclusive of Usts. (4) When the cingular lists have only the two ribs near the dorsal sagittal suture, they are said to be without ribs. (5) The right sulcal list is without structural differentiations. If not otherwise stated in the diagnosis, the species is characterized by the following features : — (1) The body is subsymmetrical, and its longitudinal axis is perpendicular to the girdle. (2) The thecal wall has scattered pores. (3) In the case of the species figured with the body viewed dorsoventrally, the flagellar pore is located about a girdle-width behind the posterior cingular list. (4) The epitheca is visible above the anterior cingular list. (5) The cmgular Usts are ribbed, subequal, about as wide as the transverse 66 THE DIXOPHYSOIDAE. furrow, subhorizontal or with but slight anterior incUnation, and run in an almost straight line across the body. Their ribs are complete or almost so, i.e., they reach the free edge of the list, equidistant, straight, and simple or almost so. (6) The right sulcal list is subtriangular, and extends to Rj or to R3 of the left sulcal list or to a point somewhere between these two ribs. (7) Left sulcal list: the ventral and posterior margins, which are straight or but slightly convex, concave, or sigmoid, form together a distinct angle. The three main ribs are present, of moderate strength, almost straight, and not club- shaped or otherwise modified; and the list has no structural differentiation besides the main ribs. (S) Accessory Usts and sails are absent, or, when present, they lack structural differentiations. The methods of measuring lengths, angles, and proportions utilized are shown in Figure 29. Subdivisions. Relationships among the Species Although Phalacroma is one of the most primitive genera of Dinophy- soidae, it exhibits a marked structural diversity. Of all the known species, P. pulchellum is considered to be, on the whole, structurally the most primitive, resembling the simple ancestors from which this genus originated. This assump- tion is based on the small size (length, 21 fi) of this species, on its approach to the fundamental spherical form, its large and high epitheca (the posterior cingular list is located about 0.40 the length of the body from the apex), the slight develop- ment of its cingular and sulcal lists, its lack of accessory lists and sails, and on the areolation and porulation of its thecal wall. A fundamental feature of the structural difTerentiation within the genus is that it nearly always has been ac- companied by an increase in size. Although the body shows a great variety of shapes, it is never characterized by extreme elongation or by processes; on the other hand, bilateral compression is a nearly universal phenomenon. The gLrdle is supposed to have been originally nearly equatorial in position (see p. 30). From the equatorial position, which still is found in some of the primitive as well as in some of the advanced species (P. lenticula, Figure 3:7; P. limbatum, Plate 3, fig. 5), the girdle has migrated anteriorly; or a reduction in the height of the epitheca has taken place. The cingular lists show increase in size and structural complexity, but remain comparatively small and simple. A somewhat greater diversity in size and structure is exhibited by the left sulcal list. Just as in Orni- thocercus and Histioneis, it is largely the portion of this list behind the fission SYSTEMATIC ACCOUXT. 67 rib that has increased in size and structural complexity. The phylogenetic de- velopment of the accessory lists and sails appears to have proceeded from the posterior end of the left sulcal list anteriorly along the dorsal side of the body, and from the right valve to the left. The structure of the thecal wall seems to have developed in most cases from areolation to reticulation; ami there seems to have been an inherent tendency for the meshes of the reticulum to become progres- sively larger. Sometimes, however, a differentiation of the thecal structure appears to have taken place (e.g., P. giganteum). In most cases the progress witliin the several groups, which have been established in speciation, is from the less to the greater diversity and specializa- tion in structural detail. This increase in complexity is expressed in several, though not often in all the characters, of tlie individual species. It is interesting to note how the structural features utilized in speciation in this genus seem to be pervaded by the aspect of factors such as those with which the geneticist deals. In some groups the speciation might be conceived as a series, partially realized, of permutations of factors with intergradations due to multiple factors giving results which simulate seriation. In other groups we find more or less clearly defined orthogenetic series of intermediate and progressively differentiated species. The factorial analyses of genetics do not compel the assumptions that the individual links in these series have originated chronologically, i.e., that they are genetically sequential, although the conventional conception of the evolu- tionary process inclines us to this interpretation. Nuclear reorganizations equiva- lent, or related to those which ensue in gametogenesis or endomixis would ap- pear to be an essential cytological basis for the distribution of genes which the kind of speciation outlined above seems to suggest. However, such reorganiza- tions are as j-et unknown in this genus. 1 . CoNTRACTUM group : — p. conlracluw, sp. nov. Figure 3:1. 2. RoTUNDATUM group: — p. jjulcliellum Lebour (1922, fig. 1-4). P. paruuluiii (Schtitt) J6rgen.sen. Figure 3: 4-6. P. ylubuhts Schutt {lS9r,, pi. 2, fig. 12). P. operculoides .Schiitt (1895, pi. 2, fig. 11: 1). P. Ttuigei Murray & Whitting (1899, pi. 31, fig. 6). P. rotundatum (Clapar«le & Lachmann) Jorgeiisen (1923, fig. 2). P. rotundatum (Claparede & Lachmann) var. laeiis (Claparede & Lachmann) ( 18.58, pi. 20, fig. 13). P. irregulare Lebour (1925, pi. 11, fig. 4). P. laiivelatum, sp. nov. Figure 3: 2, 3. P. paulseni, sp. nov. Paulsen (1911b, fig. 2). P. umtuin (Claparede & Lachmann). Jorgensen (1923, fig. 3). P. lens, sp. nov. Figure 4: 1-3. P. porosum Kofoid & Michener. I-'igure 5. P. lenticula Kofoid. Figure 3: 7. 68 THE DINOPHYSOIDAE. 3. Ahgus group: — P. porodictyum Stein. Figure C. P. vastum Schiitt (1895, pi. 3, fig. 16: 3). P. argus Stein. Figure 8:1,2. P. apicalum, sp. nov. Figure 10. P. operculatum Stein s. str. (1883, pi. 18, fig. 8). P. cirmincuictum Kofoid & Michener. Figure 8; 3. P. mniiit Schiitt. Figure 11. P. elongatum Jorgensen (1923, fig. 9). P. pyrifoTme, sp. nov. Figure 4: 4, 5. 4. CuNEUS group: — P. cuiieus Schiitt. Figure 12. P. blackmani Murray & Whitting (1899, pi. 31, fig. 4). P. striatum Kofoid. Figure 14: 3. P. gi(7a?i/e««i Kofoid & Michener. Figure 14: 1. 5. Rapa group: — p. acutuin (Schiitt) Pavillard. Schiitt (1895. pi. 3, fig. 17: 7). P. minutwn Cleve (1900c, pi. 8, fig. 10, 11). P. dolichoplerygium Murray & Whitting (1899, pi. 31, fig. 8). P. mitra Schutt (1895, pi. 4, fig. 18). P. rapa Stein. Figure 10. P. farm Kofoid & Michener. Figure 14: 2, 4, 5. P. hindmarchi Murray & Whitting. Figure 18. 6. ExPULSUM group: — P. protuberaiis, sp. nov. Figure 20: 1-5. P. expulsum (Kofoid & Michener). Figure 20: 6-9. 7. LiMBATUM group: — P. limbatwn Kofoid & Michener. Figure 21: 1. P. bipartitum, sp. nov. Figure 21 : 2. P. pulchrum Kofoid & Michener. Figure 21: 3. 8. DoRYPHORUM group : — P. mucroiiatum, sp. nov. Figure 22: 4, 6, 8. P. doryphuTum Stein. Figure 23: 1-5. P. circumsutum Karsten. Figure 23: 6. P. cviieotus, sp. nov. Figure 23: 1-3, 5, 7. P. pugiuiiculus Jorgensen (1923, fig. 19). 9. Praetextum group: — P. praetextum Kofoid & Michener. Figure 25: 4, 5. 10. Reticulum group: — p. fimbriatuin Kofoid & Michener. Figure 25: 1. P. reticulatum Kofoid. Figure 25: 2. P. turhineum Kofoid & Michener. Figure 25: 3. Discussion of Species Groups 1. CoNTRACTUM group (Figure 3:1). The sole member of this group is of uncertain generic assignment. The small size and subrotund shape of the body, the fairly large size of the epitheca, the small size and the shape of the cingular SYSTEMATIC ACCOUNT. 69 and sulcal lists, and the lack of ribs in the left sulcal list affiliate it with Pseudo- phalacroma nasuhim. The latter species is characterized especially by an anterior prolongation of the longitudinal furrow, extending two thirds the dis- tance from the girdle to the apex. This prolongation, which presumably is a primitive feature of great taxonomic significance, is not indicated in our drawing of Phalacroma contraclujn but may have been overlooked. The small size and simple structure of the cingular and sulcal lists indicate that this species is on a lower evolutionary level than the members of the Rotundatum group. 2. Rotundatum group (Lebour, 1922, fig. 1-4; Figure 3:4-6; Schlitt, 1895, pi. 2, fig. 12, 11 : 1 ; Murray & Whitting, 1899, pi. 31, fig. 6 ; Jorgensen, 1923, Argus Rapa Cuneus Limbatuni Reticulatu m Praetextum Contractum Figure 2. — Graphical representation of the relationships of the groups of species in Phalacroma. fig. 2; Claparede iu pennae). Phalacnmut opercuhihim ^CHij'n,\S95,p. 93, pi. 2, fig. 10: 1-3. Hensen, 1911, partim, p. 166, 167, tab. 15. Phalacroma porndirliim ScHRiiDEB, 1900a, p. 9, 11, 19 (/o/wk pennae); 1906a, p. 322,32.5, 327; 1911, p. 21, 37. IPhahuroma poroJictijwn Ok.\m0r.\, 1912 (non 1907, p. 134), p. IS, pi. .5, fig. 83. Diagnosis: — Body obovate or subellipsoidal in lateral outline, deepest at or somewhat behind girdle, usually somewhat deeper anteriorly than posterior^; length: depth, 1.11-1.23: 1 ; longitudinal axis deflected posteroventrally at 0°-20°. In dorsal view obovate, 1.60-1.80 times longer than wide. Posterior cingular list 0.37-0.50 the length of body from ajiex. Cingular lists without, or with but faintly indicated ribs. Right sulcal list with angular margin, concave anteriorly, and convex or straight posteriorly. Left sulcal list 0.37-0.53 the length of body; distance between Ri and R3 is 0.33-€.40 the length f)f body; R2 is 0.09-O.13, and R3 is 0.10-0.21 the greatest depth of body; margin forms angle of 75°-100° at R3: R3 inclined posteriorly at 15°-45°. Theca finely and closely areolate. Length, 72.5-81.5 M. Widely distributed in tropical, subtropical, and warm-temperate seas. Description: — A medium-sized species, the body of which is obovate or sub- ellipsoidal in lateral outline, deepest at or somewhat behind the girdle, and 1.11- 1.23 times longer than deep. Li the Expedition specimens the ratio between the length and the depth is 1.18 (1.14-1.23): 1; in Stein's (1883, pi. 18, fig. 11), 1.14: 1 ; in Schutt's (1895, pi. 2, fig. 10: 2; 13:3), 1.11-1.15: 1 ; and in Jorgensen's (1923, fig. 6), 1.13: 1. The longitudinal axis of the body either is perpendicular to the transverse furrow, as in Stehi's (1883) specimen and in some of the Expedition specimens, or it is deflected posteroventrally at an angle of l°-20°; in the Expedi- tion specimens this deflection does not exceed 8°; in Schutt's (1895, pi. 2, fig. IO2) it is about 20°, and in Jorgeiisen's (1923) about 12°. The epitheca is 0.97 (0.96-0.98) as deep as the hypotheca, strongly and al- most evenly convex, in some specimens slightly flattened ventrally or both ventrally and dorsally (Figure 6: 1), dome-shaped, highest at or in some speci- mens slightly dorsal to the center, and ^•ery prominent above the anterior cingular list. The transverse furrow is flat or slightly convex, and its width is 0.20-0.30 the greatest height of the epitheca. The posterior cingular list is 0.37-0.50 the length of the body from the apex; in the Expedition specimens this ratio is 0.47 (0.44-0.50) : 1 ; in Stein's (1883), 0.37: 1; in Schutt's (1895, pi. 2, fig. 10:2, and 13: 3), 0.38-0.50: 1 ; and in Jorgensen's (1923), 0.37: 1. The hypotheca is some- times symmetrical, as in Stein's (1883) specimen and in Figure 6: 1, but usually 100 THE DINOPHYSOIDAE. it has a more or less pronounced post ero ventral inclination (see p. 99). Its dorsal, posterior, and ventral margins are well and evenly convex and confluent. Its posterior portion, which is rather broad in some specimens and rather narrow in others, is nearly always somewhat narrower than the anterior portion of the body. In dorsoventral view (Stein, 1883, pi. 18, fig. 13) the body is regularly obovate, about 1.60-1.80 times longer than wide, and widest at or somewhat be- hind the girdle; its anterior portion is broadly, its posterior narrowly convex; and its side contours are evenly convex. The cingular lists are subhorizontal and subequal; they are about as wide as, or slightly wider or narrower than, the transverse furrow, and are "without or with faintly indicated ribs" (Jorgensen, 1923, p. 9). The right sulcal list usually ends at or somewhat behind a point midway between the fission rib and the posterior main rib of the left sulcal list; .sometimes it extends even to the last- mentioned rib; in some specimens its ventral margin, even in the widest region, does not quite extend to the ventral margin of the left sulcal list, but in most specimens it does. The anterior half to two thirds of the free margin of this list is more or less concave and sometimes strengthened by a marginal rib ; the poste- rior portion of this margin, which forms a distinct angle with the anterior portion, is more or less convex or almost straight. The greatest width of this list is located just behind the concave portion. The left sulcal list is 0.37-0.53 as long as the body; in the Expedition specimens this ratio is 0.40 (0.44-0.53): 1; in Stein's (1883), 0.40: 1; in Schiitt's (1895, pi. 2, fig. ISs), 0.39: 1; in Jorgensen's (1923), 0.37: 1. The distance between the anterior and posterior main ribs is 0.36 (0.33-0.40) the length of the body. The anterior main rib, when present, is 0.10-0.17, the fission rib 0.08-0.13, and the posterior main rib 0.10-0.21 the greatest depth of the body; in the Expedition specimens these three ratios are 0.12 (0.10-0.13) : 1, 0.10 (0.08-0.13) : 1, and 0.17 (0.14-0.21) : 1 ; behind the poste- rior main rib the list decreases suddenlj' in width. Between the anterior and posterior main ribs the free margin of this list usually is slightly sigmoid, concave anteriorly and convex posteriorly; sometimes it is nearly straight, or gently con- cave or convex, or it may be slightly angular at the fission rib (Jorgensen, 1923, fig. G) ; at the posterior main rib it forms an angle of 75°-100°; in the Expedition specimens this angle is 93° (90°-100°); in Stein's (1883) about 75°; behind the last-mentioned rib it is almost straight or slightly concave or convex. The main ribs of this list are of moderate strength, straight or nearly so, and not club- shaped or otherwise modified. The distance between the anterior main rib and the fission rib is 0.35 (0.25-0.45) the distance between the anterior and posterior SYSTEMATIC ACCOUNT. 101 main ribs. The posterior main rib has a posterior inclination of 30° (15°-45°). Sometimes this Hst has a faint reticulation (Stein, 1883; Jorgensen, 1923). There are no accessory lists or sails. The thecal wall is finely, closely, and usually rather faintly areolate, and has scattered pores. In the transverse furrow there are two rows of pores. Megacytic stages have been seen by Stein (1883) and Schiitt (1895). The proportions of eight of the Expedition specimens and of those figured by Stein (1883, pi. 18, fig. 11), Schiitt (1895. pi. 2, fig. 10: 2, and 13: 3) and Jor- gensen (1923, fig. 6) were measured. Figure 6. — 1-5, Phnhcroma porodictinim Stein, right lateral view. Porulation indi- cated only on small portion of theca in 4 and 5. X 430. 1, 2, 3, 5, from Station 4737 (300-0 fathoms); 4, from Station 4739 (300-0 fathoms). Dimensions: — Expedition specimens: Length of body, 74.8-8.15// (average, 77.5 yu). Greatest depth of body, 62.2-69.8 m (average, 65.8 ^l). The size of the type specimen (Stein, 1883, pi. 18, fig. 11) is unknown. According to Stein's (1883) information about the magnifications of his figui'es, this specimen was somewhere between 73 ^ and 112 ju long. The specimens represented by Schiitt (1895, pi. 2, fig. 10: 2, and 13: 3) are about 76-78 n long. The length of Jorgen- sen's (1923, fig. 6) specimen is about 72 /j. According to Lohmann (1902, p. 53), the length varies between 58 ^ and 76.5 fi. This statement, which implies a striking variability in size, has been disregarded in the present paper, since Lohmann does not give any figiu'es or description by means of which his con- ception of the species might be checked. Variations: — The variability of this species, as conceived here and by Jorgensen (1923), is rather striking. However, there seems to be but little doubt 102 THE DINOPHYSOIDAE. that we are dealing with a natural systematic unit. The different fnrms repre- sented by Figure 6 and by the figures of previous investigators appear to be con- nected by continuous series of intermediate forms and are probably modifications. The size of the body, the inclination of the longitudinal axis of the l)ody, the relati\'e height and the shajie of the epitheca and of the hypotheca, and the shape and size of the left sulcal Ust are the most variable characters. Comparisons: — None of the specimens found in the material of the Expedi- tion and determined as Phalacroina porodictyum agrees completely with the type specimen of this species as figured by Stein (1883, pi. 18, fig. 11). However, some of these specimens (Figure 6:1) approach the type so closely that their assign- ment can be regarded as certain. They differ from the type mainly in having the epitheca somewhat higher relatively and the left sulcal list somewhat longer. Other specimens (Figiu'e 6:3) are rather similar to Schiitt's (1895, fig. 10:2) and to Jorgensen's (1923, fig. 6) but we have not seen any specimens that show a complete agreement with any of the figures of these two authors. Plialncromn porodictyum is proliably most closely related to P. arc/us. This relationship is indicated l\y the close resemblance in the size and shape of the body, and in the shape and structure of the right sulcal list. The first species differs from the last mainly in the angularity of its left sulcal list and in the areolate structure of its thecal wall. In both these characters it probably is the more primitive. For further discussion, see p. 107. Phalacroma porodictyum is certainly specifically distinct from P. parvidum, a form which Schtitt (1895) first figured as a variety of P. porodictyum, and which Jorgensen (1923, p. 7) established as a distinct species. .S;/''0" //'".'/-■ The spofiniens figured as Pluildcmmn njiirculriluin Stein l)y Scluitt (1S0.5, pi. 2. fi^- 10) prnhahly are referable to P. porodictyum. The sjiecinien represented by Okamura (1007, pi. 4, fig. 26) under the de.^iRnation P. sp. . . . Phalacroma imrodich/um Stein? tnidonbtedly belongs to P. apiralum nob. Phalacroma porndicliiiiiii Okamura (1912, fig. S3) is probably not identical with /'. poroilirli/um Stein, since the specimen figured is very small, about 47 ^ long, and has a left sulcal list which is more than twice as wide at the posterior main rib as it is anteriorly. Henscn (191 1, p. 166) writes that he has partly confused Phalacroma porodictyum and /'. opcrculatum. These two species probably have been confused by several investigators, but the extent of this confusion cannot be established, since only a few authors give figures or descriptions. It may be mentioned in this connection that C'leve never recorded /'. porodictyum, although he carried out extensive investigations on the plankton of the seas from which this species has been recorded repeatedly. On the other hand, Clevc has numerous records of P. opcrcnlatiwi. Occurrence: — Pludarroma porodiclytnn is recorded at twenty-nine of the 127 stations. There are 5, 5, 6, 5, 8, and 0 stations on the six lines of the Expedi- tion. Of these twenty-nine stations, four (4587, 4590, 4598, 4605) are in the Mexican Current ; three (4613, 4634, 4637) in the Panamic .\rea; five (4648, 4(562, 4666, 4673, 4676) in the Peruvian Current; four (4689, 4691, 4697, 4699) in the SYSTEMATIC ACCOUNT. 103 Easter Island Eddy; one (4713) in the Galapagos Eddy; and twelve (4679, 4681, 4701, 4705, 4719, 4721, 4724, 4730, 4732, 4734, 4737, 4739) in the South Equa- torial Drift. At one station (4713) the species is recorded from 150-0 fathoms; at two stations (4662, 4666) from 800-0 fathoms. All the remaining records refer to hauls from 300-0 fathoms. The species is also recorded from surface waters in Acapulco Harbor, off the Mexican Current. This station is not included in the 127 stations mentioned above. Figure 7. — Occurrence of Pbalncroma porodicti/um Stein. Large, solid circles indicate records from vertical hauls; squares, records from surface hauls; small, solid circles, stations at which this species was not found; small, open circles, stations from which no plankton catches were examined. The temperature range of these twenty-nine stations at the surface was 67°-85°; the average was 75.5°. At Acapulco the surface temperature was 83°. The frequency is 1% at seven stations (4634, 4648, 4679, 4681, 4689, 4699, 4719). At the remaining stations it is less than 1 %. The species was first recorded by Stein (1883) "aus dem Atlantischen Meer und der Siidsee." The specimens of this species drawn by Schiitt (1895) prob- ably were taken either at Naples, or in the tropical or subtropical regions of the Atlantic. In the Atlantic it was found by the following investigators:- — Jorgensen (1923), in the Bay of Cadiz; Murray and WTiitting (1899), at lat. 37° 44' N., long. 30° 55' W.; Ostenfeld (1898a), at lat. 35° 44' N., long. 38° 03' W.; Stiiwe (1909), at lat. 28° 30' N., long. 40° 16' W.; Karsten (1906), at lat. 0° 9' S., 104 THE DIXOPHYSOIDAE. long. 8° 29' W.; Hensen (1911), localities not specified. From the Mediterranean it is recorded by Pavillard (1905, 1916), in the Gulf of Lyons; Forti (1922), in the Ligurian 8ea; Daday (1888) and Schroder (1900a), at Naples; Schroder (1906a), in the Ionian Sea; Entz (1902b, 1905) and Schriider (1911), in the Adriatic Sea; Jorgensen (1923), "throughout the entire Mediterranean." In the Red Sea it was found by Ostenfeld and Schmidt (1901); in the CUilf of Aden by Ostenfeld and Schmidt (1901) and Schroder (1906a); in the Arabian Sea by Ostenfeld and Schmidt (1901) and Schroder (1906a); off the coast of Celebes by Ostenfeld (1915); in the Peruvian Current by Lemmermann (1899a). With regard to Olcamura's (1912) records from Japanese waters, see p. 102. The species was found in waters of tlic following temperatures: — 74° (Murray and Whitting, 1899), 77.0°-80.3° (Ostenfeld and Schmidt, 1901), and 75.2° (Stiiwe, 1909). Of the investigators who ha\e contributed to a Icnowledge of the distribution of this species only Stein (1883), Forti (1922), and Jorgensen (1923) give descrip- tions or drawings of this form by means of which their determinations may be checked. Phalacroma porodiclyum is eupelagic and evidently widely distributed in tropical, subtropical, and warm-temperate seas. In the area investigated by the Expedition it appears to be of almost uniform occurrence. Its optimum habitat according to the records is in deeper waters, within the levels of photosynthesis. It was found at the surface at only one Expedition station, viz., Acapulco. Phalacroma argus Stein Figure 8: 1,2; 9 Phalacroma argxis Stein, 1883, p. 18, fig. 15-17. Schutt, 1895, p. 13, 83, pi. 3, fig. 15: 1-3. Lemmermann, 1899, p. 372; 19nia, p. 372. Cleve, 1901ii, p. Ui. Ostenfeld & Schmidt, 1901, p. 175. Entz, 1902b, p. 94; 1905, p. 111. Schroder, 190(in, p. 324, 327. K.\rsten, 1907, p. 235. P.vvillard, 1909, p. 283; 1916, 1). .52. HEN.SKN, 1911, p. 166, 167, tab. 15. Joroensen, 1923, p. 13, fig. 13. I'hahicrnina urgo Hensen, 1911, p. 167 {Uijifiti pcniiac). Diagnosis: — Body slightly obovate in lateral view, deepest at or somewhat behind girdle; length: depth, 1.15-1.48:1; longitudinal axis deflected postero- \entrally at 0°-12°. In ventral view obovate, widest somewhat behind girdle; length: width, 1.75: 1. Posterior cingular list 0.28-0.45 the length of body from apex. Cingular lists usually without ribs. Right sulcal list with angular margin, concave anteriorly and convex posteriorly. Left sulcal list 0.45-0.59 the length of body; distance between Ri and R3 is 0.33-0.43 the length of body; R2 is 0.07- 0.14 and R3 is 0.09-0.11 the greatest depth of body; margin gently sigmoid, con- SYSTEMATIC ACCOUNT. 105 cave anteriorly, convex posteriorly, seldom evenly convex throughout; R3 in- clined posteriorly at 25°-50°. Theca usually reticulate; 15-20 polygons border girdle posteriorly. Length, 83.5-110.5 11. Widely distributed in tropical, subtropical, and warm-temperate seas. Description: — A medium-sized or rather large species, the body of which is obovate in lateral view, deepest at or in most specimens somewhat behind the girdle and 1.15-1.48 times longer than deep. In the Expedition specimens the ratio between the length and the depth of the body is 1.17-1.25: 1; in Stein's (1883, pi. 18, fig. 15, 16), 1.32-1.48: 1; in Schiitt's (1895, pi. 3, fig. 15: 1), 1.15: 1; and in Jorgensen's (1923, fig. 13), 1.21: 1. The longitudinal axis either is per- pendicular to the transverse furrow, as in Stein's (1883) and in some of the Expedition specimens, or it is deflected posteroventrally at an angle of 1°-12''; in none of the Expedition specimens is it deflected more than at 4°; in those figured by Schiitt (1895) and Jorgensen (1923) the deflection is 8°-12°. The epitheca is 0.87-0.96 as deep as the hypotheca, strongly and almost evenly convex, dome-shaped, highest at or just dorsal to the center, and very prominent above the anterior cingular list. The transverse furrow is flat or slightly convex; and its width is 0.22-0.35 the greatest height of the epitheca. The posterior cingular list is 0.28-0.45 the length of the body from the apex; in the Expedition specimens the corresponding figures are 0.43-0.45; in Stein's (1883), 0.28-0.32; in Schiitt's (1895), 0.32; and in Jorgensen's (1923), 0.37. The hypotheca is sometimes .symmetrical, as in Stein's (1883) specimens and in Figure 8:2; but usually it has a more or less pronounced postero ventral inclina- tion (see above). Its dorsal, posterior, and ventral margins are well and evenly convex and confluent. The posterior end of the body is nearly always somewhat narrower than the anterior. In the specimens represented by Stein (1883, pi. 18, fig. 15) and Schiitt (1895, pi. 3, fig. 15: 1) the anterior and the posterior portions of the body are of about the same width. In dorsoventral view (Stein, 1883, pi. 18, fig. 17) the body is regularly obovate, al)out 1.75 times longer than wide, and widest at or somewhat behind the girdle; its anterior end is broadly, its posterior end narrowly, convex; and its side contours are evenly convex. The cingular lists are subhorizontal and subequal; their width usually about equals or shghtly exceeds the width of the transverse furrow; occasionally it is somewhat less. In most specimens these lists appear to lack ribs, but in speci- mens of well-developed thecal reticulation, they have basally "irregular, coarse and short radial beams, finally anastomosing into an incomplete, wide-meshed reticulation" (Jorgensen, 1923, p. 13). The right sulcal list extends to or some- 106 THE DINOPHYSOIDAE. what beyond a point midway between the fission rib and the posterior main rib of the left sulcal hst; in some specimens its ventral margin, even in the widest region, does not quite extend to the ventral margin of the left sulcal list, but in others it does. The anterior half to two thirds of the free margin of this list is more or less concave and, at least in some specimens, is strengthened by a marginal rib; the posterior portion of this margin, which forms a distinct angle with the anterior portion, is more or less convex or, seldom, almost straight. The greatest width of this list is located just behind the concave portion. The left sulcal list is 0.45-0.59 as long as the body; in the Expedition specimens this ratio is 0.45-0.53: 1; in Stein's (1883), 0.51-0.50:1; in Schiitt's (1895), 0..54: 1; in Jorgensen's (1923), 0.45 : 1 . The distance between the anterior and the posterior main ribs is 0.33-0.43 the length of the body. The anterior main rib, when present, is 0.07-0.14, the fission rib 0.07-0.14, and the posterior main rib 0.09-0.11 the greatest depth of the body; in the Expedition specimens these three ratios are 0.09-0.14:1; 0.08- 0.09: 1; and 0.09-0.11: 1. The greatest width of this list, which in most speci- mens is located about midway between the fission rib and the posterior main rib, is 0.10-0.15 the greatest depth of the body. The free margin of this list is gently sigmoid, concave anteriorly and convex posteriorly, seldom gently and evenly convex throughout (Stein, 1883, pi. 18, fig. 15, 16); in other words, it is not angular posteriorly as in most species of the genus, but gently and evenly convex. The main ribs of this list are straight and comparatively weak, and not club- shaped or otherwise modified. The distance between the anterior mam rib and the fission rib is 0.32-0.45 the distance from the anterior to the posterior main rib; in the Expedition specimens this ratio is 0.41-0.42: 1. The posterior main rib has a posterior inclination of 25°-50° ; this angle in the Expedition specimens is 30°-35°; in Stein's (1883), 40°; in Schutt's (1895), 25°; and in Jorgensen's (1923), 50°. There are no ribs except the three main ribs, but sometimes the list appears to be delicately reticulate. There are no accessory Hsts or sails. The thecal wall has a polygonate, sometimes but slightly developed, reticula- tion. There are fifteen to twenty polygons on each valve bordering the posterior margin of the girdle. Each polygon has a central pore. In the transverse fur- row there are two rows of pores (and polygons, if present), each row with fifteen to twenty pores on each valve. Megacytic stages have been found by Jorgenscn (1923, p. 13). The dimensions of two of the Expedition specimens and of the f^pecimens drawn by Stein (1883, pi. 18, fig. 15, 16), Schiitt (1895, pi. 3, fig. 15), and Jor- gensen (1923, fig. 13) were measured. SYSTEMATIC ACCOUNT. 107 Dimensions: — Expedition specimens: Length of body, 83.5-86.2 n. Greatest depth of bodj', 66.9-73.5 /i. We do not know the size of the two specimens figured by Stein (1883, pi. IS, fig. 15, 16), one of which is the type. According to Stein's information about the magnifications of his figures, these specimens were some- where between 80 ^ and 145 ^ long; the tj'pe specimen (Stein, 1883, pi. 18, fig. 16) was somewhere between 94 n and 145 /i long. The specimen represented by Schiitt (1895, pi. 3, fig. 15) is about 110.5 yu long and 96.1 /i deep; the one figure by Jorgensen (1923, fig. 13) is about 91.2 fx long and 75.7 m deep; i.e., these speci- mens are slightly larger than those found by the Expedition in the Pacific Ocean. Variations: — Phalacronia argus is rather strikuigly variable in the following characters: — the size of the body, the inclination of the longitudinal axis of the body, the relati\"e depth of the bod.y, the relative height of the epitheca, and the Fkjure 8. — 1, 2, Phalacroma argus Stein, right lateral view. Surface markings, faint reticu- lation, and ponilation, indicated only on small portion of theca in 1. X 430. 1, from Station 47.30 (300-0 fathoms); 2, from Station 4724 (300-0 fathoms). 3, Phalacroma circumcinctutn Kofoid and Michener, right lateral view of type specimen, slightly tilted. X 430. Station 4671 (300-0 fathoms). shape of the sulcal lists. The variations in the inclination of the longitudinal axis of the body are probably exaggerated due to the tilting of some of the speci- mens examined. Cotnparisons: — Stein (1883, pi. 18, fig. 15, 16) figures two specimens of Phalacroma argus, one relatively small and dorsoventrally narrow, the other relatively large and deep. The specimens in the material of the Expedition agree rather closely with the last -mentioned specimen (Stein, 1883, fig. 16). They differ mainly in the following respects: — (1) the epitheca is relatively higher; (2) the ventral margin of the left sulcal list is not evenly convex but sigmoid, being gently concave anteriorly and convex posteriorly. In most of the Expedition specimens the longitudinal axis of the body is deflected somewhat posteroventrally, while in Stein's specimen the body is symmetrical. The differences, however, are too small to be considered as specific. Moreover, it should be noted that the relative 108 THE DINOPHYSOIDAE. height of the epitheca, the shape of the ventral margin of the left sulcal list, and the inclination of the longitudinal axis of the body are characters subject to rather striking variabihty in the species of the Argus group (P. apicatum, Figure 10). The specimens of Phalacroma argns figured by Schiitt (1895, pi. 3, fig. 15) and by Jorgensen (1923, fig. 13) agree very closely with most of the Albatross specimens of this species. They differ mainly in having the longitudinal axis of the body deflected somewhat more posteroventrally, at 8°-12° as compared with l°-4°; they agree fairly closely with the specimen represented by Stein (1883, pi. 18, fig. 16). This specimen therefore should be regarded as the type of Phalacroma argns. The question as to whether or not the two forms represented by Stein (1883) under this name are different genetically cannot be settled as yet. Preliminarily, Jorgensen's (1923) broad conception of this species is adopted. Judging by the great similarities in the shape and structure of the cingular and sulcal lists and in the structural differentiation of the thecal wall, Phalacroma argus is probably more closely related to P. apicatum than to any other known member of the genus. Indeed, these two forms are so similar that even their specific separation is questionable, and the solution suggested should be regarded as tentative. Phalacroma argus differs from P. apicatum mainly in having rela- tively lower, well-rounded, and dome-shaped epitheca. Phalacroma porodictyum Stein, as conceived in the present paper and by Schutt (1895) and Jorgensen (1923), is probably, next to P. apicatum, the nearest- known relative of P. argus. This relationship is indicated by the close resem- blance in the size and shape of the body and in the shape and structure of the right sulcal list. Phalacroma- porodictyum is easily distinguished from P. argus and P. apicatum by the angularity of the posterior portion of its left sulcal list and by the structure of its thecal wall. Probably another close relative of Phalacroma argus is P. striatum. This species, which is a member of the rather highly differentiated Cuneus group, resembles P. argus in the following respects: — {1) The longitudinal axis of the body is deflected posteroventrally. (2) The free margin of the right sulcal list is concave anteriorly and convex posteriorly. Its concave anterior portion, which forms a distinct angle with the posterior portion, is strengthened by a marginal rib. (3) The left sulcal list is delicately reticulated. Its free margin is well rounded posteriori}', and not angular at the posterior main rib as in most of the species of this genus. (4) The thecal wall has a polygonate reticulation. The polygons are of moderate size and each of them has a central pore. Phalacroma SYSTEMATIC ACCOUNT. 109 striatum is easily distinguished from P. argus by its low epitheca, its broad, some- what wedge-shaped hypotheca, and its exceptionally large left sulcal list. The CuNEUS group probably originated from an ancestral form of about the same habitus as P. argus, just as this last species probably originated frcm a form re- sembling P. porodictyu7n. From the comparatively simple structure of Phalacroma argus, it seems probable that its closest relatives, next to P. apicatum, P. porodictyum, and P. striatum, should be sought among those members of the genus which are pre- FiGURE 9. — Occurrence of Phalacroma argus Stein. Large, solid circles indicate records from verti- cal hauls; squares, records from surface hauls; small, solid circles, stations at which this species was not found; small, open circles, stations from which no plankton catches were examined. sumably the most primitive. The following species are mentioned in this con- nection:— P. operculatum Stein s. str., P. onim Schiitt, P. operculoides Schiitt s. str. However, the degree of relationship between each of these simple species and P. argus is not known ; in other words, which of them is more closely related to this species than the others. Phalacroma argus is readily distinguished from these species by the shape and structure of its sulcal lists. See also p. 113. Occurrence: — Phalacroma argus is recorded at eighteen of the 127 stations. There are 1, 4, 4, 3, 4, and 2 stations on the six lines of the Expedition. Of these eighteen stations, one (4583) is in the California Current; one (4637) in the Panamic Area; three (4652, 4655, 4664) in the Peruvian Current; two (4691, no THE DINOPHYSOIDAE. 4695) in the Easter Island Eddy; ten (4679, 4681, 4683, 4701, 4705, 4722, 4724, 4725, 4730, 4740) in the South Equatorial Drift; one (4540) in the South Equa- torial Current. At two stations (4725 [Salpa], 4540), the species was taken at the surface; at one station (4052) in a haul frcnn 100-0 fathoms, and at one sta- tion (4701) in a haul from 800-0 fathoms as well as in a haul from 300-0 fathoms. All the other records refer to catches from 300-0 fathoms only. The temperature range of these eighteen stations at the surface was ()5°-83°; and the average was 73.7°. At Stations 4725 and 4540, where the species was taken at the surface, the surface temperature was 77° and 79°, respectively. At one station (4681) the frequency is 2%; at two stations (4655, 4691) it is 1%; at the remaining stations it is less than 1%. The species was first recorded by Stein (1883) "aus dem Atlantischen Meer und der Siidsee." Hensen (1911) found it at a very great number of stations throughout the tropical, subtropical, and warm-temperate regions of the Atlantic. "P. argus tritt schon im Floridastrom auf, ist im westlichen Teil der Sargossosee gut vertreten und kommt sonst ziemlich gleichmassig, aber llickenhaft vor " (Hensen, 1911, p. 167). The specimens drawn by Schiitt (1895) were probably taken either in the tropical, subtropical, or warm-temperate regions of the Atlantic or in the Ciulf of Naples. Jorgcnsen (1923, p. 13, 14), who states that this species is "known from many places in the warmer Atlantic, including the Florida Cur- rent," reports it from the Bay of Cadiz; Pavillard (1909, 1916) from the C.ulf of Lyons; Entz (1902b, 1905) from the Adriatic Sea ; Jorgensen (1923) from "through- out the whole of the Mediterranean (at fourteen out of thirty-seven statit)ns), everywhere in single specimens." In the Red Sea it has been found by Schroder (1900a); in the Ciulf of Aden by Ostenfeld and Schmidt (1901); in the Arabian Sea by Ostenfeld and Schmidt (1901), Cleve (1901a), and Schroder (1906a); in the Indian Ocean by Cleve (1901a) and Karsten (1907). 0.stenfeld and Schmidt (1901) found it in surface waters and at a depth of twenty feet; Karsten (1907) in a catch from 100 meters. Most of the remaining records mentioned in this para- graj^h probably refer to surface catches. Ostenfeld and Schmidt (1901) found the species in waters of 79.5° and 82.4°. Only Stein (1883) and Jorgensen (1923) give drawings by means of which the accuracy of their determinations may be judged. This is a eupelagic species occurring in ti'opical, subtropical, and warm- temperate regions of all seas. The Expedition records reveal an almost even distribution throughout the area investigated, with an oi)tinium habitat in deeper rather than surface waters, within the le\-els of photosynthesis. S^^TEMATIC ACCOUNT. Ill Phalacroma apicatum, sp. nov. Figure 10 Phalacroma sp. Okamura, 1907, p. 134, pi. 4, fig. 26a and b. Diagnosis: — Body irregularly obovate or sublozenge-shaped in lateral out- line, with subconical epitheca; deepest at or just behind girdle; length: depth, 1.24-1.30: 1; longitudinal axis deflected posteroventrally at 0°-10°. In ventral view subbiconical, with rounded apices, widest at girdle, 1.40 times longer than wide. Posterior cingular list 0.49-0.53 the length of body from apex. Lists resemble those of Phalacroma argits. Of the main ribs of left sulcal list R2 is Figure 10. — Phalacroma apicatum, sp. nov. 1, 3, 4, .5, in right lateral view; 2, in ventral view. Left sulcal list with traces of reticulation in specimen represented by 3; surface markings indicated only on small ])ortion of theca in 3. X 430. 1, 2, from type specimen, Station 4737 (300-0 fathoms); 3, from Station 4730 (300-0 fathoms); 4, from Station 4737 (300-0 fathoms); 5, from Station 4734 (300-0 fathoms). 0.09-0.17 and R3 is 0.12-0.17 the greatest depth of body. Surface markings of theca as in P. argus; sometimes apparently absent. Length, 91.8-102.2 /j. Tropical and subtropical Pacific. Description: — A rather large species with the body irregularly obovate or sublozenge-shaped in lateral outline, with subconical epitheca, deepest at or just behind the girdle, and 1.27 (1.24-1.30) times longer than deep. The longitudinal axis of the body is either perpendicular to the transverse furrow (Figure 10: 3), or it is deflected posteroventrally at an angle of 1°-10°. 112 THE DIXOPHYSOIDAE. The epitheca is subconical, about as deep as the hypotheca, the ratio being 0.99 (0.95-1.00): 1, highest in or just dorsally to the center and very prominent above the anterior cingular list; it has gently convex, almost straight or even slightly concave side contours, which are nearly perpendicular to each other, and rounded apex. The transverse furrow is flat or somewhat convex; and its width is 0.10 (0.14-0.19) the greatest height of the epitheca. The posterior cingular list is 0.50 (0.49-0.53) the length of the body from the apex. The hypotheca is sometimes almost symmetrical (Figure 10:3), but in most specimens it has a more or less pronounced postcroventral inclination (see above). It is somewhat irregularly, seldom regularly, ovoidal. The dorsal margin is evenly and gently convex. The ventral margin sometimes has about the same curvature as the dorsal, but in nearly all the specimens it is somewhat more strongly curved; in some, however, (Figure 10: 5) it forms a broadly rounded corner at the posterior end of the left sulcal list, in which case its anterior portion is but slightly convex. The posterior end of the body is broadly convex and always distinctly broader than the anterior. In \entral view (Figure 10: 2) the body is subbiconical, about 1.40 times longer than wide and widest at the transverse furrow; the side con- tours of the epitheca and of the hypotheca are slightly convex or almost straight ; and the anterior and posterior ends are rounded, the anterioi- slightly narrower than the posterior. The cingTilar hsts are as in PJwlacroma argus. The sulcus is about half as long as the hypotheca. The flagellar pore is about a girdle-width behind the junction of the cingulum and the sulcus. On the ventral side of the left valve a small pore is found near the sagittal suture, just in front of the anterior cingular list. The right sulcal list resembles that of P. argus, but it sometimes (Figure 10: 1) extends to the posterior main rib of the left sulcal list. The left sulcal list also resembles that of P. argus but is, on the average, somewhat wider. The ratios between the main ribs of this list and the greatest depth of the body are as follows: — anterior main rib, 0.13 (0.10-0.15): 1; fission rib, 0.12 (0.09-0.17): 1; posterior main rib, 0.14 (0.12-0.17): 1. Sometimes (Figure 10: 3) this list has a faint and irregular reticulation; and in some specimens its free margin is evenly convex throughout the whole length (Figure 10:4). The distance between the anterior main rib and the fission rib is 0.40 (0.33-0.50) the distance from the an- terior to the posterior main rib. The posterior main rib has a posterior inclina- tion of 42° (30°-50°). There are no accessory lists or sails. The surface markings of the thecal wall are the same as m Phalacroma argus; sometimes the reticulation seems to be absent. SYSTEMATIC ACCOUNT. 113 Megacytic stages have not been seen. The proportions of five of the Expedition specimens and of that drawn by Okamura (1907, fig. 26b) were measured. Dimensions: — Expedition specimens : Length of body, 91.8-102.2 ^ (average, 96.2 m; type, 101.5 m)- Greatest depth of body, 71.5-81.3 ^ (average, 75.5 m; type, 77.8 fx). Olcamm-a's (1907, fig. 26b) specimen: Length of body, 101 p. Greatest depth of body, 76 n. Variations: — This species is rather variable in the inclination of the longi- tudinal axis of the body, in the shape and relative height of the epitheca and of the hypotheca, in the shape of the right sulcal list, in the shape and structure of the left sulcal list, and in the surface markings of the thecal wall. The variations in the inclination of the longitudinal axis of the body and in the shape of the body probably are partly apparent and due to the tilting of some of the specimens examined. The epitheca generally is pronouncedly subconical, but sometimes specimens are found which in the shape of the epitheca approach, more or less, Phalacroma argus (Figure 10:5). The ventral margin of the left sulcal list is gently sigmoid in most specimens; in exceptional cases it is evenly convex (Fig- ure 10:4). Comparisons: — The structurally nearest species io Phalacroma apicatum is P. argus, from which it differs in its more contracted epitheca. For further dis- cussion, see p. 107. Phalacroma circumcinctum, which shows a rather striking resemblance to P. apicatum in the shape of the body, especially in dorsoventral view, differs in having the longitudinal axis of the body deflected posterodorsally, in having a parasagittal list, in the shape and structure of the sulcal lists, and in the surface markings of the thecal wall. The cjuestion as to whether or not these two species are closely related must be left open. The distinguishing characteristic of Phalacroma apicatnm, when compared with P. argus, i.e., the subconical shape of the epitheca, is found also in P. opercn- latum Stein s. str. (Stein, 1883, pi. 18, fig. 8), in one of the two specimens figured by Schiitt (1895, pi. 3, fig. 17: 2) under the name of P. vaslum var. acuta, in the rather highly differentiated P. turhinenm (Plate 2, fig. 3) and in P. reticulatum (Plate 4, fig. 3). In the case of P. operculatum and P. rastmn var. acuta {= P. acutum Pavillard), this similarity may be considered a sign of close relationship, i.e., as an expression of an inherent "tendency." On the other hand, the rela- tionships of P. turbineum and P. reticulatum to P. apicatum are uncertain. See also p. 191. 114 THE DINOPHYSOIDAE. Synonymy: — Okamura (1907, p. 134, fig. 26) suggested that his Phalacroma 8p. might represent P. porodictyum Stein. According to the size and the magni- fication of the drawing, this specimen was 101 /^ long. Occurrence: — Phalacroma apicatum is recorded at five of the 127 stations. There are 0, 0, 0, 1, 4, and 0 stations on the six lines of the Expedition. Of these five stations one (4G97) is in the Easter Island Eddy, and four (4730, 4732, 4734, 4737, type locality) are in the South Equatorial Drift. All the records refer to hauls from 300-0 fathoms. The temperatiu-e range of these five stations at the surface was 75.0°-81.5°; the average was 79.1°. The frequency in every case is less than 1 %. The species, which previously was known from Japanese waters, is probably eupelagic and widely distributed in tropical and subtropical seas. In the material of the Expedition it is limited to the Easter Island Eddy and the western portion of the South Equatorial Drift ; in other words, it has about the same distribution as Amphisolenia schauinslandi and A. thrinax and several other warm-water forms. Its average temperature is exceptionally high, 79.1°. Phalacroma circumcinctiim Kofoid and Michener Plate 1, fig. 4, 5. Figure 8: 3 Phalacroma circumcincta Kofoid & Michener, 1911, p. 288, 291. Diagnosis: — Body subovate in lateral view, deepest at girdle; length: depth, 1.25: 1; epitheca with slight ventral inclination, hypotheca almost symmetrical. In ventral view subbiconical, 1.54 tunes longer than wide, widest at girdle, with narrowly rounded apices. Posterior cingular list 0.46 the length of body from apex. Cingular lists without ribs. Right sulcal list reticulate. Left sulcal list : distance between posterior cingular list and Rs is 0.38 the length of body; Ro is 0.07 and Ra is 0.15 the greatest depth of body ; at R3 margin forms rounded corner at 90°; Ri is absent; R2 is T-shaped; R3 inclined posteriorly at 80°-85°. On right valve of epitheca and hypotheca there is an apparently structureless parasagittal list forming a direct continuation of left sulcal list; its maximum width but slightly exceeds half the width of transverse furrow. Theca densely areolate; 30-35 areoles border girdle posteriorly. Length 86.5 n. Eastern subtropical Pacific. Description: — A medium-sized species, with body subovate in lateral out- line, deepest at the girdle, and about 1.25 times longer than deep. The longi- SYSTEMATIC ACCOUNT. 115 tudinal axis of the body is deflected anteroventrally at an angle of about 2°, due to the ventral inclination of the epitheca. The epitheca is about as deep as the hypotheca and very prominent above the anterior cingidar list; it is inclined slightly ventrally; in other words, it is highest slightly ventrally to the center. Its ^•entral, anterior, and dorsal margins are confluent; the ventral and anterior are rather strongly convex; the dorsal is gently convex, somewhat flattened. The transverse furrow is slightly concave, and its width is about 0.26 the greatest height of the epitheca. The posterior cingular list is about 0.46 the length of the body from the apex. The hypotheca is almost symmetrical; its ventral, posterior, and dorsal margins are confluent; the posterior margin is somewhat more strongly convex than the \'entral and the dorsal. The posterior end of the body is somewhat narrower than the an- terior. In dorsoventral view the body is subbiconical, widest at the girdle, and about 1.54 times longer than wide; the side contours of the epitheca and of the hypotheca are gently concave or almost straight ; the anterior and posterior ends of the body are narrowly rounded or subtruncate, and of about the same width. The cingular lists are subhorizontal and subequal ; their width about ecjuals the width of the transverse furrow; and they appear to lack ribs. The sulcus is about half as long as the hypotheca. The flagellar pore is about a girdle-width behind the junction of the cingulum and the sulcus. On the ventral side of the left valve a small pore is found a short distance from the sagittal suture and just in front of the anterior cingular list. The right sulcal list extends to a point somewhat behind the fission rib of the left sulcal list; even in the widest region its ventral margin does not extend to the ventral margin of the left sulcal list; it is of almost equal width throughout the greater portion of its length; its free margin is gently sigmoid, being concave anteriorly and convex posteriorly. The whole list is reticulated, with two longitudinal rows of meshes, 6-8 meshes in each row. The left sulcal list is comparatively long and narrow; the distance from the posterior cingular list to the posterior main rib is 0.38 the length of the body; at the place of the anterior main rib the width of the list is about 0.10 the greatest depth of the body; the fission rib is about 0.07, the posterior main rib about 0.15 the greatest depth of the body; behind the posterior main rib the list decreases suddenlj' in width. The ventral margin of this list is gently sigmoid, being con- cave anteriorly and convex posteriorly; at the posterior main rib the margin forms a rounded corner at an angle of about 90°. The posterior margin of this list is almost straight. The anterior main rib of this list was not developed in the type specimen. The fission rib is T-shaped, being divided distally into two 116 THE DINOPHYSOIDAE. branches which form a marginal rib along the anterior half of the Ust. The posterior main rib is gently concave posteriorly and not club-shaped; it has a posterior inclination of about 80°-85°. The distance from the posterior cingular list to the fission rib is about 0.33 the distance between the posterior cingular list and the posterior main rib. With the exception of the main ribs, the left sulcal list appears to lack structural differentiation. On the right \'alve there is an apparently structureless parasagittal list. This list extends from the left sulcal list, of which it is a direct continuation, to the girdle, and along the anterior and ventral margins of the epitheca. On the epitheca the width of this list is some- what more than half the width of the transverse furrow; on the hypotheca the maximum width about ecjuals half the width of the transverse furrow. The thecal wall is areolate, and the areoles are of almost uniform size and evenly distributed. There are from thirty to thirty-five areoles on each valve bordering the posterior margin of the girdle. In the transverse furrow there are three irregular rows of areoles. Megacytic stages have not been seen. The proportions of the type were measured. Dimensions: — Length of body, 86. 5 ju. Greatest depth of body, 08.9 )u. Comparisons: — The above description is made from the oi'iginal drawings of the type specimen. The long, irregular, hyaline appendage of this specimen (Plate 1, fig. 5), apparently issuing from the sagittal suture near the posterior main rib of the left sulcal list, is of unknown nature and significance. Presum- ably it is a temporary structure, and it may be connected with reorganizing processes either independent of, or following binary fission; compare Kofoid (1908). This structure has not been considered of any systematic importance and thus has been omitted from the description of the species and disregarded in the discussion of the relationships of this form. The shape and the lack of struc- ture of the parasagittal list also appear to indicate that this specimen was taken while still in a process of skeletal reorganization (see Plate 3, fig. 4). This species has a somewhat uncertain position within the genus. It re- sembles P. apicatum in the somewhat elongate and, in dorsoventral view, sub- biconical shape of its body. Other features in which these two species agree are the subequatorial position of the girdle and the width and inclination of the cingular list; however, since these species presumably are relatively primitive in these respects, these similarities are of less importance in this connection. P. circumcincium is easily distinguishetl from P. apicatum by the following charac- ters:— {1) the epitheca is inclined ventrally instead of dorsally; (2) the right SYSTEMATIC ACCOUNT. 117 sulcal list is reticulate, and its free margin is gentlj- sigmoid and not angular; (3) the margin of the left sulcal list is angular at the posterior main rib; (4) the theca is areolate and not reticulate ; (.5) on the right valve of the epitheca as well as of the hj'potheca there is a well-developed parasagittal list. The differences be- tween these two species, therefore, appear to be more profound than the similari- ties. However, in spite of this, P. circumcinctum is probably rather closely related to P. apicatum and, preliminarily, should be placed in the Aegus group. Most of the characters of P. circumcinctum which are different from those of P. apicatum either have maintained a comparatively primitive type or they are expressions of "tendencies" inherent in the Argus group but not developed morphologically in most of the members of this group. For example, the angu- larity of the margin of the left sulcal list and the areolation of the thecal wall in P. circumcinctum probably are relatively primitive features; they are found also in P. porodictyum which is closely related to P. apicatum and P. argus. A para- sagittal list on the right valve also is characteristic of P. hlackmani, which un- doubtedly is a very close relative of P. striatum and thus also of P. apicatum and P. argus (see p. 107). The T-shaped fission rib of the left sulcal list is a character that is known only in Phalacroma circumcinctum and P. ovum. These two species resemble each other also in the width, structure, and inclination of the cingular lists, in the size and shape of the right sulcal list, and m the surface markings of the theca; they differ in the relative height of the epitheca, in the shape of the epitheca and of the hypotheca in dorsoventral view, in the relati^•e length and the inclination of the posterior main rib of the left sulcal list, and in the presence or absence of a parasagittal list. It also should be noted in this connection that Phalacrovm circumcinctum resembles P. limhalum in the subequatorial position of the girdle, in having a parasagittal list on the right valve of both the epitheca and the hypotheca, in the subbiconical shape of the body in dorsoventral view, and in the structure of the sulcus and of the adjacent regions (compare Plate 1, fig. 4, and Plate 3, fig. 3). Phalacroma limbatmn is easily distinguished from P. circumcinctum in the sub- circular shape of its body in lateral view, in the high structural differentiation of its right parasagittal list, in having two parasagittal lists, in the size and shape of its left sulcal list, in the structure of its theca, and in the narrowness of its body in dorsoventral view. Occurrence: — Phalacroma circumcincturn is recorded at only one (4671) of the 127 stations, on the second line of the Expedition and in the Peruvian Cur- 118 THE DINOPHYSOIDAE. rent, from 800-0 fathomi?, in a surface temperature of 66°, with a frequency of less than 1% (one specimen). Phalacroma ovum Schiitt Figure 11 Phalacroma ovum Schutt, 1895, p. 90. ^Phalacroma operculatum. Stein, 1883, partim, pi. 18, fig. 7. VFokti, 1922, ixirlim, p. 105, 190, 208, pi. 7, fig. 111. non Schutt, 1895, p. 93, pi. 2, fig. 10: 1-3. »io?( Jorcensen, 1923, p. 9, 10, fig. 7. Phalacroma opercvloides Schutt, 1895, partim, pi. 2, fig. 11:2. Okamura, 1912, p. 18, fig. 81, 82. JoR- GENSEN, 1923, p. 8, 9, 45, fig. 5. non Cleve, 1899(_:, p. -10. non Lohmann, 1902, p. 53. non Forti, 1922, p. 106, 190, 208, pi. 7, fig. 111. Diagnosis: — Body obovate in hit oral outhne, deepest at or somewhat behind girdle; length: depth, 1.20-1.31:1; longitudinal axis deflected posteroventrally at 0°-4°. In dorsal view narrowly oljovate, 1.59-1.64 times longer than wide. Posterior cingular list 0.28-0.36 the length of body from apex. C'ingular lists probably without ribs. Left sulcal list 0.43-0.53 the length of body; distance between Ri and R3 is 0.33-0.37 the length of body ; R, is 0.10-0.13, R3 is 0.30-0.37 the greatest depth of body; margin forms angle of 60°-90° at R3; R2 is T-shaped; R3 usually club-shaped and inclined posteriorly at about 30°. Theca finely and closely areolate. Length, 58.2-76.1 m- Widely distributed in tropical, subtropical, and warm-temperate seas. Description: — A medium-sized species, the body of which is obovate in lateral view, deepest at, or, in most specimens, somewhat behind the girdle, and about 1.26 (1.20-1.31) times longer than deep. The longitudinal axis of the body either is perpendicular to the girdle, or it is slightly (l°-4°) deflected postero- ventrally. The epitheca is 0.93 (0.91-0.96) as deep as the hypotheca, and moderately elevated or very prominent above the anterior cingular list; sometimes (Figure 11: 1) it is strongly and evenly convex, dome-shaped, sometimes (Figure 11 : 3, 2) of moderate convexity or rather fiat ; in most cases it is highest in the center, but in some specimens (Figure 11:6) somewhat dorsally to the center. The trans- verse furrow is flat or slightly convex, and its width is 0.37 (0.26-0.42) the great- est height of the epitheca. The posterior cingular list is 0.32 (0.28-0.36) the length of the body from the apex. The hypotheca sometimes is sj^mmetrical, sometimes it has a slight posteroventral mclination (see above); sometimes (Figure 11:4) it is of an oval wedge shape, with rather flat dorsal and ventral margins; sometimes (Figure 11: 1) it is regularly oval, with well or moderately rounded dorsal and \'entral margins ; its posterior margin is strongly convex and confluent with the dorsal and ventral margins. The posterior portion of the SYSTEMATIC ACCOUNT. 119 body is strikingly narrower than the anterior. In dorsoventral view the body is narrowly and regularly obovate, 1.59-1.64 times longer than wide, and widest at the girdle. The cingular lists are siibhorizontal, subequal, and about as wide as the transverse furrow; they appear to be without structural differentiation. The right sulcal list extends to a point midway between the fission rib and the poste- rior main rib of the left sulcal list, or it is slightly longer or shorter; in some specimens it is subtriangular, decreasing gradually in width posteriorly; in others it is subequal in width throughout its anterior half or even throughout the greater FiGUBE 11. — • Phalacroma ovum Schiitt. 1, 2, 3, iii right lateral view; 4, 6, in left lateral view; 5, in dorsal view. 2, from a megaoytic specimen; surface markings omitted. X 430. 1, from Station 4737 (300-0 fathoms); 2, from Station 4709 (300-0 fathoms); 3, from Station 4721 (300-0 fathoms); 4, 5, from Station 4681 (300-0 fathoms); 6, from Station 4709 (300-0 fathoms). portion of its length; anteriorly it is about as wide as the transverse furrow. The left sulcal list is of moderate length, narrow anteriorly and wide posteriorly. Its length is 0.50 (0.43-0.53) the length of the body, and the distance between the anterior and posterior main ribs is 0.35 (0.33-0.37) the length of the body. The anterior main rib, when present, is 0.12 (0.10-0.14), the fission rib 0.12 (0.10- 0.13), and the posterior main rib 0.33 (0.30-0.37) the greatest depth of the body. Behind the last-mentioned rib this Ust decreases suddenly in width. Between the anterior and posterior main ribs the free margin of this list sometimes is al- most straight or gently concave, sometimes it is gently sigmoid, concave anteri- orly and convex posteriorly; at the posterior main rib it forms an angle of 75° (60°-90°) ; behind the last-mentioned rib it is almost straight, or gently concave 120 THE DINOPHYSOIDAE. or convex. The fission rib is T-shaped, being divided distally into two branches, which form a marginal rib along the anterior third or half of the list; this marginal rib is sometimes more or less irregular. The posterior main rib is straight, or but slightly concave posteriorly, and in most cases somewhat club-shaped distally; it has a posterior inclination of 30° (25°-32°). Sometimes this list appears to have a faint reticulation (Stein, 1883, pi. 18, fig. 7). There are no accessory lists or sails. The thecal wall is finely and closely areolate and has scattered pores; the areoles are of about the same size as the pores (Jorgensen, 1923, fig. 5). Megacytic stages have been seen (Figure 11:2). Sometimes the megacytic specimen has the transdiameter subequal to the length of the body. The proportions of six Expedition specimens, antl of the specimens figured by Schiitt (1895) and Jorgensen (1923) were measured. Dimensions: — Expedition specimens: Length of body, 66.9-76.1 n (average, 71.4 n). Greatest depth of body, 55.0-58.9 ix (average, 56.9 ii). The type speci- men as figured by Schiitt (1895, pi. 2, fig. 11: 2) was 61 m long and 46 ^ deep; Jorgensen's (1923), 60-66 y. long. According to the given magnifications, the specimens figured by Okaniura (1912, pi. 5, fig. 81, 82) were about 58 y. and 34 n long, respectively (see p. 120). Varialions: — Judging by the available figures, this species appears to be rather constant. The most variable features are the size of the body, the relative height and the shape of the epitheca, and the shape of the two posterior main ribs of the left sulcal list. With regard to the porulation of the theca, see Phala- croma opcrculoidcs and P. opcrculaiu77i Jorgensen (1923, p. 8, 9). Comparisons: — Most of the Expedition specimens agree closely with the type specimen as figured by Schiitt (1895, pi. 2, fig. 11:2). They differ in having the fission rib of the left sulcal list T-shaped and the posterior main rib of this list club-shaped. It should be mentioned, however, that in some of these specimens the last-mentioned rib was not club-shaped. The specimen figured by Jorgensen (1923, fig. 5) under the name of Phala- croma operculoides shows also a close agreement with the type. Those figured by Okamura (1912, pi. 5, fig. 81, 82) are comparatively small; according to the magnifications given, they are only 58 ix and 34 n long, respectively. The last measurement may be erroneous, and due to a mistake in the magnification; it has been disregarded in the diagnosis. Judging by the simplicity of its structure, Phalacroma ovum probably is fairly closely related to the comparatively primitive members of this genus, e.g., SYSTEMATIC ACCOUNT. 121 P. porodidyum Stein, P. nperculoidcs Schiitt s. sir., P. operculalum (Stein) Jorgensen, and P. argus Stein. Due to the (simplicity of their organization, the interrelationships of these species are very difficult to establish. Ho\vc\er, it seems rather plausible that P. ovum is more closely related to P. porodidyum than to any other known species. Phalacroma ovum is most easily distinguished from all these species by the shape and structure of its left sulcal list. Phalacroma circumcindum, which differs very strikingly from P. ovum in the subequatorial position of its girdle, has the fission rib of the left sulcal list T- shaped in the same way as the last species. Judging by the great similarities in the size and shape of the body and in the shape and structure of the left sulcal list, Phalacroma ovum and P. doryphorum probably are rather closely related. The last species differs strikingly from the first in having a large, triangular posterior sail. Si/noni/my: — The sj-nonyniy of this species is complicated to such an extent that it is difficult, if not impossible, to unravel it. Stein (1883, pi. 18, fig. 7-10) figures, under the name of Phalacroma nperculaluvi, two forms which probably are specifically distinct. Jorgen.sen (1923, p. 9), the first to point out this fact, applies the name operculalum to the specimen represented by Stein (pi. 18, fig. 8). This specimen there- fore becomes the type of a species which should be named P. operculalum (Stein) Jorgensen. The speci- mens represented by Stein (i)l. 18, fig. 7, 9, 10) are assigned by Jorgensen (1923, p. 9) to P. porodidyum Stein. However, on account of the almost conijilete agreement between the specimen represented by Figure 11:1 and Stein's Plate IS, fig. 7, with regard to the shape of the body and the position of the girdle, it seems more plausible that these specimens of Stein belong to the species described and figured above under the name of P. arum Schiitt. The uncertainty of the a.ssignment is mainly due to the fact that the left sulcal li.st is incomplete (damaged ?) in Stein's Plate 18, fig. 7. The lack of porulation in the last figure probably is due to an omission; see Jorgensen (1923, p. 9). If our identification is correct, Stein (1883, pi. 18, fig. 7, 9, 10) was the first to figure a sjiecimen of P. arum, and P. operculalum Stein, parlim, is a synonym of P. arum Schiitt. Phalacroma opcrcidalum Stein is used by Daday (1888, p. 99; Phalarncoma operculalum, lapsu pcnnac), Schiitt (189.5, p. 93, pi. 2, fig. 10), Hensen "(1895, p. 190; 1911, p. 137, 138, 166, 167, .393, tab. 1.5), Leni- niermann (1899, p. 371; 1901a, p. .373; 190.5a, p. 3.5), Mvirray & Whitting (1899, p. .330, tab. 1, 2, 3, 4, .5, 6, 7, 8, 9), Schroder (1900a, p. 9, 11, 19; 1906a, p. 321, 327, 330; 1911, p. 17, 37), Rauschenplat (1900, p. 134), Cleve (1901a, p. 17; 1901c, p. 272; 1902b, p. 37; 190.3b, p. .347), Lohmann (1902, p. 53; 1920, p. 485, 492, 496, 504, 505, .507, 509, 511, 513, 515, 517, .524, 525, 529, 532, .536, .537, .544, .551, 554, 561, 572, 574, 585, 596), Entz (1902b, p. 94; 1904, p. 14; 1905. p. Ill), Pavillard (1905, p. .59, 81, 102; 1909, p. 283; 1916, p. 44, 47, 48, 49, .50, .54), Zacharias (1906, p. .509, 530, 534, 536, 544, 557), Karsten (1907, p. 236, 247, 263, 304, 349, 355, 359), Nathansohn (1908, p. 60; 1909. p. 42), Stiiwe (1909, p. 237, 252, 287), Schiller (1912, p. 27), Forti (1922, p. 105, 190, 208, pi. 7, fig. 110, 112; fig. 112 is erroneously called P. porodidyum, .and the figure representing the last species h.as no number), Jorgensen (1923, p. 9, 10, fig. 7), and Lindc- mann (1924, p. 10). Most of the investigators mentioned in this paragraph do not give any information as to which of the two forms figured by Stein (1883) under the name of P. opercidalum they refer to. Daday, Hen.sen, Ijemmermann, Murray and Whitting, Schroder, Rauschenplat, Cleve, Lohmann, Entz, Pavillard, Zacharias, Kar.sten, Nathansohn, Stiiwe, Schiller, and Lindemann have not been included in the bibliographical list of P. orum. The remaining ones give figures by means of which their determina- tions can be checked. Forti (1922) refers to both the forms figured by Stein (1883) under the name of P. operculalum. According to Jorgensen (1923, p. 8) "two apparently different species" are figured by Schiitt (1895, pi. 2, fig. 11: 1-3) under the name of Phalacroma operculoides. Jorgensen applies the name operculoidcs to the specimen represented by Schiitt (1895, pi. 2, fig. 11:2); the second figure was chosen because "un- fortimately, his [Schiitt's] first figure is uncertain, and seems to represent the same species w'hioh he notes, lower down on the same plate, as Phalacroma porodidyum var. parndum, while the second figure shows a characteristic species closely related to the foregoing [P. porodidyum], and possibly merging into the 122 THE DINOPHYSOIDAE. same" (Jorgensen, 1923, p. 8). It seems almost unquestionable that the specimens represented by Schiitt (1895, pi. 2, fig. 11: 1-3) belong to two specifically distinct forms. However, it is necessary to reject Jorgensen 's decision as to the nomenclature of these two forms, since Schiitt (1895, p. 90) refers to his Plate 2, fig. 11: 2 as P. ovum; on the other hand, except in the explanation of the plates, Schiitt never refers to this figure as representing P. npcrctdoides. The name P. operculoides therefore should be used for the species repre.seuted by Schiitt (1895, pi. 2, fig. 11:1, 3), while the name P. orum should be applied to the species represented by Plate 2, fig. 11:2. Phnlacroma operculoideK Schiitt has been used by several authors, viz., Cleve (1899c, p. 40), Lemmer- mann (1899, p. 372; 1901a, p. 373; 1905a, p. 35), Lohmann (1902, p. 53), Stiiwe (1909, p. 252, 287), Okamura (1912, p. 18, fig. 81, 82), Pavillard (1916, p. 54; 1923a, p. 879), Forti (1922, p. 106, 190, 208, pi. "7, fig. HI), and Jorgen.sen (1923, p. 8, 9, 45, fig. 5). P. operculoides Cleve (1899c) refers to P. rolunda- lutn, according to Cleve (1901c, p. 241); see also Paulsen (1908, p. 18). Lenimermann only refers to data given by other writers. The specimens referred by Lohmann (1902, p. 53) to P. npcrctdoiries were only 35 M long, which indicates that they are referable to P. npercnloidca x. xtr. Stiiwe (1909) and Pavillard (1916, 1923a) do not give any information as to which of the two forms figured by Schiitt (1895) under the name of P.opcrcidoides they refcrto. Under these circumstances these authors are not included in the bib- liography of P. ovum. Forti (1922) refers to Schiitt (1895, pi. 2, fig. 11: 1 ), not to P. ovum. P. operculoides Okamura (1912) and Jorgensen (1923) refer to P. ovum (see p. 120). Occurrence: — Phalacroma ovum is recorded at ten of the 127 stations. There are 2, 1, 1, 2, 4, and 0 stations on the six lines of the Expedition. Of these ten stations, one (4587) is in the Mexican Current; one (4617) in the Panamic Area; one (4657) in the Peruvian Current; and seven (4681, 4709, 4711, 4721, 4724, 4730, 4737) in the South Equatorial Drift. At one station (4724) the species was taken in a haul from 800-0 fathoms. All the remaming records refer to hauls from 300-0 fathoms. The temperature range of these ten stations at the surface was 68°-82° ; the average temperature was 75.8°. At one station (4657) the frequency is 1 %, at the remaining stations it is less. Schiitt (1895) did not give any information as to the localities of his speci- mens; however, his material probably was taken either at Naples, or hi the tropical or subtropical regions of the Atlantic. Jorgensen (1923) found this species "fairly common" in the Mediterranean, and "most frequent in the inner parts" of this sea. Okamura (1912) recorded it from Japanese waters. Judging by the numerous records which possibly may refer to Phalacroma ovum (see p. 121), this species appears to be eupelagic and widely distributed in tropical, subtropical, and warm-temperate seas. The ten record stations are distributed fairly evenly over the area investigated by the Expedition. The absence of surface records is noteworthy. The species probably has its optuuum habitat in deeper waters within the levels of photosynthesis. Phalacroma pyriforme, sp. nov. Figure 4:4, 5 Diagnosis: — Body obovate in lateral outline, deepest at girdle; length: depth, 1.13: 1. In dorsal view inverted pyriform, widest at girdle, broadly SYSTEMATIC ACCOUNT. 123 rounded anteriorly, narrowly rounded posteriorly; hypotheca with concave side contours; length: width, 1.30: 1. Posterior cingular list 0.43 the length of body from apex. Cingular lists somewhat wider than transverse furrow. Left sulcal Ust relatively large; its length 0.56 the length of body; throughout its anterior half it is of uniform width, 0.25 the depth of body; posteriorly it decreases gradu- ally in width; its free margin gently and evenly convex; structure unknown. Theca finely and closely areolate. Length, 48.2 m- Eastern tropical Pacific. Description: — A small species, with the body obovate in lateral outhne, deepest at the girdle. The ratio between the length and the depth is 1.13: 1. The longitudinal axis is about perpendicular to the girdle. The epitheca is about as deep as the hypotheca, highest in the center, evenly and strongly con\-cx, dome-shaped, and \-ery prominent above the anterior cingu- lar list. The transverse furrow is gently convex, and its width is 0.45 the greatest height of the epitheca. The posterior cingular list is about 0.43 the length of the body from the apex. The hypotheca is symmetrical; its dorsal and ventral margins are evenly and moderately convex and confluent with the posterior margin, which is strongly and evenly convex. The posterior portion of the body is strikingly narrower than the anterior. In dorsoventral view the body is m- verted pyriform in outlme, widest at the girdle, broadly rounded anteriorly, nar- rowly rounded posteriorly, and with the lateral contours of the hypotheca moder- ately concave; the ratio between the length and the width is 1.30: 1. The cingular lists are subhorizontal, subequal, and somewhat wider than the transverse furrow; their structure is unknown. The right sulcal list is about half as long as the left; its anterior half is subuniform in width, and about half as wide as the transversa furrow; posteriorly it decreases gradually in width. The left sulcal list is relatively large; its length is about 0.56 the length of the body; throughout its anterior half it is of uniform width and about 0.25 the greatest depth of the body; posteriorly it decreases gradually in width; its free margin is gently and evenly convex; in other words, it is not angular posteriorly, as in most species of the genus; its structure is unknown. There are no accessory lists or sails. The thecal wall is finely and closely areolate, about as in Pbalacroma dory- phorum. Pores have not been seen. Megacytic stages were not recorded. The proportions of the type were measured. Dimensions: — Length of body, 48.2 ix. Greatest depth of body, 42.5 m. 124 THE DINOPHYSOIDAE. Comparisons: — Phalncroma pi/riformc is established on two outline draw- ings made from a specimen found in the material of the Expedition. The species appears to occupy a rather isolated position. The symmetry and small size of its bod}', and the fine and close areolation of its thecal wall are features that affiliate it with the relatively primitive members of the genus, e.g., Phalacroma parmdum, P. operculoides s. str., and P. lens. In the shape of its body in lateral view and in its relatively high epitheca, it resembles P. argus, P. porodictyum, P. ovum, and some other obovoidal species. The shape of its left sulcal list recalls P. argus, P. apicatum, and the species of the Cuneiis group. The pyriform shape of its body in dorsoventral view is unique but approaches the cuneate shape of the species of the Cuneus and Rapa groups. Occurrence: — This species is recorded at one of the 127 stations. This sta- tion (4713) is on the fourth line of the Expedition and in the Galapagos Eddy. The depth is 300-0 fathoms, the surface temperature 73°, and the frequency less than 1% (one specimen). 4. Cuneus Group. Phalacmmn hlackmani is the only species of this group that is not included in the Expedition collections. Phalacroma cuneus Schiitt Figure 12, 13 Phalacroma emeus Schutt, 1895, p. 148, pi. 3, fig. 14; 1896, p. 27, fig. 38B. Ostenfeld, 1898a, p. 428. Murray & Whitting, 1899, p. 330, tab. 4. Lemmermann, 1899a, p. 372; 1901a, p. 372. Schroder, 1900a, p. 19; 1906a, p. 322, .327, 330. Cleve, 1901a, p. 16; lOOln, p. 270; 1902b, p. 3,5; 1903b, p. 347. Ostenfeld & Schmidt, 1901, p. 175. Zacharias, 1906, p. 534. Kofoid, 1907a, p. 195. Karsten, 1907, p. .325, 3.53, 355. Pavillard, 1909, p. 283; 1915a, p. 2; 1916, p. 47, 49, 52. Stuwe, 1909, p. 287. Okamura, 1912, p. 18, pi. 5, fig. 76. JoR(iEN.«EN, 1923, p. 11, fig. 11. Phalacrntiia cunciix .Stuwe, 1909, p. 252 {Inp.iiM pcnnae). Phalacroma mi/ra Oltmanns, 1922, partim, fig. 38: 1. Diagnosis: — Body cuneate in lateral outline ; epitheca very broadly rounded; posterior portion of hypotheca fairly brcadly to rather narrowly rounded; deepest at posterior cingular list; length: depth, 0.92-1.09: 1; longitudinal axis deflected posteroventrally at 4°-5°. In dorsal view cuneate ; posterior portion of hypotheca narrowly rounded to subacute; length: width, 1.22: 1. Posterior cingular list 0.23-0.28 the length of body from apex. Of- the cingular lists the anterior is reticulated, the posterior is ribbed. Left sulcal list 0.50-0.07 the length of body; distance from R, to R, is 0.40-0.47 the length of body; Ro is 0.09-0.11, and R3 is 0.07-0.10 the greatest depth of body; widest somewhat in front of R3; its ventral margin gently convex at R3, seldom subangular; R3 inclined posteriorly at 45°- SYSTEMATIC ACCOUNT. 125 60°; irregularly and incompletely reticulate. Theca reticulate; twenty-five to thirty-five polygons border girdle posteriorly. Length, 72-88 /j. Widely distributed in tropical, subtropical, and warm-temperate seas. Description: — A medium-sized species, the body of which is cuneate in lateral outline, with a rather low and very broadly rounded epitheca, with the posterior portion of the hypotheca fairly broadly to rather narrowly rounded, with the greatest depth at the posterior cingular list, and about as deep as long. The ratio between the length and the depth of the body is 0.98 (0.92-1.09): 1. The longitudinal axis is deflected posteroventrally at an angle of 4°-5°. The epitheca is about as deep as the hypotheca, very broadly and almost evenly convex, highest in or near the center, and moderately elevated above the anterior cingular list. The transverse furrow is flat to somewhat concave, and its width is 0.35-0.50 the gi'eatest height of the epitheca. The posterior cingular list is 0.23-0.28 the length of the body from the apex. The hypotheca tapers posteriorly and is somewhat deflected posteroventrally (see above); its dorsal margin is almost straight or slightly convex to concave; its ventral margin is nearly straight or gently convex; its postmargin is evenlj', and fairly broadly to rather narrowly convex, and confluent with the dorsal and ventral margins. In dorsoventral view the body is cuneate, with broadly rounded epitheca, and with the posterior portion of the hypotheca narrowly rounded to subacute; the side contours of the hypotheca are almost straight or slightly concave ; the ratio be- tween the length and the width is about 1.22: 1. The cingular lists are subhorizontal, subequal, and about as wide as the transverse furrow. Along the base of the anterior cingular list, there is a series of ribs anastomosing into a row of polygons, which have about the same size and shape as the polygons in the tj-ansverse furrow ; sometimes a more or less complete row of similar polygons is developed outside the basal row. The posterior cingu- lar list has a great number of straight and simple ribs (Schiitt, 1895, pi. 3, fig. 14: 2). The sulcus is about half as long as the hypotheca. The flagellar pore is about a girdle-width behind the junction of the cingulum and the sulcus. The right sulcal list is long and rather narrow; usually it ends at a point about half way between the fission rib and the posterior main rib of the left sulcal list, but sometimes it extends to the last-mentioned rib; anteriorly it is about as wide as or somewhat narrower than the transverse furrow, and posteriorly it tapers gradually. The left sulcal list is of moderate size. Its length is 0.50-0.(J7 the length of the body, and the distance between the anterior and posterior main ribs is 0.40-0.47 the length of the body. The anterior main rib, when pi'esent, is 126 THE DINOPHYSOIDAE. about 0.07-0.09, the fission rib 0.09-0.11, and the posterior main rib 0.07-0.10 the greatest depth of the bod3^ The greatest width of this list, which in most specimens is located somewhat in front of the posterior main ril), is about 0.12- 0.20 the greatest depth of tlic laody. In most specimens the free margm of this list is either gently sigmoid, concave anteriorly and convex posteriorly, or it is evenly to somewhat irregularly convex throughout; in other words, it is gently convex at the posterior main rib, and not angular as in most species of the genus. In exceptional cases (Figure 12:2) this margin is slightly convex between the anterior and posterior main ribs, and subangular at and gently concave behind the posterior main rib. Posteriorly this list either tapers more or less suddenly, or it is more or less decurrent (Figure 12: 2). The main ribs of this list are straight and rather weak; none of them is club->shaped, but the posterior sometimes is rather wide. The distance between the anterior main rib and the fission rib is 0.28-0.35 the distance between the anterior and posterior main ribs. The poste- rior main rib has a posterior inclination of 45°-(i0°. Besides the main ribs, the left sulcal list has an irregular and incomplete reticulation. At least in some specunens (Jorgensen, 1923, fig. 11), there is a very narrow parasagittal list ex- tending on the hypotheca from the left sulcal list to the posterior cingular list. The thecal wall has a reticulum of medium-sized polygons, and each polygon has a small pore in the center. Tlie polygons usually are about equal in size, and on each valve from twenty-fi\'e to thirty-five of them border the posterior margin of the girdle. In most specimens there are two rows of polygons in the transverse furrow, each row with twenty-five to thirty polygons on each valve; in other cases the two rows, partly or wholly, merge into a single row. Some of the polygons may be subdivided (Jorgensen, 1923, p. 12). From a point near the middle of the right sulcal list a spine-like process projects into the cytoplasm. Megacytic stages of this species have been seen by Pavillard (1910) and are contained in the Albatross collection. The dimensions of three specimens were measured. Dimensions: — Expedition specimens: Length of body, 80-87 /j (average, 84 fi). Greatest depth of body, 77-95 ^ (average, 86 yu). The largest of the speci- mens represented by Schiitt (1895, pi. 3, fig. 14: 1) is about 88 m long. The type specimen (Schiitt, 1895, pi. 3, fig. 14: 3) is somewhat smaller, but its length can- not be determined, since it is figiu'ctl in a tilted position. According to Jorgensen (1923, p. 11), the length is 72-88 /i. Variations: — This species appears to be rather remarkably constant in shape and in structm'e. A fairly large number of specimens from Station 4664 SYSTEMATIC ACCOUNT. 127 were submitted to a careful examination; most of these agreed very closely with the specimen represented by Figure 12; 1. The size and the shape of the body and of the sulcal lists are the most variable characters. The body sometimes is longer than deep, and sometimes deeper than long; sometimes (Figure 12: 1) the posterior portion of the hypotheca is rather broad, sometimes (Figure 12: 2) it is rather narrow; the dorsal margin of the hypotheca may be straight, or gently convex to concave. Posteriorly the left sulcal list either tapers more or less sud- denly (Figure 12: 1), or it is more or less decurrent (Figure 12: 2) ; its free margin Figure 12. — Phalacrotna cuneus Sehiitt, right lateral view. Structure of thecal wall and of left sulcal hst indicated onlj' in 1; 3 is somewhat tilted. X 430. 1, 3, from Station 4664 (300-0 fathoms); 2, from Station 4737 (300-0 fathoms). usually is gently concave anteriorly and convex posteriorly, but it maj' be ince versa. The variations probably are partly apparent, and due to the tilting of some of the specimens examined. Comparisons: — Sehiitt (1895) does not give any diagnosis or description of Phalacroma cuneus, and the specimens figured in lateral view by him are tlrawn in tilted positions. In spite of these discrepancies, the characteristics represented probably are sufficient for certainty of specific assignment. The Expedition specimens agree quite closely with these figured by Hchiitt (1895, pi. 3, fig. 14:1-8). Phalacroma cuneus is very closely related in structure to P. bluckniani, P. striatum., and P. giganteum. They all agree in the following respects: — their epitheca is rather low and very broadly rounded; their hypotheca is more or less cuneate in lateral as well as in dorsoventral y'ivw; in lateral x'ww the posterior portion of the hypotheca is fairly broadly to rather narrowly rounded, in dorso- ventral view it is narrowly rounded or even acute. The free margin of their left sulcal list is gently convex at the posterior main rib, and not angular as in most species of this genus. Without giving any reasons, Jorgensen (1923, p. 11) places Phalacroma hlackmani as a synonym of P. cuneus. Is this justifiable? According to the 128 THE DINOPHYSOIDAE. only drawing of this species available, tiz., that of the type specimen (Murray and Whitting, 1899, pi. 31, fig. 4), P. blackiiumi is about 125 m long, has a parasagittal list which extends on the hypotheca from the left sulcal list to the posterior cingu- lar list and is locally wider on the posterior portion of the hypotheca, and its left sulcal list lacks reticulation and has the main ribs concave anteriorlj'. P. cuneus has a maximum length of 88 fi, and its left sulcal list has an incomplete and iiTegu- lar reticulation and straight main ribs. At least in some specimens (Jorgensen, 1923, fig. 11), P. cuneus has, just as P. blackmani, a parasagittal list extending on the hypotheca from the left sulcal list to the posterior cingular list, but this para- sagittal list is very narrow throughout its whole length. Schroder (1906a, p. 330), Karsten (1907, p. 262, 473), and Stiiwe (1909, p. 238), who have recorded P. hlackmani, found also P. cuneus in the material examined by them. In short, according to the available descriptions and figures differences do exist between P. cuneus and P. hlackmani which appear to be of specific nature, and some in- vestigators have separated the two forms. Until further data are a\ailable, these two forms should be considered specifically distinct. Phalacroma cuneus differs from P. striatum in its smaller size (72-88 fi, as compared with 102-136 ;u) and its less developed left sulcal list. From P. giganteum it is easily distinguished by its smaller size (72-88 m, as compared with 148 ju). For further discussion about the relationships of these species, see p. 133. Synonymy: — The presence of a parasagittal list in the specimens deter- mined as P. cuneus by Jorgensen (1923) calls for a reexamination of this species to determine the prevalence of this structure which was not observed by Schiitt (1895) or by us. Occurrence: — Phalacroma cuneus is recorded at sixty-three of the 127 sta- tions. There are 9, 19, 7, 10, 11, and 7 stations on the six lines of the Expedition. Of these sixty-three stations, one (4583) is in the California Current; six (4587, 4590, 4594, 4598, 4005, 4545) in the Mexican Current; six (4009, 4013, 4017, 4634, 4637, 4038) in the Panamic Area; eighteen (4647, 4650, 4055, 4657, 4660, 4661, 4662, 4663, 4664, 4665, 4666, 4667, 4668, 4669, 4670, 4671, 4075, 4676) in the Peruvian Current; five (4689, 4691, 4695, 4697, 4699) in the Easter Island Eddy; one (4713) in the Galapagos Eddy; twenty-one (4679, 4681, 4683, 4701, 4705, 4706, 4707, 4709, 4711, 4719, 4724, 4728, 4730, 4731, 4732, 4733, 4734, 4736, 4737, 4739, 4740) in the South Equatorial Drift; three (4742, 4743, 4540) in the South Equatorial Current; two (4541, 4542) in the Equatorial Counter Current. There are thirteen records from the surface (Stations 4657, 4660 [Salpa], 4664, 4669, 4675, 4706 [Salpa], 4731, 4733, 4743, 4540, 4541, 4542, 4545); at ten of these SYSTEMATIC ACCOUNT. 129 stations the species was taken in surface hauls only; at three stations (4657, 4664, 4675) in hauls from 300-0 fathoms as well as at the surface. At one station (4737) the species is recorded from 100-0 fathoms and 300-0 fathoms; at one station (4713) from 150-0 fathoms and 300-0 fathoms; at three stations (4681, 4724, 4732) from 800-0 fathoms and 300-0 fathoms; at four stations (4G47, 4662, 4666, 4670) from 800-0 fathoms only. All the remaining records refer to hauls from 300-0 fathoms only. The species was taken also in surface waters in Acapulco Harbor, off the Mexican Current. This station is not included in the 127 stations mentioned above. The temperature range of these sLxty-three stations at the surface was 65°- 85°; the average was 74.7°. At the thirteen stations in the surface catches of which the species was found, the surface temperature ranged from 67° to 80°; the average was 74.5°. At Acapulco it was 83°. For the surface catches the following frequencies are recorded: — 4% at one station (4669); 3% at two stations (4542, 4545); 2% at one station (4675); 1% at two stations (4657, 4664) ; in the remaining cases the frequency was less than 1%. For the catches from 100, 150, 300, or 800 fathoms to the surface the rec- ords of frequency are as follows: — 40% at one station (4664); 10% at one sta- tion (4663) ; 6% at two stations (4613, 4666) ; 4% at one station (4662) ; 3% at two stations (4671, 4742); 2% at five stations (4609, 4634, 4689, 4713, 4740); 1% at twelve stations (4590, 4594, 4598, 4638, 4650, 4657, 4665, 4667, 4681, 4701, 4732, 4739) ; at the remaining stations the frequency is less than 1 %. For the catch made in Acapulco Harbor the frequency of 1 % is recorded. Schiitt (1895) does not give any information as to the type locality of Phala- croma cuneus; however, it is probably either Naples or in the Atlantic Ocean. Later investigators have shown this species to be widely distributed. The follow- ing authors have found it in the Atlantic, between lat. 40° N. and lat. 28° S. : — Ostenfeld (1898), Murray and Whitting (1899), Cleve (1901c, 1902b, 1903b), Stuwe (1909), and Jorgensen (1923). Cleve (1901c) recorded it from the Carib- bean Sea. Jorgensen (1923) found it at a fairly great number of stations tlirough- out the Mediterranean. It is also recorded from the Gulf of Lyons, Pavillard (1909, 1916). Naples, Schroder (1900a) ; Zacharias (1900). Ionian Sea, Schroder (1906a). Arabian Sea, Schroder (1906a); Cleve (1901a, 1903b); Ostenfeld and Schmidt (1901). Indian Ocean, Schroder (1906a); Cleve (1901a); Karsten (1907). Japan, Okamura (1912). According to the authors noted, tliis species occurs in waters of the following temperatures and salinities. 130 THE DINOPHYSOIDAE. Number and Number and Mean of Mean of Temperature Range Observations Salinity Range Observations Murray and Wliitting (1899) 79° Cleve (1901c) 56.8°- 28 73° 35.63- 22 36.49 81.5° 37.43 Cleve (1902b) 6 69.6° 6 36.24 Cleve (1903b) 60.3°- 6 75.7° 36.13- 6 36.40 82.0° 36.49 Stuwe (1909) 80.2° Only Schiitt (1895), Okamura (1912), and Jorgensen (1923) give figures or descriptions by means of which then- determinations of this form may be judged. FuiUHE 13. — Occurrence of PhaIacT07na ciineus Schiitt. Large, solid circles indicate records from ver- tical hauls; squares, record.s from surface hauls; triangles, records from both vertical and surface hauls; small, soHd circles, stations at which this species was not foimd; small, open circles, stations from which no plankton catches were examined. Phalacroma cuneus is eupelagic and widely distributed in tropical, sub- tropical, and warm-temperate seas. While previous writers have found it to be rare, it is one of the most abundant of the species of the genus found by the Expedition. In the Eastern Pacific it occurs throughout the whole area investi- gated. According to the records, there is a center of abundance in the relatively cool Peruvian Current . Out of the twenty-eight stations with frequencies of 1 % or more, not less than eleven are in that Current; and all the stations with fre- ciuencies of 4% or more are located there. The frequent occurrence of P. cuneus in the coolest (65°-72°) of the regions investigated is the more noteworthy, since SYSTEMATIC ACCOUNT. 131 the species has not been found as yet outside of tropical, sul)tropical, and warm- temperate waters. The relatively frequent occurrence of this species in our sur- face catches is another outstanding feature. Of the species of Phalacroma only P. doryphorum (twenty-six surface records) and P. rapa (twenty-two surface records) were found more frequently than P. cuneiis (thirteen surface records) in the surface catches of the Expedition. Of P. Jams, which next to P. cuneus was the most frequent species of this genus in the surface catches, there are only six surface records. Phalacroma striatitm Kofoid Plate 2, fig. 5, 8. Figui-e 14: 3; 15 Phalacroma slriala Kofoid, 1907a, p. 195, pi. 12, fig. 73. Kofoid & Michener, 1911, p. 289. Phalacroma striatum Jorgensen, 1923, p. 12, 43, fig. 12. Diagnosis: — Body cuneate in lateral view, with very broadly rounded epitheca, and with posterior portion of hypotheca fairly broadly rounded ; deepest at posterior cingular list; length: depth, 1.00-1.04: 1; longitudinal axis deflected posteroventrally at 6°-9°. In dorsal view cuneate; posterior portion of hypo- theca narrowly rounded to acute; length: width, 1.34-1.54: 1. Posterior cingular list 0.20-0.24 the length of body from apex. Cingular lists with numerous simple or branched ribs, or reticulate. Left sulcal list extends to or ahnost to antapex; distance from Ri to R3 is 0.57-0.63 the length of body; R2 is 0.13-0.18, and R3 is 0.06-0.16 the greatest depth of body; widest somewhat in front of R3; margin gently sigmoid, gently convex at R3; R3 inclined posteriorly at about 90°; irregu- larly and incompletely reticulate. Theca reticulate; 25-35 polygons border girdle posteriorly. Length, 102-136 p. Eastern tropical Pacific, Mediterranean, Guinea Current. Description: — This is a comparatively large species, the body of which is cuneate in lateral view, with a rather low and very broadly rounded epitheca, with the posterior portion of the hypotheca fairly broadly rounded, with the greatest depth at the posterior cingular list, and about as deep as long. The ratio between the length and the depth of the body is 1.02 (1.00-1.04) : 1. The longi- tudinal axis is deflected posteroventrally at an angle of 6°-9°. The epitheca is about as deep as the hypotheca, verj- broadly and usually almost evenly convex, highest at or somewhat dorsally to the center, and moder- ately elevated above the anterior cingular list. The transverse furrow is some- what concave, and its width is 0.35-0.50 the greatest height of the epitheca. The posterior cingular list is 0.22 (0.20-0.24) the length of the body from the apex. 132 THE DINOPHYSOIDAE. The hypotheca tapers posteriorly and is somewhat deflected posteroventrally (see p. 131); its dorsal margin is almost straight, slightly convex or concave, or slightly sigmoid, concave anteriorly and convex posteriorly; its ventral margin is almost straight, gently concave or convex, or slightly irregular (Figure 14: 3); its postmargin is evenly and fairly broadly convex, and confluent with the dorsal and ventral margins. In dorsoventral view the body is cuneate, with broadly rounded epitheca, and with the posterior portion of the hypotheca narrowly rounded to acute; the side contours of the hypotheca are gently concave in the middle, or gently sigmoid, concave anteriorly and convex posteriorly; the ratio between the length and the width is 1.34-1.54: 1. The cingular lists are subhorizontal and subequal; their width about equals or in some specimens slightly exceeds the width of the transverse furrow ; each of these lists has, on each valve, about 35^5 straight or more or less irregular ribs, which often anastomose and form a more or less regular reticulum (Jorgensen, 1923, p. 12, fig. 12). The sulcus is about 0.45-0.55 the length of the hypotheca. The flagellar pore is about a girdle-width behind the junction of the cingulum and the sulcus. On the ventral side of the left valve two or three rather small pores are found on the sagittal suture, just in front of the anterior cingular list (Plate 2, fig. 5). The right sulcal list extends to or somewhat beyond the point midwaj' between the fission rib and the posterior main rib of the left sulcal list ; in some specimens its ventral margin, even in the widest region, does not quite extend to the \'entral margin of the left sulcal hst, but in others it does. The anterior half to two thirds of the free margin of this list is fairly deeply concave and is, at least in some specimens (Plate 2, fig. 8), strengthened by a marginal rib; the posterior portion of this margin, which forms a distinct angle with the anterior portion, is more or less convex, or sometimes almost straight. The greatest height of this list is located just behind the concavity. Along the base of this list a reticulum of about the same type as that of the thecal wall may be developed. The left sulcal list is unusually large. In most specimens it extends to or almost to the antapex, and the distance between the anterior and posterior main ribs is 0.57- 0.63 the length of the body. The anterior main rib, when present, is about 0.08-0.10, the fission rib is 0.13-0.18, and the posterior main rib is O.OG-O.IG the greatest depth of the body. The greatest width of this list, which is located somewhere between the fission rib and the posterior main rib, is about 0.20-0.22 the greatest depth of the body. Anteriorly the free margin of this list is gently sigmoid, concave, or almost straight; posteriorly it is convex; in other words, this margin is gently convex at the posterior main rib, and not angular as in most SYSTEMATIC ACCOUNT. 133 species of the genus. The main ribs of this Hst are rather weak and not ckib- shaped or otherwise modified. The distance between the anterior main rib and the fission rib is 0.35-0.45 the distance between the anterior and posterior main ribs. The posterior main rib has a posterior inchnation of about 90°. Besides the main ribs, the left sulcal hst has a "more or less distinct structiu-e of incom- plete areoles, which often present the impression of irregular radial striation" (Jorgensen, 1923, p. 12). There are no accessory lists or sails. The thecal wall has a reticulum of medium-sized polygons, and each polygon has a small pore in the center; the reticulation often is "faint and thm-wallcd' (Jorgensen, 1923, p. 12). The polygons usually are about eciual in size, and on each valve about 25-35 of them border the posterior margin of the girdle. In most specimens there are two rows of polygons in the trans\erse furrow, each row with 25-30 polygons on each vahe. From a point near the middle of the right sulcal list a spine-like process projects into the cytoplasm. Megacytic stages have not been recorded. The proportions of three Expedition specimens were measured. Dimensions: — Expedition specimens: Length of body, 108-115^1 (average, 111 m; type, 108 m). Greatest depth of body, 105-114 ^ (average, 109 m; type, 109 n). Jorgensen (1923, p. 12) : "Length of cell (height) almost equal to breadth in dorso-ventral direction, or somewhat greater; the latter 102-136 yu." Variations: — Apart from variations in size, this species appears to be unusually constant. Cornparisons: — The description and figures of this species are based on the type material. Phalacroma striatum is structurally very close to P. cuneus, P. hlachmani, and P. giganteum (see P. cuneus, p. 127). It differs from P. cuneus in the larger size of its body (102-136 tx, as compared with 80-88 m), and in its larger left sulcal list. From P. blackmani it is easily distinguished by its lack of parasagittal lists, and by the large size and the structure of its left sulcal list. From P. giganteum it differs in the smaller size of its body (102-136 ju, as compared with 148 m), and in the large size and the structure of its left sulcal list. The four species mentioned in the last paragraph form a natural systematic group, the Cuneus group. The closest relative of Phalacroma striatum, outside of this group, presumably is P. argus. With regard to the characters in which the last two species resemble each other, see P. argus (p. 107). Phalacroma striatum is easily distinguished from P. argus by its low and broadly roundsd epitheca, its cuneate hypotheca, and its exceptionally large left sulcal list. The 134 THE DINOPHYSOIDAE. most striking feature of this relationship is that P. argus appears to be more closely related to P. striatum than to P. cuneus, although the latter species is smaller and has a smaller and less differentiated left sulcal list (compare, for in- stance, the directions of the posterior main ribs) and a simpler right sulcal list than P. striatum. The Cuneus group probably evolved from an ancestral form of about the same habitus as P. argus. Figure 14. — 1, I'halacronm giganitum Kofoid and Michencr, right lateral view of right valve of type specimen. X 430. Station 4734 (300-0 fathoms). 3, P/ratacroma s(riohrilra Schiitt as a .synonym of P. rapa Stein. Even though the extent of the confusion of Phalacroma rapa and P. milra by previous investigators caimot be fully established, there are reasons to assume that, with the exceptions mentioned in the last paragraph, it is not very great and that, therefore, the bibliographical list of /■*. rapa is fairly correct. This statement is ba.sed on the fact that P. rapa anil P. milra arc recorded as specifically distinct by most of the investigators, Ostenfeld and Schmidt (1901), Zacharias (190(i), Karstcn (1906, 1907), Stiiwe (1909), and Hensen (1911), who do not give any descriptions or figures by means of which their determinations may be checked. Occurrence: — Phalacroma rapa is recorded at fifty-three of the 127 stations. There are 17, 9, 10, 7, 6, and 4 stations on the six lines of the Expedition. Of these fifty-three stations, two (4580, 4583) are in the California Current; nine (4587, 4590, 4594, 4596, 4598, 4000, 4004, 4005, 4007) are in the Mexican Current; eleven (4009, 4013, 4015, 4017, 4019, 4024, 4034, 4035, 4037, 4040, 4044) hi the Panamic Area; six (4050, 4001, 4005, 4000, 4670, 4678) in the Peruvian Current; six (4689, 4691, 4692, 4695, 4697, 4099) in the Easter Island Eddy; seventeen (4079, 4680, 4081, 4083, 4085, 4701, 4705, 4709, 4711, 4719, 4721, 4722, 4724, 4737, 4739, 4740, 4741) in the South Equatorial Drift; two (4742, 4743) in the South Equatorial Current. There are twenty-two records from the surface (Sta- tions 4590, 4590, 4600, 4604, 4607, 4615, 4017, 4019, 4024, 4035, 4040, 4044, 4650, 4001 [Salpa], 4000, 4070, 4078, 4080, 4092, 4709 [Salpa], 4741, 4743); at eighteen of these stations the species was taken in surface hauls only; at four stations (4590, 4617, 4650, 4709) in hauls from 300-0 fathoms as well as at the surface. At one station (4737) the species is recorded from 100-0 fathoms and 300-0 SYSTEMATIC ACCOUNT. 145 fathoms; at four stations (4681, 4689, 4701, 4724) from 800-0 fathoms and 300-0 fathoms. All the remaming records refer to hauls from 300-0 fathoms only. The temperature range of these fifty-three stations at the surface was 67°- 85°; the average was 76.2°. At the twenty-two stations in the surface catches of which the species was found, the surface temperature ranged from 67° to 84°; the average was 76.0°. For the surface catches the following frequencies are recorded : — 2% at three stations (4615, 4624, 4650), 1 % at three stations (4600, 4619, 4666), and less than 1 % in the remaining cases. For the catches from 100, 300, or 800 fathoms to the Figure 17. — Occurrence of Phalacroma rapa Stein. Large, solid circles indicate records from verti- cal hauls; squares, records from surface hauls; triangles, records from both vertical and surface hauls; small, solid circles, stations at which this sjiecies was not found; small, open circles, stations from which no plankton catches were examined. surface the records of frequency are as follows: — 5% at one station (4709), 3% at one station (4695), 2% at one station (4697), 1 % at seven stations (4598, 4613, 4617, 4681, 4689, 4701, 4721), and less than 1 % at the remaining stations. The species was first recorded by Stein (1883) "aus dem Atlantischen Meer und der Siidsee." Hensen (1895, 1911) found it at a great number of stations throughout the tropical, subtropical, and warm-temperate regions of the Atlantic ; Cleve (1901c, 1902b, 1903b), at a fairly great number of stations in the Atlantic, between lat. 42° N. and lat. 20° S., in the Caribbean Sea, in the Arabian Sea, and the Indian Ocean and recorded it at lat. 36°-37° N. and long. 10° E.-5° W.; 146 THE DINOPHYSOIDAE. Karsten (1906) in the Atlantic, at lat. 5° N. and long. 13° W.; Stiiwe (1909) off the west coast of Africa, between lat. 4° N. and lat. 9° N.; Jorgensen (1923) in the Bay of Cadiz, off the coast of Portugal, and southwest of Brittany, and at a fairly great number of stations in various parts of the Mediterranean; Pavillard (1916) in the Gulf of Lyons; Forti (1922) in the Ligin-ian Sea; Schroder (1900a) at Naples; Zacharias (1906) at Naples and in the Adriatic ; Entz (1902b, 1905) in the Adriatic Sea. Ostenfeld and Schmidt (1901) found it in the Gulf of Aden; Karsten (1907) in the Indian Ocean; Whitelegge (1891) on the east coast of Australia; and Okamura (1907) in Japanese waters. According to the authors noted, this species occurs in waters of the following temperatures and salinities. Number and Number and Mean of Mean of Temperature Range Oljservations Salinity Range Observations Cleve (1901c) 56.0°-82.0° 39 70.0° 33.44-37.63 28 35.80 Cleve (1902b) 72.6° 12 36.27 Cleve (1903b) 59.2°-83.0° 13 70.0° 36.02-37.29 13 36.59 Ostenfeld and Schmidt (1901) 791)° Stuwe (1909) 80.6°-82.8° Of the writers who have contributed to our knowledge of the distribution of this species, only Stein (1883), Okamura (1907), Pavillard (1916), Forti (1922), and J5rgensen (1923) give descriptions or figures by means of which their de- terminations may be judged. It should be mentioned that Cleve (1901c, p. 273) gives Phalacroma mitra Schiitt as a synonym of P. rapa Stein. Some of the data referred to P. rapa by Cleve therefore may be referable to P. mitra. This is a eupelagic species occurring in tropical, subtropical, and temperate regions of all seas. According to the Expedition records it is almost evenly dis- tributed throughout the area investigated. It probably has its optimum habitat in deeper waters, within the levels of photosynthesis, but it appears to occur more frequently at the surface than most of the other species of Phalacroma. Indeed, of the species of this genus only P. doryphorum (twenty-six surface records) was found more frequently than P. rapa (twenty-two surface records) in the surface* catches. Of P. cuneus and P. f aims, which next to P. rapa were the most frequent in the surface catches, there are only thirteen and sLx records, respectively. Phalacroma favus Kofoid and Michener Plate 2, fig. 7. Figure 14: 4, 5 Phdlacrnmnfarus Kofoid & MicheNeh, 1911, j). 289. P.will.ujd, 1923a, p. 879. Phalacroma hindmarchii Pavill.\rd, 1916, p. 53; of. 1923a, p. 879. ?/'//(;/fjrro«ia /rtrjt-s Jorgensen, 1923, p. 15, 16, 43, fig. 16. Phalacroma simulans Jobgensen ms., 1923, p. 15. SYSTEMATIC ACCOUNT. 147 Diagnosis: — Body subcuneate in lateral view; epitheca verj' broadly rounded; posterior portion of hypotheca constricted, mammilliform; ventral margin of hypotheca broadly rounded or somewhat angular at R3 of left sulcal list; deepest at posterior cingular list; length: depth, 1.14-1.21: 1; longitudinal axis deflected posteroventrally at 5°-10°. In dorsal view cuneate ; hypotheca \'ery narrowly rounded to subacute posteriorly, and its side contours gently undu- lating; length : width, 1.45-1.50: 1. Posterior cingular list 0.21-0.27 the length of body from apex. Left sulcal list 0.50-0.65 the length of body; distance from Ri to R3 is 0.25-0.31 the length of body; unusually decurrent behind R3; R-. is 0.12-0.14, and R3 is 0.14-0.22 the greatest depth of body; margin forms angle of 100°-130° at R3; R3 inclmed posteriorly at 5°-30°; with about 8-15 short riblets. Theca reticulate; 20-25 polygons border girdle posteriorly. Length, 54-83 m- Probably widely distributed in tropical, subtropical, and warm-temperate seas. Description: — A medium-sized species, the body of which is subcuneate in lateral view, with very broadly rounded epitheca, with the posterior portion of the hypotheca constricted and mammilliform, with the ventral margin of the hypotheca broadly rounded or somewhat angular at the posterior main rib of the left sulcal list, and with the greatest depth at the posterior cingular list. The ratio between the length and the depth of the body is 1.18 (1.14 1.21) : 1. The longi- tudinal axis is deflected posteroventrally at 5''-10°. The epitheca is about as deep as the hypotheca or slightly less, very l)roadly convex, highest in or near the center, and moderately elevated above the anterior cingular list. The transverse furrow is somewhat concave, and its width is 0.50- 0.67 the greatest height of the epitheca. The posterior cingular list is 0.24 (0.21- 0.27) the length of the body from the apex. The hypotheca is somewhat deflected posteroventrally (see above). The dorsal margin, from the girdle to the antapex, is gently undulating, convex in the middle, and concave anteriorly and posteri- orly; or its anterior half is almost straight and its posterior half gently concave (Jorgensen, 1923, fig. 16). The portion of the ventral margin that is in front of the posterior main rib of the left sulcal list is almost straight or gently convex or concave, and about perpendicular to the girdle. The portion of this margin that is behind the posterior main rib of the left sulcal list is 1.7-2.7 the length of the anterior portion, and moderately to rather strikingly concave. At the mentioned rib this margin is either broadly rounded or somewhat angular. The posterior portion of the hypotheca is constricted and mammilliform, and the antapex is narrowly rounded to subacute. In dorsal view the body is cuneate, with very 148 THE DINOPHYSOIDAE. broadly convex epitheca and with the posterior portion of the hypotheca narrowly rounded to subacute; the side contours of the hypotheca are gently undulating, convex in the middle, and concave anteriorly and posteriorly; the ratio between the length and the width is 1.45-1.50: 1. The cingular lists are subhorizontal, subequal, and about as wide as or slightly narrower than the transverse furrow; each of them has, on each valve, twenty to thirty rather strong, straight, simple, and nearly equidistant ribs. The sulcus is about 0.30-0.40 the length of the hypotheca. The flagellar pore is at the junction of the sulcus and the cingulum. The right sulcal list extends to the posterior main rib of the left sulcal list; in some specimens it is subtriangular, decreasmg gradually in width posteriorly, in others it is of subequal width through- out the greater portion of its length ; anteriorly it is about as wide as or somewhat narrower than the transverse furrow. The left sulcal list is of moderate size. Its length is 0.50-0.65 the length of the body, and the distance between the anterior and posterior main ribs is 0.25-0.31 the length of the body. It extends to the base of the posterior mammilliform projection of the bod}'. The anterior main rib is 0.08-0.11, the fission rib 0.12-0.14, and the posterior main rib 0.14-0.22 the greatest depth of the body. Behind the posterior main rib this Ust is unusually decurrent. Between the anterior and posterior main ribs, the free margin of this list is nearly straight, gently convex, or gently concave between the anterior main rilj and the fission rib as well as between the fission rib and the posterior main rib; at the posterior main rib it forms an angle of 100°-130°; behind the last- mentioned rib it is straight or gently concave. The main ribs of this list generally are of moderate strength and straight or almost so; in exceptional cases, for in- stance, in the type specmien, the posterior rib is club-shaped. The distance from the anterior main rib to the fission rib is 0.33-0.50 the distance between the an- terior and posterior main ribs. The posterior main rib has a posterior mclination of 5°-30°. Besides the main ribs, this list has about eight to fifteen short riblets. At least in some specimens (Jorgensen, 1923, fig. 16) the left sulcal list is con- tinued by a very narrow parasagittal list extending around the posterior portion of the hypotheca. The thecal wall has a reticulum of medium-sized, subequal polygons, each with a small pore in the center. There are twenty to twenty-five of them border- ing the posterior margin of the girdle and two rows in the transverse furrow, each with twenty to twenty-five polygons on each valve. Megacytic stages were seen by Jorgensen (1923, p. 16). The proportions of two of the Expedition specimens and the one represented by Jorgensen (1923) were measured. SYSTEMATIC ACCOUNT. 149 Dimet^sions: — Length of body, 72-76 m (average, 74.3 m; type, 76 m). Great- est depth of body, 62.7-63.6 m (average, 63.2 )x; type, 62.7 m). The specimens measured by Jorgensen (1923, p. 15) formed two size classes, the one 81-83 m long and 68-71 m deep, the other 54-62 /i long and 54 n deep. Variations: — Judging by the material thus far exainined, this species ap- pears to be fairly constant. Most of the specimens examined and the specimen represented by Jorgensen (1923, fig. 16) are strikingly similar. The most variable characters are the size of the body and the shape of the posterior portion of the hypotheca. Comparisons: — The description and figures of Phalacroma fatms are based on the type material. Phalacroma rapa probably is structurally the closest relative of P. Jams. Indeed, these two forms sometimes are so similar that even their specific differ- entiation is difficult; compare Pavillard (1916, fig. 13A) and Figure 14: 5. How- ever, we do not agree with Jorgensen's (1923, p. 16) statement that P. faints is so closely related to P. rapa "that there might be some grounds for regarding it as a variety of that species." The most important distinguishing characteristics are : — (1) P. f aims has a relatively longer and more decurrent left sulcal list, which has a fairly great number of short riblets besides the three main ribs; {2) the posterior portion of the hypotheca in P. famis is somewhat more contracted, i.e., mam- milliform, than in P. rapa. P. favus is most easily distinguished from P. hind- marchi by its lower epitheca, its relatively longer and more decurrent left sulcal list, which has a fairly great number of short riblets besides the three main ribs, an(4 the polygonate reticulation of its theca. It differs from P. mitra, P. doKchop- terygium, and P. acutum in the shape and structure of its left sulcal list and in the mammilUform shapeof the posterior portionof its hypotheca. See P. rapa, p. 143, and Jorgensen, 1923, p. 15, 16. Occurrence: — Phalacroma favus is recorded at twelve of the 127 stations. There are 6, 1,0, 0, 2, and 3 stations on the six lines of the Expedition. Of these twelve stations, six (4587, 4590, 4592, 4594, 4604, 4545) are in the Mexican Cur- rent; one (4617) is in the Panamic .\j-ea; one (4665) is in the Peruvian Current; three (4730, 4737, 4741) are in the South Equatorial Current; one (4542) is in the Equatorial Counter Current. At one of these stations (4617) this species was taken in a surface haul as well as in a haul from 300-0 fathoms; at five stations (4592, 4604, 4741 [Salpa], 4542, 4545) in surface hauls only. All the remain- ing records refer to hauls from 300-0 fathoms only. The temperature range of these twelve stations at the surface was 68°-84°; 150 THE DINOPHYSOIDAE. the average was 80.2°. At the six stations in the surface catches of which this species was found, the surface temperature ranged from 78° to 84°; the average was 80.8°. At two stations (4604, 4542) the frequency is 1%; in the remaining cases it is less. The species was first recorded by Kofoid and Michener (1911) at Station 4737 of the Expedition. Later Jorgensen (1923) found it in the Bay of Cadiz and in the Ionian Sea. Pavillard (1916) reported it from the Gulf of Lyons under the name of Phalacroma hindmarchii. This is a eupelagic, stenothermal, and warm-water species. Presumably it is widely distributed in tropical and subtropical seas, and it occurs only occasion- ally in warm-temperate regions. According to the Expedition records, it is found in waters of very high average temperature (80.2°). Most of the record stations are in the warm Mexican Current, and in the western portion of the South Equatorial Drift, and only one is located in the relatively cooler Peruvian Cur- rent. Another outstanding peculiarity in the occurrence of this species, according to the Expedition data, is the exceptionally high percentage (50%) of surface records. In no other species of this genus has such a high percentage been re- corded. Phalacroma hindmarchi Murray and Whitting Figure 18, 19 Phalacroma hindmarchii Murray & WuiTTiNci, 1899, p. 330, tab. 4, 5, 6, S, 9, i)l. 31, fig. .5. Lemmermann, lS99a, p. 372; igoi.a, p. 372. Cleve, 1901o, p. 271; 1902b, p. 36. .Jorgensen, 1923, p. 1.5, 16. Pavil- L.\RD, 192.3a, 1). 879 (non 1916, p. 53). Phalacroma hindmarchi Stuwe, 1909, p. 251, 288. Diagnosis: — Body of inverted fig-shapo in lateral view; epitheca broadly and more or less strongly convex; posterior portion of hypotheca constricted, mammilliform; ventral margin sigmoid, convex anteriorly, concave posteriorly; deepest at or somewhat behind girdle; length: depth, 1.15-1.31: 1; longitudinal axis usually perpendicular to girdle. Dorsoventral and lateral \icws almost similar. Posterior cingular list 0.27-0.44 the length of body from apex. Cingular lists possibly without structure. Left sulcal list 0.33-0.50 the length of body; distance from Ri to R3 is 0.24-0.33 the length of body; R2 is 0.09-0.16, and R3 is 0.20-0.25 the greatest depth of body; margin forms angle of 70°-100° at R3; R3 inclined posteriorly at 15°-35°; often club-shaped; .sometimes reticulate. Theca finely and closely areolate. Length, 82-98 m- Probably widely distributed in tropical, subtropical, and warm-temperate seas. SYSTEMATIC ACCOUNT. 151 Description: — A medium-sized species, the body of which is of inverted fig- shape in lateral view, with broadly and more or less strongly convex epitheca, with the posterior portion of the hypotheca constricted and mammilliform, with the ventral margin of the hypotheca sigmoid, convex anteriorly and concave posteriorly, and with the greatest depth at or somewhat behind the girdle. The ratio between the length and the depth of the body is 1.25 (1.15-1.31): 1. The longitudinal axis usually is about perpendicular to the girdle, but sometimes it is slightly deflected posterodorsally or posteroventrally. Figure 18. — Plialacroma hindmarchi Murray and Whitting. 1, right, lateral view. 2, 3, left lateral view; surface markings, areolation, and porulation indicated only on small portion of theca in 2, 3. X 430. 1, from Station 47.37 (1.50-0 fathoms). 2, 3, from Station 4730 (300-0 fathoms). The epitheca is about as deep as the hypotheca or somewhat less, broadly and more or less strongly convex, often dome-shaped, highest in or near the center, and moderately elevated or generally \'ery prominent above the anterior cingular list. The transverse furrow is nearly flat or slightly convex, and its width is 0.20-0.40 the greatest height of the epitheca. The posterior cingular list is 0.38 (0.27-0.44) the length of the body from the apex. The hypotheca some- times is almost symmetrical, as in the type specimen and in Figure 18: 1; some- times its dorsal and ventral margins, which are sigmoid, convex anteriorly and concave posteriorly, are more or less different (Figure 18: 3). The posterior por- tion of the hypotheca is constricted and mannnilliform ; in none of the Expedition specimens was this constriction ciuite so pronounced as in the type (Murray and Whitting, 1899, pi. 31, fig. 5). The antapex is narrowly rounded to subacute. In dorsoventral view the body has about the same shape as in lateral view, with the exception that it is somewhat narrower. The cingular lists are subhorizontal, subequal, and about as wide as the transverse furrow; judging bj' the figures of the type and by Figures 18, 19, these lists are hyaline and without structure. The sulcus is about 0.4 the length of the hypotheca. The right sulcal list extends to the posterior main rib of the left 152 THE DINOPHYSOIDAE. siilcal list, is of subequal width throughout the greater part of its length, as wide as or somewhat narrower than the transverse furrow, and rounded posteriorly. The left sulcal list is of moderate size. Its length is 0.33-0.50 the length of the body, and the distance between the anterior and posterior main ribs is 0.24-0.33 the length of the body. It ends somewhat in front of the base of the posterior mammilliform projection of the body. The anterior liiain rib, when present, is 0.07-0.14, the fission rib 0.09-0.16, and the posterior main rib 0.20-0.25 the greatest depth of the body; behind the posterior main rib the list decreases sud- denly in width. Between the anterior and posterior main ribs, the free margin of this list is almost straight, gently convex or concave, sigmoid, i.e., concave an- FiGURE 19. — Occurrence of Phnlncrotiia hhidinarchi Murray and Wliitting. Large, solid circles indi- cate records from vertical hauls; squares, records from surface hauls; small, solid circles, stations at which this species was not found; small, open circles, stations from which no plankton catches were examined. teriorly and convex posteriorly, or it is somewhat irregular; at the posterior main rib it forms an angle of 70°-100°; behind the last-mentioned rib it is either straight or gently concave or convex. The anterior main rib and the fission rib of this list are of moderate strength or rather weak; the posterior main rib usually is quite heavy and club-shaped distally; all of them are straight or almost so, or the pos- terior has a gentle posterior concavity. The distance between the anterior main rib and the fission rib is about 0.33-0.50 the distance from the anterior to the posterior main rib. The posterior main rib has a posterior inclmation of 15°-35°. SYSTEMATIC ACCOUNT. 153 Besides the main ribs, this list sometimes has a well-developed reticulation (Figure 18: 3). There are no accessory lists or sails. The thecal wall has a fairly great number of scattered pores, interspersed with more numerous, small, rounded, and closely set areoles of somewhat varying size (Figure 18: 2, 3). Usually the areoles are of about the same size as the pores, but sometimes they are smaller (see Cleve, 1901c, p. 271). Megacytic stages were found. The proportions of five of the Expedition specimens and of the type speci- men as represented by Murray and Whitting (1899) were measured. Dimensions: — Length of body, 82-98 m (average, 87 n). Greatest depth of body, 65-75 m (average, 69.5 m). The type specimen as represented by Murray and Whitting (1899, pi. 31, fig. 5) was about 92 /x long and 75 n deep. Variations: — The length of the body, the relative height of the epitheca, the degree of symmetry and of posterior constriction of the hypotheca, and the structure of the theca and of the left sulcal list appear to be the most variable characters. Co7nparisons: — Most of the Expedition specimens referred to Phalacroma hindmarchi (Figure 18: 1) agree closely with the type as figured by Murray and Whitting (1899, pi. 31, fig. 5). Others are more or less aberrant (Figiu'e 18: 2, 3), but even in their case the assignment may be regarded as certain. Phalacroma hindmarchi is closely related to P. Jams, P. rapa, P. miira, and P. dolichopterygium, although this relationship appears to be collateral rather than linear. In other words, P. hindmarchi seems to be a highly differentiated member of an evolutionary branch originating from the same ancestral form as the branch represented by the last four species and partly embodying the same tendencies as the latter branch, e.g., the tendency to constrict the posterior por- tion of the hypotheca into a mammilliform projection. Phalacroma hindmarchi is easily distinguished from P. fainis, P. rapa, P. mitra, and P. dolichopterygitim by its higher and more dome-shaped epitheca, and by having its thecal wall finely areolated instead of reticulated. According to Murray and Whitting (1899, p. 330), Phalacroma hindmarchi is "closely allied to P. operculatwn." However, this relationship is not so evident as the ones suggested above. Occurrence: — Phalacroma hindmarchi is recorded at thirteen of the 127 sta- tions. There are 0, 0, 2, 3, 7, and 1 stations on the six lines of the Expedition. Of these thirteen stations, four (4689, 4691, 4697, 4699) are in the Easter Island Eddy; eight (4701, 4730, 4731, 4732, 4734, 4736, 4737, 4739) are in the South ] 54 THE DINOPHYSOIDAE. Equatorial Drift ; one (4742) is in the South Equatorial Current. At one of these stations (4731) the species is recorded from the surface; at one station (4737) from 100-0 fathoms as well as from 300-0 fathoms. The remaining records refer to hauls from 300-0 fathoms only. The temperature range of these thirteen stations at the surface was 72.0°- 81.5°; the average was 77.2°. At Station 4731, the only surface record, the surface temperature was 79.5°. At four stations (4689, 4731 , 4734, 4739) the frequency is 1 % ; m the remain- ing cases it is less. The species was first recorded by Murray and Whitting (1899), who found it at nine stations in the tropical and subtropical regions of the Atlantic, between lat. 14° N. and 31° N., and in the Caribbean Sea. Later Cleve (1901c, 1902b) reported it from the Atlantic, between lat. 9° N. and 34° N., and from the Carib- bean Sea; and Stiiwe (1909) from the Atlantic, at lat. 3° 50' N., long. 26° 15' W. (South Equatorial Current). According to these authors, this species occurs in waters of the following tem- peratures and salinities. Number and Number and Mean of Mean of Temperature Range Observations Salinity Range Observations IMurray and Whitting (1899) 67''-G9° Cleve (1901e) 70.0°-82.0° 16 75.5° 34.88-37.4.3 10 36.27 Cleve (1902b) 76.3° 37.17 Stiiwe (1909) 80.2° Of the authors who have contributed toward a knowledge of the distribution of this species only Murray and Whitting (1899) give any descriptions or figures by means of which their determinations may be checked. Jorgensen (1923, p. 16) writes as follows: "I have not noticed that the species I have noted from the cruise of the Thor as Ph. Hindmarchii ..." However, P. Inndmarchi is not mentioned by this author except in connection with the dis- cussion of P. favus. The statement quoted above therefore is either due to a lapsus pennae, or the account of P. hindmarchi was overlooked. This species is eupelagic and widely distributed, but rare, in tropical, sub- tropical, and warm-temperate seas. Its distribution in the Ea.stern Pacific, according to the Expedition data, is remarkably similar to that of Amphisolenia schauinslandi and A . thrinax (see p. 442). Although there are as many as thirteen record stations, the species was never found in the California Current, Mexican Current, Panamic >Area, and Peruvian Current. With one exception, all the record stations are in or near the Easter Island Eddy and in the western portion of the South Equatorial Drift. The average temperature of its habitat is high, 77.2.° SYSTEMATIC ACCOUNT. 155 6. ExPULSUM Group. Both the species referred to this group, PhaJacroma protuberans and P. expulsuin, were found in the material of the Expedition. Phalacroma protuberans, sp. nov. Figure 20: 6-9 Diagnosis: — Body subovate to subcircular in lateral outline, with dorsal shoulder-like constriction at posterior cingular list, deepest in or somewhat in front of the middle, and 1.08-1.18 times longer than deep; longitudinal axis per- pendicular to girdle or deflected posteroventrally at l°-4°. In dorsal view wedge- shaped, widest just behind girdle, subacute to narrowly rounded posteriorly, and with more or less pronounced rounded protuberances just behind girdle; length: width, 1 .38-1 .74 : 1 . Posterior cingular list 0.23-0.26 the length of body from apex. Cingular lists 1.7-2.0 the width of transverse furrow; their structure unknown. Left sulcal list 0.64-0.65 the length of body; Ri is 0.18-0.20, R2 is 0.14-0.16 the greatest depth of body; R3 absent; margin rounded or rounded angular poste- riorly. Length, 53.0-59.6 //• Eastern tropical Pacific. Description: — A medium-sized species, the body of which is subovate to subcircular in lateral outline, with a dorsal shoulder-like constriction at the pos- terior cingular list, and deepest in or somewhat in front of the middle. The ratio between the length and the depth of the body is 1.13 (1.08-1.18) : 1. The longi- tudinal axis is perpendicular to the girdle, or it has a posteroventral deflection of l°-4°. The epitheca is 0.74 (0.71-0.76) as deep as the hypotheca, highest in the center, of moderate convexity or more or less flat, and moderately to rather slightly elevated above the anterior cingular list. The transverse furrow is some- what concave, and about as wide as or somewhat narrower than the greatest height of the epitheca. The posterior cingular list is 0.25 (0.23-0.26) the length of the body from the apex. The hypotheca sometimes is subsymmetrical (Figm-e 20:6), sometimes it is decidedly flatter ventrally than dorsally (Figure 20:8). The dorsal margin is subuniformly, and gently to moderately convex except just behind the girdle where it is more bulging. The ventral margin, from the girdle to the posterior end of the left sulcal list, is gently convex. The postmargin is moderately to rather strongly convex and confluent with the dorsal and ventral margins. In dorsoventral view the body is wedge-shaped, widest just behind the girdle, subacute to narrowly rounded posteriorly, and with broadly rounded, more or less prominent protuberances just behind the girdle, one on each side of 156 THE DINOPHYSOIDAE. the body; the ratio between the length and the width of the body is 1.56 (1.38- 1.74) :1. The cingular lists are subhorizontal and subequal; their width is about 1.7- 2.3 the width of the transverse furrow and 0.23-0.26 the greatest depth of the body; their structure is unknown. The right sulcal list is unusually short, extend- ing to or slightly beyond the fission rib of the left sulcal list, and subtriangular, decreasing gradually in width posteriorly; anteriorly it is about as wide as the transverse furrow. The left sulcal list is about 0.64-0.65 the length of the body and of subuniform width throughout the greater portion of its length. The an- terior main rib is 0.18-0.20, the fission rib 0.14-0.16 the greatest depth of the body. The posterior main rib is lacking. The anterior two thirds of the free Figure 20. — 1-5, Phahicroma exj)ulsii»i. (Kofoid and Mirhener). 1, in left lateral view; 2, 4, in dorsal view; 3, in right lateral view; 5, in ventral view. 2, from the same specimen as 1 ; 4, from the same specimen as 3. 1-3, X 4-30; 4, of unknown magnification; 5, X 890. 1, 2, from Station 4713 (300-0 fath- oms); 3, 4, from Station 4724 (300-0 fathoms); 5, from Station 4717 (300-0 fathoms). 6-9, Phalacroma proluberans, sp. nov. 6, 8, in right lateral view; 7, 9, in dorsal view. 7, from the same specimen as 6; 8, 9, from type specimen. Porulation of theca indicated only in 8. X 430. Station 4730 (300-0 fathoms). margin of this list is almost straight, or gently sigmoid, concave anteriorly and convex posteriorly. The posterior portion of this margin, which is confluent with the anterior, is gently convex (Figure 20:6) to rounded angular (Figure 20:8). The main ribs are straight or almost so, and not club-shaped or otherwise modified. There are no accessory Usts or sails. The thecal wall is porulate. Megacytic specimens were not recorded. The dimensions of two specimens were measured. Dimensions: — Length of body, 53.0-59.6 n (type, 59.6 m). Greatest depth of body, 44.9-55.2 ^ (type, 55.2 m). Variations: — The two specimens differ strikingly from each other in the depth of the body, in the width and shape of the left sulcal list, in the development SYSTEMATIC ACCOUNT. 157 of the lateral expansions and constrictions of the body, and in the width of the posterior end of the body when seen in dorsoventral view. Figure 20 : 6 may, how- ever, represent a specimen drawn in a somewhat tilted position. Comparisons: ■ — Phalacroma protuberans is established on four outline draw- ings made from the two Expedition specimens. The structure of the cingular and sulcal lists and of the thecal wall is unknown. In spite of several rather striking differences, this species probably is closely related to Phalacrovia expulsum. It differs most strikingly from this species in the less-developed posteroventral inclination of the longitudmal axis of the body, in the greater width and subhorizontal position of the cingular lists, in the greater length of the left sulcal list, and m having a pair of broadly rounded lateral protuberances just behind the girdle. For further discussion on the relationships of these two species, see P. expulsum (p. 159). Occurrence: — Phalacroma protuberans is recorded at onlj^ one (4730, type locahty) of the 127 stations, on the fifth line of the Expedition, in the South Equatorial Drift, from 300-0 fathoms, and at a surface temperature of 79°. The frequency is less than 1 % (two specimens). Phalacroma expulsxjm (Kofoid and Michener) Plate 5, fig. 1. Figure 20: 1-5 Dinophysis exp^ilsa Kofoid & Michener, 1911, p. 268. Phalacroma stenopterygium Jorgensen, 1923, p. 11, 24, 44, fig. 10. Diagnosis: — Body sack-like in lateral outline, truncate anteriorly, with dorsal shoulder-like constriction at girdle, broadly rounded posteriorly, deepest in the middle, and 1.02-1.24 times longer than deep; longitudinal axis deflected posteroventrally at 5°-15°. In dorsal view wedge-shaped, widest at posterior cingular list or in the middle, subacute to narrowly rounded posteriorly, and with broad constriction in front of the middle; length: width, 1.42-1.82: 1. Posterior cingular list 0.11-0.17 the length of body from apex; epitheca barely if at all visible above anterior cmgular list. Cingular lists inclined anteriorly at 30°-45°, with or without ribs; the anterior about as wide as or somewhat narrower than transverse furrow; the posterior narrower than the anterior. Left sulcal list 0.42-0.57 the length of body; Ri is 0.12-0.16, R2 is 0.09-0.17 the greatest depth of body; R3 absent; margin rounded, not angular, posteriorly. Theca centrally faintly reticulate; about 17-20 polygons border girdle posteriorly. Length, 53.7- 67.5 M. Probably widely distributed in tropical, subtropical, and warm-temperate seas. 158 THE DLNOPHYSOIDAE. Description: — A medium-sized species, the body of which is sack-like in lateral outline, truncate anteriorly, with dorsal shoulder-like constriction at girdle, broadly rounded posteriorly, and deepest in or near the middle. The ratio be- tween the length and the depth of the body is 1.02-1.24: 1; in the Expedition specunens this ratio is 1.15 (1.08-1.24) : 1 ; in Jorgensen's (1923) it varied between 1.02: 1 and 1.23: 1. The longitudinal axis is inclined posteroventrally at 5°-15°; in the Expedition specimens this angle is 8° (5°-10°), in that figured by Jorgensen (1923, fig. 10) it is 15°. The epitheca is 0.60 (0.54-0.66) as deep as the hypotheca, gently convex or flat, highest in or ventrally to the center, and barely if at all visible above the anterior cingular list. The transverse furrow is somewhat concave, and about as wide as or somewhat wider than the greatest height of the epitheca. The posterior cingular list is 0.15 (0.11-0.17) the length of the body from the apex. The hy- potheca is more or less inclined posteroventrally (see above). The dorsal margin is subuniformly and gently to moderately convex. The ventral margin, from the girdle to the posterior end of the left sulcal list, is almost straight or gently con- cave or convex. The postmargin is subuniformly and broadly rounded, sub- semicircular, and either confluent with the dorsal and ventral margins, or (Jorgensen, 1923, fig. 10) forming a distinct although well-rounded angle at the posterior end of the left sulcal list. In dorsoventral view the body is wedge- shaped, widest at the posterior cingular list or in the middle, and subacute to narrowly rounded posteriorly; between the middle, where the body is more or less expanded, and the posterior cingular list, there is a broad and more or less pro- nounced constriction; the ratio between the length and the width is 1.60 (1.42- 1.82): 1. The cingular lists, which are characterized by an anterior inclination of 30°-45°, usually are subequal, but sometimes the posterior is more or less strik- ingly narro\\'er than the anterior. The anterior is about as wide as or somewhat narrower than the transverse furrow and about 0.11 (0.09-0.13) the greatest depth of the body. In the specimens examined Ijoth these lists are hyaline and lack distinct ribs. Jorgensen (1923, p. 11), on the other hand, writes about these lists : "both without or with distinct short radial transverse ribs, innermost by the cell." The right sulcal list is unusually short and extends to or slightly beyond the fission rib of the left sulcal list; it is subtriangular, or rounded anteriorly, and decreases gradually in width posteriorly; its maximum width does not exceed the width of the transverse furrow. The left sulcal list is 0.51 (0.42-0.57) the length of the body, and of subuniform width throughout the greater part of its length. SYSTEMATIC ACCOUNT. 159 The anterior main rib is 0.13 (0.12-0.16) and the fission rib 0.13 (0.09-0.17) the greatest depth of the body. The posterior main rib is usually lacking, but some- times a short rudiment may be found (Jorgensen, 1923, p. 11). The anterior half of the free margin of this list is straight or gently concave or convex; the posterior half, which is confluent with the anterior, is gently, moderately, or strongly con- vex, and not angular as in most species of the genus. The main ribs are straight or almost so, and not club-shaped or otherwise modified. With the exception of the main ribs, this list appears always to lack structural differentiation. There are no accessory lists or sails. Except along the dorsal, posterior, and ventral margins, the thecal wall of the hypotheca has a fine reticulum of polygonate meshes; the posterior margin of the girdle is bordered by seventeen to twenty polygons. Sometimes the polygons are subequal in size, sometimes they decrease somewhat m size posteriorly. In the transverse furrow there are two rows of polygons of about the same average size as the hypothecal polygons. On the epitheca no reticulation has been ob- served. The whole theca is furnished with evenly scattered pores. In the trans- verse furrow every polygon has a pore, which usually is located in or near the center. Megacytic stages were not recorded. The dunensions of six of the Expedition specimens and of the specunen figured by Jorgensen (1923, fig. 10) were measured. Dimensions: — Length of body, 53.7-64.5 n (average, 58.3 n; type, 55.8 n). Greatest depth of body, 45.7-51.2 ix (average, 49.1 m; type, 47.8 m)- The speci- mens measured by Jorgensen (1923) were, according to a statement in the text, 54-63 fx long and 53-58 m deep; the one figured (1923, fig. 10) was 67.5 yu long and 55.0 fi deep. Variations: — Phalacroma expulsum is rather variable. The following char- acters are the most variable : — the size of the body, the relative depth and width of the body, the degree of the posteroventral inclination of the longitudmal axis, the relative depth of the epitheca, the shape of the ventral margin of the hypo- theca, the development of the lateral expansions and constrictions of the body, and the width of the posterior end of the body when seen in dorsoventral view. Co?nparisons: — The description and figures of this species are based on the type material. The most aberrant specunen is represented by Figure 20:3,4. It differs from the remaining specunens in its narrower body, in the subuniform convexity of the ventral margm of its body from the girdle to the antapex, and in the shortness of its left sulcal list. Its present assignment is tentative. It has 160 THE DINOPHYSOIDAE. been referred here mainly because of the fact that its body, in dorsoventral view, has an outline of about the same shape as that of the typical specimens. The specimen figured by Jorgensen (1923, fig. 10) as Phalacroma stenoptery- gium differs from the typical representatives mainly in the following respects: its longitudinal axis has a more pronounced posteroventral inclination (15°, as compared with 5°-10°), the ventral margin of its body forms a distinct, although rounded corner at the posterior end of the left sulcal list, and its posterior cingular list is decidedly narrower. Jorgensen (1923) states that his specimens were 54- 63 M long and 53-58 m deep. From this it is evident that at least some of his speci- mens were decidedly deeper relatively than any of the Expedition specimens. The generic assignment of this species is somewhat uncertain. Kofoid and Michener (1911) allocated it to Dinophysis on account of the narrowness of the epitheca, the shape of the hypotheca in lateral view, and the position and anterior inclination of the cingular hst. Jorgensen (1923, p. 11), on the other hand, writes that: "Despite the great resemblance to the Dinophysis species, I have ascribed it to the genus Phalacroma, chiefly on account of the wedge-shaped cell and also because the lower girdle list can apparently have the same short radial ribs as the upper." With regard to the last-mentioned character it should be pointed out, first, that in the specimens examined both the cingular lists lacked ribs; and, second, that even though the ribbing of the posterior cingular list is a character frequently found in the species of Phalacroma, nevertheless, it is not limited to this genus but is found also in typical representatives of Dinophysis; see Dino- physis triacantha,Y ignve 42:2. Furthermore, Dinophysis cuneiformis, which is a typical representative of its genus, is narrowly wedge-shaped in dorsoventral view (Meunier, 1910, pi. 14, fig. 32). However, Jorgensen's (1923) decision is accepted by the close agreement between the dorsoventral outline of the body in this species and in some species of Phalacroma, e.g., in Phalacroma favus (Figure 14 : 2) ; by the regularly ovate, ellipsoidal, or bilaterally flattened dorsoventral outline of almost all the other known species of Dinophysis; and by the fact that in the closely related Phalacroma protuberans the epitheca is rather large and the cingular lists are subhorizontal. Of course, the question is of minor importance, since the genera Dinophysis and Phalacroma are so closely related that their separation is almost arbitrary. The structurally closest-known relative of Phalacroma e.rpulsu7n is P. pro- tuberans. The former species differs most strikingly from the latter in the more pronounced posteroventral inclination of the longitudinal axis of the body, in the anterior inclination of the cingular lists, and in the relative shortness of the left SYSTEMATIC ACCOUNT. IGl sulcal list. These two species occupy structurally a rather isolated position and ought to be made the only representatives of a special group. Jorgensen (1923) assigned P. expulsum to the Cuneus group. This allocation undoubtedly was made on account of the wedge-like shape of the body in dorsoventral view and because of the fact that the free margin of the left sulcal list is rounded postero- ventrally, instead of angular as in most of the species of Phalacroma. However, this allocation greatly decreases the structural uniformity of the Cuneus group, and for that reason should not be maintamed. Note, for instance, the relati\-e narrowness of the epitheca and the sack-like, instead of wedge-like, shape of the hypotheca of P. expulsum in lateral view. On the other hand, these two species show distinct affiliation to the Cuneus group and should thus be placed near to it. Jorgensen (1923, p. 11) also points out that P. expulsum "in several respects, especially in the left longitudinal furrow list," is "remarkably like" Dinophysis sphaerica (= similis). However, the relationship between these two species must be regarded as uncertain; the similarities in the shape of the hypotheca in lateral view and in the shape of the left sulcal list are rather superficial, while the differ- ence in the shape of the body in dorsoventral view is profound. It also may be worth mentioning that Dinophysis okamurai (Figure 31:5) shows a certain re- semblance to Phalacroma expulsum in the shape of the body in lateral view. Occurrence: — Phalacroma expulsum is recorded at four of the 127 stations. There are 0, 0, 0, 1, 3, and 0 stations on the six lines of the Expedition. Of these four stations, one (4713) is in the Galapagos Eddy, and three (4717, 4720, 4724) are in the South Equatorial Drift. There is one record from a surface haul (4720, from Salpa stomach). The remaining records refer to hauls from 300-0 fathoms. The temperature range of these four stations at the surface was 73°-81°, and the average was 76.8°. At Station 4720, where the species was taken at the sur- face, the surface temperature was 76°. The frequency is in every case less than 1 %. The species was first recorded byKofoid and Michener (1911) from Station 4717 of the Expedition, the type locality. Later Jorgensen (1923) reported it to be "scattered throughout the whole of the Mediterranean to the Bay of Corinth, often in deep water samples." It is probably widely distributed in tropical, sub- tropical, and warm-temperate seas. 7. LiMBATUM Group. Three species, viz., Phalacroma Ivmhalum, P. biparti- tum, and P. pulchrum, belong to this group, and all of them occurred in the col- lections of the Expedition. 162 THE DINOPHYSOIDAE. Phalacroma limbatum Kofoid and Michener Plate 3, fig. 3-5. Figure 21: 1 Phalacroma limbiila Kofoid & Michener, 1911, p. 290. Diagnosis: — -Body subcircular in lateral outline, deepest near the middle; length: depth, 1.08-1.10: 1. In dorsal view lens-shaped, 2.27 times longer than wide, with narrowly rounded apices. Posterior cingular list 0.38-0.50 the length of body from apex. Cingular lists appear to be without ribs. Left sulcal list: distance between Ri and Ra is 0.25-0.28 the length of body; R2 is 0.12-0.14 and Rs is 0.19-0.24 the greatest depth of body; margin forms angle of 70°-90° at R3; Ra inclined posteriorly at 20°-35°. Two parasagittal lists encircle body; the left of subuniform width throughout, and its maximum width is subequal to or some- what less than width of transverse furrow ; the right resembles the left but forms on antapex an acute, wedge-shaped, posterior sail directed posteriorly. Sail 0.15-0.27 the greatest depth of body in length, with central rib, whose proximal half might form more or less complex reticulum; angle at its tip, 35°-50°. Right parasagittal list forms du'ect continuation of left sulcal list. Central portion of each thecal valve with reticulum of moderate-sized mesh. Length, G8.9-7G.5 ju. Eastern tropical Pacific. Description: — A medium-sized species, the body of which is subcii'cular in lateral outline, slightly longer than deep, and deepest near the middle. The ratio between the length and the greatest depth of the body is 1.09 (1.08-1.10) : 1. The longitudinal axis is perpendicular to the girdle. The epitheca is about 0.87 as deep as the hypotheca, strongly and evenly convex, dome-shaped, highest in the center, and very prominent above the an- terior cingular list. The transverse furrow is flat, or slightly convex or concave, and its width is 0.24-0.32 the greatest height of the epitheca. The posterior cingular list is 0.38-0.50 the length of the body from the apex. The hypotheca is symmetrical; its dorsal, posterior, and ventral margins are well and evenly con- vex, and confluent. Its posterior portion is of about the same depth as the an- terior portion of the body. In dorsoventral view the body is lens-shaped, about 2.27 times wider than long, widest at the girdle, and with narrowly rounded apices; the side contours are nearly evenly convex. The cingular lists are subhorizontal and subequal; their width about equals or somewhat exceeds the width of the transverse furrow; and they are without ribs, according to available drawings (see p. 164). The sulcus is about half as long as the hypotheca. The flagellar pore is just behind the junction of the SYSTEMATIC ACCOUNT. 163 cingulum and the sulcus. On the ventral side of the left valve a small pore is found near the sagittal suture, just in front of the anterior cingular list. The right sulcal list extends to the posterior main rib of the left sulcal list; it is subtriangu- lar, decreasing gradually in width posteriorly, and anteriorly it is about as wide as the transverse furrow. The left sulcal list is of moderate width and rather short. The distance between the anterior and posterior mam ribs is 0.25-0.28 the length of the body. The anterior main rib is 0.13-0.14, the fission rib 0.12- 0.14, and the posterior main rib 0.19-0.24 the greatest depth of the body; behind the posterior main rib the hst decreases suddenly in width. The free margin of this list is gently sigmoid or nearly straight between the anterior and posterior mam ribs; at the posterior main rib it forms an angle of 70°-90°; behind the pos- terior main rib it is gently concave. The main ribs of this list are of moderate strength and almost straight; or the posterior one is rather strong and gently concave posteriorly; none of them is club-shaped or otherwise modified. The distance between the anterior main rib and the fission rib is 0.33-0.50 the distance between the anterior and posterior main ribs. The posterior main rib has a posterior inclination of 20°-35°. Besides the three main ribs, this list sometimes has a faint reticulation; and just behind the posterior main rib it has a fine rib with a posterior inclination of about 50°-60°. Two parasagittal Usts encircle the entire body; the one on the right valve of the hypotheca forms a direct contin- uation of the left sulcal list. The left parasagittal list is of subuniform width throughout its entire length, has a maximum width subequal to or somewhat less than the width of the transverse furrow, and is furnished with a moderate num- ber of sunple, free or anastomosing, incomplete riblets. The right parasagittal list resembles the left, but it forms on the antapex an acute, wedge-shaped posterior sail, directed posteriorly. WTien fully developed, the sail is 0.15-0.27 the greatest depth of the body m length and has a central rib, whose proximal half might form a more or less complex reticulum; the angle at its tip is 35°-50°. The central portion of each thecal valve, with the exception of the transverse furrow, is characterized by a reticuliun of closely set, rather heavily niai'gined polj'gons of moderate size. The polygons are subequal, and on each valve about thirty of them border the posterior margm of the transverse furrow. The entne theca is furnished with scattered pores. There is a central pore in most of the polygons; and in the transverse furrow there are three rows of pores, each row with thirty to forty pores on each valve. Megacytic stages were not found. The proportions of two specimens were measured. 164 THE DINOPHYSOIDAE. Dimensions: — Length of body, 68.9-76.5 m (average, 72.7 m; type, 76.5 a<). Greatest depth of body, 62.4-70.6 m (average, 66.5 n; type, 70.6 fi). Variations: — This species is fairly constant; the relative height of the epitheca, the width of the parasagittal lists, and the size and structure of the posterior sail of the right parasagittal list are the most variable characters. Comparisons: — The description and figures are based on the type material. Figure 21. — 1, Phalacroma limbatum Kofoid and Michener, right lateral view. X 430. Station 4722 (300-0 fathoms). 2, Phalacroma hipartitum, sp. nov., right lat- eral view of type specimen; porulation indicated only on portion of transverse furrow. X 430. Station 4736 (300-0 fathoms). 3, Phalacroma pulchrum Kofoid and Michener, right lateral view of type specimen. X 430. Station 4G99 (300-0 fathoms). We refer the specimen, Plate 3, fig. 4, tentatively to this species; the similari- ties between it and those described above are as follows: — the body is subcircu- lar in lateral outline; the girdle is subequatorial in position; only the central portions of the thecal valves are reticulated; there is an accessory rib in the left sulcal list just behind the posterior main rib. On the other hand, the differences exhibited by the first specimen also are conspicuous; for instance, it lacks para- sagittal lists except on the antapical portion of the hypotheca ; at the antapex its right parasagittal list is rather wide, but no acute, wedge-shaped projection sup- ported by a central rib is developed. This specimen has been referred to this species, since these differences probably are due to regulatory adjustments fol- lowing binary fission. However, since the regulatory structural changes that take place at binary fission are unknown, this assignment must be regarded as tenta- tive. The specimen is not included in the description given above, and its locality is omitted from the account of the occurrences. Its length is 84 ix, its greatest depth, 78 p.; its cingular lists are ribbed. It was taken at Station 4687 of the Expedition, in a haul from 300-0 fathoms; surface temperature, 73°. Phalacroma limbalum is structurally closely related to P. hipartitum. These species resemble each other in the following respects: — (1) the body is sub- cu'cular in lateral view; (2) the longitudinal axis of the body is perpendicular to the girdle; (3) the epitheca is high; (4) the cingular lists are subhorizontal, sub- SYSTEMATIC ACCOUNT. 165 equal, and about as wide as the transverse furrow; (5) the left sulcal list is angular at the posterior main rib and is contmued posteriorly by a parasagittal list that has one or two acute, wedge-shaped posterior sails with a more or less complex reticulum of irregular ribs; (6) the central portion of each valve, with the excep- tion of the transverse furrow, has a reticulum of closely placed, rather heavily margined polygons; the polygons are subequal, of moderate size, and most of them have a central pore ; in the transverse furrow and near the sagittal margins, the valves are porulate but not reticulate; (7) the size of the body is the same. It should be remembered, however, that several of these similarities are due to the fact that these two species in some respects are fairly low in the scale of the evolutionary development of the genus. The subcircular shape of the body in lateral view, the longitudinal axis of the body being perpendicular to the girdle, the relatively great height of the epitheca, the cingular lists being subhorizontal, subequal, and about as wide as the transverse furrow, and the angularity of the left sulcal list at the posterior main rib are more or less primitive characters and, therefore, must be considered to have relatively little importance in an attempt to determine the degree of relationship between these and other more or less primi- tive species of this genus. Phalacroma limhalum is more primitive than P. biparti- tum in having a relatively higher epitheca, and in the right parasagittal list having one, instead of two, acute triangular posterior sails. On the other hand, it is more advanced than the last species m having the two parasagittal lists encircle the entire body. Phalacroma limhatum is probably also rather closely related to P. pulchrum. This relationship is indicated by the following similarities: — (1) the body of these two species is subcircular in lateral view and lens-shaped m dorsoventral view; (2) on the hypotheca both have two parasagittal lists; the left parasagittal hst is narrow and of subuniform width throughout; the right forms a direct con- tinuation of the left sulcal list and has a large, acute, wedge-shaped posterior sail, which is furnished with a central rib ; (S) at least in some specimens of P. pulchrum (Jorgensen, 1923, fig. 18), the left sulcal list has behind the posterior main rib an accessory rib of the same kind as that found in P. limhalum. Phalacroma limha- tum is more primitive than P. pulchrum in having a relatively higher epitheca and subhorizontal cingular lists. It is more advanced in size of body, and in having parasagittal lists on the epitheca. With regard to the relationship between the species mentioned above and Dinophysis collaris (p. 295), see the section on comparisons. Occurrence: — Phalacroma limbalum is recorded at six of the 127 stations. 166 THE DINOPHYSOIDAE. There are 0, 1, 2, 2, 1, and 0 stations on the six hnes of the Expedition. Of these six stations, two (4667, 4676) are in the Peruvian Current; one (4699) is in the Easter Island Eddy; and three (4681, 4705, 4722) are in the South Equatorial Drift. At one station (4681) the species is recorded from 300-0 fathoms and 800-0 fathoms. All the other records refer to hauls from 300-0 fathoms only. The temperature range of these six stations at the surface was 68°-75° ; the average was 71.2°. At two stations (4676, 4705) the frequency is 1%; at the remaining stations it is less. The species has been found only in the material of the Expedition. It was first recorded by Kofoid and Michener (1911) at Station 4667 of the Expedition, which thus is the type locality. Phalacroma bipartitum, sp. nov. Figure 21:2 Diagnosis: — Body subcircular in lateral outline, deepest in the middle; length : depth, 1 .04 : 1 . Posterior cingular list 0.30 the length of body from apex. Left sulcal list: distance between Ri and R3 is 0.41 the length of body; R2 is 0.13 and R3 is 0.25 the greatest depth of body; margin acuminate at R3 at angle of 40°; R3 incUned posteriorly at 20°. Around posterior portion of hypotheca there is on right ^'alve a parasagittal list forming direct continuation of left sulcal list ; para- sagittal list forms two subequal, acute, wedge-shaped posterior sails, one on each side of midline, both with irregularly anastomosing ribs; length of sails equals length of R3; angles at their tips, 30°-40°; width of Hst between sails and between ventral sail and R3 is 0.20 the length of R3. Central portion of each valve reticu- lated with heavily margined polygons of moderate size. Length, 68 ti. Eastern tropical Pacific. Description: — A medium-sized species, the body of which is subcircular in lateral outline, about as long as deep and deepest Hear the middle. The ratio between the length and the greatest depth of the body is 1 .04 : 1 . The longitudinal axis of the body is perpendicular to the girdle. The epitheca is 0.87 as deep as the hypotheca, strongly and evenly convex, dome-shaped, highest in the center, and very prominent above the anterior cingular list. The transverse furrow is slightly convex; and its width is 0.37 the greatest height of the epitheca. The posterior cingular list is 0.30 the length of the body from the apex. The hypotheca is symmetrical; its dorsal, posterior. SYSTEMATIC ACCOUNT. 167 and ventral margins are well and evenly convex, and confluent. In the type specimen the posterior end of the body is slightly narrower than the anterior. The cingiilar lists are subhorizontal and subequal; their width about equals the width of the transverse furrow; their structure is unknown. The right sulcal list extends to a point about halfway between the fission rib and the posterior mam rib of the left sulcal list ; it is of subuniform width throughout the greater part of its length, and its greatest width is somewhat less than the length of the fission rib of the left sulcal list; its free margin is sigmoid, being slightly concave anteriorly and rather strongly convex posteriorly. The left sulcal list is of moderate length and width. The distance between the anterior and posterior main ribs is 0.41 the length of the body. The anterior main rib and the fission rib are subequal and 0.13 the greatest depth of the body; the posterior main rib is 0.25 the greatest depth of the body; behind the posterior main rib the list de- creases suddenly in width. The free margin of this list is gentlj' concave between the anterior main rib and the fission rib as well as between the fission rib and the posterior main rib ; at the posterior main rib it is acuminate and forms an angle of about 40°; behind the posterior main rib it is gently concave. The main ribs of this list are of moderate strength and almost straight, and none of them is club- shaped or otherwise modified. The distance between the anterior main rib and the fission rib is about 0.50 the distance from the anterior to the posterior main rib. The posterior main rib has a posterior inclination of about 20°. There are no ribs except the three main ribs, but, judging by the short projections from the an- terior side of the posterior main rib in the drawing of the type specimen (Figure 21:2), this list probably has a faint reticulation. On the right valve there is a parasagittal list, which forms a direct continuation of the left sulcal list, runs around the posterior portion- of the hypotheca and extends to the posterior end of the dorsal margin of the hypotheca. The parasagittal list is characterized by two subequal, acute, wedge-shaped posterior sails, one on either side of the mid- line, and both furnished with irregularly anastomosing ribs. The length of these sails about equals the length of the posterior main rilj of the left sulcal list; and the angles at their tips are about 30°-40°. The width of the parasagittal list, between the left sulcal list and the ventral sail and between the two sails, is about 0.20 the length of the posterior main rib of the left sulcal list. It is not known whether there is any parasagittal list on the left ^•alve of the hypotheca; nor whether the parasagittal list of the right valve ever extends farther on the dorsal side than in Figure 21:2 (as in the closely related Phalacroma limbaimn). The epitheca had no parasagittal lists in the type spechnen. 168 THE DINOPHYSOIDAE. The structure of the theca is about the same as in Phalacroma limhatum; i.e., the central portion of each valve, with the exception of the transverse furrow, has a reticulum of closely placed, rather heavily margined polygons. The poly- gons are subequal and of moderate size, and most of them have a central pore; on each valve about thii'ty polygons border the posterior side of the transverse furrow. In the transverse furrow and near the sagittal margins the valves are porulate, but not reticulate. In the transverse furrow there are two transverse rows of fine pores, on each valve about forty to fifty pores in each row. JNIegacytic stages were not found. The proportions of the type specimen were measured. Dimensions: — Length of body, 68.0 n. Greatest depth of body, 65.5 m- Comparisons: — The description of the type is incomplete in some respects; e.g., the shape of the body in dorsoventral view and the structure of the cingular lists are unknown ; and the number and the extension of the parasagittal lists are uncertain. The structurally closest-known relative of Phalacroma bipartitum is P. lim- hatum (see p. 164). The former is more advanced than the latter in ha^'ing a relatively lower epitheca and in having a parasagittal list with two, instead of one, acute, triangular posterior sails. It is more prmiitive in having the para- sagittal list of the hypotheca relatively shorter and in lacking parasagittal lists on the epitheca. It should be noted, however, that the extension of the parasagit- tal list of the hypotheca presumably is variable (see Dinophysis collaris, p. 295). Phalacroma bipartitum resembles Dinophysis collaris in having the parasagit- tal list of the right valve furnished with two subequal and triangular posterior sails. Whether or not this similarity is an indication of a close relationship cannot be decided. Dinophy.sis collaris is more advanced than Phalacroma bipartitum in the rather irregular shape of its body in lateral \-iew, in the reduced height of its epitheca, and in its larger and more differentiated cingular and sulcal lists. Occurrence: — This species is recorded at only one (4736, the type locality) of the 127 stations, on the fifth line of the Expedition, in the South Equatorial Drift, from 300-0 fathoms, and at a surface temperature of 81°. The frequency is less than 1 % (one specimen). Phalacroma pulchrum Kofoid and jNIichener Plate 3, fig. 1, 6. Figure 21:3 Phnlacroma pxdchra Kofoid & Michener, parlim, 1911, p. 290, 291. Phalacroma cireumxuiiim Jorgensen, ■parlim, 192.3, p. 17, 4.3, fig. 18. Phalacroyna ptdchrtun Jorgensen, 1923, p. 18. SYSTEMATIC ACCOUNT. 169 Diagnosis: — Body subcircular in lateral outline, subtruncate anteriorly, deepest near the middle; length: depth, 0.99-1.03:1. In dorsal view narrowly obovate, 1.80 times longer than wide. Posterior cingular list 0.19-0.24 the length of body from apex; epitheca sometimes hardly visible above anterior cingular list. Cingular lists somewhat inclined anteriorly, without ribs. Left sulcal list: distance between Ri and R, is 0.41-0.43 the length of body ; R2 is 0.14-0.16, and R3 is 0.33-0.38 the greatest depth of body; margin forms angle of 50°-60° at R3; R3 inclined posteriorly at 35°-40°. Two parasagittal lists on hypotheca; the left very narrow, sometimes possibly absent. The right continues left sulcal list; dorsally its maximum width is subequal to or less than half the width of trans- verse furrow; on antapex it forms an acute, wedge-shaped posterior sail directed posteriorly. Sail, when fully developed, 0.33-0.43 the greatest depth of body in length, with central rib whose base might be divided; angle at tip, 30°-40°. Ventrally to sail, width of list is 0.50-1.75 the width of transverse furrow; with about 8-11 short riblets. Theca with fine reticulation. Length, 51.2-52.5 p. Tropical, subtropical, and warm-temperate seas. Description: — A rather small species, the body of which is subcircular in lateral outline, subtruncate anteriorly, and deepest near the middle. The ratio between the length and the greatest depth of the body is 0.99-1.03: 1. The longi- tudinal axfs is perpendicular to the girdle. The epitheca is 0.77-0.80 as deep as the hypotheca, highest in the center, of moderate convexity to rather flat, and but slightly elevated above the anterior cingular list. The transverse furrow is flat, or slightly convex to concave, and its width is 0.8-2.0 the greatest height of the epitheca. The posterior cingular list is 0.19-0.24 the length of the body from the apex. The hypotheca is almost sym- metrical; sometimes, as in the type specimen, it is subcircular, with confluent dorsal, posterior, and ventral margins; sometimes, as in the specimen figured by Jorgensen (1923, fig. 18), its ventral margin is flattened, and its postmargin forms a broadly rounded, subrectangular corner. In dorsoventral view the body is narrowly obovate, about 1.80 times longer than wide, widest somewhat in front of the middle, somewhat more broadly rounded anteriorly than posteriorly, narrowly rounded posteriorly, and with evenly convex side contours. The cingular lists have an anterior inclination of 20°-40°; the anterior is about as wide as the transverse furrow, the posterior often is slightly narrower. The sulcus is about half as long as the hypotheca. The flagellar pore is located at a distance about eciualing its own diameter behind the junction of the cingulum and the sulcus. On the ventral side of the left valve a small pore is found near the 170 THE DINOPHYSOIDAE. sagittal suture, just in front of the anterior cingular list. The right sulcal Ust sometimes ends at or somewhat behmd a point midway between the fission rib and the posterior main rib of the left sulcal list ; sometimes it extends even to the last-mentioned rib; its anterior half is about as wide as the transverse furrow; posteriorly it decreases gradually in width. The left sulcal list is rather large ; the distance between its anterior and posterior main ribs is 0.41-0.43 the length of the body. The anterior main rib is 0.14-0.15, the fission rib 0.14-0.16, and the posterior main rib 0.33-0.38 the greatest depth of the body. Behind the posterior main rib this list decreases suddenly in width. Between the anterior and posterior main ribs the free margin of this Ust either is nearly straight or slightly sigmoid ; in the latter case it is slightly concave anteriorly and convex posteriorly, or vice versa; at the posterior main rib it is somewhat acuminate and forms an angle of 50°-60°; behind the last-mentioned rib it is gently concave. The main ribs of this list are of moderate strength and straight or almost so; none of them is club- shaped or otherwise modified. The distance between the anterior main rib and the fission rib is 0.33-0.37 the distance between the anterior and posterior main ribs. The posterior main rib has a posterior inclination of 35°-40°. Somewhat behind the posterior main rib there is in some specimens (Jorgensen, 1923, fig. 18) a fine rib with a posterior inclination of about 75°. Except the ribs mentioned above, the left sulcal list lacks structural differentiations. There are two para- sagittal lists on the hypotheca, but none on the epitheca. The left one of these two lists extends, on the dorsal side of the body, from the posterior cingular list to the antapex; it is very narrow, of subuniform width throughout its entire length, and furnished with a few cross-ribs; its maximum width is less than half the width of the transverse furrow; in some specimens (Jorgensen, 1923, fig. 18) it may be absent. The right parasagittal list forms a direct continuation of the left sulcal list; sometimes, as in the type specimen, it extends to the posterior cingular list; sometimes, as in the specimen figured by Jorgensen (1923, fig. 18), it is not developed on the dorsal side of the hypotheca. On the dorsal side this list is narrow, about as wide as or slightly wider than, the left list, of subuniform width throughout, and furnished with a few cross-ribs; posteriorly it forms an acute, wedge-shaped posterior sail directed posteriorly. The sail, when fully developed, is 0.33-0.43 the greatest depth of the body in length, and has a central rib which arises from the antapex, and whose base might be divided; the angle at its tip is 30°-40°. Between this rib and the posterior main rib of the left sulcal list, the right parasagittal list has a minimum width equaling 0.50-1. 75 the width of the transverse furrow, and its margin is evenly concave, or slightly undulating SYSTEMATIC ACCOUNT. 171 as in the type specimen ; along the base of this portion of the hst there are 8-1 1 short, nearly equidistant, simple or bifurcate riblets; see also the accessory rib of the left sulcal list. The thecal wall is finely reticulate (Jorgensen, 1923, fig. 18) and furnished with scattered pores. Sometimes the reticulation is so faint as to make it difficult to detect. Megacytic stages were not recorded. The dimensions of the type and of the specimen figured by Jorgensen (1923, fig. 18) were measured. Dimensions: — Length of body, 51.2 yu. Greatest depth of body, 51.6 ai. Specimen figured by Jorgensen (1923, fig. 18) : Length of body, 52.5 tx. Greatest depth of body, 51.0 m- Variations: — Very little is known about the variability of this species. The specimen figured by Jorgensen (1923) is somewhat more pointed posteriorly than the type specimen, and its right sulcal list is somewhat smaller. The struc- ture of the thecal wall is sometimes fairly well developed, sometimes almost in- visible. Comparisons: — The description and figures are based on the type material. The structurally close.st-known relative of Phalacroma inilchrum is P. lim- batum. The former can easily be distinguished by its low epitheca and by the absence of parasagittal lists from the epitheca (see P. limhatum, p. 164). Sijnonymy: — Some of the specimens on which Kofoid and Michener (1911) based their description, have in the present paper been referred to P. cimcolus, sp. nov. Phalacrovm circumsutuvi Jorgensen (1923) includes P. circumsuium Karsten and P. pulchrmn Kofoid and Michener. Occurrence: — Phalacroma pulchrum is recorded at six of the 127 stations. There are 0, 0, 0, 4, 1, and 1 stations on the six lines of the Expedition. Of these six stations, one (4699) is in the Easter Island Eddy; one (4713) is in tlie Gala- pagos Eddy; four (4701, 4705, 4730, 4740) are in the South Equatorial Drift. All the records refer to hauls from 300-0 fathoms. The temperature range of these sLx stations at the surface was 72°-81°; the average was 75.9°. The frequency is less than 1 % in all the cases recorded. The species was first recorded by Kofoid and Michener (1911) at Station 4699, the type locality. Later Jorgensen (1923) found it in the Mediterranean, the Thor Expedition, and possibly also in the Guinea Current, the German South Pole Expedition. 172 THE DINOPHYSOIDAE. This is a eupelagic species widely distributed in tropical, subtropical, and warm-temperate seas. The most outstanding peculiarities of its distribution in the Eastern Pacific are its absence from surface catches, as well as from the California Current, Mexican Current, Panamic Area, and Peruvian Current. Its optimum habitat probably is in deeper waters within the levels of photosynthesis, and in the most distinctively tropical regions. 8. DoRYPHORUM Group. Of the five species of this group only Phalacroma pugiunculus was not found by the Expedition. Phalacroma mucronatum, sp. no v. Figure 22:4, 6, 8 Diagnosis: — Body subcircular in lateral outline, sometimes slightly deeper anteriorly than posteriorly; length: depth, 1.07-1.11:1. Posterior cingular list 0.33-0.41 the length of body from apex. Left sulcal list 0.50-0.54 the length of body; distance from Ri to Rs is 0.33-0.37 the length of body; Ro is about 0.17, and R3 about 0.30 the greatest depth of body; margin forms angle of 80°-90° at Rs; R3 inclined posteriorly at 30°-40°. With triangular posterior sail, placed on antapex and directed posteriorly; when fully developed, its length is 0.28-0.40 the greatest depth of body, and it is somewhat narrower than long; without or with central rib ; well separated from left sulcal list. Theca porulate (and finely areolate?). Length, 35.0-45.4 fx. Eastern tropical Pacific. Description: — A small species, the body of which is subcircular in lateral outline, deepest near the middle, and sometimes slightly deeper anteriorly than posteriorly. The ratio between the length and the depth is 1.07-1.11:1. The longitudinal axis of the body is perpendicular to the girdle. The epitheca is 0.89-0.94 as deep as the hypotheca, evenly and moderately convex, highest in the center, and moderately elevated or rather prominent above the anterior cingular list. The transverse furrow is flat or slightly convex, and its width is 0.5-1.0 the greatest height of the epitheca. The posterior cingular list is 0.33-0.41 the length of the body from the apex. The hypotheca is symmetrical, with well or moderately rounded dorsal and ventral margins; the posterior margin is rather strongly convex and confluent with the dorsal and ventral margins. The posterior portion of the hypotheca is of about the same width as, or slightly nar- ower than, the epitheca. The shape of the body in dorsoventral view is not known with certainty, as all the specimens drawn in this aspect were megacytic. It SYSTEMATIC ACCOUNT. 173 appears to be rather narrowly obovate, widest somewhat m front of the middle, with rounded apices, and with the antapical end somewhat narrower than the epitheca. The cingular lists are subhorizontal, subequal, and about as wide as the transverse furrow; their structure is unknown. The right sulcal list is unknown. The left sulcal list is of moderate length and width. Its length is 0.50-0.54 the length of the body, and the distance between the anterior and posterior main ribs is 0.33-0.37 the length of the body. The anterior main rib is 0.15-0.20, the fission rib about 0.17, and the posterior main rib about 0.30 the greatest depth of the body. Behind the posterior main rib this list decreases suddenly in width. Between the anterior and posterior main ribs the free margin of this list is almost straight or slightly concave ; at the posterior main rib it forms an angle of 80°-90° ; behind the last-mentioned rib it is almost straight or slightly concave or convex. The main ribs of this list are of moderate strength, straight or almost so, and not club-shaped or otherwise modified. The distance between the anterior main rib and the fission rib is about 0.50 the distance between the anterior and posterior main ribs. The posterior rib has a posterior inclination of 30°-40°. Except for the three main ribs, this list appears to lack structural differentiations. On the right valve there is an acute, triangular posterior sail, which is located on the ant- apex and is directed posteriorly. When fully developed, its length is 0.28-0.40 the greatest depth of the body, and it is somewhat narrower at its base than it is long; sometimes its length is about twice the basal width. Its margins are almost straight or gently concave, convex or sigmoid, and the angle at its tip is 15°-40°. In some specimens this sail lacks structural differentiations, in others it has a central rib which may be divided at the base. The sail is well separated from the left sulcal list. The distance between these two structures is variable ; in the speci- mens examined it equals at least half the basal width of the sail. There are no parasagittal lists. The thecal wall has scattered pores. Ai'eolation was not seen, but might have been overlooked. Megacytic stages have been seen (Figure 22: 6). The proportions of three specimens were measured. Dimensions: — Length of body, 35.0-45.4 yu (average, 40.7 //; type, 38.8 m). Greatest depth of body, 32.6-41.2 m (average, 36.2 m; type, 34.9 n). Variations: — The five specimens examined are very similar. They exhibit variations mainly in the size of the body, the relative height of the epitheca, and the shape and structure of the posterior sail. 174 THE DINOPHYSOIDAE. Comparisons: — The species is established on outline drawings of five Ex- pedition specimens. It probably has about the same habitus as the ancestral form from which Phalacroma doryphorum, P. circumsutum, P. cuneolus, and P. pugiunculus origi- nated. The following of its characters are more or less prmiitive : its small size, its subcii'cular shape in lateral view, the relatively great height of its epitheca, the fact that its left sulcal list is relatively long and not unusually wide posteriorly, and that the posterior main rib of this list is not club-shaped. Phalaa'oma mucro- FiGUBE 22. — 1, 2, 3, 5, 7, Phalacroma cuneolus, sp. nov. 1, 2, left lateral view; 3, right lateral view; 5, 7, dorsal view (megacytio). Surface markings, areolation, indicated only on small portion of theca in 1. 1, 5, from type specimen. X 430. Station 4711 (300-0 fathoms). 4, 6, 8, Phalacroma mucronalum, sp. nov. 4, 8, left lateral view; 6, dorsal view (megacytic). X 430. 4, 6, typo specimen, from Station 4730 (300-0 fathoms); 8, from Station 4638 (300-0 fathoms). nalwm differs from the four species mentioned in its smaller size and in the sub- circular shape of its bod}'. Phalacroyna doryphorum appears to be its closest relative. Occurrence: — Phalacroma mucronalum is recorded at four of the 127 stations. There are 0, 1, 1, 1, 1, and 0 stations on the six lines of the Expedition. Of these four stations, one (46.38) is in the Panamic Area, and three (4681, 4711, 4730) are in the South Equatorial Drift. All records refer to hauls from 300-0 fathoms only. Station 4730 is the type locality. The temperature range of these four stations at the surface was 68°-79°; the average was 74.5°. At one station (4638) the frequency is 2 % ; in the remaining cases it is less than 1 %. SYSTEMATIC ACCOUNT. 175 Phalacroma doryphorum Stein Figure 23: 1-5; 24 Phalacroma doryphorum Stein, 1883, p. 23, pi. 19, fig. 1-1. Butschli, 1885, p. 942. Schutt, 1895, p. 50, 89, pi. 4, fig. 19; 1899, p. 42, pi. 6, fig. 17. Murr.w & Whittinq, 1899, p. 330, tab. 1-9. Lemmer- MAKN, 1899a, p. 319, 320, 372; 1901a, p. 372; 1905a, p. 35. Schroder, 1900a, p. 19; 1906a, p. 325, 327, 330; 1911, p. 25, 37. Cleve, 1900b, p. 1031; 1901a, p. 17; 1901c, p. 270; 1902b, p. 36; 1903b, p. 347. Schmidt, 1901, p, 137. Ostenfeld & Schmidt, 1901, p. 176. Entz, 1902b, p. 94; 1905, p. HI. LoHMANN, 1902, p. 53; 1908a, p. 161; 1920, p. 484, 492, 508. P.will.^rd, 1905, p. 58, 81, 102; 1909, p. 283; 1915a, p. 2, fig. B; 1916, p. 47, 49, 50, 52, 60, fig. 12, pi. 3, fig. 8. K.^jisten, 1906, p. 185, 187, 189, 191; 1907, p. 228, 238, 240, 247, 257, 285, 295, 304, 318, 321, 334, 337, 340, 341, 343, 347, 348, 349, 352, 354, 355, 421, 438, 464, 471. Nathansohn, 1908, p. 604; 1909, p. 47; 1910a, p. 14, 17, 20; 1910b, p. 61. Graf, 1909, p. 136. Stuwe, 1909, p. 252, 287. Hensen, 1911, p. 166, tab. 15. Okamitra, 1912, p. 18, pi. 5, fig. 77. Schiller, 1912, p. 27. Forti, 1922, p. 104, 190, 208, fig. 107. Johqensen, 1923, p. 16, 17, 37, 44, fig. 17. Forti & Issel, 1923, p. 3. Diagnosis: — Body obovate or subovate in lateral outline, deepest at or gener- ally somewhat behind girdle ; length : depth, 1 .05-1 .17:1. In dorsal view narrowly obovate or subcuneate; narrowly rounded or subacute posteriorly. Posterior cingular list 0.17-0.37 the length of body from apex. Cingular lists without ribs. Left sulcal list rather short, often high posteriorly; length, 0.41-0.55 the length of body; distance between Ri and Rs is 0.26-0.38 the length of body; R2 is 0.11- 0.17, R3 is 0.24-0.45 the greatest depth of body; margin forms angle of 50°-90° at R3; R3 inclined posteriorly at 15°-40°, often club-shaped. With triangular pos- terior sail, either on or usually somewhat ventrally to antapex, and inclined ventrally at 0°-35°; when fully developed, its length is 0.20-0.37 the greatest depth of body, and it is about as wide as long or slightly narrower; with or without central rib or reticulation; usually well separated from left sulcal list. Theca finely and closely areolate. Length, 54-86 m- Widely distributed in tropical, subtropical, and warm-temperate seas. Description: — A medium-sized species, the body of which is obovate or sub- obovate in lateral outline, deepest at or m most specmiens somewhat behind the girdle and 1.11 (1.05-1.17) times longer than deep. The longitudinal axis of the body is perpendicular to the girdle. The epitheca is evenly and moderately convex to rather flat, highest in the center, and moderately or but slightly elevated above the anterior cingular list. In the Expedition specimens it is about 0.90 (0.85-0.92) as deep as the hypotheca; in the specimen represented by Stein (1883, pi. 19, fig. 2), on the other hand, the epitheca and the hypotheca are of about equal depth. The transverse furrow is flat or slightly convex, and its width is about 0.33-0.50 the greatest height of the epitheca. The posterior cingular list is 0.31 (0.17-0.37) the length of the body 176 THE DINOPHYSOIDAE. from the apex. The hypotheca is symmetrical; sometimes it is, as Jorgensen (1923, p. 16) expresses it, of an "oval wedge-shape," with rather flat dorsal and ventral margins; sometimes it is regularly' oval, with well or moderately rounded dorsal and ventral margins; the posterior margin is strongly convex and confluent with the dorsal and ventral margins. The posterior portion of the body is strik- ingly narrower than the anterior. In dorsoventral view (Stein, 1883, pi. 19, fig. 4; Okamura, 1912, pi. 5, fig. 77b) the body is narrowly obovate or subcuneate, widest somewhat in front of the middle, and narrowly rounded or subacute pos- teriorly. The cingular lists are subequal and subhorizontal, or slightly inclined an- teriorly; their width about equals the width of the transverse furrow, and they are hyaline, and without ribs or other structural differentiations. The right sulcal list ends at a point about halfway between the fission rib and the posterior main rib of the left sulcal list, or it extends to or even slightly beyond the posterior main rib of the last-mentioned list ; in some specimens it is subtriangular, decreas- ing gradually in width posteriorly ; in other specimens it is of almost equal width throughout its anterior half or even throughout the greater portion of its length; anteriorly it is about as wide as the transverse furrow. The left sulcal list is rather short, and often comparatively high posteriorly. Its length is 0.46 (0.41-0.55) the length of the body, and the distance between the anterior and posterior main ribs is 0.29 (0.26-0.38) the length of the body. The anterior main rib is 0.1 1 (0.08- 0.14), the fission rib 0.13 (0.11-0.17), and the posterior main rib 0.31 (0.24-0.45) the greatest depth of the body. Behind the posterior main rib this list decreases suddenly in width. Between the anterior and posterior main ribs the free margin of this list often is almost straight ; sometimes it is gently concave or convex, or it is slightly sigmoid, concave anteriorly and convex posteriorly; at the posterior mam rib it forms an angle of 70° (50°-90°) ; behind the last-mentioned rib it is al- most straight, or gently concave or convex. The main ribs of this list are of moderate strength and straight or nearly so ; in most of the specimens examined the posterior main rib is club-shaped (see also Schiitt, 1895, pi. 4, fig. 19: 2, and Jorgensen, 1923, fig. 17); in the specimens drawn by Stein (1883), on the other hand, this rib is not club-shaped. The distance between the anterior main rib and the fission rib is 0.35-0.50 the distance between the anterior and posterior main ribs. The posterior main rib has a posterior inclination of 25° (15°-40°). Except for the three main ribs, this list seems to be without structure in most specimens; sometimes it may have a very faint reticulation. On the right valve there is an acute, triangular, posterior sail, which is situated either on or usually SYSTEMATIC ACCOUNT. 177 somewhat ventrally to the antapex, and is mcHned ventrally at 0°-35°. The size of this sail is somewhat variable, probably due to the fact that m each binary fission one of the daughter schizonts has to form this structure anew. When fully developed, its length is 0.20-0.37 the greatest depth of the body, and it is about as wide at the base as it is long or slightly narrower. Its margins are almost straight, or gently concave, convex, or sigmoid, and the angle at its tip is 40°-90°. In some specimens the sail has a central rib, in others it is more or less reticulated, especially in its central portion ; sometimes both the central rib and the reticula- tion are developed, sometimes no structure can be distinguished. In most speci- mens the sail is well separated from the left sulcal list; the distance between these two structures is variable, but usually it equals at least half the basal width of the posterior sail. Only in one specimen (Figure 23 : 5) were these structures con- nected, and this specimen appeared still to be in a condition of reorganization fol- lowing binary fission. There are no parasagittal lists. The thecal wall is finely and closely areolate and has scattered pores; the areoles are subequal or somewhat unequal in size. In the transverse furrow there are two rows of pores. Megacytic stages have been recorded (see Pavillard, 1916, fig. 12, and J5r- gensen, 1923, p. 17). The proportions of seven Expedition specimens and of the specimens drawn by Stein (1883), Schutt (1895), Okamura (1912), and Jorgensen (1923) were measured. Dimensions: — Length of body, 62.7-73.8 m (average, 67.2//). Greatest depth of body, 56.7-02.8 m- Though the size of the type specimen figured by Stein (1883, pi. 19, fig. 1) is not known it probably was somewhere between 69 fx and 104 fi. long; and the specimen represented by Stein (1883, p. 31, pi. 19, fig. 2) was somewhere between 86 ju and 128 m long. The specimen figured by Schutt (1895, pi. 4, fig. 19: 2) was about 70.5 ^ long and 61.6 ix deep; Okamura's (1912, pi. 5, fig. 77) about 54 yu long and 48.7 ^ deep; and Jorgensen's (1923) about 68 m long and 62 fi deep. According to Lohmann (1902, p. 53), this species is 86 M long. Variations: — In spite of the fact that Phalacroma doryphorum is "a very characteristic and easily recognizable species " (Jorgensen, 1923, p. 16), it exhibits a considerable variability in several respects. The following characters are more or less strikingly variable : — the size and shape of the body, the rela- tive height of the epitheca, the shape of the left sulcal list, the length and shape of the posterior main rib of this list, the position, direction, and struc- 178 THE DINOPHYSOID.\E. ture of the posterior sail, and the distance between this structure and the left sulcal list. Comparisons: — Under the name of Phalacroma doryphorum, Stein (1883, pi. 19, fig. 1, 2, 3) figures three specimens of fairly different habitus. One, the type specimen (pi. 19, fig. 1), has a regularly obovoidal body, deepest somewhat behind the girdle, and with comparatively high epitheca. The others (pi. 19, fig. Figure 23. — 1-S, Phalacroma danjiiJiorum .Stein, right lateral view. Surface markings, areolation, and jjorulation indicated only on .small portion of theca in 1 and 5. X 430. 1, from Station 4737 (300-0 fathoms); 2, from Station 4590 (300-0 fathoms); 3, 4, from Station 4664 (300-0 fathom.s); .5, from Station 4681 (300-0 fathoms). 6, Phalacroma cir- cumautuia Karsten, right lateral view. X 430. Station 46.38 (300-0 fathoms). 2, 3) are deepest at the girdle, and have the epitheca rather flat and the hypotheca of an oval wedge-shape. The question as to whether or not these differences are specific cannot be settled. Most of the specimens in the material of the Expedi- tion assigned to P. doryphorum agree fairly closely with the type specimen as far as the shape of the body is concerned, but differ in the shape and relative length of the left sulcal list and of its posterior main rilj, and in the position and direction of the posterior sail. In spite of these differences our assignments may be re- garded as fairly certain. The specimens figured by Schiitt (1895), Okamura (1912), and Jorgensen (1923) fall within the range of variations established for the Expedition specimens. The structurally closest-known relative of Phalacroma doryphorum is P. circumsutum. The first species differs from the last mainly in ha\'ing the left SYSTEMATIC ACCOUNT. 179 sulcal list and the posterior sail well separated; furthermore, the posterior main rib of its left sulcal list generally is club-shaped. Next to Phalacroma circumsutum, P. mucronatum appears to be the nearest- known relative of P. doryphormn. Phalacroma mucronatum probably agrees fairly closely with the ancestral form from which P. doryphormn evolved, as indicated by the following characteristics : — it is a small species, 35.0-45.4 m long, sub- circular in lateral view, with rather high epitheca, with the left sulcal list rela- tively long and not exceptionally wide posteriorly, and with the posterior main rib of this list not club-shaped. Phalacroma cuneolus and P. pugiunculus may possibly also belong to the same evolutionary series as P. doryphormn; the last is readily distinguished from these two species by its regularly obovoidal body and wider posterior sail. With regard to the possible relationship between Phalacroma doryphormn and P. ovum, see the section on comparisons (p. 120). Synonymy: — With regard to Jorgensen's (1923, pi. 17) assumption that figure G: 2, of Dinophysis galea Pouchet (1883), represents a specimen of Phala- croma doryphorum, see P. circumsutum, p. 184. Occurrence: — Phalacroma doryphorum is recorded at seventy-five of the 127 stations. There are 16, 21, 8, 9, 12, and 9 stations on the six lines of the Expedi- tion. Of these seventy-five stations, one (4583) is in the California Current; twelve (4587, 4588, 4590, 4592, 4594, 4596, 4598, 4600, 4604, 4605, 4545, 4546) are in the Mexican Current; nine (4609, 4611, 4613, 4617, 4619, 4631, 4634, 4637, 4644) are in the Panamic Area; twenty (4646, 4647, 4648, 4650, 4651, 4652, 4655, 4659, 4660, 4661, 4662, 4663, 4664, 4665, 4667, 4669, 4671, 4675, 4676, 4678) are in the Peruvian Current; five (4689, 4091, 4695, 4697, 4699) are in the Easter Island Eddy; two (4713, 4715) are in the Galapagos Eddy; twenty (4679,4681, 4683, 4701, 4707, 4709, 4711, 4717, 4719, 4720, 4721, 4722, 4724, 4728, 4730, 4732, 4734, 4737, 4739, 4740) are in the South Equatorial Drift; three (4742, 4743, 4540) are in the South Equatorial Current; two (4541, 4542) are in the Equatorial Counter Current; and one (4543) is in the North Equatorial Current. There are twenty-six records from the surface (Stations 4583, 4588, 4592, 4596, 4600, 4604, 4611, 4617, 4619 [Salpa], 4631, 4644, 4648, 4660 [Salpa], 4664 [Salpa], 4669, 4675, 4676, 4678, 4720 [Salpa], 4743, 4540, 4541, 4542, 4543, 4545, 4546) ; at four of these stations (4648, 4664, 4675, 4676) the species was taken in hauls from 300-0 fathoms as well as at the surface; at the other twenty-two stations in sur- face hauls only. At one station (4652) the species is recorded from 100-0 fathoms; at one station (4737) from 100-0 fathoms and 300-0 fathoms; at one station 180 THE DIXOPHYSOIDAE. (4655) from 400-0 fathoms; at two stations (4701, 4724) from 300-€ fathoms and 800-0 fathoms; at three stations (4647, 4751, 4662) from 800-0 fathoms only. All the remaining records refer to hauls from 300-0 fathoms only. The species is also recorded from surface waters in Acapulco Harbor, off the Mexican Current. This station is not included in the 127 stations mentioned above. The temperature range of these seventy-five stations at the surface was 65°- 85°; the average was 75.3°. At the twenty-six stations in the surface catches of Figure 24. — Occurrence of PhaUicroma dorijphorwn Stein. Large, .solid circles indicate records from vertical hauls; squares, records from surface hauls; triangles, records from both vertical and surface hauls; small, sohd circles, stations at which this species was not found; small, open circles, stations from which no plankton catches were examined. which this species was found, the surface temperature ranged from 07° to 84°; the average was 76.9°. At Acapulco it was 83°. For the surface catches the following frequencies are recorded: — 7% at one station (4546), 5% at one station (4619), 3% at one station (4604), 2% at two stations (4669, 4675), 1% at eight stations (4600, 4617, 4720, 4743, 4540, 4541, 4543, 4545), and less than 1% at the remaining stations. For the catches from 100, 300, 400, or 800 fathoms to the surface the records of frequency are as fol- lows:—5% at one station (4663), 4% at one station (4590), 3% at five stations (4598, 4664, 4667, 4713, 4715), 2% at three stations (4675, 4689, 4699), 1% at SYSTEMATIC ACCOUNT. 181 nineteen stations (4605, 4613, 4634, 4637, 4647, 4648, 4651, 4671, 4676, 4679, 4691, 4695, 4697, 4701, 4719, 4722, 4730, 4737, 4739), and less than 1% at the remaining stations. For the catch made in Acapulco Harbor the frequency of 1 % is recorded. The species was first recorded by Stein (1883) "aus dem Atlantischen Meer und der Stidsee." Murray and Whitting (1899), who recorded it from the At- lantic, between lat. 43°4' N. and lat. 0°16' S., and from the Caribbean Sea, found it to occur "pretty constantly over the whole region examined." According to Cleve (1901c, 1902b, 1903b) it occurs in the Atlantic, between lat. 57° N. and lat. 39° S., and in the Caribbean Sea; only one of Cleve 's several stations is located north of lat. 48° N., and this station, lat. 57° N., long. 14° W., is in the Gulf Stream. Hensen (1911) found it at a fairly great number of stations in various parts of the tropical and subtropical regions of the Atlantic. Lemmermann (1899a), Karsten (1906), and Stiiwe (1909) recorded it from the Gumea Current; Lohmann (1920) from the Brazil Current, lat. 15° S., long. 34° W.; Jorgensen (1923) from the Bay of Cadiz and off the southwest coast of Portugal. From the Mediterranean there are the followmg records. Jorgensen (1923) found it at a very great number of stations "throughout the whole of the Mediterranean." Cleve (1903b) recorded it at lat. 37° N., long. 2° W.-8° E.; Pavillard (1905, 1909, 1915a, 1916) in the Gulf of Lyons; Nathansohn (1908, 1909, 1910a) off Monaco; Forti (1922) in the Ligurian Sea; Schroder (1900a) at Naples; Entz (1902b, 1905), Schroder (1911), and Schiller (1912), in the Adriatic Sea. Cleve (1900b, 1903b) and Ostenfeld and Schmidt (1901) found it in the Red Sea; Ostenfeld and Schmidt (1901) and Schroder (l-906a) in the Gulf of Aden; Cleve (1901a, 1903b) and Schroder (1906a) in the Arabian Sea; Cleve (1901a) and Karsten (1907) in the Indian Ocean; Schroder (1906a) near Singapore; Schmidt (1901) in the Gulf of Siam; Okamura (1912) in Japanese waters; and Lemmermann (1899a) in the Pacific Ocean at lat. 6° N., long. 111° W. and lat. 12° N., long. 117° W. Most of the records referred to in the last paragraph were from surface hauls. The only records of closing-net hauls published as yet are the following by Karsten (1907) :— Station 227, lat. 2°56' S., long. 67°59' E., 1000-800 m.; Station 228, lat. 2°38' S., long. 65°59' E., 420-350 m.; Station 229, lat. 2°38' S., long. 63°37' E., 1600-1400 m.; Station 2.39, lat. 5°42' S., long. 43°36' E., 120-105 m.; Station 268, lat. 9°6' N., long. 53°41' E., 105-88 m. and 63-46 m. All the specunens recorded were reported as "living." 182 THE DINOPHYSOIDAE. According to these authors, this species occurs in waters of the following temperatures and salinities. Number and Number and Mean of Mean of Temperature Range Observations Salinity Range Observations Murray and AMiitting (1899) OO'-SS" Cleve (1901a) 79.r-86.7° 34.77-34.69 Cleve (1901c) 57.9°-82.8'' 55 71.6° 33.88-38.48 41 35.96 Cleve (1902b) 68.1° 11 36.12 Cleve (1903b) 60.3°-84.5° 14 75.9° 35.97-40.70 14 37.08 Ostenfeld & Schmidt (1901) 72.1''-S0.r Stuwe (1909) 6S.5''-75.2° Of the authors who have contributed toward our knowledge of the distribu- tion of this species only Stein (1883): Pavillard (1916), Okamura (1912), Forti (1922), and Jorgensen (1923) give descriptions or drawings by means of which their determinations of this species may be judged. This species appears to be one of the most common representatives of Dinophysoidae (Jorgensen, 1923, p. 17). It is eupelagic and widely distributed in tropical, subtropical, and warm-temperate seas. Occasionally it is carried by warm currents into cooler regions (Cleve, 1901c). According to the Expedition records it is the most common species of this tribe and is almost evenly distributed throughout the area investigated. Although it probably has its optimum habitat in deeper waters, within the levels of photosynthesis, it is rather frequent at the surface. Phalacroma doryphorum, P. rapa, P. cuneus, and P. favus occur more frequently in the surface catches than any other species of Phalacroma; the num- bers of surface records are 26, 22, 13, and 6, respectively; there are more surface records of P. doryphorum than of any other species of Phalacroma. Phalacroma circumsutum Karsten Figure 23:6 Phalacroma rirrumfiuttim Karsten, 1907, p. 421, pi. 53, fig. 8. Steuer, 1910, p. 476, 478, fig. 246a; 1911, p. 252, fig. 214a. Jorgensen, 1923, pnrlim., p. 17, 43. ?Dinoph)jsis galea Pouchet, 1883, parlim, p. 426, fig. (1 : 2. Diagnosis: — Body obovate in lateral outline, deepest somewhat in front of middle; length: depth, 1.20-1.23: 1. Posterior cingular list 0.24-0.20 the length of body from apex. Cingular lists without ribs, sometimes inclined anteriorly at 10°-30°. Left sulcal list: distance between Ri and R3 is 0.33-0.38 the length of body; R2 is 0.20-0.22, R3 is 0.42-0.55 the greatest depth of body; margin acumi- nate at R3 at angle of 35°-7()°; R3 inclined posteriorly at 25°. With acute, almost equilaterally triangular posterior sail supported by strong central rib, which is 0.43 the greatest depth of body, placed somewhat ventrally to antapex and in- SYSTEMATIC ACCOUNT. 183 clined ventrally at 15°-20°; angle at tip of sail, 50°. Left sulcal list and posterior sail connected by list with minimum width 0.3-0.5 the width of transverse furrow, with or without parasagittal list along dorsal side of hypotheca. Theca finely and closely areolate. Length, 75.7-80.0 /i. Widely distributed in tropical, subtropical, and warm-temperate seas. Description: — A medium-sized species, with the body obovate in lateral outline, deepest somewhat in front of the middle and 1.20-L23 times longer than deep. The longitudinal axis of the body is perpendicular to the girdle. The epitheca is 0.87-0.90 as deep as the hypotheca, highest in the center, of moderate convexity or rather flat, and moderately or but slightly elevated above the anterior cingular list. The transverse furrow is flat or but slightly convex; and its width is 0.50-0.60 the greatest height of the epitheca. The posterior cingular list is 0.24-0.26 the length of the body from the apex. The hypotheca is symmetrical; its ventral, posterior, and dorsal margins are confluent; the ventral and dorsal margins are of moderate convexity; the posterior margin is strongly convex. The posterior part of the body is strikingly narrower than the anterior. The cingular lists are subequal, and either subhorizontal or inclined ante- riorly at about 10°-30°; their width about equals or somewhat exceeds the width of the transverse furrow ; and they appear to lack ribs and other structural differ- entiation. The right sulcal list extends almost to the posterior main rib of the left sulcal list; its anterior half is about as wide as the cingular lists, i.e., it is decidedly narrower than the left sulcal list; posteriorly the list decreases gradually in width. The left sulcal list is of moderate length and rather wide posteriorly. The dis- tance between the anterior and the posterior main ribs is 0.33-0.38 the length of the body. The anterior main rib is 0.18-0.19, the fission rib 0.20-0.22, and the posterior main rib 0.42-0.55 the greatest depth of the body; behind the posterior main rib the list decreases suddenly in width. The free margin of this list is gently concave or almost straight between the anterior main rib and the fission rib, as well as between the fission rib and the posterior main rib ; at the posterior main rib it is more or less acuminate and forms an angle of 35°-70°; behind the pos- terior main rib it is gently concave. The main ribs of this list are of moderate strength and none of them is club-shaped or otherwise modified; the anterior main rib and the fission rib are straight or almost so, the posterior main rib is gently concave posteriorly. The distance between the anterior main rib and the fission rib is 0.30-0.45 the distance from the anterior to the posterior main rib. The pos- terior main rib has a posterior inclination of about 25°. Disregarding the three main ribs, this list appears to lack structural differentiation. On the right valve 184 THE DINOPHYSOIDAE. there is an acute, almost equilaterally triangular posterior sail supported by a strong rib which arises somewhat ventrally to the antapex and ends at the tip of the sail. This rib has a ventral inclination of 15°-20°, and its length is about 0.43 the greatest depth of the body; i.e., it is about as long as or somewhat shorter than the main rib of the left sulcal list. The margins of this sail are almost straight or slightly curved, and the angle at its tip is about 50°. The posterior sail and the left sulcal list are connected by a short, but well-developed list, which at its nar- rowest place is about 0.3-0.5 as wide as the transverse furrow. The connecting list and the posterior sail, disregarding the main rib of the latter, appear to lack structural differentiation. The type specimen (Karsten, 1907, pi. 53, fig. 8) had an apparently structureless parasagittal list extending, on the dorsal side of the hypotheca, from the posterior cingular list to the posterior sail but not forming a direct continuation of the latter; this list was widest posteriorly and decreased gradually in width anteriorly; its greatest width about equaled that of the trans- verse furrow. The structure of the theca is about the same as in Phalacroma dorypharum; i.e., the wall is finely and closely areolate and furnished with scattered pores. Megacytic stages have not been seen. Dimensions: — Length of body, 75.7 ii. Greatest depth of body, 61.5 m. Type specimen (Karsten, 1907, pi. 53, fig. 8) : Length of body, 80.0 fi. Greatest depth of body, 66.7 n. Variations: — Very little is known about the variability of this species. The Expedition specimen differs from the type mainly in being slight Ij' smaller (75.7 M as compared with 80.0 m), in having a slightly higher epitheca and a some- what longer posterior main rib in the left sulcal list, and in lacking a parasagittal list along the dorsal side of thehj^potheca. The specimen represented by Pouchet's (1883) figure G:2, of Dinophysis galea has been disregarded because of the im- possibility of bemg certain that his figure represents P. circumsutum. Comparisons: — In spite of the several differences mentioned in the last section between the type of Phalacroma circumsutum and the Expedition speci- men, the determination of the last seems justifiable. The differences in the size of the body, in the height of the epitheca and in the length of the posterior main rib of the left sulcal Ust appear of httle systematic importance, since they are within the range of variation established for the closely-related species, P. dory- pliorum. The "parasagittal list" along the dorsal side of the hypotheca in the type of P. circumsutum may be the intercalary zone not yet resorbed following binary fission. It should be noted that this Ust does not form a direct continua- SYSTEMATIC ACCOUNT. 185 tion of the posterior sail as it does in P. limhatum, P. pulchruni, and in other species of this genus. Furthermore, in the closely related P. doryphorum, Pavil- lard (1916, p. 53) frequently found specimens "en voie de reconstitution portant encore des fragments irreguliers de la zone iptercalaire le long de la ligne de suture." Phalacroma circumsutum is structurally very similar to P. doryphorum (see Karsten, 1907, p. 426; and Jorgensen, 1923, p. 17), differing mainly in having the left sulcal list and the posterior sail connected by a rather wide sail; in addition, the posterior main rib of its left sulcal list is not club-shaped. See also P. dory- phorum, p. 178. Synonymy: — Phalacroma circum^ulum Jorgensen (1923) evidently includes two systematic units. Jorgensen (192.3, p. 17) writes as follows: "Besides a larger form, answering to Karsten's figure, with pro- tracted lower end of left longitudinal fin (at the supporting spine) and large, somewhat ventrally directed terminal spine, there occurred in the 'Thor' material also a smaller and divergent form, with downward trending terminal spine, and slightly prominent corner of left longitudinal fin, shape of cell in profile broader in proportion, and epitheca higlier." Of these two "forms" the larger probably should be as- signed to P. circumtsutum Karsten; the smaller, which is figured (Jorgensen, 1923, fig. 18), belongs to P. pulchnim Kofoid and Michener. Of the four specimens represented by Pouchet (1883) under the name of Dinophysis galea, one (Pouchet, 1883, fig. G: 2) resembles P. circumsutum in the .shape of its body and in having its large, tri- angular posterior sail connected with the left sulcal list by a rather wide sail. It differs from this species in having the left sulcal list relatively narrow posteriorly, in having two instead of one posterior main rib in this list, in lacking a central rib in its posterior sail, and in having its posterior sail directed posteriorly instead of posteroventrally. Jorgensen (1923, p. 37) assigns this specimen to P. doryphorum. This iden- tification is contradicted by the fact mentioned above that, according to Pouchet's (1883) drawing, this specimen had the posterior sail connected with the left sulcal fist by a fairly wide sail and by the shape of the left sulcal list. It may be noted in this connection that Pavillard (1916) foimd P. doryphorum to be common in the Gulf of Lyons, which is the type locality of Dinophysis galea; on the other hand, he never recorded Phalacroma circumsutum from these waters in spite of very intensive investigations. This fact, of course, seems to .support Jorgensen's (1923) identification above mentioned. Pouchet's figure must be considered as insufficient for certainty of specific assignment until the question as to the occasional occur- rence in P. doryphorum of a connection between the left sulcal list and the posterior sail is settled and until the variability of the left sulcal list and of the posterior sail in P. circumsutum has been further investi- gated. The four specimens figured by Pouchet (1883) under the name of Dinophysis galea belong to four different species, two of which (fig. G: 1, 4) are referable to Ornithocercus, one (fig. G:2) to Phalacroma, and one (fig. G: 3) to Ceratocorys. Since of the.se four figures, figure G: 2 appears to be the most accurate, it is suggested that it represent D. galea. This name antedates both doryphorum and circumsutum. Occurrence: — This species is recorded at only one of the 127 stations. This station (4638) is on the second line of the Expedition and in the Panamic Area. The depth is 300-0 fathoms, the surface temperature 75°, and the frequency 1% (one specimen). Karsten (1907), who found this species in the Indian Ocean, did not give any information as to the type locality. Jorgensen (1923) recorded it from a few sta- tions in the Mediterranean (the Thor Expedition) and from the Guinea Current (the German South Pole Expedition). With regard to Jorgensen's (1923) data, see above. 186 THE DINOPHYSOIDAE. This species, which is eupelagic, appears to be very rare and limited to tropi- cal, subtropical, and warm-temperate waters. Phalacroma cuneolus, sp. nov. Figure 22:1, 2, 3,5, 7 Diagnosis: — Body subobovate in lateral outline; epitheca moderately con- vex to rather flat; ventral margin of hypotheca may form broadly rounded corner at R3 of left sulcal list; deepest somewhat behind girdle; length: depth, 1.10-1.28:1. In dorsal view cuneate; hypotheca acute to subacute posteriorly, its side contours gently concave or gently undulating; length : width, 1.44: 1. Pos- terior cingular list 0.22-0.33 the length of body from apex. Cingular lists may be somewhat inclined anteriorly. Left sulcal list of moderate length and width; distance between Ri and R2 0.45-0.52 the length of body; R2 is 0.13-0.20, and R3 is 0.26-0.31 the greatest depth of body; margin forms angle of 70°-100° at R3; R3 incHned posteriorly at 40°-55°. With rather narrow, triangular posterior sail, located on antapex and directed posteriorly; 0.18-0.31 the greatest depth of body in length, and 0.40-0.80 as wide as long; often connected with left sulcal list by narrow list. Theca finely and closely areolate. Length, 59.0-74.0 ix. Eastern tropical Pacific. Description: — A medium-sized species, the body of which is subobovate in lateral outline, with moderately convex to almost flat epitheca, with the posterior portion of the hypotheca more or less narrowly rounded, with the ventral margin of the hypotheca sometimes broadly rounded-angular at the posterior main rib of the left sulcal list, and with the greatest depth somewhat behind the girdle. The ratio between the length and the depth of the body is I.IG (1.10-1.28): 1. The longitudinal axis is perpendicular to the girdle, or it has a slight postero- ventral inclination. The epitheca is 0.88 (0.86-0.90) as deep as the hypotheca, evenly and moder- ately convex to rather flat, highest in the center, and moderately to rather slightly elevated above the anterior cingular list. The transverse furrow is flat, or slightly concave or convex, and its width is 0.60-1.10 the greatest height of the epitheca. The posterior cingular list is 0.26 (0.22-0.33) the length of the liody from the apex. Some specimens (Figure 22:3) have the hypotheca subsymmetrical. In these specimens the dorsal and ventral margins of the hypotheca are of moderate to gentle convexity, sometunes with almost even curvature, sometimes more or less flattened posteriorly; the postmargin is confluent with the dorsal and ventral margins, subrectangular, gently convex or almost straight dorsally and ventrally, SYSTEMATIC ACCOUNT. 187 and strongly rounded at the antapex. Other specimens, the type (Figure 22: 1), which are somewhat asymmetrical, have the dorsal margin of the hypotheca of about the same shapes as the subsymmetrical specimens, but the ventral margin is more or less irregular, being broadly rounded-angular at the posterior main rib of the left sulcal list; the postmargin is subrectangular, rather narrowly rounded at the antapex, and its ventral portion is gently convex to straight, or even slightly concave. Between these two types of hypotheca, transitional forms are to be found (Figure 22: 2). In dorsoventral view the body is cuneate, widest at the girdle, and about 1.44 times longer than wide. The epitheca is broadly rounded to rather flat; the hypotheca is acute to subacute posteriorly, and its side contours are gently undulating, as in the type specimen (Figiu-e 22: 5), or gently concave or convex. The cingular lists are subhorizontal or somewhat (10°-25°) mclined anteri- orly, subequal, and about as wide as or slightly wider than the transverse furrow; their structure is unknown. The right sulcal list (observed in only one of the specimens; Figure 22:3) extends to a point about halfway between the fission rib and the posterior main rib of the left sulcal list; it is subtriangular in shape, decreasing gradually in width posteriorly, and anteriorly about half as wide as the transverse furrow. The left sulcal list is of moderate length and width. The distance between the anterior and posterior main ribs is 0.47 (0.45-0.52) the length of the body. The anterior main rib is 0.16 (0.13-€.18), the fission rib 0.16 (0.13-0.20), and the posterior main rib 0.29 (0.26-0.31) the greatest depth of the body. Between the anterior and posterior main ribs the free margin of this list is straight, gently convex or concave, or shghtly irregular (Figure 22: 1); at the posterior main rib it forms an angle of about 90° (70°-100°). When this list is not connected with the posterior sail, its margin behind the posterior main rib is straight, or slightly concave or convex. The main ribs of this list are of moderate strength, straight or almost so, and not club-shaped or otherwise modified. The distance between the anterior main rib and the fission rib is 0.33 (0.29-0.40) the distance between the anterior and posterior main ribs. The posterior main rib has a posterior mclination of 40°-55°. It is not known whether this list has any structure besides the three main ribs. On the right valve there is an acute, rather narrow, triangular posterior sail, which is located on the antapex and directed posteriorly. When fully developed its length is 0.18-0.31 the greatest depth of the body, and it is, at the base, 0.40-0.80 as wide as long. Its margins are almost straight, or gently convex, concave, or sigmoid, and the angle at its tip is 20°-40°. Its structure is unknown. In some specimens (Figure 22: 3) the posterior sail is 188 THE DINOPm'SOIDAE. separated from the left sulcal list by a distance that about equals or even exceeds its basal width. In other specimens (Figure 22: 1, 2) these two structures are connected by a rather narrow Ust. The thecal wall is finely and closely areolate. Porulation was not observed. Megacytic stages were seen. The dimensions of five specimens were measured. Dimensions: — Length of body, 59.0-74.0 ix (average, 64.2 n; type, 60.3 m). Greatest depth of body, 51.3-58.0 m (average, 54.3 /x; type, 51.3 n). Variations: — This species is rather strikingly variable in the following characters : the shape and relative height of the epitheca, the shape of the ventral margin of the hypotheca, the shape of the side contours of the hypotheca in dorso- ventral view, the development of the list connecting the left sulcal list with the posterior sail, and the inchnation of the cingular list. The structurally closest-known relative of Phalacroma cuneolus is P. pugiun- culus. These two species agree rather closely in the shape of the body in lateral view, in the shape and relative size of the left sulcal list, and in the shape, direc- tion, and relative size of the posterior sail. The outline of P. pugiunculus in dorso- ventral view is unknown. The latter species can be distinguished from the former by its smaller size (49-53 n as compared with 59-74 fj.). The position of these two species is not readily established because of their smiphcity. We have placed them preliminarily with P. mucronatum, P. doryphorum, and P. circumsutum on account of the general resemblance in the shape of the body in lateral view and because of the presence of a posterior sail. • The outline of the body of Phalacroma cuneolus in dorsoventral \'iew is sug- gestive of P. cuneus and Dinophysis expulsa and their near relatives. However, this similarity probably is due to convergence and can hardlj' be considered as an indication of close relationships. Occurrence: — Phalacroma cuneolus is recorded at onlj^ one (4711, type locality) of the 127 stations, on the fourth line of the Expedition, in the South Equatorial Drift, from 300-0 fathoms, and at a surface temperature of 75°. The frequency is less than 1 % (six specimens), 9. Praetextum Group. The single member of this group, viz., Phalacroma praetextum, differs so strikingly from the remaining species of Phalacroma in having a ventral cribriform plate on the hypotheca that it is made the type of a new subgenus, Ethmophalacroma, subgen. nov. SYSTEMATIC ACCOUNT. 189 Phalacroma praetextum Kofoid and Michener Plate 4, fig. 6, 7. Figure 25: 4, 5 Phalacroma praetexla Kofoid & Michener, 1911, p. 291. Diagnosis: — Body rounded subbiconical in lateral outline, slightly broader posteriorly than anteriorly; deepest at girdle, 1.12 times longer than deep. In dorsal view subbiconical, with broadly roxinded apices, and 1.17 times longer than wide. Posterior cingular list 0.51-0.53 the length of body from apex. Cingular lists about half as wide as transverse furrow, reticulated. Left sulcal Ust 0.34 the length of body, subtriangular, gradually decreasing in width posteriorly; anteriorly about as wide as transverse furrow; Ri and Ro weak, straight, and simple, Rs absent. On ventral side of hypotheca a dumbbell-shaped, cribriform plate extending from girdle to antapex. Theca with heavy and rather wide- meshed reticulation ; about fifteen polygons border the girdle posteriorly. Length, 61-62 M. Eastern tropical Pacific. Description: — A rather small species, the body of which is rounded subbi- conical in lateral outlme, shghtly broader posteriorly than anteriorly, and deepest at the girdle. The ratio between the length and the depth of the body is 1.12: 1. The longitudinal axis is about perpendicular to the girdle. The epitheca is about as deep as or slightly deeper than the hypotheca, highest in the center, subcorneal, gently convex dorsally and ventrally, evenly and fairly broadly convex anteriorly, with the dorsal, anterior, and ventral margins confluent, and very prominent above the anterior cingular list. The transverse furrow is somewhat concave and its width is about 0.22-0.27 the great- est height of the epitheca. The posterior cingular Ust is 0.51-0.53 the length of the body from the apex. The hypotheca has almost the same shape as the epi- theca but is somewhat more rounded; its posterior portion, therefore, is somewhat broader than the anterior portion of the epitheca; see the description of the struc- ture of the theca. In dorsoventral view the body is subbiconical, with fairly broadly rounded apices, and with the side contours of the epitheca and of the hypotheca almost straight or shghtly concave ; it is widest at the girdle, and about 1.17 times longer than wide; see below, the description of the structure of the theca. The cingular lists are subhorizontal, subequal, and about half as wide as the transverse furrow. Along the base of the anterior of these lists there is a series of ribs anastomosing into a row of polygons in the same way as for Phalacroma cuneus 190 THE DINOPHYSOIDAE. Schiitt (1895, pi. 3, fig. 14:2); the number of these polygons and the structure of the posterior cingular list are unknown. The sulcus is about half as long as the hypotheca. The flagellar pore is at the junction of the cingulum and the sulcus. On the ventral side of the left valve a fairly small pore is found near the sagittal suture, just behind the anterior cingular list. The right sulcal list is about half as long as the left sulcal list; posteriorly it gradually decreases in width, or its ventral margin is gently sigmoid, convex anteriorly and concave posteriorly; its greatest width somewhat exceeds half the width of the transverse fiu"row. The left sulcal list is 0.34 the length of the body, subtriangular, gradually decreasing in width posteriorly ; anteriorly it is about as wide as the transverse furrow; it has two fairly weak, straight, and simple ribs, both of which are located in the an- terior half of the list; the posterior main rib is absent (Plate 4, fig. 7). There are no accessory lists or sails. On the ventral side of the hypotheca, just to the left of the sulcus, there is a dumbbell-shaped area, extending on both sides of the sagittal suture from the posterior cingular list to the antapex. Due to the fact that it is bordered by the posterior cingular list and by the sulcus, its anterior half is truncate and some- what asymmetrical. The constriction in the middle is rather slight. This area, which has a maximum width of about 0.33 the dorsoventral diameter of the transverse furrow, is characterized by a structure quite different from that of the remaining portion of the theca. It has a faint reticulum of very small polygons, and in the center of each or at any rate of most of these polygons there is an ex- ceedingly fine pore. In other words, this area has the structure of a cribriform plate. It is not known whether this area is permanent, or whether it is temporary and formed just before fission. The first alternative, which appears the more plausible, is suggested by the relative position of this area to the girdle, by the absence of an intercalary zone from the dorsal side of the body and from the ventral side of the epitheca, by the peculiar shape and structure of this area, and by the fact that an area of this kind was found in both the specimens. If this area is of an intercalary nature, then the non-megacytic specimens probably have the ventral margin of the hypotheca resembling that of Phalacroma reiiculatum, and the posterior portion of their hypotheca is somewhat narrower than in the speci- mens described above. The thecal wall has a heavy reticulum of rather large polygons, except in the cribriform plate described in the last paragraph. Most of the polygons are sub- uniform in size, and on each valve about fifteen of them border the po.sterior margin of the girdle. Each valve of the epitheca has about fifty-eight and each SYSTEMATIC ACCOUNT. 191 valve of the hypotheca about seventy-six polygons. In the transverse furrow there are two rows of polygons, each row with fourteen to sixteen polygons on each valve ; near the junction of the cingulum and the sulcus these two rows may merge into a single row. Except in the cribriform plate, pores have not been seen. With regard to megacytic stages, see p. 190. The dimensions of one specimen, the type, were measured. Figure 25. — 1, Phalacroiim Jiinbriatum Kofoid and Michener, right lateral view of type specimen. X 430. Station 4613 (300-0 fathom.s). 2, Phalacroma reticulatum Kofoid, right lateral view of type specimen. X 430. Station 4740 (300-0 fathoms). 3, Phalacroma turhineum Kofoid and Michener, right lateral view of type specimen. X 430. Station 4681 (300-0 fathoms). 4, 5, Phalacroma praetextum Kofoid and Michener; 4, oblique ventral view of detached left valve; 5, right lateral view of type specimen. X 430. Station 4742 (300-0 fathoms). Dimensions: — Length of body, 61-62 ^ (type, 61 m). Greatest depth of body, 53.3 M (type). Comparisons: — The small size, the subcircular shape of the body in lateral view, the subequatorial position of the girdle, the relatively slight development of the cingular and sulcal lists, and the absence of accessory lists and sails are funda- mental primitive features, which place this species among the most primitive Uving representatives of Phalacroma. On the other hand, the surface structure of the theca is better developed than in most species of the genus, and the cribriform plate of the hypotheca is unique and places the species in a rather isolated posi- tion. The species therefore appears to be a representative of a short evolutionary branch which split off at an early stage in the phylogenetic differentiation of the genus. ,, Phalacroma praetextum approaches P. fimbriatum, P. reticulatum, and P. turbineum in the tendency of its epitheca and hypotheca to taper toward the 192 THE DINOPHYSOIDAE. apices, and in the heavy and coarse reticulation of its thecal wall. However, in most respects it differs more or less strikingly from these three species, and it must be regarded as an open question whether or not the three characters men- tioned above are indicative of a close relationship. PJialacroma praetexlum re- sembles P. a-picatum in the subbiconical shape of the bodj' in dorsoventral view, in the reticulated structure of the anterior cingular list, and in the absence of posterior angularity in the free margin of the left sulcal list. Phalacroma -prae- texlum is easily distinguished from all the known species by the large cribriform plate of its hypotheca. Occurrence: — Phalacroma praetexlum has thus far been recorded at only one (4742, the type locality) of the 127 stations, on the sixth line of the Expedition, in the South Equatorial Current, from 300-0 fathoms, and at a surface tempera- ture of 77°. The frequency is less than 1% (two specimens). 10. Reticulatum Group. On account of the striking difference between the left sulcal list of the three members of this group, viz., Phalacroma fimbriatum, P. reliculatum, and P. lurbineum, and that of the typical members of Phalacroma, it appears advisable to give this group the rank of a subgenus : Retephalacroma, subgen. nov., with P. lurbineum the type. Phalacroma fimbriatum Kofoid and Michener Plate 2, fig. 1. Plate 4, fig. 4. Figure 25: 1 Phalacroma fimhriata Kofoid & Michener, 1911, p. 289. Diagnosis: — Body subbiconical in lateral view, with well-rounded apices, deepest at girdle, 1.35-1.40 tunes longer than deep. In dorsal view biconical, subacute anteriorly, narrowly rounded posteriorly; but slightly narrower than in lateral view. Posterior cingular list 0.43 the length of body from apex. Each cingular list with twelve branched ribs on each ^■alve. Margin of right sulcal list widely angular in the middle. Two parasagittal lists encircle body ; the right one, which is much wider than the left, merges so completely into left sulcal list that the boundary between these two lists cannot be estabUshed. Right parasagittal list widest and of nearly uniform width on ventral and posterior sides of hypo- theca, where it is 0.14-0.20 the greatest depth of body ; on dorsal side of hypotheca and on epitheca it is about half this width. Parasagittal lists with numerous ribs, some of which are irregular and branched. Theca with heavy reticulation, form- ing large polygons; about ten polygons border girdle posteriorly. Length, 118 m- Eastern tropical Pacific. SYSTEMATIC ACCOUNT. 193 Description: — A large species, the body of which is subbiconical in lateral outUne, with well-rounded apices, and deepest at the girdle. The ratio between the length and the depth of the body is 1.35-1.40: 1. The longitudinal axis has a very shght postero ventral inclination. The epitheca is about as deep as the hypotheca, highest in the center, sub- conical, gently concave or almost straight dorsally and ventrally, evenly and well rounded anteriorly, with the dorsal, anterior, and ventral margins confluent, and very promment above the anterior cingular list. The transverse furrow is some- what concave, and its width is about 0.19 the greatest height of the epitheca. The posterior cingular list is about 0.43 the length of the body from the apex. The hypotheca has a very slight posteroventral inclination; its dorsal margin is gently concave or almost straight; its ventral margin is sigmoid, rather strikingly con- cave anteriorly, and gently and evenly convex posteriorly; its postmargin is evenly and fairly broadly convex, and confluent with the dorsal and ventral margins. In dorsoventral view the body is biconical, subacute anteriorly and narrowly rounded posteriorly; the side contours of the epitheca and hypotheca are gently concave; the ratio between the length and the width is about 1.45- 1.50:1. The cingular lists are subhorizontal, subequal, and about as wide as or some- what wider than the transverse furrow; each of them has on each valve about twelve rather strong, branched, and nearly equidistant ribs. The sulcus is about 0.40 the length of the hypotheca. The flagellar pore is slightly more than a girdle- width behind the junction of the cingulum and the sulcus. On the ventral side of the left valve a rather small pore is found on the sagittal suture just in front of the anterior cmgular list. The right sulcal list is about 0.22 the length of the body ; the anterior half of this list is about 0.14 as wide as the body, and its free margin is gently concave or almost straight; the posterior half decreases gradually in width posteriorly, and its free margin is gently concave or nearly straight and forms a very wide but distmct angle with the free margin of the anterior half of this list. Two parasagittal lists encircle the body. The right one, which is much wider than the left, merges so completely into the left sulcal list that the boundary between these two lists cannot be established with certainty. For the sake of convenience in describing these structures, it is assumed that the first rib of this compound list behind the fission rib (easily recognized in the type specimen as it is split lengthwise) corresponds to the posterior main rib of the left sulcal list, and that it marks the posterior boundary of the last-mentioned list. Defined in this way, the left sulcal list is about as long as the right sulcal list, and has four 194 THE DINOPHYSOIDAE. simple, and almost straight and equidistant ribs. Three of these ribs would cor- respond to the three main ribs; the fourth, which is placed between the anterior main rib and the fission rib, would be an accessory structure. The width of the left sulcal list is almost uniform throughout the entire length and about equals 0.14-0.15 the greatest depth of the body. From the posterior main rib of the left sulcal list to the antapex, the right parasagittal list is of nearly uniform \\idth, about 0.14-0.20 the greatest depth of the body; anterodorsally to the antapex this list gradually becomes narrower, and its average width on the dorsal side of the hypotheca is only about 0.07 the greatest depth of the body. The right para- sagittal list of the hypotheca has about twentj'-sbc fairly strong and almost equi- distant ribs, most of which are simple and almost straight, a few irregular, or irregular and branched. The left parasagittal list of the hypotheca extends from the fission rib of the left sulcal list to the posterior cingular list on the dorsal side of the body ; it is very narrow, at most about half as wide as the transverse fur- row; its rilibing is unknown. The two parasagittal lists of the epitheca are of about the same width and structure; from the apex, where their width about equals 0.09 the greatest depth of the body, they slightly decrease in width tow ard the girdle; each of them has about twenty almost equidistant ribs, a few of which are branched. The thecal wall has a heavy reticulum of rather large i)olygons, mainly sub- uniform, ten of which on each valve border the posterior margin of the gii-dle. Each valve of the epitheca has about forty and each vaWe of the hypotheca about sixty-seven polygons. The transverse furrow is divided by cross-ridges into rectangular fields, of which there are about twelve or thirteen on each valve. The whole theca has numerous scattered pores, about ten to twenty in each polygon. On the left xaWe of the hypotheca there is the unique feature of a large postcingular plate, bordered by a heavy ridge (Plate 2, fig. 1). The type specmien is megacytic. The dimensions of the type specimen were measured. Dimeiisions: — Length of body, 118 m (140 /i, with the parasagittal lists in- cluded). Greatest depth of body, 83 ii. Comparisons: — The systematic position of Phalacroma fimbriatum is some- what uncertain. In the size of the body, the shape of the body in dorsoventral view, and in the structural differentiation of the thecal wall, both in the reduction in number of polygons, and in the increase in their size and in the weight of the reticulations, this species is one of the most ad\anced members of the genus, and with regard to the development of the parasagittal lists it may be considered as SYSTEMATIC ACCOUNT. 195 the most advanced in the genus. On the other hand, m the position of the girdle it is very primitive. Its closest structural relatives possiblj^ may be P. reticulatum and P. turbineum; see these two species, the sections on comparisons (p. 197, 200). These three species differ strikingly from one another in the shape of the body in lateral view and in the parasagittal Usts. With regard to the possible relationship between Phalacroma fimbriatum and P. ■praetextuni, see p. 191. Occurrence: — Phalacroma fimbriatum has been recorded at only one (4613, the type locality) of the 127 stations, on the first line of the Expedition, in the Panamic Area, from 300-0 fathoms, at a surface temperature of 80°. The fre- quency is less than 1 % (one specimen). Phalacroma reticulatum Kofoid Plate 4, fig. 3, 5. Figiu-e 25: 2; 26 Phalacroma relicuiata Kofoid, 1907a, p. 195, pi. 12, fig. 72. Kofoid & Michener, 1911, p. 291. Pavil- LARD, 1923a, p. 879. Diagnosis: — Body biconical in lateral view, narrowly rounded anteriorly, narrowly rounded to subacute posteriorly, deepest at girdle, 1.25-1.30 times deeper than long. In dorsal view biconical, subacute anteriorly, acute posteriorly, and 1.65 times longer than wide. Posterior cingular list 0.40-0.45 the length of body from apex. Cingular list somewhat narrower than transverse furrow and apparently without ribs. Left sulcal list about 0.33 the length of the body; fau-ly narrow; its maximum width is 0.08 the greatest depth of body; with one rib near posterior end; free edge gently sigmoid, concave anteriorly and convex posteriorly. With a narrow parasagittal list on the hypotheca; this list forms a direct continua- tion of left sulcal list and extends to posterior cingular list. Theca with heavy and wide-meshed reticulation; about seven or eight polygons border girdle posteriorly. Length, 100-115 ti. Eastern tropical Pacific. Description: — A rather large species, the body of which is biconical, top- shaped in lateral outline, narrowly rounded anteriorlj^ subacute to acute posteri- orly, and deepest at the girdle. The ratio between the length and the depth of the bod}' is 1.25-1.30: 1. The longitudinal axis is about perpendicular to the gu-dle or but slightly deflected posteroventrally. The epitheca is about as deep as or but slightly deeper than the hypotheca, highest in the center, and very prominent above the anterior cingular list; its dorsal and ventral margins are almost straight or gently convex, and about per- 196 THE DINOPHYSOIDAE. pendicular to each other; anteriorly the epitheca is narrowly rounded. The trans- verse furrow is flat or somewhat convex or concave, and its width is 0.15-0.20 the greatest height of the epitheca. The posterior cingular list is 0.40-0.45 the length of the body from the apex. The hypotheca is conical, narrowly rounded or even subacute (Figure 25: 2) posteriorly ; its dorsal margin (from the girdle to the antapex) is sUghtly and evenly convex or almost straight; its ventral margin (from the girdle to the antapex) is either gently and almost evenly convex, or it is gently undulating, concave anteriorly and posteriorly, and convex in the middle (Figiu-e 25: 2). In dorso ventral view the body is biconical, top-shaped, widest at the girdle, and 1.65 times wider than long; anteriorly it is nan-owly rounded, posteriorly it is acute; the side contours of the epitheca are almost straight and nearly perpendicular to each other; those of the hypotheca are gently sigmoid, concave anteriorly and convex posteriorly. The cingular lists are subhorizontal, subecjual, somewhat narrower than the transverse furrow, and apparently without ribs. The sulcus is about half as long as the hypotheca. The right sulcal list is low, extends to the posterior rib of the left sulcal Ust, and is reticulated posteriorly. The left sulcal list is about 0.33 the length of the body, and fairly narrow, its maximum width being about 0.08 the greatest depth of the body; its free edge is gently sigmoid, concave anteriorly and convex posteriorly. Near its posterior end this list has a short, straight, and simple rib of moderate strength; with the exception of this rib, this list appears to lack structural differentiation. On the right valve of the hypotheca there is a narrow parasagittal list, which forms a direct continuation of the left sulcal list and extends to the posterior cingular list; this parasagittal list is of almost uni- form width thi'oughout, somewhat narrower than the cingular lists, and is fur- nished with a moderate number (15-20) of sunple ribs. There are no accessory lists or sails on the left valve of the hypotheca and on the epitheca. The thecal wall has a heavy reticulum of relatively large polygons. Gener- ally speakmg, the polygons increase in size from the two apices toward the girdle, and on each valve about seven or eight of them border the posterior margin of the girdle. Each valve of the epitheca has a total of about twenty-four and each valve of the hypotheca about thirty-five to forty polj'gons. The trans\'erse fur- row is divided by cross-ridges into a number of rectangular fields, about ten fields on each valve. Pores were not seen. Megacytic stages were not recorded. The dimensions of two specimens were measured. Dimensions: — Length of body, 100-115 m (type, 100 /x). Greatest depth of body, 77-84 ^ (type, 77 n). SYSTEMATIC ACCOUNT. 197 Variations: — The shape of the hypotheca is rather variable. In some speci- mens the hypotheca is subacute posteriorly, and its ventral margin (from the girdle to the antapex) is gently undulating, concave anteriorly and posteriorly, and convex in the middle. In other specimens the hypotheca is narrowly rounded posteriorly, and its ventral margin is gently and almost evenly convex. Comparisons: — The structurally closest-known relative of Phalacroina reticulatum is P. turbineum. This relationship is especially indicated by the biconical, top-like shape of the body, by the heavy and coarse reticulation of the Figure 26. — Occurrence of Phalacroma reticulatum Kofoid. Large, solid circles indicate records from vertical hauLs; small, solid circles, stations at which this species was not found; small, open circles, stations from which no plankton catches were examined. thecal wall, and by the presence of parasagittal lists. Phalacroma reticulatum is more advanced than P. turbineum in size, but in most other characters it appears to be more primitive; for instance, its body is relatively deeper, its hypotheca is relatively shorter, the reticulation of its thecal wall is less coarse, and it has only one parasagittal list on the hypotheca, while P. turbineum has two such lists on the epitheca as well as on the hypotheca. Phalacroma jimbriatum may be fairly closely related to P. reticulatum. This relationship, which is less evident than that mentioned in the last paragraph, is indicated by the biconical, top-like shape of the body in dorsoventral view, by the heavy reticulation of the thecal wall, and by the presence of parasagittal lists. 198 THE DINOPHYSOIDAE. When compared with P. fimbriatum, P. reticulatum is less advanced with regard to the size of the body, and in the number, size, and structural differentiation of the parasagittal hsts, and it is more advanced in the top-like shape of its body in lateral view, in the position of its girdle, and in the coarser reticulation of its thecal wall. With regard to the possible relationship between Phalacroma rcliculatum and P. -praelextuvi, see the last species, the section on comparisons. Occurrence: — Phalacroma reticulatum is recorded at thirteen of the 127 stations. There are 0, 1, 2, 5, 4, and 1 stations on the six lines of the Expedition. Of these thirteen stations, one (4655) is in the Peru\ian (lUTent; three (4()95, 4697, 4699) in the Easter Island Eddy; nine (4679, 4681, 4701, 4709, 4724, 4730, 4732, 4739, 4740) in the South Equatorial Drift. At one station (4655) the species is recorded from 400-0 fathoms; at one (4681) from 800-0 fathoms and 300-0 fathoms. All the remaining records refer to hauls from 300-0 fathoms only. The temperature range of these thirteen stations at the surface was 68°-81°; the average was 74.4°. At one station (4699) the frequency is 1 % ; in the remaining cases it is less. The species was first recorded by Kofoid (1907a) from Station 4740 of the Expedition; it was found in the Mediterranean by Pavillard (1923a). The most outstanding peculiarities of the distributioii of this species in the Eastern Pacific are the absence of surface records and the fact that all the record stations except one are located either in the South Equatorial Drift or in the Easter Island Eddy. Phalacroma turbineum Kofoid and Michener Plate 2, fig. 3. Plate 4, fig. 1, 2. Figure 25: 3 Phalacroma iwbinea Kofoid & Michener, 1911, p. 291. Diagnosis: — Body biconical in lateral outline, rather broadly rounded an- teriorly, narrowly rounded posteriorly; deepest at girdle, 1.84 times deeper than long. In dorsal view of similar shape but somewhat wider. Posterior cingular list about 0.23 the length of body from apex. Cingular lists about half as wide as transverse furrow, with few ribs. Right sulcal list with T-shaped rib. Left sulcal list small, 0.38-0.39 the length of body, and with maximum width 0.12-0.13 the greatest depth of body; free margin gently and evenly convex; with cross-rib somewhat in front of its middle and submarginal rib along its entire length. Two parasagittal lists encircle body; the one on right valve of hj'potheca continues left sulcal list; maximum width about equaling that of cingular lists; with moder- SYSTEMATIC ACCOUNT. 199 ate number of short ribs. Theca with heavy and very wide-meshed reticulum; five polygons border girdle posteriorly. Length, 69.5 m. Eastern tropical Pacific. Description: — A medium-sized species, the body of which is biconical, top- shaped in lateral outline, rather broadly rounded anteriorly, narrowly rounded posteriorly, and deepest at the girdle. The ratio between the length and the depth of the body is about 1 .84 : 1 . The longitudinal axis is about perpendicular to the girdle. The epitheca is about as deep as the hypotheca, highest in the center, and very prominent above the anterior cingular Ust; its dorsal and ventral margins are almost straight, or gently concave or convex; anteriorly it is fairly broadly rounded. The transverse furrow is somewhat conca^'e, and its width is about 0.40 the greatest height of the epitheca. The posterior cingular list is about 0.23 the length of the body from the apex. The hypotheca is elongated and conical; posteriorly it is narrowly rounded; its dorsal margin (from the girdle to the antapex) is gently undulating, convex anteriorly and posteriorly, and concave in the middle; its ventral margin (from the girdle to the antapex) also is gently undulating, but it is concave anteriorly and posteriorly and convex in the middle. In dorsoventral view the Ijody has about the same shape as in lateral view, but it is somewhat wider; the ratio between the length and the width is about 1.50: 1 ; the side contours of the hypotheca are evenly and moderately concave. The cingular lists are subhorizontal, subequal, about half as wide as the transverse furrow, and have a few (about seven or eight on each valve) simple, straight cross-ribs. The sulcus is relatively short, about 0.26 the length of the hypotheca. The flagellar pore is at the junction of the cingulum and the sulcus. A large pore is to be found in the transverse furrow, just in front of the flagellar pore. The right sulcal list is about 0.20 the length of the body, subtriangular, and decreases gradually in width posteriorly; its maximum width about equals half the width of the transverse furrow ; along its entire free edge runs a marginal rib, from the middle of which arises a cross-rib to the base of the list. The left sulcal Ust is small; its length is 0.38-0.39 the length of the hypotheca, and its maximum width is 0.12-0.13 the greatest depth of the body; its free margin is gently and evenly convex; it has a weak cross-rib somewhat in front of its middle and a submarginal rib along its entire length ; with the exception of these two ribs, this list appears to lack structvu-al diff'erentiation. Two parasagittal lists encircle the body; the one on the right valve of the hypotheca forms a direct continuation of the left sulcal list. On the average the parasagittal lists are about twice as wide 200 THE DINOPHYSOIDAE. on the posterior half of the shell as on the anterior half, and their maximum width about equals that of the cingular lists; each of them has a moderate number of short, simple ribs, approximately six on the epitheca and eighteen on the hy- potheca. On the ventral side of the body the distance between the two para- sagittal lists about equals the width of the transverse furrow ; on the dorsal side of the body it is about 2.5 times as wide at the girdle, and from the girdle it decreases gradually toward the apices. The cingular and the parasagittal lists arise from narrow but well-marked ridges, which are connected by similar cross-ridges. These cross-ridges divide the transverse furrow and the zone between the parasagittal lists into rectangular fields. The number of rectangular fields in the transverse furrow is about seven or eight on each valve ; most of these fields have a fine, central pore. The number of rectangular fields in the zone between the parasagittal lists is about five or six on the epitheca, and seventeen or eighteen on the hypotheca; in these fields no pores were found. The rest of the thecal wall has a heavy reticulum of very large polygons. Generally speaking, the polygons inci'ease in size from the two apices toward the girdle, and on each \'alve five of them border the posterior margin of the girdle. Each valve of the epitheca has about ten, and each valve of the hypo- theca about twenty-nine polygons. Pores were not seen in the polygons. The nucleus is large and rounded; its diameter is about 0.5 the greatest depth of the body. It Ues at the level of the girdle, is filled with beaded chromatin, and contains some spheroidal structures. With regard to megacytic stages, see the section on comparisons, below. The dimensions of the type specimen were measured. Dimensions: — Length of body, 69.5 ti. Greatest depth of bodj', 37.8 m. Variations: — The two specimens were very similar. Comparisons: — In the specimens of this species examined the two para- sagittal lists are separated from the sagittal suture by a broad zone. In premega- cytic specimens of other species of this genus with two parasagittal lists, these Usts are located near the sagittal suture (f/. P. Umhatum, Plate 3, fig. 3) ; in mega- cytic specimens of these species on the other hand, they are separated from this suture by the broad intercalary zone {cf. P.fimbriatum, Plate 2, fig. 1). The zone between the parasagittal lists in P. iurhincum has the same position, relative size, and shape (being wider on the dorsal side of the body than on the ventral) as the intercalary zone. On the other hand, it has not the juvenile structural appearance that usually characterizes the intercalary zone but is furnished with cross-ridges which are as well developed as the ridges on the remaining portion of the thecal SYSTEMATIC ACCOUNT. 201 wall. For this reason we consider the question open as to whether or not the Expedition specimens are megacytic. This beautiful and characteristic species is structurally most closely related to Phalacroma reticulatum (see p. 197). It may also be fairly closely related to P. fimbriatum, but this relationship is less evident. The last two species resemble each other in the biconical, top-like shape of the body in dorsoventral view, in the development of two parasagittal lists extending around the body, and in the heavy reticulation of the thecal wall. Phalacroma fimhriatxmi is much more ad- vanced than P. turhineuvi in the size of the body and in the size and sti-uctural differentiation of the parasagittal lists. On the other hand, P. turbineum is more advanced than P. fimhriatum in the elongated, conical shape of the hypotheca seen in lateral view, in the position of the gii'dle, and in the coarser reticulation of the thecal wall. Indeed, the polygons of the thecal reticulation are relatively larger than in any other known species of this genus. With regard to the possible relationship between Phalacroma turbineum and P. praeiextum, see p. 191. Occurrence: — Phalacroma turbineum is recorded at two of the 127 stations. One of these two stations (4681) is on the third line of the Expedition and in the South Equatorial Drift; the other (4715) is on the fourth line and in the Gala- pagos Eddy. Both records refer to hauls from 300-0 fathoms. The temperature of these two stations at the surface was 68° and 75°, re- spectively; the average was 71.5°. The frequency is less than 1% (one specimen at each station). DiNOFURCULA, gen. nov. Phalacroma Kofoid, partim (1906c, 1907a). Diagnosis: — Body molariform in lateral outline, with two long, narrow, and acute to narrowly rounded posterior processes; strongly compressed bi- laterally; its length 1.57-1.83 the depth of midbody. Epitheca 0.62-0.78 the greatest depth of midbody, low. Transverse furrow narrow, of subuniform width throughout, and gently concave to flat. Posterior cingular list 0.14-0.26 the length of body from apex. Shortest distance from curvature between posterior processes to posterior cingular list 0.26-0.61 the greatest depth of midbody. Distance between tips of posterior processes 0.80-0.93 the greatest depth of midbody. Cingular lists subhorizontal, subequal, slightly wider to slightly narrower than transverse furrow, and ribbed. Sulcal lists about as wide as or somewhat narrower than the cingular. 202 THE DINOPHYSOIDAE. Type. — Dinofurcula ultima Kofoid. Distribution: — Up till the present time, species of this planktonic genus have been found only in the eastern warm-temperate and tropical Pacific. Representatives of Dinofurcula are recorded- at six of the 127 stations of the Expedition. Four of these six stations (4655, 4659, 4661, 4671) are within the incl. v. post. pr. ant. cing. I. post. cing. I. V. post. pr. d. pest. jrr. Figure 27. — Diagram of measurements in Dinofurcula. Measurements are made in straight lines. Proportions are measured as if all parts were in the same plane, i.e., as they apiwar in drawings. Ijongi- tudinal axis of body is perpendicular to girdle. The length of body is distance between the two perpen- diculars to longitudinal axis tliat pass through apex of body and tip of dorsoposterior process. The depth of body is measured perpendicularly to longitudinal axis. The greatest height of epitheca is the shortest distance from apex of body and base of anterior cingular hst. The depth of epitheca is measured along base of anterior cingular list. The width of transverse furrow and of cingular lists is measured dorsally. The position of posterior cingular list i.s expressed in the ratio between the shortest distance from apex of body to the mentioned list and the length of body. The inclinations of the posterior processes are the angles between the longitudinal axis of the body and the long axis of these processes. Abtrrevialions: — ant. cing. I., anterior cingular list; d. post, pr., dorsoposterior process; epi., epitheca; f.p., flagellar pore; hyp., hypotheca; incl. v. post, pr., inclination of ventroposterior process; /., length of body; l.f., longitudinal furrow; I. sulc. I., left sulcal list; post. cing. I., posterior cingular li.st; r. sulci., right sulcal list; tr.f., transverse furrow; v. po.^t. jrr., ventroposterior process. Peruvian Current; two (4681, 4711) are outside this current, but undoubtedly within its sphere of influence. The genus has not been found as yet in the surface waters. All the records refer to hauls from 300 (400, 800)-0 fathoms. Data based on hauls made with closing nets are not available. Both the species known seem to be rare. The frequency is less than 1% at all the six stations mentioned. SYSTEMATIC ACCOUNT. 203 DiNOFURCULA ULTIMA (Kofoicl) Plate 5, fig. 4, 6. Figure 28: 1 Phalacroma idtima Kofoid, 1906c, p. 95; 1907a, p. 19.5, pi. 12, fig. 6S. Diagnosis: — Body broadly molariform in lateral view; its length 1.57-1.60 the greatest depth of midbody. Epitheca 0.78 the greatest depth of midbody; with rounded ventral hump, which is somewhat higher than \\idth of transverse furrow. Ventroposterior process inclined posteroventrally at 21°-23°. Sulcus runs on right face of body and extends to dorsal side of base of ventroposterior process. Except on posterior processes, theca is finely reticulated. Length, 61. . 3-63.6 M. Eastern tropical and subtropical Pacific. Description: — A medium-sized and rather broad species, the body of w hich is molariform in lateral outlme. The ratio between the length of the body and the greatest depth of the midbody is 1 .57-1 .60 : 1 . The epitheca is about 0.78 as deep as the midbody; dorsally it is low, but ventrally it forms a large rounded hump, which includes the A-entral half of the epitheca arid is about twice as high as the width of the transverse furrow or somewhat less. The transverse furrow, which is concave, forms a descending left spiral, and its distal end is displaced posteri- orly about half its width. The posterior cingular list is about 0.24-0.26 the length of the body from the apex. The hypotheca is deepest just behind the girdle. The dorsal margin is rather strikingly sigmoid, being gently convex, almost straight or even slightly concave anteriorly (Figure 28: 1), and moderately concave pos- teriorly. The ventral margin is also sigmoid, being gently to moderately convex anteriorly and slightly concave posteriorly. The posterior processes are almost straight, slender, slightly tapering, and subacute distally; the ventral one is in- clined posteroventrally at about 21°-23°; the dorsal either is subparallel to the longitudinal axis of the body or it has a slight posteroventral inclination. The curvature between the posterior processes is comparatively deep and broad, and the shortest distance between this curvature and the posterior cingular list is about 0.26-0.29 the greatest depth of the midbody and 0.13-0.16 the length of the body. The distance between the tips of the posterior processes is 0.80-0.86 the greatest depth of the midbody. In dorsoventral \iew the ventroposterior process is inclined to the right; the two posterior processes diverge from each other at an angle of about 20°. The greatest transdiameter of the body is located just behind the transverse furrow and equals about 0.38 the greatest dorsoventral diameter of the midbody. 204 THE DINOPHYSOIDAE. The cingular lists are subhorizontal, subequal, and about as wide as or some- what wider than the transverse furrow; each of them has on each valve about fifteen to seventeen straight, simple, and almost equidistant ribs. The sulcus runs on the right face of the body toward the dorsal side of the base of the ventro- posterior process; it extends to the posterior margin of the midbody, and is nearly as wide as the transverse furrow. The flagellar pore is located just be- hind the girdle on the right (dorsal) side of the sulcus. The right (dorsal) margin of the sulcus has a low ridge or list, which continues as a short spine beyond the postmargin of the midbody. The left sulcal list extends to the posterior end of the sulcus and is throughout its entire length about as wide as the posterior cingular list ; posteriorly it is rectangular (Plate 6, fig. 6) ; its structure is unknown. Except on the posterior processes, the thecal wall is finely reticulate. The meshes are subequal and irregular, and on each valve about thirty-five to forty- five meshes border the girdle anteriorly. The suture between the valves seems to follow the left sulcal list to a point near the postmargin of the midbody, from whence it turns abruptly toward the ventral margin of the ventroposterior process. The dimensions of two specimens were measured. Figure 2S. — 1, Dinofurcula ulliina (Kofoid), right lateral view. X 430. Station 4661 (300-0 fathoms). 2, 3, 4, Dinofurcula rentralis, gen. ct spec. nov. 2, 3, in right lateral view; 4, left lateral view of type specimen. X 430. 2, from Station 4659 (300-0 fathoms); 3, from Station 4655 (400-0 fathoms); 4, from Station 4681 (800-0 fathoms). Dimensions: — Length of body, 61.3-63.6 m (type, 61.3 m). Greatest depth of midbody, 39.0-39.7 n (type, 39.0 m). Distance between tips of posterior processes, 32.3-34.0 M (type, 34.0 n). Variations: — The specimens examined are strikingly simUar. Comparisons: — Dinofurcula ultima is easily distinguished from D. vcntralis by having a large, rounded hump on the ventral side of the epitheca and by the lateral position of the sulcus. Occurrence: — Dinofurcula ultima is recorded at three of the 127 stations. There are 0, 2, 0, 1, 0, and 0 stations on the six lines of the Expedition. Of these three stations, two (4661, 4671) are in the Peruvian Current, and one (4711) is in the South Equatorial Drift. All the records refer to hauls from 300-0 fathoms. SYSTEMATIC ACCOUNT. 205 The temperature range of these three stations at the surface was 66°-75°; the average was 70.0°. The frequency is in all the cases less than 1 %. DiNOFURCULA VENTRALIS, sp. nOV. Figure 28:2, 3, 4 Diagnosis: — Body rather narrowly molariform in lateral view; its length 1.60-1.83 the greatest depth of midbody. Epitheca 0.62-0.74 the greatest depth of midbody, lower than transver.se furrow, not, or but slightly higher ventrally than dorsally. Ventroposterior process inclined postero vent rally at 30°-36°. Sulcus runs along ventral margin of midbody to ventral side of base of ventro- posterior process. Length, 56.5-60.7 ix. Eastern tropical and subtropical Pacific. Description: — A medium-sized and rather narrow species, the body of which is molariform in lateral outline. The ratio between the length of the body and the greatest depth of the midbody is 1.74 (1.60-1.83) :1. The epitheca is relatively narrow, 0.68 (0.62-0.74) as deep as the midbody, and low, 0.45-0.80 the width of the transverse furrow ; sometunes (Figure 28 : 4) it is gently convex and highest in the center, sometimes (Figure 28 : 3) flattened and highest ventrally, and some- tunes (Figure 28 : 2) gently concave in the middle and of equal height dorsally and ventrally. The transverse furrow is concave, and, accorduig to a\'ailable data, its distal end is not displaced posteriorly. The posterior cingular list is 0.18 (0.14-0.22) the length of the body from the apex. The greatest depth of the hypotheca is variable in position; sometimes it is located near the middle of the midbody, sometimes near the bases of the posterior processes, and sometunes near the distal ends of these processes. The dorsal margin either (Figure 28:2, 3) is sigmoid, being gently convex or almost straight anteriorly and gently to moder- ately concave posteriorly ; or it is (Figure 28 : 4) gently undulating. In some speci- mens (Figure 28 : 2) the ventral margin is gently and almost evenly conca\-e ; in others (Figure 28 : 3, 4) it is more or less strikingly con\-ex anteriorlj'. The posterior processes sometimes are almost straight, and sometimes (Figure 28 : 4) slightly curved toward each other. The ventral of them, which usually is very slender, tapers very slightly and is subacute to more or less rounded distally. The dorsal one, usually somewhat wider than the ventral, also is somewhat taper- ing, and subacute to more or less rounded distall}-; in one specimen (Figure 28:3) the average width of this process about equaled the width of the transverse furrow, but usually it is somewhat less. The ventroposterior process is inclined 206 THE DINOPHYSOIDAE. posteroventrall}' at about 30°-36°; the dorsoposterior either is subparallel to the longitudinal axis of the body, or it has a slight postero ventral inclination. In one specimen, the type, the tip of each of these processes had a very small spinule. The curvature between the posterior processes is fairly deep and wide, and the shortest distance between this curvature and the posterior cingular list is 0.53 (0.47-0.61) the greatest depth of the midbody and 0.30 (0.28-0.33) the length of the body. The distance between the tips of the posterior processes is 0.88 (0.80-0.93) the greatest depth of the midbody. The cingular lists are subequal, somewhat narrower than the transverse furrow, and subhorizontal or somewhat inclined anteriorly; their structure is unknown. The sulcus runs on the right face of the body, near the ventral margin, to the ventral side of the base of the ventroposterior process. The extension and width of the right sulcal list are unknown. The left sulcal list extends to the pos- terior end of the sulcus; anteriorly it is about as wide as or even somewhat wider than the transverse furrow, and it decreases gradually in width posteriorly; its structure is not known with entire certamty, but it appears to have one to three ribs in its anterior half. The structure of the theca is unknown. The dimensions of three specimens were measured. Dimensions: — Length of body, 56.6-60.7 ^ (average, 58.4 n; type, 00.7 n). Greatest depth of midbody, 31.1-36.0 m (average, 33.6 n; type, 33.7 ai). Distance between tips of posterior processes, 28.8-31.1 n (average, 29.6 /x; type, 31.1 fi). Variations: — '^Maile Dinofurcula ultima appears to be fairly constant in size and structure, this species is comparatively variable. The foUowmg char- acters are the most variable : the depth and the shape of the epitheca, the depth of the midbody, the shape of the ventral margin of the hypotheca, and the width of dorsoposterior process. Comparisons: — This species is established on outline drawings of three specimens in lateral view. The shape of the body in dorsoventral view, the structure of the cingular lists and of the theca are unknown, and knowledge of the sulcal lists is incomplete. The species is readily distinguished from Dino- furcula ultima by the lack of the large ventral hump on the ei^itheca and by the ventral position of the sulcus. Occurrence: — Dinofurcula ventralis is recorded at three of the 127 stations. There are 0, 2, 1, 0, 0, and 0 stations on the six lines of the Expedition. Of these three stations, two (4655, 4659) are in the Peruvian Current, and one (4681), the type locality, is in the South Equatorial Drift. The depth records are 400, 300, and 800 fathoms to the surface, respectively. SYSTEMATIC ACCOUNT. 207 The temperature range of these three stations at the surface was 65°-69°; the average was 67.3°. The frequency is in every case less than 1 %. DiNOPHYSis Ehrenberg Dinophysis Ehrenberg, 1840a, p. 124. Griffith & Henfrey, 1856, p. 211. Diesing, 1866, p. 98. Fromentel, 1874, p. 196. Kent, 1881, p. 458. Bergh, 18Slb, p. 226. Butschli, 1885, p. 941, 1010. Delage & Herodard, 1896, p. 385. Schutt, 1896, p. 27. Paul.sen, 1908, p. 12. Mednier, 1919, p. 79. Jorgensen, 1923, p. 18. Lebour, 1925, p. 79. Dynophysis Van Oye, 1921, p. 209, 212, 214 {lapsus pennae). Diagnosis: — Body usually subovate, subcircular, subellipsoidal, subobovate, subreniform, subtrapeziform, or subtrapezoidal in lateral outline, or very elon- gated with one or two posterior peduncles; ratio between length and depth usu- ally somewhere between 1.1 : 1 and 1.6: 1 but may be as low as 0.97: 1 or as high as 5.3 : 1 ; on the whole somewhat more compressed bilaterally than in Phala- croma. Epitheca always low, as evidenced by the fact that posterior cingular list is never more than 0.28 the length of body from apex; ratio between its depth and depth of hypotheca 0.30-0.86: 1. Transverse furrow narrow, of subuniform width throughout, flat or but slightly convex or concave; its width subequal to or but slightly greater or smaller than height of epitheca. Cmgular lists may be subhorizontal but usually with more or less pronounced anterior inclmation; anterior 1.0-4.0 the width of transverse furrow and sometimes ribbed; posterior usually somewhat narrower than anterior and without ribs in most species. Right sulcal list small, usually subtriangular, and ends at or slightly in front of or behind Rj of left sulcal list. Left sulcal list usually subtrapeziform, somewhat wider posteriorly than anteriorly, with pronounced posteroventral angle, and with three ribs, one (Ri) in front and one (Rs) behind fission rib (R2) ; R2 some- what behind girdle; length of this Ust ranges from 0.41 to 0.84 the length of hypotheca; R2 usually 0.20-0.35, seldom as little as 0.11 or as much as 1.25 the depth of body; for R3 the corresponding values are 0.30-0.60 and 0.00, 1.33; margin seldom rounded posteroventrally. Sometimes with accessory lists and sails. Type. — Dinophysis acuta Ehrenberg. Organology: — The body {theca) in Dinophysis is diverse in shape; usuallj^ however, its shape is fairly simple, and only in one group of species, viz., in the Caudata group, does it show pronounced elongation and protuberances. When seen laterally, it generally is somewhat asymmetrical, but .symmetrical species 208 THE DINOPHYSOIDAE. also have been recorded {e.g., D. sphaerica, Figure 31 : 3). In dorsoventral view it appears always to be symmetrical. Sometimes (Z). triacantha, Figure 42 : 2) the loiujiludinal axis {La., Figure 29) is perpendicular to the girdle, but in most species it is inclined either posterodorsally or posteroventrallj'. While in Phala- croma only one species has been recorded as yet in which this axis is inclined posterodorsally, this inclmation is found rather frequently in Dinophysis. In the species with the most pronounced posterodorsal deflection of this axis, viz., D. collaris (Plate 5, fig. 2), an inclination of 10°-15° was found. The maximum posteroventral inclination as yet established is 30° {D. caudata, Figure 44). The inclination is very variable even within the species. luU. cing. I. ant. 711. r. post, cing incl. I. a. I. f.p. r.r. r. sulc. I. 1. S7llc. I. I.f. met. post, incl. post, par. I. m. I. m. r. s. post. ?n. r. La. ind. post. vent, m post. s. FiGDRE 29. — Dinophysis collaris Kofoid and Michenor, rij^ht lateral view. Station 4071 (300-0 fathoms). Terminology and methods of making measurements of Phalaeroma and Dinophysis. Terminology: — The niidhne is the perpendicular to girdle which is equidistant from dorsal and ventral margins of body at place of greatest depth of body. Wlien longitudinal axis of body is not in- clined, it coincides with miflline. In specimen represented by tliis Figure the longitudinal axis of body is inclined posterodorsally and tlie posteroventral margin of body, dorsopusteriorhi, according to our terminology. The anterior main rib (Ri) is the rib at junction of left sulcul list and posterior cingular list. The fission rib (R2) is the rib of left sulcal list at place where this list is divi 1912), Pavillard (1912), Petersen (1897), Pouchet and Guerne (1885), Rauschenplat (1900), Redeke and van Breeman (1903), Reinhard (1910), Schiller (1911a, b, 1912), Schroder (1900a, b, 1906b, 1909), Schutt (1890, 1893), Schussnig (1914), Steuer (1910), Storrow (1913), Stiiwe (1909), Tempere (1898), Vanhoffen (1897a,' b), Walther (1893), Whitelegge (1891), Wright (1907), and Zernow (1904). Of these papers only the following contain figures by means of which the accuracy of the determinations may be judged: — Broch (1910a), Carisso (1911), Dixon and Joly (1898), Gran (1902), Levander (1894c), Okamura (1907, 1912), Paulsen (1911b), and VanhofTen (1897a). Forti (1922) reproduced previously published figures. Due to the great difficulties in the determination of several species and to the broad concept of species held by many investigators, much of the data on the distribution should be accepted tentatively. References to Dinophysis or minor contributions to our knowledge of this genus are also found in Balbiani (1884a, c), Beauregard and Galippe (1880), Bergh (1882, 1884), Bles (1892), Brandt (1883), Butschli (1885a), Buhigas (1918), Calkins (1902), Chun (1897), Daday (1884), Dakin (1908), Doflein (1909, 1911, 1916), Dogiel (1906), Entz (1909), Forti and Issel (1923), Goldschmidt (1907), Gran (1905, 1908, 1912a, b), Hensen (1891), Hjort (1911), Klebs (1884, 1885), Lawessan (1882), Lankester (1885), Lebour (1922), Leuckart (1861), Maggi (1874, 1880a, b), Nathansohn (1910b), Oltmanns (1922), Pavillard (1915a), Pouchet (1882, 1885a), Pritchard (1845, 1861), Stein (1878), Torrey (1902), Wallengren and Hennig (1911), West (1916), and Willey and Hickson (1909). Pavillard (1923a) gives a critical review of Jorgensen's (1923) important paper on this group. It may also be mentioned that Gourret (1883, p. 74) rejected Amphidinium as a distinct genus and united it with Dinophysis. The following typographical errors in the spelling of the name Dinophysis have been found by us: — Dixophysis Kofoid and Michener (1911) and Dyno- physis van Oye (1921). Adaptive and Systematic Value of the Characters. Principles used IN THE Description of the Species The species of Dinophysis are limited largely to tropical, subtropical, and warm-temperate waters of low buoyancy. Nevertheless, most of them are charac- terized by the fact that the body (theca) is more or less simple in shape, such as lenticular (Figure 35:1, 6), bilaterally flattened, ellipsoidal (Figure 32:13), ' Contains figures by which the identifications may be verified. 234 THE DINOPHYSOIDAE. ovoidal (Figure 31:7), obovoidal (Figure 32 : 6), etc. Only in one group of species, viz., in the Caudata group, e.g., in D. caudata and D. miles, does the body show adaptations to flotation in the form of pronounced elongation and protuberances. In D. miles, a species occurring in tropical seas, the capacity of flotation is further increased during the period of binary fission by the fact that the schizonts form a wheel-like colony, a phenomenon that facilitates the remaining of the asymmetri- cal specimens in the horizontal position in which they offer a maximum surface of resistance (Weber-van Bosse, 1901, pi. 17, fig. 3). In D. caudata two specimens have been found to adhere to each other for some time following binary fission. The species that have a body more or less simple in shape maintain themselves in the upper strata of the sea by the same means as those foinid in Phalacroma (see p. 64). It may be mentioned, however, that in Dinophysis the anterior cingular list is, on the average, somewhat wider, that in some species (Plate 6, fig. 7) the left sulcal list and the posterior sail (or sails) are very large, and that the para- sagittal lists along the dorsal margin of the body very seldom are present. Just as in Phalacroma, the species of Dinophysis that are endogenetic in cold waters (D. acuminata, D. acuta, D. arctica, D. cuneiformis, D. granulata, D. norvegica, D. semen, D. sphaerica, D. ventricosa, and D. vertex) are characterized by the relatively small size of their lists. With regard to the spiral swimming the same adaptations are found in Dinophysis and Phalacroma. All the external characters have been considered and utilized in the estab- lishment of the species; nevertheless, as in Phalacroma, a number of the specific separations should be regarded as tentative. This is true not only among the species of relatively simple organization, such as the members of the Acuta group, but also among the highest-developed forms, e.g., the representati\es of the Caudata and Hastata groups. It is due to the profound fluctuations that some- times occur in several of the most important characters, e.g., in the size and shape of the body and in the size, shape, and structure of the left sulcal list and of the posterior sail (or sails). As examples of striking variations in the size of the body, Dinophysis hastata andZ). uracantha (Figure 32, 35) may be mentioned; profound variations in the shape of the body are well illustrated by D. hastata, D. swezyi, and D. caudata (Figure 32, 39, 44) ; variations in the shape of the left sulcal list by D. swezyi (Figure 39) ; variations in the shape and structure of the posterior sail by D. hastata (Figure 32). Jorgensen (1923, fig. 30-37) gives a good illustra- tion of the wide amplitude of variation that might characterize a species of this genus. Dinophysis hastata, although highly differentiated, includes specimens that have the habitus characteristics of Dinophysis and specimens that look like SYSTEMATIC ACCOUNT. 235 rather typical representatives of Phalacroma (Figure 33). The characters most variable within the species are those which have also been most profoundly modi- fied within the genus as a whole, and it is largely upon them that the subdivisions of the genus have been founded. In the descriptions of the species the same (five) principles have been applied as in Phalacroma. Of the eight principles followed in the diagnoses of the species of Phalacroma, 1, 2, 3, 6, 7, and 8 are followed also in Dinophysis. If not otherwise mentioned, the epitheca is not visible above the anterior cingular lists, and the cingular lists are more or less funnel-hke (inclined anteriorly) and subequal or the anterior is but slightly wider than the posterior; furthermore, the anterior cingular list is ribbed and the posterior is without ribs. The methods of measuring lengths, pro- portions, and angles are shown in Figure 29. Subdivisions. Relationships among the Species Dinophysis, like Phalacroma, exhibits a marked diversity of structural types, proceeding from small, simple, and undifferentiated forms such as Dinophysis exigua (Figure 30), closely related to the members of the primitive Rotundatum group of Phalacroma, to highly specialized ones such as Dinophysis 7niles (Cleve, 1900b, fig. la) and D. jorgenseni (Figure 42: 3-8). On the other hand, while the species of Phalacroma without very great difficulty can be arranged into several groups, all of which probably represent different evolutionary lines, Dinophysis does not readily yield to a similar analysis. Accordmgly, the forty-one presum- ably valid species of the latter cannot with certainty be divided into more than three groups. Two of these include quite heterogeneous elements, and a further subdivision is desirable. However, attempts in this direction have not proved suc- cessful. The reason for this failure is largely the independent materialization in the various evolutionary lines of "tendencies" that were inherent in, but not morphologically expressed by, the siinpler common ancestors. In other words, the obstacles were the same as in the genus Histioneis. The three groups recognized in Dmophysis are as follows: — 1. Acuta group: — D. exigua, sp. nov., D. punctata Jorgensen, D. semen Meunier, D. granulata Cleve, D. recurva, nom. sp. nov., D. arctica Mereschowsky, D. acuminata Claparede and Lachmann, D. baltica (Paulsen) nobis, D. sphaerica Stein, D. sifnilis, sp. nov., D. ouuin Schiitt, D. okamurai, sp. nov., D. reniformis (Pavillard) nobis, D. sacculus Stein, D. ellipsoides Kofoid, D. dens Pavillard, D. truncata Cleve, D.fortii Pavillard, D. vertex Meunier, D. ventricosa Claparede and 236 THE DINOPHYSOIDAE. Lachmann, D. nonrgica Claparede and Lachmann, D. schroderi Pavillard, D. acuta Ehrenberg, and D. cuneiformis Meunier (Figure 30; Jorgensen, 1923, fig. 28; Meunier, 1910, pi. 3, fig. 47, 48; Cleve, 1899c, pi. 4, fig. 7; Pavillard, 1916, pi. 3, fig. 6; Mereschowsky, 1879, pi. 11, fig. 19; Claparede and Lachmann, 1858, pi. 20, fig. 17; Paulsen, 1908, fig. 17; Figure 31: 3, 4 and 1, 2; Schutt, 1895, pi. 1, fig. 6, Figure 31:5; Pavillard, 1905, pi. 3, fig. 10; Stein, 1883, pi. 20, fig. 10; Kofoid, 1907b, pi. 33, fig. 56; Pavillard, 1916, pi. 3, fig. 1; Cleve, 1900f, fig. 7; Pavillard, 1916, pi. 3, fig. 4; Meunier, 1910, pi. 14, fig. 29, 30; Claparede and Lachmann, 1858, pi. 20, fig. 20; Paulsen, 1907, fig. la; Figure 31: 6; Paulsen, 1907, fig. lb; Meunier, 1910, pi. 14, fig. 31, 32). 2. Hastata group: — D. pusilla Jorgensen, D. hastata Stein, D. alata Jorgensen, D. rudgei Murray and \Miitting, D. uracantha Stein, D. urceolus, sp. nov., D. monacantha, sp. nov., D. trajwzium, sp. nov., D. sivezyi, sp. nov., D. col- laris Kofoid and Michener, D. schtitti Murray and Wliitting, D. nias Karsten, D. jorgcnseni, sp. nov., and D. triacantha Kofoid (Jorgensen, 1923, fig. 44; Figure 32, 33; Jorgensen, 1923, fig. 45; Murray and Wliitting, 1899, pi. 31, fig. 9; Figure 35, 37: 1 and 2, 3; 38: 2, 3; 39; 3§: 1, 40, 42: 1, and 3-8 and 2). 3. Caudata group: — D. caudata Saville-Kent, D. fripos Gourret, and D. miles Cleve (Figure 44; Jorgensen, 1923, fig. 38, 39; Cleve, 1900b, fig. la). 1 . Acuta group (for references, see above) : — Body of somewhat varying shapes, usually ellipsoidal, ovate to subovate in lateral outline, never with pos- terior peduncle. Accessory sails and lists not present. The twenty-four species referred by us to this group have been arranged above, as far as possible, according to similarities. The series begins with small, simple, ellipsoidal, and usually symmetrical forms {Dinophysis exigua, D. pu7ic- tata, D. semen, D. granulata, and D. recurva; Figure 30; Jorgensen, 1923, fig. 28; Meunier, 1910, pi. 3, fig. 47, 48; Cleve, 1899c, pi. 4, fig. 7; Pavillard, 1916, pi. 3, fig. 6) , and ends with moderately large forms, the bodies of which are somewhat asymmetrical and elongated, and narrowly rounded to subacute posteriorly (D. vertex, D. ventricosa, D. norvegica, D. schroderi, D. acuta, and D. cuneiformis; Meunier, 1910, pi. 14, fig. 29, 30; Claparede and Lachmann, 1858, pi. 20, fig. 20; Paulsen, 1907, fig. la; Figure 31: 6; Paulsen, 1907, fig. lb; Meunier, 1910, pi. 14, fig. 31, 32). Between these extremes we have a practically continuous series of intermediate forms (D. arctica, D. haltica, D. ovum, and D. fortii; Mereschowsky, 1879, pi. 11, fig. 19; Paulsen, 1908, fig. 17; Schiitt, 1895, pi. 1, fig. 6; Pavillard, 1916, pi. 3, fig. 4). Indeed, the transitions between the members of the extreme groups as well as between the members of the intermediate group are so gradual SYSTEMATIC ACCOUxNT. 237 that delimitations of species are extremely difficult. D. acuminata (Claparede and Lachmann, 1858, pi. 20, fig. 17) agrees fairly closely with D. arctica, but is characterized by one or more small protuberances on the posterior end of the body. D. si'tnilis (Figure 31: 1, 2), which represents a small side branch charac- terized by the reduction of the posterior mam rib of the left sulcal list, is struc- turally closely related to D. sphaerica (Figure 31 : 3, 4), which in its turn is close to D. ovum (Schiitt, 1895, pi. 1, fig. 6). D. okamurai (Figure 31: 5), which pre- sumably also is closely related to D. ovum, is characterized by the concavitj^ of the anterior portion of the ventral margin of the hypotheca. D. reniformis and D. sacculus (Pavillard, 1905, pi. 3, fig. 10; Stein, 1883, pi. 20, fig. 10) as well as D. dens and D. truncata (Pavillard, 1916, pi. 3, fig. 1 ; Cleve, 1900f, fig. 7) form short side branches. The first two species are characterized by the slight, but dis- tinct bendmg of the body, the last two by the subtrapeziform lateral outline of the body. These four species probably originated from an ancestral form re- sembhng D. ovum or D. fortii (Pavillard, 1916, pi. 3, fig. 3, 4). D. ellipsoides (Kofoid, 1907b, pi. 33, fig. 56) is probably closely related to D. reniformis. This group corresponds to Pa vil lard's (1916) sections Acuta, Sacculus, and Sphaerica, and to Jorgensen's (1923) sections Acutae, Ovum, and Sphaericae. 2. Hastata group (for references, see above) : — Body of varied shapes, rounded, ovate, subovate, or rounded subtrapeziform in lateral outline, without posterior peduncle. With one or two posterior sails and sometimes with para- sagittal fists along dorsal side of hypotheca. The fourteen species of this group are arranged above, as far as possible, according to structural similarities. The first three of them, viz., Dinophysis pusilla, D. hastata, and D. alata (Jorgensen, 1923, fig. 44; Figure 32, 33; Jorgensen, 1923, fig. 45) are structurally, and presumably also genetically, very closely re- lated and appear to form a natm-al unit. Their most characteristic common features are as follows : — (1) only one posterior sail is present, and this is located ventrally to the midline and directed somewhat ventrally; (2) the ventral margin of the left sulcal list is nearly straight or more or less convex. D. pusilla, which is the most primiti\'e member of this group, is a very small form, subcircular in lateral outline, and its posterior sail is small and lacks structural difi"erentiation. In D. hastata and D. alata the body is somewhat elongated, the longitudmal axis deflected posteroventrally, and the posterior sail is fairly large and has a central rib or a central reticulation. Of these two species D. alata is the more highly differentiated, in so far as it has a wide parasagittal list along the dorsal side of the hypotheca. The remaining members of this group are structurally \-ery close 238 THE DIXOPHYSOIDAE. to D. pusilla, D. hastata, and D. alata, but at the same time they presumably belong to different evolutionary lines. Of these species D. uracantha and D. urceolus (Figure 35, 37: 1) undoubtedly are very closely related. Their most im- portant common characteristics are as follows : — (/ ) the longitudinal axis is somewhat deflected posterodorsally; (2) only one posterior sail is developed, and this is located somewhat dorsally to the midline and supported by marginal ribs ; (S) the ventral margin of the left sulcal Ust is straight or nearly so. D. monacantha, D. trapezium, and D. collaris (Figure 37: 2, 3; 38: 2, 3; 38: 1) are variants of the same evolutionary type as the last two species. D. monacantha differs from these mainly in two respects, viz., its posterior sail is located on the midline and para- sagittal lists are developed along the dorsal side of the hypotheca. D. trapezium is characterized especially by the fact that its left sulcal list is furnished with an accessory triangular lobe behind the posterior main rib; and D. collaris is dis- tinguished from D. trapezium chiefly by having the accessory lobe and the posterior sail well balanced, of about the same size, shape, and structure. The last five species form in most respects a well-marked orthogenetic series. D. swezyi (Figure 39), which approaches D. trapezium in having the left sulcal list furnished with an accessory posterior lobe, differs strikingly from this as well as from the other previously discussed species in the more or less pronounced con- cavity of the ventral margin of its left sulcal list. In the last respect D. swezyi resembles D. schiitti, D. jorgenseni, and D. triacantha (Figiu'e 40; 42: 3-8; 42: 2), of which the last two are characterized especially by having two posterior sails, instead of one. On the other hand, in D. nias (Figure 42: 1), which undoubtedly is very close to D. jorgenseni as indicated by the fact that it has two posterior sails, the ventral margin of the left sulcal list is straight or nearly so. Finally, D. rudgei (Murray and Whitting, 1899, pi. 31, fig. 9) represents a small side branch, char- acterized by the primitive, subcircular lateral outline of the body and by the very heavy and large-meshed thecal reticulation. This branch probably split off near the base of the evolutionary scale of this group. This group corresponds to the section Hastata of Pavillard (1916) and of Jorgensen (1923). Although it comprises at least three evolutionary lines, viz., the Pusilla, the Uracantha, and the Rudgei lines, its subdivision is hardly advis- able at the present time. 3. Caudata group (Figure 44; Jorgensen, 1923, fig. 38, 39; Cleve, 1900b, fig. la) : — Body strongly elongated, with posterior peduncle. Without accessory lists or sails. Of the three species of this group, Dinophysis caudata (Figure 44) is the most SYSTEMATIC ACCOUNT. 239 primitive, and D. miles (Cleve, 1900b, fig. la) is the most advanced. D. tripos (Jorgensen, 1923, fig. 38, 39) is structurally and presumably also genetically in- termediate between these extreme types. This group corresponds to the section Homunculus of Pavillard (1916) and Jorgensen (1923). . Each of these three groups presumably has evolved independently from small, subrotmid, and structurally undifferentiated forms, such as Dinophysis exigua (Figure 30: 1) in three diverging lines. 1. Acuta Group: — Of the twenty-four species referred to this group, only the following were found in the material of the E.xpedition: Dinophysis exigua, D. sphaerica, D. similis, D. okamurai, D. fortii, D. norvegica, and D. schroden. Their arrangement is indicative of relationships. Dinophysis exigua, sp. nov. Figure 30 Diagnosis: — Body subellipsoidal in lateral outline, deepest near the middle; length : depth, 1 .08-1 .14 : 1 ; symmetrical, or its longitudinal axis deflected postero- dorsally at 1°-10°. Posterior cingular list 0.18-0.28 the length of body from apex; epitheca barely if at all visible above anterior cingular list. Cingular list inclined anteriorly at 30°-50° ; the anterior sometimes somewhat wider than the posterior ; structure unknown. Left sulcal list 0.50-0.60 the length of body; distance be- tween Ri and R3 0.40-0.45 the length of body; R2 is 0.15-0.17, R3 is 0.37-0.50 the greatest depth of body; margin forms angle of 40°-55° at R3; R3 mclined pos- teriorly at 40°-50°. Structure of theca unknown, presumably not differentiated. Length, 28.9-30.4 m- Eastern tropical Pacific. Description: — This is a small species, the body of which is subellipsoidal in lateral outline, deepest near the middle. The ratio between the length and the depth of the body is 1.12 (0.08-1.14): 1. Either the body is symmetrical, or its longitudinal axis is deflected posterodorsally at 1°-10°. The epitheca is about 0.63 (0.58-0.67) as deep as the hypotheca and highest in the center; it is evenly, and moderately to rather strongly convex, and barely, if at all, visible above the anterior cingular list. The transverse furrow is flat or slightly convex, and its width is 1.0-1.5 the greatest height of the epitheca. The posterior cingular list is about 0.22 (0.18-0.28) the length of the body from the apex. The hypotheca is sometimes symmetrical (Figure 30: 1), sometimes more or less deflected posteroventrally (Figure 30: 2, 3) (see p. 240). Its dorsal, pos- 240 THE DINOPHYSOIDAE. terior, and ventral margins are moderately convex and confluent. The posterior portion of the body is not, or but slightly, deeper than the anterior. The cingular lists are subequal, or the anterior is somewhat wider than the posterior; the anterior is about 1.0-1.5 times wider than the transverse furrow and 0.21 (0.19-0.25) the greatest depth of the body; the anterior has an anterior in- clination of 30°-50°, the posterior sometimes a little less; their structure is un- known. The right sulcal list is unknown. The left sulcal list is 0.50-0.60 as long as the body, and the distance between the anterior and posterior main ribs is 0.40-0.45 the length of the body. The anterior main rib is 0.20-0.21, the fission rib 0.15-0.17, and the posterior main rib 0.42 (0.37-0.50) the greatest depth of the body; behind the last rib this list decreases suddenly in width. Between the anterior and posterior main ribs the free margin of this list is almost straight ; at the posterior main rib it forms an angle of 40°-55°; behind the last-mentioned rib it is gently concave or straight. The mam ribs of this list are of moderate strength, and not club-shaped or otherwise modified ; the two anterior are straight or nearly so, the posterior is gently concave posteriorly. The distance between the anterior main rib and the fission rib is 0.29-0.40 the distance between the an- terior and posterior main ribs. The posterior main rib has a posterior inclination of 40°-50°. Except for the main ribs, this list appears to lack structure. There are no accessory lists or sails. The structure of the thecal wall is unknown. Megacytic specimens were not recorded. The dimensions of three specimens were measured. / 2 Figure 30. — Dinophijs'is exiyua, sp. nov., 1, 2, in left lateral view; 3, in right lat- eral view. 1, from type specimen. X 430. 1, 2, from Station 4.545, surface; 3, from Sta- tion 4709, stomach of Salpa. Dimensions: — Length of body, 28.9-30.4 ix (average, 29.6 m; type, 29.6 n). Greatest depth of body, 25.6-28.2 m (average, 26.6 m; type, 25.9 y). Variations: — This species varies mainly in the inclination of the longitudinal axis of the body, and in the inclination and relative width of the anterior cingular list. Comparisons: — This species is established on drawings giving the lateral outlines of three specimens from the material of the Expedition. The description given above is thus incomplete in several respects. The following important char- SYSTEMATIC ACCOUNT. 241 acters are unknown : — the shape of the body in dorsoventral view, the structure of the thecal wall, the structure of the cingular lists, and the shape and structure of the right sulcal list. Dinophysis exigua is on the border line between Dinophysis and Phalacroma. It has some features that affiliate it with Dinophysis, viz., the narrowness of the epitheca, the strong anterior inclination of the cingular hsts, and the postero- dorsal inclination of the longitudinal axis of the hypotheca. The last peculiarity has not been found in any of the known species of Phalacroma, but it is char- acteristic of some species of Dinophysis, e.g., D. uracantha, D. schutti, and D. swezyi. Compare also the shape of the left sulcal list in D. exigua and D. ura- cantha. On the other hand, D. exigua is not far removed from Phalacro7na parvu- lum and its small, simple, and rounded relatives. The structurally closest-known relative of Dinophysis exigua probably is the species estabhshed by Ehrenberg (1840a, p. 151, pi. 4, fig. 15) mider the name of D. michaelis and later figured by Lohmann (1908a, fig. 7: 43, p. 201) under the name of D. rotundata; Lohmann's as well as Ehrenberg's specimens were from the Bay of Kiel. These two species are so similar that their specific separation should be regarded as tentative. They have been separated by us, smce D. michaelis appears to be somewhat larger (about 52 yu, according to Lohmann's figure, as compared with 28.9-30.4 m) and narrower. Another structurally close relative is D. lenticula Pavillard {non [Bergh] Daday, 1888), which is most easily distin- guished by the shape of its left sulcal list (see also Phalaa-oma pawulum, the section on comparisons). Occurrence: — Dinophysis exigua is recorded at two of the 127 stations. Of these two stations one (4709) is on the fourth line of the Expedition and in the South Equatorial Drift; the other (4545), the type locality, is on the sixth line and in the Mexican Current. At Station 4709 one specimen was found in the stomach of a Salpa taken in a surface haul; the other record (two specimens) refers to a surface haul. The surface temperatures of these two stations were 72° and 79°, respectively. The frequency is less than 1 % in both cases. Dinophysis sphaerica Stein Figure 31 : 3, 4 Dinophijxis sphaerica Stein, 1883, parlim, pi. 20, fig. 3, 4. Dadav, 1888, p. 99, 10-1. non ScHtixT, 1895, p. 13, 64, 119, pi. 1, fig. 7: 1, 2. Aurivillius, 1898a, p. 256, 348, 368, 418; 1899, p. 33, 64. Lemmer- MANN, 1899a, partim, p. 373; 1901a, p. 375. Murray & Whitting, 1899, p. 331, tab. 1, 2, 4-6, 8, 9. Schroder, 1900a, p. 19. Gran, 1900, p. 43. Ostenfeld & Schmidt, 1901, p. 170. Schmidt, 1901, p. 138. Entz, 1902b, p. 94; 1905, p. HI. Cleve, 1903b, p. 344. Karsten, 1905, p. 59, 62, 03; 1906, 242 THE DINOPHYSOIDAE. p. 218; 1907, p. 233. Zacharias, 1906, p. 530. Paul.sex, p. 13, 16, fig. 15. Graf, 1909, p. 139. Hensen-Ostenfeld, 1913, p. 141. nun Pavjllabd, 1916, p. 47, 60. J6RciEN.sEN, 1923, p. 11, 23, 44, fig. 29. Lebour, 1925, p. 82, fig. 21c. fPhalacroma sphaericum Zacharias, 1906, p. 530. Diagnosis: — Body ovate to subellipsoidal in lateral outline, deepest in or somewhat behind the middle; length: depth, 1.17-1.43: 1; symmetrical or with but slight posteroventral inclination. Anterior cingular hst 1.4-1.7 times wider than transverse furrow. Left sulcal list 0.52-0.66 the length of body; distance between Ri and R3 0.44-0.58 the length of body; R2 is 0.18-0.33, R3 is 0.23-0.41 the greatest depth of body; margin forms angle of 80°-90° at R3; R3 incUned posteriorly at 5°-15°; besides the three main ribs this hst has numerous anasto- mosing cross-ribs. Theca finely and closely areolate. Length, 44.5-47.0 n. Probably widely distributed in tropical, subtropical, and warm-temperate seas, seldom m colder regions. Description: — This is a rather small species, the body of which is ovate to subellipsoidal in lateral outline, and deepest in or somewhat behiml the middle. The ratio between the length and the depth of the body is 1.17-1.43: 1 ; in Stein's (1883, pi. 20, fig. 3, 4) specimens this ratio is 1.20-1.43: 1 ; in our specimens it is 1.17-1.31:1. The longitudinal axis either is perpendicular to the transverse furrow, or it is inclmed posteroventrally at l°-5°. The epitheca is 0.43-0.55 as deep as the hypotheca, of moderate convexity, highest in or near the center, and not visible above the anterior cmgular hst. The transverse furrow is flat or somewhat convex, and its width is 1.0-2.0 the greatest height of the epitheca. The posterior cingular list is 0.12-0.19 the length of the body from the apex. The hypotheca is symmetrical or almost so. The dorsal and ventral margins are subuniformly, and gently to moderately or rather strongly coiiA'ex; or they are somewhat flattened anteriorly. The postmargin is evenly and rather strongly convex, and confluent \\ith the dorsal and ventral margins. The posterior portion of the body may be of about the same depth as the anterior (Stein, 1883, pi. 20, fig. 3), but usually it is more or less strikingly deeper (Figure 31:3). The anterior cingular list is 1.4-1.7 times wider than the transverse furrow and 0.18-0.29 the greatest depth of the body; it has a well-developed funnel shape, being inclined anteriorly at 45°-60°; and on each vah-e it is furnished with about ten to twenty simple, straight, and almost equidistant ribs. The posterior cingu- lar list has about the same inclination as the anterior and appears always to lack structural differentiation. In our specimens it is about as wide as or slightly nar- rower than the anterior, but in the specimens figured by Stein (1883, pi. 20, fig. SYSTEMATIC ACCOUNT. 243 3, 4) it is very narrow, only about 0.25 the width of the anterior. It should be noticed, however, that Stein (1883) made this list very narrow in almost all his figures of species of Dinophysis. The right sulcal hst extends to or slightly beyond the fission rib of the left sulcal list and is subtriangular, decreasing gradually in width posteriorly; anteriorly it is about as wide as the transverse furrow. The left sulcal list is 0.58 (0.52-O.66) the length of the body, and the distance between the anterior and posterior main ribs is 0.51 (0.44-0.58) the length of the body. The anterior main rib is 0.20-0.28, the fission rib 0.18-0.33, and the posterior main rib 0.23-0.41 the greatest depth of the body; m the speemiens examined by us the corresponding values are 0.20-0.22, 0.18, and 0.23-0.24, respectively; in the specimens figured by Stem (1883, pi. 20, fig. 3, 4) they are 0.25-0.28, 0.29-0.33, and 0.31-0.41, respectively. Behind the posterior mam rib this list decreases sud- FiGURE 31. — 1, 2, Dinophysix similis, sp. nov., riglit lateral view (2, possibly some- what tilted). X 430. 1, from Station 4701 (.300-0 fathoms); 2, from Station 4691 (300- 0 fathoms). 3, 4, Dinophysis sphaerica Stein, right lateral view. X 430. 3, from Station 4697 (300-0 fathoms); 4, from Station 4711 (300-0 fathoms). .5, Dinophysis okamtirai, sp. nov., right lateral view of type specimen. X 430. Station 4673 (300-0 fathoms). 6, Dinophysis schroderi Pavillard, left lateral view; structure of thecal wall indicated only. X 430. Station 4574, surface. 7, Dinophysis Joriii Pavillard, right lateral view. X 430. Station 4667 (300-0 fathoms). 8, Dinophysis norvegica Claparede and Lachmann, left lateral view. X 430. Station 4638 (300-0 fathoms). denly in width. Between the anterior and posterior main ribs the free margin of this list is almost straight, or gently to moderately convex; at the posterior main rib it forms an angle of 80°-90° ; behind this rib it is almost straight or gently con- cave. The mam ribs are of moderate strength, and not club-shaped or otherwise modified; the anterior two are straight or almost so; the posterior one may be straight but usually is gently concave posteriorly. The distance between the an- terior main rib and the fission rib is 0.30-0.38 the distance between the anterior and posterior main ribs. The posterior main rib has a posterior mclination of 5°- 244 THE DINOPHYSOIDAE. 15°. Besides the three main ribs, this list has numerous anastomosing cross-ribs (Stein, 1883, pi. 20, fig. 3, 4). There are no accessory lists or sails. The thecal wall is finely and closely areolate, and at least sometimes furnished with scattered pores; compare figure 3 and figure 4 of Stem's (1883) Plate 20. Megacytic specimens have not been recorded. The dunensions of two of our specimens and of the specimens figured by Stein (1883, pi. 20, fig. 3, 4) were measured. Dimensions: — Our specimens: length of body, 44.5-47.0 fi. Greatest depth of body, 35.9-38.1 n. The size of the specimens represented by Stem (1883, pi. 20, fig. 3, 4) is unknown. Judging by Stein's (1883) information about the magni- fications of his figures, given in an introductory remark to the explanations of the plates, the lengths of these specimens were somewhere between 52.2 )j. and 80.5 ju. Variations: — This species, as here conceived, is rather variable. The size of the body, the shape of the hypotheca, the width of the cingular lists, and the size of the left sulcal lists are the most variable characters. Comparisons: — Dinophysis sphaerica, as conceived by Stem (1883, pi. 20, fig. 3-9), is certahily not a natural systematic unit. Indeed, the five specimens figured imder this name by him probably belong to at least three different species. This fact was pomted out first by Ostenfeld and Schmidt (1901) and later by Schmidt (1901), Paulsen (1908), Pavillard (1916), and Jorgensen (1923), who, in their references to Stein's (1883) figiu'es of this species, mention only his Plate 20, figure 3 and 4 (see also D. punctata Jorgensen, 1923, p. 23). Although the two specimens represented by Stein's (1883) Plate 20, figure 3 and 4, are rather differ- ent, and the question as to whether or not they belong to the same species must be regarded as open, we have considered it most advisable to accept, preliminarily, the decision of Ostenfeld and Schmidt (1901). The specimen represented by Stein's (1883) Plate 20, figure 4, is regarded as the type of D. sphaerica s. str. Of the two specimens found by us in the material of the Expedition and figured in the present paper, one (Figure 31:3) shows, in the shape of its body, a rather striking resemblance to the type spccunen; the other (Figure 31:4) ap- proaches in the same respect the specimen represented by Stem's (1883) Plate 20, figure 3. Both of our specimens differ conspicuously from Stem's (1883) speci- mens in their wider posterior cingular list and in their narrower left sulcal list. We are not able to judge the significance of these differences. However, it seems far from improbable that the difference in the width of the posterior cingular list is duo to the incomplete observation of this structure by Stein (1883), since this list is \or\- narrow in almost all the figures of species of Dmophysis to be found in SYSTEMATIC ACCOUNT. 245 Stem's (1883) iconography. Reference should also be made to the rather striking similarity in the shape of the body, between one of our specimens (Figure 31:3) and the specimen represented by Stein's (1883) Plate 20, figure 9. The structurally closest-known relatives of Dinophysis sphaerica are D. ovum and D. similis. Indeed, the specific separation of these three forms should be regarded as tentative (see D. similis, the section on comparisons). D. sphaerica differs from D. ovum mainly in having the anterior cingular list ribbed and in the shape of the posterior portion of the left sulcal list. From D. similis it differs mainly in having the posterior main rib of the left sulcal list well deAeloped, about as long as or somewhat longer than the fission rib of this list, and in having the posteroventral portion of this list angular instead of rounded to rounded angular. Other structurally close relatives of D. sphaerica are D. lenticula Pavillard (mvi [Bergh] Daday, 1888) and D. punctata. D. lenticula differs from D. sphaerica mainly m lacking ribs in the anterior cingular list and in having the ventral margin of the left sulcal list strongly recurved to the right (Pavillard, 1916, pi. 3, fig. 6). D. punctata is characterized by its smaller size and by the faint structure of its thecal wall. Synonymy: — The name of Dinophysis sphaerica was first applied by Stein (1883, pi. 20, fig. 3-9) to a compound species. Later it was used in a restricted sense by Ostenfeld and Schmidt (1901), Schmidt (1901), Paulsen (1908), Pavil- lard (1916), and Jorgensen (1923) (see the previous section). Of the rather great nmnber of writers who have used this name, only the following, besides Stein (1883), give descriptions or figures by means of which their determmations of this species may be checked: — Schiitt (1895), Paulsen (1908), Okamura (1912), Pavillard (1916), Jorgensen (1923), and Lebour (1925). With regard to D. sphaerica of Schiitt (1895), Pavillard (1916), and Jorgensen (1923), see D. similis, the sections on comparisons and synonymy. The figures of this species given by Paulsen (1908, fig. 15) and Lebour (1925, fig. 21b) are reproductions of Stein's (1883) Plate 20, figure 3. The figure given by Okamura (1912, pi. 5, fig. 85) under the name of D. sphaerica is too mcomplete for certainty of specific determina- tion. However, the small size of the figured specimen, 35.8 m, seems to indicate that D. sphaerica Okamura (1912) is not identical with D. sphaerica Stein. Zacharias (1906, p. 530) records a species from the Gulf of Naples under the name of Phalacroma sphaericum Stein. This is probal^ly a lapsus pennac for Dinophysis sphaerica, a name also used by Zacharias (1906), but no data bearhig on this question are available. Entz (1902b, 1905) writes sphaeriaca instead of sphaerica. Our bibhographical list of D. sphaerica should be regarded as rather uncertain, 246 THE DINOPHYSOIDAE. since the great difficulties in determining this and closely related species make errors or at least diversity in the assignments of specimens of these species very probable. Occurrence: — Dinophysis sphaerica is recorded at four of the 127 stations. There are 1, 0, 0, 2, 1, and 0 stations on the six lines of the Expedition. Of these four stations, one (4619) is in the Panamic Area; one (4697) is in the Easter Island Eddy; and two (4711, 4720) are in the South Equatorial Drift. There are two records from surface waters (from Salpa stomachs; Stations 4619, 4720), and two from hauls from 300-0 fathoms (Stations 4697, 4711). The temperature range of these four stations at the surface was 75°-79°; the average was 76.2°. At one station (4720) the frequency is 4%, at one station (4619) 1%, and at the two remaining stations less than 1 %. Stein (1883) recorded this species "aus dem Atlantischen Meer." In the Arctic Ocean it has not been recorded more than once, viz., by Aurivillius (1899), oh the north side of Spitzbergen, in Hinlopen Strait, lat. 80° 31' N. Aurivillius (1898a) also found it on four occasions in the Skager Rak, on the west coast of Sweden. Murray and Whitting (1899) reported it from thirteen stations in the Atlantic, between lat. 44° 30' N. and lat. 18° 20' N., and between long. 19° W. and long. 64° W.; Graf (1909) found it at lat. 35° 57' N., long. 7° 50' W. At the Cape of Good Hope it was found by Karsten (1906). From the Mediterranean it has been recorded by the following investigators : — from Naples by Daday (1888) and Zacharias (1906); from the Adriatic by Entz (1902b, 1905). In the Red Sea it was found by Ostenfeld and Schmidt (1901), and Cleve (1903b) ; in the Gulf of Aden, by Ostenfeld and Schmidt (1901); in the Indian Ocean, at lat. 34° 13' S., long. 80° 30' E., by Karsten (1907); in the Antarctic Ocean, at lat. 63° 16' S.-63° 32' S., long. 57° 51' E.-54° 46' E., by Karsten (1905); and in the Gulf of Siam, by Schmidt (1901). As stated in the last section, it is doubtful whether or not the form recorded by Okamura (1912) from Japanese waters under the name of Dinophysis sphaerica is referable to this species. Karsten (1905, 1907) reported this species as taken with closing net at the following stations: — Station 151, lat. 63° 32' S., long. 54° 46' E., 100-10 m. and 300-100 m.; Station 152, lat. 63° 15' S., long. 57° 51' E., 80-60 m.; Station 169, lat. 34° 13' S., long. 80° 30' E., 60-40 m. All the specimens recorded were re- ported as living. Most of the other records mentioned above probably refer to surface hauls. According to the writers who have contributed t(j our knowledge of the SYSTEMATIC ACCOUNT. 247 distriliution of this species, it occurs in waters of the following temperatures and salinities. Aurivillius (1898a) : — temperature: range, 54.1°-65.6°; mean of three observations, 60.2°; salinity: range, 24.0-29.5; mean of tliree observations, 26.3. Aurivillius (1899) : — temperature, 36.4°; salinity, 33.93. Murray and Wliitting (1899): — temperature: range, 58°-79°; mean of thirteen observations, 70.0°. Ostenfeld and Schmidt (1901): — temperature: range, 72.0°-80.2°; mean of four observations, 77.2°. Cleve (1903b): — temperature, 84.6°; salinity, 39.92. Unfortmiately our knowledge of the distribution of D. sphaerica must be re- garded as rather uncertain, since this species, on account of its simple structure, is very easily confused with some closely related forms. Of the authors who have recorded this species, only Stein (1883) and Okamura (1912) give descriptions or figiu-es by means of which theii' determinations may be checked. Ostenfeld and Schmidt (1901), and Schmidt (1901) refer to Stein's (1883) Plate 20, figure 3 and 4 (see above, the sections on comparisons and synonymy). This species has been recorded from waters of very different temperatures. However, it probably is a warm-water form and not endogenetic in the Skager Rak (see Aurivillius, 1898a; Gran, 1900; and Hensen-Ostenfeld, 1913) and in the Arctic Ocean. We are not in the position to discuss the question as to whether or not it is endogenetic in the ^Antarctic Ocean. DiNOPHYSis siMiLis, sp. nov. Figure 31: 1,2 Dinophysis sphaerica Scbvtt, 1895, p. 13, 119, non p. 64, pi. 1, fig. 7: 1,2. Pavill-^rd, 1916, p. 47,60. JoRGENSEN, 1925, p. 11, 23, 44, fig. 29. Diagnosis: — Body subcircular to ovate or subellipsoidal in lateral outline, deepest in or just behind the middle, and 1.03-1.29 tunes longer than deep; sym- metrical or its longitudinal axis is inclmed posteroventrally at about 1°-10°. Anterior cingiilar list 1.4-2.0 times wider than transverse furrow. Left sulcal list 0.50-0.57 the length of body, of subuniform width throughout the greater portion of its length; Ri and Ro subequal, 0.19-0.30 the greatest depth of body; R3 absent or rudimentary (very short) ; posteroventrally this list is rounded or forms a more or less rounded angle of 50°-55°. Theca finely and closely areolate. Length, 43.0-62.0 M. Probably widely distributed in tropical, subtropical, and warm-temperate seas. Description: — This is a rather small to medium-sized species, the body of which is subcii'cular to ovate or subellipsoidal in lateral outline, and deepest in or 248 THE DINOPHYSOIDAE. somewhat behind the middle. The ratio between the length and the depth of the body is 1.03-1.29: 1; in Schutt's (1895, pi. 1, fig. 7: 1) and Jorgensen's (1923, fig. 29) specimens, this ratio is 1.09: 1 and 1.03: 1, respectively; in our specimens it is 1.19-1.29: 1. The longitudinal axis either is about perpendicular to the trans- verse furrow (Figure 31: 1, 2), or it has a posteroventral inclination of about 1°-10°, as in the specimens figured by Schiitt (1895) and Jorgensen (1923). The epitheca is 0.42-0.54 as deep as the hypotheca, of moderate convexity, highest in or near the center, and not visible above the anterior cingular list. The transverse furrow is flat or somewhat convex, and it s width is 1 .5-2.0 the greatest height of the epitheca. The posterior cingular list is 0.14-0.17 the length of the body from the apex. The hypotheca either is symmetrical (Figiu'e 31 : 1, 2), or somewhat inclined posteroventrally (see above). The dorsal and ventral margins are subuniformly, and moderately to rather strongly convex; sometimes (Schiitt, 1895, pi. 1, fig. 7:1; Jorgensen, 1923, fig. 29) the ventral margin is less convex than the dorsal. The postmargin is evenly and rather strongly convex, and confluent with the dorsal and ventral margins. The posterior portion of the body may be of about the same depth as the anterior (Figure 31:2), but usually it is more or less strikingly deeper. The cingular lists are subequal, or the posterior is slightly narrower than the anterior; the anterior is 1.4—2.0 times wider than the transverse furrow and 0.22- 0.29 the greatest depth of the body. Both these lists are funnel-shaped, being in- clined anteriorly at 30°-40°. The anterior has simple, straight, and almost equi- distant ribs, about 10-15 on each valve; the posterior appears always to lack structural differentiation. The right sulcal list extends to or slightly beyond the fission rib of the left sulcal list and is subtriangular, decreasmg gradually in width posteriorly; anteriorly it is about as wide as or somewhat wider than the trans- verse furrow. The left sulcal list is 0.50-0.57 the length of the body and of sub- uniform width throughout the greater part of its length. In the specimens figured by Schiitt (1895) and Jorgensen (1923) the anterior main rib and the fission rib are about 0.19 the greatest depth of the body; in our specimens the corresponding value is 0.28-0.30. The posterior main rib is absent, or rudimentary and short; it is "usually indistinct, or only distinct for a short inner portion" (Jorgensen, 1923, p. 23, fig. 29). Posteriorly the left sulcal list is rounded (Jorgensen, 1923, fig. 29), or it forms a more or less rounded angle of about 50°-55°; one of our specimens (Figure 31 : 2) agrees in this respect rather closely with Schiitt's (1895) Plate 1, figure 7: 1; the other (Figure 31: 1) with Schiitt's (1895) Plate 1, figure 7 : 2. The ventral margin of this Ust is almost straight or gently convex. The main SYSTEMATIC ACCOUNT. 249 ribs are straight or almost so, and not club-shaped or otherwise modified. In some specimens this list is without structure except for the anterior two main ribs; sometimes (Schiitt, 1895, pi. 1, fig. 7:2) its posterior half is finely reticulated. There are no accessory lists or sails. The thecal wall is finely and closely areolate and has scattered pores. Megacytic specimens have not been recorded. The proportions of two of our specimens and of the specimens figured by Schiitt (1895) and Jorgensen (1923) were measured. Dimensions: — Our specimens: Length of body, 43.0-46.4 m. Greatest depth of body, 33.2-38.5 m- The specimen figured by Schiitt (1895, pi. 1, fig. 7: 1) was 62.0 M long and 57.0 /x deep. The type specuiien, as figiu'ed by Jorgensen (1923, fig. 29) was 61.0 n long and 59.0 /i deep. According to Jorgensen (1923, p. 24) this species is "as a rule " 57-60 ix long and 55-57 ^ deep. Variations: — Jorgensen (1923), who examined a great number of specimens, does not indicate by any statement that he had found this species to be variable. Furthermore, the specimens figured by Schiitt (1895) and Jorgensen (1923) are very similar. These cncumstances seem to indicate that this species is rather con- stant. The specimen represented by our Figure 31 : 2 should be regarded as aber- rant. The following characters are the most variable : the size of the body, the inclination of the longitudinal axis of the body, the relative depth of the body, and the shape and structure of the posterior portion of the left sulcal list. Comparisons: — This species is established on the specimens figured by Schiitt (1895, pi. 1, fig. 7) and Jorgensen (1923, fig. 29) under the name of Dino- physis sphaerica Stein, and on two specimens found in the material of the Expedi- tion. Its most outstanding characteristics are the absence or strong reduction of the posterior main rib of the left sulcal list and the rounded to rounded-angular shape of the posteroventral portion of this list. Before Jorgensen's (1923) paper was available, we were inclined to consider these peculiarities as signs of recent binary fission (see Figure 46: 2, of D. caudata f. acutiforynis) . However, the fact that Jorgensen (1923) fomid this form to be "numerous" in the Mediterranean and always characterized by these features very strongly indicates that we are concerned with characters of specific value. The reasons given by Jorgensen (1923, p. 23) for assigning to D. sphaerica Stein the specimens of D. similis found by him hardly can be considered as valid. As the type of D. similis we regard the specimen represented by Jorgensen's (1923) figure 29. The structurally closest-known relatives of Dinophijsis similis are D. sphae- rica and D. ovum. The two last species differ from the first mainly in having the 250 THE DINOPHYSOIDAE. posterior main rib of the left sulcal list well developed, about as long as or some- what longer than the fission rib of this list, and in ha\dng the posteroventral portion of this list angular, instead of rounded to rounded angular. (See also Dinophysis sphaerica and Phalacroma expuhum, the sections on comparisons.) Synonymij: — As stated above, Dinophysis similis has previously been treated by Schiitt (1895) and Jorgensen (1923) under the name of D. sphaerica Stein. D. sphaerica Pavillard (1916) is also identical with D. similis, and not with D. sphaerica Stein, as is plainly shown by Pavillard's (1916, p. 60) statement that his specimens of this species had the body of about the same shape as the specimen represented by Schiitt's (1895) Plate 1, figiu'e 7: 1, and that their left sulcal list "r^pond exactement au dessin de Schiitt; son bord inferieur est con- vex et non concave comme dans le dessin approximatif de Stein." It should be noted in this connection that D. sphaerica Schiitt (1895, p. 64) refers to Phala- croma rotundatum (Claparede and Lachmann). Occurrence: — Dinophysis similis is recorded at two of the 127 stations. Of these two stations, one (4691) is on the third line of the Expedition and in the Easter Island Eddy; the other (4701) is on the fourth line and in the South Equatorial Drift, just to the north of the Easter Island Eddy. Both records refer to hauls from 300-0 fathoms. The temperature of these two stations at the surface was 72° and 73°. The frequency is less than 1 % at both stations. Schiitt (1895) does not give any information as to the localities of his material; however, it probably was taken in the tropical and subtropical regions of the Atlantic or near Naples. Pavillard (1916) found this species to be rare in the Gulf of Lyons. Jorgensen (1916) reports it as fairly common and "widely dis- tributed m the Western Mediterranean "; he also found it in the Bay of Cadiz and in the Guinea Current. It probably is widely distributed in tropical, subtropical, and warm-temperate seas. Dinophysis okamurai, sp. nov. Figure 31:5 Dinophysis ranhoffenii Okamura, 1907, pariim, p. 131, pi. 5, fig. 41c. Diagnosis: — Body subovate in lateral outline, with dorsal shoulder-like constriction at girdle, with ventral margin somewhat concave between R2 of left sulcal list and epitheca, deepest in the middle, and 1.32-1.50 times longer than deep. Anterior cingular list about as wide as transverse furrow. Left sulcal list 0.44-0.50 the length of the body; distance between Ri and R3 0.37-0.42 the length SYSTEMATIC ACCOUNT. 251 of the body; R2 is 0.15, R3 is 0.12-0.29 the greatest depth of body; margin forms angle of 100°-110° at R3; R3 incUned posteriorly at 5°-25°. Length, 38.3-53.0 m- Eastern tropical Pacific, Japan. Description: — This is a rather small or medium-sized species, the bodj' of M hich is subovate m lateral outline, with a dorsal shoulder-like constriction at the girdle, with the ventral margin somewhat concave between the fission rib of the left sulcal list and the epitheca, and deepest in the middle. The ratio between the length and the depth of the body is 1.32-1.50: 1. The longitudinal axis is about perpendicular to the girdle. The epitheca is about 0.37-0.38 as deep as the hypotheca, slightly convex, highest in or near the middle, and not \asible above the anterior cingular list. The transverse furrow is flat or slightly concave, and about three times as wide as the greatest height of the epitheca. The posterior cingular list is about 0.10- 0.11 the length of the body from the apex. The hypotheca is somewhat asym- metrical. The dorsal margin is moderately and subuniformly convex. The ventral margin is gently to moderately concave in front of the fission rib of the left sulcal list; behmd this rib it is slightly convex or almost straight. The postmargin is strongly and subuniformly convex and confluent with the dorsal and ventral margins. The posterior portion of the body is decidedly deeper than the anterior. The anterior cingular list, which is about as wide as the transverse furrow and 0.16-0.28 the greatest depth of the body, has an anterior inclination of 30°- 40°. The posterior cingular hst is narrower than the anterior, sometimes even very narrow as in the type specimen; it has an anterior inclination of 40°-50°. The structure of the cingular lists is unknown. The right sulcal list extends to the fission rib of the left sulcal list and is subtriangular in shape, decreasing gradually in width posteriorly. The left feulcal list is comparatively small. Its length is 0.44- 0.50 the length of the body, and the distance between the anterior and posterior main ribs is 0.37-0.42 the length of the body. The anterior main rib is about 0.18, the fission rib 0.15, and the posterior main rib 0.12-0.29 the greatest depth of the body. Behind the posterior main rib this list decreases suddenly in width. Be- tween the anterior main rib and the fission rib and between the fission rib and the posterior main rib, the free margin of this list is straight or almost so; at the fission rib it forms an obtuse angle ; at the posterior main rib it forms an angle of 100°-110°; behind the last rib it is almost straight. The main ribs of this list are straight or almost so, and not club-shaped or otherwise modified. The distance between the anterior main rib and the fission rib is 0.35-0.40 the distance between the anterior and posterior main ribs. The posterior main rib has a posterior in- 252 THE DINOPHYSOIDAE. clination of 5°-25°. We do not know whether or not this list has any structural differentiation besides the main ribs. There are no accessory lists or sails. The structure of the theca is unknown; judging by Okamura's (1907, pi. 5, fig. 41c) figure, it appears to be a fine areolation. Megacytic stages have not been recorded. The dimensions of the type specimen and of the specimen figured by Oka- mura (1907, pi. 5, fig. 41c) were measured. Dimensions: — Length of body, 38.3-53.0 ^ (type, 53.0 m). Greatest depth of body, 25.4-40.0 n (type, 40.0 m). Comparisons: — Dinophysis okamurai is established on a single outline draw- ing of a specimen in lateral view found in the material of the Expedition. The shape of the body in dorsoventral view, the structure of the cingular lists and of the theca are unknown, and our knowledge of the structure of the left sulcal list is incomplete. When compared with the type specimen, the specimen figured by Okariiura (1907, pi. 5, fig. 41c) is smaller (38.3 ix as compared with 53.0 m) and narrower (ratio between length and depth of body, 1.50: 1 as compared with 1.32: 1); its ventral margin is more concave anteriorly, and its posterior cingular list is some- what wider relatively. The fission rib of its left sulcal list is comparati^■ely long, which makes the ventral margin of this list more angular at this rib than in any other known species. The difference in the shape of the ventral margin of the body may be due to the fact that Okamura (1907) made his drawing from a de- tached valve. The shape of the body in lateral view and the shortness of the left sulcal list in this species are suggestive of Dinophysis ovum and Phalacroma expulsum. How- ever, these resemblances cannot be considered as sufficient proofs of a close re- lationship. Dinophysis okamurai is easily distinguished from D. ovum by the char- acteristic sinuation of the ventral margin of its body and from Phalacroma expul- sum by the shape of its left sulcal list. Synonymy: — Under the name of Dinophysis vanhoffenii Ostenfeld, Okamura (1907) figures two specimens, one of which (his Plate 5, fig. 41c) belongs to the species described above. Later (1912) he writes D. vanhoffenii Okamura (1907) as a synonym of D. ovum Schiitt. Occurrence: — Dinophysis okaiuurai is recorded at only one (4G73, the type locality) of the 127 stations, on the third line of the Expedition, in the Peruvian Current, from 300-0 fathoms, at a surface temperature of 67°. The frequency is less than 1% (one specimen). Previously the species was fovmd in Japanese waters by Okamura (1907). SYSTEMATIC ACCOUNT. 253 DiNOPHYSis FORTH Pavillard Figure 31:7 Dinophysis forlii Pavillard, 1923a, p. 881. Dinophysis laevis Bergh, Podchet, 1883, p. 426, pi. 18, 19, fig. 6. Dinophysis intermedia Pavillard, 1916, p. 58, pi. 3, fig. 4. Forti, 1923, p. 110, 112, 190, 209, fig. 119. JoRGENSEN, 1923, p. 19, 21, 22, 44, 45. Dinophysis intermedia f. pachyderma Johgensen, 1923, fig. 21. Diagnosis: — Body narrowly subovate in lateral outline, with flattened ventral margin, and broadly and evenly rounded posteriorly ; deepest somewhat behind the middle, and 1.44-1.56 times longer than deep; longitudinal axis de- flected posteroventrally at 3°-15°. Markedly compressed bilaterally, with lateral contours almost parallel. Anterior cingular list 1.4-2.0 as wide as transverse furrow, without ribs. Left sulcal list 0.63-0.67 as long as body; distance between Ri and R3 0.50-0.53 the length of body; Ro is 0.26-0.30, R3 is 0.33-0.43 the great- est depth of body ; margm forms angle of 65°-85° at R3 ; R3 inclmed posteriorly at 10°-25°; besides main ribs this list has distinct reticulation. Theca finely and closely areolate. Length, 62.3-66.1 n. Probably widely distributed in subtropical, warm-temperate, and temperate seas. Description: — This is a medium-sized species, narrowly subovate in lateral outUne, with flattened ventral margin, broadly and evenly rounded posteriorly, and deepest somewhat behind the middle. The ratio between the length and the depth of the body is 1.48 (1.44^1.56) : 1. The longitudinal axis is deflected postero- ventrally at 9° (3°-15°). The epitheca is 0.38-0.48 as deep as the hypotheca, moderately convex (Figure 31: 7) to almost flat (Jorgensen, 1923, fig. 21), not visible above the an- terior cingidar list, and highest in or somewhat dorsally to the center. The trans- verse furrow is flat, or slightly concave or convex, and its depth is 2.0-2.5 the greatest height of the epitheca. The posterior cingular list is 0.09-0.10 the length of the body from the apex. The hypotheca is somewhat asymmetrical to sub- symmetrical, and more or less deflected posteroventrally (see above). Its dorsal margin (from the girdle to the antapex) is rather strongly and almost evenly con- vex. Its ventral margin (between the anterior and posterior main ribs of the left sulcal list) is almost straight, or slightly and evenly convex. Its postmargin is evenly subsemicircular and confluent with the dorsal and ventral margins; it does not form an angle just behind the posterior main rib of the left sulcal list. The posterior end of the body is strikingly deeper than the anterior. In dorso- ventral view the body is markedly compressed bilaterally, and its lateral con- tours are subparallel (Jorgensen, 1923, p. 19). 254 THE DINOPHYSOIDAE. The cingular lists may be subequal, but in most cases the anterior is some- what wider than the posterior ; the anterior is 1 .5-2.0 as wide as the transverse furrow, and 0.20-0.23 the greatest depth of the body; both are funnel-shaped, with an anterior incUnation of 30°-50°; sometimes the posterior is somewhat less, in other cases somewhat more inclined anteriorly than the anterior; both lack distinct structural differentiation. The right sulcal list ends at or just behind the fission rib of the left sulcal list; it is subtriangular and decreases gradually in width posteriorly; anteriorly it is about as wide as, or slightly wider or narrower than the transverse furrow; its ventral margin is slightly and evenly con^'ex, or almost straight. The left sulcal list is fairly large. Its length is 0.66 (0.63-0.67) the length of the body; and the distance between the anterior and posterior main ribs is 0.52 (0.50-0.53) the length of the body. The anterior main rib is 0.25 (0.24-0.26), the fission rib 0.28 (0.26-0.30), and the posterior main rib 0.38 (0.33-0.43) the greatest depth of the body; behind the last rib this list decreases suddenly in width. Between the anterior and posterior main ribs the free margin of this list either is almost straight (Pavillard, 1916, pi. 3, fig. 4), or it is ahnost straight anteriorly and gently convex posteriorly (Jorgensen, 1923, fig. 21) ; at the posterior main rib it forms an angle of 75° (65°-85°) ; behind the last rib it either is nearly straight (Pavillard, 1916, pi. 3, fig. 4) or more or less concave (Jorgensen, 1923, fig. 21). The two anterior of the main ribs are of moderate strength, and straight or nearly so; the posterior, which sometimes (Jorgensen, 1923, fig. 21) is club-shaped, is almost straight (Pavillard, 1916, pi. 3, fig. 4), or more or less strikingly concave posteriorly (Jorgensen, 1923, fig. 21). The distance between the anterior main rib and the fission rib is 0.44 (0.40-0.50) the distance between the anterior and posterior main ribs. The posterior main rib has a posterior in- clination of 18° (10°-25°). Besides the main ribs this list has a well-developed reticulation. There are no accessory lists or sails. The thecal wall is closely and rather finely areolate; the areoles are subequal in size. Alegacytic specimens have not been recorded. The dimensions of one specimen found in the material of the Expedition and of the specimens figured by Pavillard (1916, pi. 3, fig. 4) and Jorgensen (1923, fig. 21) were measured. Dimensions: — Our specimen: Length of body, 62.3 /u. Greatest depth of body, 40.0 /z. Type specimen (Pavillard, 1916, pi. 3, fig. 4) : Length of body, 66.1 n. Greatest depth of body, 45.7 y.. The specimen figured by Jorgensen (1923, fig. 21) was 63.8 M long and 43.7 m deep. According to a statement in the text, the speci- mens examined by Jorgensen (1923, p. 19) were about 64 yu long and 49 m deep "reckoned vertically and parallel to girdle." SYSTEMATIC ACCOUNT. 255 Variations: — This species appears to be rather variable. The most variable of its characters are as follows : the inclination of the longitudinal axis of the body, the curvature of the epitheca, the shape of the posterior portion of the left sulcal list, and the shape and structure of the posterior main rib of this list. Comparisons: — The specimen found in the material of the Expedition and assigned to Dinophysis foriii shows a close agreement with the type specimen as figured by Pavillard (1916, pi. 3, fig. 4). It differs from the type mainly in being somewhat smaller and more slender. The specimen figured by Jorgensen (1923, fig. 21) under the name of Dino- physis intermedia Pavillard, f. pachyderma has the longitudmal axis of the body rather strikingly inclmed posteroventrally ; the posterior portion of the left sulcal hst is strikingly recurved; and the posterior main rib of this list is decidedly club- shaped and has a pronounced posterior concavity. Dinophysis fortii is morphologically and probably also genetically inter- mediate between D. ovum Schiitt and D. schroderi. Of these three species D. oimm is the shortest, most rounded, and presumably most primitive ; D. schroderi is the longest, most elongate, and presumably most highly developed (see also D. schroderi, the section on comparisons). Synomjmy: — This species was established by Pavillard (1916), who de- scribed and figured it under the name of Dinophysis intermedia. Jorgensen (1923), who described it under the same name, once, on page 22, used the name D. inter- medium {lapsus pennae). Without referring to it in the text, Jorgensen (1923, fig. 21) figured a specimen under the name of D. intermedia Pavillard, f. pachy- derma. According to our opmion, this specimen is an old, thick-walled repre- sentative of the main species, and so its name has been included in our bibUo- graphical list of D. fortii. Fortf (1922) gave a reproduction of Pavillard's (1916) figure of the type specimen. Pavillard (1923a, p. 881) rejected intermedia as the specific name of this form and replaced it with fortii, on account of the fact that Dinophysis intermedia Pavillard (1916) was antedated by D. intermedia Cleve (1902b, p. 30). Jorgensen (1923, p. 19) suggests that Dinophysis atlantica Ehrenberg (1854, p. 239) may be identical with this species. This question cannot be settled as yet, since Ehrenberg's (1854a) diagnosis of D. atlantica is too incomplete for certainty of specific identification. Pouchet (1883, p. 426, pi. 18-19, fig. 6) figures, under the name of Dinophysis laevis Bergh, a specimen which very decidedly resembles D. fortii. Pouchet's assignment of this specimen is wrong. According to our opinion, D. laevis Pouchet 256 THE DINOPHYSOIDAE. (1883) may be identical with D.fortii, but Jorgensen (1899, p. 30) refers it to D. acuminata Claparede and Lachmann, and Paulsen (1908, p. 15) to D. arctica Mereschowsky. Jorgensen (1923, p. 19) suggests that the specimen which Cleve (1901c, p. 239) recorded from the Florida Current under the designation of Dinophysis acuta Ehrenberg belonged to D. intermedia { = fortii). This suggestion has not been verified as yet. Occurrence: — Dinophysis fortii is recorded at only one of the 127 stations. This station (4667) is on the second line of the Expedition and in the Peruvian Current. The depth is 300-0 fathoms, the surface temperature 68°, and the fre- quency less than 1% (one specimen). The type locality of this species is the Gulf of Lyons (Pavillard, 1916). Forti (1922) obtained one specimen in the Ligurian Sea, and Jorgensen (1923) found the species to be widely distributed throughout the Mediterranean. Out- side the Mediterranean the species has been found by Jorgensen (1923), who re- ported it from the Bay of Cadiz, and "from the Florida Current (off the northern coast of Florida, temp. 23°C., 6/5 1910) and from several places, partly in mixed water, along the northern course of the Gulf Stream, far to the south of Nova Scotia, from 39° to 43° N., and 70°^0° W. (temp. 22°-18°C., salmity abt. 36.5)." Dinophysis atlantica Ehrenberg (1854a), which may be identical with D. fortii, was taken off Newfomidland. D. laevis Pouchet (1883), which also may be identi- cal with D.fortii, was found at Concarneau, on the west coast of France. The species probably is eupelagic and widely distributed in subtropical, warm-temperate, and temperate seas. Dinophysis norvegica Claparede and Lachmann? Figure 31:8 f Dinophysis norrcgica Broch, 1910n, p. 31, fig. 1: A, B. The single specimen represented by Figure 31 : 8 was taken in the Panamic Area, at Station 4638 of the Expedition. The depth was 300-0 fathoms, and the surface temperature was 75°. The length of this specimen is 44.5 fi. The simil- arity between this specimen and the one from Spitzbergen figured by Broch (1910a, fig. 1:A) is very striking. However, since the surface markings of the theca are not indicated in our drawing, and a certain knowledge of this character is essential for a well-founded determination of the species of the Acuta group, our assignment of this specimen should be regarded as tentative. A record of this specimen is being published mamly because of the fact that Dinophysis SYSTEMATIC ACCOUNT. 257 norvegica, which supposedly is a neritic and boreoarctic species, has not been found previously in waters of tropical and subtropical character. Our specimen might have been a vagrant from polar waters. DiNOPHYSis scHRoDERi Pavilkrd Figure 31:6 Dinophijsis schroederi Pavillarh, 1909, p. 2S4, fig. 5; 1916, p. 58, pi. 3, fig. 5. Schilleb, 1911a, p. 52; 1911b, p. 90. Schroder, 1911, p. 22, 25, 36. Forti, 1922, p. 110, 112, 190, 209, fig. 120. Jokgensen, 1923, p. 21, 45, fig. 22. Diagnosis: — Body elongated in lateral outline, deepest somewhat behind the middle, 1.68-1.90 times longer than deep; dorsal margin fairly strongly con- vex; ventral margin nearly straight in front of R3 of left sulcal list and more or less angular at R3; narrowly rounded to subacute posteriorly; longitudinal axis deflected posteroventrally at 1°-12°. Markedly compressed bilaterally; lateral contours parallel or even slightly concave. Anterior cingular list 1.4-2.0 as wide as transverse furrow, without ribs. Left sulcal list 0.66-0.70 as long as body; distance between Ri and R3 0.58-0.61 the length of body; R2 is 0.20-0.31 and R3 is 0.33-0.50 the greatest depth of body; margin forms angle of 40°-55° at R3; R3 inclined posteriorly at 10°-35° ; except for main ribs this list almost lacks struc- ture. Theca finely and closely areolate. Length, 73.0-88.0 fx. Probably widely distributed in subtropical, warm-temperate, and temperate seas. Description: — This is a medium-sized species, the body of which is elongate in lateral outline, with well-convex dorsal margin, almost straight ventral margin, subacute antapical end, and the greatest depth somewhat behind the middle. The ratio between the length and the depth of the body is 1.79 (1.68-1.90): 1. The longitudinal axis is deflected posteroventrally at 7° (1°-12°). The epitheca is 0.43-0.46 as deep as the hypotheca, rather strongly convex (Figure 31:6) to almost flat (Jorgensen, 1923, fig. 22), not visible above the an- terior cingular list, and highest in or somewhat dorsally to the center. The trans- verse furrow is flat or slightly concave or convex, and its width is 1.5-2.5 the greatest height of the epitheca. The posterior cingular list is 0.07-0.11 the length of the body from the apex. The hypotheca is asymmetrical and more or less de- flected posteroventrally (see above). Its dorsal margin (from the girdle to the antapex) is rather strongly convex; sometmies (Figure 31:6) the convexity is almost even, sometunes (Pavillard, 1916, pi. 3, fig. 5) it is more pronounced near the middle than anteriorly and posteriorly. Its ventral margin (between the 258 THE DINOPHYSOID.\E. anterior and posterior main ribs of the left sulcal list) is almost straight or but slightly convex or concave. Just behind the posterior main rib of the left sulcal list the margin forms a more or less pronounced but rounded corner, and between this corner and the antapex the margin is almost straight or slightly convex. Posteriorly the body is narrowly rounded to subacute. However, according to Jorgensen (1923, p. 21), some specunens which "undoubtedly should be referred to D. schroederi" had the posterior end of the body "evenly rounded"; these specimens were from the southern part of the Atlantic. In dorsoventral view the body is markedly compressed bilaterally, with the lateral contours parallel or even slightly concave (Jorgensen, 1923, p. 21). The cingular lists may be subequal (Pavillard, 1909, fig. 5), but in most cases the posterior is somewhat narrower than the anterior; the anterior is 1 .4-2.0 as wide as the transverse furrow and 0.21 (0.17-0.25) the greatest depth of the body; both are funnel-shaped, with an anterior inclination of 20°^0°; sometimes the posterior is somewhat less inclined anteriorly than the anterior; both lack distinct structural differentiation. The right sulcal list ends at or just in front of the fission rib of the left sulcal list; it is subtriangular and decreases gradually in width posteriorly; anteriorly it is about as wide as or slightlj' wider than the transverse furrow ; its ventral margin is almost straight, gently and evenly con- vex, or gently convex anteriorly and slightly concave posteriorly. The left sulcal list is fairly large; its length is 0.68 (0.G6-0.70) the length of the body, and the distance between the anterior and posterior main ribs is 0.60 (0.58-0.61) the length of the body. The anterior main rib is 0.22 (0.17-0.27), the fission rib 0.25 (0.20-0.31), and the posterior main rib 0.41 (0.33-0.50) the greatest depth of the body; behind the last rib this list decreases suddenly' in width. Between the an- terior and posterior main ribs the free margin of this list either is almost straight, or it is almost straight anteriorly and gently convex posteriorly (Pavillard, 1916, pi. 3, fig. 5), or slightly undulating, concave anteriorly and posteriorly and convex in the middle (Pavillard, 1909, fig. 5) ; at the posterior main rib this margin forms an angle of 48° (40°-55°) ; behind the last rib it either is almost straight (Figure 31 : 6) or more or less strikingly concave (Jorgensen, 1923, fig. 22). The main ribs of this list are of moderate strength; the two anterior are straight or almost so; the posterior, which may be club-shaped (Jorgensen, 1923, fig. 22), is almost straight (Figure 31: 6), or more or less strikingly concave posteriorly. The dis- tance between the anterior main rib and the fission rib is 0.42 (0.40-0.50) the distance between the anterior and posterior main ribs. The posterior main rib has a posterior inclination of 20° (10°-35°). Except for the main ribs this list is SYSTEMATIC ACCOUNT. 259 without structure, or it has a faint reticulation near the posterior main rib. There are no accessory Usts or sails. The thecal wall is closely and rather finely areolate. Megacytic specimens have not been recorded. The dimensions of one specimen found in the material of the Expedition and of the specimens figured by Pavillard (1909, fig. 5; 1916, pi. 3, fig. 5) and Jorgensen (1923, fig. 22) were measured. Dimensions: — Our specimen: Length of body, 73.0 ^u. Greatest depth of body, 38.5 m- Type specimen (Pavillard, 1909, fig. 5) : Length of body, 88.0 n. Greatest depth of body, 52.2 /j. The specimen figured by Pavillard (1910, pi. 3, fig. 5) was 81.5 m long and 46.7 /ji deep; according to Pavillard (1916, p. 58) this species has a length of 80 ^ or more, and a depth of 45-50 n. The specimen figured by Jorgensen (1923, fig. 22) was 83.2 ^ long and 45.4 ^ deep. Variations: — Judging by the available figures, this species appears to be relatively constant. However, the specimens examined by Jorgensen (1923) seem to have been rather variable, since this investigator points out that he at times found it difficult to distinguish with certainty between this species and Dinophysis acuta Ehrenberg (Jorgensen). The most variable characters are: the size of the body, the uiclmation of the longitudinal axis of the body, the curvature of the epitheca, the shape of the dorsal margin of the body, and the shape and structure of the posterior main rib of the left sulcal list. Con^parisons: — The specimen found in the material of the Expedition and assigned to Dinophysis schroderi agrees rather closely with the specimens from the Mediterranean figured by Pavillard (1909, 1916). It differs from these specimens mainly m being somewhat smaller (73.0 ^t as compared with 81.6-88.0 m) and more slender. The specimen from the Mediterranean figured by Jorgensen (1923) also agrees closely with Pavillard 's (1909, 1916) specimens. It is noteworthy because of its flat epitheca and the somewhat club-shaped posterior main rib of its left sulcal list. Dinophysis schroderi is structurally \'ery closely related to D. acuta Ehren- berg (Jorgensen). Indeed, we should have been inclined to regard these two forms as belonging to the same species, if it were not for the fact that D. schroderi was established by such a careful investigator as Pavillard, who, moreover, reported D. acuta from the same locality as D. schroderi. According to Pavillard's (1916) data, these two forms differ mainly in the size of the body and in the shape of the posterior portion of the body. Jorgensen (1923, p. 21) seems to be somewhat in doubt as to the status of D. schroderi as an independent species; he writes as 260 THE DINOPHYSOIDAE. follows: D. schroderi "seems therefore to be a true Mediterranean species {or form)," and it "is at times hard to distinguish with certainty" from D. acuta. According to Jorgensen (1923), D. schroderi differs from D. acuta in the following respects: — (1) it is somewhat more elongated; (2) its dorsal margin is somewhat more evenly convex; (3) the posterior portion of its body, as a rule, is slightly more rounded; (4) it is markedly compressed in dorsoventral view with the lateral contours of the body parallel to each other or even slightly concave, while hi D. acuta the body in dorsoventral view "is generally narrowly elliptical, with convex lateral contours"; (5) its anterior cingular list is slightly higher relatively, and narrower at the base; and (6) its left sulcal Hst is almost without reticulation. Dinophysis schroderi is also structurally close to D.fortii. It differs from this species mainly in the subacute posterior end of its body, in the angularity of the ventral margin of its body at the posterior main rib of the left sulcal list, and in the very slight development of the reticulation of its left sulcal list (see also D. fortii, the section on comparisons). Another close relative of Dinophysis schroderi probably is D. norvegica. However, the latter species, which is easily recognized by its heavy surface mark- ings, is too insufficiently known for a detailed comparison. Synonymy: — This species was established by Pavillard (1909, p. 284, fig. 5), who described and figured it under the name of Dinophysis schroederi. Later it was described and figin-ed by Pavillard (1916) and by Jorgensen (1923) under the same name. Forti (1922) gave a reproduction of Pavillard's (191G) figure of this species. Occurrence: — Dinophysis schroderi was obtained at only one of the 127 sta- tions. This station (4574) is on the first line of the Expedition and in the Cali- fornia Current. The sample was taken at the surface. The surface temperature was 69°, and the frequency less than 1% (one specimen). The species was first recorded by Pavillard (1909, 1916) from the Gulf of Lyons, which thus is the tyjje locality. Later it was reported from the Ligurian Sea by Forti (1922) and from the Adriatic by Schiller (1911a, 1911b) and Schroder (1911). Jorgensen (1923) found it to be widely distributed throughout the Mediterranean. Outside the Mediterranean it has been recorded only by Jorgen- sen (1923), who found it in the Bay of Cadiz and at "several places in the south- ern Atlantic, partly near and west of the southernmost part of Africa (German South-Pole Expedition, 1/7-1/8 1903), partly east of South America ('Fram' 1911)." Of the writers w ho have contributed to our knowledge of the distribution of SYSTEMATIC ACCOUNT. 261 this species, Schiller and Schroder do not give any descriptions or figures by means of which their determmations may be checked. This is a eupelagic species widely distributed l)ut rare in subtropical, warm- temperate, and temperate seas. 2. Hast ATA Group. Fourteen species are assigned to this group. Of these the following have been found to occur m the Eastern Pacific : — Dinophysis hastata, D. uracantha, D. urceolus, D. monacantha, D. trapezium, D. swezyi, D. collaris, D. schutii, D. nias, D. jorgenseni, and D. triacantha. Dinophysis hastata Stein Figure 32, 34 Dinophysis hastata Stein, 1883, pi. 19, fig. 12. HENSE^, 1887, p. 77, tab. 1.3, 15; 1895, p. 190; 1911, p. 165, tab. 15. MoBins, 1887, p. 121. O.stenfeld, 1898b, p. 42, tab. 5; 1899, tab. 5, 8; 1900, p. 56, tab. 2, 5-7; 1906, p. 18; 1909, p. 22; 1913a, p. 309; 1916b, p. 13. Murhay & Whitting, 1899, p. 331, tab. 1-3, 6. JoRGENSEN, 1899, p. 32, l\-.\ii; 1912, p. 11, 16; 1923 (partim?). p. 31-33, 44. Lemmermann, 1899a, p. 319, 372; 1901a, p. 374; 1902a, p. 263; 190.5a, p. 36. Ostenfeld & Schmidt, 1901, p. 169. Cleve, 1901a, p. 15; 1901c, p. 239; 1902b, p. 29; 1903b, p. 343. Entz, 1902b, p. 94: 1905, p. 111. OsTENFELiJife Paulsen, 1904, p. 164, 171. Zacharias, 1906, p. 530. Kajisten, 1907, p.234. Paulsen, 1907, p. 5; 1908, p. 12, 13, fig. 9; 1912, p. 289. Nathansohn, 1908, p. 604; 1909, p. 46; 1910, p. 61, fig. 29: 12. Graf, 1909, p. 143, 151, 155, 159, 171, 173, 179, 182, 184, 188, 190, 192. Pavillard, 1909, p. 283, 284; 1915a, p. 2; 1916, p. 47, 53, 60; 1923a, p. 879, 880, fig. 2A. Oka.mura, 1912, p. 19, 33, fig. 7.3-75. Mangin, 1912, p. 33, tab. 2. Forti & Lssel, 1923, p. 3, 4, fig. ; 1924, p. 7. PhalacToma hastatum P.willard, 1909, p. 2S3, fig. 4; 1916, p. 53; 1923a, p. 879, 880. Schiller, 1911a, I). 52; 1912, p. 27. Schroder, 1911, p. 17, 25, 37. Forti, 1922, p. 104, fig. 108. Lebour, 1925, p. 83, fig. 21c. Dinophysis hastata var. or form, phalacromides Jorgensen, 1923, p. 31, fig. 41 (nomeri rejectutn). Dinophysis hastata var. or form, uracanihides Jorgensen, 1923, p. 31, 32, 33, fig. 40 (iiomcn njrctvm). Dinophysis uracantha var. mediterranea Jorgen.?en, 1923, p. 32, fig. 43. Dinophysis hastata var. parvula Lindemann, 1923, p. 219, fig. 6. Diagnosis: — Body subobovate, subovate, or subellipsoidal in lateral outline ; rounded, subtruncate, or truncate anteriorly*; subacute, or narrowly to broadly rounded posteriorly; deepest somewhat in front of to somewhat behind the mid- dle; length: depth, 1.10-1.60: 1; longitudinal axis deflected posteroventrally at 0°-15°. In dorsal view narrowly obovate, about 2.5 times longer than wide. Cingular lists sometimes but slightly inclined anteriorly; epitheca sometimes visible above anterior cingular list; anterior cingular list 1.5-2.5 times wider than transverse furrow; posterior sometimes ribbed. Left sulcal list 0.48-0.77 the length of body; distance between Ri and R3 0.37-0.58 the length of body; R2 is 0.11-0.29, R3 is 0.25-0.74 the depth of body; margin forms angle of 20°-90° at R3; R3 inclined posteriorly at 15°-55°; sometimes with reticulation and accessory ribs. With triangular to spine-like posterior sail, usually somewhat ventrally to midline, and inclined ventrally at 0°-40°; its length 0.19-0.61 the depth of body and 1.0-4.0 its basal width; with or without central rib or reticulation, sometimes 262 THE DINOPHYSOIDAE. with marginal ribs; separated from left sulcal list. Theca usually finely and closely areolate. Length, 42.5-90.0 m- Widely distributed in tropical, subtropical, and warm-temperate seas, seldom in colder waters. Description: — This is a medium-sized to small species, the body of which is very variable in shape. It is subobovate, subovate, or subeUipsoidal in lateral outline; rounded, subtruncate, or truncate anteriorly; subacute, or narrowly to broadly rounded posteriorly; and deepest somewhat in front of to somewhat behind the middle. The ratio between the length and the depth of the body is 1.10-1.60: 1; in our specimens this ratio is 1.27 (1.17-1.38): 1; in Stein's (1883, pi. 19, fig. 12) figure of the type specmien it is about 1.60:1; in Okamura's (1912, fig. 73-75) specimens, 1.10-1.18:1; and in Pavillard's (1909, fig. 4A), Jorgensen's (1923, fig. 40, 41), and Lindemann's (1923, fig. 6) specimens, 1.14- 1.27: 1. The longitudinal axis either is perpendicular to the transverse furrow, as in the type specimen, or it is deflected posteroventrally at an angle of 1°-15°; in our specimens this deflection is 6° (0°-ll°); in Pavillard's (1909) specimen, which is extreme in this respect, it is 15°. The epitheca is very variable in depth and shape; it is 0.67 (0.54-0.86) as deep as the hypotheca, usually gently convex, but sometimes rather strongly convex (Figure 32: 14) or more or less flat (Figiu-e 32: 9; Figure 33: 2), highest in the center or near the dorsal side (Figure 32: 9; Figure 33: 1, 3), and sometimes visible above the anterior cingular list. The transverse furrow is flat or slightly convex, and about as wide as or somewhat wider or narrower than the greatest height of the epitheca. The posterior cingular list is 0.18 (0.12-0.24) the length of the body from the apex. The hypotheca sometimes is symmetrical, e.g., in the type specimen and in the specimens represented l)y our Figure 32: 12, 13, 16; but usually it has a more or less pronounced postero ventral inclination (see above). The dorsal margin (from the girdle to the antapex) usually is subuniformly and gently to moderately convex; sometimes it is more or less flattened anteriorly (Figure 32:2) or posteriorly (Figure 32:8), or both anteriorly and posteriorly (Figure 32: 10, 11). The ventral margin (from the girdle to the antapex) usually is more or less flattened posteriorly, or it is subuniformly and gently to moder- ately convex; sometimes (Figure 32: 9) it forms a broadly rounded corner at or near the posterior main rib of the left sulcal list. The posterior poi-tion of the body is subacute, or narrowly to broadly rounded; in most specimens it is strik- ingly narrower than the anterior portion, but sometimes it is of about the same depth (Figure 32: 7) or even deeper (Figure 32: 13). In dorsoventral view (Pavil- SYSTEMATIC ACCOUNT. 263 lard, 1909, fig. 4B) the body is narrowly obovate, about 2.5 times longer than wide, and its side contours are evenly convex. The cingular lists are subequal or the posterior is somewhat narrower than Figure 32. — Dinophysis hastnta Stfin, all, except 4, .seen in right lateral view; 4, in ventral view; structure of thecal wall indicated in 2, 13, and 1.5 only; 4 and 14 from the same specimen. X 430. 1, 5, 6, 9, 10, from Station 4737 (300-0, or 100-^ fathom.s) ; 2, from Station 4732 (300-0 fathoms) ; 3, from Station 4734 (300-0 fathoms); 4, 14, from Station 4724 (300-0 fathom.s); 7, from Station 4722 (300-0 fathoms); 8, 13, from Station 4730 (.300-0 fathoms); 11, 12, from Station 4740 (.300-0 fathoms); 1.5, 16, 17, from Station 4711 (300-0 fathoms). 264 THE DINOPHYSOIDAE. the anterior. In the type specimen (Stein, 1883, pi. 19, fig. 12) and in the speci- men figured by Jorgensen (1923, fig. 43) under the name of Dinophysis uracantha var. mediterranea, the posterior list is very narrow, about 0.30-0.40 the width of the anterior (see the section on comparisons). The anterior cingular list is 1.5- 2.5 times wider than the transverse furrow and 0.20 (0.13-0.27) the greatest depth of the body. In most specimens both these lists are more or less funnel-shaped, being inclined anteriorly at 20°-50°; sometimes, however, the posterior or even both of them (Jorgensen, 1923, fig. 41; Pavillard, 1909, fig. 4A) may be sub- horizontal. The anterior one has, on each valve, about 15-25 usually simple, straight, and almost equidistant ribs; sometimes these ribs ramify and anas- tomose into an indistinct and incomplete reticulum (Jorgensen, 1923). The posterior cingular list may have the same structure as the anterior (Pavillard, 1909, fig. 4A), but usually it lacks structural differentiations. The right sulcal list sometimes ends at the fission rib of the left sulcal list, but usually it extends somewhat beyond this rib; in exceptional cases (Jorgensen, 1923, fig. 41) it ex- tends as far as to the posterior main rib of the left sulcal list ; it either is subtri- angular, decreasing gradually in width posteriorly, or it is of subuniform width throughout the greater part of its length and rounded posteriorly (Figure 32 : 9) ; anteriorly it is about as wide as the transverse furrow, sometimes somewhat nar- rower. The left sulcal list is of moderate size or rather large, and usually strikingly wider posteriorly than anteriorly. Its length is 0.58 (0.48-0.77) the length of the body, and the distance between the anterior and posterior ribs is 0.48 (0.37-0.58) the length of the body. The anterior main rib is 0.19 (0.13-0.25), the fission rib 0.22 (0.11-0.29), and the posterior main rib 0.52 (0.25-0.74) the greatest depth of the body. Behind the posterior main rib this list decreases suddenly in width. Between the anterior and posterior main ribs the free margin of this list is almost straight (Figure 32: 6, 11), evenly and gently to moderately convex (Figure 32: 11, 13), or slightly irregular (Pavillard, 1909, fig. 4A; Jorgensen, 1923, fig. 41). At the posterior main rib this margin forms an angle of 45° (20°-90°). Behind this rib the margin may be almost straight (Figure 32 : 11), but usually it is gently to moderately concave. The main ribs of this list are of moderate strength; the anterior two are straight or almost so; the posterior, which sometimes (Figure 32: 6, 10) is club-shaped, may be straight, but in most specimens it is gently to moderately concave posteriorly. The distance between the anterior main rib and the fission rib is 0.42 (0.30-0.55) the distance between the anterior and posterior main ribs. The posterior mam rib has a posterior inclination of 30° (15°-55°). In some specimens the entire left sulcal list has a fairly well-developed reticula- SYSTEMATIC ACCOUNT. 265 tion; in other specimens it has numerous fine ribs and imperfect reticulation; some tunes there is no structural differentiation except the three main ribs. On the right valve there is an acute, triangular to spine-like posterior sail, which is situated either on or generally somewhat ventrally to the midline, and is mclined ventrally at 0°-40°; in our specimens this sail has a ventral inclination of 22° (7°^0°), in Stein's (1883) specimen it was parallel to the longitudinal axis of the body. The left sulcal list usually ends at a distance from the posterior sail equal- ing about 0.30 the greatest depth of the body; in extreme cases the corresponding figure may be as high as 0.45, or the left sulcal list may extend almost to the posterior sail (Figiu'e 32: 8). The size of the posterior sail is somewhat variable, probably due to the fact that in each binary fission one of the daughter schizonts has to form this structure anew. When fully developed, its length is 0.46 (0.19- 0.61) the greatest depth of the body; i.e., its average length is slightly less than that of the posterior main rib of the left sulcal hst. It is 2.2 (1.0-4.0) times longer than it is wide at the base. In most cases it is straight or almost so; its margins are almost straight, or gently convex, concave, or sigmoid; and the angle at its tip is 22° (10°-40°). In most of our specimens this sail is supported by a central rib, which may be quite or nearly solid (Figure 32: 1, 2), more or less broken up into an irregular reticulum (Figure 32: 3, 5, 7), or bifurcate at the base (Figure 32: 9, 10). In one of our specimens (Figure 32: 11) this sail had a marginal rib besides a central rib, in another (Figure 32: 12) only a marginal rib (see also the section on comparisons). No parasagittal lists are developed. In most specimens (Figure 32 : 2) the thecal wall is finely and closely areolate and has scattered pores; some specimens (Figure 33: 1) are characterized by rather large areoles, others (Figure 33: 2) are reticulate. In northern waters the living cell, accordmg to Jorgensen (1923, p. 31), is "of a light rose, without chromatophores, or hyaline." Megacytic stages were seen by Pavillard (1916, p. 47) and by ourselves (Figure 32: 4). The dimensions of nmeteen of our specimens and of the specimens figured by Stein (1883), Pavillard (1909), Okamura (1912), Jorgensen (1923), and Linde- mann (1923) were measured. Dimensions: — Our specimens : Length of body, 50.2-80.6 ix (average, 67.2 ^i). Greatest depth of body, 37.7-63.8 n (average, 53.3 ai). The size of the type speci- men (Stein, 1883, pi. 19, fig. 12) is unknown. Judging by Stein's (1883) informa- tion about the magnification of his figures, given in an introductory remark to the explanation of the plates, the length of this specimen was somewhere between 266 THE DINOPHYSOIUAE. 71.5 fi and 110.0 jji. A specimen found by Jorgensen (1899, p. 32) on the west coast of Norway was 73 /x long and 53 ix deep. The specimens recorded by Pavil- lard (1909, fig. 4A) under the name of Phalacroma hastahun were 85-90 m long and 35-40 /x deep. The three speciiBens figured by Okamura (1912, fig. 73-75) were 59.5-77.5 M long and 50.6-70.3 n deep. The specimens represented by Jorgensen's (1923) figures 40 and 41 were 58.7 fx long and 47.5 m deep, and 78.0 n long and 61.3 M deep, respectively; the specimen figured by Jorgensen (1923, fig. 43) under the name of Dinophysis uracantha var. mediterranea was 42.5 ^ long and 38.0 n deep. Lindemann's (1923) specimen was 60 ju long. Figure 33. — ■Diiiophi/sis hastata Stein, right lateral view of three aberrant specimens; structure of thecal wall incUcated in 1 and 2. Specimen represented by 2 had left sulcal list and posterior sail heavily reticulated, but its posterior sail had no central rili. X 430. 1, from Station 4711 (300-0 fathoms); 2, from Station 4730 (300-0 fathoms); 3, from Station 4732 (300-0 fathoms). Variations: — The variabihty of Dinophysis hastata, as conceived in the present paper and by Jorgensen (1923), is amazing, and, according to our opinion, there can be but little doubt that future in\'estigators, with adequate material at their disposal, will find it necessary and feasible to divide this species into two or more systematic units. The principles which we have been forced to apply in our treatment of this species completely upset those used in almost all the other species of the genera Dinophysis and Phalacroma. Indeed, some of our specimens of this species have the habitus characteristic of Dinophysis, while others look like rather typical representatives of Phalacroma. Jorgensen (1923, p. 31) writes that this species in the Mediterranean at times varies "to an extraordinary degree." The following characters are the most variable : — the size of the body, the inclination of the longitudinal axis of the body, the shape and the relative depth of the epitheca, the shape of the hypotheca, the inclination of the cingular lists, the width and the shape of the left sulcal list, the shape and relative size of the right sulcal list, and the size, shape, structure, and position of the posterior sail. SYSTEMATIC ACCOUNT. 267 Cotnjxnisons: — The type specimen, as figured by Stein (1883, pi. 19, fig. 12), has an oblong, subo\'oidal, and symmetrical body, subtruncate anteriorly and subacute posteriorly; its anterior cingular list is wide and funnel-shaped, and its posterior cingular list is very narrow ; its posterior sail is located on the very mid- line, directed posteriorly, and furnished with a sunple, strong, central rib. Speci- mens of the same habitus as the type specimen appear not to have been found by later mvestigators. For instance, Okamura (1912) emphasizes that in his speci- mens the body was "shorter and more roundish" than in the type specimen, and he found the posterior sail to be without central rib or its central portion "is partly thickened like a spine." Pavillard (1916, p. 60), too, found the body to be shorter and more rounded than in Stein's (1883) figure of the type specimen; he also remarks that in his specimens the posterior sail was displaced ventrally and that the presence of a central rib in this sail "parait tres douteuse." Hensen (1895, p. 190) observed that his specimens of this species were "schief." It is, of course, not excluded that Stein (1883) made mistakes in his representation of some of the structures of the type specimen. However, the striking variability of this species makes it appear possible that this author has correctly figured a specimen of an unusual habitus. A glance at our figures will show that we have found specimens of this species (Figure 32: 16, 17) almost as elongated as in the figure of the type. As to the position and structiire of the posterior sail, Jorgensen (1923, p. 31) writes as follows: "Terminal spine displaced toward the ventral side, and directed obliquely ventrally, rarely almost in the middle and pointing straight down, often with distinct median rib and more or less broad sail lists, or quite or nearly solid throughout, with narrow sail lists or none." With regard to the width of the posterior cingular list of the type, it should be mentioned that this list is very narrow in Dinophysis umcantha var. mediterranea Jorgensen (1923, p. 43), a form which probably is referable to D. hastata (see below, the section on synon- ymy). As previously indicated, none of the specunens found in the material of the Expedition and assigned to Dinophysis hastata agrees exactly with the type speci- men of this species, as figured by Stein (1883, pi. 19, fig. 12). Moreover, only a few of our specimens (Figure 32: 16) show a comparatively close resemblance to the type. The specimens represented by Figure 32: 1, 2, 3, 5, 6, 9, 10, may be considered as typical of Dinophysis hastata, as found by us in the tropical and subtropical regions of the eastern Pacific. This form may be characterized briefly in the fol- lowing way. The body is somewhat asymmetrical, with posteroventrally inclined 268 THE DINOPHYSOIDAE. longitudinal axis, subovoidal, subtruncate anteriorly, and subacute to narrowly rounded posteriorly. The posterior sail is located somewhat ventrally to the midline, directed posteroventrally, and furnished with a central rib, which either is quite or nearly solid throughout or more or less broken up into an irregular reticulum. On the other hand, the specimens represented by Figure 32: 4, 7, 8, 11-17, and by Figure 33 should be regarded as unusual. A detailed discussion of each of these figures would lead too far; the mentioning of a few outstanding facts must suffice. Figure 32: 16, as previously mentioned, approaches the type specimen as figured by Stein (1883). Figure 32: 12 resembles Jorgensen's (1923, fig. 40) figure of Dinophysis hastata f. uracanthides, and approaches D. uracantha in having the posterior sail bordered with marginal ribs; this specimen would have been as- signed to D. uracantha, if its posterior sail had not been displaced ventrally. Figure 33: 2 approaches Jorgensen's (1923, fig. 41) figure of D. hastata f. phala- cromides. We would have separated the three specimens represented by I'lgure 33, from D. hastata, were it not for the following statement by Jorgensen (1923, p. 31): "I have not been able with certainty to distinguish Phalacroma hastatum Pavill. from Dinophysis hastata, not even as a separate variety, though it is often rather characteristic." Since our material is not sufficient for a well- founded solution of this problem, we have considered it advisable to accept temporarily Jorgensen's (1923) decision (see Pavillard, 1923a, p. 879, 880). Jorgensen (1923, p. 31) states that the specimens of this species, found on the west coast of Norway and in the Black Sea, are relatively small, with narrow left sulcal list without posterior main rib, and with small posterior sail. In none of the fully developed specimens of this species, found by us in the material of the Expedition and in a collection from Alaska, did the left sulcal list lack the poste- rior main rib. Dinophysis hastata is structurally closely related to D. uracantha. The ranges of variation of these two species, as conceived in the present paper, even may overlap (see, for instance, Figure 32: 12). When typical, accordmg to our con- ception of the species, D. hastata differs from the typical D. uracantha in being somewhat larger, in having the body more elongated and somewhat deflected posteroventrally, and in having the posterior sail situated somewhat ventrally to the midline. Dinophysis monacantha, D. pusilla, and D. alata are presumably also rather close relatives of D. hastata. These relationships are indicated by the presence of a posterior sail and by the relatively simple structure of the left sulcal list. The SYSTEMATIC ACCOUNT. 269 typical D. hastata differs from D. monacaniha mainly in lacking a parasagittal list, in having the longitudinal axis of the body deflected posteroventrally instead of posterodorsally, and in having the posterior sail with a central and not with a marginal rib. From D. pusilla it differs strikingly in its larger size and in its more elongated body. D. alata is characterized by the wide parasagittal list along the dorsal margin of the body. See also D. monacaniha, the section on compari- sons. Synonymy: — The species was established by Stein (1883, pi. 19, fig. 12), who figured it under the name of Dinophysis hastata. Later it was described by Paulsen (1908, p. 13) and figured by Okamura (1912, fig. 73-75) and Forti and Issel (1923) under the same name. Paulsen (1908), who bases his description on Stein's (1883) figure of the type specimen, gives a reproduction of this figure. Other reproductions of Stein's (1883) figure are to be found in Nathansohn (1910, fig. 29: 12) and Lebour (1925, fig. 21e). For reasons mentioned in the section on comparisons, the name of Phala- croma hastatum Pavillard (1909) has been included in our bibliographical list of Dinophysis hastata. However, we wish to emphasize that this should be regarded as a tentative measure, and that we consider it possible that Phalacroma hastatum is specifically distinct. It should be pointed out in this connection that Schiller (1911a, 1912) and Schroder (1911), who recorded Phalacroma hastatum from the Adriatic Sea, as well as Forti (1922), who found it in the Ligurian Sea, did not mention Dinophysis hastata as occurring in their collections. Forti (1922) gives a reproduction of Pavillard's (1909) figure of Phalacroma hastatum (see also Pavillard, 1923a, p. 879). It should be noted that if this form is specifically dis- tinct, then its name must be changed, since P. hastatum is preoccupied. This name was first used as a nomen nudum by Hensen (1895, p. 190; see also Hensen, 1911, p. 166, 167). Jorgensen (1923, p. 31) writes that he had tried to distinguish between two varieties of Dinophysis hastata, viz., var. uracanthides and var. phalacromides, the latter "answering to Phalacroma hastatum Pavill.," "but this again cannot, as far as my experience goes, be maintained with certainty." Dinophysis uracantha Stein var. mediterranea Jorgensen (1923) should be assigned to D. hastata in spite of its small size, since the longitudinal axis of its body is deflected posteroventrally, and since its posterior sail is situated ventrally to the ant apex and is furnished with a central rib and not with marginal ribs. Lindemann's (1923, fig. 6) figure of Dinophysis hastata var. parvula from the Bosporus corroborates Jorgensen's (1923, p. 31) statement that the specimens of 270 THE DINOPHYSOIDAE. this species found by him in the Black Sea were relatively small, with small left sulcal list and small posterior sail. Occurrence: — Dinophysis hasfata is recorded at fort,y-eight of the 127 sta- tions. There are 10, 9, 9, 7, 9, and 4 stations on the six hnes of the Expedition. Of these forty-eight stations, three (4571, 4580, 4583) are in the California Current; six (4587, 4588, 4590, 4594, 4596, 4605) are in the Mexican Current; four (4613, 4634, 4637, 4638) are in the Panamic Area; eight (4650, 4651, 4659, 4664, 4665, 4666, 4675, 4676) are in the Peruvian Current; five (4689, 4691, 4695, 4697, 4699) are in the Easter Island Eddy; twenty (4679, 4680, 4681, 4683, 4687, 4701, 4705, 4707, 4711, 4717, 4721, 4722, 4724, 4730, 4732, 4734, 4737, 4739, 4740, 4741) are in the South Equatorial Drift; one (4742) is in the South Equatorial Current; and one (4542) is in the Equatorial Counter Current. There are eight records from the surface (Stations 4583, 4588, 4596, 4666, 4676, 4680, 4741, 4542) ; at two of these stations (4583, 4676) the species was taken in hauls from 300-0 fathoms as well as at the surface; at one station (4666) it was taken in a haul from 800-0 fathoms as well as at the surface ; at the remaining five stations it is recorded from surface catches only. At one station (4737) the species is re- corded from 100-0 fathoms and 300-0 fathoms, and at one station (4681) from 800-0 fathoms and 300-0 fathoms. All the remaining records refer to hauls from 300-0 fathoms only. The species is also recorded from surface waters in Acapulco Harbor, off the Mexican Current. This station is not included in the 127 stations mentioned above. The temperature range of these forty-eight stations at the surface was 66°- 85°; the average was 75.3°. At the eight stations in the surface catches of which this species was found, the surface temperature ranged from 67° to 84°; the aver- age was 76.7°. At two of the surface stations (4542, 4666) a frequency of 1 % was found ; in the remaining cases frequencies of less than 1% are recorded. For the catches from 100, 300, or 800 fathoms to the surface the records of frequency are as fol- lows: 2% at one station (4689), 1 % at ten stations (4571, 4580, 4587, 4634, 4681, 4724, 4734, 4737, 4740, 4742), and less than 1% at the remaining stations. In addition it may be mentioned that we have found this species also in plankton from Yes Bay, Alaska. Stein (1883) recorded this species "aus dem Atlantischen Mcer mid der Sudsee." It was mentioned as occurring in the Cattegat by Ostenfeld (1906, SYSTEMATIC ACCOUNT. 271 1916b) and Paulsen (1912) ; and in the Skager Rak by Paulsen (1907), Jorgensen (1912), and Ostenfeld (1913a). (It has not been found in the Baltic. Lenimer- mann's, 1902a, p. 263, statement that Mobius, 1887, recorded it from the western part of the Baltic is erroneous.) It was found on the west coast of Norway by Jorgensen (1899, 1912); in the North Sea, by Ostenfeld (1906, 1909) and by Paulsen (1912). Ostenfeld (1898b, 1899, 1900) reported it from twelve localities in the northern part of the Atlantic, between lat. 59° 36' N.-63° 57' N. and long. 2° 28' W.-28°45' W. Hensen (1887) and Mobius (1887) found it west of Scot- Figure 34. — Occurrence of Dinophi/sis hastala Stein. Large, solid circle.s indicate records from vertical hauls; squares, records from siu'faee hauls; triangles, records from both vertical and surface hauls; small, solid circles, stations at which this species was not found; small, open circles, stations from which no plankton catches were examined. land, at about lat. 57° N. and long. 10°-11° W., and Ostenfeld and Paulsen (1904) at lat. 59° 41' N., long. 22° 27' W. Other papers in which this species is mentioned as occurrmg in the northern part of the Atlantic are Ostenfeld (1906) and Paulsen (1912). Mangin (1912) found it in the English Channel; Jorgensen (1923), on the southwest coast of Portugal and in the Bay of Cadiz; Murray and Whitting (1899), at six stations in the Atlantic, between lat. 45° 38' N.-18° 20' N. and long. 13° 53' W.-52° 40' W.; Cleve (1901c), at twelve localities in the Atlantic, between lat. 62° N.-19° S., and long. 1° W.-66° W.; Cleve (1903b), at lat. 37° N., long. 8° W.; Cleve (1902b), at lat. 24° S., long. 4° E.; Graf (1909), at lat. 4° 37' N., long. 9° 52' W., lat. 17° 22' S., long. 0° 13' E., and lat. 31° 12' S., long. 16° 1' E.; 272 THE DINOPHYSOIDAE. Hensen (1911), in the Gulf Stream, Florida Current, Sargasso Sea, Canary Cur- rent, North Equatorial Current, Guinea Current, South Equatorial Current, and on the coast of Brazil. In the Mediterranean this species has been found by several investigators, viz.: in the Gulf of Lyons, by Pavillard (1909, 1916); off Monaco, by Nathansohn (1908, 1909); in the Ligurian Sea, by Forti (1922); near Naples, by Zacharias (1906); in the Adriatic Sea, by Entz (1902b, 1905), Schroder (1911), Schiller (1911a, 1912), and Forti and Issel (1923); and at a rather great number of localities in various parts of the Mediterranean, bj^ Jorgensen (1923). Lindemann (1923) reported it from the Bosporus; and Jorgensen (1923) from the Marmora Sea, Bosporus, and the Black Sea. In the Red Sea it was found by Cleve (1903b); in the Gulf of Aden, by Ostenfeld and Schmidt (1901), and by Cleve (1903b) ; in the Arabian Sea, at lat. 9° N., long. 59° E., by Cleve (1901a) ; in the Indian Ocean, at lat. 2° S., long. 91° E., by Cleve (1901a), at lat. 34° 13' S., long. 80° 30' E., by Karsten (1907), and at the Suvadiva Atoll, at Rodriguez Island, and on the west coast of Madagascar by Graf (1909); in the Antarctic Ocean, south of Africa, at lat. 42° 55' S., long. 22° 53' E., by Griif (1909); in the East Indies, near Sumatra, Celebes, Amboina, and on the north coast of New Guinea, by Griif (1909); in Japanese waters, by Okamura (1912); and in the Peruvian Current by Lemmermann (lS99a). Karsten (1907) reported the species as taken with a closing net between 80 and 60 meters. Most of the remainmg records refer to surface hauls. According to the investigators who have contributed to our knowledge of the distribution of this species, it occurs in waters of the following temperatures and salinities. Ostenfeld (1898b): — temperature, 51.8°; salinity, 35.25. Ostenfeld (1899): — temperature: range, 44.6°-52.7°; mean of four observations, 48.8°. Salinity: range, 35.37-35.50; mean of four observations, 35.41. Ostenfeld (1900): — temperature: range, 45.0°-55.4°; mean of seven observations, 52.4°. Salinity: range, 35.00-35.34; mean of seven observations, 35.18. Murray and Whitting (1899): — ^ temperature : range, 61°-79°; mean of six observations, 71.9°. Cleve (1901a): — temperature, 82.3°; salinity, 32.13. Cleve (1901c): — temperature: range, 44.6°-77.2°; mean of eighteen observations, 63.7°. Salinity: range, 33.93- 37.36; mean of eighteen observations, 35.71. Cleve (1902b): — temperature, 71.6°; salinity, 35.95. Cleve (1903b): — temperature, 60.3°-84.7°; salinity, 36.44-40.70 (four samples). Ostenfeld and Paulsen (1904): — temperature, 52.2°. Griif (1909): — temperature, 82.8°-84.1° (two samples). Salinity: range 33.78-36.22; mean of eleven observations, 34.79. Of the authors who have contributed to our knowledge of the distribution of SYSTEMATIC ACCOUNT. 273 this species, only Stein (1883), Pavillard (1916), Okamura (1912), Forti (1922), Jorgensen (1923), Forti and Issel (1923), and Lindemann (1923) give descriptions or drawings by means of which their determinations of this species may be checked (see the section on synonymy). Dinophysis hastata is eupelagic and occurs in tropical, subtropical, and warm- temperate regions of all seas. It has been fomid in the northern part of the At- lantic, in the North Sea, Norwegian Sea, Skager Rak, and Cattegat, but it is probably a warm-water form and not endogenetic in these waters. The specimens recorded probably were stragglers of southern origin, carried to the north by the Gulf Stream (see Paulsen, 1912, p. 289). It is thermophilous but eurythermal and, as far as known, stenohaline. According to our records, it is almost uniformly distributed throughout the area investigated by the Expedition and has its optimum habitat in deeper waters, within the levels of photosynthesis. Dinophysis uracantha Stein Figure 35, 36 Dinophysis uracantha Stein, 1883, pi. 20, f3s. 22, 23. Hensen, 1895, p. 190; 1911, p. 105, tab. 15. non ScHiJTT, 1895, p. 16, 17, pi. 2, fig. 9. Lemmerm.\nn, 1899a, p. 373; 1901a, p. 375; 1905a, p. 30. MnnRAV & Whitting, 1899, p. 331, tab. 1-6, 8, 9. Schroder, 1900a, p. 35; 1906a, p. 326, 329. OsTENFELD & ScHMiDT, 1901, p. 171. Cleve, 1903b, p. 344. Zachari.^s, 1906, p. 557. Kofoid, 1907a, p. 196; 1910a, p. 184. Karsten, 1907, p. 234, 237, 247. Pavillard, 1916, p. 60; 1923a, p. 879, 880, fig. 2B. JoRiiENSEN, 1923, p. 31, 33, 34, fig. 42. Forti & Lssel, 1924, p. 6, 7. Lebodr, 1925, p. 83, fig. 20g. non Dinophysis uracantha var. iiiedilcrrartea Jorgen.sen, 1923, ]). 32, fig. 43. Diagnosis: — Body subcircular to subovate in lateral outline, seldom trun- cate anteriorly; deepest somewhat in front of to somewhat behind the middle; length: depth, 0.98-1.14: 1 ; longitudinal axis deflected posterodorsally at 0°-13°. Anterior cingnlar list 1.5-2.5 times wider than transverse furrow ; epitheca seldom visible above this list. Left sulcal list 0.60-0.75 the length of body; distance be- tween Ri and R3 0.50-0.57 the length of body ; Ro is 0.13-0.30, R3 is 0.50-0.88 the greatest depth of body; margin forms angle of 20°-40° at R3; R3 inclined poste- riorly at 20°-55°; sometimes with secondary ribs and reticulation. With triangu- lar or claw-like posterior sail located somewhat dorsally to midline; its length is 0.45-0.88 the greatest depth of body and 1.5-3.0 its basal width; with marginal ribs but without central rib; separated from left sulcal list. Theca finely and closely areolate. Length, 37.5-63.5 m- Widely distributed in tropical, subtropical, and warm-temperate .seas, seldom in colder waters. Description: — This is a small or medium-sized species of rather variable 274 THE DINOPHYSOIDAE. shape. The body is subcircular to subovate in lateral outline, seldom truncate anteriorly, and deepest somewhat in front of, to somewhat behind, the middle. The ratio between the length and the depth of the body is 1.07 (0.98-1.14): 1. The longitudinal axis either is perpendicular to the transverse furrow, as in the specimen represented by Stein's (1883) Plate 20, figure 23, or it is deflected posterodorsally at 8° (1°-13°). The epitheca is variable in depth and shape; it is 0.59 (0.40-0.69) as deep as the hypotheca, usually gently convex, but sometimes rather strongly convex (Figure 35:2) or flat (Figure 35:3), highest in or near the center, and seldom visible above the anterior cingular list. The transverse furrow is flat to slightly convex, and about as wide as or more or less decidedly wider than the greatest height of the epitheca. The posterior cingular list is 0.17 (0.09-0.20) the length of the body from the apex. The hypotheca sometimes is synnnetrical, e.g., in the specimens represented by Figure 35: 3 and by Stein's (1883) Plate 20, figure 23, but usually it has a more or less pronounced posterodorsal inclination. The dorsal margin (from the girdle to the antapex) sometimes is subuniformly and moder- ately to gently convex, sometimes (Figure 35 : 4) flattened anteriorly, and some- times (Figure 35: 1) somewhat more strongly convex anteriorly than posteriorly. The ventral margin (from the girdle to the antapex) sometimes, as in the type specimen and in Figure 35 : 3, is subuniformly and moderately convex, but usually it is more or less flattened posteriorly. The posterior portion of the body is broadly rounded, but seldom (Figxire 35 : 2) strikingly deeper than the anterior. The cingular lists are subequal or the posterior is slightly narrower than the anterior. In the two specimens figured by Stem (1883, pi. 20, fig. 22, 23) the posterior hst is very narrow and only about 0.30-0.50 as wide as the anterior; this is probably due to an imperfect observation, since Stein (1883) made this list very narrow m almost all his figures of species of Dinophysis. The anterior of these lists is 1.5-2.5 times wider than the transverse furrow and 0.27 (0.20-0.40) the greatest depth of the body. In most specimens both these lists are funnel- shaped, being mclined anteriorly at 20°-50°; however, the posterior, which some- tunes is somewhat less inclined than the anterior, may be subhorizontal (Jorgen- sen, 1923, fig. 42). The anterior has, on each valve, 10-20 simple, straight, and almost equidistant ribs, some of which may be incomplete ; the posterior appears always to lack structure. The right sulcal list extends to or slightly beyond the fission rib of the left sulcal list, and is subtriangular, decreasing gradually in width posteriorly, or (Figure 35: 7) of subuniform width throughout the greater part of its length and rounded posteriorly; anteriorly it is about as wide as the SYSTEMATIC ACCOUNT. 275 transverse furrow. The left sulcal list is comparatively large and usually strikingly wider posteriorly than anteriorly. Its length is 0.68 (0.60-0.75) the length of the body, and the distance between the anterior and posterior main ribs is 0.53 (0.50-0.57) the length of the body. The anterior mam rib is 0.21 (0.15-0.24), the fission rib 0.22 (0.13-0.30), and the posterior main rib 0.67 (0.50-0.88) the greatest depth of the body. Behind the posterior main rib this list decreases suddenly in width. Between the anterior and posterior main ribs the free margin of this list is nearly straight, as in the type specimen and in our Figure 35: 5, or it is gently sigmoid or somewhat irregular; when sigmoid, it may be convex an- teriorly and concave posteriorly (Figure 35:6), or vice versa (Figure 35:2, 3); when irregular, it may be straight anteriorly and concave posteriorly (Figure 35: 4), or vice versa (Figure 35: 7). At the posterior main rib this margin forms an angle of 28° (20°^0°). Behind this rib the margin usually is gently or moderately concave, but sometimes it is gently or moderately sigmoid. The main ribs of this list are of moderate strength; the two anterior are straight or almost so; the posterior, which in exceptional cases is club-shaped, may be straight, but in most specimens it is gently or moderately concave posteriorly. The distance between the anterior main rib and the fission rib is 0.38 (0.33-0.45) the distance between the anterior and posterior main ribs. The posterior main rib has a posterior in- clination of 37° (20°-55°) . Besides the three main ribs this list has either a more or less developed reticulum of anastomosing, almost parallel transverse ribs, or a varying number of almost parallel, or more or less irregular transverse rilss, not connected by anastomoses. On the right valve there is an acute, triangular or claw-like posterior sail, which is situated on or generally somewhat dorsally to the midline and directed posteriorly or slightly ventrally. The left sulcal list ends at a distance from the posterior sail equaling 0.38 (0.29-0.50) the greatest depth of the body. WTien the posterior sail is fully developed, its length is O.CO (0.45-0.88) the greatest depth of the body, i.e., its average length is slightly less than that of the posterior main rib of the left sulcal list; and it is 2.1 (1.5-3.0) times longer than it is wide at the base. Sometimes both its margins are almost straight; but generally its dorsal margin is convex and its ventral concave, or both of them are gently sigmoid (Figure 35: 6) ; the angle at its tip is 18° (10°-25°). It is supported by marginal ribs and lacks a central rib. No parasagittal lists are developed. The thecal wall is finely and closely areolate ; the size of the areoles is some- what variable. The structure varies with age; in old, thick-walled specimens it is very distinct (Jorgensen, 1923, p. 32). Megacytic stages have not been recorded. 276 THE DINOPHYSOIDAE. The dimensions of nine of our specimens and of the specimens figured by Stein (1883) and Jorgensen (1923) were measured. Dimensions: — Our specimens: Length of body, 37.5-63.5 ju (average, 50.8 p). Greatest depth of body, 33.1-55.4 n (average, 47.7 n). The size of the type speci- men is unknown. Judging by Stein's (1883) information about the magnification of his figures, given in an introductory note to the e.xplanations of the plates, the length'of this specimen was somewhere between 40.0 /x and 61.3 ix. The specimen figured by Jorgensen (1923, fig. 42) was 56.3 n long and 50.0 /j. deep. Variations: — Dinophysis uracantha, as conceived in the present paper, is rather strikingly variable, and future investigations probably will prove it neces- t'luUKE 35. — Dinophysis uracantha Stein, right lateral view. 1, 4-7, typical; 2, 3, atyiiical. In 2 tlie marginal ribs of the posterior sail maj' have been overlooked. X 430. 1, from Station 4724 (300-0 fathoms); 2, 3, from Station 4732 (300-0 fathom.'i) ; 4, 5, 7, from Station 4722 (300- 0 fathoms); C, from Station 4711 (300-0 fathoms). sary and possible to split this species into two or more systematic units. Our material is imfortunately insufficient ft)r a well-founded solution of this ques- tion. The following characters are the most variable : — the size of the body, the shape and the relative depth of the epitheca, the shape of the hypotheca, the shape and the relative width of the left sulcal list, and the shape and size of the posterior sail. Comparisons: — Stein (1883, pi. 20, fig. 22, 23) figures two specimens of fairly different habitus under the name of Dinophysis uracantha. Of these speci- mens the one represented l)y his Plate 20, figure 22, should be regarded as the SYSTEMATIC ACCOUxNT. 277 type (Jorgensen, 1923, p. 32) and may be characterized briefly in the following manner: the body is subcircular in lateral view, and its longitudinal axis is slightly deflected posterodorsally ; the anterior cingular list is very wide, with narrow base, funnel-shaped, and ribbed; the posterior cingular list is very narrow, mclined anteriorly, and without ribs; the posterior main rib of the left sulcal list has a pronounced posterior concavity, and the posteroventral portion of this list is strikingly recurved; the posterior sail is placed somewhat dorsally to the antapex and is supported by marginal ribs and not by a central ril); the theca has very fine areolation. The specunen represented by Stein's (1883) Plate 20, figure 23, differs from the type mainly in the following respects : the body is more elongated and decidedly deeper posteriorly than anteriorly, and its longitudinal axis is perpendicular to the girdle; the base of the anterior cingular list is wider; the posterior main rib of the left sulcal list has but a slight posterior concavity, and the posteroventral portion of this list is not strikingly recurved ; the posterior sail is situated at the very antapex; the areolation of the thecal wall is somewhat coarser. Both these specimens belong to D. uracantha as conceived in the present paper. Most of the specimens found in the material of the Expedition and assigned to Dinophysis uracantha (Figure 35:1, 4-7) show a striking resemblance to the type; they differ mainly in the following respects: the posterior cingular list is about as wide as the anterior, the posterior main rib of the left sulcal list has but a slight posterior concavity, and the posteroventral portion of this list is not strik- ingly recurved. In the two last characters our specimens resemble the specimen represented by Stein's (1883) Plate 20, figure 23. The specimens represented by Figure 35 : 2, 3, one of which is subovate with strongly rounded epitheca, the other subovate and strikingly truncate anteriorly, should be regarded as unusual in our material and may belong to two hitherto undescribed species, closely re- lated to D. uracantha. In the specimen represented by Figure 35: 2 the marginal ribs of the posterior sail may have been overlooked. The specimen figured by Jorgensen (1923, fig. 42) under the name of Dino- physis uracantha undoubtedly is correctly determmed. It differs from the type mainly in the f ollowmg respects : the body is less regularly subcircular, its dorsal margin being decidedly flatter than the ventral ; the base of the anterior cingular Ust, or in other words the epitheca, is somewhat deeper; the posterior cingular list is wider. The difference in the width of the posterior cingular list may be appar- ent, since in Stein's (1883) iconography almost all drawings of species of Dino- physis are characterized by the extreme narrowness of this hst. 278 THE DINOPHYSOIDAE. The structurally closest-known relative of Dinophysis uracantha is D. ur- ceolus, from which it is easily recognized by its more regularly shaped body (see the latter species, the section on comparisons). Next to D. urceolus, D. hastala appears to be most closely related to D. uracnntha (see D. hastata, the section on comparisons). According to Jorgensen (1923, p. 33), D. pusilla may be a degener- ate form of D. uracantha. The former species resembles the latter in its small size and in the subcircular shape of its body; on the other hand, the posteroventral position of its posterior sail is suggestive of a closer relationship to D. hasfatn (see also D. monacantha, the section on comparisons). Synonymy: — This species was established by Stein (1883, pi. 20, fig. 22, 23), who figured it under the name of Dinophysis uracantha. Later it was de- scribed and figured under the same name by Jorgensen (1923, p. 32, fig. 42) and Pavillard (1923a, fig. 2B). Lebour (1923, fig. 20g) published a reproduction of Jorgensen's (1923) figure. The specimens referred to Dinophysis uracantha by Schiitt (1895, pi. 2, fig. 9: 1-5) were assigned to a new species, D. schiitti, by Murray and Whitting (1899, p. 331). The specimen represented by Jorgensen's (1923) figure 42 is, in our opinion, a fairly typical D. uracantha. D. uracantha var. mediterranca Jorgen- sen (1923, fig. 43), on the other hand, probably represents D. hastata, since its body is fairly elongated and somewhat deflected posteroventrally, and since its posterior sail is located ventrally to the antapex and is supported by a central rib and not by marginal ribs. The specimen figured by Jorgensen (1923, fig. 40) under the name of D. hastata f. uracanthides and the one represented l)y our Figure 32: 12 occupy an intermediate position between D. hastata and D. ura- cantha, and their allocation to the former species should be regarded as tentative. The elongation of the body of these specimens and the posteroventral position of the posterior sail are suggestive of D. hastata, while the fact that the posterior sail is supported by marginal ribs and not by a central rib indicates a close rela- tionship to D. uracantha. Furthermore, the longitudinal axis of the body is about perpendicular to the girdle, while in typical D. hastata and D. uracantha it is in- clined posteroventrally and posterodorsally, respectively. Jorgensen (1923, p. 33) writes: "The species Dinophysis uracantfm was first recorded from the Mediterranean by Schroder, 1900 (from Naples, he does not mention the far more frequent form D. hastata, so that his note perhaps applies to the larger and similar D. hastata var. uracanthides . . .)." This suggestion, of course, is without foundation, since Schroder (1900a) recorded neither D. ura- cantha nor D. hastata from Naples. D. uracantha is mentioned only once by SYSTEMATIC ACCOUNT. 279 Schroder (1900a, p. 35), and then as an example of a warm-water species of Dinophysis. Occurrence: — Dinophysis uracantha is recorded at twenty-nine of the 127 stations. There are 4, 6, 3, 5, 10, and 1 stations on the six hnes of the Expedition. Of these twenty-nine stations, four (4588, 4590, 4592, 4605) are in the Mexican Current; two (4637, 4638) are in the Panamic Area; five (4647, 4648, 4659, 4671, 4676) are in the Peruvian Current; three (4691, 4697, 4699) are in the Easter Island Eddy; and fifteen (4679, 4701, 4705, 4711, 4717, 4719, 4721, 4722, 4724, Figure 36. — Occurrence of Dinophysis uraainlha Stein. Large, .solid circles indicate records from vertical hauls; squares, records from surface hauls; small, solid circles, stations at which this species was not found; small, open circles, stations from which no plankton catches were examined. 4728, 4730, 4732, 4734, 4739, 4740) are in the South Equatorial Drift. There are two records from the surface (Stations 4588 [Salpa], 4592); at both these stations the species was taken in sm-face catches only. At one station (4647) the species was found only in a catch from 800-0 fathoms; and at one station (4724) in a catch from 300-0 fathoms as well as in one from 800-0 fathoms. All the remaining records refer to hauls from 300-0 fathoms only. The temperature range of these twenty-nine stations at the surface was 66°-85°; the average was 75.7°. At the two surface stations the surface tempera- ture was 82° and 84°, respectively. At eight stations (4590, 4592, 4638, 4648, 4671, 4679, 4722, 4740) the fre- quency is 1%; at the remaining ones it is less. 280 THE DINOPHYSOIUAE. The specimens figured by Stein (1883) were "aus der Siidsee." Murray and Whitting (1899) recorded this species from a great number of localities in the Atlantic, between lat. 18° N. and lat. 43° N., and in the Caribbean Sea; their northernmost localities are situated in the Gulf Stream. Zacharias (1906) found it m the Atlantic, at lat. 12° N., long. 28° W. Hensen (1911) recorded it at several stations in different regions of the Atlantic, viz., in the Irminger Sea, Gulf Stream, Florida Current, Sargasso Sea, Canary Current, North Equatorial Cur- rent, Guinea Current, South Equatorial Current, and off the coast of Brazil. Jorgensen (1923) reported it from the west coast of Norway (at Bergen), the Bay of Biscay, the west coast of Portugal, the Bay of Cadiz, the Gulf of Guinea, and the southern Atlantic. In the Mediterranean this species has been found by the following investigators, viz.: at lat. 36° N., long. 5° W., by Cleve (1903b); in the Gulf of Lyons, by Pavillard (1916) ; and at several stations in different regions of this sea, by Jorgensen (1923). In the .\rabian Sea it has been recorded by Osten- feld and Schmidt (1901) and by Schroder (1906a); in the Indian Ocean by Schroder (1906a) and by Karsten (1907); Karsten found it at three localities between lat. 10° S. and lat. 34° S. Karsten (1907) reported this species as taken with a closing net between 80 and 60 meters. Most of the remaining records refer to surface hauls. According to the mvestigators who have contributed to our knowledge of the distribution of this species, it occurs in waters of the following temperatures and salinities. Murray and Wliitting (1899) : — temperature, 54°-84°. Cleve (1903b): — temperature, 66°; salinity, 36.45. Of the authors mentioned in this section only Stein (1883) and Jorgen.sen (1923) give descriptions or figures by means of which their determinations of this species may be checked. With regard to Jorgensen's (1923) determinations, see the section on synonymy. This is a eupelagic species, widely distributed in tropical, subtropical, and warm-temperate seas. It has been found as far to the north as in the Irminger Sea and on the west coast of Norway (at Bergen), but is probably a warm-water species, and the specimens found in these relati\'ely cold regions were presumably of southern origin, carried to the north by the Gulf Stream. This .supposition is borne out by the fact that this species has never been recorded by the Inter- national Council for the Investigation of the Sea. According to our records, it is almost uniformly distributed throughout the area investigated by the Expedition and has its optimum habitat in deeper waters, within the levels of photosynthesis. Its absence from the California Current may be worth emphasis. SYSTEMATIC ACCOUNT. 281 DiNOPHYSIS URC'EOLUS, Sp. nov. Figure 37:1 Diagnosis: — Body rounded subtrapeziform to pitcher-shaped in lateral out- line, deepest in or just behind the middle, and 1.13 times longer than deep; longi- tudinal axis deflected posterodorsally at 10°. Cingular lists inclined anteriorly at only 10°-20°, subequal, about twice as wide as transverse furrow. Left sulcal list 0.73 the length of body; distance between Ri and R3 0.58 the length of body; R2 is 0.21, R3 is 0.51, the greatest depth of body ; margin forms angle of 25°-30° at R3; posteroventral portion of list recurved; R3 inclined posteriorly at 50°-55°. With triangular posterior sail somewhat dorsally to midline; its length 0.44 the greatest depth of body and 1.6 the basal width; with marginal ribs but without central rib. Distance between posterior sail and left sulcal list 0.50 the greatest depth of body. Length, 48.5 m- Eastern tropical Pacific. Description: — ■ This is a small species, the body of which is rounded sub- trapeziform to pitcher-shaped in lateral outlme, and deepest in or just behind the middle. The ratio between the length and the depth of the body is 1.13: 1. The longitudinal axis is deflected posterodorsally at about 10°. The epitheca is about 0.53 as deep as the hypotheca, gently concave, highest dorsally, and not visible above the anterior cingular list. The transverse furrow is gently concave and somewhat wider than the greatest height of the epitheca. The posterior cingular list is 0.14 the length of the body from the apex. The hypotheca is asymmetrical. The dorsal margin (from the girdle to the posterior sail) is subuniformly and moderately convex. The ventral margin (from the girdle to the posterior end of the left sulcal list) is subuniformly and gently con- vex. The postmargin is but shghtly convex, inclmed dorsoposteriorly at an angle of 15°-20° to a plane parallel to the gudle. The cingular lists are subequal, about twice as wide as the transverse furrow and 0.25 the greatest depth of the body, and have an anterior inclination of only 10°-20°; their structure is unknown. The right sulcal list extends slightly beyond the fission rib of the left sulcal list, is subtriangular, decreasing gradually m width posteriorly; anteriorly it is about as wide as the transverse furrow. The left sulcal hst is about 0.73 the length of the body; and the distance between the anterior and posterior main ribs is about 0.58 the length of body. The anterior main rib is about 0.21, the fission rib 0.21, and the posterior main rib 0.51 the greatest depth of the body. Behind the posterior main rib this list decreases suddenly in 282 THE DINOPHYSOIDAE. width. Between the anterior and posterior main ribs the free margin of this hst is gently undulating, sHghtly convex anteriorly and posteriorly and slightly concave in the middle. At the posterior main rib this margin forms an angle of about 25°-30°; behind this rib it is moderately concave. The posteroventral portion of this list is recurved. The main ribs are of moderate strength and not club-shaped or otherwise modified; the anterior two are straight, and the posterior has a moderate posterior concavity. The distance between the anterior main rib and the fission rib is about 0.33 the distance between the anterior and posterior main ribs. The posterior main rib has a posterior inclination of 50°-55°. We do not know whether or not this list has any structural differentiation besides the main ribs. On the right valve there is an acute, triangular posterior sail, which is situated somewhat dorsally to the midline and directed posteriorly and slightly Figure 37. — 1, Dinophysis urceolus, sp. nov., right lateral view of type specimen. X 430. Station 46S1 (300-0 fathoms). 2, 3, LHtwphysis monacanlha, sp. nov., right lateral view. 2, from type specimen. X 430. 2, from Station 4730 (300-0 fathoms) ; 3, from Station 4691 (300-0 fathoms). dorsally. The left sulcal list ends at a distance about 0.50 the greatest depth of the body from this sail. When the posterior sail is fully developed, its length is about 0.44 the greatest depth of the body; i.e., it is somewhat shorter than the posterior main rib of the left sulcal list; and it is about 1.6 times longer than it is wide at the base. Its margins are almost straight, and the angle at the tip is about 25°-30°. It is supported by marginal ribs and lacks a central rib. No parasagittal lists are developed. Megacytic stages have not been recorded. The dimensions of the type specimen only were measured. Dimensions: — Length of body, 48.5 ii. Greatest depth of body, 43.0 n. Comparisons: — Dinophysis urceolus is established on a single outline draw- ing of a specimen in lateral view found in the material of the Expedition. The shape of the body in dorsoventral view, the struct\ire of the cingular lists and of SYSTEMATIC ACCOUNT. 283 the theca are unknown, and our knowledge of the structiire of the left sulcal list is incomplete. The species forms structurally a connectmg link between Dinophysis ura- cantha and D. collaris. It resembles the former species in the small size of the body, in the posterodorsal inclination of the longitudinal axis of the body, in the shape of the left sulcal list, and in the size, structure, and position of the posterior sail. Furthermore, it approaches this species in the anterior inclination of the cingidar lists. It recalls D. collaris very strikingly in the pitcher-like shape of the body in lateral view. It is easily distinguished from D. uracantha by the peculiar pitcher-like shape of the body in lateral view and from D. collaris in having only one posterior sail. Dinophysis monacantha resembles D. urceolus in the development of the sulcal lists and of the posterior sail, and in the shape of the ventral and posterior margins of the hypotheca. The latter species is readily distinguished from the former by the smaller size of the body, by the sigmoid shape of the dorsal margin of the body, by the absence of parasagittal lists, and by the fact that the posterior sail is situated somewhat nearer to the dorsal side of the body and is inclmed posterodorsally instead of posteroventrally. Occurrence: — Dinophysis urceolus is recorded at only one (4681, the type locality) of the 127 stations, on the third line of the E.xpedition, in the South Equatorial Drift, from 300-0 fathoms, at a surface temperature of 68°. The frequency is less than 1 % (one specimen) . Dinophysis monacantha, sp. nov. Figure 37: 2, 3 Diagnosis: — Body rounded subtrapeziform in lateral outUne, deepest in or just behind the middle, and 1.19-1.22 times longer than deep; longitudmal axis perpendicular to girdle or deflected posterodorsally at l°-3°. Cingular lists sub- horizontal, subequal, 2.0-2.5 times wider than transverse furrow. Left sulcal list 0.64-0.69 the length of body; distance between Ri and R3 0.50-0.51 the length of body; R2 is 0.24-0.29, R3 is 0.45-0.50 the greatest depth of body; margin forms angle of 30° at R3; postero ventral portion of list recurved; R3 inclined posteriorly at 30°-40°. With triangular posterior sail located on midline or but slightly dis- placed dorsally or ventrally; its length is 0.45-0.52 the greatest depth of body and 1.6-2.0 its basal width; with margmal ribs but without central rib. Distance be- tween posterior sail and left sulcal list 0.35-0.37 the greatest depth of body. Along dorsal margin a narrow parasagittal list of subuniform \^'idth, extending 284 THE DINOPHYSOIDAE. from posterior cingular list to posterior sail; its average width about half the width of transverse furrow. Length, 68.2-69.7 m- Eastern tropical Pacific. Descripiion: — This is a medium-sized species, the body of which is sub- trapeziform in lateral outline, deej^cst in or just behind the middle. The ratio between the length and the depth of the body is 1.19-1.22: 1. The longitudinal axis is either perpendicular to the gii'dle or deflected posterodorsally at about l°-3°. The epitheca is 0.64-0.72 as deep as the hypotheca, gently convex, flat or even slightly concave, highest in the middle or dorsally, and visible above the anterior cingular list. The transverse furrow is flat or gently convex, and some- what wider than the greatest height of the epitheca. The posterior cingular list is 0.13-0.16 the length of the body from the apex. The hypotheca is asymmetrical. The dorsal margin (from the girdle to the posterior sail) is subuniformly and gently to moderately convex. The ventral margin (from the girdle to the posterior end of the left sulcal list) is subuniforndy and moderately convex. The post- margin is gently convex or flattened, inclined dorsoposteriorly, forming an angle of 25°-30° to a plane parallel to the girdle, confluent with the ventral margin, and, in some specimens, forms with the dorsal margin a distinct Ijut well-rounded corner. The cingular lists are subhorizontal and subequal, and 2.0-2.5 times wider than the transverse furrow; their structure is unknown. The right sulcal list ex- tends to or slightly beyond the fission rib of the left sulcal list, and is subtriangu- lar, decreasing gradually in width posteriorly; anteriorly it is about as wide as the transverse furrow. The left sulcal list is 0.64-0.69 the length of the body, and the distance between the anterior and posterior main ribs is about 0.50-0.51 the length of body. The anterior main rib is 0.20-0.21, the fission rib 0.26 (0.24-0.29), and the posterior main rib 0.47 (0.45-0.50) the greatest depth of the body. Be- hind the posterior main rib this list decreases suddenly in width. Between the anterior and posterior main ribs the free margin of this list is gently and evenly convex, or it is somewhat concave anteriorly ; at the posterior main rib it forms an angle of about 30°, and behind this rib it is gently concave or sigmoid. The posteroventral portion of this list is somewhat recurved. The mam ribs are of moderate strength, and not club-shaped or otherwise modified ; the two anterior are straight or almost so, and the posterior one has a moderate posterior con- cavity. The distance between the anterior main rib and the fission rib is 0.38-0.40 the distance between the anterior and posterior main ribs. The posterior mam SYSTEMATIC ACCOUNT. 285 rib has a posterior inclination of 30°-40°. We do not know whether or not this hst has any structural differentiation besides the main ribs. On the right valve there is an acute, triangular posterior sail, which is located on the midline or but slightly displaced dorsally or ventrally and directed somewhat posteroventrally. The left sulcal list ends at a distance from this sail eciualing about 0.35-0.37 the greatest depth of the body. When the posterior sail is fully developed, its length is 0.45- 0.52 the greatest depth of the body; i.e., it has about the same length as the poste- rior main rib of the left sulcal list; and it is 1.6-2.0 times longer than its width at the base. Its margins are almost straight, and the angle at the tip is 25°-30°. It is supported by marginal ribs and lacks a central rib. Along the dorsal margin of the body there is a narrow parasagittal list of subuniform width, extending from the posterior cingular list to the posterior sail; its average width is about half the width of the transverse furrow, and at least m some specimens (Figure 37 : 3) it has a fairly great number of cross-ribs. Megacytic stages have not been recorded. The dimensions of two specmiens were measured. Dimensions: — Length of body, 68.2-69.7 n (type, 68.2 ju). Greatest depth of body, 57.0-57.2 ^ (type, 57.2 n). Variations: — Judging by the two specimens exammed, this species is rela- tively constant. Comparisons: — Dinophysis monacantha is established on outline drawings of two specunens in lateral view, found in the material of the Expedition. The shape of the body in dorsoventral view, the structure of the cingular lists and of the theca are unknown, and the structure of the left sulcal list needs further examination. The species combines characteristics of Dinophysis hasiata and D. uracantha. In the elongation and relatively large size of the body, and in the posteroventral inclination of the posterior sail, it approaches D. hastata; and in the tendency to develop a posterodorsal rather than a posteroventral mclination of the longi- tudinal axis of the body and in the structure of the posterior sail, it recalls D. uracantha. In the position of the posterior sail, it is intermediate between these two species. It differs from both these species in having a parasagittal list along the dorsal margin of the hypotheca. It resembles Dinophysis alata in having a posterior sail and in having a para- sagittal list along the dorsal margin of the hypotheca. It is easily distinguished from this species by the more irregular shape of the body, bj' the subhorizontal position of the cingular lists, by the narrowness of the parasagittal list, and by the 286 THE DINOPHYSOIDAE. structure of the posterior sail. With regard to the relationships between D. monacanthu, and D. urceolus and D. trapezium, see the last two species, the sections on comparisons. Occurrence: — Dinophysis monacantha is recorded at two of the 127 stations. Of these two stations, one (4C91) is on the third line of the Expedition and in the Easter Island Eddy; the other (4730), the type locality, is on the fifth line and in the South Equatorial Drift. The depth is 300-0 fathoms, the surface temperature 73°-79°, and the frequency less than 1% (one specimen at each station). Dinophysis trapezium, sp. nov. Figure 38:2, 3 Diagnosis: — Body subtrapeziform in lateral outline, protracted postero- dorsally, deepest in or just behind the middle, and 1.23-1.31 times longer than deep; longitudinal axis deflected posterodorsally at about 15°. Cingular lists sub- horizontal, subequal, and 1.5-2.5 times wider than transverse furrow. Epitheca visible above anterior cingular list. Left sulcal list: distance between Rj and R3 about 0.50 the length of body; R2 is 0.17-0.21, R3 is 0.56-0.70 the greatest depth of body; margin forms angle of 35°-55° at R3; R3 inclined posteriorly at 15°-25°; reticulated; behind R3 this list has an acute, wedge-shaped accessory lobe, which has a height 0.28-0.32 the greatest depth of body and an angle at the Vip of 30°- 40°. With triangular posterior sail somewhat dorsally to midline; length is 0.43- 0.50 the greatest depth of body; angle at tip 25°-35°; with marginal ribs; some- tmies connected with left sulcal list by very narrow list, sometimes unconnected. Length, 65.2-67.5 m- Eastern tropical Pacific. Descriplior}: — This is a medium-sized species, the body of which is sub- trapeziform in lateral outline, protracted posterodorsally, and deepest in or just behmd the middle. The ratio between the length and the depth of the body is 1 .27 (1 .23-1 .31 ) : 1 . The longitudinal axis is deflected posterodorsally at about 15°. The epitheca is deep, 0.70-0.72 as deep as hypotheca, flat to slightly convex, highest in the middle or ventrally, and visible above the anterior cingular list. The transverse furrow is flat or gently convex, and about as wide as or somewhat narrower than the greatest height of the epitheca. The posterior cmgular list is 0.19-0.22 the length of the body from the apex. The hypotheca is asymmetrical. The dorsal margin (from the girdle to the posterodorsal sail) is almost straight or subuniformly and gently convex. The \-entral margin (from the gii'dle to a point somewhat behind the posterior main rib of the left sulcal Ust) is gently to SYSTEMATIC ACCOUNT. 287 moderately convex. The postmargin is gently to moderately convex, inclined dorsoposteriorly at an angle of 25°-30° to a plane parallel to the girdle, confluent with the ventral margin, and forms with the dorsal margin a distinct but well- rounded corner. The cingular lists are subhorizontal and subequal, and 1.5-2.5 times wider than the transverse furroAv; their structure is unknown. The right sulcal list extends to a point somewhat beyond the fission rib of the left sulcal list; it is either subtriangular, decreasing gradually in width posteriorly, or of subuniform width throughout its anterior half and gently rounded posteriorly; anteriorly it is about as wide as the transverse furrow. The left sulcal list is relatively long and strikingly wider posteriorly than anteriorly. The distance between the anterior and posterior main ribs is about 0.50 the length of the body. The anterior main FiGDHE 38. — 1, Dinophysis collaris Kofoid and Michener, right lateral view. X 430. Sta- tion 4711 (300-0 fathoms). 2, 3, Dinophysis trapezium, sp. nov., right lateral view. 2, from type specimen. X 430. Station 4709 (300-0 fathoms). rib is 0.22 (0.20-0.25), the fission rib 0.19 (0.17-0.21), and the posterior main rib 0.63 (0.56-0.70) the greatest depth of the body. Behind the posterior main rib the list decreases suddenly in width. The free margm of this list is almost straight (Figure 38: 2) or slightly irregular (Figure 38: 3) between the anterior and poste- rior main ribs; at the posterior main rib it forms an angle of 35°-55°; and behind this rib it first is gently concave and then forms an acute, wedge-shaped accessory lobe. This lobe has a height 0.28-0.32 the greatest depth of the body; and the angle at its tip is 30°-40°. The main ribs of this list are of moderate strength and not club-shaped; or the posterior may be slightly club-shaped (Figm-e 38:2); the anterior two are straight or almost so, the posterior is slightly to moderately concave posteriorly. The distance between the anterior main rib and the fission rib is 0.30-0.43 the distance between the anterior and posterior main ribs. The posterior main rib has a posterior inclination of 15°-25°. Besides the main ribs this list has many cross-ribs anastomosmg into an ii-regular reticulum. On the 288 THE DINOPHYSOIDAE. right valve there is an acute, triangular posterior sail, which is located somewhat dorsally to the midline and has a dorsal inclination of 15°-20° to the midline. When this sail is fully developed, its length is 0.43-0.50 the greatest depth of the body; its margins are almost straight, and the angle at its tip is 25°-35°; it is supported by marginal ribs ; sometimes it is connected with the left sulcal list by a very narrow list (Figure 38: 2), and sometimes it is unconnected. As far as we know, there are no parasagittal lists along the dorsal margin of the hypotheca and on the epitheca. The structure of the theca is unknown. Megacytic specimens have not been recorded. The dimensions of two specimens were measured. Dimensions: — Length of body, 65.2-67.5 m (average, 66.4 ix; type, 67.5 n). Greatest depth of body, 51.5-52.8 m (type, 51.5 ^). Variations: — The two specimens examined are rather similar. They differ mainly in the shape of the hypotheca viewed laterally, in the relative length of the posterior main rib of the left sulcal list, and in regard to the connection be- tween the posterior sail and the left sulcal list. Comparisons: — Dinophysis trapezium is established on two drawings repre- senting the lateral outlines of two specimens from the material of the Expedition. The shape of the body in dorsoventral view, as well as the structure of the cingu- lar hsts and of the thecal wall are unknown. The structure of the left sulcal list and of the posterior sail needs to be further examined. The species is on the border line between the genera Phalacroma and Dino- physis, and its present generic allocation should be regarded as tentative. Its structurally closest-known relative is Dinophysis collaris (see this species, the section on comparisons). Next to D. collaris, D. monacantha appears to be its nearest relative. It shows a striking resemblance to the last species in the shape of the body in lateral view, in the position of the cingular lists, and in the size, shape, structure, and position of the posterior sail. D. trapezium is easily dis- tinguished from D. monacantha in having a parasagittal list along the dorsal margin of the hypotheca and in lacking the accessory lobe behind the posterior main rib of the left sulcal list. Another species which probably is rather closely related to Dinophysis trapezium is D. swezyi. This relationship is indicated bj' the similar shape of the body in lateral view (compare Figure 38 : 2, 3, and Figure 39 : 2, 3) and by the fact that in Ixith these species the left sulcal list has an accessory lobe behind the posterior main rib. D. sicezyi, which is a typical representative of its genus, is easily recognized by the anterior inclination of its cingular list, by the shape of its SYSTEMATIC ACCOUNT. 289 left sulcal list, and by the fact that its posterior sail is supported by a central rib and not by marginal ribs. Occurrence: — Dinophysis trapezium is recorded at only one (4709, the type locality) of the 127 stations, on the fourth line of the Expedition, in the South Equatorial Drift, from 300-0 fathoms, at a surface temperature of 72°. The fre- quency is less than 1% (two specimens). Dinophysis swezyi, sp. nov. Plate 5, fig. 9. Figure 39 Diagnosis: — Body broadly ovate to rounded subtrapeziform in lateral out- Une, deepest somewhat behind the middle, 1.12-1.31 times longer than deep; longitudinal axis perpendicular to girdle or deflected posterodoi sally at 1°-10°. Cingular lists ribbed, the anterior 2.0-4.0 times wider than transverse furrow. Left sulcal list 0.74-0.83 the length of body; distance between Ri and Rs 0.55- 0.64 the length of body; R. is 0.41-0.71, Rj is 0.73-1.07 the greatest depth of body; margin often strikingly concave just behind R2 and forms angle of 25°-35° at R3; R3 inclmed posteriorly at 25°-45° and sometimes club-shaped; behind R3 this list forms a rounded to subacute accessory lobe, the height of which is 0.25- 0.41 the greatest depth of body; between Rj and R3 there are a few short basal ribs, and the accessory lobe supported by similar ribs, one of which extends to tip of lobe. With triangular posterior sail, located dorsally to midline, and sup- ported by central rib; its length is 0.69-0.83 the greatest depth of body and 1.9- 2.7 its basal width; separated from left sulcal list. Theca reticulate; 11-12 poly- gons border girdle posteriorly. Length, 47.2-56.4 n. Eastern tropical, subtropical, and warm-temperate Pacific. Description: — This is a medium-sized species, the body of which is broadly ovate to rounded subtrapeziform in lateral outline, and deepest somewhat behind the middle. The ratio between the length and the depth of the body is 1.23 (1.12- 1.31): 1. The longitudinal axis is either perpendicular to the girdle or deflected posterodorsally at 8° (1°-10°). The epitheca is 0.55 (0.50-0.61) as deep as the hypotheca, of moderate to gentle convexity, highest in or somewhat dorsally to the center, and not visible above the anterior cingular list. The transverse furrow is flat or gently convex, and about as wide as or somewhat wider than the greatest height of the epitheca. The posterior cingular list is 0.15 (0.11-0.17) the length of the body from the apex. The hypotheca is very variable in shape. Sometimes (Figure 39: 1) it is symmetrical and broadly ovate; its dorsal and \entral margins are of moderate 290 THE DINOPHYSOIDAE. and subuniform convexity or either of them is somewhat flattened anteriorly; its postmargin is broadly and subuniformly convex and confluent with the dorsal and ventral margins. In most specimens, however, it is more or less strikingly asymmetrical; its dorsal margin (from the gii'dle to the posterior sail) is sub- uniformly and slightly to gently convex, or (Figure 39: 2) it is more or less bulging posteriorly; its ventral margin (from the girdle to the posterior main rib of the left sulcal list) is gently convex; its postmargin is gently to moderately convex, inclined dorsoposteriorly, forming an angle of 25°-35° to a plane parallel to the girdle, confluent with the ventral margin, and forms with the dorsal margin a fairly distinct but well-rounded corner. The anterior cingular list is 2.8 (2.0-4.0) tunes wider than the transverse furrow and 0.30-0.33 the greatest depth of the body; it is inclined anteriorly at 20°-30° and has, on each valve, about 13-14 simple, straight, and almost equi- distant ribs. The posterior cingular list resembles the anterior but in some speci- mens it is somewhat narrower or somewhat less inclined anteriorly. The right sulcal list is rather variable ; it extends to or somewhat beyond the fission rib of the left sulcal list; sometmies (Figure 39: 3) it is subtriangular, decreasing gradu- ally in width posteriorly; sometimes it Ls of subuniform width throughout the greater portion of its length and rounded to angular (Plate 5, fig. 9) posteriorly; anteriorly it is about as wide as or somewhat wider than the transverse furrow. The left sulcal list is large. Its length is 0.79 (0.74-0.83) the length of the body, and the distance between the anterior and posterior main ribs is 0.59 (0.55-0.64) the length of the body. The anterior main rib is 0.25 (0.21-0.29), the fission rib 0.59 (0.41-0.71), and the posterior main rib 0.89 (0.73-1.07) the greatest depth of the body. Behind the posterior main rib this list decreases suddenly in width. Between the anterior main rib and the fission rib the free margin of this list is almost straight or slightly concave, convex, or sigmoid; between the fission rib and the posterior main rib it is conca^'e, and the concavity usually is very striking just behind the fission rib. At the posterior main rib this margin forms an angle of 28° (25°-35°), and behind this ril) it first is gently concave and then forms a more or less narrowly rounded to subacute accessory lobe, the height of which is 0.32 (0.25-0.41) the greatest depth of the body. The main ribs are of moderate strength, and the fission rib and the posterior main rib sometimes are club-shaped. The anterior main rib is straight or nearly so ; the fission rib and the posterior main rib usually are gently concave posteriorlj-. The distance between the anterior main rib and the fission ril) is 0.48 (0.45-0.50) the distance between the anterior and posterior main ribs. The posterior main rib has a posterior inclination of 36° SYSTEMATIC ACCOUNT. 291 (25°-45°). Between the fission rib and the posterior main rib there are a few (about three) short ribs, and the posterior accessory lobe is supported by similar ribs, one of which extends to the tip of the lobe (Plate 5, fig. 9). On the right valve there is an acute posterior sail, situated somewhat dorsally to the midline, and supported by a central rib, which is gently concave ventrally and which may be club-shaped (Figure 39: 2) or split lengthwise (Plate 5, fig. 9). WTien this sail is fully developed, its length is 0.75 (0.69-0.83) the greatest depth of the body and 2.3 (1.9-2.7.) its basal width; except for the central rib it lacks structural differ- entiation, and it is separated from the left sulcal list by a distance equaling 0.34 (0.27-0.42) the greatest depth of the body. There are no parasagittal Usts along the dorsal margin of the hypotheca and on the epitheca. The thecal wall is reticulate and porulate. About 11-12 polygons border the girdle posteriorly. In the girdle there are two rows of meshes. Figure 39. — Dinophysis sii^ezyi, sp. nov., right lateral view. X 430. 1, from Station 4679 (300-0 fathomsj; 2, from Station 4638 (300-0 fathoms); 3, from Station 4732 (300-0 fathoms). Megacytic stages have not been reported. The dimensions of three specimens from the material of the Expedition and of the type specunen were measured. This species is named for Dr. Olive Swezy, the collaborator with the senior author in the preparation of "The Free-living Unarmored Dmoflagellata," who during her investigations on the dinoflagellates of the San Diego region made our drawing of the type specimen (Plate 5, fig. 9) as well as sketches of several other specimens of this species. Dimensions: — Length of body, 47.2-56.4 n (average, 52.1 m; type, 55.0 yu). Greatest depth of body, 39.2-44.8 m (average, 42.3 /^; type, 44.8 ju). Variations: — This species is rather constant except in the shape of the body in lateral view and m the shape of the free margin of the left sulcal list between the fission rib and the posterior mam rib, in which respects it is strikingly variable. The body usually is more or less asymmetrical, rounded subtrapeziform, and pro- 292 THE DINOPHYSOIDAE. tracted posterodorsally, but sometimes (Figure 39: 1) it is symmetrical and broadly ovate. The margin of the left sulcal list Between the fission rib and the posterior main rib usually is deeply concave just behmd the fission rib, but some- times (Figure 39: 1) it is gently and subuniformly concave. Other characters ex- hibiting variations are : the shape and size of the right sulcal list, the shape and relative length of the fission rib and of the posterior mam rib of the left sulcal list and of the central rib of the posterior sail ; and the shape of the accessory lobe of the left sulcal list. The specimen represented by Figure 39: 1, may be specifically different from the other specimens assigned to this species. Comparisons: — This species is rather closely related to Dinophysis trapezium and D. schutti. Its relationship to D. trapezium is indicated by the shape of the body in lateral view (compare Figure 39: 2, 3, and Figure 38 : 2, 3) and by the fact that in both these species the left sulcal list has an accessory lobe behind the posterior main rib. Its relationship to D. schutti is suggested by the shape of the body in lateral view (compare Figure 39: 2, 3, and Figure 40: 3, 5), by the shape of the ventral margin of the left sulcal list, and by the shape, structure, and rela- tive size of the posterior sail. D. swezyi is easily distinguished from D. trapezium by the shape of the ventral margin of its left sulcal list, by the anterior inclination of its cingular lists, and by its posterior sail being supported by a central rib and not by marginal ribs. It differs from D. schutti especially in having the left sulcal list with an accessory lobe behind the posterior main rib. Occurrence: — Dinophysis swezyi is recorded at four of the 127 stations. There are 0, 1, 1, 0, 2, and 0 stations on the six lines of the Expedition. Of these four sta- tions, one (4638) is in the Panamic Area; and three (4679, 4732, 4737) are in the South Equatorial Drift. At one of these stations (4737) the species was taken in a haul from 100-0 fathoms. All the other records refer to hauls from 300-0 fathoms. The temperature range of these four stations at the surface was 69.0°-81.5°; the average was 76.1°. The frequency is in every case less than 1 %. The type specimen (Plate 5, fig. 9) was taken at San Diego, California, where this species has been found on several occasions. Dinophysis collaris Kofoid and Michener Plate 5, fig. 2, 8. Figure 38: 1 Dinophysis collaris KoroiD & MicheNek, 1911, j). 292. Diagnosis: — ^ Body rounded trapeziform in lateral outline, deepest in the middle, 1.05-1.07 times longer than dee]); longitudinal axis deflected postero- SYSTEMATIC ACCOUNT. 293 dorsally at 10°-15°. In dorsal view about three times longer than wide, widest at girdle, with almost straight side contours and rounded apices. Cingular lists sub- horizontal, subequal, 1.5-2.5 times wider than transverse furrow, and ribl;ed. Epitheca visible above anterior cingular list. Left sulcal list: distance between Ri and R3, 0.55-0.59 the length of body; R. is 0.21-0.22, R3 is 0.39-0.43 the great- est depth of body; margin forms angle of 35°-65° at R3; R3 mclined posteriorly at 30°-40° ; reticulated. Two parasagittal lists encircle hypotheca. The left is of sub- uniform width throughout, 0.5 as wide as transverse furrow. The right continues left sulcal list and forms two subequal, acute, wedge-shaped posterior sails, one on each side of midline, both with irregadarly anastomosing ribs; length of sails somewhat less than R3; angles at their tips 30°-50°; width of list between sails and between ventral sail and R3 subequal to w idth of transverse furrow or less. Theca reticulate; 25-30 polygons border girdle posteriorly. Length, 58.0-67.0 n. Eastern tropical and subtropical Pacific. Description: — This is a medium-sized species, the body of which is rounded trapeziform in lateral outline, and deepest near the middle. The ratio between the length and the depth of the body is 1.05-1.07: 1. The longitudinal axis is de- flected posterodorsally at 10°-15°. The epitheca is very deep, 0.73-0.74 as deep as the hypotheca, flat or even gently concave, highest dorsally, and visible above the anterior cingular list. The transverse furrow is flat or gently concave, and about as wide as or slightly wider than the greatest height of the epitheca. The posterior cingular list is 0.17-0.21 the length of the body from the apex. The hypotheca is asymmetrical. The dorsal margin is sigmoid, being shghtly to moderately concave anteriorly, gently to moderately convex in the middle, and more or less flattened posteriorly. The ventral margin (from the girdle to the posterior end of the left sulcal list) is sub- uniformly, and gently to moderately convex. The postmargin is gently to moder- ately con\'ex and is inclined dorsoposteriorly, forming an angle of about 10°-25° to a plane parallel to the girdle. In dorsoventral view the body is bilaterally compressed, widest at the girdle, about three times longer than wide, with almost straight side contours, and with rounded apices. The anterior cingular list is 1.5-2.5 times wider than the transverse furrow and 0.17-0.22 the greatest depth of the body; usually it is subhorizontal, but sometimes it has an anterior inclmation of about 10°-20°; on each valve it has 15-20 simple, straight, and almost equidistant ribs. The posterior cingular list may be somewhat narrower than but usually it is about as wide as the anterior; it is subhorizontal and has about the same ribbing as the anterior. The sulcus is 294 THE DINOPHYSOIDAE. about 0.40-0.45 as long as the hypotheca. The flagellar pore is located at a dis- tance from the girdle subequal to the width of the transverse furrow. The right sulcal list extends to a point somewhat beyond the fission rib of the left sulcal list, and is subtriangular, decreasing gradually in width posteriorly; anteriorly it is about as wide as the transverse furrow. The left sulcal list is of moderate width and rather long. The distance between the anterior and posterior main ribs is 0.55-0.59 the length of the body. The anterior main rib is 0.19-0.24, the fission rib 0.21-0.22, and the posterior main rib 0.39-0.43 the greatest depth of the body. Behind the posterior main rib the list decreases suddenly in width. Between the anterior and posterior main ribs the free margin of this list is almost straight, or gently to moderately convex; the convexity is either subuniform or somewhat irregular (Plate 5, fig. 2). At the posterior main rib this margin forms an angle of 35°-65°; behind this rib it is gently concave. The main ribs of this list are of moderate strength and not club-shaped; the anterior two are straight or almost so, the posterior, which sometimes is branched (Plate 5, fig. 2), is slightly to moderately concave posteriorly. The distance between the anterior main rib and the fission rib is 0.30-0.33 the distance between the anterior and posterior main ribs. The posterior main rib has a posterior inclination of 30°-40°. Besides the main ribs this list has many cross-ribs anastomosing into a \ery irregular reticu- lum. The epitheca has no accessory lists. The hypotheca, on the other hand, has two parasagittal lists, which sometimes are limited to the postmargin (Figure 38: 1), but usually continue to the posterior cingular list. The left one of these lists is of subuniform width throughout, and its average width is about half the width of the transverse furrow; sometunes it has no distinct structiu'al differentia- tion, sometimes it has a few cross-ribs. The right, which forms a direct continua- tion of the left sulcal list, resembles the left but is characterized by two subeciual, acute, wedge-shaped posterior sails, one on either side of the midline, and both furnished with irregularly anastomosing ribs. The length of the dorsal sail is 0.31 (0.25-0.37), that of the ventral sail 0.31 (0.27-0.35) the greatest depth of the body; in other words, these sails are somewhat shorter than the posterior main rib of the left sulcal list. The angles at their tips are about 30°-50°. The width of the right parasagittal list, between the left sulcal Ust and the ventral sail and between the two sails, is subequal to the width of the trans^"crse fmTow or less. The surface of the theca is uniformly covered with small, irregular polygons of heavy mesh. On each valve about 25-30 polygons border the girdle posteriorly. In the transverse furrow there are two rows of polygons, each row with 15-25 polygons on each valve. SYSTEMATIC ACCOUNT. 295 Megacytic specimens have not been recorded. The dimensions of two of our specimens were measured. Dimensions: — Length of body, 58.0-67.0 ^ (average, 62.5 m; type, 67.0 /j, not 58 ix, as stated in the original description). Greatest depth of body, 55.4- 62.5 M (average, 58.9 n; type, 62.5 m, not 55 m, as stated in the original description). Variations: — Judging by the few specimens examined, this species appears to be rather constant. The most variable character is the extension of the para- sagittal lists. Sometimes these lists are limited to the postmargin of the body, sometimes they continue along the whole dorsal margin of the hypotheca to the posterior cingular list. These variations presumably are related to skeletal re- organization consequent upon binary fission. Comparisons: — Our description and figures of Dinophysis collaris are based on the type material. The generic assignment of this species is uncertain, and its present allocation should be regarded as tentatiAe. The great relative depth of the epitheca, the subhorizontal position of the cingular lists, the ribbing of the posterior cingular list, the presence of parasagittal lists on the dorsal side of the hypotheca, and the reticulation of the thecal wall affiliate it with Phalacroma. On the other hand, the posterodorsal inclination of the longitudinal axis of the body, the shortness of the epitheca, the shape of the hypotheca in lateral view, and the great width of the cingular lists suggest relationship to Dinophysis. Two posterior sails are to be found in Dinophysis as well as in Phalacroma. The species shows that the separation of these two genera is almost arbitrary. The structm'ally closest-known relative of Dinophysis collaris is D. trapezium. The former species differs from the latter mainly in the less-pronounced trapezi- form shape of its body, in havmg its body less protracted posterodorsally, in the shorter posterior main rib of its left sulcal list, in the subequal size of its two pos- terior sails, and in having these sails supported by irregularly anastomosing ribs (see also D. trapezium, the section on comparisons). Another close relative of Dinophysis collaris is D. urceolus. This relationship is indicated by the almost complete agreement in the lateral shape of body in these two species. The latter species, which is a rather typical representative of Dinophysis, is easily distinguished from the former by having only one posterior sail, viz., the one situated dorsally to the midline of the body. Among the species of Phalacroma, P. bipartitum resembles Dinophysis col- laris in the structural differentiation of the parasagittal list of the right valve. Dinophysis collaris is easily distinguished from Phalacroma hipartituyn by the 296 THE DINOPHYSOIDAE. rather irregular shape of its body, by its much lower epitheca, and by its larger and more differentiated cingular and sulcal lists. Occurrence: — This species is recorded at six of the 127 stations. There are 0, 1, 2, 2, 1, and 0 stations on the six lines of the Expedition. Of these sLx stations one (4671) is in the Peruvian Current; one (4713) is in the Galapagos Eddy; and four (4679, 4681, 4711, 4730) are in the South Equatorial Drift. All the records refer to catches from 300-0 fathoms. The temperature range of these six stations at the surface was 66°-79°; the average was 71.7°. The frequency in every case is less than 1 %. The species has been found only in the material of the Expedition. It was first recorded by Kofoid and Michener (1911) at Station 4671 of the Expedition, which thus is the type locality. The prevalence of this species in the Peruvian Current and in the regions directly influenced by this current, as well as its absence from the California Cur- rent, the Mexican Current, and the Panamic Area may be noteworthy. DiNOPHYSis scHUTTi Murray and Wliitting Figure 40, 41 Dinophysis schuetiii MnRRAY & Whitting, 1899, p. 331, tab. 8, pi. 31, fig. 10. Lemmermann, 1899a, p. 373; 1901a, p. 375. Ostenfeld, 1900, p. 56, tab. 2. Cleve, 1901a, p. 15. Kofoid, 1907a, p. 196. Karsten, 1907, p. 421, 473. Paulsen, 1908, p. 12, 18, fig. 19. Nathansohn, 1908, p. 604; 1909, p. 46. Meunier, 1910, p. 58. P.\vill.U!D, 1916, p. 60. Jobgensen, 1923, p. 34, 44, fig. 46. Dinophysis uracanlha Schutt (non Stein), 1895, p. 16, 17, pi. 2, fig. 9. Dinophysis schuettii var. uracanthoidcs Fokti & Issel, 1924, p. 6, fig. 2. Diagnosis: — Body subovate to subellipsoidal m lateral outline, deepest near the middle ; length : depth, 1 .09-1 .23 : 1 ; longitudinal axis deflected posterodorsally at 0°-6°. In dorsal view rather narrowly ellipsoidal, 1.45 times longer than wide. Anterior cingular Ust 2.0-3.0 times wider than transverse furrow. Left sulcal list 0.68-0.84 the length of the body ; distance between Ri and R, 0.58-0.68 the length of body; R2 is 0.85-1.25, R3 is 0.89-1.33 the greatest depth of body; margin strongly concave between R2 and R3 and forms angle of 5°-15° at R3; R3 inclined posteriorly at angle of 30°-55°; R2 and R3 often club-shaped; sometimes slightly reticulated. With triangular posterior sail located somewhat dorsally to midline; its length is 0.81-1.35 the greatest depth of body and 1.7-4.0 its basal width; supported by central rib, which often is club-shaped distally and sometimes bi- furcate or trifurcate proximally; separated from left sulcal list. Thcca finely and closely areolate. Length, 29.6-62.0 m. SYSTEMATIC ACCOUNT. 297 Widely distributed in tropical, subtropical, and warm-temperate seas, sel- dom in colder waters. Description: — This is a small to medium-sized species, the body of which is subovate to subellipsoidal in lateral outline and deepest near the middle. The ratio between the length and the depth of the body is 1.09-1.2.3: 1; in our speci- mens this ratio is 1.21 (1.19-1.23): 1; in the specimens figured by Schiitt (1895), Murray and WTiitting (1899), and Jorgensen (1923) it varies between 1.09: 1 and 1.11:1. The longitudinal axis either is perpendicular to the transverse furrow, as in our Figiu'e 40: 1, 6, or it is deflected posterodorsally at an angle of 2° (1°~()°). The epitheca is 0.30-0.50 as deep as the hypotheca, moderately to rather strongly convex, highest in or just dorsally to the center, and not visible above the anterior cingidar list. The transverse furrow is flat or gently convex, and in most specimens about 2.0-3.0 times wider than the greatest height of the epitheca. The posterior cingular list is 0.15 (0.13-0.17) the length of the bodj' from the apex. The hypotheca sometimes is symmetrical, as in Figure 40: 1, 6, but usually it has a slight posterodorsal inclination (see above). The dorsal margin (from the gu'dle to the antapex) sometmies is subuniformly and moderately to gently convex, and sometimes (Figure 40 : 5) it is flattened anteriorly and boldly convex posteriorly. The ventral margin (from the girdle to the antapex) is subuniformly and moder- ately to rather strongly convex, or it is somewhat flattened posteriorly (Figure 40: 5) ; its curvature either is about the same as that of the dorsal margin, or it is more pronounced, in which case the body appears to be flattened dorsally (Fig- ure 40: 3). The apices sometmies are of about the same depth, but usually the body is distmctly deeper posteriorly than anteriorly. In dorsoventral \ie\\ the body is rather narrowly ellipsoidal, widest in the middle, about 1.45 tunes longer than wide, and with the apices well rounded and of about the same width. The anterior cingidar list is 2.0-3.0 times wider than the trans\erse fm'row and 0.30-0.50 the greatest depth of the body; it has a well-developed funnel- shape, bemg inclmed anteriorly at 20°-40°, and, on each valve, is furnished with about ten to twenty simple, straight, and almost equidistant ribs, some of which may be incomplete. The posterior cingular list may be about as wide as the anterior, but usually it is somewhat narrower; its inclination usually is about the same as that of the anterior, and it appears always to lack structural differentia- tion. The right sulcal list extends to or slightly beyond the fission rib of the left sulcal list, and is subtriangular, decreasmg gradually in width posteriorly, or of subuniform width throughout the greater part of its length and rounded poste- riorly; anteriorly it is about as wide as or somewhat wider than the transverse 298 THE DINOPHYSOIDAE. furrow. The left sulcal list is unusually large, with very long fission rib and posterior main rib. Its length is 0.74 (0.68-0.84) the length of the body, and the distance between the anterior and posterior main ribs is 0.63 (0.58-0.68) the length of the body. The anterior main rib is 0.34 (0.27-0.42), the fission rib 1.04 (0.85-1.25), and the posterior mam rib 1.07 (0.89-1.33) the greatest depth of the body. Behind the posterior main rib this list decreases suddenly in width. The ventral margin of this list sometimes (Figure 40: 2, 4) extends to the tips of the fission rib and the posterior main rib, but usually it terminates at some distance from the tips, and the distal portions of these two ribs are free ; the length of the free portions may even slightly exceed half the length of these ribs. Between the anterior main rib and the fission rib the margin is almost straight or gently con- cave; between the fission rib and the posterior main rib it is strongly concave; at the posterior main rib it forms an angle of 9° (5°-15°), and behind this rib it is gently to moderately concave or almost straight. The main ribs of this list some- times are straight, sometimes the fission rib and the posterior main rib are con- cave posteriorly; the two last ribs frequently are club-shaped distally. The dis- tance between the anterior main rib and the fission rib is 0.43 (0.33-0.50) the distance between the anterior and posterior main ribs. The posterior main rib has a posterior inclination of 47° (30°-55°). Besides the three main ribs this list sometimes has a slight indication of reticulation (Schiitt, 1895, pi. 2, fig. 9:3). On the right valve there is an acute, triangular posterior sail, which is situated somewhat dorsally to the midline and usually directed posterodorsally at an angle of about 10°. The left sulcal list ends at a distance from the posterior sail equal- ing 0.51 (0.42-0.61) the greatest depth of the body. WTien the posterior sail is fully developed, its length is 1.00 (0.81-1.35) the greatest depth of the body; i.e., its average length is subequal to that of the posterior main rib of the left sulcal list; and it is 2.9 (1.7-4.0) times longer than it is wide at the base. Somethnes both its margins are almost straight or slightly concave, but usually its dorsal margin is gently convex and its ventral margin gently concave; the angle at its tip is about the same as that at the tip of the posterior main rib of the left sulcal list. The posterior sail is supported by a central rib. This rib, which frequently is club-shaped distally, sometimes is straight and sometimes gently concave ventrally; in some specmiens its base is bifurcate or trifurcate; sometimes the sail extends to the tip of this rib, sometimes it terminates at some distance from the tip, and the distal portion of the rib is free ; the length of the free portion, which is variable, may exceed half the length of the rib. No parasagittal Usts are devel- oped. The thecal wall is finely and closely areolate and has scattered pores SYSTEMATIC ACCOUNT. 299 (Schiitt, 1895, pi. 2, fig. 9:3). According to Schutt (1895, p. 64), this species lacks chromatophores. Megacytic stages have not been recorded. The dimensions of eight of our specimens and of the specimens figured by Schutt (1895), Murray and Whitting (1899), and Jorgensen (1923) were measured. Figure 40. — Dinophysis schiilti Murray and WTiitting, all, except 7, in right-lateral view; 7, in ventral view; 6 and 7, from the same specimen. X 430. 1, from Station 4730 (300-0 fathoms); 2, from Station 4732 (300-0 fathoms); 3, from Station 4697 (300-0 fathoms); 4, 6, 7, from Station 4679 (300-0 fathoms); 5, from Station 4701 (300-0 fathoms). Dimensions: — Four of our large specimens: Length of body, 49.8-57.7 /j (average, 52.6 m). Greatest depth of body, 40.5-47.8 m (average, 43.2 n). Four of our small specimens: Length of body, 29.6-35.5 m (average, 32.2 ^i). Greatest depth of body, 24.7-29.1 m (average, 27.7 ^u). The specimen represented by Schutt's (1895) Plate 2, figure 9: 1, was 45.3 m long and 41.4 m deep. The type specimen, as figured by Murray and WTiitting (1899, pi. 31, fig. 10), was 37.5 m long and 34.1 /x deep. The specmiens measured by Jorgensen (1923) formed two size classes, the larger specimens being 54-62 m long and 45-47 m deep, the smaller 40-43 M long and 32-36 m deep. Variations: — Dinophysis schiitti, as conceived in this paper, is rather strik- ingly variable, and a future subdivision of this species is not excluded. The follow- ing characters are the most variable : — the size of the body, the shape of the hypotheca, the width of the cingular lists, the shape of the left sulcal list and the shape and relative lengths of its ribs, and the structure of the midrib of the pos- terior sail. Comparisons: — Jorgensen (1923, p. 34) writes about this species as follows: " It occurs in two forms, a larger, 54-62 m long by 45-47 m broad, often with 300 THE DIXOPHYSOIDAE. downward cur\'ing \'entral spines, and a smaller, 40-43 /u by 32-36 n, generally with straight ventral spines." Our specimens of this species also form two distinct size classes : the specimens of the larger class are 52.6 (49.8-57.7) n long and 43.2 (40.5-47.8) IX deep; the smaller class comprises specimens 32.2 (29.6-35.5) m long and 27.7 (24.7-29.1) m deep. The type specimen, figured by Murray and Whitting (1899, pi. 31, fig. 10), is 37.5 m long and 34.1 ^ deep and thus belongs to the small size class. The specimen figured by Schiitt (1895, pi. 2, fig. 9:1) is 45.3 ^ long and 41.4 n deep and should be referred to the larger size class. Do these size classes represent two different systematic units, or only devel- opmental or cyclic stages of one and the same species? The very marked dis- continuity in the size-frequency distribution of Jorgensen's (1923) specimens as well as of those examined by us undoubtedly seems to indicate that these classes are systematically distinct. On the other hand, no correlation has been found be- tween any morphological character and the size of the body. In the following the main results of an attempt to establish a correlation of this kind will be given. In our large specimens the ratio Ijctween the length and the greatest depth of the body is 1.22 (1.21-1.23): 1; in our small specimens, 1.20 (1.19-1.22): 1. In the specimens figured by Schiitt (1895), Murray and Whitting (1899), and Jorgensen (1923) this ratio varies between 1.09: 1 and 1.11: 1. In other words, in our large specimens this ratio is about the same as in our small specimens, and the difference established has no systematic importance as is shown by a com- parison with the correspondmg ratio in the specimens figured by Schiitt (1895), Murray and ^Vllitting (1899), and Jorgensen (1923). The ratios between the lengths of the three main ribs of the left sulcal list and the greatest depth of the body are to be found in the following table. Ri Our large specimens: 0.33 (0.27-0.40) Our small specimens: 0.35 (0.31-0.42) Schutt (1895, fig. 9:1): 0.33 Murray and Whitting (1899, pi. 31, fig. 10): 0.29 Jorgensen (1923, fig. 46): 0.31 Our large specimens thus have, on the average, relatively shorter ribs in the left sulcal list than have our small specimens. However, that the difference es- tablished has no systematic value is again indicated by a comparison with the corresponding ratios in the specimens figured by Schiitt (1895), Murray and Whitting (1899), and Jorgensen (1923). For instance, in spite of its small size, i?2- Rz 0.90(0.8.5-1.07) 1.06(1.00-1.10) 1.17(1.07-1.2.5) 1.15 (1.00-1.33) 0.89 0.89 0.S7 0.98 0.93 1.08 SYSTEMATIC ACCOUNT. 301 the specimen figured by IVIurray and Whitting (1899) has these ribs compara- tively shorter than in our large specimens. The ratio between the width of the anterior cingular list and the greatest depth of the body, the curvature of the ribs of the left sulcal list (mentioned by Jorgensen, 1923), the shape of the left sulcal list, and other characters exhibiting variations in this species have also been testedmorder to establish a morphologi- cal difference between the two size classes. In these cases, too, the results were negative. Under these circumstances we have judged it most advisable to accept Jorgensen's (1923) decision; and so we have treated these two size classes as systematically identical. Dmophysis schiltti is structm-ally closely related to D. swezyi, from which it is distinguished mainly by the fact that its left sulcal Hst lacks an accessory lobe just behind the posterior main rili. Other close relatives of D. schiitii probably are D. hastata, D. uracantha, D. triacantha, D. jorgenseni, and D. nias. D. schutti is easily distinguished from D. hastata and D. uracantha by the great length of the fission rib of its left sulcal list and the shape of the ventral margin of this Ust. From D. triacantha, D. jorgenseni, and D. riias, it is easily distinguished by having only one posterior sail. Synonymy: — This species was established by Murray and Whitting (1899), who described and figured it under the name of Dinophysis schiittii. However, it had been figured previously by Schiitt (1895) under the name of D. uracantha Stein. Jorgensen (1923) described and figured this species, Paulsen (1908) de- scribed it; both used the name D. schutti. Paulsen (1908) gave a reproduction of Murray and Whitting's (1899) figure of the type specimen. Murray and Whitting (1899), Paulsen (1908), and Jorgensen (1923) wrote D. uracantha Schutt (1895) as a synonym of D. schutti. The form described and figured by Forti and Issel (1924) under the name of D. schuettii var. uracanthoides, is, according to our opinion, not systematically different from the main species but should be regarded as one of the several modifications occurring in this. According to Jorgensen (1923, p. 34), Ostenfeld's (1900) record of Dinophysis schutti from the northern Atlantic "is perhaps due to confusion with some other form {Dinophysis uracanthat)." There are no facts available supporting this assumption. Ostenfeld (1900) does not give any figure, but he refers his readers to Murray and Whitting's (1899) Plate 31, figure 10, of D. schutti. Under these circumstances we accept Ostenfeld's (1900) assignment and consider the speci- men recorded by him as a straggler of southern origin carried to the north by the Gulf Stream. 302 THE DINOPHYSOIDAE. Occurrence: — Dinophysis schiiUi is recorded at twenty-two of the 127 sta- tions. There are 1, 1, 7, 8, 4, and 1 stations on the six lines of the Expedition. Of these twenty-two stations, one (4587) is in the Mexican Current; two (4652, 4676) are in the Peruvian Current; five (4689, 4691, 4695, 4697, 4699) are in the Easter Island Eddy; two (4713, 4715) are in the Galapagos Eddy; and twelve (4679, 4681, 4683, 4687, 4701, 4705, 4711, 4722, 4730, 4732, 4736, 4740) are in the South Equatorial Drift. At one of these stations (4676) the species was taken in a sur- face haul as well as in a haul from 300-0 fathoms; at one station (4652) in a haul from 100-0 fathoms; at one station (4689) in a haul from 800-0 fathoms; at one Figure 41. — Occurrence of Dinophysis schiilH Murray and 'WTiitting. Large, solid circles indicate records from vertical hauls; triangles, records from both vertical and surface hauls; small, solid circles, stations at which this species was not found; small, o])en circles, stations from which no plankton catclies were examined. station (4681) in hauls from 800-0 fathoms and 300-0 fathoms. All the remaining records refer to hauls from 300-0 fathoms only. The temperature range of these twenty-two stations at the surface was 66°- 82°; the average was 74.0°. At Station 4676, where the species was taken in a surface haul, the surface temperature was 69°. The frequency is 2% at one station (4695), 1% at four stations (4681, 4689, 4699, 4701), and less than 1% at the remaining stations. Schiitt (1895), who was the first to treat this species, did not give any in- formation as to the localities of his specunens; however, his material probably SYSTEMATIC ACCOUNT. 303 was taken either in the tropical or subtropical Atlantic or at Naples. Murray and Whitting (1899) recorded the species from the Sargasso Sea, lat. 29° 20' N.- 31° 20' N., long. 44° 50' W.-42° 30' W. Ostenfeld (1900) found it at lat. 59° 47' N., long. 25° 53' W., i.e., about midway between the northern point of Scotland and the southern point of Cireenland; Jorgensen (1923), in the Bay of Cadiz. According to Jorgensen (1923, p. 34), the species "occurs in the tropical and sub- tropical Atlantic, especially south of the equator"; however, there is no informa- tion as to what material this statement is founded upon. In the Mediterranean it has been found by the following authors : — in the Gulf of Lyons, by Pavillard (1916); off Monaco, by Nathansohn (1908, 1909); in the Adriatic Sea, by Forti and Issel (1924) ; and at several stations m different regions of this sea, by Jorgen- sen (1923). From the Marmora Sea it is reported by Jorgensen (1923), and from the Indian Ocean, at lat. 7°-3° N., long. 73°-86° E., by Cleve (1901a). Murray and WTiitting (1899) recorded a temperature of 67° for this species; Ostenfeld (1900), a temperature of 45.3° and a salinity of 35.21. Of the authors mentioned in connection with the distribution of this species, only Schutt (1895), Murray and \^Tiittmg (1899), Jorgensen (1923), and Forti and Issel (1924) give descriptions or figures by means of which their determina- tions of this form may be checked. This is a eupelagic species, widely distributed in tropical, subtropical, and warm-temperate seas. On one occasion it was recorded as far to the north as in the Irminger Sea, but the specimen found in this relatively cold region was pre- sumably a straggler of southern origin, carried to the north by the Gulf Stream. Its distribution in the Eastern Pacific, accordmg to our data, is peculiar in so far as it has never been found in the California Current and in the Panamic Area and only once in the Mexican Current; its relatively frequent occurrence in the Easter Island Eddy is also noteworthy. Its optimum habitat presumably is in deeper waters, within the levels of photosynthesis. DiNOPHYSis NiAS Karsten Figure 42: 1 Dinophijsis nias Karsten, 1907, p. 262, 421, 473, 540, pi. 47, fig. 7. Dinophysis triacantha Karsten, 1907, p. 421, 540, pi. 47, fig. 7. Diagnosis: — Body subovate in lateral outline, deepest somewhat behind the middle; length: depth, 1.18-1.23: 1 ; longitudinal axis deflected posteroventrally at 10°-15°. Both cingular lists ribbed; the anterior 3.0 times wider than transveree furrow. Left sulcal list 0.62-0.67 the length of body; distance between Ri and Rj 304 THE DINOPHYSOIDAE. 0.50-0.53 the length of body; R2 is 0.25-0.31, R3 is 0.84-0.92 the greatest depth of body; margin forms angle of 10°-15° at R3; R3 incHned posteriorly at 30°. With two acute posterior sails, one on each side of midlme, each supported by central rib which may be bifurcate basally. Ventroposterior sail 0.82-0.83 the greatest depth of body in length, 3.0-3.3 times longer than its basal width, and separated from left sulcal list and dorsal posterior sail by distance eciualing or somewhat exceeding its basal width. Dorsoposterior sail 0.63-0.78 the greatest depth of body in length and 2.2-3.0 times longer than its basal width. Except for central ribs posterior sails appear to lack structure. Theca finely and closely areolate. Length, 52.2-56.0 m. Probably widely distributed in tropical seas. Description: — This is a medium-sized species, subovate in lateral outline, subtruncate anteriorly, and deepest somewhat behmd the middle. The ratio between the length and the depth of the body is 1.18-1.23: 1. The longitudinal axis is deflected posteroventrally at about 10°-15°. The epitheca is about 0.54 as deep as the hypotheca, gently convex, highest in the center, and not visible above the anterior cingular list. The transverse furrow is flat or gently convex and about twice as wide as the greatest height of the epitheca. The posterior cmgular list is about 0.13 the length of the body from the apex. The hypotheca has a rather pronounced posteroventral inclination (see above). The dorsal margin (from the girdle to the antapex) has a subuniform and moderate convexity. The ventral margin (from the girdle to the antapex) is rather strongly convex posteriorly and more or less flattened anteriorly. The pos- terior end of the body is distinctly deeper than the anterior. The anterior cingular list is about 3.0 times wider than the trans^•erse furrow and 0.33-0.40 the greatest depth of the body; it is inclined anteriorly at about 40°-50° and has, on each valve, about 11-12 simple, straight, and almost equi- distant ribs (Karsten, 1907, pi. 47, fig. 7). The posterior cingular list is about as wide as or somewhat narrower than the anterior, inclined anteriorly at about 40°, and has, on each valve, about 13-14 simple, straight, and nearly equidistant ribs (Karsten, 1907, pi. 47, fig. 7). Judging by Karsten's (1907) figure, the right sulcal list extends to the fission rib of the left sulcal list and is subtriangular, decreasing in width posteriorly. The left sulcal list is strikingly wider pcsteriorly than an- teriorly. Its length is 0.62-0.67 the length of the body, and the distance between the anterior and posterior main ribs is 0.50-0.53 the length of the body. The an- terior main rib is 0.22-0.28, the fission rib 0.25-0.31, and the posterior main rit) 0.84-0.92 the greatest depth of the body. Behind the posterior main rib this list SYSTEMATIC ACCOUNT. 305 decreases suddenly in width. Between the anterior and posterior main ribs the free margin of this list is almost straight or slightly undulating (Karsten, 1907, pi. 47, fig. 7). At the posterior main rib the list is acuminate, and the margin forms an angle of 10°-15°. Behind this rib the margin is moderately convex or sigmoid. The mam ribs are of moderate strength and not club-shaped or other- wise modified; the two anterior are straight or almost so, the posterior is gently concave posteriorly. The distance between the anterior main rib and the fission rib is 0.45-0.50 the distance between the anterior and posterior main ribs. The posterior main rib has a posterior inclination of about 30°. Except for the main ribs, this list appears to lack ribs. On the right valve there are two acute posterior sails, one on each side of the midlme, each supported by a central rib which is gently concave ventrally and which may be bifurcate basally. The ventropos- terior sail is 0.82-0.83 the greatest depth of the body in length, 3.0-3.3 times longer than its basal width, and directed posteroventrally; its ventral margin is gently concave, its dorsal margin gently convex, and the angle at its tip about 10°-15°; it is separated from the left sulcal list and from the dorsal posterior sail by distances equaling or somewhat exceeding its basal width. The dorsopos- terior sail is 0.63-0.78 the greatest depth of the body in length, 2.2-3.0 times longer than its basal width, and directed posterodorsally; its ventral margin is gently concave, its dorsal margin gently convex, and the angle at its tip about 15°. With the exception of the central ribs, the posterior sails appear to lack struc- tural differentiation. There are no parasagittal lists along the dorsal margin of the hypotheca and on the epitheca. The thecal wall is finely and closely areolate and has scattered pores. The areoles are of about the same size as the pores. Megacytic stages have not been recorded. The dimensions of one-specimen from the material of the Expedition and of the type specimen, as figured by Karsten (1907, pi. 47, fig. 7), were measured. Dimensions: — Our specimen: Length of body, 56.0 m- Greatest depth of body, 47.2 fx. The type specimen, as figured by Karsten (1907, pi. 47, fig. 7), was about 52.2 ^ long and 42.6 ^ deep. Variations: — Judging by the two specimens hitherto figured, one of which was taken off the coast of Sumatra, the other in the eastern tropical Pacific, this species appears to be rather constant. Comparisons: — The specimen found in the material of the Expedition re- sembles closely the type specimen, as figured by Karsten (1907). It differs mainly in having the two posterior sails subequal and (possibly) in its somewhat nar- rower anterior cmgular list. 306 THE DINOPHYSOIDAE. The structurally closest-known relatives of Dinophysis nias are D. triacantha and D.jorgenseni (see the last two species, the sections on comparisons). D. nias is easily recognizable by the straight or nearly straight ventral margin of its left sulcal list, and by having its two posterior sails free, i.e., connected neither with each other nor with the left sulcal list. Figure 42. — 1, Dinophysis nias Karsten, left lateral view; porulation indicated on small portion of thecal wall. X 430. Station 4691 (300-0 fathoms). 2, Dinophysis triacantha Kofoid, right lateral view of type specimen. X 430. Station 4722 {30O-0 fathoms). 3-8, Dinophysis jorgenseni, sp. nov., right lateral view; porulation indi- cated on small portion of thecal wall in 7. X 4.30. 3, from Station 4713 (300-0 fathoms); 4, from Station 4609 (300-0 fathoms); 5, from Station 4732 (300-0 fath- oms); 6, from Station 4637 (300-0 fathoms); 7, from Station 4709 (300-O fathoms); 8, from Station 4730 (300-0 fathoms). Synonymy: — This species was established by Karsten (1907), who de- scribed and figured it under the name of Dinophysis nias. While his paper was in press, Karsten received Kofoid's (1907a) preliminary report on the dinoflagellates of the Expedition to the Eastern Tropical Pacific. Identifying Kofoid's D. tri- acantha with D. nias, Karsten (1907), on page 421 and in the explanation to SYSTEMATIC ACCOUNT. 307 Plate 47, wrote the latter name within parenthesis as a synonym of the former. We do not accept this identification but consider these two forms as specifically distinct. Occurrence: — Dinophysis nias is recorded at only one (4691) of the 127 stations, on the third line of the Expedition, in the Easter Island Eddy, from 300-0 fathoms, and at a surface temperature of 73°. The frequency was less than 1% (one specimen). It was first recorded by Karsten (1907) from the west coast of Sumatra, and probably is widely distributed, but rare, in tropical seas. Dinophysis jorgenseni, sp. nov. Plate 5, fig. 3, 7. Figure 42: 3-8, 43 Dinophysis trincantha Jorgensen, 1923, partim, p. 34, 43, fig. 47. Diagnosis: — Body subellipsoidal or subovate in lateral outline, deepest near the middle; length: depth, 0.99-1.08: 1; longitudmal axis perpendicular to girdle or somewhat deflected posterodorsally or posteroventrally. In dorsal view ovate. Anterior cingular list 1.5-2.5 times wider than transverse furrow. Left sulcal list: distance between Ri and R3 0.52-0.58 the length of body; R2 is 0.35- 0.43, R3 is 0.58-0.82 the greatest depth of body; margin often rather strikingly concave just behind Ro and forms angle of 10°-35° at R3; R3 inclined posteriorly at 30°-45° and often club-shaped; usually without accessory structure. With two subequal, acute posterior sails, one on each side of midline, each supported by central rib, which is 0.36-0.55 the greatest depth of body m length and sometimes bifurcate or trifurcate basally; the dorsal usually connected with ventral by list, the minimum width of which is 0>5-l .3 the width of transverse furrow ; the ventral usually connected with left sulcal list by similar sail; except for central ribs, posterior sails appear to lack structure. Theca areolate or reticulate. Length, 51.1-58.8 M. Widely distributed, but rare, in tropical, subtropical, and warm-temperate seas. Description: — This is a medium-sized species, the body of which is sub- ellipsoidal to broadly subovate m lateral outline, and deepest m or near the middle. The ratio between the length and the depth of the body is 1.05 (0.99- 1 .08) : 1 . The longitudinal axis is sometunes perpendicular to the girdle and sometimes deflected posterodorsally at 3° (1°-15°); seldom it has a slight (l°-3°) posteroventral inclination. The posterodorsal deflection of this axis is unusually 308 THE DINOPHYSOIDAE. pronounced in the t3rpe specimen, viz., 15°; in the other specimens examined it does not exceed 5°. The epitheca is 0.57 (0.53-0.61) as deep as the hypotheca, highest in or near the center, usually evenly and moderately convex, seldom flat (Plate 5, fig. 7), and not visible above the anterior cingular list. The transA-erse furrow is flat or slightly con\-ex, and about as wide as or more or less decidedly wider than the greatest height of the epitheca. The posterior cingular list is 0.17 (0.11-0.20) the length of the body from the apex. The hypotheca is sometimes subsymmetri- cal (Figiu'e 42 : 5) , sometunes more or less asymmetrical (Plate 5, fig. 7) . The dorsal and ventral margins are moderately to rather strongly convex ; sometimes they have about the same shape, but either of them may be somewhat more or less convex than the other; in some specunens their convexity is subuniform, in others (Figure 42 : 4, 6) they are more or less flattened anteriorly. The post- margm is broadly rounded, sometimes (Plate 5, fig. 7) flattened ventralty, and confluent with the dorsal and ventral margms. The posterior end of the body is either about as deep as the anterior or more or less strikingly deeper. In dorso- ventral view the body is ovate, widest somewhat behind the middle, with evenly convex side contours, and with the posterior end somewhat more broadly rounded than the anterior. The anterior cingular list is 2.1 (1.5-2.5) tunes wider than the transverse furrow and 0.27 (0.23-0.29) the greatest depth of the body. It has on each valve 14-20 ribs, which usually are sunple, straight, and almost equidistant; sometimes some of these ribs are branched (Plate 5, fig. 7) or incomplete (Jorgensen, 1923, fig. 47). The posterior cmgular list is either about as wide as or somewhat nar- rower than the anterior and lacks structural dilTerentiation. Both these lists are funnel-shaped and have an anterior mcliiiation of 30°-40°. The right sulcal list extends to or somewhat beyond the fission rib of the left sulcal list; it is either sub- triangular, decreasing gradually in width posteriorly, or of subuniform \\idth throughout the greater portion of its length and more or less well rounded pos- teriorly; its ventral margm is either straight, gently convex, or sigmoid, concave anteriorly and convex posteriorly. The left sulcal list is large. The distance be- tween the anterior and posterior mam ribs is 0.55 (0.52-0.58) the length of the body. The anterior main rib is 0.24 (0.20-0.28), the fission rib 0.40 (0.35-0.43), and the posterior main rib 0.67 (0.58-0.82) the greatest depth of the body. Be- hind the posterior main rib this list decreases suddenly in width. Between the anterior main rib and the fission rilj the free margm of this list is straight or slightly couAex, concave, or sigmoid; between the fission rib and the posterior SYSTEMATIC ACCOUNT. 309 main rib it is concave, and the concavity often is rather striking just behind the fission rib. At the posterior mam rib this margm forms an angle of 20° (10°-35°) ; behind this ril) it usually is gently to moderately concave, but sometimes it is sigmoid (Figure 42:6, 8) or somewhat irregular (Figure 42:7). The main ribs are of moderate strength, and the posterior is often somewhat club-shaped. The anterior main rib is straight or almost so; the fission rib is straight, gently con- cave posteriorly, or sigmoid (Plate 5, fig. 7) ; the posterior main rib is gently con- cave posteriorly. The distance between the anterior main rib and the fission rib is 0.51 (0.48-0.60) the distance between the anterior and posterior main ribs. The posterior mam rib has a posterior inclination of 37° (30°-45°). Except for the three mam ribs, this Ust usually lacks structural differentiation, but in some specimens traces of an irregular reticulation may be found; furthermore, m the type specimen (Plate 5, fig. 7) there is an accessory longitudinal rib between the anterior main rib and the fission rib. On the right valve there are two subequal, acute posterior sails, one on each side of the midlme, and each supported by a central rib, which may be bifurcate or trifurcate basally and which usually has a gentle ventral concavity. The rib of the ventral of these sails is 0.45 (0.40-0.52), that of the dorsal 0.46 (0.36-0.55) the greatest depth of the body; in other words, these ribs are somewhat shorter than the posterior main rib of the left sulcal list. The angle at the tip of the \'entral of these sails is 23° (10°-35°) ; that at the tip of the dorsal, 25° (10°-40°). In the specimens examined by us the dorsal of these sails is always connected with the ventral by a list, the minimum width of which is 0.5-1.3 the width of the transverse furrow ; in the specimen figured by Jorgensen (1923, fig. 47) this connection is broken. The ventroposterior sail usually is con- nected with the left sulcal list by a list similar to the one between the dorsal and the ventral posterior sails, but this connection too may be broken (Figure 42: 7, 8; Jorgensen, 1923, fig. 47). Except for the central ribs, the posterior sails appear always to lack structural differentiation. There are no parasagittal hsts along the dorsal margin of the hypotheca and on the epitheca. The thecal wall is areolate or reticulate and has scattered pores. The areoles, which are closely set, sometunes are about as small as the pores, and sometimes (Plate 5, fig. 7) they widen into polygonate meshes. The areoles and the poly- gonate meshes grow fainter toward the margins of the valves, and at least some- times they are lacking on the epitheca, m the transverse furrow, and along the dorsal, posterior, and ventral margins of the hypotheca (Plate 5, fig. 7). In the transverse furrow two rows of pores are present. Megacytic stages have not been recorded. 310 THE DINOPHYSOIDAE. The dimensions of eight of our specimens and of the specimen figured by Jorgensen (1923, fig. 47) were measured. Dimensions: — Our specimens: Length of body, 51.1-58.8 m (average, 55.0 n; type, 54.8 fi). Greatest depth of body, 48.6-56.2 ^ (average, 52.7 n; type, 53.5 n). The specimen figured by Jorgensen (1923, fig. 47) was about 52.9 m long and 50.0 n deep. Variations: — As conceived in the present paper, this species is rather vari- able. The most variable characters are as follows: — the shape and structure of the body, the shape of the ventral margin of the left sulcal list, the structure of the central ribs of the two posterior sails, and the development of the hsts between the posterior sails and between the ventral posterior sail and the left sulcal list. Comparisons: — Although fairly variable in several respects, Dinophysis jorgenseni is so uniform in size and structure that its systematic unity hardly can be doubted. The allocation of only one of our specimens Is questionable. In this specimen (Figure 42: 8) the dorsoposterior sail is indicated only by the bulging of the dorsal margin of the ventroposterior sail and by a rudimentary central rib. Furthermore, the posterior main rib of the left sulcal list of this specimen is more club-shaped and has a more pronounced posterior concavity than in any of the other specimens examined. This specimen is included m the species, since the slight development of the dorsoposterior sail may be due to recent binary fission. The specimen of this species figured by Jorgensen (1923, fig. 47) differs from our spechnens mainly in having no connecting lists between the two posterior sails and between the ventral posterior sail and the left sulcal list. As to the systematic value of this difference, it should be mentioned that in the specimens examined by Jorgensen (1923, p. 34) the posterior sails were either "free, or joined by the sail lists, or the ventral alone connected with the lower part of the large (left) longitudinal list." The closest-known relatives of Dinophysis jorgenseni probably are D. Iri- acantha and D. 7iins. This relationship is suggested especially by the presence of two posterior sails. D. triacantha is closer to D. jorgenseni than is D. nias in the shape of its left sulcal list and in having the ventroposterior sail connected with the left sulcal list. It differs most strikingly from D. jdrgenseni in having the posterior sails supported by marginal instead of by central ribs. D. nias ap- proaches D. jdrgenseni more than does D. triacantha in having the posterior sails supported by central ribs, which may be bifurcated basally. It differs from this species especially in having the posterior sails well separated from each other and SYSTEMATIC ACCOUNT. 311 from the left sulcal list and in the almost straight ventral margin of its left sulcal list (see also D. triacaniha, the section on comparisons) . Synonymy: — This species was first described and figured by Jorgensen (1923) under the name of Dinophysis triacantha Kofoid. Under this name Jorgensen (1923, p. 35) also included D. nias Karsten. Occurrence: — Dinophysis jorgenseni is recorded at twenty-three of the 127 stations. There are 3, 8, 2, 5, 4, and 1 stations on the six Imes of the Expedition. Of these stations one (4598) is in the Mexican Current; three (4609, 4613, 4637) are m the Panamic Area; seven (4650, 4651, 4652, 4655, 4659, 4663, 4665) are in the Peruvian Current; two (4713, 4715) are in the Galapagos Eddy; ten (4679, Figure 43. — Occurrence of Dinophysis jorgenseni, sp. nov. Large, solid circles indicate records from vertical hauls; squares, records from surface hauls; small, solid circles, stations at which this species was not found; small, open circles, stations from which no plankton catches were examined. 4681, 4701, 4709, 4711, 4717, 4727 [type locality], 4730, 4732, 4742) are in the South Equatorial Drift. There is only one record from surface hauls (Station 4727). At one station (4652) the species was taken in a haul from 100-0 fathoms; at one station (4655) in a haul from 400-0 fathoms; and at two stations (4651, 4715) in hauls from 800-0 fathoms. All the remaining records refer to hauls from 300-0 fathoms. The temperature range of these twenty-thi-ee stations at the surface was 65°-84°; the average was 73.3°. At Station 4727, where this species was taken in a surface haul, the svu-face temperature was 77°. 312 THE DINOPHYSOIDAE. At thi'ee stations (4613, 4663, 4715) the frequency is 1%; at the reniaming stations it is less. This species was first recorded by Jorgensen (1923), who found it at five stations in different regions of the Mediterranean, at one station on the coast of Portugal, and at lat. 3° S., long. 16° W., Gulf of Guinea. In the Mediterranean it seems to occur only below 200 meters (Jorgensen, 1923, p. 35). This is a eupelagic species, widely distributed, but rare, in tropical, sub- tropical, and warm-temperate regions. The most outstanding peculiarities about its distribution in the Eastern Pacific, according to our data, are its rare occur- rence in surface waters, its absence from the California Current and the northern portion of the Mexican Current, and its relatively frequent occurrence m the Peruvian Current and m the regions under the direct influence of this current. Its optimiun habitat probably is in deeper waters, within the levels of photo- synthesis. DiNOPHYSIS TRIACANTHA Kofoid Figure 42:2 Dinophijsis triacanlhn Kofoid, 1907a, p. 196, pi. 12, fig. 74. non K.^rsten, 1907, p. 421, .540, pi. 47, fig. 7. Kofoid & Michener, 1911, p. 292. 7ioyi Jorgensen, 1923, p. 34, 43, fig. 47. Diagnosis: — Body subcircular in lateral outline; length: depth, 0.97:1. Both cingular lists ribbed; the anterior 1.5 times wider than transverse furrow. Left sulcal list: distance between Ri and R3 0.59 the length of body; R2 is 0.32, R3 is 0.48 the greatest depth of body; margin rather .strikingly concave just be- hind R2 and forms angle of 30° at R3; R3 inclined posteriorly at 35°; besides main ribs this list has a few short ribs. With two subequal, acute posterior sails, one on each side of midline, each supported by marginal ribs and lackmg central rib; height of these sails, 0.37-0.44 the greatest depth of body ; angles at their tips, 20°; no connectmg list between posterior sails; ventroposterior sail connected with left sulcal list by list, the minimum width of which is 0.7 the width of transverse fur- row; except for marginal ribs, posterior sails lack structure; theca with fine, irregular reticulation. Length, 49.4 /i. Eastern tropical Pacific. Description: — This is a medium-sized species, subcircular in lateral outline, and deepest in the middle. The ratio between the length and the depth of the body is 0.97: 1. The longitudinal axis is perpendicular to the girdle. The epitheca is about 0.58 as deep as the hypotheca, of moderate convexity, highest in the center, and not \isible above the anterior cingular Ust. The trans- SYSTEMATIC ACCOUNT. 313 verse furrow is flat or gently concave, and about twice as wide as the greatest height of the epitheca. The posterior cingular Hst is about 0.21 the length of the body from the apex. The hypotheca has a subcircular outline with almost uniform convexity. The anterior cingular list is about 1 .5 times wider than the transverse furrow and 0.25 the greatest depth of the body; it is inclined anteriorly at 25°-30° and has, on each valve, about 18-20 complete or incomplete ribs, most of which are straight and unbranched. The posterior cingular list is slightly narrower than the anterior, inclined anteriorly at about 25°, and has, on each valve, about 19-20 simple, straight, and almost equidistant ribs. The right sulcal list extends to or slightly beyond the fission rib of the left sulcal list, is subtriangular in shape, de- creases gradually in width posteriorly, and is anteriorly about half as wide as the transverse furrow. The left sulcal list is rather large. The distance between the anterior and posterior main ribs is about 0.59 the length of the body. The an- terior main rib is about 0.21, the fission rib 0.32, and the posterior main rib 0.48 the greatest depth of the body. Behind the posterior main rib this list decreases suddenly in width. Between the anterior main rib and the fission rib the free margin of this list is straight or almost so; just behind the fission rib it is rather strikingly concave; at the posterior main rib it forms an angle of about 30°; and just behind the last rib it is gently concave. The main ribs are of moderate strength and not club-shaped or otherwise modified. The anterior main rib is straight or almost so ; the fission rib and the posterior main rib are gently concave posteriorly. The distance between the anterior main rib and the fission rib is about 0.37 the distance between the anterior and posterior main ribs. The pos- terior main rib has a posterior inclination of about 35°. Besides the main ribs this list has a few short ribs, some of which may be branched, or a feeble reticula- tion; at the base of the posterior main rib there is a short posterior branch. On the right valve there are two subequal, acute posterior sails, one on each side of the midline, each supported by marginal ribs and lacking a central rib. The ventral of these sails has a length 0.44 the greatest depth of the body and an angle at the tip of about 20°; the dorsal has a length 0.37 the greatest depth of the body and an angle at the tip of about 20°. There is no connecting list between the pos- terior sails, but the ventroposterior sail is connected with the left sulcal list by a list, the minmium width of which is about 0.7 the width of the transverse fuiTow. Except for the marginal ribs, which may have a few short branches, the posterior sails lack ribs. There are no parasagittal lists along the dorsal margin of the hypotheca and on the epitheca. 314 THE DINOPHYSOIDAE. Megacytic stages have not been recorded. The dimensions of one specimen only, the type, were measured. Dimensions: — Length of body, 49.4 ju. Greatest depth of body, 50.7 ju. Comparisons: — Our description and figure of this species are based on the type specimen. Dinophijsis triacantha appears to be structurally most closely related to D. nias and D. jorgeyiseni (see the last two species, the sections on comparisons). It is easily distinguished from these two species by having its two posterior sails supported by marginal ribs instead of by central ribs. The structure of its poste- rior sails is suggestive of a fairly close affiliation to D. uracantha, D. urceolus, D. monacantha, and D. trapezium, but this must be regarded as uncertain. Synonymy: — The species was established by Kofoid (1907a), who described and figured it under the name of Dinophysis triacantha. D. triacantha Karsten (1907) is identical with D. nias Karsten and not with D. triacantha Kofoid. D. triacantha Jorgensen (1923) refers both to D. nias Karsten and to D. jorgen- seni nobis. Occxirrence: — Dinophysis triacantha is recorded at only one (4722, type locality) of the 127 stations, on the fifth line of the Expedition, in the South Equatorial Drift, from 300-0 fathoms, and at a surface temperature of 75°. The frequency is less than 1% (one specimen). 3. Caudata Group. — Dinophysis caudata., D. tripos, and D. rniles belong to this group. Of these only the first was found in the material of the Expedition. Dinophysis caudata Saville-Kent Figure 44, 45 Dinophysis caudata Saville-Kent, 18S1, p. 455, 400. Jorgensen, 1923, p. 22, 24-30, 43-45. Pavillard, 1923a, p. 881. Dinophysis acuta var. geminata Podchet, 1883, p. 27, pi. IS, 19, fig. 5. Dinophysis honmnculus Stein, 1883, partim, p. 3, 24, jil. 21, fig. 1, 2, 5-7. Bergh, 1884, p. 385. Butschli, 1885, p. 995, 996, pi. 55, fig. 36. Imhof, 1886, p. 102. Hensen, 1887, p. 77, tab. 13-16; 1911, p. 164 165. MoDius, 1887, p. 121. Daday, 1888, p. 99. Whitelegge, 1891, partim, p. 184, pi. 28, fig. 14 W.\LTHER, 1893, p. 129. Scui-TT, 1895, p. 65, 112, pi. 2, fig. 8. Cleve, 1897a, p. 26, tab. 1; 1899a p. 17; 1899c, p. 4; 1900b, p. 1031; 1900e, p. 21, 35, 39; 1900f, p. 925, 936; 1901a, p. 15; 1901c, p. 239: 1902a, p. 10, 25; 1902b, p. 29; 1903a, p. 38; 1903b, p. 344. Ostenfeld, 189Sa, \>. 428; 1898b, tab. 3-5: 1899, tab. 2-5; 1902, p. 20; 1906, ]>. 18; 1915, p. 6; 1916b, p. 13. Lemmebmann, lS99a, p. 319, 373: 1901a, p. 374; 1902a, p. 263; 190.5a, p. 36. Minkiewitsch, 1899, p. 357, 361. Murray & Whitting, 1899, p. 331, tab. 1, 3-6, 8, 9. Schroder, 1900a, partim, p. 19, 21, 35; 1900b, p. 3; 1906a, p. .321 325, 326, 374; 1906b, p. 262; 1909, p. 211; 1911, p. 12, 17, 18, 25, 36, 37, fig. 7b, Schmidt, 1901, p 138. Ostenfeld & Schmidt, 1901, p. 169. Cleve, Ekman & Pettersson, 1901, p. 36. Gran, 1902 p. 184. LoHMANN, 1902, p. 53; 1908n, p. 168, 169. Entz, 1902a, fig. 1; 1902b, p. 94; 1905, p. Ill 1907, p. 19; 1909, p. 255. Paulsen, 1904, p. 29; 1908, p. 12, 19, fig. 20; 1912, p. 290. Pavillard, 1905, p. 59, 100, 102; 1909, p. 284; 1916, p. 56, fig. 15A. Kar.sten, 1905, p. 8, 34, 35; 1906, p. 180; SYSTEMATIC ACCOUNT. 315 183, 193, 195, 196, 198, 205-207, 210-213, 215, 217, 218; 1907, p. 225, 240, 209, 278, 332, 333, 335, 359, 424, 445, 449, 451, 471, 472. Apstein, 1905a, tal). 2, 10. Cl.\rk, 1905, p. 179, 182, 188. Jorgen- SEN, 1905, p. 50, 108; 1923, p. 24, 29. Zachari.^s, 1906, p. 509, 530, 531, 533, 534, 536, 538, 540, 544, 545, 549, 551, 552, 5.54. Linko, 1907, p. 109, 111. Nathan.sohn, 1908, p. 604, 612; 1909, p. 46, 59; 1910a, p. 14, 15, 20. Czapek, 1909, p. 8, fig. 5. Stuwe, 1909, p. 243, 254, 288. Graf, 1909, p. 139, 157, 1.59, 161, 163, 171, 173, 184, 192. Reinhard, 1910, p. 8, 12, 14, 15, 16, 23, 26. Schiller, 1911a, p. 52; 1911b, p. 90; 1912, p. 27. Carazzi & Grandori, 1912, p. 48, tab. F. Mangin, 1912, p. 5-7, 10, 12-18, 20, 21, 58, 60, tab. 2. Faria & Cunha, 1917, p. 74. Cunha & Fonseca, 1917, |>. 141. FoRTi, 1922, p. 107, 109, 190, 208, fig. 114. Dinophysis allien Gourret, 1883, p. 75, 76, 78-80, pi. 3, fig. 54, 54a. Dinophysis inaequalis Gourret, 1883, p. 77, 78, 80, pi. 1, fig. 21. Dinophysis homunculus and caudata Grenfell, 1887, p. 559. Dinophysis semicarinata Grenfell, 1887, p. 560, pi. 11, fig. 12. Dinophysis homunculus f. pedunculata Schmidt, 1901, p. 138, fig. S. Okamura, 1907, p. 131, fig. 40; 1912, p. 19. VAN Oye, 1921, p. 215. Dinophysis homunculus var. carinata Zacharia,?, 1906, p. 540. Dinophysis homunculus var. pedunculaia Schroder, 1906a, p. 335. Jorgensen, 1923, p. 25, 29. Dinophysis homunculus var. gracilis Schroder, 1911, p. 5, 7, 22-24, 36, fig. 7a. Dinophysis geminata Kofoid & Rigden, 1912a, p. 337. Dinophysis pedunculata O.stenfeld, 1915, p. 6, 7. Dinophysis caudata var. allieri Jorgensen, 1923, p. 25. Dinophysis caudata allieri f. speciosa Jorgensen, 1923, p. 25, fig. 30. Dinophysis caudata var. abbreriata Jorgensen, 1923, p. 25, 27, 28, fig. 31. Dinophysis caudata f. marmarae Jorgensen, 1923, p. 26, 45, fig. 33. Dinophysis caudata f. pontica Jorgensen, 1923, p. 26, 45, fig. 34. Dinophysis homunculus var. latus Lindemann, 1923, p. 220, fig. 11, 12; 1925, p. 101. Diagnosis: — Body very variable in shape, elongated; deepest somewhat in front of to somewhat behind the middle, and 1.65-3.23 times longer than deep; anterior portion irregularly obovate in lateral outline, subtruncate anteriorly; posterior portion more or less slender, pedunculate, of varying length; transition between anterior and posterior portions gradual or more or less sudden; longi- tudinal axis inclined posteroventrally at 7°-30°. In dorsal view strongly flattened, 7-8 times longer than wide. Anterior cingular list 1.5-2.5 as wide as transverse furrow. Left sulcal list extends to base of peduncle; distance between Ri and R3 0.28-0.38 the length of body; R2 is 0.23-0.59, R3 is 0.31-0.77 the greatest depth of body; margin forms angle of 70°-110° at R3, or this portion of list is recurved; R3 about parallel to girdle ; with more or less developed reticulation. Theca finely and closely areolate. Length, 71.8-107.0 yu. Widely distributed in tropical, subtropical, and warm-temperate seas, seldom in colder waters. Descriptio7i: — This is a medium-sized to rather large species, very variable in shape, elongated, and deepest somewhat in front of to somewhat behind the middle. The anterior portion of the body is irregularly obovate in lateral outline, subtruncate anteriorly; the posterior portion is more or less slender, pedunculate, of varying length. The ratio between the length and the depth of the body is 1.65-3.23: 1; in our specimens this ratio is 2.06 (1.65-2.46): 1; in the specimens 316 THE DINOPHYSOIDAE. figured by Stein (1883, pi. 21, fig. 1, 2, 5) which are unusually slender, it is 2.82- 3.23: 1. The longitudinal axis is inclmed posteroventrally at 7°-30°; in our speci- men this angle measured 20° (14°-30°). Seldom almost symmetrical specimens are found (see Stein, 1883, pi. 21, fig. 1, and Jorgensen, 1923, p. 2,5). The epitheca is gently convex to very slightly concave, highest in the middle or near the dorsal side, not visible above the anterior cingular list, and 0.43-0.58 as deep as the hypotheca; in our specimens this ratio is 0.51(0.45-0.58) : 1 ; in Stein's (1883) it is 0.43-0.46: 1. The transverse furrow is flat or gently con- cave, and generally somewhat wider than the greatest height of the epitheca. The posterior cingular list is 0.04-0.08 the length of the body from the apex. The hypotheca is very variable in shape; its anterior portion is irregularly obovate, and its posterior portion is more or less slender, pedunculate. The transition be- tween the anterior portion and the posterior (the pedvmcle) sometimes is gradual, sometimes more or less sudden; in forma pedunculata it is very sudden. The ratio between the length of the body and that of the peduncle is variable, due to the great variability of the length of the peduncle. In oiir specimens this ratio is 2.73 (2.35-4.21) : 1 ; in the specimens figured by Stein (1883) it is 2.21-2.50: 1; in the other specimens of this species figured by previous authors it falls within this range of variation. The dorsal margin of the anterior portion of the hypotheca either is gently concave (Stein, 1883, pi. 21, fig. 2, 5), straight (Stein, 1883, pi. 21, fig. 1), almost uniformly and gently to moderately convex (Figure 44: 2), or its anterior and posterior parts are gently convex or almost straight and together form a broadly rounded but distinct corner (Figure 44: 4) ; this corner has a vary- ing relative position (Figure 44: 5, 6), but usually it is located behind the middle of this margin. The ventral margin of the anterior portion of the hypotheca usually is straight or almost so, or it is gently undulating, being convex anteriorly and posteriorly and concave in the middle. Proxinially the peduncle is 0.30-0.55 the greatest depth of the body and about 1.9-3.0 times wider than distally. In most specimens its dorsal and ventral margins are gently convex, but sometimes they are almost straight (Figure 44:6) or somewhat irregular (Figure 44:7). Posteriorly the peduncle is truncate, rounded, or pointed; ^\■hen truncate or rounded, the antapex frequently has one to four tubercles of varying size. In dorsoventral view the body is very narrow, about seven to eight times longer than wide (Stein, 1883, pi. 21, fig. 6, 7), and with subparallel sides. The cingular lists may be subequal, but in most specimens the posterior is somewhat narrower than the anterior; the anterior is 1.5-2.5 times wider than the transverse furrow and 0.26 (0.18-0.35) the greatest depth of the body; both are SYSTEMATIC ACCOUNT. 317 funnel-shaped, with an anterior mclmation of 20°-50° ; sometimes the posterior is somewhat less inclined than the anterior. In most cases the anterior has, on each valve, about ten to fifteen simple, straight, almost equidistant, and sometimes partly incomplete ribs ; sometunes the posterior may have about the same struc- ture as the anterior, but in most specimens it lacks structural differentiation. The right sulcal list ends from just in front of to somewhat behind the fission rib of the left sulcal list; it either is subtriangular, decreasing gradually in width pos- teriorly, or it is of subuniform width throughout its anterior half; anteriorly it is about as wide as, or usuallj' somewhat wider than, the transverse furrow. The left sulcal list extends to the base of the pedimcle ; the distance between the an- terior and posterior main ribs is 0.33-0.42 the length of the body. In our speci- mens the anterior main rib is 0.29 (0.21-0.40), the fission rib 0.32 (0.23-0.40), and the posterior main rib 0.40 (0.31-0.50) the greatest depth of the body. In the specimens figured by Stein (1883, pi. 21, fig. 1, 2, 5) this list is unusually wide and the corresponding values are 0.40 (0.34-0.50), 0.48 (0.44-0.59), and 0.66 (0.59- 0.77). Behind the posterior mam rib this list decreases suddenly in width. Be- tween the anterior and posterior main ribs the free margin of this list is gently concave (Figure 44: 6), almost straight throughout (Figure 44: 3), almost straight anteriorly and posteriorly and convex in the middle (Jorgensen, 1923, fig. 34), gently to moderately convex (Figure 44: 2), or gently sigmoid, straight to gently concave anteriorly and convex posteriorly (Figure 44 : 7) . At the posterior mam rib this margin forms an angle of 70°-l 10° ; or this portion of the list is more or less strikingly recurved (Stein, 1883, pi. 21, fig. 1; Jorgensen, 1923, fig. 31). Behind the posterior main rib this margm either is almost straight, somewhat convex or irregular (Figure 44: 1), or it is more or less concave (Jorgensen, 1923, fig. 31). The main ribs of this list are of moderate strength ; the two anterior are straight or almost so; the posterior, which in exceptional cases (Stein, 1883, pi. 21, fig. 5) is club-shaped, is sometimes straight but generally more or less concave posteriorly. The distance between the anterior main rib and the fission rib is 0.40-0.55 the distance between the anterior and posterior main ribs. The posterior main rib is about parallel to the girdle. In some specimens the whole list has a well-developed reticulation, in others the reticulation is more or less reduced or even totally absent. There are no accessory lists or sails. The thecal wall is areolate. The areoles are usually numerous and rather small (Figure 44:4), but their size and number are somewhat variable (Stein, 1883, pi. 21, fig. 1, 2, 5). Megacytic stages have not been recorded. 318 THE DINOPHYSOIDAE. The species is characterized by the fact that after binary fission the two schizonts remain for some time attached to each other by a ridge along the posterior portion of the dorsal margin of the anterior part of the hypotheca. This ridge sometimes is rather narrow (Figure 44:6), sometimes it is rather wide (Figure 44 : 2) ; in some specimens it is without structural differentiation ; in others it is areolate. The dmiensions of ten of our specimens and of the specimens figured by previous investigators were measured. Figure 44. — Dinophysis caxidata Saville-Kent, right lateral view, .structure of tliccal wall and of loft suleal li.st indicated only in 4. 2, two .schizonts, not yet detached from each other. 6, two schizonts, one of which is indicated only by a portion of its dorsal margin. X 430. 1, 3, from Station 4G37 (300-0 fathoms); 2, from Station 4.')94 (300-0 fathoms); 4, from Station 4,571, surface; .5, S, from. Station 4652 (300-0 fathoms); 6, 7, from Panama Harbor, surface. Dimensions: — Our specimens: Length of body, 71.8-107.0 m (average, 94.3 ix). Greatest depth of body, 29.2-56.5 m (average, 46.5 n). The specimens figured at known magnifications by previous uivestigators fall within the range of variation given above for our specunens. The figures of this species given by Stein (1883, pi. 21 , fig. 1, 2, 5) are of unknown magnifications. Judging by Stein's (1883) information about the magnifications of his figures, giA'en in an intro- SYSTEMATIC ACCOUNT. 319 ductory remark to the explanations of the plates, the length of the largest of his specimens (pi. 21, fig. 1) was somewhere between 132 ^ and 201 ijl. Variations: — As conceived in the present paper, this species is almost protean. Previous authors have subdivided it into a number of forms, varieties, or even species, but this subdivision is based on rather limited material and in- adequate knowledge of the species, and not on extensive biological and statistical investigations. The systematic value of these "forms," "varieties," and "spe- cies" is therefore not known as j-et. Future investigators with adequate material at their disposal may be able to show that at least some of these units are the expressions of genetic processes. On the other hand, it is probable that many if not all of them are modifications, due to changes in the surroundmg medium, to regulatory processes consequent upon binary fission, or to metabolic phenomena, e.g., the ratio between the rate of growth of the protoplasm and the rapidity of fission. It should be noted that in the inner region of the Aegean Sea, in the Marmora Sea, and in the Black Sea, where the salinity is low in comparison with that of the open Mediterranean, as well as in the brackish Porkljan Lake on the Dalmatian coast, this species is represented by more or less aberrant forms (Jorgensen, 1923, p. 26; Schroder, 1911, p. 36). With regard to the possible connection between some differences in the shape and size of the body and binary fission, a "degenerate" specimen of this species figured by Jorgensen (1923, fig. 37) may be significant. This specimen evidently was taken shortly after the separation of the two schizonts in binary fission. Its right (old) valve shows a striking resemblance to Pavillard's (1916, fig. 15B) Dinophysis Jwmuncxdus var. rentricosa ( = caudata xslv. acutij'ormis). Its left (new) valve, on the other hand, is decidedly smaller than the right and has a shape which is very similar to that of D. caudata var. subdiegensis Jorgensen (1923, fig. 35). If in this specimen the next fission had taken place before the new valve had reached the size of the old (supposing that such a postfission increase in size does take place, which is not known), one of the daughter individuals probably would have been referable to D. caudata var. acutij'ormis, the other to D. caudata var. subdiegensis. The smallness of the left (new) valve of this specimen probably was due to a disproportion between the rate of growth of the protoplasm and the rapidity of fission. This supposed disproportion might have been brought about either by an increased rate of fission not accompanied by a correspondingly in- creased growth of the protoplasm, or by a decrease m the growth of the proto- plasm without a proportionally low rate of fission. In the latter case we would be 320 THE DINOPHYSOIDAE. dealing with a case of degeneration, possibly called forth by adverse hydro- graphical conditions. It should be remembered in this connection that specimens of D. caudata var. acutiformis and var. subdiegensis never have been recorded except from near the coast line, where the hydrographical conditions are more or less changeable and different from those of the open sea. We should also remem- ber "the exceedmgly great sensitiveness of pelagic organisms to even slight changes in their environment" (Kofoid, 1909, p. 245). On the other hand, the specimen mentioned above may be comparable to the heteromorphic chains found in the genus Ceratium by Lohmann (1908) and Kofoid (1909), an assumption that has already been voiced by Jorgensen (1923, p. 29) . About the systematic significance and possible causes of such chains, see Kofoid (1909) and Jorgensen (1911). A third possibility is that fission was completed before the regulatory forma- tion of the new half of the theca was accomplished, and that this would have later ensued, restoring the organism to the ancestral form. The value of this specimen as evidence in the problem of the variability of this species is therefore rather uncertain. Compare also the lengths of the peduncles in the two schizonts from Karrachee represented by Czapek (1909, fig. 5). Under these circumstances it has been considered advisable in the present paper to treat most of the above-mentioned "forms," "variations," and "spe- cies" under the name of Dinophysis caudata, and, with one exception, not to assign the specimens found m the material of the Expedition and considered to belong to this species, to any one of these smaller units. The following characters are more or less strikingly variable in this species, as conceived in the present paper : — the size of the body, the inclination of the longitudinal axis of the body, the depth and the shape of the anterior portion of the body, the width, shape, and size of the pedimcle, i.e., the posterior portion of the body, the tuberculation of the antapical end of the body, and the width, shape, and structure of the left sulcal list. According to the conception of Dinophysis caudata adopted in the present paper and by Jorgensen (1923, p. 25), the specimen represented by Jorgensen's (1923) figure 30 under the name of D. caudata Kent a allieri (Gourret) f. speciosa Jorgensen may be considered as typical of this species. Kent's (1881, p. 460) original description of D. caudata is very fragmentary and is not accompanied by any illustrations. It does not suffice for a perfectly certain determination of species, but it conforms fairly well with the mentioned figui'e of Jorgensen (1923). SYSTEMATIC ACCOUNT. 321 This form may be characterized briefly in the following way : — the body is slender, and its longitudinal axis is somewhat inclined posteroventrally; the transition between the anterior portion of the body and the pedmicle is rather gradual than sudden; the peduncle is rather long, and attains about the length of the left sulcal list; the left sulcal list is broad, and decidedly broader posteriorly than anteriorly. None of the specimens of this species found in the material of the Expedition agrees completely with this form, and, as far as our records go, only one of them (Figure 44: 2) approaches it rather close lj\ The great majority of our specimens (Figure 44: 4) resemble very closely the specimen from Japanese waters figured by Okamura (1907, fig. 40a) under the name of Dmophysis homunculus f. pedunculata. In other words, they have a habitus which in several respects is intermediate between Grenfell's (1887) D. semicarinata and Schmidt's (1901) D. caudata f. pedunculata. This form may be given the following short characterization: — the body is somewhat less slender than in the typical form of this species, and its longitudinal axis is somewhat in- clined posteroventrally; the transition between the anterior portion of the body and the peduncle is rather sudden than gradual, but not quite so sudden as in Schmidt's (1901) D. caudata f. pedunculata; the peduncle is of moderate length, being somewhat shorter than the left sulcal list; the left sulcal list is rather broad, and somewhat broader posteriorly than anteriorly. Some of our specimens (Figure 44: 7) show a still more pronounced resem- blance to Dinophysis caudata f. pedunculata Schmidt than those mentioned in the last paragraph. Indeed, the agreement sometimes was almost complete. As far as our records go, the form represented by Figure 44 : 3 was found but once in the material of the Expedition. It is easily distinguished from the typical form of this species by the rather sudden transition between the anterior portion of the body and the peduncle, by the rather broad, tuberculate peduncle, and by the two tubercles on the dorsal protuberance of the body just in front of the peduncle. In the last character this form approaches Dinophysis tripos, a form which in the present paper is considered as specifically distinct from D. caudata, and which never was found in the material of the Expedition. Some of our specimens of this species (Figure 44: 5, 8) differed rather strik- ingly from the forms discussed above, and in several respects approached the forms figured by Jorgensen (1923, fig. 33, 34) under the names of Dinophysis caudata f. marmarae and D. caudata f. pontica. These specunens may be briefly characterized as follows: — the longitudinal axis of the body is somewhat in- clined posteroventrally; the anterior portion of the body is deep and bulky, and 322 THE DINOPHYSOIDAE. its transition to the peduncle is sometimes rather sudden, sometimes more or less gradual; the peduncle is more or less reduced in size, in some cases being only about half as long as the left sulcal list; the left sulcal list is rather broad. In the material of the Expedition we also have found specimens which to varymg extents are intermediate between the form (Figure 44 : 4) to which be- longs the majority of our specimens of this species on the one hand, and (Figure 44: 3) the form approaching Dinophtjsis tripos and the forms treated in the last paragraph. (See also D. caudata f. acutiformis, the section on comparisons, and Jorgensen, 1923, p. 24-27.) Comparisons: — Dinophysis tripos is the structurally closest-known relative of D. caudata. Indeed, transitional specimens between these two forms have been found (Figure 44: 3), and even their specific separation has been questioned; see- the section on synonymy. D. tripos is a connecting link, presumably genetically as well as morphologicallj', between D. caudata and D. miles. The last species occurs in at least two rather different varieties (whether or not specificallj^ dis- tinct is unknown), viz., D. miles niiles (Cleve, 1900b, tig. la) and D. miles f. schroteri (Forti) Ostenfeld. (A third variety, established by Schroder, 1906a, under the name D. miles f. maris jonii, has not been figured or sufficiently de- scribed and thus cannot be considered in this connection.) Of these two varieties the former is less excessively differentiated and therefore presumably more closely related to D. caudata than the latter. D. caudata differs from D. tripos and D. miles in not having a well-developed dorsal process just in front of the peduncle. The three species mentioned in the last paragraph occupy a rather isolated position, and we do not know which of the remahiing species of Dinojjhysis are their closest relatives. See also the section on variations and D. caudata f. acuti- formis, the section on comparisons. Synonymy: — This species, as conceived in the present paper, has appeared under several different names, which will be discussed briefly in the following paragraphs. It was first described by Saville-Kent (1881), and later described and figured by Jorgensen (1923) under the name of Dinophysis caudata. These investigators, besides Grenfell (1887, p. 559, 560), are the only ones who have used this name. The first author, after Saville-Kent (1881), to treat this species is Pouchet (1883, p. 27, pi. 18-19, fig. 5; 1885a, p. 22), who, under the name of Dinophysis acuta var. geminata, figured two schizonts not yet detached from each other. The similarity between these two specunens and D. caudata a allieri f. speciosa Jorgen- sen (1923, fig. 30), i.e., the typical form of D. caudata, is very striking. D. acuta SYSTEMATIC ACCOUNT. 323 var. geminata is regarded as identical with D. caudata by Bergh (1884, p. 385), Butschli (1885, p. 995), Paulsen (1908, p. 19), and Jorgensen (1923, p. 24). Lem- mermann (1899a, p. 373) identified it with D. allieri Gourret (see below). Kofoid and Rigden (1912a, p. 337) use the name D. geminata without discussing the sys- tematic position of this form. Shortly after the publication of Pouchet's (1883) work, referred to in the last paragraph, Dinophysis caudata was excellently figured in Stein's (1883, pi. 21, fig. 1, 2, 5, 6, 7) big iconography under the name of D. homuncidus. The name last-mentioned was accepted for this species by almost everybody, and the two older names were forgotten or disregarded. Under the name of D. homunculus this species was later figured or described by the following authors : — Whitelegge (1891, p. 184, pi. 28, fig. 14), Schutt (1895, p. 65, 112, pi. 2, fig. 8), Entz (1902a, fig. 1), Czapek (1909, fig. 5), Schroder (1911, p. 36, fig. 7b), and Pavillard (1916, p. 56, fig. 15a). ButschU (1885, p. 995, 996, pi. 55, fig. 36) reproduces Stein's (1883) Plate 21, figure 1; Paulsen (1908, p. 19, fig. 20) reproduces Stein's (1883) Plate 21, figure 2; and Forti (1922, fig. 114) reproduces Pavillard's (1916) figure 15a. Almost simultaneously with the publication of Stein's (1883) iconography, Gourret (1883) issued a paper in which were established three species belonging to the genus Dinophysis, viz., D. allieri, D. inaequalis, and D. tripos. There seems to be but little doubt that Dinophysis allieri and D. inaequalis should be referred to D. caudata. Indeed, D. allieri probably represents the typical form of this species; see Jorgensen (1923, p. 25). D. allieri Gourret (1883, p. 75, 76, 78, 79, 80, pi. 3, fig. 54, 54a) is considered as identical with D. caudata by Whitelegge (1891, p. 185), Schroder (1900a, p. 19), Pavillard (1905, p. 59), Paulsen (1908, p. 19), Okamura (1912, p. 19), Faria and Cunha (1917, p. 74), and Forti (1922, p. 108). Bergh (1884, p. 385), Lemmermann (1899a, p. 373), and Jorgensen (1923, p. 24, 25, fig. 30) treat it as a variety of D. caudata. D. inaequalis Gourret (1883, p. 77, 78, 80, pi. 1, fig. 21) is regarded as identical with D. caudata by \Mntelegge (1891, p. 185), Pavillard (1905, p. 59), Paulsen (1908, p. 19), Okamura (1912, p. 19), Faria and Cunha (1917, p. 74), and Jorgensen (1923, p. 24, 25). Bergh (1884, p. 385) regards it as a variety of D. caudata. Schroder (1900a, p. 19) is doubtful about its systematic status. The systematic position of Dinophysis tripos Gourret (1883, p. 76, 80, pi. 3, fig. 53) must be regarded as unsettled. This form is treated as a variety of D. caudata by Bergh (1884, p. 385), Lemmermann (1899a, p. 373; 1901a, p. 374; 1905a, p. 36), Ostenfeld and Schmidt (1901, p. 170), Entz (1902b, p. 94; 1905, 324 THE DINOPHYSOIDAE p. 112), Schroder (1906a, p. 322, 340), and Lebour (1917a, tab. 1; 1917b, p. 184). Paulsen (1908, p. 19) and Okamura (1912, p. 19) suggest that D. tripos and D. caudata may be identical, and Stein (1883, pi. 21, fig. 3, 4) does not separate them systematically. Some of the authors mentioned hi this paragraph change tlieir attitude in this question and treat D. tripos as a distinct species; these authors are : Ostenfeld (1906, p. 18; 1909, p. 23; 1915, p. 6; 1916, p. 13), Schroder (1911, p. 25), and Paulsen (1912, p. 290). Other investigators who treat D. tripos as a distinct species are : Pavillard (1905, p. 59; 1909, p. 284; 1916, p. 57), Cough (1907, p. 190, 192), Griif (1909, p. 155), Hensen (1911, p. 165), Schiller (1912, p. 27), Faria and Cunha (1917, p. 74), Forti (1922, p. 109), and Jorgensen (1923, p. 29, 30, fig. 38, 39). Pavillard (1909, p. 283) writes that the comparison of a great number of specunens collected during all seasons of the year "me permet de considerer comme necessaire la separation definitive de D. Homunculus Stein et D. tripos Gourret." Jorgensen (1923, p. 31) calls D. tripos: "this easily recognizable and very constant species." In the material of the Expedition a specunen (Figure 44: 3) was found which approached D. tripos f. brevicauda Jorgensen (1923, fig. 39) in the structure of the dorsal projection at the base of the peduncle, and which, on the other hand, showed a great sunilarity to some of the specimens (Figure 44 1) which in the present paper have been referred to D. caudata. In other words, specimens do exist which are more or less intermediate between D. tripos and D. caudata, as conceived in the present paper. Compare also Figure 40a and 40b, Plate 5, Okamura (1907). However, the evidence seems to be in favor of the opinion of those who regard D. tripos as a distinct species. The names of this form, therefore, have not been included in our bibliographical list of D. caudata. Dinophysis hoviunculus var. appendiculata Zacharias (1906, p. 538, 540, 574) is a synonym of D. tripos (see Faria and Cunha, 1917, p. 74, and Jorgensen, 1923, p. 29). Dinophysis homunculus var. incurva is a name apphed by Lemmermann (1899a, p. 373) to the two specunens of D. iripos figured by Stem (1883, pi. 21, fig. 3, 4); see Jorgensen (1923, p. 30). In this connection it may be mentioned that Whitelegge (1891, p. 185) sup- posed that " D. homunculus is identical with D. inaequalis Gourret, and D. allieri Gourret, and that all three are simply innnature forms of D. tripos Gourret." This statement is explained by the fact that this investigator has confused the phylogenetic and the ontogenetic processes in the genus Dinophysis. Dinophysis scmicarinata Grenfell (1887, p. 560, pi. 11, fig. 12) agrees closely with D. caudata f. pontica Jorgensen (1923, fig. 34). It is mentioned only by Jor- SYSTEMATIC ACCOUNT. 325 gensen (1923, p. 24), and should preliminarily be referred to D. caudata (see the section on the variations of this species) . The systematic status of Dinophysis homunculus f. pedunculata Schmidt (1901, p. 138, fig. 8) is somewhat doubtful. Lemmermann (1901a, p. 374), Schroder (1906a, p. 335), Okamura (1907, p. 131, pi. 5, fig. 40; 1912, p. 19), van Oye (1921, p. 215), and Jorgensen (1923, p. 25, 29) consider this form as a variety of D. caudata. Ostenfeld (1915, p. 6), on the other hand, writes that it "no doubt" is an independent species. The fact that this form seems to have a geographical distribution different from that of the other forms of D. caudata indicates that it is different genetically. It was first recorded by Schmidt (1901) from the Gulf of Siam; later van Oye (1921) reported it from Batavia, Java; Ostenfeld (1915) from the coast of Celebes; Schroder (1906a) from the Formosa Straits; and Okamura (1907, 1912) from Japanese waters. Jorgensen (1923, p. 29) foimd it on the coast of Arabia, and some of the specimens recorded by WTiite- legge (1891) from Sydney, Australia, under the name of D. homunculus, probably belong to this form. The great majority of the specimens of D. caudata found in the material of the Expedition also belong to this form. However, in spite of its peculiar distribution, it seems most advisable to refer D. pedunculata to D. cau- data until further evidence is available. See also the section on the variations. Dinophysis homunculus var. carinata Zacharias (1906, p. 540) undoubtedly refers to a recently divided specimen of D. ca^^data in which the fission ridge on the dorsal side of the body has not yet been absorbed. The two specimens figured from the Golden Horn and the Bosporus by Linde- mann (1923, p. 220, fig. 11, 12) under the name of Dinophysis homunculus var. latus are referable to D. caudata f . inarmarae and D. caudata f. pontica, respectively, figured by Jorgensen (1923) from the same waters. The reader is referred to the section on the variations of this species for our reasons for including in the bibliographical list of Dinophysis caudata the following names: — Schroder's (1911) D. homunculus var. gracilis (also mentioned by Jorgensen, 1923, p. 27) and Jorgensen's (1923) D. caudata var. allieri, D. caudata var. abbreviata,- D. caudata f . marmarae, and D. caudata f . pontica. Under Dinophysis caudata f. acutifornns will be discussed the systematic status and synonymy of the following forms belonging to or at any rate close relatives of D. caudata: — D. diegensis Kofoid and its varieties curvata Kofoid and caudata Pavillard, D. homunculus var. ventricosa Pavillard, and D. caudata var. subdiegensis Jorgensen ( = D. kofoidi Jorgensen). Reinhard (1910, p. 26) wrote Dinophysis homumnculus for D. homunculus. 326 THE DINOPHYSOIDAE. Hensen (1911, p. 164, tab. 15) used D. hormunculus and D. hornunculus for D. homunculus. Occurrence: — Dinophysis caudata is recorded at fifty of tbe 127 stations. There are 16, 23, 3, 5, 0, and 3 stations on the six hnes of the Expedition. Of these fifty stations, four (4571, 4574, 4580, 4583) are in the California Current; nine (4588, 4594, 4596, 4598, 4600, 4604, 4605, 4545, 4546) are in the Mexican Current; twelve (4609, 4611, 4613, 4615, 4617, 4631, 4634, 4637, 4638, 4639, 4640, 4644) are in the Panamic Area; nineteen (4647, 4649, 4650, 4651, 4652, 4655, 4657, 4659, 4660, 4663, 4666, 4667, 4668, 4669, 4670, 4671, 4675, 4676, Figure 45. — Occurrence of Dinophysis ctnulata Saville-Kent. Large, solid circles indicate records from vertical hauls; sq\iares, records from surface liauls; triangles, records from both vertical and sur- face hauls; small, solid circles, stations at which this species was not found; small, open circles, stations from which no plankton catches were examined. 4678) are in the Peruvian Current; two (4701, 4705) are m the South Equatorial Drift; three (4713, 4715, 4716) are in the Galapagos Eddy; one (4544) is in the North Equatorial Current. There are twenty-eight records from the surface (Stations 4571, 4574, 4580, 4583, 4588, 4596, 4600, 4604, 4611, 4615, 4631, 4638, 4639, 4640, 4644, 4649, 4650, 4657, 4659, 4660 [Salpa], 4666, 4669, 4676, 4678, 4716, 4544, 4545, 4546); at eight of these stations (4571, 4574, 4580, 4583, 4638, 4650, 4657, 4659) the species was taken in hauls from 300-0 fathoms as well as at the surface; and at one station (4676) it was taken in hauls from 300-0 fathoms and 800-0 fathoms as well as at the surface; at the remaining nineteen stations it SYSTEMATIC ACCOUNT. 327 was taken in surface hauls only. At one station (4652) the species is recorded from 100-0 fathoms; at one station (4655) m hauls from 300-0 fathoms and 400-0 fathoms; and at three stations (4647, 4651, 4670) in hauls from 800-0 fathoms. All the remaining records refer to hauls from 300-0 fathoms onlj'. The species is also recorded from surface waters in Panama Harbor, in the Panamic Area. This station is not included in the 127 stations mentioned above. The temperature range of these fifty stations at the surface was 65°-85°; the average was 74.3°. At the twenty-eight stations in the surface catches of which this species was found, the surface temperature ranged from 67° to 84°; the average was 75.1°. For the surface catches the following frequencies are recorded: — 72% at one station (4571), 29% at one station (4611), 11% at one station (4649), 10%, at one station (4669), 9% at one station (4639), 6%, at one station (4615), 5% at one station (4640), 3% at three stations (4657, 4659, 4676), 2% at three stations (4644, 4650, 4544), 1 % at four stations (4574, 4588, 4666, 4716), and less than 1 % at the remaining stations. For the catches from 100, 300, 400, or 800 fathoms to the surface the records of frequency are as follows: — 12% at one station (4659), 9% at one station (4571), 3% at one station (4650), 2% at nine stations (4574, 4598, 4609, 4638, 4651, 4655, 4671, 4675, 4676), 1% at nine stations (4613, 4617, 4637, 4647, 4657, 4663, 4667, 4670, 4715), and less than 1% at the remaining stations. In the catch from Panama Harbor thu'ty specimens were found. This species was first recorded by Saville-Kent (1881) from the Adriatic Sea. From the North Sea it is reported by Cleve (1900e, 1903a), Apstein (1905a), Clark (1905), Ostenfeld (1906), and Paulsen (1912). It has never been found in the Baltic. Lemmermann's (1902a, p. 263) statement that Mobius (1887) re- corded it from the western part of the Baltic is erroneous. Jorgensen (1905) found it near Bodo, on the west coast of Norway. It is mentioned as occurring in the Norwegian Sea by Cleve (1902a), Gran (1902), Ostenfeld (1906), Linko (1907), and Paulsen (1912). From the northern part of the Atlantic {i.e., west of Scotland, south of the Faroe Islands and of Iceland, and in the Irminger Sea) it is reported by Hensen (1887, 1911), IMobius (1887), Ostenfeld (1898b, six stations; 1899, seven stations), and by Paulsen (1904, 1912). In the English Channel it was found by Cleve (1899a, 1900e) and Ostenfeld (1916b); on the west coast of France by Mangin (1912) and Jorgensen (1923); on the west coast of the Py- renean peninsula by Jorgensen (1923); and west of Gibraltar by Graf (1909). In the northern and middle parts of the Atlantic it has been recorded by the fol- lowing mvestigators: — Cleve (1897a), at lat. 49° 53' N., long. 28° 13' W., and at 328 THE DINOPHYSOIDAE. lat. 37° 47' N., long. 47° 31' W.; Cleve (1901c) at about fifty stations between lat. 63° N. and lat. 10° N.; Cleve (1902b), at four stations between lat. 39° N. and lat. 21° S.; Cleve (1903b) at lat. 36° N., long. 6° W., and at lat. 37° N., long. 8° W.; Ostenfeld (1898a) at lat. 35° 44' N., long. 38° 03' W., and at the Canary Islands; Murray and Whitting (1899) at forty stations between lat. 48° N. and lat. 20° N.; Zacharias (1906) near the Azores and at lat. 31° N., long. 38° W.; Karsten (1906) at six stations between lat. 33° N. and 1° N.; Lohmann (1908a) between lat. 39° N., long. 46° W., and lat. 35° N., long. 53° W.; Stuwe (1909) at lat. 28° N., long. 40° W.; and Hensen (1911) at a great number of stations in the Labrador Current, Gulf Stream, Florida Current, Sargasso Sea, Canary Cur- rent, North Equatorial Current, Guinea Current, South Equatorial Current, and oflf the northeastern coast of Brazil. In the Caribbean Sea and among the West Indies this species has been found by Grenfell (1887), Murray and Whitting (1899), who record it at fourteen stations, and by Cleve (1901c) and Schroder (1909). From the South Atlantic we have the following records: — Cleve (1900f), "not rarely, in almost all samples from 37° S., 23° W. to 44° S., 4° E."; Cleve (1901c) off Rio de Janeiro; Cleve (1902b) at lat. 38° S., long. 20° W. to 44° S., long. 4° E.; Karsten (1906) at eleven stations between lat. 5° S. and lat. 35° S.; and Faria and Cunha (1907), Cunha and Fonseca (1907), off Rio de Janeiro. General statements about the occurrence of this species in the Atlantic are to be found in Stein (1883), Cleve (1899c, 1900f, 1901c), and Cleve, Ekman, and Pettersson (1901). In the Mediterranean this species has been observed on a great nmnber of occasions, viz.: at lat. 36° N., long. 5° W., by Cleve (1903b); at lat. 36° 46' N., long. 2° 25' E., by Ostenfeld (1902); in the Gulf of Lyons by Pouchet (1883), Gourret (1883), and Pavillard (1905, 1909, 1916); off Monaco by Nathansohn (1908, 1909, 1910a) ; in the Ligurian Sea by Forti (1922) ; ofT Naples by Daday (1888), Schroder (1900a), and Zacharias (1906); off Palermo and Algiers by Zacharias (1906); in the Adriatic Sea by Saville-Kent (1881), Stein (1883), Imhof (1886), Entz (1902b, 1905), Schroder (1906a, 1911), Zacharias (1906), Schiller (1911a, 1911b, 1912), and Carazzi and Grandori (1912). Jorgensen (1923) found it to be comparatively well represented throughout the Mediterra- nean. From the Marmora Sea, Golden Horn, and Bosporus it has been recorded by Zacharias (1900), Lindemann (1923), and Jorgensen (1923). In the Black Sea it was found by Minkiewitsch (1899), Reinhard (1910), and Jorgensen (1923). From the Red Sea the species has been reported by Ostenfeld and Schmidt (1901), Cleve (1903b), and Karsten (1907); from the Gulf of Aden, by Ostenfeld SYSTEMATIC ACCOUNT. 329 and Schmidt (1901), Cleve (1903b), Schroder (1906a), and Karsten (1907); from "the coast of Arabia," by Jorgensen (1923, p. 29); from the Arabian Sea, by Ostenfeldand Schmidt (1901), Cleve (1903b), Schroder (1906a), Karsten (1907), and Czapek (1909); from the Indian Ocean, by Cleve (1900f, 1901a), Karsten (1907), and Graf (1909). In the Antarctic Ocean, south of South Africa, it was found by Karsten (1905) at two stations, viz., at lat. 37° 31' S., long. 17° 1' E., and at lat. 41° 5' S., long. 14° 51' E.; and by Graf (1909) at three stations, viz., lat. 39° 56' S., long. 20° 7' E., lat. 43° 39' S., long. 23° 30' E., and lat. 42° 21' S., long. 31° 0' E. Lemmermann (1899a) found it m the Peruvian Current. Cleve (1901a) recorded this species from the Malay archipelago; van Oye (1921) from Batavia, Java; Graf (1909) from the north coast of New Guinea; Schmidt (1901) from the Gulf of Siam ; Ostenfeld (1915) from the coast of Celebes; Schroder (1906a) from the Formosa Strait; Okamura (1907, 1912) from Japanese waters; Stein (1883) from the "Siidsee"; and Whitelegge (1891) from Sydney, on the east coast of Australia. Karsten (1907) reported the species to have been taken with a closing net between 200 and 100 meters. Most of the remaining records refer to surface hauls. Some investigators, e.g. Jorgensen (1923), have found this species to be "remarkably numerous" or even "abundant" at some localities. According to the investigators who have contributed to our knowledge of the distribution of this species, it occurs in waters of the following temperatures and salinities. Ostenfeld (1898b): — temperature: range, 47.5°-53.6°; mean of six observations, 50.7°. Salinity: range, 35.25-35.50; mean of six observations, 35.39. Ostenfeld (1899): — temperature: range, 48.6°-53.4°; mean of seven ob- servations, 50.6°. Salinity: range, 35.33-35.55; mean of seven observations, 35.43. Ostenfeld (1902) : — temperature, 73.4°; salinity, 36.9. Murray and Whitting (1899) : — temperature : range, 55°-86°; mean of 53 observations, 71.0°. Ostenfeld and Schmidt (1901) : — temperatures, 80.1° and 79.1°. Cleve (1901a) : — temperatures, 53.9° and 80.3°; salinities, 33.71 and 34.69. Cleve (1901c) : — temperature: range, 42.8°-81.0°; mean of 100 observations, 65.3°., Salinity: range, 31.01-36.54; mean of 77 observations, 35.54. Cleve (1902a) : — tempera- ture, 48.2°; salinity, 35.42. Cleve (1902b): — temperatm-e, 61.4° and salinity, 35.77 (means of 26 observations). Cleve (1903b) : — temperature: range, 60.3°- 84.7°; mean of seven observations, 73.2°. Salinity: range, 35.77-42.81; mean of seven observations, 37.95. Cleve, Ekman, and Pettersson (1901): — average temperature, 65.3°; average salinity, 35.54. Jorgensen (1905) : — temperatures, 330 THE DINOPHYSOIDAE. 39.7° and 39.9°. Stuwe (1909) : — temperature, 75.2°. Graf (1909) : — salinity: range, 34.09-36.33; mean of nine observations, 35.13. Van Oye (1921): — salinity, 31.5. Of the authors mentioned in connection with the distribution of this species, only the following have given descriptions or figures, by means of which their detei'minations of this species may be checked: — Saville-Kent (1881), Pouchet (1883), Stein (1883), Gourret (1883), Entz (1902a), Czapek (1909), Reinhard (1910), Schroder (1911), Pavillard (1916), Forti (1922), Lindemann (1923), and Jorgensen (1923). Dinophysis caudata is eupelagic and occurs in tropical, subtropical, and warm-temperate regions of all seas. It has been found in the northern part of the Atlantic, in the North Sea, and in the Norwegian Sea, but it is probably a warm- water species and not endogenetic in these relatively cool regions. The records of the International Council for the Investigation of the Sea "decidedly support the notion that the specimens found in these northern waters are of southern origin, carried to the north by the Gulf Stream" (Paulsen, 1912, p. 290). The species is eurythermal, with an exceptionally wide amplitude of thermal adaptation, viz., 39.7°-85.0°, and euryhaline, occurring in brackish water as well as in waters with a salinity as high as 42.81% (see Jorgensen, 1923, p. 27). It is probably one of the most common representatives of this family (see Cleve, 1901c, p. 240, and Jorgensen, 1923, p. 27). Its distribution in the Eastern Pacific, according to our data, is very peculiar, since almost all the record stations are located in the coastal currents. Its almost complete absence from the South Equatorial Drift is not directly due to the temperature, since this region is thermally intermediate be- tween the relatively cool Peruvian Current and the very warm Mexican Current and Panamic Area, where the species is rather common. The comparatively frequent occurrence of this species in our surface catches (twenty-eight surface records) is another outstanding feature. Of D. hastala, which next to D. caudata is the most common species of this genus in our surface catches, we have only eight surface records. Furthermore, it should be remembered that some of our surface records of D. caudata showed exceptionally high frequencies. Dinophysis caudata Saville-Kent f. acutiformis, nom. f. nov. Figure 46 Dinophysis homunculus var. venlricosa Pavillard, 1916, p. 56, fig. 1.5B. Forti, 1922, p. 108, 190, 208, fig. 115. Dinophysis caudata v.ar. ventrico.ia Jorgensen, 1923, p. 25, 27-29, fig. 32, 37. Diagnosis: — This form, which is rather variable in shape, differs from the main species chiefly in its smaller size and its less-developed peduncle. The body SYSTEMATIC ACCOUNT. 331 measures 66.7-86.6 m in length. The peduncle sometimes is fairly well developed, sometimes it is hardly distinguishable ; when best developed, its length is about 0.25 the length of the body. The eingular and sulcal lists, as a rule, have less de- veloped structural differentiation than in the main species. Coastal waters in the Atlantic, Mediterranean, Arabian Sea, and the eastern tropical Pacific. Dimensions: — Our specimens: Length of body, 71.5-86.6 n (average, 81.4 y). Greatest depth of body, 38.5-47.4 n (average, 44.2 ii). The type specimen, as figiu'ed by Pavillard (1916, fig. 15B), was 82.5 ^ long and 41.8 m deep. The speci- mens figured by Jorgensen (1923, fig. 32, 37) were 66.7-74.0 m long and 32.0-38.1 m deep. Figure 46. — Dinophysis ca-udata Saville-Kent f acutiformis, nom. f. nov., right lateral view. 2, two schizonts, not yet detached from each other; 3, two schizonts, one of which is indicated only by a portion of its dorsal margin. X 430. Station 4673 (300-0 fathoms). Variations: — Judging by the seven specimens as yet figiu'ed, this form is rather ^'ariable in the size of the body, m the depth and shape of the anterior portion of the body, and in the size and shape of the peduncle. The anterior portion of the body may be rather deep and bulky (Jorgensen, 1923, fig. 32; our Figure 46: 3) or relatively slender (Figure 46: 2), and the peduncle may be well developed (Pavillard, 1916, fig. 15B; Jorgensen, 1923, fig. 32) or hardly distin- guishable (the left specimen m Figure 46 : 2) . Comparisons: — The specimens found in the material of the Expedition and in the present paper referred to Dinophysis caudata f. acutiformis agree rather closely with Pavillard's (1916, fig. 15B) figure of the type specunen of this form. They differ from the type and from the specimens figured by Jorgensen (1923, fig. 32, 37) mainly in having, to varying degrees, somewhat less-developed ped- uncles. This form has an uncertain systematic status. It may be a distinct species; or it may be a subspecies of Dinophysis caudata as assumed m this paper. The evidence bearing on this question is anything but conclusive. Our tentative deci- 332 THE DINOPHYSOIDAE. sion is based on the following facts: — (1) Wliile the typical D. caudata is rather commonly found in waters far from the coast as well as in coastal waters, D. caudata f. acutiformis has never been recorded except from near the coast line, where the hydrographical conditions are more or less changeable and different from those of the open sea; (S) a series of specimens of D. caudata has been found, which is characterized by a gradual decrease in the size of the peduncle. As far as available records go, such specimens, too, have not been taken except near the coast; (3) Jorgensen (1923, fig. 37) has figured a specimen in which one valve has a fairly well-developed peduncle and the other valve only a slight indication of a peduncle. For further discussion of the systematic significance of this specimen, see D. caudata, the section on variations. The systematic status of the following two species must also be regarded as uncertain : — Dinophysis diegensis Kofoid and its varieties curvata Kofoid and caudata Pavillard (the name of the last variety is preoccupied and therefore must be changed), and D. subdiegensis Jorgensen. These two species have been sup- posed to be nothing but modifications of D. caudata, but they may be independent species. Just as in the case of D. caudata f. acutiformis, the evidence bearing on the question is of uncertain value. It should be mentioned that, just as D. caudata f. acutiformis, these two species are not known to have been taken except near the coast line. Dinophysis diegensis Kofoid is considered a distinct species by Kofoid (1907b, p. 313, pi. 33, fig. 57, 59, 60), Ostenfeld (1915, p. 7), Pavillard (1916, p. 57), and, with hesitation, by Jorgensen (1923, p. 28, fig. 36). It is held as a de- generate form of D. caudata by Jorgensen (1923, p. 22, 25, 27; see also Forti, 1922, p. 109). D. diegensis var. ciirvaia Kofoid (1907b, p. 313, pi. 33, fig. 58, 61) and D. diegensis var. caudata Pavillard (1916, p. 57, pi. 3, fig. 2; 1923a, p. 881) are considered as degenerate forms of D. caudata by Jorgensen (1923, p. 27, 28). Jorgensen (1923, p. 27, 28, 43, 44, fig. 35) is, at present, the only one who has treated Dinophysis caudata var. subdiege7isis. He first established this form as a distinct species under the name of D. kofoidi; then, on page 28 of the mentioned paper, he changed his opinion and treated it as a degenerate form of D. caudata under the name of D. caudata var. subdiegensis (the last name has the pri- ority). WTiatever the systematic status of Dinophysis caudata f. acutiformis, D. diegensis, and D. subdiegensis may be, there seems to be but little doubt that they are very closely related to D. caudata. Indeed, these forms most probably have overlapping ranges of variation. D. caudata f. acutiformis is in several respects SYSTEMATIC ACCOUNT. 333 intermediate between the tj-pical D. caudata on the one hand and D. diegensis and D. subdiegensis on the other. Synonymy: — This form was estabUshed by Pavillard (1916), who described and figured it under the name of Dinophysis homunculus var. ventricosa. Later Jorgensen (1923, fig. 32) figured it under the name of D. caudata var. ventricosa and (1923, fig. 37) as "a degenerate form" of D. caudata. Forti (1922, fig. 115) reproduced Pavillard's (1916) figure of the type specimen. Unfortunately, the name ventricosa, suggested by Pavillard (1916), must be rejected on account of preoccupation by Dinophysis ventricosa Claparede and Lachmann (1858, p. 408). Occurrence: — Dinophysis caudata f. acutiformis is recorded at only one (4673) of the 127 stations, on the second Ime of the Expedition, in the Peruvian Current, and very near the coast. The depth is 300-0 fathoms, the surface temperature 67°, and the frequency less than 1% (three specimens). This form was first recorded by Pavillard (1916) from the Etang de Thau, an enclosed and very shallow part of the Gulf of Lyons. Later it was found by Forti (1922) in the Ligurian Sea, and by Jorgensen (1923) in the harbor of Lisbon, "m the innermost, less salme region" of the Mediterranean, and on the coast of Arabia, near Ras Dschibsch. There are no records of this form from waters far from the coast line. HisTioPHYSis, gen. nov. Dinophysis Kofoid & Michener, 1911, p. 293 partim. Diagnosis: — Body ovate m lateral outline. Epitheca low, about 0.26 as deep as hypotheca. Dorsally transverse furrow is about 0.19 the greatest depth of body and about twice as wide as ventrally. Posterior cingular list about 0.11 the length of body from apex. Anterior cingular list funnel-shaped, ribbed; its dorsal height about 0.30 the greatest depth of body. Posterior somewhat nar- rower and more inclined than anterior, reticulate but not ribbed. Right sulcal list long, narrow, and reticulated. Left sulcal list ends somewhat ventrally to antapex and increases gradually in width posteriorly, where it is angular and about 0.46 the greatest depth of body; the entire list reticulated; with no mam rib behind fission rib. The type species is Histiophysis rugosa. 334 THE DINOPHYSOIDAE. HiSTioPHYSis RUGOSA (Kofoid and Michener) Plate 5, fig. 5. Figure 93:1 Dinophysis rugosa Kofoid & Michener, 1911, p. 293. Diagnosis: — Body broadly ovate in lateral view; length: depth, 1.08:1. Posterior cingular list about 0.11 the length of body from apex. C'ingular lists funnel-shaped; anterior ribbed, 0.30 the depth of body; posterior finely reticulate, somewhat narrower. Left sulcal list ends somewhat ventrally to antapex; in- creases gradually in width posteriorly; dorsal margin gently sigmoid, 0.4G the depth of body, forming acute angle of 60° with ventral margin; with only one rib, the fission rib, which is 0.20 the depth of body; entire Ust reticulate; meshes de- creasing in size posteriorly, where they are very small. Theca finely and deeply areolate. Length, 45.7 ju. Eastern tropical Pacific. Description: — This is a medium-sized species, the body of which is broadly ovate in lateral ^'iew, broadly rounded posteriorly and rather narrowly rounded anteriorly. It is deepest just behind the middle, and the ratio between the length and the depth is about 1.08: 1. The longitudinal axis is subperpendicular, slightly deflected posterodorsally. The epitheca is about 0.26 as deep as the hypotheca; its shape is unknown but is probably gently convex. Dorsally the transverse furrow is about 0.19 the greatest depth of the body and about twice as wide as ventrally ; its shape is un- known. The posterior cingular list is about 0.11 the length of the body from the apex. The hypotheca is subcircular, somewhat narrower anteriorly than poste- riorly, and its margins are confluent. The anterior cingular list is sessile, with a well-developed fuimel shape and an anterior inclination of about 35°; it flares about twice as much dorsally as ventrally, with the dorsal side more concave than the ventral; with flat, slightly incUned top; its height is about 0.30 the greatest depth of the body, and it has on each valve about 11-13 ribs; most of these are complete and simple, others are incomplete or anastomose proximallj'. The posterior cingular list is somewhat narrower and more inclined than the anterior; it lacks ribs but is finely and rather heavily reticulated throughout. The right sulcal list is long, narrow, and finely and rather heavily reticulated throughout; it extends to the fission rib of the left sulcal list, i.e., to a point about two thirds the way from the apex to the antapex; it is of subuniform width throughout, and its width about equals 0.06 the greatest depth of the body. The left sulcal list ends somewhat ventrally to the antapex SYSTEMATIC ACCOUNT. 335 and increases gradually in width posteriorly. Its ventral margin is strongly and somewhat irregularly convex except anteriorly and posteriorly, where it is gently concave. Its dorsal margm (postmargin) has a postero ventral inclmation of about 25°, is gently sigmoid, concave anteriorly and convex posteriorly, forms an acute angle of about 60° with the ventral margin, and its length is about 0.46 the greatest depth of the body. The list has only one rib, the fission rib, located at a point about two thirds the way from the apex to the antapex, deflected posteriorly at about 20°, and measuring 0.20 the greatest depth of the body. In front of this rib the list is irregularly and rather coarsely reticulated, most of the meshes being elongated transversely, and the margin is finely undulatmg. Behind this rib the list is finely and rather heavily reticulated, the meshes decreasing in size poste- riorly, and the margin is not undulating but smooth. The thecal wall is finely, deeply and regularly areolate. Across the middle of each valve there are about twenty-two areoles. Some phaeosomes were found in the girdle of the type specunen. The dimensions of one specimen only, the type, were measured. Dimensions: — Length of body, 45.7 n. Greatest depth of body, 42.8 m- Comparisons: — Our description and figure are based on the type specimen. Synonymy: — This species was established by Kofoid and Michener (1911) under the name of Dinophysis rugosa. Occurrence: — Histiophysis rugosa is recorded from only one (4705) of the 127 stations, on the fourth line of the Expedition, in the South Equatorial Drift, from 300-0 fathoms, and at a surface temperature of 72°. The frequency is less than 1% (one specunen). The species has been found only in the material of the Expedition. It was first recorded by Kofoid and Michener (1911) from Station 4705 of the Expedi- tion, which thus is the type locality. 2. AMPHISOLENIDAE, fam. nov. Diagnosis: — Body divided, for purpose of description, into five regions: — head, neck, anterior process, midbody, and antapical (or antapicals). Sometimes strikingly elongated, straight, or somewhat sigmoid; with two long and narrow- extensions arising from midbody, one anterior, made up of head, neck, and an- terior process, the other posterior, forming the antapical; and with longitudinal axis of body passhig through regions from which anterior process and antapical originate. Sometimes seemingly tripartite, with three long and narrow extensions arising from midbody, one anterior, made up of head, neck, and anterior process, 336 THE DIXOPHYSOIDAE. and two posterior, the antapicals, one dorsal, the other ventral in origin ; the three extensions generally subequal, approximately balanced, and their length 1.0-3.4 the dorsoventral diameter of midbody measured between bases of antapicals. Head encircled by girdle; its axial length 0.11-1.00 the length of neck; its dorsoventral diameter (width) 0.5-10.0 its axial length. Epitheca very small; its axial length less than half that of head; its anterior face convex to concave. Neck elongated, cylmdrical or slightly flattened bilaterally, 0.02-0.28 the length of body and 2.5-22.0 times longer than wide. Anterior process somewhat wider than neck ; when set off from midbody, it is about as long as, or generally some- what shorter than, neck. Midbodj' more or less compressed bilaterally, in lateral view of varying shapes, fusiform, ovoidal, subtriangular, subrotund, subellip- soidal or sack-shaped. Antapical (or antapicals) of varying lengths, 0.9-33.0 the length of neck, and about as wide as or slightlj^ wider than neck. Distal portion of girdle not displaced posteriorly. Transverse furrow flat or somewhat convex or concave, of subuniform width throughout and somewhat narrower than and slightly oblique to head. Longitudinal furrow at most but sUghtly impressed, extendmg from girdle to base of neck; it does not extend beyond the girdle, i.e., into the epitheca. Cingular Usts subequal or nearly so, of moderate height, 0.5-3.0 the width of transverse furrow, hyalme, and generally ribbed, the anterior one with midventral gap. Sulcal lists also of moderate height, extendmg to or somewhat beyond flagellar pore, simple, less differentiated than in other families, hyaline, and frequently without ribs; fission rib, when present, near flagellar pore. Flagellar pore displaced to posterior end of neck. Transverse flagellum extending from flagellar pore into girdle and around head. Longitudinal flagellum not observed with certainty. Thecal wall in most cases apparently without structure, seldom striate, pitted or locally reticulate. Nucleus located in midbody. Pusule developed, opening into flagellar pore. Chromatophores, if present, pale j^ellowish green, l^rownish, or colorless. Length of body ranging from 92 m to 1520 ^l. Marme and eupelagic, probably of wide distribution in tropical, subtropical, and warm-temperate waters. Subdivisions. Relationships between the Genera In the present paper all the known species of the family Aniphisolenidae have been assigned to either of two genera, Amphisolenia and Triposolenia. These two genera, which probably represent natural systematic units, are sharply set off SYSTEMATIC ACCOUNT. 337 from each other. Ahhough, at the first glance, they appear to present two rather different types, there seem to be good reasons in favor of the supposition that they are closely related. Their close relationship is indicated by the following struc- tural characteristics. (1) The anterior portion of the body, which for the purpose of description is diA'ided into three regions, head, neck and anterior process, is pro- longed and attenuated. (2) The head as a rule is short and is surrounded by the girdle. The epitheca is therefore very short. (S) The neck is comparatively long and narrow. (4) The flagellar pore is displaced to the posterior end of the neck. (5) The transverse furrow is somewhat narrower than and slightly oblique to the head. The longitudinal furrow extends from the girdle to the posterior end of the neck and does not extend into the epitheca. (6) The cingular lists are subequal or nearly so, of moderate height, 0.5-3.0 the width of the transverse furrow, hyaline, and generally ribbed. The sulcal lists are also of moderate height, simple, less differentiated than in the other families of this tribe, hyaline, frequently without ribs, and extend to or somewhat beyond the flagellar pore. The fission rib, when present, is near the flagellar pore. (7) Around the flagellar pore a cytopharyngeal protuberance generally is developed. (5) The portion of the body which is just behind the anterior process is inflated and contains the nucleus. (9) The posterior portion of the body is prolonged and attenuated. {10) In Triposolenia the middle portion of the body is always bifurcated. In the most highly differentiated species of Amphisolenia, viz., those belonging to the Bifurcata group, the posterior portion of the body has one or more branches. This constellation of char- acters differentiates these two genera from all other known genera of Dino- physoidae. The distinctive characters of Triposolenia, when compared with those of Amphisolenia, are: — (1) the shape of the strongly bilaterally compressed mid- body, (3) the presence of two antapical processes arishig from the midbody, one dorsal, the other ventral in position, both in most species being almost similar in size and the one the mirror image of the other, and (3) the fact that the longi- tudinal axis of the body does not pass through the region of the midbody from which arises the antapical process that corresponds to the original posterior por- tion of the body. The relation which the midbody in Triposolenia bears to its three approximately balanced processes, viz., the anterior process with the neck and the head, and the two antapicals, strikingly distinguishes this genus from all the other known genera of Dinophysoidae. In Amphisolenia, {!) the midbody in most cases is decidedly less compressed bilaterally than in Triposolenia, {2) only one antapical process arises from the midbody, and (3) the longitudinal axis of 338 THE DINOPHYSOIDAE. the body passes through the region of the midbody from which the antapical originates. With regard to the shape of the midbody transitional types between these two genera do exist, e.g., AmpJmolenia inflata; but with regard to the number and position of antapical processes no transitional types have been found as yet, but accessory antapical processes in Amphisolenia arise solely from the antapical and not from the midbody. There seems to be but little doubt that, among the species of Amphisolenia hitherto described, A. inflata isthe one that is most closely related to Triposolenia. It can be regarded as ec^ually certain that, among the known species of Tripo- solenia, T. truncata is the one most closely related to Amphisolenia. These two species show important similarities in the following respects: — (1) the concavity of the anterior face of the epitheca, a character not found in any other species of these two genera; (2) the very distinct structural differentiation of the thecal wall which is almost identical in the two species iji spite of its relative complexity and is unique as far as these two genera are concerned; (3) the well set off anterior process, and tlie relatively large, bilaterally compressed midbody; (4) the absence of antapical tubercles; (5) small size of the body. The coexistence of all these characters is most probably not the result of convergence but of commt)n in- heritance. If this statement is correct, then Amyhisolenia inflata represents a type rather similar to the ancestral type of the genus Triposolenia; and Triposo- lenia truncata is the most primitive of all the known members of its genus. On the other hand, it does not follow that Amphisolenia inflata is necessarily the most primitive species of its genus. It is, of course, possible that Triposolenia originated from Amphisolenia after the latter had passed through a period of differentiation. The position of Amphisolenia inflata in the natural system must be judged in the light of facts gathered from a comparative morphological study of Dinophysoidae. Such a study shows that tliis species occupies a rather isolated position in Amphisolenia, and that it probably is one of the most primitive, pos- sibly the most primitive, of the known species of this genus. The following of its characters appear to be more or less primitive: — {1) the small size of the body; {2) the straight longitudinal axis of the body; (3) the exceptionally large, bilater- ally compressed midbody; (4) the simple, unbranched antapical; (5) the well- developed areolation and reticulation of the thecal wall; {6) the reticulation of the sulcal lists; (7) the continuation of the sulcal lists beyond the flagellar pore to the posterior end of the anterior process. The question as to whether or not the con- cavity of the anterior face of the epitheca of this species is primitive cannot be SYSTEMATIC ACCOUNT. 339 settled as yet. It is a peculiarity of this species and of Triposokma truncata which has not been found in any of the primitive genera of Dhiophysoidae. We also leave the question open as to whether or not the peculiarity of havuig the anterior process well set off from the midbody is a prunitive feature. The result of this analysis is therefore : — (/) that the genera Amphisolenia and Triposolenia are more closely related to each other than to any other genus of this tribe ; {2) that Triposolenia probably originated from an ancestral form of about the same habitus as Amphisolenia inflata; and (3) that of the known species of these two genera Amphisolenia inflata and Triposolenia truncata probably are the most primitive. Did Triposolenia originate from Amphisolenia by a sudden duplication of the antapical, or by a gradual development of the accessory antapical, in other words, in the same way as Dinophysis miles Cleve probably arose from the Dinophysis caudata type? This problem cannot be settled as yet. The first alternative appears to be supported by the fact that the antapicals are better balanced in Triposolenia truncata than in most of the remaining species of this genus. Furthermore, sudden duplication of parts is a phenomenon frequently met with in Natm-e. The second alternative is not supported by the shape and the relative size of the antapicals m any of the known species of Triposolenia. On the other hand, the geniculation at the junction of the anterior process and the neck may possibly indicate that the anterior portion of the body, in order to maintain a state of equilibrium, may have been tilted first ventrally and then dorsally during a gradual grow th of the acces- sory antapical. The ventral antapical, which in this case would correspond to the original posterior portion of the botly, for the same reason may ha\e shifted ventrally at the same time. Distribution The two genera of the family Amphisolenidae probably are exclusively marme and eupelagic and seem to be confined to waters of tropical, subtropical, and warm-temperate nature or origin. Both of them probably have a world-wide distribution in these waters, but as yet the genus Triposolenia has been recorded only from the eastern portion of the Pacific, from the southwest coast of Portugal, from the Mediterranean and from the Marmora Sea. Most of the species appear to be rare. Some of the species of Amphisolenia have been recorded rather frequently from surface waters, but the optimum habi- tat of both genera appears to be in the deeper levels of photosynthesis, predomi- nantly above 200 meters. 340 THE DINOPHYSOIDAE. Key to the Genera 1. Midbody with one ant;i])ical which is sometimes divided distally AmpliisolenLa Stein, 1883. 2. Midbody with two autapicals Triposolenia Kofoid, 1907. Amphisolenia Stein Amphisolenia Stein, 1883, p. 24. BOtschli, 1889, p. 94.5, 1010, SchCitt, 1896, p. 28. Delage & HArouard, 1896, p. 385. Diagnosis: — Body strikingly elongated, straight or somewhat sigmoid, often asymmetrical posteriorly. With two long and narrow extensions arising from midbody, one anterior, made up of head, neck, and anterior process, the other posterior, forming the antapical. Longitudinal axis of body passing through re- gions from which anterior and posterior extensions originate. In some species the posterior part of the antapical has one to four ventral branches. The type species is Avifhisolenia globifera. Organology: — While for the sake of simplicity in quantitative determinations of structural features, in Triposolenia a more or less arbitrary longitudinal axis of the body has to be chosen, in Amphisolenia the original and, for most genera of Dinophysoidae, the typical axis can be utilizetl. In the species of Amphisolenia with an unbranched antapical, the longitudinal axis of the body either is straight or nearly so, or it has a slight or moderate sigmoid curvature, with the neck somewhat deflected dorsaliy and the antapical more or less concave ventrally. In some of these species the posterior portion of the antapical is bent ventrally, as in ^. astragalus (see this species, the section on comparisons), or to the right, as in A. clavipes and A. lemmermanni, at an angle of about 25°-50°. In the species with a branched antapical, e.g., in A. quinquecauda, the antapical often is char- acterized by a more or less strikmgly developed sigmoid curvature in the dorso- ventral plane; in other words, the longitudmal axis of the body may have a double sigmoid curvature. Sometunes the diversifications in the longitudinal axis of the body are accompanied by sinistral torsion. The body is strikingly elongated, with two long and narrow extensions aris- ing from midbody, one anterior made up of head, neck, and anterior process, the other posterior, forming the antapical. Just as in Triposolenia, the body in Amphisolenia frequently is characterized by a combination of asymmetry, both bilateral and dorsoventral, with a condition of balance. The head {hd., Figure 47) is composed of the very short epitheca, the girdle region, and a short, generally expanded region at the anterior end of the neck. It is encircled by the transverse furrow, which covers about two thirds or even more of its axial length. Its form and size ^•ary rather strikingly in the diff'erent species. It is bilaterally compressed, and its transdiameter is not much greater SYSTEMATIC ACCOUNT. 341 Mlt. p m. h. ' 1 int. cing. I. . post. cing. I. . r. sulc. I. I. sulc. I. than that of the neck. In most species it is small, its axial length being between 0.11 and 0.30 the length of the neck and its dorsoventral diameter (its "width," according to the terminology utilized in this paper) between 1.0 and 4.0 its axial length. In Amphisolenia inflata it is 0.38-0.50 the length of the neck and about twice as wide as long. In A. laticincta it is about 1.0 the length of the neck and 0.5 as wide as long. In A. asymmctrica it is 0.13-0.14 the length of the neck and not less than 9-10 times wider than long. In lateral view, the anterior face of the head, i.e., the epitheca, is more or less con- vex in most species ; sometimes it is flat and in one species, A. inflata, even quite con- cave. In a few species the head is sub- horizontal, but generally it is more or less inclined, about 5°-45°, anteriorly, that is, its dorsal end is elevated. The neck in.. Figure 47) is an elon- gated, slender, and almost cylindrical ex- tension of the anterior process. Its length, measured in a straight line from the pos- teroventral point of the head to the flagellar pore is 0.02-0.26 the length of the body. It is 2.5-22.0 times longer than wide, of nearly uniform caliber throughout, almost straight or gently curved dorsally, and gen- erally sHghtly, l°-20°, deflected dorsally from the longitudinal axis of the body. In most species the neck passes more or less abruptly into the h.GSid;m Amphisolenia lati- cincta, which has a head but slightly wider than the neck, the transition is gradual. The anterior process {ant. p., Figin-e 47), or the proabdomen according to Stein's (1883) terminology, is an anterior ex- tension of the midbody. Contrary to the ant Figure H .— Amphisolenia inflata Mur- ray and '^'hitting, right lateral view. X 52.5. Station 4681 (300-0 fathoms). The length of body is measured in a straight line from fore- most point of head to tip of antapical (in species with branched antapical, to tip of main stem of antapical). Our method of measuring the (a.xial) length of head is indicated by bi- capitate barbed line crossuig the head. The length of neck is the distance in a straight line from posteroventral point of head to flagellar pore. The length of anterior process is the dis- tance in a straight line from midpoint of flagel- lar pore to midpoint of base of process. The lengths of midbody and antapical are measured in straight lines. (The lengths of antapical branches, on the other hand, are measured along curvatures from midpoints of bases of branches to tips.) Our method of measuring the anterior inclination of head is mdicated by two dotted lines and a dotted arc; one of the lines is perpendicular to the longitudinal axis of body, the other runs parallel to dorsoventral midline of head. Ahhrei' with more or less pronounced sigmoid curvature; antapical spinules, if present, lateral in ])Osition, one on each valve 15. 14. Longitudinal axis of body straight or with but slightly develoi)ed sigmoid curvature; antapical spinules, if present, sagittal in po.sition (belonging to left valve?) I'i- 15. Antapical spinules comparatively strong; ratio between length and greatest width of body, 3.5-40: 1. bispinosa Kofoid. 366 THE DIXOPHYSOIDAE. 15. Antapical spinules, if present, very small; ratio between length and greatest width of body, 12-13: 1. curvaia Kofoid. 10. .4ntai)ioal without spinules truncata Kofoid and Miehener. 10. .^ntaiiieal with two spinules 17. 17. Head about three times wider than long complanala, sp. nov. 17. Head spheroidal, about as long as wide 18. 18. Length of body, 230-203 ii\ posterior part of antapical more or le.ss .strongly bent to the right. clavipes Kofoid. 18. Length of body, aliout olO m; posterior part of antapical but slightly bent to the right. schroderi Kofoid. 19. Antapcx with two spinules; right valve without antapical spinule 20. 19. Antapex with three spinules; right valve with one antapical spinule 21. 20. Length of body, 716-990 mI ratio between length and greatest width of body, 30.3-07.1 : 1. hideittnia Schroder. 20. Length of body, 522-009 /i; ratio between length and greatest width of body, 13.4-33.8: 1. leininermaimi Kofoid . 21. Head 9-10 times wider than long asymmelrica Kofoid. 21. Head about 1 ..5-3.0 times wider than long 22. 22. Antapical 1.5-2.3 times longer than anterior process and midbody palmata Stein. 22. Antapical shorter than anterior process and midbody palueotlieroides Kofoid. 23. Antapical with one branch 24. 23. Anta])ical with more than one branch 25. 24. One of the antajjical limbs strikingly .shorter than the other, davate, without spinules. projectn Kofoid. 24. Antapical limbs subetiual, both with sjjinules hifurculn Murray and \Miitting. 25. Antapical with two branches tlirinax Schiitt. 25. ,\ntapical with three branches quadricauda Kofoid ami Miehener. 25. .\ntapical with four branches qmnquecanda Kofoid. 1. Species of Uncertain or More or Less Isolated Position. Out of the seven species that belong to this group, six have been found in the material of the Expedition. These six species have been treated in the following order, which is partly indicative of their relative positions in the scale of the evolution- ary development of the genus: — Amphisolenia inflata, A. laticincta, A. hrevi- cauda, A schauinslandi, A. recta ngidala, and .4. iidragalus. Amphisolenia infl.\ta Murray and Whitting Plate 6, fig. 2, 7. Figure 47, 49: 1 Amphisolenia inflata Murray & Whitting, 1899, p. 332, tab. 8, pi. 31, fig. 2a, b. Lemmerm.^nn, 1899a, p. 373; 1905a, p. 37. Schroder, 1900a, p. 20; 1906a, p. 324, 328. Cleve, 1903b, p. 3;?9. Osten- FELD & Paulsen, 1904, p. 164, 173. Paulsen, 1908, p. 21, fig. 24. Diagnosis: — Straight. Head twice as wide as long. Epitheea concave. Midbody very large, ovoidal, well set off from anterior process and antapical; subequal in length to antapical and somewhat longer than anterior part of body. Ratio between length and width of body, 4.2-4.3: 1. Antapical straight, not in- flated posteriorly. Antapex with two lateral spinules. Wall profusely pitted, neck reticulate. Length, 143-233 p. Probably of world-wide distribution in tropical, subtropical, and warm- temperate regions of the sea. SYSTEMATIC ACCOUNT. 367 Descriptioji: — The longitudinal axis of the body is almost straight. The head is subhorizontal or inclined anteriorly at an angle of about 10°; its axial length is 0.38-0.50 the length of the neck; and it is about twice as wide as long. The anterior face of the epitheca is distinctly concave. The transverse furrow is 0.4 as wide as the dorsoventral diameter of the head, somewhat concave, and without cross-ridges. The cingular lists are about twice as wide as the transverse furrow; the anterior flares anteriorly at an angle of 10°-20° and has, dorsally and ventrally only, a few branched ribs; the posterior is subhorizontal and lacks ribs. The neck is short and thick, 0.10-0.11 the length of the body, and 2.5-3.5 times longer than wide. The sulcal lists continue beyond the flagellar pore to the midbody. They are reticulated anteriorly, and the left has, near the flagellar pore, a characteristic thickened area with a cross-rib in front of the pore. The anterior process is 0.G4-O.69 the length of the neck and about as long as wide. The midbody is relati\-ely very large, ovoidal, compressed laterally, and well set off from the anterior process and from the antapical. Its length is 3.5-3.6 and its width 2.1-2.2 the length of the neck. The ratio between the length and the greatest width of the body is 4.2-4.3: 1. In lateral view the midbody is about twice as wide as in dorsoventral view. The antapical is unbranched, straight, 3.1-3.8 the length of the neck and 7-12 times longer than its average width. When viewed laterally, it tapers to a point posteriorly; when seen dorsoventrally, it tapers \'ery little and terminates in two spreading, rather stout lateral spinules. The thecal wall is characterized by a very distinct structural differentiation. On the antapical and on the midbody it has numerous closely placed, scattered, minute pits (pores?). On the anterior process the pits are larger and more widely spaced; on the neck they widen to a coarse reticulum, about three meshes w ide on each valve. The nucleus is large and oval. The chromatophores are numerous, of various sizes, and dark brown in color. The proportions of two specimens, the type (Murray and Whitting, 1899, pi. 31, fig. 2a) and one taken by the Expedition, were measured. Ditnensiotis: — Our specimen: Length of body, 143 m- Length of head, 8.0 ii. Length of neck, Hj ii. Length of anterior process, 11 n. Length of midbody, 58 n. Width of midbody, 33 yu. Length of antapical, 50 ^u. Type specimen (Murray and Whitting, 1899, pi. 31, fig. 2a): Length of body, 233 m- Length of head, 9.5 fi. Length of neck, 25 fi; apparently it is somewhat shorter because of 368 THE DINOPHYSOIDAE. the fact that the thickened area of the left sulcal Ust has been drawn as if it be- longed to the anterior process. Length of anterior process, 1(5 fi. Length of mid- body, 88 fx. Width of midbody, 56 ii. Length of antapical, 94 ix. Variations: — Judging by the two specimens thus far figiu-ed, the type speci- men (Murray and Wliitting, 1899, pi. 31, fig. 2a) and the spechnen represented by our Plate 6, figure 7, this species appears to be rather constant. Most variable are the length of the body, and the relative width of the midbody and of the antapical. Comparisons: — The three specimens referred to Amphisolenia inflata in this paper agree rather closely with the type of this species as figured by Murray and Whitting (1899). They differ from the type specimen mainly in the length of the body, and in the relative width of the midbody and of the antapical. Amphisolenia inflata occupies a rather isolated position. We do not know which of the described species of Amphisolenia is its closest relative. It appears to be one of the most primitive, possibly the most primitive, species of this genus at present known. The following of its characters appear to be more or less primi- tive:— (1) the small size of the body; (2) the straight longitudinal axis of the body; (3) the exceptionally large, bilaterally compressed midbody; (4) the simple, unbranched antapical; (5) the well-developed structural differentiation of the thecal wall, fineareoles, and reticulation; (6') the reticulation of the sulcal lists; (7) the continuation of the sulcal lists beyond the flagellar pore. The question as to whether or not the concavity of the anterior face of the epitheca is primitive cannot be settled as j'et. This is a peculiarity of this species and of Triposolenia truncata that has not been found in any other species of these two genera or in the primitive genera of this tribe. As to the relationship between Amphisolenia in- flata and Triposolenia truncata, see the section on the subdivisions of this family. Synonymy: — The s])ecies was established by Murray and Whitting (1899) under the name of Amphisolenia inflata. Paulsen (1908) reproduces the figure of the type specimen. Occurrence: — ^ The species is recorded at three of the 127 stations with a total of three specimens. There are 0, 0, 1,1,1, and 0 stations on the six lines of the Expedition. One of these three stations (4713) is in the Galapagos Eddy; the two others (4681, 4724) are in the South Eciuatorial Drift. One of the catches (4681) was from 800-0 fathoms, the two others from 300-0 fathoms. The temperature range of these three stations at the surface was 68°-79°; the average was 73.3°. The frequency is less than 1%. The species was first recorded by jMurray and Whitting (1899) from the SYSTEMATIC ACCOUNT. 369 Sargasso Sea, lat. 34°-39° N., long. 39°-32° W., the type locality. Later it was reported from lat. 59° N., long. 21° W. (the Gulf Stream) by Ostenfeld and Paulsen (1904) ; from lat. 15° N., long. 52° E. (the Arabian Sea) by Cleve (1903b) ; and from the Gulf of Aden and the Indian Ocean, between Ceylon and Singapore, by Schroder (1906a). All the samples in which the species was found were from surface waters. The species occiu's in waters of the following temperatures and salmities: — Murray and Wliitting (1899) : temperature, 66°-60°; Ostenfeld and Paulsen (1904) : temperature, 49.3°; salinity, 35.30; Cleve (1903b): temperature, 74.3°; salinity, 36.02. Ostenfeld and Paulsen (1904), Cleve (1903b), and Schroder (1906a) do not give any information as to the structure of their specimens. The species is probably eupelagic and of world-wide distribution in tropical and subtropical waters and only occasionally is carried by currents to the colder regions of the sea. Amphisolenia laticincta Kofoid Plate 6, fig. 6, 8, 9. Figure 49: 2, 50: 10 Amphisolenia lalicincia Kofoid, 1907a, p. 198, pi. 13, fig. 80. Diagnosis: — Straight. Head but slightly wider than neck, almost twice as long as wide, hourglass-shaped. Epitheca boldly convex. Midbody fusiform, gradually merging into anterior process but set off from antapical. Distance from flagellar pore to antapical not quite half as long as body and somewhat longer than antapical. Ratio between length and width of body, 12: 1. Antapical straight, not inflated posteriorly. Antapex round with one short spinule (in typical specimens probably with one spinule on each valve). Length, 117-136 n. Eastern tropical Pacific. Description: — The longitudinal axis of the body is almost straight. The head is but slightly wider than the neck, of which it forms a continuation in a nearly straight line; its axial length is almost twice its width and subequal to or rather slightly shorter than the length of the neck. The epitheca is oblique and much more convex dorsally than ventrally. The transverse furrow is \ery broad, somewhat wider than the dorsoventral diameter of the head; it is rather strik- ingly concave, which gives to the head the shape of an hourglass, and without cross-ridges. The anterior cingular list is subhorizontal or it flares anteriorly at an angle of 10°; its width is 0.5 the width of the transverse furrow and subequal to the width of the neck; it has, on each valve, two to three simple ribs. The poste- 14 IS 16 Figure 49. — 1, Amiiliinnlenia inflntn Murray and Whifting. Station 4081 (300-0 fatlioms). 2, A. latichicla Kofoid. Station 4740 (300-0 fathoms). 3, A. hrencmtda Kofoid. Station 4740 (300-0 faUioni.s). 4, A. schuuinslundi Lomnicrniann. Station 4736 (300-0 fathoms). 5, A. rectangulata Kofoid. Station 4740 (300-0 fathom.s). 0, .4. aslraguhis Kofoid and Michener. Station 4713 (300-0 fathoms). 7, A. elongata, sp. nov. Station 4681 (800-0 fathoms). 8, A. exiensa Kofoid, posterior portion of antapical i.s a direct continuation of anterior. Station 4099 (300-0 fathoms). 9, A. glohifera Stein. Station 4742 (300-0 fathoms). 10, A. timuhixinna Kofoid. Station 471 1 (300-0 fathoms). 11, .4. currala Kofoid. Station 400.5 (300-0 fatlioms). 12, A. biaputoja Kofoid. Station 4605 (300-t) fathoms). 13, .4. clanpcs Kofoid. Station 4736 (300-0 fathoms). 14, A. Irimcitta Kof length of the body, and about seven times longer than wide. On each valve, near the middle, it has a low, narrow, somewhat irregidar, sometimes broken, longitudinal ridge, which in some indi- viduals extends from the girdle to near the flagellar pore. The sulcal lists extend a little behind the flagellar pore and have no ribs; the fission lib has not been seen. The midbody is fusiform, and merges very gradually into the anterior process and into the antapical. It is very slender and witlest at a distance from the apex equaling two fifths of the length of the body or at a distance from the flagellar pore four times longer than the neck; its greatest width is 0.43 the length of the neck. The ratio between the length and the greatest width of the body is 35- 40: 1. The antapical is unbranched, and gently and evenly concave ventrally. Its length cannot be determined, even approximately, because of the fact that it merges so gradually into the midbody. It grows gradually narrower posteriorly, but its posterior end, which is somewhat wider than the neck, is sUghtly inflated. In lateral view the antapex is rounded; in dorsoventral view it is truncate. It has two stout, pointed, lateral antapical spinules, one on each valve; the length of these spinules approaches the distal width of the antapical. The thecal wall is almost without structure. Along the base of each of the SYSTEMATIC ACCOUNT. 397 sulcal lists and on either side of the dorsal suture line of the neck there is a row of closely set pores. On the posterior part of the antapical there is a row of pores on either side of the dorsal as well as of the ventral suture line and also some scattered pores (see also the description of the neck). The nucleus is elongated and somewhat irregular. The chromatophores are numerous, small, and ellipsoidal. The proportions of one specimen, the type, were measured. Dimensions: — Length of body, 646-652 n (type, 652 n, not 670 /i as stated in the original description). Length of head, 11 yu. Length of neck, 43 n. Length of anterior process, midbody, and antapical, 587 n. Width of midbody, 20 ju. Variations: — In the few specimens examined as yet, this species appears to be rather constant. Comparisons: — The description given above is based on the type material. The simple, unbranched type of the antapical appears to be a primitive feature of this species. On the other hand, the slenderness of the midbody, the very gradual merging of the midbody into the antapical, and the sigmoid curva- ture of the longitudinal axis of the body probably are later differentiations. Amphisolenia bispinosa presumably is closely related to A. curvata. This relationship is indicated by the sigmoid cur^'atlU'e of the longitudinal axis of the body, by the shape of the posterior part of the antapical, by the lateral position of the spinules on the antapex, and by the longitudinal ridges on the neck. A. bispinosa differs from A. curvata mainly in the small size of its body, in the nar- rowness of its midbody, and in the relatively large size of its antapical spinules. Occurrence: — Amphisolenia bispinosa is recorded at eight of the 127 stations. There are 3, 2, 0, 1, 2, and 0 stations on the six lines of the Expedition. Of these eight stations, one (4583) is in the California Current; one (4605) is in the Mexican Current; one (4617) is in the Panamic Area; two (4662, 4668) are in the Peru\ian Current; one (4713) is in the Galapagos Eddy; two (4736, 4737) are in the South Eciuatorial Drift. At one station (4662) the species is recorded from 800-0 fathoms. All the remaining records refer to catches from 300-0 fathoms. The temperature range of these eight stations at the surface was 67°-85°; the average was 77.2°. The frequency is less than 1 % except at one station (4668) where it is 1 %. The species has been found only in the material of the Expedition. It was first recorded by Kofoid (1907a) from Station 4605 of the Expedition, which thus is the type locality. 398 THE DINOPHYSOIDAE. Amphisolenia curvata Kofoid Plate 9, fig. 5, 9, 10. Figure 49: 11 Atnpliisolem'ii currata KoFoii), l()07u, p. 197, pi. 14, fig. 87. Diagnosis: — With rather strong sigmoid curvature. Head 1.5 times wider than long. Epitheca flat or slightly convex. Midbody fusiform, merging gradu- ally into anterior process but rather well set off from antapical. Approximate distance from flagellar pore to antapical, 4.7 the length of neck. Ratio between length and width of body, 12-13: 1. Antapical with rather strong, even, ventral concavity, approximately four times longer than neck; its posterior end very slightly widened. Antapex rounded with minute spinule on left valve (normally with one on each valve?). Length, 463 ti. Eastern tropical Pacific. Description: — The longitudinal axis of the body has a rather strong sigmoid curvature, due to the moderate dorsal deflection of the neck and the rather strong and even ventral concavity of the antapical. The ventral deflection of the ant- apical is about 7°. The head is inclined anteriorly at an angle of 20°; its axial length is 0.24 the length of the neck, and it is 1.5 times wider than long. The anterior face of the epitheca is flat or slightly convex. The transverse furrow is 0.25 as wide as the dorsoventral diameter of the head and flat; it has, on each valve, five to six almost equidistant cross-ridges. The cingular lists are twice as wide as the transverse furrow and flare anteriorly at an angle of 10°; each of them has, on each valve, six fairly regularly spaced, simple ribs. The neck is of moderate length, 0.10 the length of the body, and seven to eight times longer than wide. On each valve, near the middle, it has a low, narrow, somewhat irregular, longitudinal ridge, which in the type specimen extends from the girdle to the flagellar pore. The sulcal lists extend to the flagellar pore and have no ribs. An accessory list, arising from the originally free edge of the left sulcal list, extends along the po.sterior thirtl of the left sulcal list and ends some- what behind the flagellar pore; its greatest width, which is subequal to that of the neck, is at the flagellar pore; anteriorly and posteriorly it tapers, rather sud- denly posteriorly, very slowly anteriorly; at the flagellar pore it has a cross-rib, the fission rib. The midbody is fusiform and merges gradually into the anterior process but is rather well set off from the antapical. The approximate distance between the flagellar pore and the anterior end of the antapical is 4.7 the length of the neck. SYSTEMATIC ACCOUNT. 399 The midbody is of moderate width, widest at a distance from the apex equaUng two fifths of the length of the body, or at a distance from the flagellar pore three times the length of the neck ; its greatest width is 0.8 the length of the neck. The ratio between the length and the greatest width of the body is 12-13: 1 . The antapical is unbi'anched, with a rather strong and even ventral con- cavity, and approximately four times longer than the neck or 0.8 the total length of the anterior process and the midbody. Posteriorly it grows gradually narrower, except near the posterior end where it widens very slightly; its posterior end is about 1.5 times wider than the neck. In lateral view the antapex is rounded; in dorsoventi'al \ie\\ it probably is truncate in completely developed specimens with a minute lateral spinule on each valve; in the type specimen the right valve is rounded and without any spinule, the left truncate with a minute lateral spinule. The thecal wall is almost without structure. On the posterior end of the antapical there is a row of pores on either side of the dorsal as well as of the ven- tral suture line and also some scattered pores. In front of these, there is an irregu- lar reticulum of low and narrow ridges (see also the description of the neck). The porulation of the neck is probably as in Amphisolenia bispinosa, but there is no information on this point. The nucleus is small and ellipsoidal. The chromatophores are numerous, ellipsoidal and spheroidal. The proportions of one specimen, the type, were measured. Dimensions: — Length of body, 4G3 ,u. Length of head, 11 ;u. Length of neck, 45 ju. Length of anterior process, midbody, and antapical, 397 n. Approxi- mate length of anterior process and midbody, 220 m- Width of midbody, 35 /i. Approximate length of antapical, 177 /j. Comparisons: — The description given above is based on the type specimen. As to the relationships between this species and the other members of this genus, see Amphisolenia bispinosa, the section on comparisons. Occurrence: — Amphisolenia curvata is recorded at only one of the 127 sta- tions. This station (4G05) is on the first line of the Expedition and in the Mexican Current. The depth is 300-0 fathoms, the surface temperature 85°, and the fre- quency less than 1% (one specimen). The species has been found only in the material of the Expedition. It was first recorded by Kofoid (1907a) from Station 4605 of the Expedition, which thus is the type locality. 400 THE DINOPHYSOIDAE. 5. ScHRODERi Group. All the four species that belong to this group have been founil in the material of the Expedition. They have been treated in this paper in the followmg order indicative of relationships: — Amphisolenia schroderi, A. clavipes, A. complanata, and A. Iruncata. Amphisolenia schroderi Kofoid Plate 10, fig. 2-4. Figure 49: 15 Ainphi>:ule.nia nchroedcri Kofoid, 1907a, ]). 201, i)l. 13, fig. 81. Diagnosis: — Straight. Head spheroidal, about as long as wide. Epitheca boldly convex. Midbody fusiform merging very gradually into anterior process and antapical. Ratio between length and width of body, 25: 1. Antapical almost straight ; its posterior part slightly bent ventrally and to the right and not or only slightly widened. Antapex rounded in dorso\entral view and truncate in lateral view; with two short spinules, one dorsal and one ventral, both probably belong- ing to left valve. Length, 510 ju. Tropical and subtropical regions of Eastern Pacific. Description: — The longitudinal axis of the body is almost straight; only the posterior part of the antapical is slightly bent. The head is spheroidal, about as wide as long or but slightly wider than long; its axial length is 0.18-0.19 the length of the neck. The anterior face of the epitheca is boldly convex. The trans- verse furrow is about 0.5 as wide as the dorsoventral diameter of the head, some- what convex, and has on each valve one to four cross-ridges. The anterior cingular list is about twice as wide as the transverse furrow, flares anteriorly at an angle of 30°-40°, and has, on each valve, five to six equidistant, simple ribs. The posterior c-ingular list is slightly narrower, subhorizontal or inclined anteriorly about 10°, and has no ribs. The neck is rather long, 0.13 the length of the body, and nine times longer than wide. The sulcal lists end at the flagellar pore and have no ribs. The midbody is fusiform and merges \-ery gradually into the anterior process and into the antapical. It is rather slender and widest at a distance from the apex eciualing two fifths of the length of the body or at a distance from the flagellar pore about twice as long as the neck ; its greatest width is 0.30 the length of the neck. The ratio between the length and the greatest width of the body is 25 : 1 . The antapical is unbranched and almost straight ; its posterior part is slightly bent ventrally and to the right. Its length cannot be determined even approxi- mately, because of the fact that it merges so gradually into the midbody. It SYSTEMATIC ACCOUNT. 401 grows gradually narrower posteriorly, and its posterior part is not or only slightly widened; its width near the posterior end is about 1.5 the average width of the neck. In dorsoventral view the antapex is rounded, in lateral view it is truncate. It has two short spinules, one dorsal and one ventral, both probably belonging to the left valve ; the length of the spinules is about 0.25 the dorsoventral width of the posterior end of the ant apical. The thecal wall is apparently without structure. On the posterior part of the antapical there is a row of fine pores on either side of the ventral suture line, and a similar row is found along the base of each of the sulcal lists. The nucleus and the chromatophores are as in Aviphisolenia clampes. The proportions of one specimen, the type, were measured. Dimensions: — Length of body, 510 n. Length of head, 10 yu. Length of neck, 68 ix. Length of anterior process, niidbody, and antapical, 431 //. Width of midbody, 20 n. Comparisons: — The description given above is based on the type material The straight longitudinal axis of the body and the simple, unbranched type of the antapical are primiti\'e features of this species. The slenderness and the very gradual merging of the midbody into the antapical probably are advances in differentiation. The closest-known relative of Amphisolenia schroderi is undoubtedly A. clavipes. Indeed, it even may be questionable whether or not these two forms are specifically distinct, and their specific separation should be regarded as tentative. A. schroderi differs from .4. clavipes mainly in the following respects: — (1) its midbody is slightly narrower, with the ratio between length and width of body 25: 1, as compared with 15-19: 1; (2) the posterior portion of its antapical gener- ally is somewhat less inflated antl less strongly bent to the right (see A. clavipes, the section on variations) ; (3) its transverse furrow and its epitheca appear to lack porulation; (4) it is decidedly larger, 510 yu as compared with 230-263 m- The difference in size appears to be the most important. However, it should be remembered that dwarfs and giants have been found in a rather large number of species in Dinophysoidae. Another close relative of Amphisolenia schroderi is A. complanata. The former species differs from the latter mainly in the smaller width of its head (ratio between length and width of its head about 1 : 1 as compared with 1 : 3) and in the smaller size of its antapical spinules. Amphisolenia schroderi is probably also rather closely related to A. truncata, from which it is easily distinguished by the narrowness of its head and by having 402 THE DINOPHYSOIDAE. antapical spinules (as to the systematic value of the latter difference, see A. truncata, the section on comparisons). Amphisolenia schroderi, A. clavipes, A. complanata, and A. truncata form a presumably natural systematic group, the Schroderi group. Their closest rela- tives, outside the group, probably are to be found in the Palmata group. Occurrence: — Amphisolenia schroderi is recorded at seven of the 127 stations. There are 2, 1, 0, 1, 3, and 0 stations on the six lines of the Expedition. Of these seven stations, two (4598, 4605) are in the Mexican Current; one (4659) is in the Peruvian Current; four (4701, 4736, 4737, 4739) are in the South Equatorial Drift. At one station (4737) the species is recorded from 100-0 fathoms as well as from 300-0 fathoms. All the other records refer to catches from 300-0 fathoms only. The temperature range of tliese seven stations at the surface was 69°-85°; the average was 78.8°. The frequency is less than 1 % at all stations. The species has been found only in the material of the Expedition. It was first recorded by Kofoid (1907a) from Station 4737 of the Expedition, which thus is the type locality. Amphisolenia clavipes Kofoid Plate 11, fig. 8-11. Figure 49: 13, 50: 7 Amphisolenia clavipes Kofoid, 1907a, p. 197, pi. 14, fig. 90. Doflein, 1909, p. 464, fig. 403c]; 1911, p. 531, fig. 462d; 1910, p. 437, fig. 374d. Diagnosis: — Almost straight except near antapex. Head spheroidal, about as long as wide. Epitheca boldly convex. Midbody fusiform, merging very gradually into anterior process and antapical. Ratio between length and width of body, 15-19: 1. Antapical almost straight except near antapex where it is slightly curved ventrally and more or less strongly bent to the right ; posterior end but slightly or more or less strikingly widened. Antapex rounded in dorso ventral \-iew and truncate in lateral view; with two short spinules, one dorsal and one ventral, both probably on left valve. Length, 230-263 ix. Eastern tropical Pacific. Description: — The longitudinal axis of the body is almost straight; the neck is slightly deflected dorsally and the posterior end of the antapical is more or less bent. The head is spheroidal, aliout as wide as long, or rather slightly wider than long (width: length, 1.0-1.3: 1); its axial length is 0.23 (0.19-0.27) the length of SYSTEMATIC ACCOUNT. 403 the neck. The anterior face of the epitheca is boldly convex. The transverse furrow is about 0.5 as wide as the dorsoventral diameter of the head, somewhat convex or flat and has, on each valve, three to four cross-ridges. The anterior cingular list is about as wide as or somewhat wider than the transverse furrow, flares anteriorly at an angle of 30°-40°, and has, on each valve, four to six equi- distant, simple ribs. The posterior cingular list is generally slightly narrower, subhorizontal or inclined anteriorly at an angle of 10°-20° and has no or only a few simple ribs. The neck is rather long, 0.14-0.15 the length of the body, and six to seven times longer than wide. The sulcal lists end at or just behind the flagellar pore and have no ribs; the fission rib has not been seen. Sometimes they have about the same shape as in Plate 10, figure 3, of Amphisolenia schroderi; sometimes a rather broad accessory list arises near the flagellar pore from the originally free edge of the left sulcal list and continues somewhat behind the pore (Plate 11, fig. 8). The midbody is fusiform and merges very gradually into the anterior process and into the antapical. It is rather slender or of moderate width and widest somewhat in front of the middle ; its greatest width is 0.38-0.45 the length of the neck. The ratio between the length and the greatest width of the body is 1 5-1 9 : 1 . The antapical is unbranched and almost straight except posteriorly; its posterior part is slightly curved ventrally and more or less strongly bent to the right. Its length cannot be determined even approximately because of the fact that it merges so gradually into the midbody. It grows gradually narrower posteriorly, except near the antapex where it sometimes is slightly or moderately widened; its width near the antapex somewhat exf'eeds the average width of the neck. In dorsoventral view the antapex is rounded, in lateral view it is truncate; it has two short spinules, one dorsal and one ventral, both probably belonging to the left valve. The length of the spinules is about 0.25 the dorsoventral width of the posterior end of the antapical. The thecal wall has the same structure as in Amphisolenia schroderi, but the transverse furrow and the epitheca are porulate. The nucleus is large, elongated, and somewhat irregular. The chromato- phores are rather large and cylindrical (Plate 11, fig. 11). The proportions of five specimens from the material cf the Expedition were measured. Dimensions: — Length of body, 230-263 m (average, 248 m; type,. 230 yu). Length of head, 7.0-8.5 yu (average and type, 8.0 n). Length of neck, 31-39 ^ 404 THE DINOPHYSOIDAE. (average, 36 m; type, 31 yu). Length of anterior process, midbody, and antapical in the type specunen, 190 fi. Width of midbody, 12-10 n (average, 15 fi; type, 12 A"). Variations: — The species appears to be relatively constant. The shape of the posterior portion of the antapical is the most variable character. Sometimes, as in the type specimen (Figure 50: 7), this portion is rather strikingly inflated, and strongly and rather abruptly bent to the right. In other cases these pecuUari- ties are less developed, and this portion exen may ha\e almost the same shape as in Amphisolenia schroderi, i.e., it may be but slightly or even not at all inflated and only slightly bent to the right (Plate 10, fig. 4). Comparisons: — The description given above is based on the type material. With regard to the relationships of this species, see Amphisolenia schroderi, the section on comparisons (p. 401). Sijnonyvnj: — The species was established by Kofoid (1907a) under the name of Amphisolenia claripes. Reproductions of Kofoid's (1907a) figiu'e of the type specimen are given by Doflein (1909, 1911, 1910). Occurrence: — Amphisolenia clavipes is recorded at two of the 127 stations. These two stations (4736, 4737) are on the fifth line of the Expedition and in the South Equatorial Drift. At Station 4737 tjie species is recorded from 100-0 fathoms and from 300-0 fathoms; at Station 4730 from 300-0 fathoms only. The surface temperatures of these two stations were 81.0° and 81.5°, re- spectively. The frequency is less than 1 %. The species has been found only in the material of the Expedition. It was first recorded by Kofoid (1907a) from Station 4736 of the Expedition, which thus is the type locality. Amphisolenia complanata, sp. nov. Plate 10, fig. 5, 7, 9. Figure 49: 16 Diagnosis: — With slight sigmoid curvature. Head three times wider than long. Epitheca flat. Midbody fusiform, merging very gradually into anterior process and antapical. Ratio between length and width of body, 32-39: 1. Ant- apical with very slight ventral concavity; approximately five to eight times longer than neck; its posterior end slightly widened, almost straight. Antapex rounded in dorsoventral view and truncate in lateral view ; with two rather strong spinules, one dorsal and one ventral, both probably on left valve. Length, 600-785 n. Eastern tropical Pacific. Description: — The longitudinal axis of the body is slightly sigmoid, the neck SYSTEMATIC ACCOUNT. 405 somewhat deflected dorsally, and the antapical evenly and very sUghtly concave ventrally. The head is mcUned anteriorly at an angle of 30°; its axial length is about 0.12 the length of the neck, and it is three times wider than long. The anterior face of the epitheca is flat or but slightly convex. The transverse furrow is 0.25 as wide as the dorsoventral diameter of the head, somewhat convex, and has, on each valve, six to seven equidistant cross-ridges. The cingidar lists are about twice as wide as the transverse furrow or somewhat less; the anterior is inclined anteriorly at an angle of 20° and has, on each valve, eight equidistant, simple ribs; the posterior is subhorizontal and has, on each valve, six to seven smiilar ribs. The neck is of moderate length, 0.06-0.09 the length of the body, and eight to nine times longer than wide. The sulcal lists sometimes extend to the flagellar pore and have no ribs (Plate 10, fig. 5); sometimes an accessory Ust of moderate width arises near the flagellar pore from the originally free edge of the left sulcal list and continues somewhat behind the pore. The accessory list has a fission rib at the pore (compare Plate 11, figure 8, of Amphisolenia clampes). The midbody is fusiform and merges very gradually into the anterior process and into the antapical. The approximate distance between the flagellar pore and the anterior end of the antapical is five to eight times the length of the neck. The midbody is very slender, widest at a distance from the apex equaling one third of the length of the body, or at a distance from the flagellar pore 2.5-3.5 times the length of the neck; its greatest width is 0.35-0.40 the length of the neck. The ratio between the length and the greatest width of the body is 32-39 : 1 . The antapical is unbranched, with a very slight ventral concavity; its poste- rior part is almost straight. It is approximately five to eight times longer than the neck and subequal to the total length of the anterior process and the midbod}'. Posteriorly^ it grows slightly narrower, except near the posterior end where it widens slightly; its posterior end is about 1.8-2.2 times wider than the neck. In dorsoventral view the antapex is rounded; in lateral view it is truncate; it has two rather strong spinules, one dorsal and one A-entral, both probably on the left valve; the length of the spinules is about 0.50 the dorsoventral width of the posterior end of the antapical or somewhat less. The thecal wall is apparently without structure. On the posterior part of the antapical there is a row of fine pores on either side of the dorsal as well as of the ventral suture line and some pores between these rows. On the neck no pores have been observed. 406 THE DINOPHYSOIDAE. The nucleus and the chromatophores are probably as in Amphisolenia clavipes. The proportions of two specimens were measured. Dimensions: — Length of body, 600-785 m (type, 600 m). Length of head, 7.0-7.5 fi (type, 7.0 n). Length of neck, 46-53 ju (type, 53 n). Length of anterior process, midbody, and antapical of type specimen, 432 ix. Width of midbody, 20/.. Variations: — In the two specimens examined, this species was constant, except in the size of the body and the relative length of the neck. Comparisons: — The species certainly is very closely related to Aiuphi- solenia schroderi and A. clnmpes, as is shown by the similar differentiation of the posterior portion of the antapical in these three species. It is easily distinguished from these two species by the greater width of its head and by the larger size of its two antapical spinules. However, A. complanata probably is most closely re- lated to A. truncata, from which it differs mainly in having antapical spinules. For further discussion of the relationships of these four species, see A. schroderi^ the section on comparisons (p. 401). Occurrence: — Amphisolenia complanata is recorded at two of the 127 sta- tions. These two stations (4737, 4739) are on the fifth line of the Expedition and in the South Equatorial Drift. The depth is 300-0 fathoms. Station 4739 is the type locality. The surface temperatures of these two stations were 79.0° and 81.5° re- spectively. The frequency is less than 1 %. Amphisolenia truncata Kofoid and Michener Plate 11, fig. 1, 12. Figure 49:14 Amphisolenia Iruncala Kofoid & Michener, 1911, p. 294. Jorgensen, 1923, p. 40, 43, fig. 58. Diagnosis: — Almost straight or with slight sigmoid curvature. Head 2.0- 2.5 times wider than long. Epitheca moderately convex. Midbody fusiform, merging very gradually into anterior process and antapical. Approximate dis- tance from flagellar pore to antapical four to five times longer than neck. Ratio between length and width of body, 28-38: 1. Antapical straight or with slight ventral concavity; approximately 6.0-6.5 times longer than neck; its posterior end not widened, straight. /Vntapex truncate in lateral \-iew, without spinules. Length, 650-663 a. Eastern tropical Pacific, and Mediterranean. SYSTEMATIC ACCOUNT. 407 Description: — The longitudinal axis of the body is almost straight or gently sigmoid, with the neck somewhat deflected dorsally and the antapical straight or with a shght ventral concavity. The head is inclined anteriorly at an angle of 30°; its axial length is 0.19-0.20 the length of the neck; and it is 2.0-2.5 times wider than long. The anterior face of the epitheca is moderately convex. The transverse furrow is 0.25 as wide as the dorsoventral diameter of the head, flat, and has, on each valve, six to seven equi- distant cross-ridges. The cingular lists are 1.5-2.0 as wide as the transverse fur- row ; the anterior is inclined anteriorly at an angle of 20° and has, on each valve, seven or eight simple, equidistant ribs; the posterior is subhorizontal and may possibly lack ribs (Jorgensen, 1923, fig. 58b). In the type specimen no ribs were observed in the cingular lists; they probably were overlooked. The neck is of moderate length, 0.08 the length of the body, and eight to nme times longer than wide. The sulcal lists are as m Amphisolenia complanala. The midbody is fusiform and merges very gradually into the anterior process and into the antapical. The approximate distance between the flagellar pore and the anterior end of the antapical is four to five times longer than the neck. The midbody is very slender, widest at a distance from the apex equaling 0.29-0.33 the length of the body, or at a distance from the flagellar pore two to three times longer than the neck; its greatest width is 0.34-0.44 the length of the neck. The ratio between the length and the greatest width of the body is 28-38 : 1 . The antapical is unbranched, straight, or with a very slight ventral concavity ; its posterior part is straight. It is approximately 6.0-6.5 times longer than the neck and 1.2-1.8 times the total length of the anterior process and the midbody. Posteriorly it grows slightly narrower; its posterior end is not widened and is 1.3-1.8 times wider than the neck. The antapex is truncate in lateral view and lacks ribs. Sometimes it is surrounded by a (protoplasmic?) sheath (see Jorgen- sen, 1923, fig. 58c). The thecal wall is apparently without structure. On the posterior part of the antapical there is a row of fine pores on either side of the dorsal as well as of the ventral suture line. A row of fine pores is also found along the base of each of the sulcal lists. The proportions of the type specimen and of the specimen drawn by Jorgen- sen (1923, fig. 58) were measured. Dimensions: — Type specimen: Length of body, 650 n. Length of head, 9 //. Length of neck, 50 fi. Length of anterior process, midbody, and antapical, 584 m- Approximate length of antapical, 317 n. Width of midbody, 17 m- Jorgensen's 408 THE DINOPHYSOIDAE. (1923) specimen: " Length 663 m, max. breadth 24 fi, or 36% of the length." Ac- cording to measurements made from his figures 58a and b : Length of body, 675 ix. Length of head, 11m- Length of neck, 55 fi. Length of anterior process, midbody, and antapical, 600 ix. Approximate length of antapical, 380 ii. Width of body, 28 n. Variations: — The two specimens examined by us exhibit variations in the degree of curvature of the longitudinal axis of the body, in the ratio between the length and the width of the body, and in the relative length of the antapical. Comparisons: — The description given above is based on the type specimen. The specimen described and figured by Jorgensen (1923) under the name of Amphisolenia iruncata agrees very closely with the type specimen of this species. It is, of course, an open question as to whether or not the absence of antapical spinules in Amphisolenia Iruncata is primitive. However, the fact that antapical spinules are present in A. inflnla, which probably is the most primitive known species of this genus, and in almost all the other known members of this genus, representing different evolutionary lines, and the fact that A. Iruncata in several respects appears to be highly differentiated rather than primitive seem to indicate that the absence of antapical spinules is secondary in this species. Amphisolenia truncata probably is closely related to A.schroderi, A. clavipes, and A. complanata. It differs from these three species in not having any antapical spinules. From A. schroderi and A. clavipes it also differs in the greater width of its head. The closest-known relative of A mphisolenia truncata is undoubtedly A . com- planata. It may even be questionable whether or not the two forms are specifically distinct, and their systematic separation should be regarded as tentative. They differ mainly in the presence or absence of antapical spinules. Both the antapical spinules in A. complnnaln probably belong to one valve. Binary fission in this species might therefore be followed by a stage of short duration in which one of the daughter individuals is without antapical spinules. Is A. truncata based on a representative of A. complanata in which the antapical spinules have not yet developed following binary fission? It may be of a certain interest in this con- nection that in the specimen of A. truncata figured Ijy Jorgensen (1923, fig. 58c) the posterior portion of the antapical appears to be surrounded by a hyaline sheath. Such a sheath ma,y be a sign of a recent cell division. For further discussion al)out the relationships of these species, see the sec- tion on the comparisons of Amphisolenia schroderi (p. 401). Occurrence: — Amphisolenia truncata is recorded at only one of the 127 stations. This station (4733) is on the fifth line of the Expedition and in the South SYSTEMATIC ACCOUNT. 409 Equatorial Drift. The frequency is less than 1%; only one specimen is recorded. This specimen was found in the stomach of a Salpa from surface waters, at a temperature of 80°. The species was first recorded by Kofoid and Michener (1911) from Station 4733 of the Expedition, which thus is the type locality. Later, Jorgensen (1923) reported one specimen from the Balearic Sea, off the coast of Africa. 6. Palmata Group. All the five species that belong to this group have been found in the material of the Expedition. They have been treated in this paper in the following order indicative of relationships: — Amphisolcnia bidentala, A. lemmerrnanni, A. palmata, A. palaeotheroides, and A. asymmeirica. Amphisolenia bidentata Schroder Figure 54: 1-4, 55, 56: 1 Amphisolenia bidentata Schrodkr, 1900a, p. 20, 35, pi. 1, fig. 16a-c; 1900b, ji. :i; 190(ia, p. .iTS, :524, .326, 328, .331, 332, 333, .338, 3.39, 371, 374; 1906b, p. 262; 1909, p. 211 ; 1911, p. 3, 4, 25, 38. Ostenfeld & Schmidt, 1901, p. 162. Schmidt, 1901, p. 1.38. Lemmermann, 1901a, p. 376. OsteN'feld, 1902, p. 21; 1915, p. 4. Entz, 1902b, p. 94; 1905, p. 112. Cleve, 1903b, pnrtim, p. 3:50. Okamura, 1907, ixirliii^, p. 127, non pi. 3, fig. 15; 1912, j). 20. Karsten, 1907, p. 2.52, 250, 2.59, 264, 267, 269, 270, 271, 274, 280, 284, 287, 288, 292, 297, 306, 307, 308, 311, 315, 325, 334, 335, 336, 337, 339, 342, 344, .345, 346, .347, 350, 3.52, 353, 354, 3.50, .357, 359. Stuwe, 1909, p. 2.54, 288. Kofoid & Michener, 1911, p. 293. Schiller, 1911a, p. 52; 1912, p. 27. P.-^vill-^rd, 1916, partiiii, p. 61. Forti, 1922, p. 112, 113, 192, 209, fig. 123. Jorcjensen, 1923, p. 39, 40, fig. 56. Amphisolenia sp., Chun, 1903, p. 76, fig. e. Diagnosis: — Slightly or moderately sigmoid. Head 2.0-3.0 times wider than long. Epitheca gently convex. Midbody fusiform, merging very gradually into anterior process and antapical. Approximate distance from flagellar pore to ant- apical, 3.5-5.5 times longer than neck. Ratio between length and width of body, 30.3-67.7: 1. Antapical approximately 8.0-13.5 the length of neck, with gentle ventral concavity, its posterior part bent to the right 25°-40°, and twisted \-en- trally 90°; at point of bending, left valve has strong, short, pointed spinule, which makes posterior part of antapical foot-like. Seen ventrally (torsion), antapex is widened, two to three times wider than average width of antapical, truncate; with two spinules on left valve, one on each corner, but none on right \ahe. Length, 716-990 m- Of world-wide distribution, in tropical, subtropical, and warm-temperate waters. Description: — The longitudinal axis of the body is slightly or moderately sigmoid, with the neck somewhat deflected dorsally, and the antapical having a gentle ventral concavity. The posterior part of the antapical is always bent to the right and twisted ventrally. 410 thp: dinophysoidae. The head is incUned anteriorly at an angle of 30°-50°; its axial length is 0.16 (0.15-0.19) the length of the neck; and it is 2.0-3.0 times wider than long. The anterior face of the epitheca is gently convex, or in some cases rather flat. The transverse furrow is 0.20-0.30 as wide as the dorsoventral diameter of the head, gently convex flat, or slightly concave and has, on each valve, about three to five cross-ridges. The cingular lists are 1.2-2.0 as wide as the transverse furrow; both may be inclined anteriorly at an angle of 10°-25° or both may be subhorizontal; each of them has, on each valve, five to nine simple, equidistant ribs. The neck is of moderate length and width, 0.05-0.08 the length of the body and eight to fourteen times longer than wide. The sulcal lists either end at the cytopharyngeal pore or they continue somewhat behind it ; in some specimens they appear to lack ribs, in others the left one has a fission rib at the flagellar pore. The midbody is fusiform and merges very gradually into the anterior process and into the antapical. The approximate distance between the flagellar pore and the anterior end of the antapical is 3.5 5.5 times longer than the neck. The mid- body is very slender, widest at a distance from the apex equaling about one fifth to one fourth the length of the body, or at a distance from the flagellar pore 1.5- 3.5 the length of the neck; its greatest width is 0.27-0.67 the length of the neck. The ratio between the length and the greatest width of the body is 46.3 (30.3- 67.7) : 1 (21 : 1 in a parasitized specimen). The antapical is unbranched, approximately 8.0-13.5 the length of the neck and 1.5-3.0 times longer than the total length of the anterior process and the mid- body. It is gently concave ventrally, and its posterior part is bent to the right at an angle of 25°-40° and twisted ventrally 90°. Our data are hisufficient to estab- lish whether or not this torsion is a constant characteristic of this species. At the point of bending the left valve has a short, strong, pointed, heel-like spinule, which gives to the posterior part of the antapical a foot-like appearance. In lateral view (torsion) the foot-like part is not as \\ide as or is rather slightly wider than the average width of the antapical, about 2.5-4.5 times longer than wide and more or less rounded distally. Seen ventrally (torsion) the posterior part of the "foot" is truncate and about two to three times wider than the antapical just in front of the foot. The antapex has two fairly strong and short, pointed spinules on the left valve, one on each corner, but no spinule on the right valve. The average width of the antapical is subequal to or somewhat exceeds that of the neck. Near the antapex the antapical has what appears to be a protoplasmic sheath. The thecal wall is apparently without structure. The posterior part of the SYSTEMATIC ACCOUNT. 411 antapical is porulate. Along the base of each of the sulcal lists there is a row of pores, and on the head a number of scattered pores are to be found. The chromatophores are located near the center of the midbody and are light yellowish brown (Ostenfeld and Schmidt, 1901). P^iCiURE 54. — 1-4, Aniphisohnia bidcntala iSchroder. X 430. l:i-c, from Station 4737 (100-0 fathoms), la, right lateral view of anterior portion of body; lb and Ir, posterior portion of antapical of specimen represented by la; lb is seen in right lateral view and shows ventral deflection; Ic is seen in dor.sal view and shows a comparatively long "foot"; torsion disregarded. 2, right lateral view of anterior portion of body; left sulcal list has a fission rib and continues for a coiLsiderable distance be- yond flagellar pore. Station 4740 (300-0 fathoms). 3, right lateral view of posterior portion of ant- apical ; straight. Station 4730 (300-0 fathoms). 4, dorsal view of posterior jiortion of antapical ; "foot " relatively short. Station 4713 (300-O fathoms). 5, A. palmata Stein, dorsal view of posterior j^ortion of antapical following fission; right (old) valve with its distal spinule fully developed; left (new) valve without heel spinule and with but .slightly developed distal spinules. X 900. Station 4734 (.300-0 fathoms). 6, A. asi/mmctrica Kofoid, dorsal view of posterior portion of antapical of tyi)e specimen of A. dolichocephalica Kofuid, i.e., of a specimen folhjwing fission; light (old) valve has its distal spinule fully developed; left (new) valve without heel spinule and with rather slightly developed distal spin- ules. X 900. Station 4728 (300-0 fathoms). 7-9, A. thrinax Schtitt. X 430. 7-8, from a specimen from Station 4594 (300-0 fathoms); 7, in right lateral ww; 8, in ventral view. 7a and 8a represent posterior portion of antapical stem, with relatively short "foot," and with heel-spinule and three distal sijinules; 7b and Sb represent distal portion of second antapical branch, without heel-spinule and with three distal spinules; 7c and Sc, distal iiortion of first antapical branch, with relatively short "foot," and with heel-spinule and three distal si)inules. 9, from a specimen, in right lateral view, from Station 4737 (.300-0 fathoms). 9a, posterior portion of antapical stem, with relatively long "foot," and with heel-spinule and three distal spinules; 9b, distal portion of second antapical branch, without heel-spinule and with two distal spinules on left valve; 9c, distal jiortion of first antapical branch, with relatively long "foot," and with heel-spinule and three distal spinules. The length and the width of the body were measured in twenty of our speci- mens and the proportions of the head and of the neck in fourteen of our specimens. All the proportions were measured in only thi-ee of our specunens and in the type specimen and the specimen reproduced by Jorgensen (1923, fig. 56). 412 THE DINOPHYSOIDAE. Dimensions: — Our specimens: Length of body, 716-990 ^ (average, 822 m). Length of head, 6-10 ^ (average, 8 yu). Length of neck, 43-53 ^ (a\-erage, 47 n). Width of midbod}', 14.0-29.6 m (average, 18.8 /x) ; in a parasitized specimen it was 34.8 IX. Type specimen (Schroder, 1900a, pi. 1, fig. 16a): Length of body, 773 n. Length of head, 8.5 /x. Length of neck, 48 ix. Length of anterior process and mid- body, 178 M- Length of antapical, 537 /x. Width of midbody, 18 fi. Jorgensen's (1923, fig. 56) specimen: Length of body, 920 ix. Length of head, 10 ix. Length of neck, 53 ix. Length of anterior process and midbody, 280 yu. Length of antapical 580 m. Width of midbody, 20 a/. ]'ariations: — The length of the body and the widtli of the midbody are rather variable. Variability is also fomid in other characters, e.g., in the width and relati\-e length of the foot-like posterior jmrt of the antapical. It may be pertinent to mention in this connection that in our narrowest specimen (ratio between the length and the width of the body, 67.7: 1) one of the two thecal valves was new; in other words, this specimen had been taken shortly after the separation of the two daughter cells in binary fission. This may indicate that the width of the mid- body increases with the age of the indi\iduals (see the section on the reproduction of this genus). Comparisons: — The specimens foimd in the material of the Expedition show, as previously mentioned, some variability. Howe\'cr, most of them agree closely with Amphisolenia bidentata as figured and described by Schroder (1900a). With regard to the specimen figiu'ed by Okamura (1907, pi. 3, fig. 15) under the name of Amphisolenia bidentata, see the following section. The specunen represented by Jorgensen (1923, fig. 56) under this name agrees very closely with the type specimen. The unbranched antapical is a i)rimitive feature of this species. The slight sigmoid curvature of the longitudinal axis of the body, the narrowness, and the gradual merging of the midbody into the antapical, as well as the differentiation of the posterior part of the antapical, are probably later acquisitions. Amphisolenia bidentata is certainly very closely related to A. lemmermanni. Indeed, these two forms are so similar that their systematic separation should be regarded as tentative. They differ from each other mainly in the length of the body and in the width of the midbody. The length of A. bidentata is 716-990 n (average, 822 n); that of A. lemmermanni, 522-669 m (average, 572 fi). The ratio between the length and the width of the body is 30.3-67.7 (average, 46.3) : 1 in A. bidentata; in A. lemmermanni it is 13.4-33.8 (average, 23.4) : 1 (see the previous section). Furthermore, the posterior part of the antapical is twisted ventrally at SYSTEMATIC ACCOUNT. 413 an angle of about 90° in A. bidentata and appears always to be without torsion in A. lemmermanni. However, our data are not sufficient to establish whether or not the last difference is constant. In order to settle the question as to whether or not these two forms are specifically distinct, it will be necessary to resort to biometric determinations on more extensiA-e material. Our statements of the lengths and the relative widths of these forms are based on thirty-six specimens, twenty of which were referable to A. bidentata and sixteen to A. lemmermanni. Another close relative of Ampkisolenia bidentata is undoubtedly A. pahnata. The former species differs from the latter mainly in its slightly larger size and in having two, instead of three, spinules on the antapex; in other words, in not hav- ing any antapical spinule on the right \-ah'e. The systematic separation of these two forms should be regarded as an open question (see Schroder, 1906a, p. 371, and Ostenfeld, 1915, p. 4). This statement is based on the following facts. The antapical mid-spinule in A. pahnata belongs to the right valve, the two remaining spinules on the antapex to the left valve. Plgure 54: 5 shows that, in this species, there is a stage following binary fission in which the antapical spinules of the new valve are very slightly or not at all developed. We have, therefore, reason to sup- pose that one of the two daughter individuals passes through a stage character- ized by the same number and arrangement of the antapical spinules as in A . bi- dentata. It also may be mentioned that in A. bidentata and in ^4. lemmermanni the posterior part of the antapical is characterized by what appears to be a hyaline sheath, while such a sheath has not been found in A. palmata. The presence of a sheath of this kind may possibly be a sign of recent binary fission (see A. asym- meirica, the section on comparison. These facts seem to indicate that A. palmata may be an older developmental stage of A. bidentata. However, this assumption appears to be contradicted by the following facts: — (1) A. bidentata is more frequent than A. palmata in the material of the Expedition, although the diffei- entiation of thecal structures following binary fission has been found to take place very rapidly in all the dinoflagellates that have been studied in this respect; (^) in A. schroderi, A. ckmpes, and A. complanata the antapex appears to have the same number and arrangement of spinules as in ^ . bidentata. Other close relatives of Amphisolenia bidentata are .4. palaeother aides and .4. asymmetrica, both of which are easily distinguished from the first-mentioned species by having three, instead of two, spinules on the antapex. .4. asymmetrica is also characterized by the extreme width of its head. Amphisolenia bidentata, A. lemmermanni, A. pahnata, A. palaeotheroides, and A. asymmetrica form a natural systematic group, the Palmata group. Their 414 THE DINOPHVSOIDAE. closest relatives, outside the group, are probably to be found on the one hand in the ScHRODERi group, on the other in the Bifurcata group (see the section on the subdivisions of the genus, p. 360). Synonymy: — The species was established by Schroder (1900a) under the name of Amphisolenia bidentata. Of the writers who have treated this species only Jorgensen (1923) gives drawings by means of which the determinations may be checked. Forti (1922) gives a reproduction of Schroder's figure of the type speci- men. Cleve (1903b, p. 339) wrote Amphisolenia bidentata as a synonym of A. globifera. This decision was justly rejected by Schroder (1906a, p. 371) as well as by other writers who have treated these species. Cleve's (1903b) data have ac- cordingly not been included in our section on the distribution of A . bidentata. The specimen figiired by Chun (1903, p. 76, fig. e) under Amphisolenia sp. probably belongs to A. hidentdUi, l)ut the drawing does not permit a certain de- termination. The specimen reproduced by Okamura (1907, pi. 3, fig. 15) under the name of Amphisolenia bidentata is referable to A. lemmermanni. It is only 666 ^ long, the ratio between the length and tlie width of its body is 29: 1, and the posterior part of its antapical is not twisted ventrally. Accordhig to statements in the text, the specimens examined by Okamura were 0.1 nun. long and 0.02 mm. wide. These measurements probably are wrong; they presumably stand for 1000 n and 20 /i. The latter figures indicate that some of the specimens examined by Okamura (1907) belonged to A. bidentata. However, their assignment must be regarded as uncertain. Ostenfeld (1915, p. 4) suggested as a possibility that Amphisolenia paltnata Cleve (1901a) from the Malay Archipelago may be identical with A. bidentata. Cleve (1901a) recorded only A. palmata from these waters, Ostenfeld (1915) only A. bidentata. Since no data bearing on this question have yet been published, Cleve's determination has been accepted in this paper. Among the specimens of Amphisolenia bidentata found by Pavillard (1916, p. 61) there was one that agreed with A. pulinala. "Cependant une recolte du 27 decembre 1914 m'a donne, au milieu d'une douzaine d'echantillons typiques, un individu muni d'un troisieme denticule antapical, reproduisant ainsi la dis- position caracteristique de VA. palmata." This specimen, which was referred to A. bidentata by Pavillard (1916), has been referred to A. palmata in the present paper. Pavillard (1916) considered the independence of these two species as probable. SYSTEMATIC ACCOUNT. 415 With regard to Jorgensen's (1923, p. 40) statement that Avi-phisolenia pal- mata and A. bidentata have been confused by previous authors, see below, the section on the distribution. Occurrence: — The species is recorded at sixtj^-three of the 127 stations. There are 13, 7, 4, 10, 21, and 8 stations on the six hues of the Expedition. Of these sixty-three stations, three (4574, 4580, 4583) are in the Cahfornia Current; seven (4587, 4588, 4590, 4592, 4594, 4596, 4605) are in the Mexican Current; six (4609, 4613, 4617, 4634, 4638, 4640) are in the Panamic Area; fi\'e (4646, 4648, 4659, 4667, 4676) are in the Peruvian Current; seven (4689, 4691, 4694, 4695, 4697, 4699, 4700) are in the Easter Island Eddy; one (4713) is in the Galapagos Eddy; twenty-eight (4679, 4701, 4705, 4708, 4712, 4717, 4718, 4719, 4720, 4721, 4722, 4723, 4724, 4725, 4727, 4728, 4729, 4730, 4731, 4732, 4733, 4734, 4736, 4737, 4738, 4739, 4740, 4741) are in the South Equatorial Drift; three (4742, 4743, 4540) are in the South Equatorial Current; two (4541, 4542) are in the Equatorial Counter Current; one (4543) is in the North Equatorial Current. At twent.v-four stations (4588, 4592, 4596, 4640, 4676, 4694, 4700, 4708, 4712, 4718, 4720, 4723 [Salpa stomachs], 4725, 4727, 4729, 4731, 4733 [Salpa stomach], 4738, 4741, 4743, 4540, 4541, 4542, 4543) the species was found in surface catches only. At four stations (4583, 4638, 4648, 4740) it was found in surface catches as well as in catches from 300-0 fathoms; at one station (4737) in a catch from 100-0 fathoms as well as in one from 300-0 fathoms; at one station (4713) in a catch from 150-0 fathoms as well as in one from 300-0 fathoms; at .seven stations (4689, 4701, 4717, 4721, 4724, 4728, 4732) in catches from both 800-0 fathoms and 300-0 fathoms. All the remaining records refer to catches from 300-0 fathoms only. The species is also recorded from Acapulco Harbor near the Mexican Cur- rent; this station is not included in the 127 stations mentioned above. The temperature range of these sixty-three stations at the surface was 68°- 95°; the average was 77.0°. At the twenty-eight stations in the surface catches of which the species was found, the surface temperature ranged from 69° to 84°; the average was 77.7°. At Acapulco the temperature was 83°. For the surface samples the following frequencies are recorded : — 2% at four stations (4592, 4720, 4725, 4741); 1% at twelve stations (4583, 4596, 4638, 4640, 4648, 4676, 4712, 4718, 4731, 4738, 4743, 4543); in the remaining cases less than 1 %. For the samples from 100, 150, 300, or 800 fathoms to the surface the records of frequency are as follows: 8% at one station (4648); 6^ at one station (4739); 4% at three stations (4713, 4719, 4740); 3% at six stations (4587, 4590, 4689, 4697, 4728, 4737) ; 2% at eight stations (4583, 4659, 4701, 4717, 4721, 4724, 4734, 416 THE DINOPHYSOIDAE. 4736); 1% at fourteen stations (4580, 4605, 4609, 4613, 4617, 4638, 4646, 4667, 4691, 4695, 4722, 4730, 4732, 4742); in the remaining cases less than 1%. The species was first recorded by Schroder (1900a) from the Gulf of Naples, w hich thus is the tj'pe locality. In the Atlantic Ocean it was found by Stiiwe (1909) at lat. 3° 50' N., long. 26° 15' W.; by Schroder (1909) at St. Thomas in the West Indies; and l)y Jorgensen (1923) in the Bay of Cadiz. Jorgensen (1923, p. Figure 55. — Occurrence of Amphisolenia bUleidata Schroder. Large, .solid circles indicate records from vertical haiiLs; squares, records from surface hauls; triangles, records from both vertical and sur- face hauls; small, solid circles, stations at which this species was not found; small, open circle.s, stations from which no plankton catches were examined. 40) states that Amphisolenia bidentuta "has a wide distribution in the Atlantic, both in the northern and the southern parts," and that the Atlantic distribution of A. palmata is unknown, as this species "has been confused with and included under A. bideijtata." These statements, although very plausible, are not sup- ported by any jiublished records (see also the section on the comparisons of A. hidentuUi, \). 412). Jorgensen (1923) found A. bidentata to be the most common species of this genus and distributed throughout the Mediterranean. In the Gulf of Lyons it was found by Pavillard (1916); in the Ligurian Sea by Forti (1922); in the Adriatic Sea by Entz (1902b, 1905), Schroder (1911), and Schiller (1911a, 1912). In the Red Sea it was found by Ostenfeld and Schmidt (1901), Schroder (1906a), and Karsten (1907) ; in the Gulf of Aden by Ostenfeld and Schmidt (1901), Schroder (1906a), and Karsten (1907); in the Arabian Sea by Ostenfeld and SYSTEMATIC ACCOUNT. 417 Schmidt (1901) and by Schroder (1906a); in the Indian Ocean by Ostenfeld (1902), Schroder (1906a), and Karsten (1907). Karsten recorded it from twenty- six of the stations of the Valdivia Expedition in the Indian Ocean, between lat. 8° N. and lat. 10° S., and from near the west coast of Sumatra to near the east coast of Africa. From the East Indies it was recorded by Schroder (1906a) and Ostenfeld (1915); from the Gulf of Siam by Schmidt (1901); from Japanese waters by Schroder (1900a) and Okamura (1907, 1912); from the eastern part of the Pacific, at lat. 33° N., long. 135° W., and lat. 25° N., long. 140° W., by Schroder (1906a). Most of the records referred to above were from surface catches (see the section on the distribution of the genus, p. 349). The only records of closing-net hauls published as yet are those of Karsten (1907) from Station 268 of the Val- divia Expedition, lat. 9° 6' N., long. 53° 41' E., viz., hauls from 17-0 m., 24-4 m., 30-U m., 63-46 m., and 105-88 m.; all the individuals are reported as living. The species occurs in waters of the following temperatures and salinities : — Ostenfeld and Schmidt (1901): temperature range of four stations, 79.0°-84.5°; average, 80.5°. Ostenfeld (1902): temperature, 82.4°; salmity, 33.8. Okamura (1912): temperature, 76.5°-76.8°. Stuwe (1909): temperature, 80.1°. Forti (1922) : the temperature range of four stations, 55.6°-65 .6°, and the average, 61.0°. Of the writers who have contributed to our knowledge of the distribution of this species only Schroder (1900a), Okamura (1907), and Jorgensen (1923) give any information or figures by means of which the accuracy of their deternuna- tions may be judged. The figure given by Forti (1922) is a reproduction of Schroder's (1900a) drawing of the type specimeh. With regard to Okamura's (1907) determinations, see above. For a discussion of Ostenfeld's (1915) suggestion that the specimens recorded by Cleve (1901a) from the Malay Archipelago under the name of Amphisolenia palmata would belong to A. bidentata, see above. Amphisolenia bidentata evidently is one of the most common species of its genus, of world-wide distribution in tropical, subtropical, and warm-temperate seas and sometimes fairly rich in individuals (e. g., it was found to be "common" in the Arabian Sea by Schroder, 1906a). In the material of the Expedition it is by far the most common representative of this genus. The most outstanduig peculiarities of its distribution in the Eastern Pacific, according to the records of the Expedition, are its eupelagic habitat and its freciuent occurrence in all regions except in the Peruvian Current and in the part of the South Equatorial Drift which is nearest to this current. jAjiiong all the species of this genus it is the most frequent in surface waters. 418 THE DINOPHYSOIDAE. Figure 56. — 1, Amphisolenia Mdentata Schroder. Station 4730 (300-0 fathonxs). 2, .4. Icminer- marini Kofoid. Station 4730 (300-0 fathom.'i). 3, A. ^mlmata St«in. Station 4734 (300-0 fathoms). 4, A. jalaeolheroides Kofoid. Station 4732 (300-0 fathom.s). ,5, A. asymmetrica Kofoid. Station 4732 (300-fl fathom.s). 6, yl. frz/wm/Zo Murray and Whitting. Station 4699 (300-0 fat hoin.s). 7, .4. projecta Kofoid. Station 4701 (300-0 fatliom.s). 8, A. tliririax Schiitt. Station 4737 (300-0 fathoms). 9, A. qitudricawhi Kofoid. Station 409.5 (300-0 fathoms). 10, .4. q-uitiiigeal protuberance, and its left lip is some- what higher than the right. It is therefore dii'ected anteriorly and somewhat to the right. The transverse flagellum (tr. fl., Kofoid, 1906c, pi. Ki, fig. 6) arises in the flagellar pore and passes anteriorly between the sulcal lists and encircles the head from the left aromid to the right as in other girdled dmoflagellates. SYSTEMATIC ACCOUNT. 453 The longitudinal flagellu7n has not been seen by us, nor is there any clear statement in the literature that this flagellum has been seen by any other in- vestigator. Hjort (1911, pi. 12: 3) gives a reproduction of Kofoid's (1906c, pi. 15, fig. 1 ) drawing of the type specimen of Triposolenia bicornis to which a long trail- ing flagellum has been added. However, this addition probably is purely dia- grammatical. The posterior displacement of the flagellar pore is suggestive of a possible reduction of this flagellum to- a rudimentary state. The thecal ivall is of nearly uniform thickness except on the conca^'e faces of the antapicals, where, as in Ceratium, it is somewhat thickened, especially in the more robust individuals. In most species it is hyaline and under low magnifica- tions apparently structureless. When these species are examined under high magnifications, however, the wall is found to be very faintly and minutely spotted or flecked; sometimes it also is striate. In but a smgle species, Triposolenia trun- cala, are there plainly visible pits and reticulation; this is undoubtedly to be con- sidered the more primitive condition. Pores are found only in the antapical tubercles and along the sagittal suture on the ventral side of the neck. The sagittal suture is not structurally marked except by a faint Ime, or a few scattered pores along the ventral side of the neck or in the tubercles on the major flexures of the antapicals. The three plates, epithecal, cingular, and hypothecal, of each valve have never been found parted in the specimens thus far observed. The protoplasmic contents (Plate 15, fig. 2) are very hyaline and coarsely granular. The nucleus, which is located in the midbody, is compressed and ellip- soidal, and has a moniliform chromatin reticulum. In the subgenus Ramiciformia, the nucleus and the greater part of the protoplasm are found in the pendent region of the midbody. Near the nucleus is found a pusxile, which opens by a slender canal into the flagellar pore. Oiie or more accessory pusules, also opening into the flagellar pore, may Lie near it in the plasma. Occasionally the plasma is crowded with highly refractive ellipsoidal metaplasmic bodies of amyloid character. In many individuals it is possible to find small irregular, very pale yellowish-green, sometimes nearly colorless chromatophores. No species appears to possess the pro- toplasmic structures with a characteristic pattern in either form or distribution. The length of the body, measured from the foremost point of the head per- pendicularly to the line connecting the tips of the antapicals (Figure 61), is a fairly ^'ariable character even within the various species. The range of variation in length for the genus as a whole thus far established is 92 /x to 243 //. Reproduction: — Neither sexual reproduction nor binary fission has been observed as yet. Nor is there any evidence of a megacytic phase preceding binary fission. 454 THE DINOPHYSOIDAE. The entire course of the line of fission is not known with certainty. Most probably it is as follows: — On the dorsal side of the body, along the middle of the head, neck, anterior process, niidbody and antapicals. On the ventral side of the body, along the middle of the heail, along the left side of the sulcus, to the left of the flagellar pore, crossing the left sulcal list just behhid this pore, and along the middle of the anterior process, midbody, and antapicals. This assumption is sup- ported by the conditions in Amphisolenia, as established in the present paper (Plate 13, fig. 1-6, 8), and by the following facts which have been indicated in the section on the organology. The left sulcal Ust swings in toward the median line just behind the flagellar pore. Near the sagittal plane, just behind the pore, it has usually a fine rib, behind \\hich it continues for a short distance along the ventral side of the anterior process. This rib is, according to our opinion, homologous with the fisson rib of the left sulcal list in other genera of this tribe, e.g., Amphiso- lenia, Phalacroma, and Dinophysis (Plate 13, fig. 6; Plate 1, fig. 6; Plate 5, fig. 3). iSince no representative of Triposolenia has been observed in a stage of division, it cannot be stated positively that the left sulcal list is parted at this rib, and that its anterior and posterior parts are going to the left and to the right valve respec- tively, but the homologies here indicated make such a division probable. In any case, the continuation of the left sulcal list behind " the fission rib " in Triposolenia runs near the sagittal plane and may well belong to the right valve. Distribution: — Our present knowledge of the general oceanic distribution of Triposolenia must be regarded as very incomplete. The distributional data avail- able, given by Kofoid (1906c, 1907a) and Jorgensen (1923), are accompanied by drawings and thus can be checked. The genus is marineand probablyexclusivelyeupelagic. Wehavefound it to be extensively distributed throughout the tropical, subtropical, and warm-temperate regions of the Eastern Pacific, and it has been recorded from the southwest coast of Portugal, from the Mediterranean, and from the Marmora Sea by Jorgensen (1923). It is therefore probably of world-wide distribution in tropical, sub- tropical, and warm-temperate waters. The fact that it has not been found as yet except in the regions mentioned above is presumably largely due to its very rare occurrence at the surface and the limitation of most plankton collections to sur- face levels. The typical habitat of the genus appears to be the deeper levels of photo- synthesis, predominantly between 200 and 100 meters (Jorgensen, 1923, p. 42, 43). Only one of the eight known species, Triposolenia bicornis, has been recorded SYSTEMATIC ACCOUNT. 455 as yet from surface waters (Kofoid, 190Gc, p. 107). No closing-net data have been published as yet. All the species seem to be rare. Even the most frequent members of the genus, Triposolenia bicornis and T. ramiciformis, have never been found up till the present time except in small numbers. The most outstanding peculiarity of the distribution of Triposolenia accord- ing to the data of the Expedition is that the genus did not occur in any of the eighty-one surface catches nor in any of the twenty-four samples of Salpa stom- achs from specunens taken m surface waters, although it is recorded from as many as fifty-nine (46.5%) out of the 127 stations of the Expedition at which Dino- flagellates were found. Most of the records are from vertical hauls from 300-0 fathoms, some from vertical hauls from 800, 400, 150, 100-0 fathoms. Vertical hauls were made at only sixty-eight out of the 127 stations. Representatives of the genus thus were found at no less than 86.8% of the stations at which vertical hauls were made. Disregarding the stations at which surface catches only were made, the record stations of Triposolenia are distributed over the area covered by the Expedition in the following mamier (Plate 27): — 4 (100.0%) out of the 4 stations in the California Current 5 (100.0%) out of the 5 stations in the Mexican Current 6 (100.0%) out of the 6 stations in the Panamic Area 16 (76.2%) out of the 21 stations in the Peruvian Current 3 (60.0%) out of the 5 stations in the Easter Island Eddy 2 (100.0%) out of the 2 stations in the Galapagos Eddy 22 (95.7%) out of the 23 stations in the South Eciuatorial Drift 1 (100.0%) out of the 1 station in the South Equatorial Current No vertical collections were made in the Equatorial Counter Current or in the North Equatorial Current. The numbers of these fifty-nine stations are as follows: — 4571, 4574, 4580, 4583 (California Current) ; 4587, 4590, 4594, 4598, 4605 (Mexican Current) ; 4609, 4613, 4617, 4634, 4637, 4638 (Panamic .4rea); 4648, 4650, 4652, 4655, 4657, 4659, 4662, 4663, 4665, 4666, 4667, 4668, 4670, 4671, 4675, 4676 (Peruvian Current); 4691, 4697, 4699 (Easter Island Eddy) ; 4713, 4715 (Galapagos Eddy) ; 4679, 4681, 4683, 4685, 4687, 4701, 4705, 4707, 4709, 4711, 4717, 4719, 4721, 4722, 4724, 4728, 4730, 4732, 4734, 4737, 4739, 4740 (South Equatorial Drift) ; 4742 (South Equatorial Current). The fifty-nine stations at which Triposolenia was found (Plate 27) are scat- 456 THE DINOPHYSOIDAE. tered over the whole area covered by the Expedition and do not show any pro- nounced concentration in any of the regions into which this area has been divided. In the California Current, the Mexican Current, the Panamic Area, the Gala- pagos Eddy, and the South Equatorial Current the genus occurred at all the sta- tions at which vertical hauls were made; in the South Equatorial Drift it was found at no less than twenty-two out of the twenty-three stations with vertical hauls. In the Peruvian Current it occurred at only sixteen out of the twenty-one stations with vertical hauls and in the Easter Island Eddy at only three oiit of the five stations with vertical hauls. If, on the other hand, the number of records of species, instead of the number of record stations, is considered, there appears to be a moderate concentration of the genus in the Peruvian Current and in the region of the South Equatorial Drift that is under the direct influence of this cur- rent, but in the Easter Island Eddy the genus is strikingly scarce. In the Ea.ster Island Eddy the genus is recorded three times only; two of these three records are of T. hicornis, the remaining one of T. ramiciformis. When the distribution of Triposolenia (Plate 27) in the region investigated by the Expedition is compared with that of the closely related Amphisolenia (Plate 26), we find the following outstanding similarities and differences. (/) Both genera have been found throughout the whole region investigated. (2) While in Triposolenia no species has been recorded from the surface, several spe- cies of Amphisolenia have been found in surface catches. However, most of the species of Amphisolenia appear also to be more or less limited to the deeper levels of photosynthesis. (3) WTiile in Triposolenia there is a relatively large number of records of species in the Peruvian Current and in the portion of the South Equa- torial Drift that is under the direct influence of this current, a moderate number of records of species in the western portion of the South Equatorial Drift, and a very small number of records of species in the Easter Island Eddy, there is in Amphisolenia a strikingly small number of records of species in the Peruvian Current and in the portion of the South Equatorial Drift that is under the direct influence of this current and a remarkably large number of records of species in the western and middle portions of the South Equatorial Drift and in the Easter Island Eddy. In order to understand the differences between the horizontal distribution of Triposolenia and that of Amphisolenia in the area investigated by the Expedi- tion, it is necessary to consider the horizontal distributions of the individual species of these two genera according to the data of the Expedition. Disregarding the species of rare occurrence, the species of Triposolenia be- SYSTEMATIC ACCOUNT. 457 long to any one of three main distributional types. Type 1 : This type is charac- terized by an even distribution throughout the area investigated by the Expedi- tion. It is represented by one species onlj', \iz., T. bicornis, which according to available data is the most common representative of this genus (Figure 67). Type 2 : This type is characterized by a distribution in waters of high average temperature. It is represented by one species only, viz., T. truncata (Figure 63). Most of the record stations of this species are in the Panamic Area and in the western portion of the South Equatorial Drift. In the relatively cool Peruvian Current and in the portion of the South Equatorial Drift that is under the direct mfluence of this current, this species appears to be either very rare or absent. Type 3: This type is characterized by a comparatively frequent occurrence in the relatively cool Peruvian Current and tJie portion of the South Equatorial Drift that is imder the direct influence of this current. It comprises three of the most frequent species of this genus. Its most typical representative is T. longicornis (Figure 70). This species is recorded at a rather large number of stations in the Peruvian Current and the eastern portion of the South Equatorial Drift but has never been found in the Easter Island Eddy and in the western portion of the South Equatorial Drift. Another typical member of this group is T. ramiciformis, which next to T. bicornis is the most common species of this genus. T. ramici- formis (Figure 74) is recorded at forty of the 127 stations of the Expedition. Of these forty stations no less than twenty-six are in the Peruvian Current and in the portion of the South Equatorial Drift that is under the direct influence of this current, and two are in the Galapagos Eddy, which also is directly influenced by this current. In the western portion of the South Equatorial Drift and in the Easter Island Eddy, this species was found altogether at two stations only. T. depressa (Figure 65) also belongs to this group, but its distribution is .somewhat atypical. It prevails in the Peruvian Current and in the eastern portion of the South Equatorial Drift and is absent in the Easter Island Eddy, but there are also some record stations in the western portion of the South Equatorial Drift. The most frequent species of Amphisolenia belong to any one of three main distributional types. Type 1 : This type is characterized by an almost even dis- tribution throughout the area investigated by the Expedition. As examples A. extensa (Figure 52) and A. globifera (Figiu'e 53) may be mentioned. Type 2: The species of this distributional type occur throughout the whole area mvestigated, but they are less frequent in the relatively cool Peruvian Current and in the east- ern portion of the South Equatorial Drift, which is duectly mfluenced by this current, than they are in the western portion of the South Equatorial Drift. This 458 THE DINOPHYSOIDAE. type includes A. bidentata (Figure 55) and .4. lernmermanni (Figure 57). Type 3: The species of this type occur in waters of high average temperature. Their main habitat appears to be the western portion of the South Equatorial Drift and the Easter Island Eddy, and they have not been found as yet in the Peru^'ian Cur- rent and in the eastern portion of the South Equatorial Drift. This type includes A. schauinslandi (Figure 51), .4. thrinax (Figure 59), and probably A. redangtilata , A. bifurcata, A. quadricauda, and A. quinquecauda. The differences between the horizontal distribution of Triposolenia and that of .Vmphisolenia mentioned above are therefore due to the fact that in Triposo- lenia most of the species of relatively frequent occurrence prevail ui the compara- tively cool Peruvian Current and in the portion of the South Equatorial Drift that is directly influenced by this current, while in Amphisolenia most of the rela- tively common species are warm-water forms with theii* principal habitat in the middle and western portions of the South Equatorial Drift and in the Easter Island Eddy and are absent or of rare occurrence in the Peruvian Current and in the eastern portion of the South Equatorial Drift. The similarities in the hori- zontal distribution of Triposolenia and Amphisolenia are due to the fact that both genera mclude species of rather wide amplitude of thermal adaptation and there- fore of more or less even distribution throughout the area investigated by the Expedition. The prevalence of the most frequent species of Triposolenia in the Peruvian Current and in the eastern portion of the South Equatorial Drift indicates that these forms have their optunum habitat in relatively cool waters. This is pre- sumably the explanation of the peculiar fact that this genus never was found in the surface catches of the Expedition, although it appears to be structurally well adapted to waters of low buoyancy. Its species are also less adapted by structure to vertical migration than are the slender, linear species of Amphisolenia. There are 15t) records of species of Triposolenia from vertical catches from 300 (800, 400, 150, 100)-0 fathoms. Out of these 156 records only three (1.9%; Stations 4571, 4648, 4663) showed a frequency of 2%; thirty-six (23.1%; Stations 4574, 4580, 4583, 4598 [3 records], 4605, 4613, 4617 [2 records], 4634, 4637, 4648, 4650 [2 records], 4652, 4655, 4657, 4659, 4662, 4666, 4668 [2 records], 4671 [2 records], 4675, 4701, 4713 [2 records], 4715 [2 records], 4724, 4730, 4739, 4742 [2 records]) showed a frequency of 1%.; 117 (75.0%) showed a frecjucncy of less than 1%. Coincident occurrence of different species of Triposolenia is recorded at the following of the fifty-nine stations mentioned above : six species occurred coinci- SYSTEMATIC ACCOUNT. 459 dently at two stations (3.4%; Stations 4580, 4711); five species at five stations (8.5%; Stations 4583, 4598, 4613, 4709, 4713); four species at seven stations (11.9%; Stations 4650, 4652, 4659, 4667, 4721, 4724, 4742); three species at six- teen stations (27.1%; Stations 4587, 4617, 4637, 4662, 4668, 4671, 4676, 4681, 4683, 4701, 4705, 4715, 4717, 4730, 4732, 4740); and two species at fourteen sta- tions (23.7%; Stations 4594, 4605, 4634, 4638, 4648, 4655, 4665, 4670, 4675, 4679, 4707, 4722, 4737, 4739). Historical Discussion The genus Triposolenia was established by Kofoid (1906c), who founded it on five new species, T. iniucata, T. depressa, T. bicornis, T. raTinciformis, and T. exilis. In a later paper (1907a) the same author described three more species, T. longicornis, T.fahda, and T. ambulatrix. These eight species, besides a new one, T. intermedia, described in the present paper, are the only species of this genus known at the present time. In Kofoid's works (1906c, d, e) an extensive description and discussion of the genus as a whole are gi\-en, and a subdivision of the genus in two subgenera, Posterocornia and Ramiciformia, is suggested. The only writer besides Kofoid, who has recorded the genus, is Jorgensen (1923). This author describes, figures, and gives the distribution of three of the species established by Kofoid but does not treat the genus as a whole. Kofoid (1906c, p. 93) suggests that Amphisolenia tripos (nomen nudum) Schiitt (1893) may refer to a species of Triposolenia. However, there 'is no in- formation in the literature as to the structure of this species (see the section on the historical discussion of the genus Amphisolenia, p. 353). Adaptive and Systematic Value of the Characters. Principles Used in THE Descriptions of the Species The genus Triposolenia, as previously mentioned, is confhied to waters of tropical, subtropical, and warm-temperate nature or origin, in other words, to waters of comparatively low viscosity. Its peculiar organization is probably in part to be interpreted as an adaptation to the relati^•ely low carrying capacity of these waters. The asymmetry caused by the deflection to the left of the posterior parts of the antapicals has been analyzed by Kofoid (1906e) and shown to be adaptive. Its effect is to turn the organism upon its right face, as soon as it begins to descend while oriented with the anterior end of the body uppermost. In this sidewise 460 THE DINOPHYSOIDAE. position the body offers a maximal surface of resistance to the downward move- ment and thus facilitates the remaining of the organism on a level of optimum illumination. For further information on this interesting point, reference is made to the mentioned article by Kofoid (1906e). For further discussion of the adaptive significance of the organization of Triposolenia, see Kofoid (1906c, p. 95, 96; 1906d, p. 121; 1906e, p. 129). The characters by which the species of this genus have been distinguished from each other are principally the shape of the head, the inclination, curvature, and relati\'e length and width of the neck, the shape of the midbody, and the origin, relati\-e length, cur\atin-e, spread, and structure of the antapicals. Pro- nounced structural difTerentiation of the thecal wall appears only in Triposolenia truncata among the species of this genus thus far known. Although the size of the body is a character of rather great variability within the species, it is not without systematic value. In support of this statement, the foUowmg examples may be given: — Triposolenia depressa and T. truncata, 92- 122 fi and 109-145 n long respectively, are comparatively small species; while T. Jongicornis and T. fatula are relatively large, measuring 210-243 n and 175 ii in length, respectively. The structures least modified within the genus are the anterior process and the cingular and sulcal lists. Differences in the position, size, and shape of nucleus and pusules and in the number, form, and color of the chromatophores and meta- plastic inclusions appear to have little significance in the matter of specific dis- tinctions. The individual species characters are all subject to \-ariations of the fluctuat- ing kind, in varying degrees of amplitude. Thus the relati\'e length of the neck and the shape of the margins of the midbody, in lateral view, are rather variable. The greatest fluctuations are perhaps found in the antapicals, in their relative length and width, in their spread, in the degree and localization of their curvature, in the number and distribution of their tubercles, and in the form and structure of their tips. Fluctuations are least evident in the anterior process and in the head, but, on the other hand, these parts exhibit, generally speaking, rather slight differences in the various species. A comparison between the list of the characters used in this paper for the differentiation of the species and the list of the characters subject to decided variations of the fluctuating kind shows that these lists are almost identical. From this follows that it is not possible, as a rule, to separate the species by single characters; the totality of the characters must be considered (cf. Kofoid, 1906d). SYSTEMATIC ACCOUNT. 461 The great variability and the scarcity of the material available have rendered the establishment of the species very difficult. The division into species as pre- sented in this paper should in some cases be regarded as tentative. The f(jllowing principles have been used in describing the species of this genus : — {!) Wlien not otherwise stated, all characteristics refer to specimens in lateral view. (2) In describing the proportions of the body, the length of the neck, meas- ured m a straight Ime from the posteroventral point of the head to the flagellar pore (Figure 61), has been used as a unit. (3) The methods of measuring axes and angles used in this paper are shown in Figure 60. (4) The methods used in measuring the length of the body and the propor- tions of the different parts of the body are shown in Figure 61. (5) The thecal wall is without structure in all the species in whose diagnoses this character is not mentioned. Subdivisions. Rel.\tionships between the Subgenera and among the Species The genus Triposolenia may conveniently be divided into two subgenera, accordmg to the position of the anterior process, the points of origin of the ant- apicals, and the shape of the midbody. These subgenera, Triposolenia and Rami- ciformia, most probably represent natural systematic units and are, as far as our present knowledge goes, not connected by any transitional forms. They may be defined as follows : — Subgenus 1. Triposolenia nobis, nom. subgen. nov. — Anterior process arises at or near the middle of anterior side of midbody. Midbody centrally lo- cated ; in lateral view subtriangular, more or less rounded triangular, or subellip- soidal; with three distinct margins; postmargin not excessively developed. Ant- apicals arising from posterior part of midbody. Distance in a straight line from center of midbody to tip of ventral antapical always exceeds distance from center of midbody to apex of head. Subgenus 2. Ramiciformia Kofoid. — Point of origin of anterior process shifted to near ventral margin of midbody. Midbody in lateral view ranging from oblong to rounded sack-shaped ; with only two distmct margins, the anteroventral margin of the previous subgenus being suppressed; postmargin extended in a pendent lobe. Antapicals arismg from anterior part of pendent midbody. 462 THE DINOPHYSOIDAE. The subgenus Triposolenia corresponds to the subgenus Posterocornia Kofoid (1906c, p. 101), which thus becomes a synonym. It comprises seven of the nine species hitherto known, viz., T. (runcata, T. depressa, T. bicornis, T. longi- cornis, T. ambidatrix, and T. fatula, all described by Kofoid in previous papers (1906c and 1907a), and T. intermedia, established in the present paper. The type of this subgenus (and thus of the genus) is T. truncata. The subgenus Ramiciformia was estabhshed by Kofoid (1906c, p. 101) and comprises thus far only two species, viz., Triposolenia ramiciformis, which is the type of the subgenus, and T. exilts, both previously described by Kofoid (1906c). By a lapsus calami this subgenus was named Ramiciformis by Kofoid (1906c, p. 97) ; cf. Article 19 of the International Rules of Zoological Nomenclature. If Triposolenia truncata be the most primitive species of this genus thus far known (p. 463), then the subgenus Triposolenia is more primitive than the sub- genus Ramiciformia. The latter prolxabl}' arose from a member of the former. However, as previously mentioned, no transitional forms between the two sub- genera are known at the present time. The present knowledge of the subgenus Triposolenia is too limited to allow any well-founded statements as to the relationships between the species. T. depressa, T. bicornis, T. intermedia, and T. longicornis appear to be very closely connected. T. depressa and T. bicornis intergrade; so probably do T. bicornis and T. intermedia; and T. intermedia appears to form a comiecting link between T. bicornis and T. longicornis. Specimens are sometimes found which in the light of our present knowledge may be assigned to either T. depressa or T. bicornis. The sizes of these four species probably give a good indication of their mutual rela- tionships. T. depressa is 92-122 m long; T. bicornis is 120-153 ix long; T. inter- media, 165-177 ix\ T. longicornis, 210-243 jx. T. fatula is probably rather closely related to T. longicornis. T. truncata and T. ambulatrix, especially the former, occupy rather isolated positions. As has previously been shown, T. truncata is probably the most primitive known member of its genus. Key to the Species of Triposolenia 1. Midbody subtriangular; its postmaigin not extended in a pendent lobe (subgenu.s Triposolenia) . . . .2. 1. Postmargin of midbody extended in a pendent lobe (subgenus Ramieiformia) 8. 2. Epitheca quite concave; thecal wall profusely and strikingly pitted truncata Kofoid. 2. Epitheca convex or more or less flattened: thecal wall structureless or almost so 3. :•(. Antapicals strikingly asymmetrical, both deflected dorsally, especially the dorsal. ami)u/a(n';r Kofoid. 3. Antapicals almost symmetrical or of moderate asymmetry, balanced or almost so 4. 4. Antapicals very long and slender, the ventral three times or more longer than neck 5. 4. .\ntapicals less than three times longer than neck 6. 5. Head 1.3-2.0 times longer than wide; margins of midbody subequal; antai)icals subsymmetrical. longicurnis Kofoid. SYSTEMATIC ACCOUNT. 463 5. Head subspheroidal; anteroventral margin of raidbody shorter than the two others; antapicals of moderate asymmetry faltila Kofoid. 6. Rather large, more than 160 ii; head wider than long intermedia, sp. nov. 6. Medium-sized or small, less that IGO ii; head subspheroidal 7. 7. Length almost always less than 120 /j; longitudinal axis of midbody in most cases foreshortened; geniculation scarcely evident in antapicals depressa Kofoid. 7. Length almost always more than 120 /u; longitudinal axis of midbody not foreshortened; antapicals generally geniculate bicornis Kofoid. 8. Distance between antapicals at level of tips 2.0 and at level of midbody 2.5 the length of midbody. ramiciformii Kofoid. 8. Distance between antajiicals at level of tips as well as at level of midbody 1.6 the length of midbody. exilic Kofoid. Subgenus 1. Triposolenia nobis, nom. subgen. nov. — All the seven species that belong to this subgenus have been found in the material of the Expedition. They have been treated in this paper in the following order indicative of relation- ships:— Triposolenia truncata, T. depressa, T. bicornis, T. intermedia, T. longi- cornis, T.fatula, and T. ambulatrix. ■ Triposolenia truncata Kofoid Plate 14, fig. 1-3. Figure 62, 63 Triposolenia Iruncala Kofoid, 1906c, p. 96, 98, 101, 102, 105, 106, 107, pi. 16, fig.5; 1906d, p. 117, 118, 119, 120. JoRGENSEN, 1923, p. 4:3, fig. 64. Diagnosis: — Head 1.5-2.0 times wider than long. Epitheca concave. Dis- tance between bases of antapicals 1.8-2.3 the length of neck which is three to five times longer than wide and straight or slightly cur\ed. Midbody subtriangular or subrotund; its three margins subequal and con\-cx. Antapicals 2.7-3.4 times longer than neck, generally stout, subsymmetrical ; geniculation scarcely evident; no tubercles; tips truncate and spinulate. Wall profusely pitted; neck reticulated. Length, 109-145 p. Tropical, subtropical, and warm-temperate regions of Eastern Pacific; Medi- terranean; Marmora Sea. Description: — The head is 0.37 (0.31-0.45) the length of the neck and 1.5- 2.0 times wider than long. The anterior face of the epitheca is quite conca^'e. The transverse furrow is flat or generally somewhat concave. The anterior cingu- lar list is 1.5-2.0 times wider than the transverse furrow and has, on each valve, five to six generally slightly developed ribs. The neck is comparatively short and thick. The distance between the mid- points of the bases of the antapicals is 2.1 (1.8-2.3) the length of the neck which is three to five times longer than wide. The neck is rather slightly curved dorsally, exceptionally almost straight, with a dorsal deflection of 5° (1°-14°) from the longitudinal axis of the body. The left sulcal list continues beyond the flagellar 464 THE DINOPm'SOIDAE. pore nearly to the midbody. It has, at least in some specimens, three ribs behind this pore; the anteriormost of these ribs is the fission rib. The anterior process arises at or near the middle of the anterior margin of the midbody and is deflected ventrally 34° (28°-41°) from the longitudinal axis of the body. It measures 0.81 (0.74-0.8G) the length of the neck. The distance from the flagellar pore to the midpoint of the postmargin of the midbody is 2.6 (2.3-2.8) times longer than the neck. FnaiRE 62. — • Triposolenia trnncata Kofoid, lateral view. 1, 3, from the left; the rest from the right. X 430. 1, 2, 3, 5, from Station 4742 (300-0 fathoms); 4, from Station 4037 (300-0 fathoms); (), from Station 4634 (300-0 fathoms). The midbody is subtriangidar or subrotund (Jorgenscn, 1923, fig. 64). Its three margins are subequal, and moderately, or rather boldly, and often nearly ecjually convex. The convexity of the anteroventral margin decreases toward the anterior process; sometimes this margin is almost straight. In some cases the pos- terior margin is somewhat longer than the anterodorsal one; and the latter is somewhat longer than the antero\-entral. SYSTEMATIC ACCOUNT. 465 The antapicals are of moderate length or comparatively short. In most cases the ventral is the longer, being 3.0 (2.7-3.4) times longer than the neck; the corresponding figures for the dorsal are 2.9 (2.7-3.2). Out of ten specimens two had subequal antapicals ; and in the specimen with the greatest difference in this respect the ventral was 3.2 and the dorsal 2.9 times longer than the neck. The distance between the center of the midbody and the tip of the ventral antapieal is subequal to the distance from the midpomt of the postmargin of the midbody to the anterior border of the neck. In most cases the antapicals are rather stout and taper gradually from the bases to the tips; sometimes they are widest at some distance from the bases. The ventral seems generally to be a little more slender than the dorsal; the former is, in the middle, 7-14, in most cases 8-11, the latter 6-13 (7-10) times longer than wide. The dorsoventral diameter of the tips is 0.4-0.7 that of the respective bases. The antapicals are almost synmietrical, the regions of major flexures and the tips being, respectively, approximately equi- distant from the center of the midbody and m most cases also from the middle of the epitheca. The curvature is, as a rule, greatest in the middle third of the length ; in this region it is, in most cases, nearly uniform, but sometmies more or less localized in a knee. In some specimens the antapicals are rather strongly curved proxinially, and their distal halves are almost straight (Jorgensen, 1923, fig. 64). The distance between the tips is 3.7 (3.2-4.5) and between the outer margins at the place of greatest spread 4.3 (3.9-5.0) times longer than the neck. The distal fourth of the antapicals is deflected slightly to the left. The tips are truncate, and each tip has, on the dorsal and ventral angles, a pair of short, fine spinules. Sometimes the outer spinule alone is present and is larger than usual. No tuber- cles are developed. The thecal wall is distinctly and profusely pitted (pores?), more sparingly so toward the base of the anterior process and toward the tips of the antapicals. On the neck the pits gradually widen to a rather coarse reticulum, 2-4 meshes wide on each valve. The nucleus is small (m some cases observed) and is located posteriorly. There are a few irregular chromatophores, or numerous elUpsoidal ones gathered along the ventral margin of the midbody. The proportions of ten specimens were measured. The length of these speci- mens was 130 (119-145) M. Dimensions: — Our specimens: Length of body, 113-145 /u (type, 131 m)- Distance from apex to midpoint of postmargin of midbody, 77-101 ^ (average, 84 m; type, 81 n). Length of head, 6.5-10.0 /i (average, 7.7 ai; type, 7.2 ;u). Length 466 THE DINOPHYSOIDAE. of neck, 19.0-23.5 m (average, 21 n\ type, 21 n). Length of anterior process, 14- 19 ij. (average, 17 n; type, 17 /x). Distance between flagellar pore and midpoint of post margin of midbodj', 49-65 ix (average, 54 /j; type, 51 m). Distance between midpoints of bases of antapicals, 39-50 ^ (average, 44 /i; type, 42 /i). Length of \entral antapical, 50-71 yu (average, 64 ^ ; type, 65 ;i). Length of dorsal antapical, 50-69 yu (average, 61 yu; type, 61 n). Distance between tips of antapicals, 67-105 fi (average, 78 /i; type, 67 m)- Distance between outer margins of antapicals at place of greatest spread, 78-117 n (average, 91 /j; type, 82 /j). Jorgensen's (1923) specimens: Length of body of figured specimen, 109 /u. "Distance between points of the antapicals, 62-81 ai, midbody 34 m high and 38-42 // long; distance from apex to lowest (convex) boundary line of the midbody 68 ju." Variations: — Judgmg by the specimens thus far observed, this species ap- pears to be fairly uniform in proportions, except with regard to the antapicals. These vary considerably in thickness but less in length, thus giving to the organ- ism different degrees of robustness. The spread and the shape of the antapicals and the size of the body are also subject to rather considerable variations. Comparisons: — There can hardly be any doubt that the specimens assigned to Triposolenia truncaUi in the present paper belong to the same species as those from San Diego treated under this name by Kofoid (1906c, d). The type speci- men (Kofoid, 1906c, pi. 16, fig. 5) falls within the range of variation of the speci- mens found m the material of the Expedition. In the specimen figured by Jorgensen (1923, fig. 64) the midbody is more rounded and there are antapicals straighter than in our specimens. In spite of these difTerences, Jorgensen's determination of this form is probably correct. Triposolenia truncata is probably the most primitive member of its genus and occupies a rather isolated position. It differs from all other species of the genus in its relatively stouter facies, the concave anterior side of its epitheca, and the very distinct structural differentiation of its thecal wall. It also is characterized l)y its small size, by its symmetrical and strikingly balanced antapicals, which, as a rule, have little or no geniculation, by the truncate tips of its antapicals, and by the absence of antapical tubercles. Its small size and relativelj' short and thick neck recall T. deprcssa. With regard to its relationship to Amphisolenia inflata, see the section on the subdivisions of the family (p. 336). The rugose surface of the thecal wall of this species, which probably is a primitive character, may be interpreted as an adaptation to flotation. In other species of the genus increase in surface is acquired by a greater elongation of processes. Synonymy: — The species was established by Kofoid (1906c) under the name SYSTEMATIC ACCOUNT. 467 of Triposolenia truncata. The only other writer to treat it, viz., Jorgensen (1923), used the same name. This is the type species of the genus. Occurrence: — The species is recorded at nineteen of the 127 stations. There are 4, 5, 0, 2, 0, and 2 stations on the six lines of the Expedition. Of these nineteen stations, two (4580, 4583) are in the California Current; five (4613, 4617, 4634, 4637, 4638) are in the Panamic Area; two (4650, 4655) are in the -Peruvian Cur- rent; one (4715) is in the Galapagos Eddy; eight (4711, 4717, 4724, 4730, 4732, FifiURE 63.— Occurrence of Triposolenia Inmcala Kofoid. Large, solid circle.s indicate records from vertical havils; small, solid circles, stations at which this species was not found; small, open circles, stations from which no plankton catches were examined. 4737, 4739, 4740) are in the South Equatorial Drift; and one (4742) is in the South Equatorial Current. At one station (4655) the species is recorded from 400-0 fathoms. All the other records refer to catches from 300-0 fathoms. The temperatiu-e range of these nineteen stations at the surface was 65°-83°; the average was 77.1°. The frequency does not exceed 1%, which is recorded at three stations (4634, 4637, 4715). The species was first recorded by Kofoid (1906c) from San Diego, California, the type locality, where it occurred sparingly in plankton from 180-0 fathoms during both winter and sunmier. Jorgensen (1923) found it in the Balearic Sea, east of Corsica, near Euboea, and in the Marmora Sea, "only in deeper water." This is probably a eupelagic species of wide distribution in tropical, sub- 468 THE DINOPHYSOIDAE. tropical, and warm-temperate Seas. In the Eastern Pacific it has, according to the records of the Expedition, a distribution that is more in accordance with the dis- tribution of the genus Amphisolenia than that of any other species of the genus Triposolenia. This similarity in distribution is very interesting because this spe- cies is more closely related to Amphisolenia than any other known member of its genus. According to our records, it occurs in waters of high average temperature (77.1°). Most of the record stations are in the warm Panamic Area and in the South Equatorial Drift ; only two record stations are in the relatively cool Peru- vian Current. Its occurrence at San Diego and in the Mediterranean shows, however, that it has a rather wide amplitude of thermal adaptation in warm- water seas. Its absence from the very warm Mexican Current may be worth mentioning. Triposolenia depressa Kofoid Plate 14, fig. 4-9. Figure 64, 05 Triposolenia depressa Kofoid, 1906c, p. 98, 99, 101, 102, 104, pi. 16, fig. 3,4; 1906d, p. 118, 119, lUO; 1906P, p. 130, fig. A, B. Triposolenia sp., Steuer, 1910, fig. 106; 1911, fig. 82. Triposolenia ambulalri.v Joroensen, 1923, p. 42, 43, fig. 63. Diagnosis: — Head small, subspheroidal or subcylLndrical. Epitheca con- vex. Distance between bases of antapicals 1.3-1.6 the length of neck which is 6-11 times longer than wide and almost straight. Midbody subellipsoidal or subtriangular, its longitudinal axis often foreshortened; its margins convex, the anteroventral somewhat shorter than the others and sometimes straight or even concave. Antapicals 2.1-2.8 times longer than neck, slender, subsymmetrical or more or less asymmetrical; geniculation scarcely evident; tuberculate; tips rounded, sometimes truncate and spinulate. Length, 92-122 pi. Tropical, subtropical, and warm-temperate regions of Eastern Pacific; Marmora Sea. Description: — The head is 0.24 (0.21-0.33) the length of the neck, sub- spheroidal, in most cases about as long as wide, sometunes narrower, even sub- cylindrical. The anterior face of the epitheca is more or less convex. The trans- verse furrow is flat or slightly concave. The anterior cingular list is 2.0-2.5 times wider than the transverse furrow and has, on each valve, fi\e to eight sometunes but slightly developed ribs. The neck is comparatively short and of moderate width. The distance be- tween the midpoints of the bases of the antapicals is 1.4 (1.3-1.6) the length of the neck which is 6-11 (in most cases 8-10) times longer than wide. The neck is SYSTEMATIC ACCOUNT. 469 almost straight or slightly curved dorsally, with a dorsal deflection of 21° (13°- 25°) from the longitudinal axis of the body. The left sulcal list has a fission rib and, anteriorly, several other ribs. The anterior process arises at a distance equaling or slightly exceeding half the dorsoventral width of the process, ventrally to the middle of the anterior margin of the midbody, and is deflected ventrally 28° (23°-33°) from the longi- tudmal axis of the body. It measures 0.52 (0.46-0.59) the length of the neck. The distance from the flagellar pore to the midpoint of the postmargin of the midbody is 1.6 (1.5-1.8) the length of the neck. FiGDRE 64. — Triposolenia depressa Kofoid, lateral view. 1, seen from the left; 2 and 3, from theright. X 430. 1, 2, from Station 4742 (300-fl fathoms); 3, from Station 4732 (300-0 fathoms). The midbody is often characterized by the foreshortening of its longitudinal axis, which gives it a subellipsoidal shape with its major axis in the dorsoventral position. However, sometimes this foreshortening does not occur, and the mid- body has a subtriangular shape. The three margins are more or less convex, some- times, due to parasitic infection (?), even to such an extent as to give the midbody a subrotund shape ; in a few cases the anteroventral margin is nearly straight or even slightly concave. The anteroventral margin is always somewhat shorter than the two others, due to the ventral dislocation of the anterior process; and the postmargin is always the longest. The antapicals are of moderate lengths. In most cases the ventral is the longer, being 2.5 (2.2-2.8) times longer than the neck; the corresponding values for the dorsal are 2.2 (2.1-2.6) ; see the section on the comparisons of this species. Out of seven specimens two had subequal antapicals; and in the specimen with the greatest difference in this respect the ventral antapical was 2.5 and the dorsal 2.1 times longer than the neck. In three out of these seven specimens the distance between the center of the midbody and the tip of the ventral antapical was sub- 470 THE DINOPHYSOIDAE. equal to the distance from the midpoint of the postmargin of the midbody to the middle of the transverse furrow. The four other specimens deviated only slightly from this type. The antapicals are of almost the same width, in most cases slender, and nearly uniform in width throughout their entire lengths. The ventral is 15-29 times longer than its average width. The two processes are fairly strongly curved; sometimes they are curved to about the same extent, sometimes either one of them is more cur\ed than the other. The curvature is sometimes almost evenly distributed, but in most cases it is mf)re or less localized in the proximal half of the antapical; in the latter case geniculation is scarcely evident. Distally the antapicals exhibit in most specimens a slight sigmoid flexure. The distance between the tips is 2.6 (2.3-3.1) and between the outer margins at the place of greatest spread 3.2 (3.0-3.7) times longer than the neck. In dorsoventral view the antapicals are slightly curved; sometimes they are bent to the right proximally, and to the left distally, the dorsal having more distal curvature than the ventral; sometimes the distal curvature is hardly developed at all. The tips are mostly truncate in dorsoventral view, with minute terminal spinules. In lateral \iew the tips are more or less rounded in most specimens, sometimes truncate. There are two to seven scattered tubercles along the outer margins of the antapicals, all of them in most cases on the distal hah'es of the antapicals. The thecal wall is hyaline, apparently without structure, and with a few pores along the midventral line of the neck and in the tubercles of the antapicals. The nucleus is large. Besides the principal pusule there is an accessory pusule. The chromatophores are few, minute, and spheroidal. The proportions of seven specimens were measured. The length of these specimens was 105 (92-114) n. Dimensions: — Our specimens: Length of body, 92-122 ^ (type, 108 m)- Distance from apex to midpoint of postmargin of midbody, 57-71 ^ (average, 66 n; type, 67 ix). Length of head, 5.0-7.5 ^ (average, 6.0 m; type, 7.5 m)- Length of neck, 21-28 m (average, 25 m; type, 23 n). Length of anterior process, 10-15 n (average, 13 m; type, 13 m). Distance between flagellar pore and midpoint of post- margin of midbody, 34-44 /x (average, 40 n; type, 41 fi). Distance between mid- points of bases of antapicals, 31-89 n (average, 35 n; type, 30 m). Length of ven- tral antapical, 47-74 m (average, 62 m; type, 63 tif- Length of dorsal antapical, 46-66 M (average, 56 n; type, 56 n). Distance between tips of antapicals, 53-75 n (average, 66 m; type, 72 m). Distance between outer margins of antapicals at place of greatest spread, 64-92 ^ (average, 79 m; type 77 n). Jorgensen's (1923) specimen: Length of body, 96 p. "Distance between points of posterior horns SYSTEMATIC ACCOUNT. 471 89 M, midbody 25 n high and 38 broad, distance from apex to lo\\er part of body 65 M." Variations: — The species exhibits a remarkable \ariabiUty in several char- acters. The head is either about as long as wide, subrectangular or subrotund, or it is narrower than long, subcylindrical. In some specimens the midbody is fore- shortened, its major axis being in the dorsoventral position, and its posterior margin is boldly convex and much longer than the anterior margins. In other specimens the midbody is either subrotund, or it is almost equilaterally sub- triangular as in Triposolenia bicoryii's with only slightly con^'ex posterior margin. The relative width of the neck, the cur\ature and spread of the antapicals, the deflection of the dorsal antapical, and the number of antapical tubercles are also rather strikingly variable characters. Comparisons: — Some of the specimens, referred to Triposolenia depressa in this paper (Plate 14, fig. 4, 5), are very similar to the type specimen of this species as drawn by Kofoid (1906c, pi. 16, fig. 3). Others (Figiu-e 64) differ more or less strikingly from the type specimen, and their determmation must be regarded as tentative. Some of these specimens (Figure 64: 2) are much smaller (about 92 ;u) than T. bicornis (120-153 m) and have relati\-ely wide neck and antapicals with- out geniculation as in T. depressa. Their midbody, ho^\■e^•er, is subtriangular with only sUghtly con\-ex posterior margin as in T. bicornis. The specimen of this species found by Jorgensen (1923) and called Triposo- lenia ambulatrix by him differs from the type specimen especially in the shape, deflection, and relative length of the dorsal antapical. The ventral antapical is 2.4, the dorsal only 1.8 times longer than the neck. In the pronounced dorsal de- flection of its dorsal antapical this specimen approaches T. antbulatrix. Triposolenia depressa probably is closely related to T. bicornis. Jorgensen (1923, p. 42, fig. 63) even suggested that it might be but a ''form" of this species. As conceived in this paper, it differs from T. bicornis especially in the smaller size of its body, the foreshortenmg of its midbody, the greater con\exity of the post- margin of its midbody, the relative shortness and greater width of its neck, and the almost complete absence of geniculation in its antapicals. However, as is shown in the section on variability, the two species tend to intergrade in some of the distinctiA'e characters. Synonymy: — The species was established l)y Kofoid (1906c) imder the name of Triposolenia depressa. The form treated by Jorgensen (1923, fig. 63) under the name T. ambulatrix is probably identical with T. depressa. Kofoid (1906e, fig. A, B) used this species to illustrate his discussion on the adaptive 472 THE DINOPHYSOIDAE. nature of the asymmetry in this genus. His figures were reproduced by Steuer (1910, fig. 106; 1911, fig. 82). Kofoid did not make any reference to species in his explanation of these figures; Steuer used the designation Triposolenia sp. Occurrence: — The species is recorded at twenty-two of the 127 stations. There are 2, 7, 3, 5, 4, and 1 stations on the six lines of the Expedition. Of these twenty-two stations, one (4580) is in the California Current; one (4613) is in the Panamic Area; eight (4650, 4652, 4655, 4659, 4662, 4667, 4671, 4676) are in the Peruvian Current; ten (4681, 4683, 4701, 4705, 4709, 4711, 4721, 4724, 4730, Figure 65. — Occurrence of Triposolenia depressa Kofoid. Large, solid circles indicate records from vertical hauls; small, .solid circles, stations at which this species was not found; small, open circles, stations from which no plankton catches were examined. 4732) are in the South Equatorial Drift; one (4713) is in the Galapagos Eddy; and one (4742) is in the South Equatorial Current. At one station (4652) the species is recorded from 100-0 fathoms; at two stations (4662, 4681) from 800-0 fathoms. All the other records refer to samples from 300-0 fathoms. The temperature range of these twenty-two stations at the surface was 65°- 80°; the average was 72.2°. The frequency does not exceed 1 %, which is recorded for two stations (4650, 4655). Triposolenia depressa was first recorded by Kofoid (1906c) from San Diego, California, the type locality, where it was taken in a vertical haul from 95 fathoms. SYSTEMATIC ACCOUNT. 473 Later it was recorded by Jorgensen (1923) from the Marmora Sea as Triposolenia ambulatrix. The species probably is eupelagic and of subtropical and tropical nature, although it sometimes occurs in warm-temperate waters. In the material of the Expedition it is recorded mainly from the relatively cool Peruvian Current and from the part of the South Equatorial Drift that is under the influence of this current. Triposolenia bicornis Kofoid Figure 66, 67 Triposolenia bicornis Kofoid, 1906c, p. 96, 99, 101, 102, 104, 105, 107, pi. 1.5, fig. 1,2, pi. 16, fig. 6; 1906d, p. 118, 119, 120. Hjobt, 1911, p. 367, fig. 12:3. Gran, 1912a, p. 935, fig. 9; 1912b, fig. 2.33. Jorgen- sen, 1923, p. 41, 42, 43, fig. 62. Diagnosis: — Head subspheroidal. Epitheca convex. Distance between bases of antapicals 1.1-1.4 the length of neck which is 9-15 times longer than wide and almost straight. Midbody subtriangular ; its margms subequal or almost so and convex. Antapicals 1.8-2.4 times longer than neck, rather slender, sub- symmetrical, mostly geniculate, tuberculate ; tips rounded, mostly without spin- ules. Length, 120-153 n. Tropical, subtropical, and warm-temperate regions of Eastern Pacific; Mediterranean and off west coast of Portugal. Description: — The head is 0.18 (0.1.5-0.20) the length of the neck and as long as wide, in some cases slightly narrower, in others slightly wider. The an- terior face of the epitheca and the transverse furrow are as in Triposolenia de- pressa. The anterior cmgular list is two to three times wider than the transverse furrow and has, on each valve, ten to twelve delicate ribs. The neck is rather long and narrow. The distance between the midpoints of the bases of the antapicals is 1.2 (1.1-1.4) the length of the neck which is 9-15 times longer than wide. The neck is almost straight or slightly curved dorsally, with a dorsal deflection of 20° (15°-25°) from the longitudinal axis of the body. The left sulcal list has a fission rib. The anterior process arises at about or slightly ventrally to the middle of the anterior margin of the midbody. It is deflected ventrally 17° (12°-21°) from the longitudinal axis of the body. It measures 0.41 (0.36-0.48) the length of the neck. The distance from the flagellar pore to the midpoint of the postmargin of the midbody is 1.4 (1.1-1.5) the length of the neck. The midbody is subtriangular. Its margins are more or less convex, some- times, due to parasitic infection (?), even to an extent so as to give the midbody a 474 THE DLNOPHYSOIDAE. subrotund shape. In some cases they are almost subeqiial, in others the antero- ^•elltral margin is somewhat shorter than the anterodorsal, due to the ventral dis- location of the anterior process; the postmargin is always the longest. The antapicals are of moderate length, rather slender, subsymmetrical, and in most cases with fairly distinct knees near or somewhat distally to the middle. The ventral antapical is 2.1 (1.9-2.4), the dorsal 2.0 (1.8-2.3) times longer than the neck. Out of eight specimens one had subequal antapicals; and in the speci- men with the greatest difference in this respect the ventral was 2.1, the dorsal 1.8 Figure 66. — Triponvlcnia hicmnU Kofoid, lateral vipw. 1, srcn from Ihe left; 2 and 3, from the right. X 430. 1, from Station 4613 (300-0 fathoms); 2, from Station 4638 (300-0 fathoms); 3, from Station 4730 (300-0 fathoms) . times longer than the neck. The ventral is 1(3-26 times longer than its average width. The distance between the tips is 2.4 (1.9-2.9), and between the outer margins at the place of greatest spread 2.8 (2.5-3.1) times longer than the neck. The tips are rounded, mostly without spinules. For characteristics not men- tioned above, see the description of the antapicals of Triposolenia depressa (p. 469). The thecal wall is as in T. drprcsso. The nucleus is large, ellipsoidal, and usually located near the posterior face of the midbody. A centrosome is found on the anterior face of the nucleus. One or two pusules are present along the ventral wall, each with efferent canal leading to the flagellar pore. The chromatophores arc small, irregular, almost colorless, with superficial distribution. The proportions of eight specimens were measured. The length of these specimens was 134 (126-141) fx. SYSTEMATIC ACCOUNT. 475 Dimensions: — Our specimens: Length of body, 120-153 yu (type, 140 /i). In the original description (Kofoid, 1906c, p. 106) a maximum length of 185 ii is given. This high value is probably due to the fact that specimens of Triposolenia intermedia were included in T. bicornis. Distance from apex to midpoint of post- margin of midbody, 76-91 m (average, 84 m; type, 91 fi). Length of head, 5-7 ^ (average, 6 ix; type, 7 m). Length of neck, 30-37 fx (average, 34 ii; type, 35 m). Length of anterior process, 12-16 m (average, 14 yu; type, 16 m)- Distance between flagellar pore and midpoint of postmargin of midbody, 40-52 n (average, 46 n; type, 52 yu). Distance between midpoints of bases of antapicals, 34-44 fi (average, 40 fi; type, 44 /i). Length of ventral antapical, 65-78 ^ (average, 72 m; type, 68 n). Length of dorsal antapical, 65-71 yu (average, 69 n; type, 68 m)- Distance between tips of antapicals, 71-89 n (average, 83 n; type, 89 ix). Distance betw^een outer margins of antapicals at place of greatest spread, 89-101 n (average, 95 yu; type, 95 yu). Jorgensen's (1923) specimens: Length of body of figured specimen, 136 m. "Distance between points of horns almost constantly 102 yu, the midbody (32-) 34 n high by (38-) 43-47 yu long, distance from upper end of cell to lowest boundary line of the midbody 80-85 yu." Variations: — .\s conceived in this paper, Triposolenia bicornis is fairly con- stant in structure and proportions. Comparisons: — Most of our specimens showed a very striking resemblance to the type specimen as drawn by Kofoid (1906c, pi. 15, fig. 1). The specimen figured under the name of Triposolenia bicornis by Jorgensen (1923, fig. 62), also agrees closely with the type and is certainly correctly determined. With regard to the position of this species within its genus, see Triposolenia depressa, T. intermedia, and the section on the subdivisions of the genus (pp. 471, 461). Synonymii: — The species was established by Kofoid (1906c) and later treated by Jorgensen (1923) under the name of Triposolenia bicornis. Reproduc- tions of Kofoid's (1906c) drawing of the type specimen are to be found in Hjort (1911) and in Gran (1912a, b). Occurrence: — The species is recorded at forty-five of the 127 stations. There are 9, 11, 4, 9, 10, and 2 stations on the six lines of the Expedition. Of these forty- five stations, four (4571, 4574, 4580, 4583) are in the California Current; two (4598, 4605) are in the Mexican Current; five (4609, 4613, 4617, 4637, 4638) are in the Panamic Area; nine (4648, 4650, 4652, 4657, 4659, 4667, 4668, 4670, 4671) are in the Peruvian Current; two (4697, 4699) are in the Easter Island Eddy; two (4713, 4715) are in the Galapagos Eddy; twenty (4679, 4681, 4683, 4687, 4701, 476 THE DINOPHYSOIDAE. 4705, 4707, 4709, 4711, 4717, 4721, 4722, 4724, 4728, 4730, 4732, 4734, 4737, 4739, 4740) are in the South Equatorial Drift; and one (4742) is in the South Equatorial Current. At one station (4652) the species is recorded from 100-0 fathoms; at one station (4713) from 150-0 fathoms; at one station (4070) from 800-0 fathoms; at four stations (4681, 4701, 4717, 4724) from 800-0 fathoms as well as from 300-0 fathoms. All other records refer to catches from 300-0 fathoms only. The temperature range of these forty-five stations at the surface was 66°-85°; the average was 74.4°. Figure 67. — Occurrence of Triposnlcnia f)icormx Kofdid. Large, solid circles indicate records from vertical hauls; small, solid circles, stations at which this species was not found; small, open circles, stations from which no plankton catches were examined. At one station (4571) the frequency is 2%. At sixteen stations (4574, 4580, 4583, 4605, 4613, 4617, 4648, 4652, 4657, 4668, 4671, 4713, 4724, 4730, 4739, 4742) the frequency is 1 %. At the remaining twenty-eight stations the frequency is less than 1 %. Previously the species has been taken l)y Kofoid (1906c) near San Diego, California, the type locality, where it frequently occurred in small numbers in vertical hauls from 135-0 fathoms, and rarely in surface catches. Jorgensen (1923) reported it as widely distributed in the Mediterranean, "especially, it would seem, between 100 and 200 m." He also found it off the southwest coast of Portugal. SYSTEMATIC ACCOUNT. 477 This appears to be the most widely distributed and the most abundant spe- cies in the genus. It is probably eupelagic and of world-wide distribution in tropical, subtropical, and warm-temperate seas. According to the records, it is fairly evenly distributed throughout the region investigated by the Expedition. Triposolenia intermedia, sp. nov. Plate 14, fig. 10. Figure 60: 1, 01, 68 Diagnosis: — Head 1.3-2.0 times wider than long. Epitheca convex or flat. Distance between bases of antapicals 1. 3-1.-1 the length of neck, which is 7-11 times longer than wide and gently curved. Midbody subtriangular ; its margins almost subequal, more or less convex. Antapicals 2.1-2.5 times longer than neck, rather slender; strongest curvature in proximal half; not geniculate; tuberculate; tips rounded, not spinulate. Wall without structure or finely striated. Length, 165-177 fi. Tropical, subtropical, and warm-temperate regions of Eastern Pacific. Description: — The head is 0.18 (0.16-0.2-4) the length of the neck and 1.3- 2.0 times wider than long. The anterior face of the epitheca is slightly convex or flat. The transverse furrow is flat or slightly concave. The anterior cingular list is about twice as wide as the transverse furrow and has five to eight ribs on each valve. The neck is rather long and of moderate width. The distance between the midpoints of the bases of the antapicals is 1.3 (1.3-1.4) the length of the neck which is 7-11 times longer than wide. The neck is gently curved dorsally, with a dorsal deflection of 20° (15°-24°) from the longitudinal axis of the body. The left sulcal list has a fission rib and' several other ribs. The anterior process arises near or somewhat ventrally to the middle of the anterior margin of the midbody ; in some cases it is displaced \'entrally more than its own width. It is deflected ventrally 24° (21°-29°) from the longitudinal axis of the body. It measures 0.47 (0.44-0.50) the length of the neck. The distance from the flagellar pore to the midpoint of the postmargin of the midbody is 1 .6 (1 .5-1 .8) the length of the neck. The midbody is subtriangular. Its three margins are convex, perhaps, on the average, slightly more so than in Triposolenia bicornis. The anteroventral and the anterodorsal margins are sometimes subequal, sometimes the former is some- what shorter than the latter, due to the ventral displacement of the anterior process. The posterior margin is somewhat longer than the two others. 478 THE DINOPHYSOIDAE. The antapicals are of moderate length, rather slender, almost symmetrical. The strongest curvature is, as a rule, in the proximal half of the antapical, some- times in the middle, and not localized in a knee. There is no, f)r only very slight, indication of sigmoid curvature distally. The ventral antapical is 2.3 (2.2-2.5) times, and the dorsal 2.2 (2.1-2.4) times longer than the neck. Out of six specimens three had subequal antapicals; and in the specimen with the greatest difference in- this respect the ventral was 2.5 times and the dorsal 2.2 tmies longer than the neck. In one of these specimens the distance between the center of the midbody and the tip of the ventral antapical was sub- ecjual to the distance from the midpoint of the Ijost margin of the midbody to the middle of the transverse furrow; in four cases the former, and in one case the latter distance was some- what shorter than the other. The antapicals are of about the same width, and subuniform in width throughout their whole length. The ventral is 12-20 times longer than its a^■erage width. The distance between the tips is 2.9 (2.6-3.2) times, that between the outer margins at the place of greatest spread 3.1 (2.9-3.3) times longer than the neck. The tubercles and the tips of the antapicals are as in Triposolenia bicornis. The thecal wall is as in Triposolenia depressa, or it is characterized by a fine striation (see Plate 14, fig. 10). The proportions of six specimens were measured. Divicnsions: — Length of body, 1()5-177 ju (average, 173 /z; type, \1\ y). Distance from apex to midpoint of postmargin of midbody, 105-124 /j (average, 113 m; type, 105 Ai). Length of head, 6.5-10.0 ju (average, 7.8 m; type, 7.0^). Length of neck, 40-42 ii (average, 41 tx; type, 40 n). Length of anterior process, 18-21 IX (average, 20 n\ type, 18 n). Distance between flagellar pore and mid- point of postmargin of midbody, 60-75 ix (average, 67 m; type, 63 ju). Distance between midpoints of bases of antapicals, 50-56 ju (average, 54 /j; type, 50^). Length of ventral antapical, 92-99 m (average, 95 /u; type, 99 m)- Length of dorsal antapical, 89-99 m (average, 92 y.; type, 90 n). Distance between tips of antapi- cals, 104-133 M (average, 118 yu; type, 107 m). Distance between outer margins Figure 68. — Triposolenia intermedia, sp. nov., right lateral view. X 430. Station 4709 CitHVO fathoms). SYSTEMATIC ACCOUNT. 479 of antapicals at place of greatest spread, 117-137 m (average, 128 m; type, 117 ;u). Variations: — As conceived in this paper, Triposolenia intermedia is fairly uniform in structure and proportions. The most variable characters are the shape of the antapicals and of the midbody. Some specimens are rather slender, others are more or less robust. Coitiparisons: — The species is established on outline drawings of six speci- mens. It is in size and structui'e intermediate between Triposolema bicornis and T. longicornis. In the shape and proportions of the midbody and in the relati\'e length of the antapicals it is nearer to T. bicornis; in the shape and relative size of the head and in the curvature of the neck it resembles T, longicornis. Some of our specimens of T. intermedia are more robust than any of the recorded specimens of T. bicornis and T. longicornis. See also the section on the subdivisions of this genus (p. 461). Synouij)ny: — See the section on the dimensions of Triposolema bicornis (p. 475). Occurrence: — Triposolenia intermedia is recorded at four of the 127 stations. There are 3, 0, 0, 1, 0, and 0 stations on the six lines of the Expedition. Of these four stations, two (4580, the type locality, 4583) are in the California Current; one (4013) is in the Panamic ^\rea; one (4709) is in the South Equatorial Drift. The catches in which the species was found are from 300-0 fathoms. The temperature range of these four stations at the surface was 72°-83°; the average was 77.7°. The frequency is less than 1 %. This appears to be a rare species. However, it is possible that several of our records of Triposolenia bicornis refer to T. intermedia. In the material of the Expedition the species is limited to catches from tropical waters of very high average temperature (77.7°). Triposolenia longicornis Kofoid Plate 15, fig. 1-6. Figure 69, 70 Tnposolenia longicornis Kofoid, 1907a, p. 201, pi. 17, fig. 101 ; 1906o, p. 102; 1906d, p. 1 18, 1 19, 124. Diagnosis: — Head 1.3-2.0 times wider than long. Epitheca flattened or convex. Distance between bases of antapicals 1.0-1.2 the length of neck, which is 10-15 times longer than wide and moderately curved. Midbody subtriangular; its three margins almost subequal, the postmargin convex, the anterior concave or convex. Antapicals 2.6-3.3 times longer than neck, slender, subsymmetrical ; 480 THE DINOPHYSOIDAE. geniciilation scarcely evident; tuIxTculate; tips truncate or rounded and spinu- late. Length, 210-243 m- Tropical and subtropical regions of Eastern Pacific. Description: — The head is O.IG (0.14-0.17) the length of the neck and 1.3- 2.0 times wider than long. The anterior face of the epitheca is flat, seldom con- vex. The transverse furrow is somewhat concave, seldom flat. The anterior FiGDRE 69. — Triposolenia longicornis Kofoid, riglit lateral view. X 430. 1, type .spcrimen, from Station 4711 (300-0 fathoms); 2 and 3, from Station 4.5S3 (300-0 fathom.s). cingular list is 1.5-2.0 times wider than the transverse furrow and has, on each valve, five to eight fairly well-developed ribs. The neck is long and narrow. The distance between the midpomts of the bases of the antapicals is 1.0 (1.0-1.2) the length of the neck which is 10-15 times longer than wide. The neck is gently curved dorsally, with a dorsal deflection of 18° (14°-22°) from the longitudinal axis of the body. No fission rib has been seen, but the sulcal lists have several fine ribs anteriorly. The anterior process arises at or near the middle of the anterior margin of the midbody and is deflected ventrally 15° (12°-19°) from the longitudinal axis of the SYSTEMATIC ACCOUNT. 481 body. It measures 0.68 (0.61-0.70) the length of the neck. The distance from the flagellar pore to the midpoint of the postmargin of the midbody is 1.5 (1.4-1.6) the length of the neck. The midbody is comparatively small and subtriangular. The postmargin is more or less convex and somewhat longer than the two anterior margins, which are subequal and slightly concave or convex. The antapicals are long. In most cases the ventral is the longer, being 3.2 (3.1-3.3) times longer than the neck; the corresponding values for the dorsal are 3.0 (2.6-3.1). Out of five specimens one had subequal antapicals; and in the specimen with the greatest difference in this respect the ventral was 3.1 and the dorsal 2.6 times longer than the neck. The distance between the center of the midbody and the tip of the ventral antapical exceeds the distance from the mid- point of the postmargin of the midbody to the apex to an extent equaling the half or the whole length of the midbody. The antapicals are almost of the same width, slender, and subuniform in width throughout their entire lengths or (as in the type specimen) decidedly narrower distally than proximally. The ventral is 34- 42 times longer than its average width. The antapicals are rather strongly curved, both to about the same extent. The curvature is sometimes almost evenly dis- tributed, but in most cases it is more or less localized in the proximal half of the antapical. Only exceptionally the antapicals have a slight sigmoid flexure dis- tally; in most cases their tips are incurved. The distance between the tips is 3.0 (2.7-3.3) and between the outer margins at the place of greatest spread 3.4 (3.2- 3.5) times longer than the neck. In dorsoventral view the antapicals have a very slight sigmoid curvature, the tips being slightly curved to the left. The tips are truncate or rounded, and spinulate. There are two to five tubercles, sometimes but slightly developed, all of them on the distal halves of the antapicals. The thecal wall, nucleus, and pusules are as in Triposolenia depressa. The proportions of five specimens were measured. The length of these speci- mens was 232 (215-243) fi. Dimensions: — Length of body, 210-243 n (type, 225 m; not 275 n, as stated in the original description). Distance from apex to midpoint of postmargin of midbody, 118-132 /x (average, 125 fi; type, 127 /i). Length of head, 6-8 ix (aver- age, 7.5 m; type, 7.5 m). Length of neck, 44-50 m (average, 47 m; type, 50 m). Length of anterior process, 29-35 m (average, 32 m; type, 35 m). Distance between flagellar pore and midpoint of postmargin of midbody, 69-77 m (average, 72 m; type, 73 m). Distance between midpoints of bases of antapicals, 44-57 m (average, 49 m; type, 50 m). Length of ventral antapical, 135-157 m (average, 151 m; type. 482 THE DL\OPH\'SOIDAE. 155 n). Length of dorsal antapical, 135-148 ^ (average, 140 ix; type, 129 n). Distance between tips of antapicals, 129-163^1 (average, 142 m; type, 163 yu). Distance between outer margins of antapicals at jilace of greatest spread, 151- 171 M (average, 158 (i; type, 171 /x). Variation: — ■ Judging by the specimens thus far exammed, this species is uniform in dimensions and proportions, but the shape of its midbody is somewhat variable. The anterior margins of the midbody are either almost straight or gently concave or convex. Comparisons: — The description given above is based on the typo material. FiGUKE 70. — Uocuiiuiui' of Triimaulentu lonyicurnis Kol'oid. Large, solid circles indicate records from vertical hauls; small, solid circles, stations at which this species was not found; small, open circles, stations from which no jilankton catches were examined. Triposolenia longicornis probably is very closely related to T. intermedia, which in size and structure is intermediate between this species and T. hicornis (see T. intermedia, the section on comparisons, p. 479). Another close relative of this species probably is T. fatula, which it resembles in its long and narrow pro- cesses and in having a midbody with straight or but slightly concave or convex margins. It differs from this species especially in its relatively larger head, in the more pronounced dorsal deflection of its neck, and in not having the longitudinal axis of the midbody foreshortened. See also the section on the subdivisions of the genus (p. 461). The species exhibits a very high degree of adaptation to flotation in its very SYSTEMATIC ACCOUNT. 483 long and slender processes and in the relatively small size of its midbody. Its processes are longer relatively than in any other species of this genus. Occurrence: — The species is recorded at nineteen of the 127 stations. There are 5, 6, 3, 3, 2, and 0 stations on the six lines of the Expedition. Of these nine- teen stations, two (4580, 4583) are in the California Current; two (4594, 4598) are in the Mexican Current; one (4617) is in the Panamic Area; eight (4652, 4659, 4662, 4665, 4667, 4668, 4675, 4676) are in the Peru\'ian Current; one (4713) is in the Galapagos Eddy; five (4685, 4709, 4711, 4719, 4721) are in the South Equa- torial Drift. At one station (4652) the species is recorded from 100-0 fathoms and at one (4662) from 800-0 fathoms. All the other records refer to samples from 300-0 fathoms. The temperature range of these nineteen stations at the surface was 66°-84° ; the average was 72.2°. The frequency never exceeds 1%, which is recorded at three stations (4617, 4675, 4713). The species was first recorded by Kofoid (1907a) from Stations 4583 (not 4385 as originally stated) and 4711 of the Expedition. Of these two stations the last mentioned is the type locality. The frequent occurrence of this species in the relatively cool Peruvian Cur- rent is remarkalile. In this current the species is recorded at eight (30.8%) out of the twenty-six stations; in the South Equatorial Drift at only five (10.9%) out of the forty-six stations. Moreover, all the record stations in the South Equatorial Drift are in the eastern part of this region, which is inider the influence of the Peruvian Current. Triposdlenia fatula Kofoid Figure 71 Triposolenia faiula Kofoid, 1907:1, p. 202, pi. 17, fig. 102; 190(3c, p. 9(3, 102; 1900il, p. 119. Diagnosis: — Head subspheroidal. Epitheca convex. Distance between bases of antapicals 1.4 the length of neck which is 14 times longer than wide and constricted anteriorly. Midbody subtriangular ; its longitudinal axis foreshort- ened; its margins straight or very slightly convex or conca^■e, the anteroventral one shortest. Antapicals 3.3-3.4 times longer than neck, slender; curvature as in Triposolenia ambulatrix but less asymmetrical; tuberculate; tips truncate and spinulate. Length, 175 m- Tropical and subtropical regions of Eastern Pacific. Description: — The head is 0.20 the length of the neck, subspheroidal, about 484 THE DINOPHYSOIDAE. as long as wide. The anterior face of the epitheca is convex. The transverse furrow is flat or slightly concave. The anterior cingular list is 2.0-2.5 times wider than the transverse furrow and has, on each valve, fi^■e to eight well-developed ribs. The neck is long, narrow, and suberect. The distance between the midpoints of the bases of the antapicals is 1 .4 the length of the neck which is 14 times longer Figure 71. — Triposolenia faluhi Kofoid, right lateral view of type speeinien. X 4.30. Station 4587 (300-0 fathoms). than wide. Anteriorly, near the head, the neck is constricted. It is almost straight, with a dorsal deflection of 7° from the longitudinal axis of the body. The left sulcal list has a fission rib and, anteriorly, several well-developed ribs. The anterior process arises at a distance equaling half the dorsoventral width of the process, ventrally to the middle of the anterior margin of the mid- body, and is deflected ventrally 27° from the longitudinal axis of the body. Its length almost equals that of the midbody and measures 0.73 the length of the neck. The distance from the flagellar pore to the midpoint of the postmargin of the midbody is 1..5 the length of the neck. The midboily is subtriangular, w ith foreshortened longitudinal axis and with SYSTEMATIC ACCOUNT. 485 its major axis in the dorsoventral position. The three margins are almost straight. The postmargin is very sUghtly convex, the two anterior ones are straight or very slightly concave or convex. The posterior margin is longer and the anteroventral shorter than the anterodorsal. The shortness of the anteroventral margin is due to the ventral dislocation of the anterior process. The antapicals are long; the ventral is 3.4 and the dorsal 3.3 times the length of the neck. The distance between the center of the midbody and the tip of the ventral antapical exceeds the distance from the midpoint of the postmargin of the midbody to the apex to an extent somewhat exceeding the length of the midbody. The antapicals are of about the same width, slender, subuniform in width through- out their entire length. The ventral is about 50 times longer than its average width. They are fairly strongly curved, somewhat asymmetrical; the dorsal does not converge or show a sigmoid flexure distally, as does the ventral; the curvature is not locahzed in a knee but fairly evenly distributed. The distance between the tips {i.e., the distance between the outer margins at the place of greatest spread) is 4.0 times longer than the neck. In dorsoventral view, the dorsal antapical bends, distally, to the left; so does the ventral one but only very slightly and at the very end. The tips are truncate and minutely spinulate. The thecal wall is as in Triposolenia depressa. The cytoplasm contains a number of double-contoured plasmosomes. A large sack-shaped pusule is located in the anteroventral part of the midbody ad- jacent to the nucleus. From it a slender canal runs to the flagellar pore, expand- ing into a small vesicle near the pore. The nucleus is ellipsoidal and lies m a sub- central position. The proportions of one specimen, the type, were measured. Dimensions: — Length of body, 175 ju (not 190 //, as stated in Kofoid, 1907a). Distance from apex to midpoint of postmargin of midbody, 87 fi. Length of head, 6.5 n. Length of neck, 33 n. Length of anterior process, 24 n. Distance between flagellar pore and midpoint of postmargin of midbody, 49 /x. Distance between midpoints of bases of antapicals, 47 yu. Length of ventral antapical, 112 /ii. Length of dorsal antapical, 110 fi. Distance between tips of antapicals (distance between outer margins of antapicals at place of greatest spread), 132 ix. Comparisons: — The description given above is based on outline drawings of the type specimen. Triposolenia fatula combines structural features of T. Inngiconiis and T. ambulatrix. It resembles T. longicornis in its long and narrow processes and in having a midbody with straight or but slightly concave or convex margins. It 486 THE DIXOPHYSOIDAE. differs from this species in its relatively smaller head, its more erect neck, and in the foreshortened longitudinal axis of its midbody. It resembles T. ambulatrix in the foreshortening of the longitudinal axis of its midbody and in the asymmetry of its antapicals. It differs from this species in the almost straight margins of its midbody, its relatively longer and more slender processes, and its less pronounced asynmietry. Occurrence: — The species is recorded at only three of the 127 stations. These three stations (4587, 4594, 4598) are on the first line of the Expedition and in the Mexican Current. The catches in which the species was found are from 300-0 fathoms. The temperature range of these three stations at the surface was 82°-84°; the average was 83.3°. At Station 4598 the frequency is 1%, at the remaining stations less. The species has been found only in the material of the Expedition. It was first recorded by Kofoid (1907a) from Station 4587 of the Expedition, which thus is the type locality. Judging by the available records, it appears to be one of the rarest of the species of the genus and restricted to waters of very high temperatures. Its adaptation to the low \iscosit3' of its recorded habitat is suggested by the extreme elongation and marked slenderness of its processes, its relatively broad cingular lists, and its large pusule. Triposolenia ambulatrix Kofoid Plate 14, fig. 11-14. Figure 60: 3, 72 Tnpotiulenia ambulalrix Kofoid, 1907a, p. 203, pi. 4, fig. 24; 1906c, p. 9G, 102; 1906d, p. US, 119, 124. non Triposolenia ambulatrix Jorgensen, 1923, p. 42, 43, fig. 63. Diagnosis: — Head subspheroidal or subcylindrical. Epitheca con\ex. Dis- tance between bases of antapicals 1.5-1.7 the length of neck which is 12-17 times longer than wide and almost straight or gently curved. Midbody subtriangular or subellipsoidal, its longitudinal axis foreshortened; with convex margins. Ant- apicals, 2.2-2.6 times longer than neck, slender, decidedly asymmetrical; deflected dorsally, more so than in other known species; not geniculate; with or without tubercles; tips rounded or truncate and spinulate. Length, 120-159 m. Tropical and subtropical regions of Eastern Pacific. Description: — The head is 0.19 the length of the neck. Its structure is about the same as in Triposolenia depressa. The neck is rather long and narrow. The distance between the midpoints of the bases of the antapicals is 1.5 (1.5-1.7) the SYSTE.MATIC ACCOUNT. 487 length (if the neck which is 12-17 times longer than \\ide. The neck is almost straight or gently curved dorsally, with a dorsal deflection of 5° (0°-8°) from the longitudinal axis of the body. The ribs of the sulcal lists are very weak, scarcely noticeable ; the fission rib was found in one specimen. The anterior process arises vent-rally to the middle of the anterior margin of the midbody; its distance from the middle equals or is somewhat greater or smaller than half its dorsoventral width. It is deflected ventrally 22° (17°-30°) from the longitudinal axis of the body. It measures 0.57 (0..56-0.58) the length of the neck. The distance from the flagellar pore to the midpoint of the postmargin of the midbody is 1.6 (1.4-1.7) the length of the neck. The midbody is subtriangular or subellipsoidal with foreshortened longitudinal axis and with its major axis in the dorsoventral position. Its three margms are more or less convex. The anteroventral margin is always some- r-,^,T„^ ?•> r,.;„. what shorter than the two others, due to the \'entral dis- solerUa ambulatrix Kofoid, right lateral view. X 430. location of the anterior process; and the postmargm is station -4.598 (.300-o fath- always the longest. °™®'- The antapicals are pronouncedly asymmetrical and of moderate lengths. In one case the ventral was 2.6, the dorsal 2.3 times longer than the neck; in two other cases the correspondmg figures were 2.4 and 2.5, and 2.3 and 2.2. In two out of three specimens the distance between the center of the midbody and the tip of the ventral antapical was subequal to the distance from the middle of the post- margin of the midbody to the apex; in the third case the latter distance was some- what longer. The antapicals are of about the same width and taper gradually from the bases to the tips. The ventral is, in the middle, 19-22 times longer than wide. The dorsoventral diameter of the tips is 0.25-0.50 that of the bases. The ventral antapical is abruptly deflected posterodorsally, and has a more or less pronounced sigmoid curvature, its tip being deflected ventrally. The dorsal is not bent so as to form a balanced horn, in reversed symmetry, as m most of the other species of this genus, but is thrown posterodorsally with a moderate dorsal con- vexity; its tip is sometimes deflected dorsally, which gives it a sigmoid shape. The distance between the tips (i. c, the distance between the outer margins at the place of greatest spread) is 3.2 (2.7-3.5) times longer than the neck. In dorso- ventral view the antapicals are somewhat curved, the tip of the dorsal being de- flected to the left, that of the ventral to the right . The tip of the ventral is rounded 488 THE DINOPHYSOIDAE. or more or less acute, that of the dorsal truncate and furnished with two minute spinules. The tubercles are very slightly developed and few in number ; sometimes they are absent . The thecal wall and the nucleus are as in Triposolenia depressa. The cyto- plasm of the type specimen was crowded with numerous highly refractive double- contoured bodies. The proportions of three specimens were measured. The length of these specimens was 146 (120-159) m- Dimensions: — Length of body, 120-159 /i (average, 146 ju; type, 158 m)- Distance from apex to midpoint of postmargin of midbody, 77-96 n (average, 88 n; type, 96 n). Length of head, 5.0-6.5 m (average, 6.0 ^i; type, 6.5 n). Length of neck, 26-35 m (average, 32 ix; type, 34 n). Length of anterior process, 15-20 /x (average, 18 m; type, 19 ix). Distance between flagellar pore and midpoint of post- margin of midbody, 45-55 ^ (average, 50 m; type, 55 n). Distance between mid- points of bases of antapicals, 39-58 m (average, 49 m; type, 58 y). Length of ven- tral antapical, 59-90 /i (average, 77 m; type, 90 /i). Length of dorsal antapical, 57-88 M (average, 74 m; type, 77 m)- Distance between tips of antapicals (distance between outer margins of the latter at place of greatest spread), 70-120 ii (average 101m; type, 120 m). Variations: — The species is rather constant in the few specimens thus far examined. The midbody varies in the degree of convexity of the anterior margins and in the degree of foreshortening. The decided convexity of the anterior mar- gins of the midbody in the type specimen might be correlated with the fact that the midbody of this specimen was croA\ded with a great number of highly refrac- tive double-contoured bodies. The degree of dorsal deflection of the antapicals is also variable. Comparisons: — The description given above is based on the type material. This is the most asymmetrical of the species of this genus. It has the most marked contrast between the dorsal and the ventral antapicals. The dorsal ant- apical is deflected dorsally to such a degree that its balance with the ventral one, so apparent in other species of this genus, is destroyed. With regard to the relationships between this species and the other species of this genus, see Triposolenia fatula and the section on the subdivisions of this genus (p. 461). Synonymy: — The species was established l^y Kofoid (1907a) under the name of Triposolenia ambulatrix. The form figured by Jorgensen (1923, fig. 63) as T. amhulatrix seems to be T. depressa Kofoid. SYSTEMATIC ACCOUNT. 489 Occurrence: — The species is recorded at four only of the 127 stations. There are 2, 0, 0, 2, 0, and 0 stations on the six lines of the Expedition. Of the.se four stations, two (4587, 4598) are in the Mexican Current ; one (4711) is in the South Equatorial Drift; one (4713) is in the Galapagos Eddy. All the samples in which the species was found are from 300-0 fathoms. The temperature of these four stations at the surface was 73°-84° ; the aver- age was 78.5°. The frequency is less than 1 % except at Station 4598, where it is 1 %. The species was first recorded by Kofoid (1907a) from Station 4711 of the Expedition, which thus is the type locality. It is eupelagic and appears to be one of the rarer species of the genus. In the material of the Expedition it is limited to catches from tropical waters. Subgenus 2. Ramiciformia Kofoid. — Of the two species that belong to this subgenus only one, Triposolenia ramiciformis, has been found in the material of the Expedition. Triposolenia ramiciformis Kofoid Figure 60: 2, 73, 74 Triposolenia rnmidformis Kofoid, 1906c, p. 94, 99, 101, 102, 106, 107, 108, 109, 110, pi. 17, fig. 7; 1906d, p. 118, 119, 120. Diagnosis: — Head small, capitate. Epitheca convex. Distance between bases of antapicals 0.77-0.97 the length of neck which is 11-17 times longer than wide and almost straight. Midbody pendent, subellipsoidal. .\ntapicals 2.6-3.4 times longer than neck, slender, subsymmetrical or more or less asymmetrical, geniculate, tuberculate; tips rounded. Length, 144-169 n. Tropical, subtropical, and warm-temperate regions of Eastern Pacific. Description: — The head is 0.17 (0.15-0.19) the length of the neck, capitate, in most cases about as long as wide. The anterior face of the epitheca is more or less con^'ex. The transverse furrow is flat or slightly concave or convex. The anterior cmgular list is 1.5-2.5 times wider than the transverse furrow and has many fine ribs. The neck is rather long and narrow. The distance between the midpoints of the bases of the antapicals is 0.89 (0.77-0.97) the length of the neck which is 11-17 times longer than wide. The neck is almost straight, with a dorsal deflection of 15° (5°-21°) from the longitudmal axis of the body. The left sulcal list has a fission rib. The anterior process arises from the midbody at the base of the ventral 490 THE DINOPHYSOIDAE. antapical. It is deflected ventrally 29° (22°-38°) from the longitudinal axis of the body. It measures 0.47 (0.41-0.54) the length of the neck. The distance from the flagellar pore to the hindmost point of the postmarghi of the midbody is 1.7 (1.6- 1.9) the length of the neck. The midbody is pendent, sack-shaped, and generally broadly rounded poste- riorly; usually it tapers but little near its posterior end, but several specimens have been seen in which it tapers gradually from the bases of the antapicals to a Figure 73. — Triimsolenia rainiciformis Kofoid, left Lateral view. X 430. 1, from Station 4681 (300-0 fathoms) ; 2, 3, from Station 4613 (300-0 fathoms). rather narrowly rounded posterior end. The ventral antapical and the anterior process arise together from the anterovent ral angle of the midbody and the dorsal antapical from the anterodorsal angle. The anterior edge of the midbody, be- tween the anterior process and the dorsal antapical, is almost straight or slightly convex or concave. The antapicals are rather long. In most cases the ventral is the longer, being 3.2 (2.7-3.4) times longer than the neck; the corresponding figures for the dorsal are 2.9 (2.6-3.1). Out of twelve specimens one had the ventral shorter than the SYSTEMATIC ACCOUNT. 491 dorsal, the former being 2.7, the latter 2.9 times longer than the neck. In the specimen with the greatest difference in this respect the ventral was 3.3, the dorsal 2.8 times longer than the neck. The antapicals are of almost the same width, slender, and subuniform in width throughout their entire lengths. The ventral is 30-40 times longer than its average width. They are strongly curved. Sometimes they are curved to about the same extent, sometimes either one of them is more cur\-ed than the other. The cur^'ature is sometimes almost evenly distributed, but in most cases there is a distinct knee near or somewhat proximally to the middle of the antapical. Distally the antapicals exhibit in most cases a more or less developed sigmoid flexure. The distance between the tips is 2.3 (1.8-2.8) and between the outer margins at the place of greatest spread 3.1 (2.9-3.2) times longer than the neck. In the region of the major flexure and behind it, each ant- apical has, on its outer surface, two to six tubercles. The tips are rounded or more or less bluntly pointed. The thecal wall is as in Triposolenia depressa. The nucleus is lenticular in shape, bilaterally compressed, and lies in the posterior part of the midbody. The chromatophores are few, irregular in form, and yellowish green. The proportions of twelve specimens were measured. The length of these specimens was 154 (144-169) yu. Dimensions: — Length of body, 144-1(39 /i (average, 154 yu; type, 152 yu). Distance from apex to hindmost point of postmargin of midbody, 100-111 n (average, 105 ^; type, 107 n). Length of head, 6-7 n (average, 6.5 yu; type, 6.0 n). Length of neck, 35-42 yu (average, 38 m; type, 40 yu). Length of anterior process, 15-22 yu (average, 18 yu; type, 17 m). Distance from flagellar pore to hindmost point of postmargin of midbody, 61-70 n (average, 65 yu; type, 65 //). Distance between midpoints of bases of antapicals, 30-36 /x (average, 34 yu; type, 36 yu). Length of ventral antapical, 106-133 n (average, 121 n; type, 119 n). Length of dorsal antapical, 103-126 m (average, 110 /x; type, 108 ,u). Distance between tips of antapicals, 64-114 yu (average, 89 yu; type, 96 yu). Distance between outer margins of antapicals at place of greatest spread, 110-130 yu (average, 119 ,u; type, 117 m). Variations: — The species is strikingly \'ariable with regard to the shape of the midbody. In some specimens the midbody tapers uniformly to a blunt pos- terior tip, while in others its dorsal and A'entral margins are subparallel, and its posterior end is broadly rounded. The antapicals vary in degree of sigmoid flex- ure, in the distance between the tips, and in the amount of tuberculation. There 492 THE DINOPHYSOIDAE. are some differences in the amount of lateral displacement of the origin of the anterior process and in the obliquity of the neck. Comparisons: — Most of the specimens assigned to Triposoknin ramicijormis in this paper agree very closely with the type specimen of this species as drawn by Kofoid (1906c, pi. 17, fig. 7). The species differs from Triposolenia exilis, the only other species of this sub- genus, in the greater obliquity of the neck, in the greater width of the anterior part of the midbody, and in the less immediate posterior deflection of the ant- FiGUEE 74. — Occurrence of Triposolenia vatniciffirniia Kofoid. Large, solid circles indicate records from vertical hauls; small, solid circles, stations at which this species was not found; small, open circles, stations from which no iilankton catches were examined. apicals. It is probably better adapted to flotation than T. exilis. The tubercula- tions on the antapicals are farther in front in the subgenus Ramiciformia than in the subgenus Triposolenia. Occurrence: — The species is recorded at forty of the 127 stations. There are 7, 14, 6, 7, 4, and 2 stations on the six lines of the Expedition. Of these forty stations, two (4580, 4583) are in the California Current; four (4587, 4590, 4598, 4605) are in the Mexican Current; three (4613, 4634, 4637) are in the Panamic .\rea; fourteen (4648, 4650, 4652, 4659, 4662, 4663, 4665, 4666, 4667, 4668, 4670, 4671, 4675, 4676) arfe in the Peruvian Current; one (4691) is in the Easter Island Eddy; two (4713, 4715) are in the Galapagos Eddy; thirteen (4679, 4681, 4683, SYSTEMATIC ACCOUNT. 493 4701, 4705, 4707, 4709, 4711, 4717, 4721, 4722, 4724, 4740) are in the South Equa- torial Drift; one (4742) is in the South Equatorial Current. At one station (4652) the species is recorded from 100-0 fathoms, at three stations (4C62, 4670, 4715) from 800-0 fathoms, and at one station (4681) from both 800-0 fathoms and 300-0 fathoms. All the other records refer to catches from 300-0 fathoms only. The temperature range of these forty stations at the surface was 66°-85°; the average was 73.4°. At two stations (4648, 4663) a frequency of 2% is recorded. At t-en stations (4598, 4650, 4659, 4552, 4666, 4668, 4671, 4701, 4715, 4742) a frequency of 1% was found. At the remaining stations the frequency is less than 1%. Previously the species has been taken by Kofoid (1906c) off San Diego, Cali- fornia, the type locality, where it occurred in catches from 100 fathoms. This is one of the two most widely distributed and abundant species of the genus in the region investigated by the Expedition, standing second to Triposo- lenia bicornis with forty-five record stations. It appears to be more concentrated than any other species of this genus in the Peruvian Current and in the eastern part of the South Equatorial Drift, which is under the influence of this current. 494 THE DINOPHYSOIDAE. 3. ORNITHOCERCIDAE, fam. nov. Diagnosis: — Body of diverse shapes, usually about as long as deep or more or less decidedly deeper than long. Epitheca low, more or less disk-like; its depth 0.07-0.88 the depth of hypotheca. Transverse furrow, as a rule, wide and de- cidedly wider dorsally than ventrally; its dorsal width 0.33-0.77 the greatest depth of body. C'ingidar lists large, inclined anteriorly, funnel-shaped; dorsal width, 0.30-2.00 the depth of body. Usually with phaeosomes in girdle. Length of body, 13.3-88.7 p. Marine and eupelagic, widely distributed in tropical, subtropical, and warm- temperate waters. Subdivisions. Relationships among the Genera The species of the family Ornithocercidae ha\'e in tlio present paper been assigned to one of three genera, viz., Ornithocercus, Parahistioneis, and His- tioneis. Of these genera, Ornithocercus and Histioneis were established by Stein (1883), while Parahistioneis is new. Following the suggestion of Murray and Whitting (1899), a fairly great numl)er of investigators rejected Stem's (1883) genus Ornithocercus and referred all the known species, among which were representatives of all the three genera mentioned above, to one single genus, Histioneis (see the historical section of Ornithocercus, p. 509). We are not able to accept this decision, which was caused by the discovery of Histioneis ( = Para- histioneis) francescae and H. { = Parahistioneis) para, since Ornithocercus and Histioneis are structurally very distinct and evidently represent two e\'olutionary lines embodying very different structural developments. For instance, in Ornitho- cercas (/) the anterior cingular list does not become stalked; (2) the posterior cin- gular list has no submarginal cross-rib but a great number of radial ribs ; and (3) the left sulcal list has usually more than one main rib behind the fissifin rib. In Histioneis, on the other hand, (1) the anterior cingular list is stalked in all but one of the twenty-eight known species; (2) the posterior cingular list has a submar- ginal cross-rib but only two pairs of radial ribs, viz., those next to the sagittal suture; and (3) the left sulcal list has n(>\cr more than one main ril) behind the fission rib. Parahistioneis occupies in several respects a position intermediate be- tween Ornithocercus and Histioneis. For example, it agrees with Ornithocercus in not having any submarginal cross-rib in the posterior cingular list, and all Init one SYSTEMATIC ACCOUNT. 495 of its species resemble Histioneis in having a single main rib in the left sulcal list tjehind the fission rib. In regard to the stalking of the anterior cingular list and to the radial ribs of the posterior cingular list, some of the members of Parahis- tioneis exhibit tendencies inherent in Histioneis, while others approach Ornitho- cercus. A further generic subdivision of this family will probably be desirable and feasible after a thorough reexamination of the right sulcal list which is now unknown in most of the species. Parahistioneis includes quite heterogeneous ele- ments, and at least one member of Ornithocercus, viz., O.fortnosus, is likely to have its generic allocation changed. Lindemann (1924, p. 7) suggested that Ornithocercus, "die Krone aller Peridineen," evolved from Histioneis. This, how^ever, can hardly be the case for the following reasons: — (1) the primitive species of Ornithocercus (0. splendidus) have a deeper epitheca than Histioneis and the width of their transverse furrow is subuniform throughout; (2) even the most primitive members of Histioneis, viz., H. codata and H. paulseni, have the posterior cingular list of a much more ad- vanced type than the one characteristic of Ornithocercus. In other words, in these three fundamental respects Ornithocercus is more primitive than Histioneis. Neither can we derive Histioneis from Ornithocercus. The former genus is in some respects, especially in the structure of the left sulcal list, more primitive than the latter. In considering the relationship between these two genera, the structure of the right sulcal list should also be taken into account. It seems to be most in conformity with known facts to assume that the three genera of this family evolved independently from a common ancestor approaching the Dinophysis type and embodying the following characteristics : — (1) the body was subrotund in lateral outline ; (2) the transverse furrow was located anteriorly and of subuniform width throughout ; (S) the anterior cingular list was wide and funnel-shaped but not stalked and furnished with a great number of radial ribs; (4) the posterior cingular list was fairly wide, inclined anteriorly, open ventrally, and had a rather great number of radial ribs but no submarginal rib ; (5) the right sulcal list was fairly small and its ventral edge free ; {6) the left sulcal list ended on the posteroventral side of the body, was angular posteroventrally, and had but one main rib behind the fission rib; (7) the thecal wall was areolate or finely reticulate. From this ancestral type the present genera evolved partly by progressive, partly by retrogressi\e e\'olutionary processes. Distribution: — The three genera of Ornithocercidae are exclusively marine, eupelagic, and widely distributed in tropical, subtropical, and warm-temperate waters. Most of the species have been found only in the warmest regions of the 496 THE DINOPHYSOIDAE. seas. Wliile some of the species of Ornithocercus occasionally may be found in large numbers, most members appear to be exceedingly rare. The optimum habi- tat of most species is the deeper levels of photosynthesis, but some species of Ornithocercus are rather frequent in surface waters. Key to the Genera 1. Posterior cingular list with siibmarginal cross-rib Histioneis Stein (1883). 1. Posterior cingular list without submargiual rib 2. 2. Posterior cingular list with less than six radial ribs Parahistionds, gen. nov. 2. Posterior cingular list with six or more radial rib.s Ornithocercus Stein (1883). Ornithocercus Stein Ornithocercus Stein, 1883, p. 2,5. Buti^chli, 188.5, p. 944, 1011. SohOtt, 1890, p. 28. Del.\ge & H^ROU.^RD, 1896, p. 380. Parelion Schmidt, 1888, pi. 144, fig. 59-01. Ornithocerus Schutt, 1893, fig. S3 (lapsus pennae). Ornithoceras Graf, 1909, p. 139 (lapsus pennae). Diagnosis: — Body subcircular, subellipsoidal, subobovate, rounded sub- triangular, or subtrapeziform in lateral outline, usually about as deep as long (length: depth, 0.76-1.24: 1), compressed bilaterally. Epitheca low, disk-like, its depth usually 0.37-0.67, seldom as little as 0.28 or as much as 0.88 the depth of body. Transverse furrow, as a rule, wide dorsally and wider dorsally than ven- trally; dorsal width, 0.33-0.55 the depth of body; ventral width, 0.31-0.70 the dorsal width (exception Ornithocercus splendidus) . Cingular lists large, anterior not stalked, posterior usually slightly narrower than anterior; width, 0.37-1.46 the depth of body; anterior with 4-24 radial ribs, inclined anteriorly at 25°-50°; posterior with 6-24 usually simple ribs, no cross-rib, inclined anteriorly at 30°- 85°; besides ribs sometimes reticulation. Right sulcal list small, its ventral margin free (except in 0. formosus). Left sulcal list large, extending at least to near antapex, sometimes to posterior cingular list; two- to five-lobed, rounded, or squarish; width at lobes 0.27-1.92 the depth of body; behind fission rib 4-15 strong radial ribs (except in 0. formosus, that has only one) and often reticulation. Type. — Ornithocercus magnificus Stein. Organology: — The body {thcca) in Ornithocercus is simple and more uniform in shape than in any of the other large genera of this tribe. Usually it is some- what asymmetrical in lateral view, due largely to the shape of the transverse fur- row and to the inclination of the longitudinal axis, but symmetrical or subsym- metrical specimens may be found in a few species, for instance, in 0. splendidus (Figure 77: 4; 85: 3) and in 0. quadratus (Figure 87: 1-4). In dorsoventral view it appears always to be symmetrical. Sometimes the longitudinal axis is perpendicu- SYSTEIVIATIC ACCOUNT. 497 lar to the girdle (see the figures just mentioned), but usually it is slightly (1°-10°) deflected posterodorsallj' ; the maximum deflection as yet recorded is about 20° (0. heteroporus). \Vlien seen in lateral view, the body usually is about as deep as long, the typical ratio between the length and the depth being 0.90-1.10: 1. In exception- ally deep specimens {cf. Ornithocercus splendidus) this ratio may be 0.76: 1 ; and in the specimen with the most elongated body found in the material of the Expedi- tion (belonging to 0. splendidus) it was as high as 1.24: 1. .\s a rule, the body is deepest in or near the middle, antl onlj- in exceptional cases (some specunens of 0. splendidus) the greatest depth is located at the girdle. In most species it is subcircular in lateral outline, but it may be subellipsoidal, subobovate, rounded subtriangular, or subtrapeziform (0. heteroporus, 0. splendidus). In dorsoventral view the body usuallj' is subellipsoidal, widest in the middle, 1.4-1.7 times longer than wide, well romided posteriorly and narrowly rounded conical anteriorly (Plate 17, fig. 4, 6) ; but it may be broadly ellipsoidal, with broadly rounded apices, and about 1.14 times longer than wide (0. heteroporus, Plate 18, fig. 1), or again its outline may be obovate (0. splendidus, Figure 77: 7). The epitheca forms an ellipsoidal, gently vaulted to flat disk, the dorso- ventral diameter of which decidedly exceeds the transdiameter (Schiitt, 1895, pi. 5, fig. 22). In lateral view it usually is somewhere between 0.37 and 0.67 as deep as the hypotheca, gently convex to flat, highest in or dorsally to the center (only seldom ventrally to the center), and inclined ventroposteriorly at 5°-15°. Some- times {Ornithocercus heteroporus) its depth maj' be onlj' 0.28, sometimes (0. splendidus) as much as 0.88 the greatest depth of the hypotheca; and it may be horizontal (0. heteroporus and 0. splendidus). Due to the small height of the epitheca, the transverse furroiv is situated near the anterior end of the body. Its distal portion is not displaced posteriorly, in other words, it does not form a spiral about the body; and its postmargin is always straight or nearly so. In the relatively primitive species, Ornithocercus splendidus, it is of subvmiform or nearly subuniform width throughout, and its width is 0.11- 0.28 the greatest depth of the body. In the remaining species it is more or less strikingly wider dorsally than ventrally; the dorsal width is 0.33-0.55 the greatest depth of the body, and the \entral width is 0.31-0.70 the dorsal width. Usually it is gently or moderately conca\e dorsallj', but it maj' be flat or even slightly convex. Ventrally it is, as a rule, gently convex to flat. The longitudinal furrow is about half as long as the hypotheca or slightly more and at most but slightly impressed. 498 THE DINOPHYSOIDAE. While in Phalacroma, Dinophysis, and most of the other genera of this tribe, the lists of the furrow are comparatively narrow and simple, they are characterized in Ornithocercus by their very large size and high structui-al differentiation. They arise from low and narrow basal ridges (Plate 17), and in spite of their large size they are hyaline and very delicate. Of the cingular lists the anterior usually is slightly wider than the posterior; and the posterior is, as a rule, somewhat more inclined anteriorly than the an- terior. In most specimens the width of these lists is somewhere between 0.37 and 0.90 the greatest depth of the body, and the anterior inclination somewhere be- tw^een 25° and 50° in the case of the anterior list and between 30° and 85° in the case of the posterior. Ornithocercus splendidus, which is unique in the excessive width of these lists (0.77-1.46 the greatest depth of the body), is also characterized by the fact that the two lists have about the same anterior inclination (30°-50°). The dorsal and ventral inclination are sometimes about equal; sometimes the ventral is slightly steeper. The anterior cingular list has on each valve between four and twenty-four, often between five and ten, complete, subequidistant radial ribs. These ribs may be simple or more or less anastomosing, and sometimes (Plate 16, fig. 1) they have distally candelabra-like branches. In 0. formosus (Plate 17, fig. 4) the dorsal of these ribs is branched in the same manner as in many of the species of Histioneis. Besides these complete ribs, there is frequently a varying number of short, marginal radial ribs, which may be simple or anasto- mosing. The tips of the ribs may extend slightly beyond the margin of the list (Plate 17, fig. 1). The posterior cingular list is on each valve furnished with be- tween six and twenty-four, as a rule between eight and nineteen, complete, sub- equidistant radial ribs. These ribs usually are smiple or one or a few of them are branched proximally. In exceptional cases (Ornithocercus splendidus, Plate 17, fig. 3) they anastomose, or (Plate 17, fig. 1, 5) the dorsal and ventral of them are connected by reticulation. In a few specimens (Plate 16, fig. 3) this list has besides the complete ribs some short, marginal radial ribs. The anterior cingular list is always closed ventrally. On the other hand, we are not able to decide as to whether or not the posterior is always open ventrally. The sulcal lists form a direct continuation of the posterior cingular list and run parallel to each other on either side of the longitudinal furrow. The entire right sulcal list and the portion of the left sulcal list which is behind the fission rib belong to the right valve; and the portion of the left sulcal hst in front of the fission rib belongs to the left valve. The right sulcal list, which always is strikingly smaller than the left, exhibits SYSTEMATIC ACCOUNT. 499 fairly decided variations in size and shape. In most species its ventral margin is free, but in Ornithocercus formosus this margin appears to be attached to the left sulcal list, just as in Histioneis, so that these two lists thus form a canal open posteriorly and presumably also anteriorly (Plate 17, fig. 5). In some species {e.g., Ornithocercus magnificus) this list is very small, ending at or near the fission rib of the left sulcal list, and subtriangular or rounded triangular, decreasing gradually in width posteriorly. In other species (Figure 85:5; Plate 16, fig. 1; Plate 17, fig. 1, 7) it is comparatively large and extends far beyond the fission rib of the left sulcal list (but never quite as far as to the midline of the body). In these species it either is of subuniform width throughout the greater portion of its length and gently (Plate 17, fig. 1) or strongly (Plate 16, fig. 1) convex posteriorly, or (Plate 17, fig. 7) it decreases strikingly in width in its posterior half and has an irregularly sigmoid ventral margin. Its maximum width does not, as a rule, ex- ceed the dorsal width of the transverse furrow. In some species this list appears to lack structural differentiation ; in others it has one or more regular or irregular ribs (Plate 16, fig. 4, 1), or it may be reticulated (Plate 17, fig. 1, 5). Its size, shape, and structure are frequently quite variable within the species. The left sulcal list, which is very large, has imdergone in this genus more striking modifications than any other structure; and the species hitherto estab- lished are based largely on the differences in this list. As a rule, it is decidedly variable even within the species. In the most primitive members of this genus {Ornithocercus heteroporus and 0. splendidus, Figiu'e 75, 77) this list is two-lobed and ends at or near the antapex of the body; the lobes are subequal or nearly so, and narrowly rounded or subacute. In the remaining species this list extends to the dorsal side of the body and sometunes (Plate 16, fig. 1; Plate 17, fig. 7) it en- circles the hypotheca, ending at the posterior cingular list. Among these species one, 0. formosus (Plate 17, fig. 5), is characterized by having this list strongly two-lobed, the lobes being narrowly rounded or subacute; the others have it three-lobed (0. magnijicus, 0. thurnt, and 0. carolinae, Figure 79, 81, 89), four- or even five-lobed (0. steini, Figure 83), rounded (0. orbiculatus, Plate 17, fig. 7), or squarish (0. quadratus, Figure 86). Wlien it is three-lobed, the lobes usually are narrowly rounded and either subequal, as in 0. magnificus and 0. thurni, or the middle (antapical) one of them is strikingly larger than the others, as in 0. carolinae. The lobes either are well marked, as in Figure 81: 6 of 0. thurni, or nearly indistinguishable, as in Figure 83: 10, 11, of 0. steini. The width of this list, measured at the tips of the lobes, is usually somewhere between 0.50 and 1.92 the greatest depth of the body; in 0. carolinae the width at the dorsal lobe may 500 THE DINOPm'SOIDAE. amovint to only 0.27 the greatest depth of the body. The fission rib, which is made up of two members, one belonging to the left, the other to the right valve, and which arises at or near the middle of the body, usually is straight or nearly so, and is subhorizontal or somewhat inclined posteriorly. As a rule, its inclination lies between 1° and 25°; its maximum inclination is about 55° (see 0. carolinae and 0. formos2is, Plate 17, fig. 1, 5). In exceptional cases (0. quadratus f. inter- media, Figiu'e 87: 15-20) its posterior member is recurved posteriorly. In front of the fission rib this list usually has, besides the anterior main rib, i.e., the rib at the junction of this list and the posterior cingular list, one or a few regular or irregular radial ribs and frequently a more or less developed reticulation (Plate 17, fig. 1, 3, 5, 7, 9). Behind the fission rib it sometimes {0. Jormosus, Plate 17, fig. 5) has only one strong radial rib, but usually {e.g., 0. magnificus, 0. thurni, 0. steini, Figure 79, 81, 83) there are from four to seven ribs of this kind; and in one species (0. carolinae, Plate 17, fig. 1) the number of these ribs is 13-15. These ribs, which afford ver,y important taxonomic characters, and wliicli in the present paper have been named the a-, b-, c-, etc., ribs, the a-rib being the one situated most dorsally, either are simple and nearly straight {e.g., in 0. quadratus f. schiitti. Figure 86:7-14), or they are more or less branched and irregular {e.g., in 0. quadratus f. assimilis. Figure 87: 1-8). The tips of these ribs are in several spe- cies connected by a submarginal rib, which may be fairly heavy {e.g., in 0. mag- nificus, Figure 79) or very fine {0. steini. Figure 83), and which may be present or absent in the same species. Besides the radial and submarginal ribs a more or less developed reticulation frequently characterizes this portion of the list; reticula- tion is especially found in the tips of the lobes and dorsally to the dorsal radial rib (Plate 17, fig. 1, of 0. carolinae). The great variations observed by us in the struc- tm-al differentiations of this list seem to be due largely' to age, to irregularities in the regeneration of local injuries, and to processes of resorption and growth in adjustment to the requu-ements of flotation. Generally speaking, the older a specimen is, the more complicated is the structure of this list (see Schiilt, 1900a). Parasagittal lists have not been recorded in this genus. The flagellar pore is a rounded or somewhat elongated opening of moderate size, located in the transverse furrow, in other words, on the right valve, at or just behind the posterior cingular list. Its largest diameter usuall.y does not exceed half the ventral width of the transverse furrow. No counterparts have been found to the pores recorded in Phalacroma (Plate 3, fig. 1, 2, of Phalacroma pulchrum and P. giganteum) oil the ventral side of the left valve, near the sagittal suture, SYSTEMATIC ACCOUNT. 501 and just in front of the anterior cingxilar list. It is, howe^•er, not impossible that such pores do exist in Ornithocercus. The longitudinal flagellum has been figiu-ed only in Ornithocercus magnificns (by Schiitt, 1895, pi. 5, fig. 21 : 1). In this species it originates in the flagellar pore and has a length about 1.5 the greatest depth of the body. The transverse flagellum has not been observed as yet ; presumably it is well developed, lies in the gii'dle, and encircles the body from the left around to the right. The structure of the thecal wall is subject to but slight variations in this genus. In all the species the entire theca has a fairly great number of scattered pores. The most primitive species {Ornithocercus heteroporus and 0. splendidus) may have no other structural differentiation, but in all the higher members the hypotheca is areolate. The areoles are rounded or angular, of moderate and often subuniform size, and each of them has a pore in the middle ; sometimes (Plate 17, fig. 5, 7) they are close-set, sometmies (Plate 17, fig. 1) more or less spaced. In some species (e.g., 0. magnificus, Plate 16, fig. 3) the (10-20) areoles that border the girdle posteriorly are larger than the others. Sometimes (Plate 16, fig. 1) the areoles are developed only on the central portion of the hypotheca. The protoplasmic contents have been figured in one species only, viz., in Ornilhocercus magnificus (Schiitt, 1895, pi. 5, fig. 21:2). In this species the nucleus is large, ovoidal, is located posterodorsally, and has a moniliform chroma- tin reticulum. Two fairly large pusules open at the flagellar pore by a slender canal. Chromatophores are not present (Schiitt, 1895, p. 64). Metaplasmic in- clusions of various kinds, such as globules of fats and oils, have been found. See also Schiitt (1895) and Biitschli (1885). Phaeosomes, ovoidal or spheroidal, small or of medium size, are often found in the girdle. Their color is green or yellowish green. The length of the body ranges within the genus from about 27 fx (Ornithocercus heteroporus) to about 68 n {0. thurni), according to our own measurements made on the material of the Expedition. Measurements made on pre\iously published figures give a maximum length of more than 80 ix. Reproduction: — Just as in Phalacroma and Dinophysis and in the other genera of this tribe, binary fission is the only known mode of reproduction in Ornithocercus. The details of this procedure are not known, but in all probability they are about the same as in Phalacroma (see the corresponding section of this genus, p. 51). At least in some species (see, for instance, Plate 18, fig. 4-0 of Ornithocercus thurni) the two schizonts remain attached to each other for some time by means 502 THE DINOPmsOIDAE. of an accessory ridge along the anterior or middle portions of the dorsal margin of the hypotheca. No data are available as to the length of this period of attach- ment, but presumably it is comparatively short. Shice very few free specimens with traces of this ridge have been found, it appears plausible that this structure is resorbed shortly after the separation of the schizonts. Usually this ridge is fairly wide ; its width may even be subequal to half the greatest depth of the body. Its shape is squarish, and its dorsal margin is deeply serrated in most specimens. It may be nearly without structural differentiation, but usually it is reticulated; sometimes (Plate 18, fig. 2) the reticulation is fine, sometimes (Plate 18, fig. 6) the meshes are large and regular in arrangement. WTiile in Dinophysis iniles as many as eight adhering specimens have been found (Weber-van Bosse, 1901, pi. 17, fig. 3), in Ornithocercus never more than two specimens have been recorded attached to each other. The adaptive significance of this phenomenon possibly is that it facilitates the remaining of the schizonts in the horizontal position during the period when their lists are relatively undeveloped. In the horizontal position these organisms offer the maximum surface of resistance. As a ride, the schizonts are of about the same size, shape, and structure as their parent. Sometimes, however, there is a more or less pronounced difference in the size or in the shape as well as in the structure. This phenomenon, which is illustrated by Plate 18, figure 2, 7-9, appears to be due to an increase in the rate of division beyond the rate of growth of the cytoplasm. Indeed, sometimes (Plate 18, fig. 9) the fission takes place so rapidly that parts of the theca of the grandparent still remain unabsorbed. A striking feature of this mode of fission is that the ventral side of the body remains unchanged and that the reduction in size takes place largely dorsally, or dorsallyand posteriorly (Plate 18, fig. 9, 2). In some speci- mens the length as well as the depth of the body are reduced, the schizonts being more or less strikingly smaller but of about the same shape as the mother. In other cases only the depth is reduced, and the schizonts become more or less oblong. Plate 18, figiu-e 9, shows a specimen about as long as its grandparent but only about half as deep. In this manner specimens may originate which are so strikingly different from their ancestral type that a specific allocation is impossi- ble or at least conjectural. The significance and causes of the phenomenon described in the last para- graph are unknown. It is probably not an abnormality brought about by unfav- orable factors in the surrounding medium, since most of the recorded specimens of this kind were fovmd not in turbulent or relatively cold regions but in the South Equatorial Drift, ?'. e., in the area characterized by hydrographic conditions more SYSTEMATIC ACCOUNT. 503 stable and presumably more favorable to this genus than any of the remaining areas investigated by the Expedition. Hensen (1911, p. 152) probably refers to small specimens originated in this manner when he speaks of " Schwiirmlinge . . . wie solches von Ornithocercus bekannt geworden ist." What happens to these small individuals? Do they succumb after they have reached a minimum size, or do they regain the size of their ancestors by the process of ecdysis or regulatory enlargement during growth? Or, again, do they form a stage preparatory to sexual reproductions? No data bearing on these interesting and important ques- tions are available at the present writing. Megacytic specimens have been found in some species, for instance, in Orni- thocercus magnificus s. str., but the intercalary zone appears never to attain any considerable width. It may finally be mentioned in this connection that we are unable to furnish data on the mode of reproduction in the closely related genera Parahistioneis and Histioneis, and that no information on this point is published at present. How- ever, in all probability these genera resemble Phalacroma in this respect. Distribution: — • Ornithocercus is marine, eupelagic, of circumequatorial dis- tribution, and all its species are true warm-water forms, being restricted to tropical, subtropical, and warm-temperate seas. The genus is a characteristic element of the group of beautiful and luxuriant plankton organisms that has aroused the admiration and been the source of the enthusiastic descriptions of the life of tropical waters by many investigators. It is one of the groups of organ- isms causing the striking contrast between the plankton of the equatorial and the polar regions. In the Atlantic Ocean, Ornithocercus has been recorded be- tween lat. 45° N. and lat. 42° S.; in the Pacific Ocean its northern limit appears to be at about lat. 35° N. Some species, as 0. splendidus, seem to be restricted to tropical and subtropical seas ; others, as 0. thurni, are predominantly tropical and subtropical but occur also in small numbers in warm-temperate waters ; still others, as 0. magnificus, are almost as frequent in warm-temperate as in tropical and subtropical seas. The distribution of the genus as a whole appears to be fairly well established. The general distribution of several of the species, on the other hand, must be regarded as rather uncertain, due to the broad concept of species held by several investigators (see, for instance, 0. magnificus, the section on synonymy, p. 534), to the great difficulties of specific determinations, and to the fact that but few of the distributional data are accompanied by figures or de- scriptions by means of which the determinations of species by planktonologists may be checked and compared. 504 THE DINOPHYSOIDAE. Although Ornithocercus appears always to lack chromatophores (Schiitt, 1895, p. 64), it is found more commonly in surface waters than any other genus of Dinophysoidae. Most of the records of this genus published at the present writing are from surface hauls. It should l)e emphasized, however, that our knowledge of its vertical distribution is still ^'ery fragmentary. The only records based on catches made with closing nets available as yet are those of the Valdivia Expedi- tion (see 0. splendidus, 0. magnificus, and 0. qundrntus, the sections on distribu- tion, p. 526, 536, 570). Most of the species seem to be rare. Some of them, however, as Ornithocercus magnificus, 0. Ihurni, and 0. quadrntus, when compared with species of other genera of this tribe, are relatively common, and a sudden appearance in great numbers of 0. magnificus has been recorded (Schiller, 1914, p. 14). Representatives of Ornithocercus were found at 100 (78.7%) out of the 127 stations of the Expedition from which dinoflagellates were recorded. These 100 stations are distributed o\er the whole area covered by the Expedition in the following manner (Plate 28) : — 2 ( 50.0%) out of the 4 stations in the California Current 13 (100.0%) out of the 13 stations in the Mexican Current 17 (100.0%) out of the 17 stations in the Pananiic Area 8 ( 29. (i%) out of the 27 stations in the Peruvian Current 9 ( 9O.O%0 out of the 10 stations in the Easter Island Eddy 4 (100.0'>f ) out of the 4 stations in the (Jalapa^os Eddy 40 ( 89.0%) out of the 45 stations in the South Equatorial Drift 3 (100.0%) out of the 3 stations in the South Equatorial Current 2 (100.0%) out of the 2 stations in the Equatorial Counter Current 2 (100.0%) out of the 2 stations in the North Equatorial Current The mmibers of these 100 stations are as follows : — 4580, 4583 (Cahfornia Cur- rent) ; 4587, 4588, 4590, 4592, 4594, 4596, 4598, 4600, 4604, 4605, 4607, 4545, 4546 (]\Iexican Current); 4609, 4611, 4613, 4615, 4617, 4619, 4623, 4624, 4627, 4631, 4634, 4635, 4637, 4638, 4639, 4640, 4644 (Panamic .Area) ; 4646, 4647, 4648, 4650, 4659, 4664, 4671, 4676 (Peruvian Current); 4689, 4691, 4692, 4695, 4696, 4697, 4698, 4699, 4700 (Easter Island Eddy) ; 4713, 4714, 4715, 4716 (Galapagos Eddy) ; 4679, 4680, 4681, 4682, 4683, 4684, 4685, 4686, 4687, 4701, 4702, 4704, 4705, 4707, 4709, 4711, 4712, 4717, 4718, 4719, 4720, 4721, 4722, 4723, 4724, 4725, 4727, 4728, 4729, 4730, 4731, 4732, 4733, 4734, 4735, 4736, 4737, 4739, 4740, 4741 (South Equatorial Drift); 4742, 4743, 4540 (South Equatorial Current); 4541, 4542 (Equatorial Counter Current); 4543, 4544 (North Equatorial Current). At fifty-one out of these 100 stations Ornithocercus was taken in \-ertical hauls and at fiftv-seven in surface hauls. Most of the vertical hauls are from SYSTEMATIC ACCOUNT. 505 300-0 fathoms, some are from 800, 150, 100-0 fathoms. Vertical hauls were made at only sixty-eight out of the 127 stations. The genus was thus found at 75.0% of the stations at which vertical hauls were made. Disregarding the stations at which siu-face catches only were made, the record stations are distributed over the area covered by the Expedition in the following manner: — 2 ( 50.0%) out of the 4 stations in the Cahfornia Current 5 (100.0%) out of the 5 stations in the ^Mexican Current 6 (100.0%) out of the 6 stations in the Panamic Area 7 ( 31.8%) out of the 22 stations in the Peruvian Current 5 (100.0%) out of the 5 stations in the Easter Island Eddy 2 (100.0%) out of the 2 stations in the Galapagos Eddy 23 (100.0%) out of the 23 stations in the South Equatorial Drift 1 (100.0%) out of the 1 station in the South Equatorial Current There were no \-ertical hauls made in the Equatorial Counter Current or in the North Equatorial Current. The numbers of these fifty-one vertical stations are as follows: — 4580, 4583 (California Current); 4587, 4590, 4594, 4598, 4605 (Mexican Current); 4609, 4613, 4617, 4634, 4637, 4638 (Panamic .\rea); 4646, 4647, 4648, 4650, 4659, 4671, 4676 (Peruvian Current); 4689, 4691, 4695, 4697, 4699 (Easter Island Eddy); 4713, 4715 (Galapagos Eddy); 4679, 4681, 4683, 4685, 4687, 4701, 4705, 4707, 4709, 4711, 4717, 4719, 4721, 4722, 4724, 4728, 4730, 4732, 4734, 4736, 4737, 4739, 4740 (South Equatorial Drift); 4742 (South Equatorial Current). Surface catches and stomachs of Salpa taken in surface waters (eighty-one sm-face catches and twenty-four samples of Salpa stomachs) were examined from eighty-two stations. As pre\'iously mentioned, the genus was taken at fifty-seven, i. e., at 69.5% of all the surface stations. Taking into consideration the surface stations only, these fifty-seven record stations are distributed m the following manner : — 1 ( 25.0%) out of the 4 stations in the California Current 9 (100.0%) out of the 9 stations in the Mexican Current 13 (100.0%) out of the 13 stations in the Panamic Area 3 ( 15.0%) out of the 20 stations in the Peruxian Current 4 ( 80.0%) out of the 5 stations in the Easter Island Eddy 2 (100.0%) out of the 2 stations in the Galapagos Eddy 19 ( 82.6%) out of the 23 stations in the South Equatorial Drift 2 (100.0%) out of the 2 stations in the South Equatorial Current 2 (100.0%) out of the 2 stations in the Equatorial Counter Current 2 (100.0%) out of the 2 stations in the North Equatorial Current The numbers of these fifty-seven surface stations are as follows : — 4583 (California Current); 4588, 4590, 4592, 4596, 4600, 4604, 4607, 4545, 4546 (Mexi- 506 THE DINOPHYSOIDAE. can Current); 4611, 4615, 4617, 4619, 4623, 4624, 4627, 4631, 4635, 4638, 4639, 4640, 4644 (Panamic Ai-ea) ; 4646, 4648, 4664 (Peruvian Current) ; 4692, 4696, 4698, 4700 (Easter Island Eddy) ; 4714, 4716 (Galapagos Eddy) ; 4680, 4682, 4684, 4686, 4702, 4704, 4709, 4712, 4718, 4720, 4723, 4725, 4727, 4729, 4731, 4733, 4735, 4740, 4741 (South Equatorial Drift); 4743, 4540 (South Equatorial Current); 4541, 4542 (Equatorial Counter Current) ; 4543, 4544 (North Equatorial Current). Thus if we use either the records from ^•ertical hauls or the records from surface hauls as indicators of the distribution of Ornithocercus, our data show that this genus is about evenly distributed throughout the area examined by the Expedi- tion except in the California Current and in the Peruvian Current, for the genus is relatively rare in waters entering the tropics from temperate regions. The genus occurred at all the vertical as well as at all the surface stations in the Mexican Current, Panamic Area, Galapagos Eddy, South Equatorial Current, Ecjuatorial Counter Current, and North Equatorial Current. In the Easter Island Eddy and in the South Equatorial Drift it occurred at all the vertical stations and at 80.0% and 82.6%, respectively, of the surface stations. In the California Current it was absent from the northern stations; and in the Peruvian Current it occurred at only 31.8% of the vertical stations and at 15.0% of the surface stations. The scarcity of the genus in the Peruvian Current is also evident if the nimiber of species found at each record station is considered. At not less than three out of the eight record stations, only one species was found, while in other regions, as a rule, several species were found at each station (Plate 28). The rare occurrence of Ornithocercus in the Peruvian Current is probably due to the fact that this current is relatively cool while the species of this genus are distinctly warm-water forms. The cause of its absence from the northern stations in the California Current is probably that this current comes from the north and has not been warm for a sufficient length of time until it reaches the middle of Lower California. Our temperature records for the stations of this current, viz.. Stations 4571, 4574, 4576, 4580, and 4583, the last located near Point Lucas, the southernmost point of Lower Cahfornia, are as follows: — 71°, 69°, 69°, 76°, and 83°. It should be observed, however, that the genus occurs, al- though rarely, as far to the north as at San Diego, California, where it has been taken repeatedly by the senior author in the autumn when other tropical elements appear in the plankton. According to their distribution within the area investigated by the Expedi- tion, the species of this genus can be divided into two groups, viz., those relatively frequent in the Mexican Current and the Panamic Area, and those absent from SYSTEMATIC ACCOUNT. 507 or very rarely occurring in these regions. Ornithocercus magnificus (Figure 80), 0. thurni (Figiu-e 82), 0. steini (Figure 84), and 0. quadratus (Figure 88) belong to the first of these two groups; while 0. heteroporus (Figiu'e 76), 0. splendidus (Figure 78), and 0. carolinae (Figure 90) belong to the second. The reasons why the last three species avoid the Mexican Current and the Panamic Area are ob- scure. It cannot be due to too low temperature, since this is about the same as in the South Equatorial Drift where these species are relatively frequent. It may be that the unusually turbulent current conditions in these regions are responsible for this phenomenon. There are 236 records of species of Ornithocercus from vertical catches. Out of these 236 records, one (0.4%; Station 4613) showed a frequency of 16%; one (0.4%,; Station 4598) showed 15%; one (0.4%; Station 4613) showed 12%; one (0.4%; Station 4721) showed 11%; three (1.3%; Stations 4713, 4740, 4742) showed 10%; two (0.8%; Stations 4695, 4699) showed 9%; four (1.7%; Stations 4605, 4609, 4617, 4728) showed 8%; six (2.5%; Stations 4634, 4638 [2 records], 4715, 4721, 4722) showed 7%; three (1.3%; Stations 4637,4638, 4689) showed 6%; five (2.1%; Stations 4613, 4617, 4691, 4724, 4736) showed 5%; twelve (5.1%; Stations 4590 [2 records], 4598, 4605, 4609, 4638, 4683, 4697, 4722, 4724, 4730, 4740) showed 4%; ten (4.2%; Stations 4598, 4634 [2 records], 4637, 4681 [2 records], 4732, 4734, 4737, 4739) showed 3%,; thirty (12.7%,; Stations 4580, 4583, 4587, 4605, 4617, 4634, 4646, 4650, 4681, 4687, 4689 [2 records], 4691, 4695, 4697, 4701 [2 records], 4709, 4715, 4717, 4719, 4728, 4730, 4734, 4737 [2 records], 4739, 4740 [2 records], 4742) showed 2%; forty-four (18.6%; Stations 4583, 4594, 4617, 4637, 4647, 4648, 4679, 4683, 4685 [2 records], 4687, 4689, 4691, 4695 [2 records], 4697, 4699 [2 records], 4701 [2 records], 4705 [2 records], 4707, 4711, 4715, 4717, 4722, 4724, 4728 [2 records], 4732 [4 records], 4734, 4736 [3 records], 4737 [3 rec- ords], 4739 [2 records], 4742) showed 1%; 113 (47.8%) showed a frequency of less than 1 %. There are 156 records of species of this genus from surface catches. Out of these 156 records, one (0.6%,; Station 4619) showed a frequency of 37%; one (0.6%; Station 4546) showed 21%; one (0.6%; Station 4607) showed 15%; one (0.6%; Station 4607) showed 14%; two (1.3%; Stations 4604, 4543) showed 13%; three (1.97o; Stations 4607, 4615, 4545) showed 10S"o; two (1.3%; Stations 4635, 4696) showed 9%; one (0.6%; Station 4543) showed 8%; two (1.3%,; Stations 4590, 4604) showed 7%; four (2.6%; Stations 4600, 4604, 4619, 4546) showed 6%; six (3.8%; Stations 4596, 4600, 4617, 4692, 4696, 4545) showed 5%; eight (5.1%; Stations 4607, 4615, 4619, 4635, 4638, 4714, 4720, 4741) showed 4%; eight (5.1 %; 508 THE DINOPHYSOIDAE. Stations 4588, 4600, 4615, 4640, 4733, 4542, 4545, 4546) showed 3%; 26 (16.7%; Stations 4592 [2 records], 4596, 4600, 4615, 4617, 4619, 4624, 4627, 4631 [2 records], 4638 [2 records], 4639, 4648, 4682, 4692, 4712, 4714, 4716, 4720, 4731, 4733, 4741, 4540, 4543) showed 2%; 29 (18.6%; Stations 4590 [2 records], 4596, 4604, 4617, 4638, 4639, 4640 [3 records], 4644, 4702, 4712, 4716 [2 records], 4720, 4725, 4729, 4731 [2 records], 4733 [3 records], 4735, 4541 [3 records], 4542 [2 records]) showed 1%; 61 (39.1%) showed less than 1%, or did not have the fre- quency estabhshed. As will be seen from a comparison between the last two paragraphs and the corresponding paragraphs in the sections on the distribution of the other genera of Dinophysoidae, the average frequency of the indiA'idual species of Ornithocercus is higher than in any other genus of this family. Fiuihermore, the relative num- ber of surface records is decidedly higher than in other genera, as will be seen from the following values: Triposolenia vertical records, 156 surface records, 0 ratio. , X : 1 llistioueis a 94 it i( 17 5.5:1 Aniphisolenia a 210 ti u 42 5.0: 1 Phalacroma a :^n9 u u 74 4.2:1 1 tiiKipliysis a 170 a a 44 3.9:1 Ornithocercus u 236 u u 156 1.5:1 The genus Parahistioneis is not considered in this connection since most of its siu'face records refer to specimens found in Salpa stomachs. Coincident occiu-rence of different species of Ornithocercus in catches from 300 (800, 150, 100)-0 fathoms is recorded at the following of the fifty-one stations mentioned above: — eight species occurred coincidently at four stations (7.8%; Stations 4701, 4722, 4724, 4737); seven species at four stations (7.8%; Stations 4730, 4732, 4734, 4742) ; six species at eight stations (15.7%; Stations 4590, 4691, 4699, 4705, 4721, 4728, 4739, 4740) ; five species at nine stations (17.7%; Stations 4583, 4587, 4613, 4681, 4695, 4697, 4709, 4717, 4736); four species at fourteen stations (27.4%; Stations 4598, 4605, 4609, 4017, 4634, 4637, 4638, 4646, 4647, 4683, 4685, 4689, 4713, 4719); three species at seven stations (13.7%; Stations 4594, 4650, 4671, 4679, 4687, 4707, 4715); two species at one station (2.0%; Station 4711). Coincident occm^rence of different species of this genus in surface hauls is recorded at the following of the fifty-seven surface stations mentioned above : five species occurred coincidently at three stations (5.3%; Stations 4617, 4733, 4743); four species at fourteen stations (24.6%; Stations 4588, 4592, 4600, 4604, 4607, SYSTEMATIC ACCOUNT. 509 4615, 4619, 4638, 4640, 4731, 4741, 4540, 4541, 4542); three species at sixteen stations (28.1%; Stations 4583, 4590, 4596, 4631, 4635, 4639, 4644, 4692, 4712, 4716, 4720, 4727, 4729, 4543, 4545, 4546) ; two species at thirteen stations (22.8%; Stations 4611, 4623, 4646, 4682, 4686, 4696, 4700, 4702, 4714, 4718, 4723, 4725, 4735). Of tlie nine species of tliis genus found in tlie material of the Expedition not less than eight were recorded from surface hauls. The number of surface records of each of these eight species is as follows: — Ornithocercus magnificus, 42 records, i.e., at 55.3% of all the record stations of this species ; 0. stcini, 37 records (54.4%) ; 0. thurni, 36 records (48.6%); 0. quadratus, 33 records (43.4%); 0. splendidus, 4 records (11.5%); 0. heteroporus, 2 records (13.5%); 0. caroUnac,! record (4.2%); 0. orbiculatus, 1 recoi'd (100.0%); 0. formosus, no record (0.0%). A fact worth emphasizing in this connection is that 0. splendidus is comparati\'ely rare in sur- face waters in spite of the fact that it appears to be more highly adapted to flota- tion than any other species of this genus. The distribution of Ornithocercus within the area investigated by the Expedition is such that we may expect to find that this genus is rather evenly distributed in all tropical and subtropical seas. Historical Discussion and Systematics Ornithocercus was established by Stein (1883, p. 25), who did not gi\'e any diagnosis but limited himself to a few remarks on the general organization and the systematic position of this genus and to the presentation of figures of five speci- mens, all, according to him, members of a single species, 0. magnificus. Diagnoses of this genus were published later by Biitschli (1885, p. 1011), Delage and He- rouard (1896, p. 386), and Schutt (1896, p. 28). All these diagnoses are fairly satisfactory; the one by Schutt (1896) is by far the most nearly complete and displays the best evaluation of the systematic importance of the individual char- acteristics. No extensive description and discussion of the morphology of this genus ha\'e been published at the present writing, and the amount of original observations presented is relati\ely small. The last fact is due to the limitation of most of the species to tropical and subtropical seas and to the complete absence of this genus from the northern European waters, which are the only areas that have been submitted as yet to intensive investigations. Besides magnificus the following specific names of Ornithocercus are to be found in the literature: — 0. splendidus Schutt (1893) ; 0. splendens Schutt (1896) ; 0. quadratus and 0. steini Schiitt (1900a); 0. carolinae, 0. Iietervporus, and 0. 510 THE DINOPHYSOIDAE. serratus Kofoid (1907a); 0. formosus and 0. orbiculatus Kofoid and Michener (1911); 0. com-pactus, 0. horridus, and 0. papillosus Hensen (1911); 0. assimilis and 0. minor Jorgensen (1923). Representatives of this genus were figured also by Pouchet (1883) under the name of Di7iophysis galea and by Schmidt (1888) as Parelion thurnii (Schmidt considered his species a diatom). In the present paper the following new names have been established : Ornithocercus quadratus f . schutti, 0. quadratus f. simplex, and 0. quadratus f. intermedia. Of the species mentioned in the last paragraph, Ornithocercus compactus, 0. horridus, and 0. papillosus were introduced by Hensen (1911) as nomina nuda, and there is no information published as yet as to their organization. 0. splendens Schiitt (1896) is a lapsus pennae for 0. splendidus. 0. minor, a name used by Jorgensen (1923) in his manuscript, is a synonym of 0. magnificus s. str. The specimens represented by Pouchet (1883, fig. G: 1, 4) as Dinophysis galea cannot be assigned with certainty to species, due to the sketchy natm'e of the figures. Ornithocercus magnificus Stein (1883) is not a homogeneous systematic unit. Schiitt (1900a), who was the fii'st to pomt out this fact, applied the name 7nag- nificus to the specimen represented by Stein's (1883) Plate 23, figure 1, and made the specimen represented by Stein's (1883) Plate 23, figure 3, the type of a new species, which he called 0. steini. However, since specimens of the last species had been figured previously as Parelion thurnii by Schmidt (1888) in a manner leaving no doubt as to their specific allocation, steini must be replaced by thurni as the name of this species. The specific name steini thus should have been dis- carded as a sjTionym, had it not been for the fact that out of the three specimens figured by Schiitt (1900a) as Ornithocercus steini, only one, his figure 7, is referable to this species as conceived in the present paper. The two other specimens, his figures 5 and 6, are, according to our opinion, fairly young and imdifferentiated members of a species later described and figured by Kofoid (1907a) as 0. serratus. Under these circumstances the name 0. sicini s. str. has to be maintained for these two specimens of Schiitt (1900a), and 0. serratus Kofoid must be discarded as a synonym. 0. orbiculatus Kofoid and Michener (1911) may be based on an ab- normal specimen of 0. steini s. str. and thus its specific status must be regarded as uncertain. 0. assimilis Jorgensen (1923) has in the present paper been re- ferred to 0. quadratus under the designation 0. quadratus f. assimilis. However, it should be emphasized that there are fairly strong reasons suggesting the specific independence not only of this forma but also of the remaining four formae in- cluded by us in this species, viz., f. qnadrata, f. sclmtti, f. simplex, and f. intermedia. With regard to the specific differentiation between Ornithocercus carolinae and SYSTEMATIC ACCOUNT. 51 1 Histioneis [Parahistioneis] francescae, see the first of these two species, the section on comparisons fp. 575). Some of the specimens of Ornithocercus carolinae are transitional to Histioneis francescae. Jorgensen (1923, p. 38), who accepted Orni- thocercus carolinae as a valid species, even suggested that these two forms "may perhaps be identical." We consider it necessary to treat them tentatively as speci- fically distinct. Thus out of the species of Ornithocercus established as j'et, only eight have been regarded in the present paper as valid, viz., 0. carolinae Kofoid (1907a), 0. formosus Kofoid and Michener (1911), 0. heteroporus Kofoid (1907a), 0. mag- nificus Stein (1883) s. str., 0. quadratus Schiitt (1900a), 0. splendidxis Schiitt (1893), 0. steini Schiitt (1900a) s. str., and 0. thurni (Schmidt, 1888); and of 0. quadratus there are fiveformae, which may be specifically independent. All the descriptions of species published up to the present writing are very incomplete and include no or but few data of variation. The descriptions published by Kofoid (1907a) and Kofoid and Michener (1911) are preliminary and based on the same material as those of the present paper. The extraordinary variabil- ity, partly due to regulatory processes, and the uncertainty of specific delimita- tions have rendered the taxonomy of this genus very difficult, and the synonymies of some of the common species, viz., Ornithocercus magnificus s. str. and 0. thurni are tangled almost beyond the possibility of unravelling. No subdivision of this genus has been attempted as yet, with the exception of the establishment by Murray and Whitting (1899, p. 332) of the subgenus Paraschiittia, which was based on a single species, Ornithocercus splendidus. A few contributions to the morphology of this genus were made by Biitschli (1885, p. 944) and Schiitt (1895, p. 19, 23, 54, 64, 112, 114, 116, 130, 131, 136, 138). Schutt (1899, 1900a) submitted the growth and differentiation of the thecal wall and of the lists to a quite detailed analysis. A few remarks as to the adaptive- ness of the lists in flotation were made by Carisso (1911), Gran (1912a, 1912b), Schiitt (1893), and Steuer (1910, 1911). Some suggestions as to the functions of the furrows and lists were made by France (1923). Hensen (1911, p. 152) men- tions the formation in this genus of small specimens, "Schwarmlinge," the results of rapid binary fission. Contributions to our knowledge of the distribution of Ornithocercus are to be found in the following papers, not specifying those previously mentioned in this section:— Cleve (1897a, 19C0b, 1901a, 1901c, 1902b, 1903b), Cleve, Ekman, and Pettersson (1901), Daday (1888), Entz (1902b, 1905), Faria and Cunha (1917), Forti (1922), Graf (1909), Hensen (1895), Issel (1921), Jorgensen (1911), 512 THE DINOPHYSOIDAE. Karsten (1906, 1907), Kofoid (1911), Lemmermann (1899a, 1901a, 1904, 1905), Lindemann (1924, 1925), Lohmann (1902, 1908, 1920), Mangin (190G), Murray and \\Tiitting (1899), Okamura (1907, 1912), Ostenfeld (1898a, 1915), Ostenfeld and Schmidt (1901), Pavillard (1905, 1916), Schiller (1912, 1912a, 1914), Schmidt (1901), Schroder (1900a, 1906a, 1906b, 1909, 1911), Stuwe (1909), Whitelegge (1891), and Zacharias (1906). Of these investigators only Murray and Wliitting (1899), Okamura (1907), and Schroder (1900a) give original figures by means of which the correctness of the determinations may be checked. Forti (1922) gives reproductions of previously published figures of the species recorded. Many, perhaps even most, of the data on the distribution of this genus should be ac- cepted tentatively on account of the great difficulties implied in the determina- tion of the species and to the broad concept of species held by several of the in- vestigators. References to Ornithocercus or mmor contributions to our knowledge of this genus are found also in Balbiani (1884c), Bergh (1884), Calkms (1902), Chun (1886, 1903), Doflein (1909, 1911, 1916), Effenberger (1911), Entz (1902a), Fricke (1902), Gadeceau (1909), Goldschmidt (1909), Hensen (1891), Hjort (1911), Klebs (1912), Meunier (1910), Nathansohn (1910b), Oltmanns (1922), Schvitt (1900b), Stiasny (1913), Theel (1909), Wallengren and Hennig (1911), Walther (1893), Willey and Hickson (1909), and Zacharias (1906a, 1907, 1911). Accepting the decision of Murray and WTiitting (1899), the following investi- gators use Histioneis instead of Ornithocercus as the generic name of the species of this genus: — Cleve (1900b, 1901a, 1901c, 1902b, 1903b), Cleve, Ekman, and Pettersson (1901), Entz (1902b, 1905), Forti (1922), Gadeceau (1909), Issel (1921), Lenm.crmann (1899a, 1901a, 1904, 1905a), Pavillard (1905, 1916), and Schroder (1900a, 1906a). With regard to this usage, see our treatment of the family Ornithocercidae, the section on the subdivisions (p. 494). The following misspellings of the name Ornithocercus have been found by us : Ornithocerus Schutt (1893, fig. 83), Fricke (1902, p. 55), Th^el (1909); Ornitho- ceras Graf (1909, p. 139). Adaptive and Systematic Value of the Characters. Principles used in THE Description of the Species Among all the genera of Dinophysoidae, Ornithocercus is the one that occurs most frequently in the surface waters of the tropical and subtropical seas. Since these waters ai-e characterized by a lower buoyancy than the deeper strata in the same regions and the waters nearer to the poles, it appears reasonable to expect SYSTEMATIC ACCOUNT. 513 this genus to be more highly differentiated for flotation than any other genus of this tribe. This expectation, however, is hardly borne out by a comparative morphological study. Ornithocercus undoubtedly is well adapted to flotation, but at the same time it has remained relatively primitive in many respects. The most striking examples of adaptations to flotation are not found within this genus but in the closely related genus Histioneis, in which the evolutionary processes have been carried to the extreme, resulting in a multitude of bizarre species, in most respects profoundly different from what might be called the basic type of this family. In regard to the shape of the theca, comparisons should also be made with the extreme genera Triposolenia and Amphisolenia. Perhaps the most outstanding feature in the evolution of Ornithocercus is the fact that the theca has remained largely undifferentiated. Indeed, in none of the other large genera of this tribe has the theca been subject to so little change. It is, generally speakmg, still subcircular in lateral outline, somewhat compressed bi- laterally, and relatively large when compared with the whole organism. In other words, while in some of the other genera, e.g., Triposolenia and Amphisolenia, the shape of the theca has been modified in response to the needs of flotation, Ornitho- cercus has remained nearly unchanged in this respect. Furthermore, while the members of Phalacroma and Dinophysis and of the other more or less primitive genera are fairly active swinuners, the organization of Ornithocercus indicates that the species of this genus are relatively slow. How, then, do these species keep themselves within their optimum habitat? We suggest the following answer to this question. (1) The specific gravity is comparatively low, approachmg that of the surrounding medium. This has been brought about in a twofold manner: first, the protoplasm contains fairly large pusules and mclusions of low specific gravity, such as fats and oils (Schutt, 1895, pi. 5, fig. 21: 2); second, the transverse furrow and the thin cingular lists have been strikingly widened and include a large space filled principally with water. Thus the volume of the organism has been increased decidedly without a corre- spondmg increase in the absolute weight. (2) In all the species the cingular lists are very large and undoubtedly act as parachutes, when the organism is in an upright position. In Ornithocercus splendidus these lists appear to be the principal means of flotation ; they have an excessive development and a very wide spread (Figure 77: 7). (5) The left sulcal list, which is very large and in most species extends nearly around the entire hypotheca, has at least two main functions: first, it increases very greatly the surface of resistance whenever the organism is turned upon one of its lateral faces, and the asymmetry and the lobes of this list 514 THE DINOPHYSOIDAE. probably cause the organism in this position to descend m a spiral or zigzag line ; second, it acts as a keel and a rudder, thus stabilizing the progressive movement. (4) The body is fairly small. In the largest specimens of this genus found in the material of the Expedition it was only 67.6 yu long, while of Phalacroma specimens were found by us measuring nearly 150 /j in length (P. giganteum). (5) In most species the surface friction is hicreased by structural differentiations in the theca and by the ribs and reticulation of the lists. These five points are arranged according to their probable relative import- ance, the first being the most important. The investigators who previously have touched upon the phenomena of adaptation to flotation in Ornithocercus (see the historical discussion of this genus, p. 509) have mentioned only the second of these points. The small size of the right sulcal list and the oblique position of the left sulcal list (the anterior portion of this list belongs to the left valve, the posterior portion to the right) should be correlated with the spiral mode of swimming. France (1923, p. 31), who interprets these organisms as mmiature turbines, emphasizes the fact that the transverse fiu'row is narrower ventrally than dorsally and that the longitudinal furrow is narrowest posteriorly. "By this the instreamhig water is retarded slightly in its course, producmg a back-pressure, an economical over- pressure which can be observed in the accelerated motion of the cell." It may also be mentioned in this connection that besides decreasing the specific gravity of the organism, the large chamber enclosed by the girdle also functions as a container for the phaeosomes. However, this appears to be a sec- ondary function, since most specimens have very few (or no) phaeosomes, occupy- ing only a small portion of the enclosed space. All the external characters have been utilized in our descriptions and separa- tions of the species of this genus. However, since the size of the body varies greatly within the species, due to the mode of fission illustrated by Plate 18, figure 2, 7-9, and since most of these species agree fairly closely in the shape and structure of the body and m the size and shape of the cingular list, the most im- portant distinguishmg characteristics have been taken from the left sulcal list. Indeed, most of the species, e.g., Ornithocercus magnificus, 0. thurni, 0. sleini, 0. orbiculatus, 0. quadratus, and 0. formosus, are based almost exclusively on differ- ences in the size, shape, and structure of this list. Unfortunately, the characters that are most profoundly modified within the genus as a whole also are subject to the most extreme fluctuations. These fluctua- tions, which are due to regulatory processes following binary fission, to adjust- SYSTEMATIC ACCOUNT. 515 ments to flotation, and to irregularities in the regeneration following local injuries, render the establishment and characterization of the species exceedingly difficult. For instance, in the case of 0. qnadratus (Figure 86, 87) we have had to establish fixeformae, but these may possibly represent five different species. Furthermore, we found some specimens in the material of the Expedition that were too aberrant for certainty of specific assignment (Figure 92) . The followmg principles have been used by us in describing the species of this genus: — {1 ) Wlien not otherwise stated, all characteristics refer to specimens in lateral view. (2) The terminology and the methods of measurement are largely the same as in Phalacroma and Dinophysis; see Figure 29. The angles are meas- ured as in Figure 94 of Histioneis. It should be mentioned that the length of the fission rib and of the dorsal main rib (the a-rib) of the left sulcal list is measured from the base of the rib to the margin of the list; while the lengths of the re- maining main ribs of this list behind the fission rib (the b-, c-, d-, and e-ribs) are measured from the bases of the ribs to the submarguial rib. (3) The terms length and depth of body refer to theca exclusive of lists. (4) The number of ribs in the cingular lists given in the diagnoses is the number on each valve. Subdivisions. Relationships among the Species No attempt at subdividing Ornithocercus has been carried out hitherto, "with the exception of the establishment by Murray and Wliittmg (1899, p. 332) of the subgenus Paraschiittia, which was based on a single species, Ornithocercus splendidus. According to our opinion, this genus may be divided into three groups as follows. 1. Splendidus group: — 0. heteroporus and 0. splendidus (Figure 75, 77). 2. Magnificus group: — 0. magnificus s. str., 0. thurni, 0. steini s. str., 0. orbiculatus, 0. quadratus, and O.carolinae (Figure 79, 81, 83; Plate 17, fig. 7; Figure 86, 87, 89). 3. FoRMOSUS group: — 0. fonnosus (Figure 91). 1. Splendidus group (syn. Paraschiittia). Epitheca subhorizontal. Theca porulate and but seldom areolate. Width of transverse furrow and inclination of cingular lists variable. Left sulcal list two-lobed and ends at or near antapex of body. Ventral margin of right sulcal list free. Of the two species of this group, Ornithocercus heteroporus is the more primi- X'we in the size of the body and in the size and structure of the cingular lists. On the other hand, it is slightly more advanced in the width of the transverse furrow 516 THE DINOPHYSOIDAE. and in the inclination of the Usts. This species may be considered, on the whole, as the most primitive member of this genus. £. Magnificus group. Transverse furrow decidedly wider dorsally than ventrally, and epitheca thus somewhat inclined ventroposteriorly. Theca areo- late as well as porulate. Posterior cingular list somewhat more inclined anteriorly than the anterior. Left sulcal list three- to five-lobed, squarish, seldom two- lobed, and ends on dorsal side of body. Ventral margin of right sulcal Ust free. Within this group, which comprises the typical members of this genus, Orm- thocercus carolinae occupies structurally a somewhat isolated position, being char- acterized especially by the very large antapical lobe and the numerous radial ribs of the left sulcal list. The remaining species of this group are fairly closely relateil and have reached about the same evolutionary level. We are unable to decide which of them, all characters considered, is structurally closest to the ancestral type. 0. orbiculatus may be found to be an aberrant specimen of 0. steini s. str. 3. FoRMOSUS group. Transverse furrow decidedly wider dorsally than ven- trally; epitheca somewhat inclined ventroposteriorly. Theca areolate and poru- late. Posterior cingular list inclined anteriorly somewhat more than the anterior. Left sulcal list two-lobed and ends on dorsal side of body. Ventral margin of right sulcal list attached to left sulcal list. Of these three groups the Splendidus group appears to be the most primitive. The obovate shape of the body in lateral as well as in dorsoventral view character- istic of most specimens of Ornithocercus splendidus, the subunifor'm width of the transverse furrow and the subequal inclination of the two cingular lists also found in this species, the subhorizontal position of the epitheca, and the ending of the left sulcal list at or near the antapex of the body are relatively primitive features suggestive of this assumption. On the other hand, it must be remembered that 0. splendidus is more advanced than the other species of this genus in the exces- sive width of the cingular lists. The FoRMOSUS group is somewhat isolated and occupies structurally in some respects a position intermediate between Ornithocercus and Histioneis (see Orni- thocercus formosus, the section on comparisons, p. 579). Key to tJie Species of Ornithocercus 1. Left .sulcal list jjiangular 2. 1. Left sulcal list not biangular 5. 2. Left sulcal list ends at or near antapex 3. 2. Left sulcal list ends on dorsal side of body 4. •i. Lengt h of body 34 m or less; .interior cingular list 0.48-0.68 the depth of body ...(). helerojmni.i Kofoid. 3. Length of body more than 34 /i; anterior cingular list 0.77-1.46 the depth of body. O. splendidus Schiitt. SYSTEMATIC ACCOUNT. 517 4. With only one strong rib in left sulcal list behind fission rib 0. formosm Kofoid :ind Michener. 4. With more than one strong rib in left sulcal list behind fission rib 0. qmdralv.s Schiitt . 5. Left sulcal list with three lobes; n-rib ends in dorsal lobe 6. 5. Left sulcal list not with three lobes; o-rib does not end in dorsal lobe 8. (5. Left sulcal list with more than eight radial ribs behind fission rib; its antapical lobe decidedly larEcr than the two other lobes O. carolinne Kofoid. 6. Left sulcal list with less than eight radial ribs behind fission rib; its three lobes subequal or nearly so. 7. 7. c- and e-ribs of left sulcal list end at vertices of antapical and posteroventral lobes, . O. thurni (Schmidt). 7. c- and e-ribs of left sulcal list do not end at vertices of antapical and posteroventral lobes. 0. 7iiagmfinn Stein s. ftr. 8. Left sulcal list with four or more usually distinct lobes behind fission rib; ribs usually regular. O. sleini Schiitt i. atr. 8. Left sulcal list with rounded margin behind fission rib; its ribs irregular and partly incomplete. 0. orbiculaius Kofoid and Micheuer. 1. Splendidus Group. Both the species belongmg to this group, viz., Ornithocercus heteroponis and 0. splendidus, were found in the material of the Expedition. Ornithocercus heteroporus Kofoid Plate 18, fig. 1, 3. Figure 75, 76 Ornithocercus heteroporus Kofoid, 1907a, p. 206, pi. 12, fig. 70. Jorgensen, 1923, p. 37, 38, fig. 54. Diagnosis: — Lateral outline obliquely subellipsoidal or subcircular. Cingii- lar lists: anterior 0.48-0.68, posterior 0.37-0.60 the greatest depth of body; an- terior with five to seven radial ribs, the bases of which may be connected by re- ticulation; posterior with nine to ten usually simple ribs. Left sulcal list usually ends at or somewhat ventrally to antapex but may extend somewhat farther ; with tivo lobes, one of which is postero\'entral, the other posterior; width at postero- ventral lobe 0.56-1.06, at posterior lobe 0.54-1.06 the greatest depth of body; lobes narrowly rounded to subacute, large to fairly small; distance between their tips 0.81-1.41 the greatest depth of body; dorsal margin straight, or gently con- vex or concave ; with three to six radial ribs and submarginal rib behind fission rib. Length, 27.4-34.0 m- Widely distributed in tropical, subtropical, and warm-temperate seas. Description: — This is a small species, obliquely subellipsoidal or subrotund in lateral outline, widest near the middle. The ratio between the length and the depth of the body is 1.04 (0.99-1.08) : 1. The longitudinal axis is deflected postero- dorsally at 10° (5°-20°). The epitheca is 0.37 (0.28-0.45) as deep as the hypotheca, gently convex, highest in or near the center, and subhorizontal or but slightly tilted %-entro- posteriorly. The transverse furrow is gently convex or flat; its dorsal width is 0.38 (0.36-0.42) the greatest depth of the body; its ventral width is 0.57 (0.50- 518 THE DINOPHYSOIDAE. 0.66) its dorsal width. The posterior cingular hst is 0.23 (0.19-0.26) the length of the body from the apex. The dorsal, posterior, and ventral margins of the hypo- theca are well rounded and confluent ; or any of them may be slightly flattened. ^\'hen seen in dorsoventral view, the body is broadly ellipsoidal and widest near the middle; the anterior end is almost as broadly rounded as the posterior; the ratio between the length and the width is 1.14: 1. The anterior cingular list is 0.56 (0.48-0.68) the greatest depth of the body and has an anterior inclination of 30° (25°-40°) ; it has on each valve five to seven complete radial ribs, the bases of which may be connected by reticulation. The posterior cingular list is 0.48 (0.37-0.60) the greatest depth of the body and is inclined anteriorly at 60° (55°-75°) ; it has on each ^'alve nine to ten complete, subequidistant radial ribs, which usually are simple, but some of which may be proxunally branched or connected by reticulation. The right sulcal list is of moderate size and ends near the fission rib of the left sulcal list ; it either is sub- triangular, decreasing gradually in width posteriorly, or it is of subuniform width throughout the greater portion of its length and gently rountled posteriorly. The left sulcal list usually ends at or somewhat ventrally to the antapex of the body, but it may extend somewhat farther (Jorgcnsen, 1923, fig. 54). It has two nar- rowly rounded to subacute lobes, one posteroventral, the other posterior, and these usually are of moderate size but may be fairly small to quite large. The width of this list at the anterior main rib is 0.49 (0.36-0.60), at the fission rib 0.46 (0.32-0.54), at the posteroventral lobe 0.80 (0.56-1.06), at the narrowest place between the lobes 0.46 (0.24-0.61), and at the posterior lobe 0.76 (0.54-1.06) the greatest depth of the body. The distance between the tips of the lobes is 1.13 (0.81-1.41) the greatest depth of the body. The margin of this list between the anterior main rib and the tip of the posteroventral lobe may be nearly straight, but usually it is gently concave. Between the tips of the lobes the margin is moderately concave in most specimens, but it may be slightly or strikingly con- cave. The dorsal margin of this list is straight, or gently convex or concave. In front of the fission rib this list generally has one or two ribs; when two ribs are present, they may either be simple, or they may anastomose, and their bases may be connected by reticulation. Behind the fission rib there are three to six simple radial ribs, the tips of which are connected by a submarginal rib. The submar- ginal rib extends to the fission rib, which it joins at any point between the tip and the base (Figure 75: 1, 2, 4). The radial ribs may be furnished with very short branches (Figure 75: 1) ; and some of them may not quite reach the submarginal rib. Reticulations have not been found behind the fission rib. SYSTEMATIC ACCOUNT. 519 The thecal wall is populate with evenly distributed pores; these may either be all alike, or one in six or eight may be darker than the others. Sometimes areo- lation is developed of about the same type as in Ornithocercus carolinae. Fairly large phaeosomes have been observed in the transverse furrow. The dimensions of seven of our specimens, including the type, and of the specimen figured by Jorgensen (1923, fig. 54) were measured. Figure 75. — Ornithocercun heteroporus Kofoid, right lateral view, illiis- f rating variations in the shape and structure of left sulcal list. X 3-10. 1, from Station 4681 (300-0 fathoms); 2, from Station 4732 (300-0 fathoms); 3, from Station 4722 (300-0 fatlioms); 4, from Station 4734 (300-0 fathoms); 5, from Station 4737 (300-0 fathoms). Dimensions: — Our specimens: Length of body, 27.4-34.0 ix (average, 30.2 Mi type, 21.1 n). Greatest depth of body, 25.7-31.4 m (average, 28.9 m; type, 25.7 ix). Jorgensen's (1923) specimens were 32-34 jx long and 31-32 ix deep. Variations: — WTiile the size of the body and the size and shape of the cingu- lar lists are quite constant, the shape and structure of the left sulcal list are sub- ject to considerable variations. The two lobes of the left sulcal list sometimes (Figure 75:4) are very large, sometimes of moderate size (Figure 75: 1). The number of radial ribs in this list behind the fission rib ranges from three (Figiu'e 75: 1-3) to six (Figure 75: 5). The shape of the body also exhibits some vari- ability; usually it is obliquely subellipsoidal, but it may be subrotund (Figure 75:3). Comparisons: — Our description and figures are based on the type material. The specimen figiu'ed by Jorgensen (1923, fig. 54) agrees so closely with some of our typical representatives of this species (Figure 75 : 3) that its allocation may be considered as certain. This species, as conceived in the present paper, includes specimens of quite different habitus, and its systematic unity may appear some- what questionable. That these specimens have been lumped together is due to the fact that their variation seems to be continuous. The small size, the shape of the body, and the fact that the left sulcal list usually ends at or somewhat ventrally to the antapex of the body mdicate that this species is fairly low in the evolutionary scale of this genus. Its closest relative possibly is Ornithocercus splendidus, which it resembles in the shape of the body 520 THE DINOPHYSOIDAE. and in having a two-lobed left sulcal list usually ending at or somewhat ventrally to the antapex. These two species are easily distinguished from each other by the width of the cingiilar lists and by the structure of the left sulcal list. 0. splen- didus has much more highly elaborated and expanded lists, though of the same fundamental type. Synonymy: — This species was established by Kofoid (1907a) under the name of Ornithocercus heteroporus, a name also used by Jorgensen (1923). Figure 76. — Occurrence of Ornithocercus hetcroporiis Kofoid. Large, solid circles indicate records from vertical hauls; squares, records from surface hauls (Salpa) ; small, solid circles, stations at which this species was not found; small, open circles, stations from which no plankton catches were examined. Occurrence: — Ornithocercus heteroporus is recorded at fifteen of the 127 stations. There are 2, 0, 2, 2, 8, and 1 stations on the six lines of the Expedition. Of these fifteen stations, two (4588, 4.590) are in the Mexican Current; two (4691, 4699) are in the Easter Island Eddy; ten (4681, 4701, 4717, 4722, 4724, 4730, 4732, 4733, 4734, 4737) are in the South Equatorial Drift; and one (4742) is in the South Equatorial Current. At two stations (4588, 4733), the species was found in stomachs of Salpa taken in surface waters; all the other records refer to hauls from 300-0 fathoms. The temperature range of these fifteen stations at the surface was 68°-83°; the average was 77.3°. The frequency is in all cases less than 1 %. SYSTEMATIC ACCOUNT. 521 This species was first recorded by Kofoid (1907a) from Station 4699 of the Expedition, which thus is the type locahty. Later Jorgensen (1923) found it to be widely spread in the Mediterranean and he also recorded it from the Bay of Cadiz, from a station "south-southeast of St. Helena," from the Guinea Current, and from off the southeast coast of >\rabia. This is probably a eupelagic, stenothermal, and warm-water species, widely distributed but rare in tropical, subtropical, and warm-temperate seas. Its ab- sence from the California Current, Panamic Area, and Peruvian Current as well as its rare occurrence in the Mexican Current, according to our data, are note- worthy. Its prevalence in deeper waters should also be observed. Ornithocercus splendidus Schiitt Plate 16, fig. 2, 4; Plate 17, fig. 3. Figure 77, 85: 3, 78 Orniihocercus splendidus Schutt, 1893, p. 272, fig. 82 (as Ornithocerus splendid^is in fig. 83); 1895, p. 19, 130, 131, 150, pi. 5, fig. 22; 1899, p. 14, 16, 50, pi. 6, fig. 11; 1900a, p. 9, 20, fig. 11. Hensen, 1895, p. 189; 1911, tab. 15. Karsten, 1906, pi. 187; 1907, p. 240, 247, 248, 249, 253, 256, 268, 272, 273, 274, 276, 287, 292, 294, 295, 296, 298, 299, 301, 303, .304, 305, 306, 309, 310, 311, 312, 316, 318, 319, 336, .338, 340, 347, 348, 351, 353, 354, 355, 356, 357, 427, 471, 473. Doflein, 1909, p. 457, fig. 404A; 1911, p. 522, fig. 463A; 1916, p. 423, fig. 375A. Stuwe, 1909, p. 236, 237, 2.54, 288. Steuek, 1910, p. 197, fig. 107; 1911, p. 103, fig. 83. Nathansohn, 1910b, p. 60, 62, fig. 30. Carisso, 1911, p. 89. Effenberger, 1911, fig. 27: B. Hjort, 1911, pi. 367, fig. 11:5. Gran, 1912a, p. 935, fig. 7a; 1912b, p. 328, 329, fig. 234a. Stiasny, 1913, p. 46, 77, fig. 12. France, 1923, fig. 11 : 5. Lindemann, 1924, p. 10;1925, p. 96. Ornithocercus splendens Schutt, 1896, pi. 29, fig. 41B. Graf, 1909, p. 138, 149, 165, 173, 176, 179, 184, 186, 188. Histioneis splendida Murray & Whitting, 1899, p. 332, tab. 3, 68, pi. 32, fig. la-c. Lemmermann, 1899a, p. 320, 374; 1901a, p. 377. Schroder, 1906a, p. 329. G.^deceau, 1909, pi. 9, fig. "2 en haut." Hislioneis splendidus Cleve, 1901c, p. 252; 1902b, p. 32. Ornilhocercxis magnificus, Doflein, 1909, p. 465, fig. 417:8; 1911, p. 531, fig. 478:8; 1916, p. 436, fig. 396: 8. Oltmanns, 1922, fig. 38: 8. Ornithocerus splendidus Thbel, 1909, p. 239, fig. 39. Diagnosis: — Lateral outline subobovate, subeUipsoidal, subcircular, subtri- angular, or subtrapeziform. Cingular lists are huge: anterior 0.77-1.46, posterior 0.81-1.41 the greatest depth of body; both with 6-24 complete, sometimes irregu- lar ribs, connected by more or less developed anastomoses. Left sulcal list ends at or somewhat ventrally to antapex; with two lobes, one of which is postero- ventral, the other posterior; width at posteroventral lobe 0.96-1.92, at po,sterior lobe 0.80-1.53 the greatest depth of body; lobes narrowly rounded distally, and sometimes long and narrow, sometimes small; distance between their tips 0.82- 2.07 the greatest depth of body; dorsal margin straight, moderately convex or concave, or sigmoid; with two to five, often incomplete, ribs behind fission rib; reticulations in lobes and often at tip of fission rib and between lobes. Submargi- nal rib usually absent. Length of body, 37.7-114.0 p.. 522 THE DINOPHYSOIDAE. Widely distributed in tropical, subtropical, and warm-temperate seas. Description: — This species is usually medium-sized but may be fairly small or very large. The shape of the body in lateral outline is quite variable : some- times it is subsymmetrically or obliquely subobovate, sometimes subrotund or obliquely subellipsoidal, sometimes rounded subtriangular or rounded sub- trapeziform; and these various shapes are connected by transitions. The greatest depth is located at or somewhat behind the posterior cingular list, or near the middle of the body. The ratio between the length and the depth of the body is 1.05 (0.76-1.24) : 1. In our material two groups can be distinguished: one in which the body is deep (ratio 0.76-0.85: 1), the other with the body of moderate depth (ratio 1.04-1.24: 1). The longitudinal axis may be perpendicular to the posterior cingular list, but usually it is somewhat deflected posterodorsally; the maximum deflection as yet observed is about 20°; the average deflection is 10°. The epitheca is 0.68 (0.50-0.88) as deep as the hypotheca, usually gently convex, but sometimes flat or strongly convex, highest in, or dorsally or ventrally to, the center, and subhorizontal. The transverse furrow is flat or gently convex; its dorsal width is subequal to or but slightly greater than the ventral, and 0.22 (0.11-0.28) the greatest depth of the body. The posterior cingular list is 0.30 (0.17-0.40) the length of the body from the apex. The dorsal, posterior, and ^'entral margins of the hypotheca are confluent, and any one of them may be somewhat flattened. WTien seen in dorsoventral view, the body is regularly ob- ovate, widest at or somewhat behind the posterior cingular list; the ratio between the length and the width is 1.4-1.7: 1. The cingular lists are wider relatively than in any other known species and resemble a pair of everted parachutes. The anterior of them is 1.17 (0.77-1.46), the posterior 1.15 (0.81-1.41) the greatest depth of the body; both have an an- terior inclination of 30°-50°, and both are furnished with 6-24 complete radial ribs, which are connected by more or less developed anastomoses, and which may be quite iiTegular. The right sulcal list, which is small, ends somewhere between the fission rib of the left sulcal list and the second rib behind the fission rib. It either is subtriangular, decreasing gradually in width posteriorly ; or it is of sub- uniform width throughout the greater portion of its length and strongly rounded or rounded subrectangular (Plate 17, fig. 3) posteriorly. Its width in the middle is 0.14-0.20 the greatest depth of the body; and it may be furnished with one to several, regular or more or less irregular ribs. The left sulcal list, which ends at or somewhat ventrally to the antapex of the body, has two lobes, one of which is posteroventral, the other posterior in position. At the posteroventral lobe the SYSTEMATIC ACCOUNT. 523 width of this Ust is 1.40 (0.96-1.92), at the posterior lobe 1.23 (0.80-1.53) the greatest depth of the body; at the anterior main rib and between these two lobes the corresponding values are 0.75 (0.67-1.00) and 0.74 (0.48-0.93), respectively. The two lobes are narrowly rounded distally, and sometinies long and narrow, sometimes small. The distance between their tips is 1.38 (0.82-2.07) the greatest depth of the body. The margin between the anterior main rib and the tip of the posteroventral lobe may be almost straight, but usually it has a more or less pro- nounced concavity. The concavity between the two lobes sometimes is rather shallow, sometimes it is very deep. The dorsal margin of this list is nearly straight in some specimens, in others it is gently concave, moderately convex, or strikingly sigmoid. There are two to five radial ribs behind the fission rilj. These ribs usu- ally are simple, straight or gently cur^'ed, and the posterior ends either in or some- what ventrally to the posterior lobe. Some may extend to the margin of the list, but most of them usually end before the margin is reached. The posterior (jne may have a dorsal branch near its base (Murray and Whitting, 1899, pi. 32, fig. la). A submarginal rib may be present between the two lobes (Figure 85:3). More or less developed reticulation is to be found in the two lobes, at the end of the fission rib, and sometimes at the end of one of the ribs ending between the lobes. In front of the fission rib this list has one to three ribs. The thecal wall has numerous pores ; on the hypotheca each of these some- times is surrounded by a small areole. Near the middle of the body there are about 10-15 pores across the face of each valve. We have no record of the pres- ence of phaeosomes. The dimensions of twenty of our specimens and of the specimen figured by Murray and Whitting (1899, pi. 32, fig. 1) were measured. Dimensions: — Our specimens: Length of body, 37.7-49.8 m (average, 43.2 fi). Greatest depth of body, 32.9-62.2 ju (average, 43.3 ii). In our large specimens the greatest depth is 43.0-62.2 n (average, 57.5 m) ; in our small specunens it is 32.9-39.2 ju (average, 35.6 m). The type specimen, according to Schiitt (1893, fig. 83), was about 114 m long. The specimen figured by Murray and Wliitting (1899, pi. 32, fig. la) was 87.2 ^ long and 84.0 m deep. Variations: — This species shows a pronounced variability in most of its characters. The length of the body varies from 37.7 to 114.0 n. In our material two very distinct size-groups can be distinguished; in one of these the body is very deep (43.0-62.2 m; average, 57.5 m), in the other it is of moderate depth (32.9-39.2 m; average, 35.6 ix). The shape of the body is also subject to striking variations ; sometimes it is subsymmetrically (Plate 16, fig. 4) or obliquely (Figure 524 THE DINOPHYSOIDAE. 77:8) subobovoidal, sometimes subrotund (Figure 77:4), sometimes obliquely subellipsoidal (Figure 77:3), sometimes rounded subtriangular (Figure 77: 11), sometimes rounded subtrapeziform (Figure 77: 9, 10), and these various forms are FioUKK 77. — Ornithdcemis splcmlUhis Schiitt. 1-6, 8-11, in rinlit hilenil view; 7, in ventral view, illustrating variations in size of body and in shape and structure of left sulral list. X JMO. 1, 2, '.i, /), 6, 7, from Station 47:37 (:300-0 fathoms); 4, from Station 47:37 (100-0 fathoms); 8, from Station 4730 (300-0 fathoms);9, 11, from Station 4740 (300-0 fat horns); 10, from Station 4695 (300-0 fat Ik mis). connected by transitions. Both the cingular lists may have relatively few (Plate 16, fig. 2) or numerous (Schiitt, 1893, fig. 82) radial ribs. The development of the anastomoses connecting these ribs also varies; compare our Plate 17, figure 3, SYSTEMATIC ACCOUNT. 525 and Schiitt's (1893) figure 82. The left sulcal list, which sometimes ends at, some- times ventrally to the antapex of the body, varies in size, shape, and structure. In some specimens it is comparatively narrow (Figure 77:8), in others wide (Figure 77: 10, 11). Its two lobes may be short (Figure 77: 8, 9), or long and nar- row (Figure 77:5, 11). The concavity between these lobes may be relatively shallow (Figure 77:9) or deep (Figure 77:5), narrow (Figure 77:2) or wide (Figure 77 : 6). The dorsal margin of this list is nearly straight in some specimens (Figure 77: 1), but it may be gently concave (Figure 77: 6), moderately convex (Plate 17, fig. 3), or strikingly sigmoid (Figure 77: 3). The number of the radial ribs behind the fission rib ranges from two to five (Figure 77: 2, 10), and their shapes also vary. A submarginal rib may be present between the two lobes (Fig- ure 85 : 3) ; and the reticulation in these lobes shows various degrees of develop- ment. The eleven specimens represented by Figure 77, Figure 85:3, Plate 16, figure 2, 4, and Plate 17, figure 3, illustrate the range of variation exhibited by this species in the material of the Expedition. Unfortunatelj', we are not able to give the relative frequency of these various forms, since only twenty of our speci- mens were subjected to a careful study. Co7nparisons: — In spite of the fact that Schiitt (1893) figures the type specimen of Ornithocercus splendidus from the ventral and posterior sides only and our knowledge of its structure thus is quite incomplete; and in spite of the apparent striking difference in size (114.0 jx as compared to 37.7-49.8 n) between this specunen and the specimens from the material of the Expedition assigned in the present paper to this species, this allocation of ours may be regarded as faii'ly certain. The dimensions of Schiitt's (1893) specimens are from a single specimen, and are based on his stated magnification (250 diameters). If we utilize another magnification (650 diameters) based on his figures, the dimensions of his specimen will fall within ours. The specimen figured by Murray and Whitting (1899, pi. 32, fig. 1) agrees closely with some of our specimens except in size (the length of its body being 87.2 m as compared to 37.7-49.8 m) and in the fact that the pos- terior rib of its left sulcal list has a short, reticulated branch near the base. A similar shift in the magnification cited by Murray and WTiitting (1899) from 310 to 620 brings his specimen within our dimensions. The species, as conceived in the present paper, is strikingly variable. We have considered it most advisable to treat these various forms as belonging to a single systematic unit for the following reasons. (1) Striking differences in size, if such exist, are characteristic of several species of this genus, and small specimens are known sometimes to arise from large by rapidly succeeding fissions (Plate 18, 526 THE DINOPHYSOIDAE. fig. 2, 7). Our large specimens evidently were in the megacytic stage. {2) The structure of the cingular lists, and the size, shape, and structure of the left sulcal list are subject to pronounced variations in other species of this genus, e.g., in Orniihocercus steini and 0. thurni. Furthermore, these variations evidently are partly due to differences in age and to adaptive regulation. (3) The variations are apparently continuous, with the exception of those in size. Ornithocercus splendidus occupies a rather isolated position; indeed, IVIurray and Whitting (1899, p. 332) established it as the sole member of a new subgenus, Paraschiittia. We do not know which of the other known species of this genus is its closest relative. It maybe mentioned, however, that 0. heteroporus resembles it in having a two-lobed left sulcal list usually ending at or somewhat ventrally to the antapex of the body. 0. splendidus differs strikingly from 0. heteroporus in its larger size and in ha\'ing the cingular list very wide and structurally highly differentiated. The similarity between 0. splendidus and 0. quadratus f. inler- media should also be observed. Synonymy: — This species was established by Schiitt (1893) under the name of Ornithocercus splendidus, a specific name also used by most of the other in- vestigators who have treated this form. Schiitt's (1896) name 0. splendens, evi- dently due to a slip of the pen, was applied to this species by Graf (1909). In Doflein (1909, 1911, 1916) and Oltmanns (1922) this form is figured as 0. mag- nificus, a typographical error. Murray and WTiitting (1899), Lemmermann (1899a, 1901a), Cleve (1901c, 1902b), Schroder (1906a), and Gadeceau (1909) use Histioneis as the generic name of this species. Reproductions of Schiitt's (1893) figures of the type specimen are found in Schiitt (1896), Doflein (1909, 1911, 1916), Theel (1909), Nathansohn (1910b), Steuer (1910, 1911), Effenberger (1911), Stiasny (1913), Oltmanns (1922), and France (1923). Reproductions of Murray and Wliitting's (1899) figures are given by Gadeceau (1909), Hjort (1911), and Gran (1912a, b). Occurrence: — Ornithocercus splendidus is recorded at thirty-five of the 127 stations. There are 2, 1, 8, 8, 11, and 5 stations on the six lines of the Expedition. Of these thirty-five stations, one (4583) is in the California Current; one (4587) is in the Mexican Current; one (4613) is in the Panamic Area; one (4676) is in the Peruvian Current; five (4689, 4691, 4695, 4697, 4699) are in Easter Island Eddy; twenty-two (4679, 4681, 4683, 4685, 4687, 4701, 4702, 4705, 4707, 4709, 4717, 4721, 4722, 4724, 4728, 4730, 4732, 4734, 4736, 4737, 4739, 4740) are in the South Equatorial Drift; three (4742, 4743, 4540) are in the South Equatorial Current; and one (4541) is in the Equatorial Counter Current. At four stations (4702, SYSTEMATIC ACCOUNT. 527 4743, 4540, 4541) the species was taken in surface waters only; at one station (4737) in a haul from 100-0 fathoms as well as in a haul from 300-0 fathoms; at eight stations (4681, 4689, 4701, 4721, 4724, 4728, 4732, 4740) in hauls from both 800-0 fathoms and 300-0 fathoms. All the remaining records refer to hauls from 300-0 fathoms only. The temperature range of these thirty-five stations at the surface was 68°- 83°; the average was 75.8°. At the four stations in the surface catches of which this species was found, the surface temperature ranged from 73° to 80°; the aver- age was 77.5°. Figure 78. — Occurrence of Ornithocercus splendidus Schtitt. Large, solid circles indicate records from vertical hauls; squares, records from surface hauls; small, solid circles, stations at which thi.s species was not found; small, open circles, stations from which no plankton catches were examined. At one of the surface stations (4541) the frequency was 1 % ; at the remaining surface stations it was less. For the catches from 100, 300, or 800 fathoms to the surface the records of frequency are as follows: 4% at one station (4724) ; 3% at two stations (4681, 4739) ; 2% at five stations (4689, 4697, 4737, 4740, 4742) ; 1 % at eight stations (4683, 4685, 4687, 4695, 4701, 4722, 4728, 4732) ; at the remain- ing stations the frequency is less than \%. The type specimen of this species (Schiitt, 1893) was probablj- taken in the tropical or in the subtropical region of the Atlantic Ocean. Later investigators have sho\\n it to be widely distributed. The following authors have found it in the Atlantic, between lat. 41° N. and lat. 21° S. : — Hensen (1895, 1911), Murray 528 THE DINOPHYSOIDAE. and Whitting (1899), Lemmermann (1899a), Cleve (1901c, 1902b), Karsten (1906), Stiiwe (1909), and Griif (1909). It has been recorded from the Gulf of Naples by Lindemann (1924, 1925). In the Red Sea and the Gulf of Aden it was found by Karsten (1907); in the Indian Ocean, by Schroder (1906a), Karsten (1907), and Graf (1909). Most of the records referred to in the last paragraph were from surface hauls. The only records of closing-net hauls published as yet are the following by Karsten (1907) : — Station 182, lat. 10° 8' S., long. 97° 14' E., 200-100 m., dead; Station 229, lat. 2° 38' S., long. 63° 37' E., 600-400 m. and 200-20 m., no statement as to condition of specimens; Station 208, lat. 9° 6' N., long. 53° 41' E., 80-67 m., living; 105-88 m., dead. According to the authors mentioned in the next but last paragraph, this species occurs in waters of the following temperatures and salinities. Number and Number and Mean of Mean of Temperature Range Observations Salinity Range Ob.'°-84.2° Of the investigators who have contributed to our knowledge of the distribu- tion of this species only Schiitt (1893) and Murray and Whitting (1899) give figures by means of which their determinations may be checked. Ornithocercus splendidus is eupelagic and widely distributed in tropical and subtropical seas. In most cases it is rare, but sometimes (Graf, 1909, p. 173) it is fairly common. In the Eastern Pacific it occurs in the whole area investigated by the Expedition, but it is strikingly rare in the Mexican Current, Panamic Area, and Peruvian Current, and is not recorded from the northern portion of the Cali- fornia Current ; however, it occurs occasionally as far to the north as at San Diego. Another striking feature about the distribution of this species, according to our data, is its rare occurrence in surface waters ( four stations out of 35) . 2. Magnificus Group. All the species belonging to this group were found in the material of the Expedition. They have been treated in this paper in the following order, indicative of relationships: — Ornithocercus magnificus s. sir., 0. thurni, 0. steini s. sir., 0. orbiculatus, 0. quadratus, and 0. caroUnae. SYSTEMATIC ACCOUNT. 529 Ornithocercus magnificus Stein, s. sir. Schiitt Plate 16, fig. 3. Figure 79, 80 Ornithocercus magnificus Stein, 1883, pariim, \A. 23, fig. 1, 2. Chun, 1886, p. 58. Dadat, 1888, p. 99. ScHUTT, 1895, p. 112, 136, 138, pi. 4, fig. 21:9, pi. 5, fig. 21: 1, 3-8; 1895, pi. .5, fig. 21:2(?); 1896^ p. 29, fig. 41 A; 1899, p. 14, 16, pi. 6, fig. 7, 12(?); 1900a, p. 9, 16, 18, fig. 8-10. Hen.sen, 1895, p. 189 1911, p. 137, 138, 163, 393, tab. 15. Cleve, 1897a, iiariim, p. 26, fab. 1. Ostenfeld, 189Sa, p. 428 1915, p. 7. Ostenfeld & Schmidt. 1901, p. 173. Schmidt, 1901, p. 138. Lohmann, 1902, p. 53 190Sa, p. 168, 169; 1920, p. 484, 492, 539, 545, 547, 565. Entz, 1902a, p. 122. Zac'h.^rias, 1906, p 509, 518, 530, 531, 532, 534, 536, 552, 557, 566, 569; non 1906a, p. 274, fig. 7; non 1907, fig. 14; non 1911, fig. 51. Karsten, 1906, p. 180, 181, 182, 183, 187, 189, 190, 191, 200, 201, 202, 204, 205, 207 1907, p. 236, 238, 240, 241, 243, 245, 246, 247, 248, 249, 250, 253, 254, 255, 256, 258, 259, 260, 263 265, 267, 268, 269, 270, 271, 274, 275, 276, 277, 279, 283, 289, 290, 291, 292, 295, 297, 298, 300, 304 305, 308, 311, 313, 314, 315, 316, 317, 321, 322, 323, 324, 325, 330, 331, 332, 333, 335, 3.36, 337, 338, 339, 340, 341, 342, 343, 344, 345, 346, 347, 348, 349, 350, 352, 353, 354, 355, 356, 357, 358, 427, 440 445, 450, 471. Schroder, 1906a, p. 321, 322, 323, 325, 326, 329, 335; 1911, p. 15, 18. KofoiD; 1907a, p. 205, 206; 1911, p. 237. Ok.\mura, 1907, p. 132, partimCf); 1912, p. 20. Doflein, 1909, p 465, fig. 417: 7 {.non 417: 8); 1911, p. 531, fig. 478; 7 {non 478: 8); 1916, p. 436, fig. 396: 7 (non 396: 8) Graf, 1909, p. 173. Schiller, 1912, p. 27; 1912a; 1914, p. 14. Gran, 1912b, p. 328, 330, fig. 235 Faria & CuNHA,1917, partim,p. 79. Oltm.\nns, 1922, fig. 38: 7 {non fig. 38: 8). Jorgensen, 1923 p. 35, 36, 37, 44, fig. 48. Lindemann, 1924, p. 7 (non p. 5). non Ornithocercus magnificus Butschli, 1885, pi. 55, fig. 7. Chdn, 1903, p. 75, fig. b. Whitelegge, 1891, p. 185. Fkicke, 1902, p. 55. Willet & Hickson, 1909, p. 184, fig. 8. Hjort, 1911, p. 367, fig. 12: 4. France, 1923, fig. 11:2. Histioneis magnifica Mvruxy & Whitting, 1899, partim, p. 332, 333, tab. 1-9. Entz, 1902b, p. 94; 1905, p. 112. P.wiLLARD, 1905, partun, p. 60, 81, 102; 1916, partim, p. 61. Meunier, 1910, p. 59. IssEL, 1921, p. 17. FoHTi, 1922, partim (?), p. 113, 192, 209, 213, pi. 7, fig. 124. Histioneis magnificus Lemmerm.Wn, 1899a, partim, p. 319, 320, 323, 330, 331, 332, 374; 1901a, partim, 376; 1905a, p. 37 (as H. magnifica). Cleve, 1900b, partim, p. 1031; 1901a, p. 16; 1901c, p. 250 (a.s H. magnifica); 1902b, p. 32 (as H. magnifica); 1903b, p. 345 (as H. magnifica). Cleve, Ekm.\n, & Pettersson, 1901, partiin, p. 16, tab. 2. non Histioneis magnificus LEMnrERM.u' Ornithocercidae, the section on the sub- divisions (p. 494). Occurrence: — OrniOiocercus thurni is recorded at se\'enty-four of the 127 stations. There are 18, 10, 11, 12, IG, and 7 stations on the six lines of the Expedi- tion. Of these seventy-four stations, one (4583) is in the California Current ; eleven (4587, 4588, 4590, 4592, 4594, 4596, 4598, 4600, 4604, 4605, 4607) are m the Mexican Current; twelve (4609, 4613, 4615, 4617, 4619, 4623, 4631,4634, 4637, 4638, 4640, 4644) are in the Panamic Ai-ea; four (4646, 4647, 4664, 4671) are in the Peruvian Current ; eight (4689, 4691, 4692, 4695, 4696, 4697, 4699, 4700) are in the Easter Island Eddy; four (4713, 4714, 4715, 4716) are in the Galapagos Eddy; twenty-nine (4679, 4680, 4681, 4682, 4683, 4684, 4685, 468(), 4701, 4705, 4709, 4717, 4719, 4721, 4722, 4724, 4727, 4728, 4730, 4731, 4732, 4733, 4734, 4735, 4736, 4737, 4739, 4740, 4741) are in the South Equatorial Drift; three (4742, 4743, 4540) are in the South Equatorial Current; and two (4541, 4542) are in the Equatorial Counter Current. There are thirty-six records from the surface (Stations 4583, 4588 [Salpa], 4590, 4592, 4596, 4600, 4604, 4607, 4615, 4617, 4619, 4623, 4631, 4638, 4640, 4644, 4646, 4664 [Salpa], 4680, 4682, 4684, 4686, 4692, 4696, 4700, 4714, 4716, 4727, 4731, 4733, 4735, 4741, 4743, 4540, 4541, 4542). At thirty-one of these thirty-six stations the species was taken in surface waters only; at five stations (4583, 4590, 4617, 4638, 4646) in hauls from 300-0 fathoms as well as at the surface ; at Stations 4619 and 4646 in Salpa stomachs as well as in surface hauls. At one station (4737) it is recorded from 100-0 fathoms and 300-0 fathoms; at one station (4713) from 150-0 fathoms and 300-0 fathoms; at four stations (4681, 4701, 4715, 4724) from 800-0 fathoms and 300-0 fathoms; at two stations (4647, 4728) from 800-0 fathoms only. All the remaining records refer to hauls from 300-0 fathoms only. The species was taken also in surface waters in Acujiidco Harlnir, off the Mexican Current. This stalion is not included in the 127 stations mentioned above. SYSTEMATIC ACCOUNT. 549 The temperature range of these seventy-four stations at the surface was 66°-85°; the average was 76.8°. At the thirty-six stations in the surface catches of which the species was found, the surface temperature ranged from 68° to 84°; the average was 77.4°. At Acapulco it was 83°. For the surface catches the following frequencies are recorded: — 37% at one station (4619, Salpa); 23% at one station (4619); 10% at two stations (4607, 4615); 7% at one station (4604); o%: at two stations (4617, 4696); 2% at nine stations (4592, 4596, 4600, 4631, 4638, 4682, 4714, 4716, 4540); 1% at three sta- tions (4590, 4640, 4733) ; in the remaining cases the frequency is less than 1 %. Figure 82. — Occurrence of OrnithocerctiB thurni (Scliniidt). Large, solid circles indicate records from vertical hauls; squares, records from surface hauls; triangles, records from both vertical and surface hauls; small, solid circles, stations at which this species was not found; small, open circles, stations from which no plankton catches were examined. For the catches from 100, 150, 300, or 800 fathoms to the surface the records of frequency are as follows: — 16% at one station (4613) ; 10% at one station (4713) ; 7% at one station (4715); 6%, at one station (4638); 4% at five stations (4598, 4609, 4683, 4730, 4740) ; 2% at eight stations (4617, 4634, 4681, 4695, 4701, 4717, 4734, 4737) ; 1 % at eleven stations (4647, 4679, 4689, 4691, 4699, 4705, 4728, 4732, 4736, 4739, 4742) ; at the remaining stations the frequency is less than 1 %. For the catch made in Acapulco Harbor a frequency of less than 1 % is recorded. Schiitt (1900a) does not give any information as to the type locality. The following investigators have found this species in the Atlantic between lat. 45° N. 550 THE DINOPHYSOIDAE. and lat. 30°S: — Cleve (1897a), Murray and Whitting (1899), Stiiwe (1909), Faria and Cunha (1917), and Jorgensen (1923). In the Caribbean Sea it was found by Murray and Whitting (1899). Jorgensen (1923) found it at a few sta- tions in the eastern as well as in the western Mediterranean. Other investigators who have found it in the Mediterranean are as follows: — Pouchet (1883) and Pavillard (1905, 1916) in the Gulf of Lyons; and Entz (1902b, 1905) in the Adri- atic Sea. Ostenfeld (1915) recorded it from the Malay archipelago; Whitelegge (1891) from Sydney, Australia; Stein (1883) "aus der Siidsee"; Lemmermann (1899a) between the island of Laysan and Hawaii. Most of the records referred to in the last paragraph were from surface hauls. No records of closing-net hauls have been published as yet. According to Murray and WTiitting (1899), the species occurs in temperatures ranging from 57.0° to 86.0°; mean of 125 observation.s, 75.5°. Of the investigators who have contributed to our knowledge of the dis- tribution of this species. Stein (1883), Whitelegge (1891), Lemmermann (1899a), Entz (1902b, 1905), Stuwe (1909), Ostenfeld (1915), and Jorgensen (1923) show by figiu'es, references to figures, or otherwise that their data refer to Ornithocercus ihurni as conceived in the present paper. Cleve (1897a), Murray and AMiitting (1899), Pavillard (1905, 1916), Okamura (1907), and Faria and Cunha (1917), on the other hand, show that they have the same broad conception of 0. magnifi- cus as Stein (1883), and their data of distribution thus may refer either to any or to all of the three species 0. magnificus, 0. Ihurni, and 0. steini s. str. With regard to the data of distribution of Pouchet (1883), see the section on synonymy (p. 548); compare also the account of 0. magnifiais (p. 534). Ornithocercus ihurni is eupelagic and widely distributed in tropical, sub- tropical, and warm-temperate waters. While in tropical seas this species is about as common as 0. magnificus, it is rare in warm-temperate and subtropical waters {e.g., in the Mediterranean), when compared with that species. In the Eastern Pacific it occurs in the whole area investigated by the Expedition, but it is strik- ingly rare in the Peruvian Current and is not recorded from the northern portion of the California Current ; however, it occurs occasionally as far to the north as at San Diego, California. Most of the record stations of high frequencies are in or near the Panamic .\rea, in the Galapagos Eddy, and in the Easter Island Eddy. This species is relatively frequent in surface waters ; it should be observed that the highest records of frequency are from surface catches. SYSTEMATIC ACCOUNT. 551 Ornithocerctjs steini Schiitt, s. str. Plate 16, fig. 1. Figure 83, 84 Ornithocercus tnagnificus Stein, 1883, parlim, pi. 23, fig. 3. Omithocercus steini Schutt, 1900a, partim, fig. 5, 6. Ornithocercus serralus Kofoid, 1907a, p. 206, pi. 15, fig. 93. Kofoid & INIicheneh, 1911 , p. 300. Faria & CuNHA, 1917, p. 79. JciHGENSEN, 1923, p. 38, 43, 44, fig. 52. Ornithocercus magnificus var. 6 Stuwe, 1909, p. 254, 275, 288. Diagnosis: — Subcircular in lateral outline. Cingular lists: anterior, 0.49- 0.73, posterior 0.45-0.72 the greatest depth of body; anterior with seven to nine complete ribs, sometunes branched distally, and with some short, marginal ribs; posterior with ten to nineteen complete and often .simple ribs. Left sulcal list ends on dorsal side of body, at or near posterior cingular list; with four nearly equidistant, narrowly to fairly broadly rounded posterior lobes, viz., the postero- ventral, midventral, middorsal, and posterodorsal lobes; width at posteroventral lobe 0.53-1.12, at midventral lobe 0.51-1.04, at middorsal lobe 0.50-1.00, at posterodorsal lobe 0.48-0.97 the greatest depth of body; typically with five ribs behind fission rib and with reticulum in each lobe ; submarginal rib may be present ; a-rib ends near middle of dorsal margin of list ; each of the b- to c-riljs ends at the vertex of one of the four lobes. Length of body, 54.8-67.6 n. Widely distributed in tropical, subtropical, and warm-temperate seas. Description: — This is a fairly large species, subcircular in lateral view, deep- est in or near the middle. The ratio between the length and the depth of the body is 0.92 (0.82-1.00) : 1. The longitudinal axis may be perpendicular to the posterior cingular list, but usually it is deflected posterodorsally at 7° (1°-15°). The epitheca is 0.61 (0.54-0.67) as deep as the hypotheca, slightly convex or nearly flat, highest in or near the center or dorsally, and tilted ventroposteriorly at 5°-15°. The transverse furrow is moderately or but slightly concave ; its dorsal width is 0.40 (0.34-0.53) the greatest depth of the body; its ventral width is 0.47 (0.31-0.62) its dorsal width. The posterior cingular list is 0.36 (0.29-0.46) the length of the body from the apex. The dorsal, posterior, and ventral margins of the hypotheca are well rounded and confluent; or any of them may be slightly flattened. The anterior cingular list is 0.49-0.73 the greatest depth of the body and has an anterior inclination of 40° (30°-50°) ; in our specimens its width is 0.57 (0.49- 0.67); it has on each valve seven to nine complete, radial ribs, which, when fully developed, may be furnished distally with candelabra-like branches (Kofoid, 1907a, pi. 15, fig. 93) ; between these main ribs there is a varying number of short, 552 THE DINOPHYSOIDAE. marginal radial ribs, which sometimes are connected by anastomoses. The poste- rior cingular list is 0.56 (0.45-0.72) the greatest depth of the body and is inclined anteriorly at (55° (50°-85°) ; it has on each vah'c ten to nmeteen complete, sub- equidistant ribs, some of which may be branched proximally. The right sulcal list is comparatively large, of subuniform width throughout the greater portion of its length and rounded posteriorly; it appears always to end near the d-rib of the left sulcal list; its average width is 0.16-0.20 the greatest depth of the body; at least in some specimens it has a few ribs (Figure 83: 5; Plate 16, fig. 1). The left sulcal list ends on the dorsal side of the body, at or near the posterior cingular list ; the maximum length of this distance is about 0.33 the greatest depth of the body. The list is characterized by four (in exceptional cases by five. Figure 83: 7) nearly equidistant, narrowly to fairly broadl.y rounded posterior lobes, which we desig- nate, according to their positions, the posteroventral, the midventral, the mid- dorsal, and the posterodorsal lobes. In most specimens these four lobes are well marked, but they may be but slightly indicated (Figure 83: 10) or almost absent (Figure 83: 11). In our specimens the width of this list at the anterior main rib is 0.48 (0.42-0.57), at the fission rib 0.54 (0.45-0.67), at the posteroventral lobe 0.69 (0.53-0.88), at the midventral lobe 0.67 (0.51-0.85), at the middorsal lobe 0.65 (0.50-0.83), and at the posterodorsal lobe 0.62 (0.48-0.79) the greatest depth of the body. In the specimen figured by Jorgensen (1923, fig. 52), which has this list somewhat wider than in any other figured specimen, the corresponding values are: 0.55, 0.81, 1.12, 1.04, 1.00, and 0.97. The ventral margin of this list usually forms a fairly distmct lobe at the fission rib (Figure 83: 2), but it may be almost straight, gently convex (Figure 83: 10, 11), or somewhat irregular (Figure 83 : 5) . The dorsal margin is gently con\-ex in most specimens, but it may be somewhat angular at the end of the o-rib (Figure 83: 10, 12). Between the an- terior main rib and the fission rib this list often has one to three ribs or an irregular reticulum (Figure 83: 11). Behind the fission rib it typically (Figure 83: 2) has five complete radial ribs, the distal ends of which may or may not be connected by a very weak submarginal rib, reticulations in the four lobes as well as dorsally to the a-rib and at the distal end of the fission rib, and in some cases it may have a few short, irregular ribs between the bases of the main ribs (Figure 83: 1 ; Kofoid, 1907a, pi. 15, fig. 93). Of the five main radial ribs behind the fission rib, the a-rib (i.e., the dorsal rib) ends near the middle of the dorsal margin of this list, the 6-rib ends at the vertex of the posterodorsal lobe, the c-rib at the vertex of the middorsal lobe, the d-rib at the vertex of the midventral lobe, and the e-rib at the vertex of the posteroventral lobe. In our specimens the lengths of these ribs are SYSTEMATIC ACCOUNT. 553 Figure 83. — Ornithocercus steini Schiitt, s. sir., right lateral view, illu.strating variation.s in shape and structure of left sulcal list. X .340. 1, from Station 4742 (300-0 fathoms); 2, from Station 4722 (300-0 fathoms); 3, from Station 4607 (300-0 fathoms); 4, 5, 6, 8, from Station 4607 (surface); 7, from Station 4619 (surface ) ; 9, from Station 4721 (300-0 fathoms) ; 10, from Station 4605 (300-0 fathoms) ; 11, from Station 4638 (300-0 fathoms); 12, from Station 4637 (300-0 fathoms). 554 THE DINOPHYSOIDAE. as follows, if the a-rib is measured in a straight line to the margin of the list, and the b- to e-ribs are measured in straight lines to the points where they join the submarginal rib or, in the case of the absence of the last rib, to the beginnmg of the reticulations of the lobes: — the a-rib is 0.37 (0.28-0.53), the 6-rib 0.52 (0.39- 0.66), the r-rib 0.56 (0.41-0.74), the d-rih 0.57 (0.45-0.78), and the e-rib 0.59 (0.45-0.77) the greatest depth of the body. In the specimen figured by Jorgensen (1923, fig. 52), these ribs are somewhat longer; compare the measurements of the width of this list given above. In no case have less than five main radial ribs been found; but in a few specimens six or seven of them have been recorded; two rf-ribs may be present, and in one specimen (Figure 83: 10) a further rib was developed next to the e-rib. An incomplete rib may be present between the b- and c-ribs, between the c- and d-ribs, and between the d- and c-ribs. In young specimens the radial ribs are simple, but soon they develop short branches, which probably in- crease in number and complexity with age. The central portion of the hypotheca is areolate and porulate ; the rest of the theca is porulate (Plate 16, fig. 1 ; Kofoid, 1907a, pi. 15, fig. 93). There is one pore in each areole. The areoles are of moderate and subuniform size. Phaeosomes of a rich greenish-yellow or olive-yellow color are often found in the transver.se furrow; sometimes a few of these bodies are found attached to the left sagittal list. Speci- mens of normal and healthy appearance have been found completely lacking phaeosomes. The cytoplasm is of a reddish or geranium-pink tint and lacks every trace of chromatophores. The dimensions of seventeen of our specimens and of the specimens figured by Stein (1883, pi. 23, fig. 3), Schiitt (1900a, fig. 6), and Jorgensen (1923, fig. 52) were measured. Dimensions: — Our specimens: Length of body, 54.8-67.6 m (average, 62.4 p). Greatest depth of body, 62.4-77.8 m (average, 67.8 m). The size of the type speci- men (Schiitt, 1900a, fig. 5) is unknown. The length of the specimen figured by Stein (1883, pi. 23, fig. 3) is also unknown. However, according to Stein's (1883) information about the magnifications of his figures given in an introductory re- mark to the explanations of the plates, it was somewhere between 42.0 and 64.5 m- Schiitt's (1900a, fig. 6) specimen was 65.0 m long and 66.6 m deep; Jorgensen's (1923, fig. 52), 64.5 m long and 67.6 m deep. Variations: — The great majority of the recorded specimens show a fairly pronounced constancy in all characters. A few are subject to rather striking variations in the shape and structure of the left sulcal list. In some specimens the number of radial ribs in this list is six (Figure 83: 5, 7) or seven (Figure 83: 10), SYSTEMATIC ACCOUNT. 555 or there may be, besides the normal number of five complete radial ribs, one more or less incomplete rib (Figure 83:3, 4, 6, 11). Sometimes (Figure 83:3, 4) the radial ribs are simple, but usually they are furnished with short, more or less de- veloped branches. At least the c- and d-ribs may be branched proximally (Figure 83 : 8) . In one of our specimens (Figure 83 : 9) the d-rib was very aberrant, its distal portion forming an irregular reticulum (due to regeneration following local injury?). The subniarginal rib may be absent (Figure 83: 1-3) or present (Figure 83: 4-11); when present, it usually, but not always (Figure 83: 11), is exceedingly weak. Sometimes (Figure 83: 1; Jorgensen, 1923, fig. 52) the four lobes of this list are very pronounced; sometimes (Figure 83:2) they are moderately devel- oped; or they may be but slightly indicated (Figure 83: 10) or almost absent (Figure 83: 11). The variations in the development of the branches of the radial ribs are presumably due to age. The same appears to be true in regard to the subniarginal rib. On the other hand, we do not know whether or not the estab- lished variations in the number of the ribs and in the development of the marginal lobes are due wholly to this cause, to resorption in adaptation to flotation, or to regulation following fission. It should be mentioned in this connection that in some recently divided specimens, probably belonging to this species, the left sulcal list had five well-developed and simple radial ribs behind the fission rib, and its posterior margin was almost evenly rounded, i.e., without lobes. The twelve specunens of Figure 83, and the two represented by Plate 16, figiu-e 1, and by Kofoid (1907a, fig. 93) illustrate the range of variation exhibited by this spe- cies in the material of the Expedition. The great majority of our specimens (125, i.e. 74.0% out of the 169 that were carefully examined) agreed with the specimen represented by Figure 83: 2; sixteen specimens (9.5%) agreed with Figiu-e 83: 4; seven (4.1%,) with Figure 83: 3,; five (3.0%o) with Figure 83: 7; two (1.2%,) with Figure 83: 5; two (1.2%c,) with Figure 83: 6; two (1.2%) with Figure 83: 2, but they had somewhat narrower left sulcal hsts with slightly less marked lobes; one (0.6%) with Figure 83: 3, but without the short rib between the b- and c-ribs; one (0.6%) with Figure 83: 3, but with a short rib between the c- and d-ribs instead of between the b- and c-ribs; and of each of the forms represented by Figure 83: 1, 8, 9, 10, 11, 12, by Plate 16, figure 1, and by Kofoid's (1907a) Plate 15, figure 93, only one specimen was found. Comparisons: — The type specimen as figured by Schiitt (1900a, fig. 5), as well as the specimen represented by Stein's (1883) Plate 23, figure 3, although young and thus not fully differentiated, are typical members of this species as con- ceived in the present paper. Schutt's (1900a) figure 6 is somewhat less repre- 556 THE DINOPHYSOIDAE. sentative, having but slightly indicated lobes in the sulcal list, but its present allocation may be considered as fairly certain. Jorgensen's (1923, fig. 52) speci- men is an unusually well-developed example of this species. The typical members of this species (Figure 83: 2) are characterized espe- cially by their left sulcal list, which is four-lobed, with or without an exceedingly fine submarginal rib, and with five complete radial ribs behind the fission rib. Of these five radial ribs the dorsal one, the o-rib, ends near the middle of the dorsal margin of this list, while each of the remaining four ends at the vertex of one of the four lobes. Besides these typical specimens several more or less aberrant ones have been assigned by us to this species. Our reasons for including these aberrant forms are the relatively great A-ariability ui several of the other species of this genus and the fact that the extremes appear to be connected with the typical ones by fairly complete series of intermediate specimens. No specimens have l^een recorded as yet which are distinctly intermediate between Ornithocercus steini and any of the other species of this genus. Regarding the specimen represented by Figure 81: 11, of 0. thurni, see the last species, the section on comparisons. The closest relative of 0. sleini appears to be 0. thurni. The former species is easily distinguished from the latter by having four, instead of three lobes in the left sulcal list, by having the o-rib end at about the middle of the dorsal side of this list, and by haA-ing the fe-rib end at the vertex of the dorsal lobe. Synonymy: — This species was established by Kofoid (1907a) under the name of Ornithocercus serratus but had been figured previously by Stein (1883, pi. 23, fig. 3) and by Schiitt (1900a, fig. 5, 6). Stein (1883, pi. 23, fig. 1-6) figured under the name of 0. rnagnificus three different species. Schiitt (1900a), who was the first to subdivide 0. rnagnificus Stein, applied the name 0. magnificv^ s. str. to Stein's (1883) Plate 23, figure 1, and the new name 0. steini to Stein's (1883) Plate 23, figiu'e 4. The species to which the last figiu'e belongs had, however, previously been established by Schmidt (1888, pi. 144, fig. 59-61) as Parelion thurnii. The name Ornithocercus steini Schiitt should therefore have been dis- carded as a synonym of 0. thurni (Schmidt), had it not been for the fact that out of the three specimens figured by Schiitt (1900a, fig. 5-7) as 0. steini two are not referable to this species but to one later established by Kofoid (1907a) as 0. serratus. Under these circumstances the name 0. steini s. str. has to be maintained for these two specimens of Schiitt (1900a) and 0. serratus Kofoid must be rejected as a synonym. Stiiwe (1909) recorded 0. steini s. str. as 0. rnagnificus var. b. Later this species was figured by Jorgensen (1923, fig. 52) as 0. serratus. SYSTEM.\TIC ACCOUNT. 557 Occurrence: — Ornithocercus steini is recorded at sixty-eight of the 127 sta- tions. There are 16, 12, 1, 10, 19, and 10 stations on the six Unes of the Expedi- tion. Of these sixty-eight stations, one (4583) is in the California Current; ten (4587, 4590, 4592, 4598, 4600, 4604, 4605, 4607, 4545, 4546) are in the Mexican Current; fourteen (4609, 4611, 4613, 4615, 4617, 4619, 4623, 4634, 4635, 4637, 4638, 4639, 4640, 4644) are in the Panamic Area; five (4646, 4647, 4648, 4650, 4659) are in the Peruvian Current; one (4692) is in the Easter Island Eddy; four (4713, 4714, 4715, 4716) are in the Galapagos Eddy; twenty-seven (4701, 4704, 4705, 4709, 4711, 4712, 4717, 4718, 4719, 4720, 4721, 4722, 4723, 4724, 4727, 4728, 4729, 4730, 4731, 4732, 4733, 4734, 4736, 4737, 4739, 4740, 4741) are in the South Equatorial Drift; tlu-ee (4742, 4743, 4540) are in the South Equatorial Current; two (4541, 4542) are in the Equatorial Counter Current; and one (4543) is in the North Equatorial Current. There are thirty-seven records from the surface (Stations 4590, 4592, 4600, 4604, 4607, 4611, 4615, 4617, 4619, 4623, 4635, 4638, 4639, 4640, 4644, 4646, 4648, 4692, 4704, 4712, 4714, 4716, 4718, 4720 [Salpa], 4723 [Salpa], 4727, 4729, 4731, 4733, 4741, 4743, 4.540, 4541, 4542, 4543, 4545, 4546). At thirty-three of these thirty-seven stations the species was taken in surface waters only; at four stations (4590, 4617, 4638, 4646), in hauls from 300-0 fathoms as well as at the surface; at Stations 4619, 4712, and 4741, in Salpa stomachs as well as in surface hauls. At one station (4737) it is recorded from 100-0 fathoms and 300-0 fathoms; at one station (4713), from 150-0 fathoms and 300-0 fathoms; at five stations (4715, 4721, 4724, 4728, 4732), from 800-0 fathoms and 300-0 fathoms; at two stations (4647, 4717), from 800-0 fathoms only. All the remaining records refer to hauls from 300-0 fathoms only. The temperature range of these sixty-eight stations at the surface was 69°- 85°; the average was 77.5°. At the thirty-seven stations in the surface catches of which this species was found, the surface temperature ranged from 71° to 84°; the average was 77.8°. For the surface catches the following frequencies are recorded: — 14^ at one station (4607) ; 13% at one station (4543) ; 10% at one station (4545) ; 6% at one station (4600); 4% at four stations (4615, 4635, 4714, 4741); 3% at two stations (4640, 4546); 2% at five stations (4619, 4648, 4692, 4712, 47.33); 1%. at eleven stations (4590, 4604, 4617, 4638, 4639, 4716, 4720, 4729, 4731, 4.541, 4542) ; in the remaining cases the frequency Ls less than 1 %. For the catche.s from 100, 150, 300, or 800 fathoms to the surface the records of frequency are as follows: — 15% at one station (4598); 12%, at one station (4613); 11%, at one station (4721); 8% at two stations (4605, 4617); 7%, at two stations (4638, 4722); 3% at two stations 558 THE DINOPHYSOIDAE. (4634, 4637); 2% at two stations (4715, 4719); 1% at five stations (4711, 4724, 4732, 4736, 4737) ; at the remaining stations the frequency is less than 1 %. The type locality is unknown, but probably somewhere in the Atlantic Ocean. iStein (1883) figured one specimen " aus der Slidsee" under the name of Ornitho- cercus magnificus. Stiiwe (1909) found this species at two stations in the Atlantic, viz., at lat. 33° 29' N., long. 21° 43' W., and at lat. 4° 3' N. and long. 26° 0' W., in surface waters of 68.5°-82.8°. Faria and Cunha (1917) found it ofT Rio de Janeiro. Jorgensen (1923) recorded it from the Bay of Cadiz and from the eastern FKitTRE 84. — Oppurrenoe of Ornithocercus sleini Sehiitt, s. str. Large, solid circles indicate records from vertical haids; squares, records from surface hauls; triangles, records from both vertical and sur- face hauls; small, solid circles, stations at which this species was not found; small, open circles, stations from which no jilankton catches were examined. portion of the Mediterranean (the Thor Expedition) and stated that "it has been found, rarely, in the Florida Current, (30°-32° N., coast of America, May, 1910), far west of the Straits of Gibraltar (14° W., 29/8 1901) and in the Guinea Current 1/10 1903 (German South Pole Exped.) and fairly common in the southern At- lantic between South Africa and South .\merica 11° W. to 38° W., 31° N. to 17° N. June to August 1911 ('Fram')-" Kofoid (1907a) recorded this species from the Eastern Pacific, viz., from Stations 4613 and 4742 of the Expedition. See also the account of Ornithocercus magnificus. Of the investigators who have contributed to our knowledge of the distribu- tion of this species, Stein (1883), Kofoid (1907a), and Jorgensen (1923) give figures by means of which the determinations may be judged. SYSTEMATIC ACCOUNT. 559 Ornithocercus steini is eupelagic and widely distributed in tropical, subtropi- cal, and warm-temperate seas. In the Eastern Pacific it occurs in the whole area investigated by the Expedition but is strikingly rare in the Peruvian Current and in the Easter Island Eddy and is not recorded from the northern portion of the California Current; however, it occurs occasionally as far to the north as San Diego. Like most of the remainhig species it is relatively frequent in surface waters, and the frequency percentages are on the average somewhat higher in the surface catches than in the vertical catches. Ornithocercus orbiculatus Kofoid and Michener Plate 17, fig. 7 Ornilhocerciis orbiculatus Kofoid & Micheneb, 1911, p. 300. Diagnosis: — Lateral outline subcircular. Anterior cingular list 0.52, poste- rior 0.51 the greatest depth of body; anterior either with about twelve complete ribs and numerous short marginal ribs, or with a number of irregular ribs; poste- rior with about fifteen complete, simple ribs. Left sulcal list ends on dorsal side of body at posterior cingular list; margin subcircular, with small lobe dorsally; with five ribs, corresponding to a-, b-, c-, d-, and e-ribs in related species; a-rib reticulated distally, the remaming ribs irregular; width of hst at a-rib 0.34, at 6-rib 0.46, at c-rib 0.50, at d-rih 0.52, and at e-rib 0.49 the greatest depth of body; submarginal rib absent. Length of body, 66.6 n. Eastern tropical Pacific. Description: — On account of the uncertain status of this species, we refrain from giving the usual elaborate specific description. Dimensions: — Length of body, 66.6 m- Greatest depth of body, 70.4 m- Comparisons: — Our diagnosis is based on the type specimen. This species undoubtedly is very closely related to Ornithocercus steini. Indeed, the only specimen observed as yet may even be an abnormal member of the mentioned species. We maintain 0. orbiculatus tentatively, until further evidence is avail- able. Kofoid and Michener (1911, p. 301) considered the type as "possibly a phase in thecal reconstruction." Occurrence: — Ornithocercus orbiculatus has been recorded at only one of the 127 stations (4617), on the finst line of the Expedition, in the Panamic Area, in surface waters of 78°, with a frequency of less than 1 % (one specimen). 560 THE DINOPHYSOIDAE. FicDKE H'l. — 1, HisHoneix jmtnlii Kofoiil anil Micln'iior, obliciui' ri^lit latoial view of type speci- men. X lt3o. Station 4724 (:i(IO-n fathoms). 2, IlistioncU iiuiriKild Kofoid and Michener, right lateral view of type specimen. X 93;'). Station 4720 (Salpa). 3, Ornilhncernis splemlvhix Schiitt, right lateral SYSTEMATIC ACCOUNT. 561 Ornithocercus quadratus Schiitt Plate 17, fig. 2, 8. Figure 85: 5, 86, 87, 88 ? Dinophysis galea Pouchet, 1883, parlim, p. 426, fig. G:4. Ornithocercus qimdrnius Schiitt, 1900a, p. 5, 10, 17, 26, 27, fig. 1-4, 12, 13. Ostenfei.d & Schmidt, 1901, p. 173. Karsten, 1906, p. 185, 186, 190, 193, 194, 195, 197, 198, 199, 200, 202, 203, 205, 206; 1907, p. 235, 237, 238, 240, 241, 249, 286, 288, 299, 302, 306, 307, 309, 312, 315, 318, .338, 341, 344, 346, 427, 447. Schroder, 1906a, p. .321, 326, 329, 3.33, 335, 3.39; 1909, p. 211; 1911, p. 12. Kofoid, 1907a, p. 206. Schiller, 1912, p. 27; 1912a. Gran, 1912b, p. 328. Jorcen.^en, 1923, p. 37, 38, fig. 50. HisHoneis magnifica Schroder, 1900a, p. 20, pi. 1, fig. 15. Hislioneis quadrata Lemmermann, 1901a, p. 376; 1904, p. 611, 643; 190.5a, p. 37. HisHoneis magnijlcus var. qiuidratus Entz, 1902b, p. 94; 1905, \>. 112 (as magnifica var. quadratn). Ornithocercus magnificus var. e Stuwe, 1909, p. 235, 238, 239, 240, 242, 254, 275, 288. Hislioneis magnifica var. quadrata FoRTi, 1922, p. 114. Ornithocercus assi7nilis Jorgensen, 1923, p. 37, fig. 51. Diagnosis: — Lateral outline subcircular. Anterior cingular list 0.50-1.00, posterior cingular list 0.45-0.86 the greatest depth of body; inclination of these lists about the same as in Ornithocercus steini. Left sulcal list ends on dorsal side of body at a distance from the posterior cingular list equaling 0.37 (0.05-0.65) the greatest depth of body; either squarish or with two lobes, one of which is posteroventral, the other posterodorsal in position; width at postero ventral corner 0.58-1.10, at posterodorsal corner 0.53-0.88 the greate.st depth of body; with three to eight ribs behind fission rib, and usually with submarginal rib ; ribs very variable, sometimes with, sometimes without, branches. Length of body, 37.8- 73.4 M. Widely distributed in tropical, subtropical, and warm-temperate seas. Description: — On account of the uncertain specific unity of this form, we have refrained from giving the usual elaborate specific description. Dimensions: — Our specimens: Length of body, 64.0-37.8 n. Greatest depth of body, 78.5-37.1 ti. In forma quadrata the corresponding values were 64.0- 55.3 fi and 78.5-68.1 m; m forma schiitti they were 54.0-44.3 // and 58.5-46.0 n; in forma assimilis, 52.8^6.7 m and 55.0-46.5 m; in forma simplex, 45.2-43.0 m and 43.4-42.0 m; in forma intermedia, 45.0-37.8 m and 42.9-37.1 fi. The type specimen (Schiitt, 1900a, fig. 4) was 73.4 m long and 82.7 ix deep. The specimens figured by Schroder (1900a, pi. 1, fig. 15) were only about 36.0 ^ long and 40.1 m view. X 450. Station 4740 (300-0 fathoms). 4, Hislioneis carinala Kofoid (?), right lateral view. X 450. Station 4681 (300-0 fathoms). 5, Ornithocercus quadratus Schiitt f. quadrata, right lateral view. X 450. Station 4740 (300-0 fathoms). 6, Parahistioneis para (Murray and Whitting), right lateral view. X 450. Station 4699 (300-0 fathoms). 7, Ornithocercus sp., right lateral view of divi.sion stage, in which the left sulcal list of the right valve is in a state of development, and which shows that the lists of the old valve are not resorbed. X 450. Station 4619 (surface). 8, Histioneis panaria, sp. nov., ventral view of type specimen. X 935. Station 4724 (300-0 fathoms). 9, Hislioneis panaria, sp. nov., oblique right lateral view of type sjiecimen. X 935. Station 4724 (:500-0 fathoms). 562 THE DINOPHYSOIDAE. deep. Whether these specimens actually were so small or whether there is a mistake in the magnifications cannot he decided. It should be observed that according to the shape and structure these specimens appear to belong to forma quadrata. The specimen represented by Jorgensen's (1923) figure 50 (forma quadrata) was about 59.9 yu long and GO. 5 m deeji; the one represented by Jorgen- sen's (1923) figure 51 (forma assimilis) was 44.4 /i long and 42.5 m deep. Variations: — Ornithocercus quadratus, as conceixed in the present paper, is characterized by an extraordmary variability. It is indeed an eloquent example of Schiitt's (1893, p. 269, 270) statement that in tropical waters "selbst die einzelnen Individuen desselben Typus weichen so weit von einandcr ab, dass es aussieht, als woUte jedes von ihnen sich als eigene Varietat etabliren." Almost every one of the characteristics used in distinguishing the species of this genus appears to be in a plastic, ever-changhig condition in 0. quadratus. WTiether or not our decision to treat this multitude of forms as a single species is correct can- not be decided at the present time. We wish to emphasize, however, that our solution is regarded as tentative and that it is the result of many futile efforts at a usable and defensible subdivision. In Figures 86 and 87 the specimens have been arranged according to certain degrees of similarity. In the following discussion of these figures we shall first present the results of our attempts to classify our speci- mens and then give the reasons why the groups established have not been treated as independent species. 0. quadratus f. quadrata: — The specimens represented by Figure 86: 1-6, and by Figure 85: 5, forma distinct group, the specific unity of which appears to be beyond doubt in spite of the rather pronounced variations in the structure and shape of the left sulcal list. This group, which comprises also the type specimen (Schutt, 1900a, fig. 4) and the specimens represented by Schiitt's (1900a) figure 3 and Jorgensen's (1923) figure 50, is characterized especially by the large size, by the gently concave or flat trans\-erse furrow, and by the highly reticulated margin of the left sulcal list ; it should also be mentioned that the radial ribs of this list often are furnished with short, more or less developed branches. The number of radial ribs in this list behind the fission rib ranges from five (Figure 86: 6) to eight (Figure 86: 2), six being the typical. The arrangement of these ribs is rather vari- able. The shape of the left sulcal list is also variable; sometimes, e.g., in the type specimen (Schutt, 1900a, fig. 4), in Jorgensen's (1923) figm-e 50, and in our Figure 86: 1, this list is squarish and comparatively broad; sometimes (Figure 86: 4) its posterior margin is characterized by a fairly well-developed antapical lobe, in other words, its posterior portion is rather distinctly three-lobcd; some- SYSTEMATIC ACCOUXT. 563 times (Figure 86 : 3) it is relatively narrow and has two pronounced posterior lobes, its postmargin being strikingly concave. To this group belonged fifty- seven (i.e., 18.7%) out of the 305 specimens of this species that were carefully exammed by us. Out of these fifty-seven specimens not less than thirty-seven (64.9%) agreed with Figure 86: 5; eight (14.0%o) also resembled Figure 86: 5, but their left sulcal list had the postmargin somewhat straighter, an extra rib was developed between the fission rib and the rib ending near the posteroventral cor- ner (as in Figure 86: 4), and in some of these specimens the margin outside the submarginal rib was very narrow; 4 (7.0%) agreed with Figure 86: 3, but their left sulcal list had the postmargin shaped as in Figure 86 : 2, and its submarginal rib was marginal; one (1.8%) agreed with Figure 86: 3, but all the ribs of the left sulcal list were complete; one (1.8%,) agreed with Figure 86: 4, but its left sulcal list had the ribs ending near the antapical and posteroventral lobes incomplete, and its antapical lobe was slightly less developed ; of each of the forms represented by Figure 86: 1, 2, 3, 4, 6, and by Figure 85: 5, only one specimen was found. 0. quadratus f. schutti: — A second, fairly distinct group is formed by the specimens represented by Figure 86 :7-14, and by Plate 17, figure 8. In the case of these specimens, too, the unity appears to be certain, although striking variations occur. This group, which also includes Schiitt's (1900a) figure 2, is characterized mainly by the intermediate size, by the gently or moderately concave transverse furrow, and by the shape and structure of the left sulcal list. This list is squarish ; the margin outside its submarginal rib is compai-atively narrow, and either plain or furnished with but slightly developed structural differentiations; and its ribs are nearly always simple, in other words, without branches. The number of radial ribs in this list behind the fission rib ranges from three (Figure 86: 14) to seven (Figure 86: 7), six being the normal, and it evidently increases with age as in the case of Ornithocercus magnificus and 0. thurni. The arrangement of these ribs is variable; e.g., in some specimens (Figure 86: 12) one of these ribs ends at the posteroventral corner of this list, while in others (Plate 17, fig. 8) this corner lacks ribs. The posterior margin of this list sometimes (Figure 86: 11) is nearly straight, sometimes moderately convex (Figure 86: 7) or concave (Figure 86: 12) ; a slight indication of an antapical lobe may also be present. To this group be- longed 145 (47.6%) out of the 305 specimens of this species that were carefully examined Ijy us. Out of these 145 specimens seventy-eight (53.8%) agreed with Figure 86:8; twenty-three (15.9%) agreed with Figure 86:10; twenty-two (15.2%) with Figure 86:7; eight (5.5%£,) with Figure 86:14; five (3.A%) with Figure 86: 8, but their left sulcal list had an extra, incomplete rib between the 564 THE DINOPHYSOIDAE. I'ldUHE 80. — Oridthocernwi gvadralus Sihiitt, right latoral view, illustrating variations in size of body and in shajje and structure of left suloal list. X 340. 1-6 belong to forma qmdrata; 7-14, to f. xchutli, f. nov. 1, 5, from Station 4687 (300-0 fathom.s); 2, from Station 4742 (300-0 fathoms); 3, 13, from Station 4737 (100-0 fathoms); 4, 7, 8, from Station 4722 (300-0 fatlioms); 6, 9, from Station 4740 (300-0 fathoms); 10, 14, from Station 4732 (300-0 fatlioms); 11, from Station 4701 (300-0 fathoms); 12, from Station 4730 (300-0 fathoms). SYSTEMATIC ACCOUNT. 565 dorsal rib and the three antapical ribs, the margin outside its submarginal rib was somewhat broader and had a slight reticulation, and its postmargin was more lobed in the middle; two (1.4 7c') agreed with Figure 86:9; one (0.7%), with Figure 86: 8, but it lacked the submarginal rib in the left sulcal list; one (0.7%), with Figure 86:8, but it lacked the submarginal rib and the radial rib to the posteroventral corner of the left sulcal hst; one (0.7%) agreed with Figin-e 86: 9, but its left sulcal list lacked the submarginal rib, and its radial ribs were complete ; there was only one specimen of each of the forms represented by Figvu'e 86 : 11, 12, 13, and by Plate 17, figure 8. 0. qiiadralus f. assimilis (Jorgensen) : — A third group, the unity of which is equally probable, is formed by the specimens represented by Figure 87: 1-8. This group, which includes also Jorgensen's (1923, fig. 51) Ornilhocercus assimilis, may be characterized as follows. The size is intermediate. The transverse furrow- is flat or gently convex. The left sulcal list is squarish and nearly as wide dorso- ventrally as in the second group. The margin of this list outside the submarginal rib is of a moderate width and as a rule not structurally differentiated except dorsally to the dorsal rib and at the corners where it often is reticulated. The radial ribs behind the fission rib, the number of which varies (with age?) from five (Figure 87: 7) to eight (Figure 87: 1), usually are fiu-nished with short, more or less developed branches, or their proxunal ends may be connected by reticular anastomoses (Figure 87: 1, 2). Their arrangement is variable; for instance, some- times (Figure 87: 2) one of them end.s at the posteroventral corner of this list, while in other cases (Figure 87: 6) this corner lacks ribs. The distal end of the posterior fission rib generally is bent posteriorly, joining the submarginal rib (Figure 87: 5, 6; Jorgensen, 1923, fig. 51). The posterior margin of this list some- times (Figure 87:3) is nearly straight, sometimes (Figure 87: 1) gently convex, and sometimes (Figure 87: 5, 6) it has an mdication of an antapical lobe. To this group belonged only twenty-nine (9.5%) out of the 305 specimens of this species that were carefully examined by us. Out of these twenty-nine specimens twehe (41.4%o) agreed with Figure 87:7; ten (34.4%) agreed with Figure 87:8; two (6.9%) with Figiu-e 87: 2; and of each of the forms represented by Figure 87: 1, 3, 4, 5, 6, there w'as only one specimen. 0. quadratus f. simplex, nov.: — A fourth and quite distinct group is formed by the specimens represented by Figure 87 : 11-13 and by Plate 17, figure 2. This group is characterized especially by the relatively small size, by the gently convex transverse furrow, and by the shape and the structure of the left sulcal list. This list is squarish and relatively narrow when compared to this structure in the prcvi- 566 THE DINOPH YSO IDA K Figure 87.— Ornithocercus quadratus Schiitt, right lateral view, illustrating variations in size of body and in shape and structure of left sulcal list. X 340. 1-8 belong to forma nsKimilis (Jorgensen) ; 1 1-13, to f. Mmplex, f. nov. ; 15-20, to f. intermedia, f. nov. ; 9, 10, 14 are of uncertain allocation. 1, 3, 6, from Station 4f)38 (300-0 fathoms); 2, from Station 460.'> (.300-0 fathoms); 4, 7, 8, from Station 4(507 (surface); .5, 14, from Station 4610 (surface); 9, from Station 4713 (300-0 fathoms); 10, 11, 12, from Station 4681 (300-0 fathoms); 13, from Station 4722 (300-0 fathoms); 1.5, from Station 4732 (300-0 fathoms); 16, 18, from Station 4737 (300-0 fathoms); 17, 19, 20, from Station 4730 (300-0 fathoms). SYSTEMATIC ACCOUNT. 567 ous groups. The margm outside its submarginal rib, when this is present, is fairly narrow and lacks structural differentiation. Its ribs appear always to be simple, in other words, to lack branches. The number of the radial ribs behind the fission rib ranges from three (Figure 87:13) to five (Figure 87:11, 12); presumably these ribs increase in number with age. At least in some specimens (Figure 87 : 13) the distal portion of the fission rib is bent backward. The postmargin of this list is nearly straight. To this group belonged sixty-one (20.0%) out of the 305 speci- mens of this species carefully examined by us. Out of these sixty-one specimens not less than fifty-one (83.5%) agreed with Figure 87: 12; seven (11.5%) re- sembled Figure 87: 13, fairly closely; two (3.3%) agreed with Figure 87: 11; and one (1.6%,) with Plate 17, figure 2. 0. quadratus f. intermedia, nov. : — A fifth, and last, group is made up of the specimens represented by Figure 87: 15-20. Although quite pronounced varia- tions occur within this group, it forms a fair unit. It is characterized chiefly by the relatively small size, by the gently convex or flat transverse furrow, and by the shape and structure of the left sulcal list. This list is strikingly bilobed, its postmargin being of pronounced concavity and relatively narrow dorsoventrally. The margin outside the submarginal rib, when this is present, is of moderate width (Figure 87: 15) or narrow (Figure 87: 16, 17), and it lacks structural differ- entiation except in the two lobes, which sometimes are reticulated (Figure 87 : 15). Its ribs are always simple, i.e., lack branches. The number of radial ribs behind the fission rib is from three (Figure 87: 18, 20) to five (Figure 87: 15, 16). The distal portion of the fission rib is curved backward. To this group belonged only eight (2.6%) of the 305 specimens of this species that we examined carefully. Two of these eight specimens agreed with Figure 87: 19; one more resembled this figure, but the postmargin of the left sulcal list was but slightly concave; and of each of the forms represented by Figure 87: 15, 16, 17, 18, 20, only one specimen was found. The specimens represented by Figure 87: 9, 10, 14, are of uncertam alloca- tion. The form represented by Figiu'e 87 : 9, of which three specimens are re- corded, resembles the specimens of the third group in the shape of the body and in the shape of the left sulcal list; in the simplicity of the radial ribs of this list, on the other hand, it agrees with the specimens of the second group. The speci- men of Figure 87: 10, embodies characteristics of the third, fourth, and fifth groups. It resembles the specimens of the third group in having the radial ribs of the left sulcal list furnished with short branches ; the shape of this list recalls the fourth group ; and the shape of the posterior fission rib suggests relationship to the 568 THE DINOPHYSOIDAE. fifth group. The specimen of Figure 87:14 presumably is abnormal. It ap- proaches most closely the fourth group, but the radial ribs of the left sulcal list behind the fission rib are A'ery irregular (possibly due to regeneration following damage). Unfortunately, we are not able to give a complete account of the distribution of these five forms in the area investigated by the Expedition, since a fairly great number of our specimens were not identified to forma. However, since the dis- tribution of the 305 specimens of this species that were carefully examined by us shows certain interesting and possibly indicative features, it will be presented. The fifty-seven specimens of forma quadrala were restricted to the South Equa- torial Drift and the South Equatorial Current. Two of them occurred at Station 4687, the remaining ones were taken at Stations 4709, 4721, 4722, 4724, 4730, 4732, 4736, 4737, 4740, and 4742; most of them (36) were found at Station 4740. The 145 specimens of forma schutti were found at the following stations : 4605, 4607, 4638, 4679, 4697, 4701, 4709, 4713, 4721, 4722, 4730, 4732, 4736, 4737, 4740, and 4742. In other words, most of them were recorded from the South Equatorial Drift, but some were taken in the Mexican Current, Panamic Area, Easter Island Eddy, Galapagos Eddy, and South Equatorial Current. This form was most common at Stations 4701, 4707, 4709, 4721, 4722, and 4740. The twenty-nine specimens of forma assimilis occurred at Stations 4605, 4607, 4619, 4()38, 4713, and 4737; in other words, in the Mexican Current, Panamic .\rea, Galapagos Eddy, and South Equatorial Drift. Most of these specimens were found in the Mexican Current and the Panamic Area. The sixtj'-one specimens of forma simplex were taken at Stations 4605, 4607, 4638, 4681, 4691, 4695, 4697, 4701, 4713, 4722, 4724, 4730, 4732, 4737, 4740, and 4742; i.e., they were fairly evenly distributed in the area occupied by forma schiltti. The eight specimens of forma intermedia were restricted to three stations in the eastern portion of the South Equatorial Drift, viz., to Stations 4730, 4732, and 4737. According to these data, therefore, two of these five forms, viz., forma schutti and forma simplex, appear to be fairly uniformly distributed over the area occupied by the species; two, viz., forma quadrata and forma intermedia, appear to have their center of distribution in the eastern portion of the South Equatorial Drift; and one, forma assimilis. appears to predominate in the Mexican Current and the Panamic .\rea. Our reasons for treating these five forms as formae and not as distinct species are as follows: — (/) Differences in size appear to be of little consequence in this genus, due to the peculiar mode of division illustrated by Plate 18. {2) The struc- ture and shape of the left sulcal list are strikingly variable also within other spe- SYSTEMATIC ACCOUNT. 569 cies of this genus. This Ust appears to change quite readily in response to changes in the surrounding medium; and the area investigated by the Expedition is very large, includes regions of quite different physical conditions, and some of these regions are very turbulent. (.5) Forma schiitti, f. assimilis, f. simplex, and f. inter- ynedia are not clear-cut units, but all their characteristics show overlapping varia- tions. Intermediate specimens sometimes are found, the allocation of which is uncertain. On the other hand, it cannot be too strongly emphasized that there are fairly good reasons in favor of the systematic independence of these forms. First, the variations exhibit distinct modes corresponding to these formae. For instance, out of the fifty-seven specimens of f. quadrata not less than tliii'ty-seven (64.9%) agreed quite well with Figure 86:5; and out of the 145 specimens of f. schiitti seventy-eight (53.8%) agreed with Figure 86: 8. Second, the distribution of these forms also appears to support the assumption of systematic differentiation. It is far from impossible that we are concerned here with the incipient stages of speciation. Comparisons: — The type specimen as figured by Schiitt (1900a, fig. 4), as well as the specimens represented by Schroder's (1900a) Plate 1, figure 15, and by Jorgensen's (1923) figure 50, resemble our Figure 85: 5; Schiitt's (1900a) figures 2 and 3 agree fairly well with our Figure 86 : 8 and Figure 86 : 4, respectively ; and Jorgensen's (1923, fig. 51) figure of Ornithocercus assimilis approaches rather closely our Figure 87: 7. On the other hand, we have not found any specimens with the posteroventral corner of the left sulcal list as well rounded as in Schiitt's (1900a) figure 1. The closest-known relative of Ornithocercus quadratus appears to be 0. thurni. The former species is easily distinguished from the latter by its squarish or two- lobed left sulcal list, although this list may have a fairly well-developed middle (antapical) lobe (Figure 86: 4). It also should be mentioned that 0. quadratus f. intermedia is almost mtermediate between this species and 0. splendidus in the shape of the left sulcal list. This list of this forma also recalls 0. heteroporus. Synomjmy: — This species was established by Schiitt (1900a) under the name of Ornithocercus quadratus. In the same year it was figured by Schroder (1900a) as Histioneis magnifica. Ostenfeld and Schmidt (1901), Karsten (1906, 1907), Schroder (1906a, 1909, 1911), Kofoid (1907a), Schiller (1912, 1912a), Gran (1912b), and Jorgensen (1923) either record or mention this species under the name applied by Schiitt (1900a); of these investigators only Jorgensen (1923) gives any figiu'e. Lemmermann (1901a, 1904, 1905a) uses the name Histioneis o/ 0 THE DINOPHYSOIDAE. quadrala; Entz (1902b) uses H. mngnificus var. quadratus; Entz (1905) and Forti (1922) use H. magnifica var. quadrata; Stiiwe (1909) uses Ornithocercus magnificus var. e; none of these investigators gives any figures. Under the name of Dino- physis galea Pouchet (1883, p. 426) figures four specimens, one of which (fig. G: 4) resembles this species but is not so figured as to permit positive identification. Ornithocercus assimilis Jorgensen (1923, p. 37, fig. 51), which by its author was considered as possibly a "form" of 0. quadratus, has in the present paper been inchided in this species. Occurrence: — Ornithocercus quadratus is recorded at seventy-six of the 127 stations. There are 17, 11, 8, 11, 19, and 10 stations on the six lines of the Expedi- tion. Of these seventy-six stations, two (4580, 4583) are in the Cahfornia Current ; thirteen (4587, 4588, 4590, 4592, 4594, 4596, 4598, 4600, 4604, 4605, 4607, 4545, 4546) are in the Mexican Current; eleven (4609, 4615, 4617, 4619, 4631, 4634, 4635, 4637, 4638, 4639, 4640) are in the Panamic Area; four (4646, 4647, 4650, 4671) are in the Peruvian Current; five (4689, 4691, 4695, 4697, 4699) are in the Easter Island Eddy; two (4713, 4715) are in the Galapagos Eddy; thirty-three (4679, 4681, 4682, 4683, 4685, 4687, 4701, 4705, 4707, 4709, 4711, 4712, 4717, 4718, 4719, 4720, 4721, 4722, 4723, 4724, 4725, 4728, 4729, 4730, 4731, 4732, 4733, 4734, 4736, 4737, 4739, 4740, 4741) are in the South Equatorial Drift; three (4742, 4743, 4540) are in the South Equatorial Current; two (4541, 4542) are in the Equatorial Counter Current; and one (4543) is in the North Equatorial Current. There are thirty-three records from the surface (Stations 4583, 4588, 4592, 4596, 4600, 4604, 4607, 4615, 4617, 4619, 4631, 4635, 4638, 4639, 4640, 4682, 4709 [Salpa], 4712, 4718, 4720 [Salpa], 4723 [Salpa], 4725, 4729, 4731, 4733, 4741, 4743, 4540, 4541, 4542, 4543, 4545, 4546). At twenty-nine of these thirty-three sta- tions the species was taken in surface waters only; at three stations (4583, 4617, 4638), in hauls from 300-0 fathoms as well as at the surface; at one station (4709), in a haul from 300-0 fathoms as well as from a Salpa; at Stations 4619 and 4725, in Salpa stomachs as well as in surface hauls. At one station (4737) it is recorded from 100-0 fathoms and 300-0 fathoms; at eight stations (4681, 4701, 4715, 4717, 4721, 4724, 4728, 4732) from 800-0 fathoms and 300-0 fathoms; at one station (4647) from 800-0 fathoms only. .\11 the remaining records refer to hauls from 300-0 fathoms only. The species was taken also in surface waters in Acapulco Harbor, off the Mexican Current. This station is not included in the 127 stations mentioned above. The temperature range of these se^■enty-six stations at the surface was 66°- SYSTEMATIC ACCOUNT. 571 85°; the average was 77.0°. At the thirty-three stations in the surface catches of which this species was found, the surface temperature ranged from 69° to 84°; the average was 78.8°. At Acapulco it was 83°. For the surface catches the following frequencies are recorded: — 15% at one station (4607); 8% at one station (4543); 6% at two stations (4604, 4546); 4% at three stations (4619, 4638, 4720) ; 3% at three stations (4600, 4615, 4545) ; 2% at one station (4741); 1% at six stations (4596, 4640, 4731, 4733, 4541, 4542); in the remaining cases the frequency is less than 1 %. For the catches from 100, 300, or 800 fathoms to the surface the records of frequency are as follows: — 10% at two stations (4740, 4742); 8% at one station (4728); 7% at two stations (4638, 4721) ; 5% at one station (4724) ; 4%, at three stations (4590, 4605, 4722) ; 3%, at two stations (4634, 4681); 2% at seven stations (4580, 4587, 4689, 4691, 4701, 4709, 4730); 1% at sixteen stations (4583, 4594, 4617, 4637, 4695, 4697, 4699, 4705, 4707, 4715, 4717, 4732, 4734, 4736, 4737, 4739) ; at the remaining stations the frequency is less than 1%. For the catch made in Acapulco Harbor a fre- quency of less than 1% is recorded. The type locality of this species is unknown. Karsten (1906) found the spe- cies at sixteen stations in the Atlantic, between lat. 8° 58' N. and lat. 5° 47' S., and between long. 16° 27' W. and long. 8° 4' E. ; Stuwe (1909) at fifteen stations in the Atlantic, between lat. 34° 53' N. and lat. 3° 50' N., and between long. 20° 22' W. and long. 40° 16' W.; Schroder (1909) in the West Indies; Jorgensen (1923) off the east coast of .America, between lat. 30° N. and lat. 36° N., at lat. 30° 8., long. 13° W., and on the southwest coast of Portugal. In the Mediterranean it has been found by the following investigators: — Pouchet (1883) [?] in the Gulf of Lyons; Schroder (1900a) in the Gulf of Naples; Entz (1902b, 1905), Schroder (1906a), and Schiller (1912, 1912a) in the Adriatic Sea; Jorgensen (1923) at several sta- tions, nearly all of which were located in the eastern portion of the Mediterranean. In the Red Sea and in the Gulf of Aden it has been found by Ostenfeld and Schmidt (1901); in the Arabian Sea by Schroder (1906a) and Jorgensen (1923); in the Indian Ocean by Schroder (1906a), and by Karsten (1907) who recorded it from nineteen stations, between lat. 32° 53' S. and lat. 9° 6' N. and between long. 45° 29' E. and long. 98° 21' E. ; in the East Indies, in the South China Sea, and in Japanese waters by Schroder (1906a) ; and between Laysan Island and Ha\\aii by Lemmermann (1904). Most of the records referred to in the last paragraph were from surface hauls. Only one closing-net record has been published as yet, viz., from Station 229 of the Valdivia Expedition, lat. 2° 38' S.,long. 63° 37' E., 200-20 m.,dead (Karsten, 1907). 572 THE DINOPHYSOIDAE. Stiiwe (1909) found this species in waters ranging from 65.5° to 82.7°. Of the investigators who have contrilnited to our Icnowledge of the distribu- tion of this species, only Pouchet (1883), Schroder (1900a), and Jorgensen (1923) give figures by means of which their determinations may be checked. Ornithocercus quadratus is eupelagic and widely distributed in tropical, sub- tropical, and warm-temperate waters. In the Eastern Pacific it occurs in the FioURE 88. — Occurrence of Ornilhocercus quadratus Schiitt. Large, solid circles indicate records from vertical hauls; squares, records from surface hauls; triangles, records from both vertical and surface hauls; small, solid circles, stations at which this species was not found; small, open circles, stations from which no plankton catches were examined. whole area investigated by the Expedition, but it is strikingly rare in the Peruvian Cm-rent. It is not recorded from the northern portion of the California Current but occurs occasionally as far to the north as San Diego, California. It is one of the most common species of Dinophysoidae. See also the section on variations. Ornithocercus carolinae Kofoid Plate 17, fig. 1, 6. Figure 89, 90 Ornithocercus carolituie Kofoid, 1907a, p. 205, pi. 1.1, fig. 92. Jorgensen, 1923, pi. 38, fig. 53. Diagnosis: — Lateral outline subcircular. Cingular lists: anterior 0.46-0.67, posterior 0.43-0.58 the greatest depth of body; anterior with 12-19 complete ribs, sometimes with basal and distal anastomoses, and a number of short, marginal SYSTEMATIC ACCOUNT. 573 ribs; posterior with 14-19 complete, simple ribs, some of which may be comiected by anastomoses. Left sulcal list ends on dorsal side of body, 0.37-0.71 the great- est depth of body from posterior cingidar list; with three lobes, one of which is postero ventral, one posterior, and one posterodorsal ; the posterior larger than the others and usually narrowly rounded; width at posteroventral lobe ( = fission rib) 0..34-0.73, at posterior lobe 0.65-1.11, at posterodorsal lobe 0.27-0.59 the greatest depth of body; with 11-15 radial ribs behind fission rib and usually with reticulation in posterodorsal and posterior lobes. Submarginal rib may be pres- ent. Length, 30.5-52.0 m. Widely distributed in tropical, subtropical, and warm-temperate seas. Description: — This is a medium-sized or small species, subrotund in lateral outline, deepest at or near the middle. The ratio between the length and the depth of the body is 0.98 (0.91-1.07) : 1. The longitudinal axis may be perpendic- ular to the posterior cingiilar list, but usually it is deflected posterodorsally at 1°-10°. The epitheca is 0.49 (0.38-0.64) as deep as the hypotheca, slightly convex or almost flat, highest in the center or dorsally, and sometimes subhorizontal but usually tilted \'entroposteriorly at 5°-12°. The transverse furrow is gently con- cave or nearly flat; its dorsal width is 0.42 (0.36-0.46) the greatest depth of the body ; its ventral width is 0.56 (0.45-0.70) its dorsal width. The posterior cingular list is 0.29 (0.22-0.36) the length of the body from the apex. The dorsal, posterior, and Ventral margins of the hypotheca are well rounded and confluent ; or any of them may be slightly flattened. In dorsoventral view the body is subellipsoidal, well rounded posteriorly, subacute anteriorly, widest in the middle, and 1.6 times longer than wide. The anteri(jr cingular list is 0.59 (0.46-0.67) the greatest depth of the body and has an anterior inclination of 35° (25°-45°) ; it has on each valve 12-19 com- plete ribs, which may be connected by anastomoses basally as well as distally, and a varying number of short, marginal radial ribs. The posterior cingular list is 0.53 (0.43-0.58) the greatest depth of the body and is inclined anteriorly at 65° (50°-85°) ; it has on each valve 14-19 complete, simple radial ribs, some of which may be connected by anastomoses (Jorgensen, 1923, fig. 53). The right sulcal list is fairly large, ending at about the third or fourth rib behind the fission rib of the left sulcal list; its anterior half is of subuniform width; posteriorly its width decreases gradually; in the middle its width is about 0.17-0.20 the greatest depth of the body ; it may be furnished with some ribs that may anastomose. The left sulcal list ends on the dorsal side of the body at a distance from the posterior 574 THE DINOPHYSOIDAE. cingular list equaling 0.51 (0.37-0.71) the greatest depth of the body and is char- acterized by three usually well-developed lobes, one of which is posteroventral, one posterior, and one posterodorsal in position. The posterior of these lobes, which always is larger than the others, usually is narrowly rounded. The postero- ventral and posterodorsal lobes also are narrowly rounded in most specimens, but the former may be but slightly indicated and the latter may be subrectangular. In our specimens the width of this list at the anterior main rib is 0.4(j (0.37-0.56), at the posteroventral lobe {i.e., at the fission rib) 0.53 (0.34-0.73), at the posterior lobe 0.80 (0.65-1.03), and at the posterodorsal lobe 0.43 (0.27-0.59) the greatest depth of the body. In the specimen figured by Jorgensen (1923, fig. 53), which had the posterior lobe better developed than in any of our specimens, the width at this lobe was 1.11 the greatest depth of the body. The margin between the anterior main rib and the fission rib may be nearly straight, but usually it is gently concave. Between the tip of the posterior lobe and the tips of the postero- ventral and posterodorsal lobes, the posterior margin usually is moderately or rather strongly concave, but the margin between the posteroventral and posterior lobes may be nearly straight or even gently convex. The dorsal margin of this list is gently convex or nearly straight. In front of the fission rib this list has a vary- ing number of ribs which may be partly connected by anastomoses. The fission rib ends in the posteroventral lobe. Behind the fission rib there are 11-15 radial ribs, some of which may be branched proximally. The dorsal one of these ribs is somewhat stronger than the others and ends in the posterodorsal lobe. Some- times the bases of these ribs are connected by reticulation. In the posterodorsal and posterior lobes and dorsally to the dorsal radial rib reticulation is usually present. Reticulation may also be developed in the posteroventral lobe. A sub- marginal rib may be present along the posterior margin. The thecal wall has numerous pores; on the hypotheca each of these is sur- rounded by a small areole. Near the middle of the body there are about 10-17 areoles across the face of each valve. The areoles bordering the transverse furrow posteriorly are somewhat larger than the rest. Small, ellipsoidal or ovate, green- ish-yellow phaeosomes may be present in the transverse furrow ; sometimes their number is very large. The cytoplasm is pink. The dimensions of fourteen of our specimens, including the tyjie, and of the specimen figured by Jorgensen (1923, fig. 53) were measured. Dimensions: — Our specimens: Length of body, 30.5-52.0 n (average, 41.9 /j; type, 44.9 m). Greatest depth of body, 31.9-57.0 ^ (average, 42.9 /x; type, 49.4 m)- The specimen figured by Jorgensen (1923, fig. 53) was 47.5 ^ long and 49.0 ^ deep. SYSTEMATIC ACCOUNT. 575 Variations: — Most of the members of this species (Figure 89: 1-4), although exhibiting a remarkable variability in size, are structurally rather constant. How- ever, since we have been forced to include some specimens de\iating ^•er3' con- siderably from these typical ones, the range of variation of the species as a whole is quite pronounced. Next to the size of the body, the most variable characters are the shape and structure of the left sulcal list. The posterior lobe of this list is usually rather narrowly rounded, but it may be subrectangular (Figure 89: 7). The posteroventral lobe is in most cases very well developed, due to the pro- FiGURE 89. — Ornithocercus carolinae Kofoid, right lateral view, illustrating variations in size and in shape and structure of the left sulcal list. X 340. 1, 2, from Station 4740 CBOO-O fathoms); 3, from Station 4719 (300-0 fathoms), 4, from Station 4691 (300-0 fathoms); 5, 7, from Station 4722 (300-0 fathoms) ; 6, from Station 4732 (300-0 fathoms). nounced concavity of the margin between this lobe and the posterior one ; some- times, however, this margin may be nearly straight, in which case the postero- ventral lobe is but slightly marked (Figure 89: 7). The posterodorsal and poste- rior lobes are reticulated in most specimens, but this reticulation may be absent (Figure 89: 5, 7). A submarginal rib may be present or absent. The bases of the radial ribs are usually free but may be connected by a fine reticulation. The fission rib is either subhorizontal or inclined posteriorly at l°-40°. The depth and inclination of the epitheca and the shape of the hypotheca also exhibit variations. Comparisons: — Our description and figures are based on the type material. The specimen figured by Jorgensen (1923, fig. 53) agrees so closely with the type that there can be no doubt as to the correctness of its allocation. The systematic unity of this species is somewhat uncertain. Some of our specimens (Figure 89: 1-4), although varying in size, are quite uniform in struc- ture and resemble the type (Kofoid, 1907a, fig. 92). Their specific unity appears 576 THE DINOPHYSOIDAE. unquestionable. Others (Figure 89: 7) are very different from the type and ap- proach in a striking manner Hisiioneis francescae Murray and WTiitting (1899, pi. 32, fig. 3) ; in some of them the radial ribs of the left suleal list were even con- nected by fine reticulation just as in the type of this last species. Our assignment of these aberrant specimens to Ornithocercus carolinae is tlue to the fact that transitional forms have been found (Figure 89: 5, G). Ornilhocercus carolinae is characterized especially by the three lobes of the left suleal list, the middle (posterior) one of which is the largest; and by the Figure 90. — Occurrence of Ornithocemis caroliitae Kofoid. Large, solid circles indicate records from vertical hauls; squares, records from surface hauls; small, solid circles, stations at which this species was not found; small, open circles, stations from which no plankton catches were examined. numerous radial ribs in this list behind the fission rib. Its closest-known relative is Hisiioneis [Parahistioneis] francescae, into which it e\'en maj' merge (Jorgensen, 1923, p. 38). According to Murray and Whitting's (1899, pi. 32, fig. 3) figure, the type of Hisiioneis francescae differs from our at j-pical members of Ornitho- cercus carolinae mainly in having the entire posterior cingular list finely and evenly reticulated, while in our specimens this list is ribbed. A reinvestigation of the relationship between these two species is necessar}'. Within Ornithocercus the present species occupies a somewhat isolated position; and there is no clear indication as to which of the members of this genus is its closest relative. Synonymy: — This species was established by Kofoid (1907a) under the SYSTEMATIC ACCOUNT. 577 name of Ornithocercus carolinae, and this specific name was used also by Jorgen- sen (1923), who is the only investigator besides Kofoid to record this form. Occurrence: — Ornilhocercus carolinae is recorded at twenty-four of the 127 stations. There are 2, 0, 3, 6, 10, and 3 stations on the six lines of the Expedition. Of these twenty-four stations, one (4590) is in the Mexican Current; one (4613) is in the Panamic Area; three (4691, 4695, 4699) are m the Easter Island Eddy; seventeen (4681, 4683, 4701, 4705, 4707, 4709, 4719, 4721, 4722, 4724, 4728, 4730, 4732, 4734, 4737, 4739, 4740) are in the South Equatorial Drift; and two (4742, 4743) are in the South Equatorial Current. At one station (4743) the species was taken in a surface haul; at one station (4737) m a haul from 100-0 fathoms as well as in a haul from 300-0 fathoms; at two stations (4681, 4728) in hauls from both 800-0 fathoms and 300-0 fathoms. All the other records refer to hauls from 300-0 fathoms only. The temperature range of these twenty-four stations at the surface was 68°-83°; the average was 76.1°. At Station 4743 the surface temperature was 78°. The frequency is less than 1 % at all stations except at Station 4737 where it isl%. The species was first recorded by Kofoid (1907a) from Stations 4719, 4721, 4722, 4724, and 4740 of the Expedition. The type specimen is from Station 4730 of the Expedition. Later Jorgensen (1923) reported it from the Bay of Cadiz, from a few localities in the central and eastern Mediterranean, and from the coast of Brazil. This is a eupelagic, stenothermal, and warm-water species, widely dis- tributed but rare in tropical, subtropical, and warm-temperate regions. Its ab- sence from the California and Peruvian Currents, according to our data, as well as its rare occurrence in surface waters are noteworthy. 3. FoRMOSUS Group. The only member of this group, Ornithocercus formo- sus, was recorded from the material of the Expedition. Ornithocercus formosus Kofoid and Michener Plate 17, fig. 4, 5. Figure 91 Ornithocercus formosiis Kofoid & Michener, 1911, p. 300. Diagnosis: — Lateral outlme subcircular. Cingular lists: anterior 0.53-0.62, posterior 0.48-0.53 the greatest depth of body; anterior with 14-17 complete radial ribs, of which some of the dorsals originate from a common stem, a few in- complete ribs, and basal reticulation; posterior with 12-15 complete ribs, some of 578 THE DINOPHYSOIDAE. which may anastomose. Left sulcal list ends on dorsal side of body, 0.54-0.64 the greatest depth of body from posterior cingular list ; with two narrowly rounded or subacute lobes, one of which is posterodorsal, the other posteroventral; width at former lobe 0.82-O.8G, at latter lobe 0.75-0.90 the greatest depth of body; be- hind fission rib only one big rib, which ends at tip of posterodorsal lobe ; with sub- marginal rib along postmargin and from posteroventral lobe to fission rib, and with fine reticulation in lobes and dorsally to big rib; between big rib and fission rib there are 10-12 weak, irregular, anastomosing or incomplete ribs. Length, 4L1-43.8M. Eastern tropical Pacific. Description: — This is a medium-sized species, subrotund in lateral outline, widest at or near the middle. The I'atio between the length and the depth of the body is 0.99 (0.97-LOl): 1. The longitudinal axis is deflected posterodorsally at about 15°. The epitheca is 0.44 (0,41-0.48) as deep as the hypotheca, slightly convex or flat, highest in or near the center, and has a slight \'entroposterior tilting. The transverse furrow is flat or gently convex; its dorsal width is 0.41 (0.33-0.47) the greatest depth of the body ; and its ventral width is 0.57 (0.50-0.67) its dorsal width. The posterior cingular list is 0.24 (0.21-0.27) the length of the body from the apex. The hypotheca may be somewhat flattened dorsally and postero- ventrally; its dorsal, posterior, and ventral margins are confluent. In dorso- ventral view the body is subellipsoidal, broadly rounded posteriori}-, subacute anteriorly, widest in the middle, and about 1.4 times longer than wide. The anterior cingular list is 0.58 (0.53-0.62) the greatest depth of the body and has an anterior inclmation of 40°-45°; on each valve it has 14-17 simple, complete radial ribs, some of the dorsal of which originate from a common stem, a few incomplete ribs, and basal reticulation. The posterior cingular list is 0.51 (0.48-0.53) the greatest depth of the body and inclined anteriorly at 70° (65°-80°) ; on each valve it has 12-15 complete, subequidistant radial ribs, some of the dorsal and ventral ones of which anastomose into a reticulum. The right sulcal list has been observed in one specimen only ; in this specimen, the type, it was subrect- angular and reticulate, ended at the fission rib, had a maximum width about 0.12 the greatest depth of the body, and its ventral margin appeared to be attached to the left sulcal list as in Histioneis (Plate 17, fig. 5). The left sulcal list ends on the dorsal side of the body at a distance from the posterior cingular list equaling 0.59 (0.54-0.64) the greatest depth of the body and is characterized by two nar- rowly rounded or subacute posterior lobes, one of which is posteroventral, the SYSTEMATIC ACCOUNT. 579 other posterodorsal. The width of this list at the anterior main rib is 0.55 (0.52- 0.58), at the fission rib 0.48 (0.36-0.58), at the postero ventral lobe 0.84 (0.82- 0.86), and at the posterodorsal lobe 0.84 (0.75-0.90) the greatest depth of the body. The ventral margin of this list usually is almost straight and subparallel to the midline of the body, but it may bulge out into a low and broad lobe at the fission rib. Between the posterior lobes the margin is deeply concave, the mini- mum width of the list in this region being 0.31-0.37 the greatest depth of the body. Dorsally the margin is gently convex. Between the anterior main rili and the fission rib this list is heavily reticulated. Behind the fission rib it has only one big rib, which ends at the tip of the posterodorsal lobe. A feeble submarginal rib extends along the posterior margin, between the tips of the posterior lobes, and along the ventral margin, from the tip of the postero ventral lobe to the fission rib. Both lobes as well as the portion of the list dorsally to the big rib are finelj^ and heavily reticulated; and between the big rib and the fission rib there are about 10-12 weak, irregular radial ribs, most of which anasto- mose. The thecal wall of the hypotheca is re- ticulate ; the meshes are rounded or subangu- lar, subuniform, of medium size, and each of them has a pore in the middle. Near the mid- dle of the body there are about 20-25 meshes across the face of each valve. The meshes bordering the gu'dle posteriorly are not dif- ferent in size from the others. The transverse furrow is porulate but lacks re- ticulation. Phaeosomes have not been observed. The dimensions of three specimens were measured. Dimensions: — Length of body, 41.1-43.8 ix (average, 42.9 n; type, 41.1 ii). Greatest depth of body, 40.6-45.2 m (average, 42.9 ix; type, 40.6 n). Variations: — The five specimens of this species examined as yet, were re- markably constant in size, shape, and structure. The most striking variation was found in the shape of the ventral margin of the left sulcal list. Usually this margin is almost straight, but in one specimen (Figure 91 : 1) it bulged out into a low and broad lobe at the fission rib. Comparisons: — The description and figures given in the present paper are based on the type material. This species occupies a rather isolated position and is in several respects inter- FiGURE 91. — ■ Ornithocercus formo- sus Kofoid and Michener, right lateral view. X 340. 1, from Station 4722 (300-0 fathoms); 2, from Station 4737 (100-0 fathoms). 580 THE DINOPHYSOIDAE. mediate between Ornithocercus and Histioneis. The shape and structure of its body, the size and shape of its cingular Usts, the structure of its posterior cingular Hst, and the endmg of its left sulcal list on the dorsal side of the body are charac- teristics typical of Ornithocercus. It resembles Histioneis in the following re- spects: — (1) the dorsal radial ribs of the anterior cingular list originate from a common stem; (^) the ventral margin of the right sulcal list appears to be at- tached to the face of the left sulcal Ust ; (3) there is only one strong and complete radial rib in the left sulcal list behind the fission rib, and this strong rib runs some- what inside the dorsal margin of this list. We do not know which of the described members of Ornithocercus is the closest-known relative of this species (see also Histioneis costata, the section on comparisons). Occurrence: — Ornithocercus formosus is recorded at five of the 127 stations. There are 0, 0, 0, 2, 3, and 0 stations on the six lines of the Expedition. Of these five stations, one (4697) is in the Easter Island Eddy, and four (4701, 4722, 4724, 4737) are in the South Equatorial Drift. At one station (4737) the species was taken in a haul from 100-0 fathoms; the remaining records refer to hauls from 300-0 fathoms. The temperature range of these five stations at the surface was 72.0°-81.5°; the average was 76.5°. The frequency is in all cases less than 1 %. The species has been found only in the material of the Expedition. It was first recorded by Kofoid and Michener (1911) from Station 4697 of the Expedi- tion, which thus is the t^'pe locality. Its absence from the coastal currents and from the surface waters is noteworthy. Specimens of Ornithocercus of Questionable Specific Allocation A few of the specimens of this genus found in the material of the Expedition are of questionable specific allocation (Figure 92) . These specimens appear to be abnormal, and their aberrant structure is probably, at least in most cases, due to irregularities in the regenerative processes following local injuries or to disturb- ances in regulatory adaptations to flotation. Instead of assigning them more or less arbitrarily to one or the other of the known species, or establishing them as types of new species, we have deemed it most advisable to treat them separately and to restrict ourselves to pointing out in which respects they approach certain known species. Future investigators may be able to solve the question of the specific allocation in a more definite manner by the application of experimental methods to form-regulation. SYSTEMATIC ACCOUNT. 581 Figure 92: 1 . This specimen resembles Ornithocercus thurni, but it differs from that species in having two antapical ribs in the left sulcal list, each ending in a separate (antapical) lobe. This list thus has fom- lobes instead of three. Figure 92: 2. The squarish shape of the left sulcal list is suggestive of Orni- thocercus quadratus. On the other hand, the ribs of this list behind the fission rib form an irregular reticulum and show no resemblance to those of this species. Figure 92. — Ornithocercus spp. Specimens of questionable specific allocation, right lateral view. X 340. 1, from Station 4691 (300-0 fathoms); 2, 7, from Station 4638 (300-0 fathoms); 3, from Station 4732 (300-0 fathoms); 4, from Station 4619 (surface); 5, from Station 4722 (300-0 fathoms) ; 6, from Station 4742 (300-0 fathoms) ; 8, from Station 4637 (300-0 fathoms) ; 9, from Station 4605 (300-0 fathoms) . Figure 92: 3. The a- and b-ribs of the left sulcal list are of about the same type as in Ornithocercus steini, but in other respects this specimen resembles some of our representatives of 0. thurni. Figure 92: 4- The squarish shape of the left sulcal list reminds us of Orni- thocercus quadratus, but the radial ribs of this list, although quite irregular and partly incomplete, indicate relationship to 0. steini. 582 THE DINOPHYSOIDAE. Figure 92: 5. The shape of the postero ventral portion of the left sulcal list, as well as the o-rib of this list indicate that this is a specimen of Ornithocercus steini, but the irregularities in the shape of the posterodorsal portion of this list and in its b- and c-ribs prevent a definite specific allocation. Figure 92: 6. Judging by the shape of the left sulcal list and by the a- and 6-ribs of this list, this specimen belongs to Ornithocercus thurni. The c-, d-, and e-ribs of this list, however, are too irregular and incomplete to permit the assign- ment of this specimen to this or to any other known species. Figure 92: 7, 9. We must refrain from attempting any specific assignment of these two specimens on account of the pronounced irregularities in the shape and structure of the left sulcal list. Figure 92: 8. This specimen probably belongs to either Ornithocercus steini or 0. thurni, but the radial ribs of the left sulcal list are too irregular and incom- plete to permit even a tentative choice between these two species. The shape of this list suggests relationship to 0. thurni. The rounding-off of the angles of the left sulcal list and the irregularities in the ribs seen in 0. orbiculatus are suggested in the individuals shown in Plgure 92:4, 6, and 8. Another group of specimens cannot be determined specifically, either on account of the fact that they are not sufficiently differentiated following fission (Figure 85: 7), or due to the profound structural changes attending fission (Plate 18, fig. 2, 7-10). Parahistioneis, gen. nov. Plate 19, 29. Figure 93: 2-7, 85: 6 Flistioiiris AucTORUM parlim. Diagnosis: — Body subrotund, subobovate, rounded squarish, carafe- or gourd-like in lateral outline, usually slightly longer than deep (ratio between length and depth, 0.86-1.44: 1), and more or less compressed bilaterally. Epi- thcca small, nearly flat to rather strongly vaulted; its depth 0.14-0.48 the depth of hypotheca. Transverse furrow wide, and, as a rule, decidedly wider dorsally than ventrally; dorsal width 0.37-0.63 the depth of body. Cingular lists large; anterior funnel-shaped but usually not stalked, with 1-14 complete radial ribs and a dorsal height 0.30-0.72 the depth of body; posterior usually forms a closed, cylindrical, upright collar, with less than six radial ribs and without a submarginal cross-rib, and with a dorsal height of 0.34r-0.57 the depth of body. Right sulcal list small; its ventral margin probably in most species attached to the left (excep- SYSTEMATIC ACCOUNT. 583 tion, Parahistioneis rotundata). Left sulcal list usually large; with but one pos- terior lobe and one complete radial rib behind the fission rib (exception, P. fran- cescae, which has two posterior lobes and several complete radial ribs behind the fission rib) ; its greatest width 0.27-0.97 the depth of body. Type. — Parahistioneis diomedeae. Organology: — The body {theca) of Parahistioneis, when seen laterally, is always more or less asymmetrical, due especially to the shape of the transverse furrow; the hypotheca usually is subsymmetrical. In dorsoventral view, on the other hand, it appears always to be symmetrical. The longitudinal axis is per- pendicular or nearly so, except in Parahistioneis francescae and P. crateriformis (Murray and Whitting, 1899, pi. 32, fig. 3; Stein, 1883, pi. 22, fig. 5, 6), in which it has a posterodorsal inclination of about 15°. The ratio between the length and the depth ranges from 0.86: 1 to 1.44: 1; usually it is somewhere between 1.08: 1 and 1.15 : 1 ; in only one species does the depth exceed the length. In other words, the body is usually slightly longer than deep. The shape of the body in lateral view ranges from subrotund (P. rotundata, Figure 93: 3), subovate (F. garretti, Figure 93:5), and rounded squarish {P. paraformis, Figure 93:4) to carafe- (P. kar- steni, Figure 93:2) and gourd-like (P. diomedeae, Figure 93:6). The outline in dorsoventral view is unknown in most species. When known, it sometimes is sub- rotund (P. rotundata, Plate 19, fig. 9) or urn-shaped, widest at the posterior cingu- lar list, well and evenly rounded posteriorly, and constricted anteriorly (P. para- formis, Plate 19, fig. 6) ; or it is widest and strongly convex m the middle, gently concave anteriorly and posteriorly, and its apices are narrowly rounded (P. diomedeae, Plate 19, fig. 4). The ratio between the length and the width ranges from 1.10:1 to 1.58:1. The epitheca is small, and nearly flat to rather strongly vaulted. Sometimes {Parahistioneis reticulata. Figure 93 : 7) its curvature is subsymmetrical, some- times (P. diomedeae, Figure 93: 6) it is more curved dorsally than ventrally. Its depth ranges from 0.14 to 0.48 the greatest depth of the hypotheca (P. crateri- formis. Stein, 1883, pi. 22, fig. 5, 6; P. francescae, Murray and \\liitting, 1899, pi. 32, fig. 3). In some species (e.g., P. reticulata, Figure 93: 7) it is subhorizontal ; in others it is more or less inclined ventroposteriorly, its maximum inclination being about 20° (P. diomedeae, Figure 93: 6). Due to the small size of the epitheca, the transverse furrow is always situated near the anterior end of the body. Its distal portion is not displaced posteriorly; in other words, it does not form a spiral about the body. Its postmargin some- times is straight {Parahistioneis rotundata. Figure 93: 3), sometimes concave an- 584 THE DINOPHYSOIDAE. teriorly (P. karsteni, Figure 93:2), and sometimes more or less midulating (P. diomedme, Figure 93: 6). It is always very wide; its dorsal width (measured in a straight line between the dorsal points of the epitheca and the hypotheca) ranges from 0.37 to 0.63 the greatest depth of the body (P. para, Murray and WTiitting, 1899, pi. 32, fig. 4a; P. crateriformis, Stein, 1883, pi. 22, fig. 5, 6) ; and the typical range is from 0.43 to 0.54. In P. paraformis (Figure 93: 4) the dorsal and ventral widths are subequal, but usually the dorsal is somewhere between 1.3 and 2.5 times wider than the ventral (measured along the curvatures). Dorsally it has a dorsoposterior inclination of somewhere between 20° and 55° and is either gently convex (P. francescae, Murray and WTiitting, 1899, pi. 32, fig. 3), or gently, moderately, to strongly concave (P. garretti, P. karsteni, P. diomedeae; Figifre 93:5, 2, 6). Ventrally it is gently convex, flat, or gently concave. Just as in Histioneis, the concavity and great width of this furrow are correlated with the tendency to increase the size of the girdle. This increase has two purposes: it decreases the specific weight of the organism and increases the capacity of the girdle as a container of the phaeosomes. The distance from the apex to the posterior cingular list is somewhere between 0.23 and 0.50 the length of the body. The length of the longitudinal furrow usually is subequal to or somewhat less than half the length of the hypotheca. Anteriorly this list does not extend be- yond the girdle and it is at most but slightly impressed. The lists of the furrows are large but delicate and arise from low and narrow basal ridges (Plate 19). The anterior cingular list is closed, funnel-shaped, and never notched ven- trally. On the dorsal side its anterior inclination ranges from 15° to 60° (usually between 20° and 40°); on the ventral side this inclination either is subequal to that of the dorsal side, or it is somewhat steeper (Parahistioneis reticulata and P. karsteni; Figure 93: 7, 2). In most species {e.g., P. diomedeae, Figure 93: 6) this list is sessile; in others (P. karsteni and P. rotundata; Figixre 93: 2, 3) it either has an incipient or a short stalk. It always is characterized by a distal flare, which either is symmetrical (P. paraformis, Figure 93 : 4) or more pronounced dorsally than ventrally (P. rotundata, Figure 93: 3). Distally this list is from two to six times wider than proximally. Its dorsal height ranges from 0.30 to 0.72 (usually from 0.50 to 0.65) the greatest depth of the body; and its right and left sides are of about equal height. Its anterior margin is even, except in P. dentata (Murray and WTiitting, 1899, pi. 33, fig. 4), in which it is deeply serrated, the teeth, accord- ing to the figure, being scale-like and overlapping. WTien sessile, this list has a SYSTEMATIC ACCOUNT. 585 fairly great number of radial ribs (about eight to fourteen), most of which are complete, and all of which are simple except the dorsal which has one to many branches (Plate 19, fig. 1, 10, 4). When stalked, it is characterized by a more or less reduced number of well-developed radial ribs, the dorsal of which are branched (P. karsteni and P. rotundata; Plate 19, fig. 2, 8). Besides the complete ribs, this list may have some incomplete, simple, marginal, radial ribs (P. diomedeae, Plate 19, fig. 5) and a basal reticulation (P. garretti and P. paraformis; Plate 19, fig. 1,3). The tips of the ribs may project slightly beyond the free edge of the list. The posterior cingular list is closed except posteroventrally, where it probably opens into the canal formed by the right sulcal list. Parahisiioneis rotundata (Plate 19, fig. 8) is presumably an exception to this rule; in this species this list appears to be open ventrally. (This character is not known in P. francescae.) In most species this list forms a cylindrical, upright collar of moderate height and about as high dorsally as ventrally ; sometunes, however, its anterior inclma- tion does not amount to more than about 65° (P. paraformis and P. garretti; Figure 93: 4, 5). It may flare anteriorly (P. reticulata and P. diomedeae; Plate 19, fig. 7, 10, 4), but it never forms lateral pouches. Its dorsal height ranges from 0.34 to 0.57 the greatest depth of the body. The structural differentiation of this list usually is very simple. In some species (P. crateriformis, Stein, 1883, pi. 22, fig. 5, 6) this list has only a pair of simple, dorsal ribs, one rib on either side of the sagittal suture, and a similar pair ventrally; others (P. paraformis and P. ro- tundata; Plate 19, fig. 3, 8) have besides the sagittal ribs only one to four smiple, radial ribs and sometimes a slight reticulation dorsally and ventrally; others, again, have no radial ribs except the sagittal ribs, but a more or less developed, fine reticulation (P. diomedeae and P. reticulata; Plate 19, fig. 5, 7). The number of radial ribs in this list is thus always small. In contradistmction to Histioneis, a submarginal cross-rib is never present, but in one species (P. dentata, Murray and Whitting, 1899, pi. 33, fig. 4) the anterior margin is thickened. In Parahistioneis rotundata (Plate 19, fig. 8) the right sulcal list is fairly large and ends at the fission rib of the left sulcal list ; its ventral margin is sigmoid and probably free, and it lacks structural differentiation. On the other hand, m P. diomedeae and P. dentata (Plate 19, fig. 5; Murray and ^^^litting, 1899, pi. 33, fig. 4a) this list is relatively very small, finely reticulated, and its ventral margin is attached to the left sulcal list, thus forming a canal, which opens posteriorly near the fission rib of the left sulcal list and presumably also into the girdle. Just as in Histioneis, the water caused by the movement of the transverse flagelhmi to enter the girdle probably passes out by this canal. The condition of this list 586 THE DINOPHYSOIDAE. in the last two species may be similar to that of the remaining members of this genus, in which this structure is unknown. This assumption is made probable by the fact that in these species the posterior cingular list is closed; furthermore, the small size and the attachment of the ventral margin make this list very difficult to detect, which would explain why it has not been recorded. The left sulcal list usually is rather large. In some species {Parahistioneis para- formis and P. reticulata; Figure 93 : 4, 7) it ends at or just ventrally to the antapex of the body ; in others it extends more or less beyond this point and may end on the dorsal side of the body at a distance from the posterior cingular list equaling only about 0.47 the greatest depth of the body (P. diomedeae, Figure 93: 6). The shape and structure of this list are quite diverse within this genus and afford im- portant taxonomic characteristics. The most aberrant type is found in P. fran- cescae (Murray and Whitting, 1899, pi. 32, fig. 3). In this species this list ends on the dorsal side of the body, at a distance from the posterior cingular list equaling about 0.55 the greatest depth of the body, and is characterized by two well- developed lobes, one of which is posterodorsal, the other antapical. At the latter lobe, which is by far the larger, the list reaches its greatest width, viz., about 0.67 the greatest depth of the body. The ventral margin (to the tip of the antapical lobe) is moderately convex; and the margin between the vertices of the lobes is rather strongly concave, subrect angular. The fission rib, which arises a fairly short distance behind the girdle, is about 0.43 the greatest depth of the body, nearly straight, and slightly deflected posteriorly. A strong rib, the posterior main rib, ends in the posterodorsal lobe. Dorsally to the latter rib and anteriorly to the fission rib, this list is reticulate. Between these ribs there are fifteen weak radial ribs partly connected by reticulations and with their tips joined by a sub- marginal rib. This is the only species of this genus in which this list has two pos- terior lobes and more than one complete radial rib behind the fission rib. In the remaining ones this list has one posterior lobe and two complete radial ril)s, viz., the fission rib and the posterior main rib, and it is always widest at the tip of the posterior lobe. The fission rib, which arises at or near the posterior end of the longitudinal furrow, is double, straight or nearly so, and subhorizontal (P. ro- tundata, Figure 93:3) or deflected posteriorly at l°-45°; its length ranges from 0.12 to 0.35 the greatest depth of the body. The posterior main ril), which arises at or somewhat ventrally to the antapex of the body and ends at or near the vertex of the posterior lobe, either is marginal {P. para, Figure 85: 6) or it arises at some distance from the dorsal end of the list ; this distance may be as long as 0.54 the greatest depth of the body (P. diomedeae, Figure 93: 6). This rib is di- SYSTEMATIC ACCOUNT. 587 rected posteriorly or deflected ventrally at l°-30° from the longitudinal axis of the body, and its length ranges from 0.27 to 0.97 the greatest depth of the body; it may be of moderate st rength or st rong ; usually it is simple and tapers distally, but its distal end may be reticulate or club-shaped {P. paraformis and P. dio- medeae; Figure 93:4, 6); and it is straight, gently concave dorsally, or gently sigmoid. In one species (P. diomedeae, Figure 93 : 6) the ventral margin of this list forms a rounded rectangular lobe at the fission rib ; in the others this margin (from the anterior edge of the posterior cingular list to the tip of the posterior lobe) does not form any distinct lobe but is gently to moderately convex (P. para- formis, Figure 93: 4; P. crateriformis, Stein, 1883, pi. 22, fig. 5, 6), moderately to strongly sigmoid (P. rotundata, Figure 93: 3; P. dentata, Murray and Whitting, 1899, pi. 33, fig. 4a), or strongly undulating (P. karsteni, Figure 93:2). At the tip of the posterior main rib the margin is narrowly rounded or acute, and dorsally to this rib it is gently concave, straight, gently convex, or gently to strongly sigmoid. In front of the fission rib this list usually is reticulate, but it may have but a few short radial marginal ribs (P. rotundata, Figure 93: 3) or be without structural differentiation (P. crateriformis, Stein, 1883, pi. 22, fig. 5, 6). Between the two main ribs it may be entirely or partly reticulate (P. reticulata and P. garretti; Plate 19, fig. 7, 1), or it may have a few short radial basal ribs (P. crateri- formis, Stein, 1883, pi. 22, fig. 5) or a submarginal rib as well as short radial basal ribs (P. rotundata, Figure 93: 3). Dorsally to the posterior main rib this list may be reticulate or it may lack structural differentiation. Parasagittal lists have thus far not been found in this genus. The flagellar pore, which is located to the right in the transverse furrow at or just behind the posterior cingular list, is rounded or slightly elongated and has a maximum diameter of about 4 ii. There are no records of pores of the kind found in Phalacroma (Plate 3, fig. 1, 2, of P. pulchrum and P. giganteum) on the ventral side of the left valve, near the sagittal suture, and just in front of the anterior cingular list. The flagclla , which have not been observed as yet, presumably are similar to those in Dinophysis and Phalacroma. The thecal wall exhibits but slight variations in structure. It is always poru- late, and in all species, except Parahistioneis rotundata, it is strongly areolate or reticulate. There is either one pore in each mesh and areole, or there is a smaller number of pores than areoles and meshes. In P. rotundata (Plate 19, fig. 8) the wall of the hypotheca is finely and very faintly reticulate and has but eleven pores on the right face, six of which are located along the girdle. The protoplasmic contents have thus far not been observed. Large, rounded 588 THE DINOPHYSOIDAE. and small, ovoidal phaeosomes have been found in the girdle (.Plate 19, fig. 5,7). The length of the body ranges within this genus from 17.7 n {Parahistioneis rotundata) to 53.5 ^ {P. diomedeae). The methods of measurement are largely the same as in Histioneis. Distribution: — Parahistioneis belongs to the same distributional group as Histioneis; in other words, it is marine, presumably of circumequatorial distribu- tion, avoids coastal waters, is limited to tropical and subtropical seas, and all its species are very rare. In the Atlantic as well as in the Pacific Ocean its northern limit seems to be near lat. 35° N.; its southern extension is unknown. It has not been found as yet in the Mediterranean. The distribution of the genus as a whole and of the individual species is very uncertain due to the scarcity of distributional data. The vertical distribution of this genus must be regarded as unknown. All the records published as yet are from surface hauls. Representatives of Parahistioneis were found at sixteen (12.6%) out of the 127 stations of the Expedition from which dino flagellates were recorded. These sixteen stations were distributed in the following manner (Plate 29) : — 0 (0.0%) out of the four stations in the California Current; 2 (15.4%) out of the thirteen stations in the Mexican Current; 1 (5.9%) out of the seventeen sta- tions in the Panamic Area; 0 (0.0%) out of the twenty-seven stations in the Peruvian Current; 2 (20.0%) out of the ten stations in the Easter Island Eddy; 0 (0.0%) out of the four stations in the Galapagos Eddy; 11 (24.4%) out of the forty-five stations in the South Equatorial Drift; 0 (0.0%) out of the three sta- tions in the South Equatorial Current; 0 (0.0%) out of the two stations in the Equatorial Counter Current; 0 (0.0%) out of the two stations in the North Equatorial Current. The numbers of these sixteen stations are as follows: — 4598, 4604 (Mexican Current); 4619 (Panamic .\rea); 4697, 4699 (Easter Island Eddy); 4701, 4709, 4711, 4720, 4722, 4724, 4725, 4732, 4733, 4734, 4741 (South Equatorial Drift). At nine out of these sixteen stations the genus was taken in hauls from 300-0 fathoms or from 800-0 fathoms. One (4598) of these nine stations is located in the Mexican Current; two (4697, 4699) are in the Easter Island Eddy; and six (4701, 4711, 4722, 4724, 4732, 4734) in the South Equatorial Drift. At seven stations the genus was found in surface waters. These stations are as follows: — 4604 (Mexican Current); 4619 (Panamic Area); 4709, 4720, 4725, 4733, 4741 (South Equatorial Drift). SYSTEMATIC ACCOUNT. 589 These data show that Parahistioneis was not found by the Expedition in the CaUfornia and Peruvian Currents, both of which are of temperate origin, in the Galapagos Eddy, which is directly influenced by the Peruvian Current, or in the South Equatorial Current, Equatorial Counter Current, and the North Equa- torial Current. Most (eleven out of sixteen) of the record stations are in the South Equatorial Drift. The species of this genus evidently are stenothermal, warm- water forms limited to high salinities. There are seventeen records of species of Parahistioneis from vertical catches. Out of these seventeen records, three (17.7%; Stations 4598, 4699, 4722) showed a frequency of 1 %; the remaining fourteen (82.3%) showed a frequency of less than 1 %. There are eleven records of species from surface catches. Out of these eleven records, one (9.1%; Station 4619) showed a frequency of 2%; one (9.1%; Station 4720) showed 1%; the remaining nine (81.8%) showed less than 1% or did not have the frequency established. The average frequency of the individual species is about as low as in the genus Histioneis. Coincident occurrence of different species of Parahistioneis in catches from 300 (800)-0 fathoms is recorded at the following of the nine stations mentioned above: — three species occurred coincidently at one station (11.1%; Station 4699); two species at six stations (66.7%; Stations 4697, 4701, 4722, 4724, 4732, 4734). Coincident occurrence of different species of this genus in surface catches is recorded at the following of the seven stations mentioned above : — three spe- cies occurred coincidently at one station (14.3%; Station 4733); two species at two stations (28.6%; Stations 4619, 4720). It should be mentioned that during the Expedition only two, Parahistioneis diomedeae and P. reticulata, out of the six species were found free-living in surface waters. The remaming surface records refer to specimens found in the stomachs of Salpa taken near the surface. The following species were found in Salpa stomachs: — P. diomedeae, P. karsteni, P. paraformis, P. reticulata, and P. rotundata. Out of our seven species only two have been found outside the area investi- gated by the Expedition, a fact undoubtedly due to the incompleteness of our knowledge of the distribution of this genus. Historical Discussion Of the nme species referred by us to the genus Parahistioneis, one, P. para- formis, is proposed in the present paper, and the remaining ones were previously described and figured in the genus Histioneis. These species are as follows: — 590 THE DINOPHYSOIDAE. Histioneis crateriformis Stein (1883), H. dentata Murray and Whitting (1899), H. diomedeae Kofoid and Michener (1911), H. francescae Murray and Whitting (1899), H. garreUi Kofoid (1907a), H. karsteni Kofoid and Michener (1911), H. para Murray and WTiitting (1899), H. reticulata Kofoid (1907a), and H. rutun- data Kofoid and Michener (1911). Tlie following investigators have contributed to our knowledge of the distri- bution of this genus, not specifying those mentioned above: — Cleve (1901a, 1901c), Daday (1888), Lemmermann (1899a, 1901a), Lohmann (1908), Ostenfeld and Schmidt (1901), and Schroder (1900a). None of these writers gives figures by means of which the determinations may be checked. Minor contributions are found in Biitschli (1885) and Jorgensen (1923). The cytological aspect of the morphology of this genus has not been treated as yet. Subdivisions. Relationships among the Species The ten species referred to Parahistioneis in the present paper may be di- vided as follows: 1 . Species of uncertain generic allocation : — P. francescae, P. crateri- formis, and P. roiundata (Figure 93: 3; Murray and Whitting, 1899, pi. 32, fig. 3; Stein, 1883, pi. 22, fig. 5, 6). 2. Garretti group: — P. garretti, P. paraformis, and P. para (Figure 93: 5, 4;85:G). 3. Reticulata group: — P. karsteni, P. reticulata, P. diomedeae, and P. dentata (Figure 93: 2, 7, 0; Murray and Whitting, 1899, pi. 33, fig. 4). 1. Species OF UNCERTAIN generic allocation. Parahistioneis francescae (Murray and Whitting, 1899, pi. 32, fig. 3) : — The structurally closest-known re- lative of this species is Ornithocercus carolinae (Figure 89). Indeed, these two forms approach each other so closely that even their specific separation has been ciuestioned (Jorgensen, 1923, p. 38), and our material indicates that their ranges of \'ariation overlap. The only reason for the present (tentative) generic assign- ment is that the posterior cingular list of P. francescae is finely and evenly reticu- lated, instead of ribbed, according to Murray and \Miitting's (1899) figm-e. It is possible, however, that this peculiarity of the type specimen is due to advanced ontogenetic differentiation; in other words, that the reticulation is preceded by ribbhig, tlie ribs later being connected by numerous fine anastomoses (see Orni- thocercus carolinae, the section on comparisons). If this is the case, then Para- histioneis francescae should be referred to Ornithocercus. Murray and \Miitting SYSTEMATIC ACCOUNT. 591 (1899), who did not separate Ornithocercus from Histioneis, referred it to the latter genus. Parahistioneis craterijormis (Stein, 1883, pi. 22, fig. 5, 6) : — This species is structurally fairly close to some of the primitive members of Histioneis, and the only character that prevents its assignment to this genus is the lack of a cross-rib in its posterior cingTilar list. If this rib is absent, in other words, if Stein (1883) did not overlook it, then this species should be assigned to the Reticulata group and be placed near Parahistioneis karsteni. Stein (1883) assigned it to Histioneis. Parahistioneis rotundata (Figure 93:3): — This species, which was referred to Histioneis by Kofoid and Michener (1911), occupies a somewhat isolated posi- tion and embodies characteristics of Ornithocercus and Histioneis, as well as of Parahistioneis. It is characterized especially by the following features: — its body is subrotund and lacks an anterior process; its anterior cingular list is stalked and has but few (five or less) ribs on each valve ; its posterior cingular list appears to be open ventrally and has about five radial ribs on each valve ; its right sulcal list is comparatively large, and its ventral edge appears to be free, i.e., not attached to the left sulcal list. If by future investigations it is shown : (1) that the posterior cingular list of this species is open ventrally, {2) that the ventral edge of its right sulcal list is not attached to the left sulcal list, and {3) that these peculiar- ities are not found in the Garretti and Reticulata groups, then this species should be remo\'ed from Parahistioneis and made the type of a new genus. We thus regard its present generic allocation as tentative. The three members of this group thus may be representatives of three ditTer- ent genera. 2. Garretti group (Figure 93:5, 4; 85:6). Body subovate to rounded squarish in lateral outline, without a well-developed anterior process. Anterior cingular list sessile, with eight or more radial ribs on each valve. Posterior cingu- lar list appears always to be closed ventrally and has two to four radial ribs on each valve besides the dorsal and ventral pairs next to the sagittal suture. Right sulcal list small; its ventral edge attached to left sulcal list (?). The three members of this group are structurally verj' uniform and have reached about the same evolutionary level. 3. Reticulata group (Figure 93 : 2, 7, 6 ; Murray and ■V\Tiitting, 1899, pi. 33, fig. 4). Body carafe- or gourd-shaped in lateral outline with well-developed an- terior process. Anterior cingular list may ha\'e an incipient stalk but is usually sessile ; with five or more radial ribs on each valve. Posterior cingular list appears always to be closed ventrally and has no radial ribs except the dorsal and ventral 592 THE DINOPHYSOIDAE. pairs next to the sagittal suture. Right sulcal Ust small; its ventral edge attached to left sulcal list (?). Of the four species assigned to this group, Parahistioneis reticulata, P. diome- deae, and P. dentata are structurally very close to each other and ha\'e reached about the same evolutionary level. P. dentata is somewhat aberrant in having the anterior edge of the anterior cingular list deeply serrated and the anterior edge of the posterior cingular list supported by a marginal rib. (These features are unique and need to be verified.) P. karsteni is somewhat more primitive than these three species in being smaller and in ha\'ing the anterior process of the body somewhat less developed. On the other hand, it is more advanced in having the anterior cingular list characterized by an incipient stalk and by a small number of radial ribs. The Reticulata group, which is quite primitive in the shape and structure of the anterior cingular list, is more advanced than the Garretti group in the shape of the body and in the structure of the posterior cingular list ; the body has developed a large anterior process, and in the posterior cingular list the primitive ribbing has disappeared. Key to the Species of Parahistioneis 1. Posterior cingular list with no radial ribs besides the dorsal and ventral pairs 2. 1 . Posterior cingular list with radial ribs besides the dorsal and ventral pairs 7. 2. Left sulcal list with two distinct posterior lobes, one of which is dorsoposterior, the other antapical; Ra ends in dorsoposterior lobe P. francescae (Murray and Whitting). 2. Left sulcal list not of this type 3. 3. Anterior edge of anterior cingular list deeply serrated P. dentata (Murray and Whitting). 3. .interior edge of anterior cingular list smooth 4. 4. Left sulcal list angular at R? P. diomedeae (Kofoid and Michener). 4. Left sulcal list not angular at R2 5. 5. Left sulcal list wide; R3 0.67 the depth of body P. karsleni (Kofoid and Michener). 5. Left sulcal list fairly narrow; R3 less than 0.45 the depth of body 0. 6. Anterior cingular list sessile, its dorsal height 0.32-0.40 the depth of body. P. reticulata (Kofoid). 6. Anterior cingular list stalked, its dorsal height 0.70-0.75 the depth of body . . P. crateriformis (Stein). 7. R3 of left sulcal list ari.ses about 0.50 the depth of body from dorsal end of list. .P. garretii (Kofoid). 7. R3 of left sulcal list submarginal 8. 8. R3 of left sulcal list with heavy posterior reticulation P. ■parajormis, sp. nov. 8. R3 of left sulcal list not with heavy posterior reticulation 9 . 9. .\nterior cingular list stalked, with five ribs or less on each valve. P. rutnndala (Kofoid and Michener). 9. .interior cingular list not stalked, with ten ribs or more on each valve. P. para (Murray and Whitting). 1. Species of Uncertain Generic Allocation. Of the three species of this group, viz., Parahistioneis francescae, P. crateriformis, and P. rotundata, only the last was found in the material of the Expedition. SYSTEMATIC ACCOUNT. 593 Parahistioneis rotundata fKofoid and Michener) Plate 19, fig. 8, 9. Figure 93: 3 Hislioneis rotundata Kofoid & Michener, 1911, p. 299. Diagnosis: — Body subcircular in lateral outline; length: depth, 1.13:1. Dorsally the transverse furrow is gently concave, 0.48 the depth of body, 1.3 times wider than ventrally, and inclined dorsoposteriorly at 20°. Distance from apex to posterior cingular list 0.23 the length of body. Anterior cingular list with short stalk, flaring in its distal 0.6G with straight sides to an asymmetrical funnel 1.5 wider dorsally than ventrally; its height 0.()4 the depth of body. Posterior cingular list appears to be open ventrally; with five to six simple, complete ribs; dorsal height 0.39 the depth of bod.v. Left sulcal list ends somewhat ventrally to antapex; narrowly rounded at Rs; margin rather strongij' convex just behind R2; dorsal margin gently concave; R-. 0.32 the depth of body; R3 submarginal, of moderate strength, 0.78 the depth of body; from near the middle of R. a rib runs posteriorly, subparallel to margin of list; between Ro and R3 there are four incom- plete ribs. Length, 17.7 m. Eastern tropical Pacific. Description: — This is a minute species, the body of which is subcircular in lateral outline, and deepest near the middle. The ratio between the length and the depth of the body is about 1.13: 1. The longitudmal axis is perpendicular. The epitheca is about 0.17 as deep as the hypotheca, convex, highest some- what dorsally to the center, and subhorizontal. Dorsally the transverse furrow is gently concave, about 0.48 the greatest depth of the body, about 1.3 times wider than ventrally, and inclined dorsoposteriorly at alwut 20°; ventrally it is gently convex. The distance from the apex to the post-erior cingular list is about 0.23 the length of the body. The anterior margin of the hypotheca is nearly straight and horizontal. The dorsal, posterior, and ventral margins are confluent and subcircular. Seen in dorso ventral view, the body is regularly and broadly obovoidal, widest near the middle, and about 1.10 times longer than wide. The anterior cingular Hst has a short but distinct stalk, is inclined anteriorly' at 35°-40°, and is about 5.5 times wider di.stally than proximally; it flares in its distal 0.G6 abruptly with straight sides to an asymmetrical funnel 1.5 times wider dorsally than ventrally, inclined slightly dextrallj- with no ventral notch: its height is about 0.64 the greatest depth of the body; and it has on each valve two radial ribs, the ventral of which is simple, the dorsal furnished with three branches. The posterior cingular list appears to be open ventrally, and its dorsal 594 THE DINOPHYSOIDAE. margin is inclined anteriorly at 80°-85°; the dcirsal height is about 0.39, the ventral 0.48 the greatest depth of the body; it has on each valve five to six simple, complete radial ribs and no reticulation. The right sulcal list ends at the fission rib of the left sulcal list, and its ventral margin, which appears to be free, is strikingly sigmoid, concave in the middle, and convex posteriorly. The left sulcal list ends somewhat ventrally to the antapex, the ventral margin is nearly straight in front of the fission rib, between the fission rib and the posterior main rib the margin is rather strongly convex anteriorly and nearly straight posteriorly, at the posterior main rib the list is narrowly rounded, and the dorsal margin is gently concave. The fission rib is straight, subhorizontal, and about 0.32 the greatest depth of the body. The posterior mam rib is submarginal, of moderate strength, gently concave dorsally, deflected 5°-10° posteroventrally, and about 0.78 the greatest depth of the body ; on the ventral side it has about six knob-like processes. In front of the fission rib there are about four short, marginal radial ribs. From near the middle of the fission rib, a rib runs posteriorly and about parallel to the ventral margin, ending near the posterior main rib. Between the fission rib and the posterior main rib there are four incomplete basal ribs, the anterior of which is faii-ly long, bent posteriorly, and branched distally. The thecal wall of the hypotheca is finely and very faintly reticulated and has eleven pores on the right face, six of which are located along the girdle. There are no records of phaeosomes. The dimensions of one specimen only, the type, were measured. Dimensions: — Length of body, 17.7 /u. Greatest depth of body, 15.7 m- Total length, 37.1 n. Comparisons: — Our description and figures are based on the type speci- men. This species occupies a somewhat isolated position. It embodies character- istics of Ornithocercus, Histioneis, and Parahistioneis, and hence even its generic assignment should be regarded as tentative. The subcircular outline of the body in lateral view, the ribbing of the posterior cingular list, and the development of the right sulcal list indicate relationships to Ornithocercus. It differs from this genus in having comparatively few ribs in the posterior cingular list, in the shape and structure of the anterior cingular list, and in having only two main ribs in the left sulcal list. The shape and structure of the anterior cingidar list remind us of Histioneis and the shape of the left sulcal list of Parahistioneis karsteni. There is no certain indication as to wliich of the known members of these three genera is the closest relative of the present species. The minute size, the subspheroidal SYSTEMATIC ACCOUNT. 595 shape of the body, the size of the right sulcal list, and the fact that the left sulcal list ends on the ventral side of the body, place it fairly low in the evolutionary scale. Occurrence: — Parahistioneis rotundata is recorded at five of the 127 stations. One station (4709) is on the fourth, three (4720, 4725, 4733) are on the fifth, and one (4741) is on the sixth line of the Expedition, and all of them are in the South Equatorial Drift. Five specimens were found, all occurring in Salpa taken in Figure 93. — Histiophysis and Parahistioneis, right lateral view. X 340. 1, H istiopJiysis rugosa (Kofoid) from Station 4705 (300-0 fathom.s); 2, Parahislioneis karsteni (Kofoid and INIic-hener) from Station 4619 (Salpa); 3, P. rohindata (Kofoid and Miehener) from Station 4720 (Salpa); 4, P. para- formis, sp. nov., from Station 4724 (300-0 fathoms); 5, P. garretti (Kofoid) from Station 4732 (300-0 fathoms); 6, P. diomedeae (Kofoid and Michener) from Station 4699 (300-0 fathoms); 7, P. rcticuluta (Kofoid) from Station 4699 (300-0 fathoms). surface waters. The temperature of these five stations at the surface ranged from 72° to 80°; the average was 77.0°. The species has been found only in the material of the Expedition. It was first recorded by Kofoid and Michener (1911) from Station 4720 of the Expedi- tion, which thus is the type locality. The restriction of all the record stations to the South Equatorial Drift is noteworthy. 596 THE DINOPHYSOIDAE. 2. Garretti Group. All the three members of this group were fouiul by us in the material of the Expedition. They have been treated in the following order: — Parahistioneis garretti, P. paraformis, and P. para. Parahistioneis garretti (Kofoid) Plate 19, fig. 1. Figure 93: 5 llislioneiK garretii KoFOiD, 1907a, p. 20:3, pi. 16, fig. 97. Diagnosis: — Body broadly subobovoidal in lateral outline; length: depth, 1.15: 1. Dorsally the transverse furrow is gently concave, 0.52 the depth of body, 2.3 times wider than ventrally, and inclined dorsoposteriorly at 35°-40°. Distance from apex to posterior cingular list 0.3G the length of body. Anterior cingular list sessile, subequal dorsally and ventrally; its height 0.60 the depth of body. Poste- rior cingular list closed; with two simple ribs in the middle and one ventrally; dorsal height 0.52 the depth of body. Left sulcal list ends on dorsal side of body, 0.59 the depth of Ijody from posterior cingular list; narrowly rounded at R3; dorsal margin and margin between Ro and R3 gently sigmoid; R.. 0.31 the depth of l)ody; R3 arises 0.50 the depth of body from dorsal end of list, fairly weak, 0.60 the depth of body ; reticulated in front of R,.; behind Rj reticulations are restricted to basal portion. Length, 37.3 /i. ^ Eastern tropical Pacific. Description: — This is a medium-sized species, with body broadly subobo- voidal in lateral outline, deepest just behind the transverse furrow. The ratio between the length and the depth of the body is about 1.15: 1. The longitudinal axis is about perpendicular. The epitheca is about 0.55 as deep as the hypotheca, convex, highest dor- sally, and tilted ventroposteriorly at about 10°. Dorsally the transverse furrow is gently concave, about 0.52 the greatest depth of the body, about 2.3 times wider than ventrally, and inclined dorsoposteriorly at 35°-40°; ventrally it is gently convex. The distance from the apex to the posterior cingular list is about 0.36 the length of the body. The anterior margin of the hypotheca is nearly straight and subhorizontal. The dorsal, posterior, and ventral margins are confluent; the posterior is somewhat more convex than the others. The anterior cingular list is sessile, inclined at the base anteriorly at 15°-25°, and al)ou1 2.5 times wider distally than proximally; it flares into a broad funnel with nearly eciual ilorsal and ventral extent, straight sides, and with the ventral arishig farther posterior than the dorsal and therefore steeper. Its top, somewhat flattened in Figure 93: 5, is nearly horizontal, its height is about 0.60 the greatest SYSTEMATIC ACCOUNT. 597 depth of the body, and it has on each valve 11 or 12 subequidistant radial ribs, most of which are complete, and all of which are simple except the dorsal one, which has two branches; proximally these ribs are connected by reticulation. The posterior cingular list is closed, somewhat flaring distally, and its dorsal and ventral margins are inclined anteriorly at 50°-70°; the dorsal height is about 0.52, the ventral 0.57 the greatest depth of the body; besides the ribs at the sagittal suture, this list has on each valve three simple, straight radial ribs, two of which are near the middle and one ventrally; dorsally a slight reticulation is present. The left sulcal list is of medium size and ends on the dorsal side of the body at a distance from the posterior cingular list etiualing about 0.59 the greatest depth of the body. It is narrowly rounded at the posterior main rib ; and the dorsal margin and the margin between the fission rib and the posterior main rilj are gently sigmoid. The portion of this list in front of the fission rib is finely reticulated; in the type specimen it is but slightly developed (following fission?), its ventral margin is gently concave, and its posterior width is but slightly more than half the length of the fission rib. The fission rib is straight, about 0.31 the greatest depth of the body, and deflected posteriorly at about 25°. The posterior main rib arises near the antapex of the body and about 0.50 the greatest depth of the body from the dorsal end of the list, is fairly weak, gently concave dorsally, and deflected posteroventrally at 5°-10°; its length is about 0.60 the greatest depth of the body. Behind the fission ril) the basal portion of this list is reticulated, the meshes being rather small to medium-sized. The thecal wall is porulate and reticulated, the meshes being subiuiiform and medium-sized, and each of them having one pore in the middle. On each vahe about eighteen meshes border the gu-dle posteriorly. There is no record of phaeo- somes. The dimensions of one specimen only, the type, were measured. Dimensions: — Length of body, 37.3 ju. Greatest depth of body, 32.3 m- Total length, 72.5 n. Comparisons: — Our description and figure are based on the type specimen. Judghig by the size, shape, and structure of the body, of the cingular lists, and of the left sulcal list in front of the fission rib, this species is most closely re- lated to Parahislioneis paraformis. These two species are readily distinguished from each other by the shape and structure of the left sulcal list behind the fission rib. Another close relative evidently is P. para. This relationship is indi- cated by pronounced similarities in the shape and structure of the liody and of the cingular lists. P. garretti differs strikingly from P. para in having the left 598 THE DINOPHYSOIDAE. sulcal list end on the dorsal side of the body and in having the posterior main rib of this list arise at a distance from the dorsal end of the list equaling 0.50 the greatest depth of the body. Occurrence: — Parahistioneis garretti is recorded at only one of the 127 stations, 4732, on the fifth line of the Expedition in the South Equatorial Drift, from a haul from 300-0 fathoms, a surface temperature of 79°. The frequency is less than 1% (one specunen). The species has been found only in the material of the Expedition. It was first recorded by Kofoid (1907a) from Station 4732 of the Expedition, which thus is the type locality. Parahistioneis paraformis, sp. nov. Plate 19, fig. 3, 6. Figure 93: 4 ? Histioneis para Murray & Whitting, 1899, partim, yi\. 32, fig. 4c. Histioneis para Okamura, 1912, p. 20, pi. 3, fig. 55. Diagnosis: — Body rounded, squarish in lateral outline; length: depth, 1.08- 1.11 : 1. Dorsally the transverse furrow is moderately concave, 0.48 the depth of body, but slightly wider than ventrally, and inclined dorsoposteriorly at 40°-45°. Distance from apex to posterior cingular list 0.43 the length of body. Anterior cingular list sessile, low, subequal, broadly flaring; its height 0.61 the depth of body. Posterior cingular list closed, with one simple rib in the middle and one ventrally; dorsal height 0.53 the depth of body; line of attachment subhorizontal. Left sulcal list ends at antapex ; narrowly rounded posteriorly, gently convex be- tween R2 and R3, gently concave dorsally; R2 0.35 the depth of body; R3 sub- marginal, strong, heavily reticulated distally, 0.95 the depth of body; Length, 38.0-39.0 M. Tropical and subtropical Pacific. Description: — This is a medium-sized species, with body rounded squarish in lateral outline, deepest in the middle. The ratio between the length and the depth of the body is 1.08-1.1 1 : 1. The longitudinal axis is perpendicular. The epitheca is about 0.35 as deep as the hypotheca, rather strongly convex, somewhat higher dorsally than ventrally, and subhorizontail. Dorsally the trans- verse furrow is moderately concave, about 0.48 the greatest depth of the body, but slightly wider than ventrally, and inclined dorsoposteriorly at 40°-45°; ven- trally it is gently convex, nearly flat. The distance from the apex to the posterior cingular list is about 0.43 the length of the body. The anterior margin of the hypotheca is gently undulating and subhorizontal. The dorsal, posterior, and SYSTEMATIC ACCOUNT. 599 ventral margins of the hypotheca are confluent and well convex ; or the hypotheca may be somewhat flattened posterodorsally and posteroventrally. In dorsoven- tral view the body is urn-shaped, widest at the posterior cingular list; narrowly rounded anteriorly, due to the pronounced concavity of the transverse furrow; and the hypotheca is fairly broadly rounded posteriorly, with evenly convex lateral outlines ; the ratio between the length and the width is about 1 .34 : 1 . The anterior cingular list is sessile, widely flaring, mclined anteriorly at 20°- 25°, and about four times wider distally than proximally; its dorsal and ventral extent are subequal, sides concave, equal dorsally and ventrally, and arising at the same level, with horizontal top; its height is about 0.61 the greatest depth of the body, and it has on each valve 8-13 subequidistant, complete radial ribs, all of which are simple except the dorsal one which may have three branches; proximally these ribs are connected by reticulation. The posterior cingular list is closed, somewhat flaring distally, and its dorsal and ventral margins are inclined anteriorly at about 70°; the dorsal and ventral heights are about 0.53 the greatest depth of the body; besides the ribs at the sagittal suture, this list has on each valve two usually simple, straight radial ribs, one of which is in the middle and one ventrally; no reticulation is present. The left sulcal list is of narrow tongue- shape and ends at or near the antapex of the body ; it is narrowly rounded pos- teriorly, the ventral margin between the fission rib and the posterior main rib is gently convex, and the dorsal margin is gently concave. The portion of this list in front of the fission rib is finely reticulated; m the type specimen it is but slightly developed (following fission?), its ventral margin is gently concave, nearly straight, and its posterior width is but slightly more than half the length of the fission rib (Plate 19, fig. 6). The fission rib is straight, about 0.35 the greatest width of the body, -and deflected posteriorly at 25°-30°. The posterior main rib is submarginal, very strong, gently concave dorsally, about 0.95 the depth of the body, and deflected posteroventrally at 5°-10°; its distal end is heavily reticulated and fills the posterior portion of the list. Between the fission rib and the posterior main rib this list is faintly reticulated, the meshes being rather large. The thecal wall is porulate and reticulated, the meshes subuniform and medium-sized, and each of them has one pore in the middle. On each valve about 16 meshes border the girdle posteriorly. There is no record of phaeosomes. Large, round chromatophores (?) are present. All the dimensions were measured in only one specunen, the type. The length and depth of the body and the total length were measured in three more specimens. 600 THE DINOPHYSOIDAE. Dimensions: — Length of body, 38.0-39.0 m (type, 38.0 m)- Greatest depth of body, 35.0-36.0 m (type, 35.0 m). Total length, 82-95 m (type, 83 n). Comi>arisons: — The specimen represented by Okaniura (1912, pi. 3, lig. 55) is figured in a \'ery sketchy manner but nevertheless agrees quite closely with the type except in the ventral margin of the left sulcal list, which is strikingly concave just behind the fission rib (due to recent binary fission?). The specimen figured by Murray and Whitting (1899, pi. 32, fig. 4c) as Histioncis para may possibly belong to this species, but it tliffcrs in the relative depth of the epitheca, in the ribbing of the posterior cingular list, and in having a well-developed rib in the left sulcal list just behind the fission rib. The closest-known relatives of this species presumably are rarahisiioncis garretli and P. para (regarding the relationship to P. garretti, see this species, the section on comparisons). P. parajormis recalls P. para in having but two ribs in the posterior cingular list, in the narrowly tongue-shaped and reticulated left sulcal list, and in the urn-shape of the body in dorsoventral view. It is easily dis- tinguished from this species by the Ijroader shape of the body seen laterally and by the fact that the strong posterior main rib of the left sulcal list is heavily reticulated distally. Synonymy: — This species has been recorded and figured previously by Okamura (1912, p. 20, pi. 3, fig. 55) under the name of Histioneis para. One of the two specimens represented by Murray and Whitting (1899, pi. 32, fig. 4c) under this name may also be referable to the present species, but this allocation is somewhat uncertain, as mentioned in the last section. Occurrence: — Parahistioneis parajormis is recorded at seven of the 127 stations. There are 0, 0, 0, 2, 5, and 0 stations on the six lines of the Expedition. Of these seven stations one (4697) is in the Easter Island Eddy; and six (4701, 4722, 4724, 4732, 4733, 4734) are in the South Equatorial Drift. At one station (4733) the species was found in the stomach of a Salpa taken in surface waters. All the other records refer to hauls from 300-0 fathoms. The temperature range of these seven stations at the surface was 75°-81°; the average was 77.7°. At Station 4733 the .surface temperature was 80°. The frequency is in each case less than 1 %. The species has pre\iously been recorded from Japanese waters by Okanmra (1912) and is presumably widely distributed in tropical and subtropical seas. SYSTEMATIC ACCOUNT. 601 Parahistioneis para (Murray and ^^^litting) Figui-e85:6 Hinlioncis jinm MunRAY & Whitting, 1899, pmihn, \). 333, pi. 32, fig. -la, b, tab. 3, 4, 5, (>, 8, 9. Lemmir- MANN, 1899a, pnrtim, p. 374; 1901a, p. 376. Cleve, 1901c, ]). 2.52. Kofoid A: Micheneu, 1911, ]i. 298, 299. non Histioneis para Okamuka, 1912, p. 20, pi. 3, fly;. 55. Diagnosis: — Body subobovoidal in lateral outline, obliquely truncate an- teriorly; length: depth, 1.13-1.44: 1. Dorsally the transverse furrow is moder- ately to rather strongly concave, 0.37-0.43 the greatest depth of Ixnly, 1.5-2.0 times wider than ventrally, and inclined dorsoposteriorly at 35°-40°. Distance from apex to posterior cingular list 0.24-0.27 the length of body. Anterior cingular list sessile, subequal in dorsal and ventral extension, with unequal concave sides; its height 0.50 the depth of body. Posterior cingular list closed, with two simple ribs, one on either side of the middle ; its dorsal height 0.39-0.45 the depth of body, with sinuous line of attachment. Left sulcal list narrowly tongue-shaped, ends at antapex, reticulated; Rg marginal, directed posteriorly, straight or gently sigmoid, of moderate strength, and 0.76-0.97 the depth of body. Length, 43.7-47.0 yu. Widely distributed in tropical and subtropical seas. Description: — This is a medium-sized species, with body subobovoidal in lateral outline, obliquely truncate anteriorly, and deepest at the posterior cingular list. The ratio between the length and the depth of the body is 1.13-1.44: 1. The longitudinal axis is perpendicular. The epitheca is 0.46-0.55 as deep as the hypotheca, flat except dorsally where it is angular or subangular in lateral outlme, and tilted ^'entroposteriorly at 10°-15°. Dorsally the transverse furrow is moderately to rather strongly con- cave, 0.37-0.43 the greatest depth of the body, 1.5-2.0 times wider than ventrally, and mclined dorsoposteriorly at 35°-40°; ventrally it is flat or gently con\'ex. The distance from the apex to the posterior cingular list is 0.24-0.27 the length of the body. The anterior margin of the hypotheca is nearly straight and .subhori- zontal. Posteriorly the hypotheca tapers quite decidedly, its posterior end bemg rather narrowly rounded. The dorsal margin of the hypotheca is gently con\-ex or flattened, the ventral margin gently con\ex. In dorsoventral view the body is urn-shaped, widest somewhat behmd the girdle ; narrowly rounded anteriorly, due to the pronounced concavity of the transverse furrow; the hypotheca tapers posteriorly, with evenly and gently convex lateral outlines, and rather narrowly rounded posterior end ; the ratio between the length and the width is about 1 .66 : 1 . The anterior cingular list is sessile and flaring, inclined anteriorly at 20°-25°, and 2.3-2.8 times wider distally than proximally; its dorsal extension is 1.3 the 602 THE DINOPHYSOID.\E. ventral, sides concave, the ventral steeper than the dorsal; posteriorly its line of attachment declines abruptly, and top horizontal; its height is about 0.50 the depth of the body ; and it has on each valve about 10-12 subequidistant, complete radial ribs, all of which are simple (except possibly the dorsal one, which may be branched as in Parahistioneis paraformis). The posterior cingular Hst is closed, shghtly flaring distally, and its dorsal and ventral margins are inclined anteriorly at 55°-65°; the dorsal and ventral heights are 0.39-0.45 and 0.43-0.52 the greatest depth of the body; besides the ribs at the sagittal suture, this list has on each valve two simple, straight radial ribs, one on either side of the middle ; no reticu- lation is present. The left sulcal list is of narrow tongue-shape, ends at or near the antapexof the body, and is entirely covered by reticulation. Its ventral margin is gently convex or sigmoid, its dorsal margin is nearly straight or gently sigmoid, and it is narrowly rounded or subacute posteriorly. The posterior main rib is marginal, directed posteriori}-, of moderate strength, and 0.76-0.97 the depth of the body. The thecal wall is reticulate; about 13 meshes border the girdle posteriorly. There is no record of phaeosomes. The dimensions of one specimen from the material of the Expedition and of the type specimen (Murray and Whitting, 1899, pi. 32, fig. 4a) were measured. Dimensions: — Our specimen: Length of body, 47.0 ;i/. Greatest depth of body, 32.6 ii. Total length, 85.2 n. Type specimen, according to Murray and Whitting (1899, Plate 32, figure 4a): Length of body, 43.7 m- Greatest depth of body, 38.6 m- Total length, 87.0 m- Comparisons: — Hislioneis [Parahistioneis] para, as conceived by Murray and \Miitting (1899), evidently includes two species, the one represented by their Plate 32, figure 4a and b, the other by their Plate 32, figure 4c. The specimen represented by the first two (fig. 4a, b) of these figures should be regarded as the type of this species. It is characterized especially by the fact that the hypotheca tapers posteriorly and that the entire left sulcal list is reticulated. The specimen represented by the last (fig. 4c) of these three figures may be referable to Para- histioneis paraformis, sp. nov., and is characterized by having the hypotheca sub- semicircular in lateral outline and by lacking reticulation in the left sulcal list except at the distal end of the posterior main rib. The specimen of P. para found in the material of the Expedition agrees quite closely with the type, except in the smaller relative depth of the body and in the smaller relative width of the cingular lists. We have no records of the structure of the cingular and sulcal lists except what is shown in Figure 85: 6. SYSTEMATIC ACCOUNT. 603 The closest-known relatives of Parahistioneis para presumablj' are P. pnra- formis and P. garretti (see these two species, the sections on comparisons). Synonymy: — This species was established by Murray and Whitting (1899, pi. 32, fig. 4a-c) under the name of Histioneis para. This specific name should, accordmg to our opinion, be applied to the specimen represented only by their Plate 32, figure 4a and 4b. Their Plate 32, figure 4c, appears to be referable to Parahistioneis paraformis (see the last section). Cleve (1901c) does not state whether he found both these forms or only one of them in his material. The specimen figured by Okamura (1912) as Histioneis para belongs to Parahistioneis paraformis. Occurrence: — Parahistioneis para is recorded at only one (4699) of the 127 stations, on the fourth line of the Expedition in the Easter Island Eddy, from 300-0 fathoms, at a surface temperature of 75°. The frequency is 1 %. Murray and Whitting (1899), who were the first to record this species, found it at sixteen stations in the Atlantic Ocean and in the Caribbean Sea, between lat. 33° 40' N. and lat. 13° 6' N. and between long. 31° 20' W. and long. 78° 44' W., in surface waters of 75° (62°-82°). Cleve (1901c) found it in the Atlantic, be- tween lat. 33° N. and lat. 24° N. and between long. 31° W. and long. 51° W.; temperature, 70.7°; salinity, 37.22. It should be remembered that this species, as conceived by Murray and Whitting (1899), appears to be heterogeneous (see the section on synonymy). Cleve (1901c) does not give any figures, and so his de- terminations cannot be checked. 3. Reticulata Group. Of the four members of this group, Parahistioneis dentata was not recorded from the material of the Expedition. The remaining ones have been treated in the following order: — P. karsteni, P. reticulata, and P. diomedeae. Parahistioneis karsteni (Kofoid and Michener) Plate 19, fig. 2. Figure 93:2 Histioneis karsteni Kofoid & Michener, 1911, p. 298. Diagnosis: — Body rounded carafe-shape in lateral outline; with fairly long, nearly erect anterior process ; hypotheca semicircular; length: depth, 1.09: 1. Dorsally the transverse furrow is rather strongly concave, 0.52 the depth of body, 2.5 times wider than ventrally, and inclined dorsoposteriorly at 40°-45°. Dis- tance from apex to posterior cingular list 0.41 the length of body. Anterior cingu- lar list with incipient stalk, flaring asymmetrically with twice as much dorsal as 604 THE DINOPHYSOIDAE. ventral flare and more posterior origin ventrally; its height 0.54 the depth of body. Posterior cingular list closed, lacks structural ditTerentiation; dorsal height, 0.57 the depth of body. Left sulcal list ends on dorsal side of body, 0.04 the depth of body from posterior cingular list; rather strongly convex just behind R2, acute at R3; R- 0.25 the depth of body; R3 arises 0.27 the depth of body from dorsal end of list, fairly weak, 0.67 the dei)tli of body; reticulated, reticulation weak behind R,. Length, 28.2 ja. Eastern tropical Pacific. Description: — This is a fairly small species, with body of rounded carafe- shape in lateral outline; with a rather long, slightly oblique anterior process, broadly rounded posteriorly and deepest in the midtlle. The anterior process is slightly deeper than long and has an anterodorsal inclmation of about 5°. The ratio between the length and the depth of the body is about L09: L The longi- tudinal axis is about perpendicular. The epitheca is about 0.35 as deep as the hypotheca, well convex, highest near the center, and tilted ventroposteriorly at 15°-20°. Dorsally the transverse furrow is rather strongly concave, about 0.52 the greatest depth of the body, 2.5 times wider than ventrally, and inclined dorsoposteriorly at 40°-45°; A-entrally it is gently concave. The distance from the apex to the posterior cingular list is about 0.41 the length of body. The anterior margin of the hj^potheca is gently concave and inclined dorsoposteriorly at 5°-10°. The dorsal, posterior, and ventral margins of the hypotheca are confluent and semicircular. The anterior cingular list has an hicipient stalk, is inclined anteriorly at about 40° and is approximately 2.5 times wider distally than proximally; it flares asymmetrically, arising farther posterior ventrally than dorsally, with straight ventral and concave dorsal side, twice as much dorsal as ventral extension, and top with slight posteroventral slope; its height is about 0.54 the greatest depth of the body; and it has on each valve five to six complete, subequidistant radial ribs, all of which are simple except the dorsal one, which has three branches. The posterior cingular list is closed, and its dorsal and ventral margins are subparallel to the longitudinal axis; its dorsal height is about 0.57 the greatest depth of the body; it has no ribs except the dorsal and ventral ones, and it lacks structural difTerentiation with the exception of a slight ventral reticulation. The left sulcal list is of medium size and ends on the dorsal side of the body at a distance from the posterior cingular list equaling about 0.64 the greatest depth of the body. The ventral margin is gently sigmoid in front of the fission rib; between this rib and the posterior main rib the margin is rather strongly convex anteriorly and SYSTEMATIC ACCOUNT. 605 gently sigmoid posteriorly; at the tip of the posterior main rib it is acute, forming an angle of about 85°; the dorsal margin is gently sigmoid, convex anteriorly and concave posteriorly. The fission rib is straight, about 0.25 the greatest depth of the body, and deflected posteriorly at about 25°. The posterior main rib arises at the antapex of the body and about 0.27 the greatest depth of the body from the dorsal end of the list, is fairly weak, gently sigmoid, and directed posteriorly; its length is about 0.67 the greatest depth of the body. Almost the entire list is re- ticulated, the meshes being of medium size to rather small; behind the fission rib the reticulation is very faint; in front of this rib it is rather pronounced. The thecal wall is porulate and deeply reticulated, the meshes being sub- uniform and medium-sized. The pores are located in the meshes, but by far the greater number of the meshes lack pores. On each valve about 15 meshes border the girdle posteriorly. There is no record of phaeosomes. The dimensions of the type specimen only were measured. Dimensions: — Type specimen: Length of bodj', 28.2 yu. Greatest depth of body, 25.8 fi. Total length, 56.G m- Co77iparisons: — Our description and figure are based on the type specimen. The shape of the body and the structure of the left sulcal list indicate that this species is rather closely related to Parahistioneis diomedeae and P. reticidata, two species from which it differs strikingly m being markedly smaller, in having the anterior cingular list characterized by an incipient stalk, and in the shape of the left sulcal list. Synomjmij: — This species was established by Kofoid and Michener (1911) as Histioncis karsteni. Occurrence: — Parahistioneis karsteni is recorded at only one of the 127 stations (4619) on the first line of the Expedition in the Panamic Area. Only one specimen was found in a Salpa taken in surface waters of 79°. The species has been found only in the material of the Expedition. It was first recorded by Kofoid and Michener (1911) from Station 4619 of the Expedi- tion, which thus is the type locality. Parahistioneis reticulata (Kofoid) Plate 19, fig. 7, 10. Figure 93: 7 Histioncis reticuhtla Kofoid, 1907a, p. 205, pi. 15, fig. 95. Okamura, 1012, p. 21, pi 3, fig. 5Ca, b. Diagnosis: — Body gourd-shaped in lateral outline; with long oblique an- terior process; hypotheca semicircular; length: depth, 1.15-1.35:1. Dorsally the transverse furrow is very strongly concave, 0.48-0.54 the depth of body, 1.5 606 THE DINOPHYSOIDAE. as wide as ventrally, and inclined dorsoposteriorly at 50°. Distance from apex to posterior cingiilar list 0.50 the length of body. Anterior cingular list sessile; its height 0.32-0.40 the depth of body. Posterior cingular list closed, a large part of it finely reticulated; dorsal height 0.34-0.43 the depth of body. Left sulcal list narrow; ends at or somewhat ventrally to antapex; strikingly sigmoid between R2 and R3; R2 0.12 the depth of body; R, arises 0.09-0.20 the depth of body from dorsal end of list, strong but not club-shaped, 0.27-0.37 the depth of body; reticu- lated. Length, 35.9-44.0 n. Tropical and subtropical Pacific. Description: — The body of this medium-sized species is gourd-shaped in lateral outline; with long, oblique anterior process, broadly rounded posteriorly and deepest in the middle. The anterior process is about 1.30 times longer than deep, and has an anterodorsal inclination of about 25°-30°. The ratio between the length and the depth of the body is 1.15-1.35: 1. The longitudinal axis is perpendicular. The epitheca is about 0.31 as deep as the hypotheca, well convex, highest in or near the center, and subhorizontal. Dorsally the transverse furrow is very strongly concave, 0.48-0.54 the greatest depth of the body, about 1.5 times wider than ventrally, and inclined dorsoposteriorly at about 50°; ventrally it is gently concave. The distance from the apex to the posterior cingiilar list is about 0.50 the length of the body. The anterior margin of the hypotheca is moderately and somewhat irregularly concave, and inclined dorsoposteriorly at 5°-10°. The dorsal, posterior, and ventral margins of the hypotheca are confluent and semi- circular. WTien seen in dorsoventral view, the body is carafe-shaped, widest in the middle, and strikingly angular at the junction of the transverse furrow and the hypotheca; the transverse furrow is fairly strongly concave; the antapex is rather narrowly rounded; the outlines of the hypotheca are moderately convex, and flattened or even slightly concave posteriorly; length: width, 1.42: 1. The anterior cingular list is sessile, of equal width and flare ventrally and dorsally, and with no posterior extension ventrally, inclined anteriorly at about 25°, and about 2.7 times wider distally than proximally; its height is 0.32-0.40 the greatest depth of the body; and it has on each valve about 11-14 subequi- distant, complete radial ribs, all of which are simple except the dorsal one which has two branches. The posterior cingular list is closed; the membrane that closes the midventral cleft has five equidistant transverse ribs; the dorsal and ventral margins are subparallel to the longitudinal axis, except anteriorly, where the list is flaring: the dorsal height is 0.34-0.43 the greatest depth of the body, the ven- SYSTEMATIC ACCOUNT. 607 tral height is slightly less; it has no ribs except the dorsal and ventral ones, but the larger portion of it is finelj' reticulated. The left sulcal list is relatively nar- row and ends at or somewhat ventrally to the antapex of the body. Its ventral margin is nearly straight in front of the fission rib; between this rib and the pos- terior main rib the margin is strongly sigmoid; at the tip of the posterior main rib it forms an angle of 40°-50°; the dorsal margin is gently convex, nearly straight. The fission rib is nearly straight, about 0.12 the greatest depth of the body, and deflected posteriorly at about 10°. The posterior main rib arises 0.09- 0.20 the greatest depth of the body from the dorsal end of the list, is strong but not club-shaped distally, deflected about 30° from the longitudinal axis, and 0.27-0.37 the greatest depth of the body. The entire list is strikingly reticulated, with the meshes of medium to rather small size. The thecal wall is porulate and areolate, the areoles being medium-sized. The pores are located in the areoles, but not all the areoles have pores. On each valve 20-25 areoles border the girdle posteriorly. Phaeosomes, some small and ellipsoidal, some large and globular, are found in the transverse furrow. The dimensions of the type only were measured. Dimensions: — Type specimen : Length of body, 44.0 ix. Greatest depth of body, 38.2 fi. Total length, 55.5 n (not 115 n, as stated in the original descrip- tion). The two specimens figured by Okamura (1912, pi. 3, fig. 56a, b) were 35.9-39.7 M long and 27.0-29.5 ^ deep. Cotnparisons: — Our description and figures are based on the type material. The two specimens represented by Okamura (1912, pi. 3, fig. 56a, b) as Histioneis reticulata (?) are figured in a very sketchy manner but nevertheless sufficiently detailed for certainty of specific assignment. Although they are somewhat smaller (35.9-39.7 II, as compared with 44.0 m) and narrower (ratio between length and depth, 1.33-1.35: 1, as compared with 1.15: 1) than the type specimen, their allocation undoubtedly is correct. With regard to the position of Parahistioneis reticulata within its genus, see P. diomedeae, which is the closest-known relative of this species. Synonymy: — This species was established by Kofoid (1907a) under the name of Histioneis reticulata, and this name was used also by Okamura (1912). Occurrence: — Parahistioneis reticidata is recorded at nine of the 127 stations. There are 2, 0, 0, 4, 3, and 0 stations on the six lines of the Expedition. Of these nine stations, two (4598,4604) are in the Mexican Current; two (4697, 4699) are in the Easter Island Eddy ; and five (4701, 4711, 4722, 4733, 4734) are in the South Equatorial Drift. At one station (4604) the species was taken in a surface haul; 608 THE DINOPHYSOIDAE. at one station (4733) it occurred in a Salpa taken in surface waters; the remaining records refer to hauls from 300-0 fathoms. The temperature range of these nine stations at the surface was 72°-84°; the average was 77.9°. At the two stations in the surface waters of which this species was found, the surface temperatures were 84° and 80°. The frequency is less than 1 % except at two stations (4598, 4722), where it is 1%. The species was first recorded by Kofoid (1907a) from Station 4699 of the Expedition which thus is the type locality. Later it was found by Okamura (1912) in Japanese watei-s. Its absence from the California and PeruAian Currents is noteworthy. Parahistioneis diomedeae (Kofoid and Michener) Plate 19, fig. 4, 5. Figure 93: 6 IliKliimrU lUninciknc Kofoid & MiciiENEii, 1911, p. 294. Diagnosis: — Body asymmetrically gourd-shaped in lateral outline, with long, oblique, dorsally deflected anterior process; hypotheca semicircular; length: depth, 1.11:1. Dorsally the transverse furrow is very strongly concave, 0.43 the depth of body, twice as wide as ventrally, and inclined dorsoposteriorly at 55°. Distance from apex to posterior cingular list 0.47 the length of body. Anterior cingular list sessile, of equal length and flare dorsally and ventrally; its height 0.30 the depth of body. Posterior cingular list closed, in part finely reticulated; dorsal height 0.3G the depth of body. Left .sulcal list ends on dorsal side of body, 0.47 the depth of body from posterior cingular list; with two lobes, one rounded, subrectangular at R,-, one narrowly rounded at Rs; R2 0.33 the depth of body; R3 arises 0.54 the depth of body from dorsal end nf list, club-shaped distally, 0.68 the depth of body; reticulated. Length, 53.5 p. Eastern tropical Pacific. Description: — This is a large species with the body asynnnetrically gourd- shaped in lateral outline, with long, oblique anterior process, broadly rounded posteriorly, and deepest in the middle. The anterior process is about 1.25 times longer than deep and has an anterodorsal inclination of about 20°-25°. Th(> ratio between the length and the depth of the body is about 1.11:1. The longitudinal axis is perpendicular. The epitheca is about 0.40 as deep as the hypotheca, well convex, somewhat higher dorsally than ventrally, and tilted ventroposteriorly at 15°-20°. l')orsally the transv(>rse furniw is very strongly concave, 0.43 the greatest depth of the SYSTEMATIC ACCOUNT. 609 body, about twice as wide as ventrally, and inclined dorsoposteriorly at about 55°; ventrally it is gently concave. The distance from the apex to the posterior cingular list is about 0.47 the length of the body. The anterior margin of the hypotheca is rather strikingly undulating and is inclined dorsoposteriorly at 5°- 10°. The dorsal, posterior, and ventral margins of the hypotheca are confluent and semicircular. When seen in dorsoventral view, the body is widest and strongly convex in the middle and has narrowly rounded apices, the outline being gently concave anteriorly and posteriorly; the ratio between the length and the width is 1.58:1. The anterior cingular list is sessile, of equal length and flare dorsally and ven- trally, inclined anteriorly at about 20°, and about twice as wide distally as proxi- mally; its height is 0.30 the greatest depth of the body; and it has on each valve about 12-14 subequidistant, complete, radial ribs and a number of incomplete, radial marginal ribs; of the complete ribs all are simple except the dorsal one, which has about six branches (Plate 19, fig. 4). The posterior cingular list is closed, and its dorsal and ventral margins are subparallel to the longitudinal axis; its dorsal height is about 0.36 the greatest depth of the body; it has no ribs except the dorsal and ventral ones, and is finely reticulated dorsally and ventrally. The right sulcal list is crescent-shaped and extends to a point somewhat beyond the fission rib of the left sulcal list ; its greatest width is about 0.09 the greatest depth of the body, and it is finely reticulated. The left sulcal list ends on the dorsal side of the body at a distance from the posterior cingular list equaling about 0.47 the depth of the body. It is characterized by having two lobes, one rounded, sub- rectangular at the fission rib, the other narrowly rounded at the tip of the poste- rior main rib. The ventral margin of this list in front of the fission rib is almost straight throughout ; between the fission rib and the posterior main rib the margin is moderately concave ; and the dorsal margin is gently sigmoid, con\-ex anteriorly, and concave posteriorly. The fission rib is nearly straight, about 0.33 the greatest depth of the body, and deflected posteriorly at about 15°-20°. The posterior main rib arises somewhat ventrally to the antapex and about 0.54 the greatest depth of the body from the dorsal end of the list, is club-shaped distally, directed posteriorly, and measures about 0.68 the greatest depth of the body. Dorsally to the posterior main rib this list is finely reticulated; its remaining portion is characterized by a somewhat coarser reticulation except anteriorly, and ven- trally to the posterior main rib where no structural differentiation is to be found. The thecal wall has numerous pores, each of which is surrounded by an areole of moderate and subuniform size. On each valve about 30-35 areoles border the 610 THE DINOPHYSOIDAE. girdle posteriorly. Phaeosomes, some small and ellipsoidal, some large and globu- lar, are found in the transverse furrow. The dimensions of the type only were measured. Dimensions: — Type specimen: Length of body, 53.5 ^l. Greatest depth of body, 48.2 m- Total length, 91 n. Compariso)is: — Our description and figures are based on type material. The closest-known relative of this species undoubtedly is Parahistioneis reticulata, which it resembles in the peculiar gourd-like shape of the outline of the body in lateral view and in the shape and structure of the cingular lists. It is very easily distinguished from this species by its larger size, and by its large, two-lobed left sulcal list, the posterior main rib of which is very long and club-shaped dis- tally. Another close relative is Hislioneis [Parahistioneis] dentata Murray and Wliitting (1899, pi. 33, fig. 4a, b), which it resembles especially in the shape of the body in lateral as well as in dorsoventral view, and to a less degree in the shape of the posterior portion of the left sulcal list. Parahistioneis dentata differs strikingly from the present species in the peculiarly serrated anterior edge of the anterior cingular list and in having the anterior margin of the left sulcal list thickened. The shape of its body and the structure of its left sulcal list also suggest com- paratively close relationship to P. karstcni, a very small species lacking a lobe in the left sulcal list at the fission rib and having a slightly stalked anterior cingular list. Occurrence: — Parahistioneis diomedeae is recorded at four of the 127 stations. There are 1, 0, 0, 1, 2, and 0 stations on the six lines of the Expedition. Of these four stations, one (4619) is in the Panamic Area; one (4699) is in the Easter Is- land Eddy; and two (4720, 4724) are in the South Equatorial Drift. At one sta- tion (4720) the species was taken in a sin-face haul; at one station (4619) it oc- curred in a Salpa taken in surface waters ; the two remaining records refer to hauls from 300-0 fathoms. The temperature range of these four stations at the surface was 75°-79°; the average was 77.3°. At Stations 4619 and 4720 the temperature was 79° and 76°, respectively. The frequency is less than 1 % except at Station 4619 where it is 2%. The species has been found only in the material of the Expedition. It was first recorded by Kofoid and Michener (1911) from Station 4720. The type speci- men is from Station 4699. SYSTEMATIC ACCOUNT. 611 HisTiONEis Stein HistioneisSTEiN, 1883, p. 25. But.schli, 1885, p. 943, 1010. Schutt, 1896, p. 29. Delage & H£rouard, 1896, p. 386. Histoneis Ostenfeld & Schmidt (lapsus pennae), 1901, p. 173. Diagnosis: — Body of diverse shapes, usually deeper than long; ratio be- tween length and depth, 0.32-1.33: 1. Epitheca small, low, disk-like; its depth 0.07-0.24 the depth of body. Transverse furrow wide dorsally and decidedly wider dorsally than ventrally; dorsal width 0.37-0.77 the depth of body; ventral width 0.09-0.50 the dorsal width. Cingular lists large, closed; anterior funnel-like and stalked (except in Histioneis paulseni), its dorsal height 0.57-2.00 the depth of body, its structure variable ; posterior usually subcylindrical, its dorsal height 0.47-1.10 the depth of body, and always with one dorsal and one ventral pair of ribs and one cross-rib. Left sulcal list usually large, of diverse shapes, in most species narrowly to broadly tongue-shaped, its greatest width varying from 0.30 to 2.14 the depth of body, and never with more than one main rib behind fission rib. Type: — We designate as the type species of the genus, Histioneis remora Stein (1883, explanation of hLs pi. 22, fig. 11). Organology: — The body (theca) in Histioneis is more diverse in shape than in any of the other large genera of Dinophysoidae, ranging from quite simple to very aberrant types. WTien seen laterally, it always is more or less asymmetrical, due largely to the shape of the transverse furrow and to the posterodorsal inclination of the longitudinal axis. The hypotheca, however, is symmetrical or subsym- metrical in some of the primitive species. In dorsoventral view the body appears always to be symmetrical just as in the other genera of this tribe. The minimum posterodorsal inclination of the longitudinal axis (La., Figure 94) is about 5°-10° {H. inornata and H. inclinata; Figure 95: 10, 13); the maximum inclination is 50°-60° {H. josephinae, Figure 9(): 7). The inclination, which is somewhat vari- able even within the species, is diflficult to determine with accuracy, since the antapex is not clearly marked in most species. The information given on this point in our descriptions of species should thus be regarded as approximative. WTien seen laterally, the body sometimes {e.g., Histioneis inclinata, Figure 95: 13; H. remora. Stein, 1883, pi. 22, fig. 11) is about as long as deep, but usually it is deeper than long. In the species with the most pronounced dorsoventral elongation (H. mitchellana, Figure 96:4), the ratio between the length and the depth of the body is 0.32-0.53: 1. H. panda, in which this ratio is about 0.36: 1, has a curved, sausage-shaped body with an average height only about 0.23 the depth, measured along the curvature of the body (Figure 95: 9). In only two of the known species {H. biremis and H. highkyi; Figure 95: 11, 14) the body is 612 THE DINOPHYSOIDAE. longer than deep; in the latter species the ratio between the length and the depth may be as high as 1.33: 1. There is no shape of body that can be said to be tj'pical of this genus. In most of the more or less primitive and in some of the highly differentiated species, int. dug. t. pos(. //(. r. incl. post. m. r. Figure 94. — Terminology and methods of measurement in Histioneis. Terminology: — The midUne is the Une drawn through the center of the epitheca and the central axis of the proximal portion of the anterior cingular list. If the longitudinal axis of the body were not inclined, it would coincide with the midline. In tlie specimen represented by this figure the longitudinal axis of body is inclined po.'iilchellana (Plate 21, fig. 2). In others of these highly developed forms this list may have, besides some irregular and some- times reticulated ribs, a submarginal rib with a series of subequidistant anterior branches {H. reginella and H. hippoperoidcs; Plate 20, fig. 9; Plate 23, fig. 1, 3). Or, again, it may be nearly without structural differentiation, as in H. longicolUs (Plate 21, fig. 5). In all these advanced species the ribbing is limited to the distal portion of the list; the stalk lacks structural differentiation. It may finally be mentioned that in some species {e.g., H. costala and H.josephinae; Plate 20, fig. 3; Plate 22, fig. 2) the tips of the ribs extend slightly beyond the edge of the list. When fully developed, the posterior cingular list (post. cing. I., Figiu'e 94) is closed except posteroventrally, where it probably always opens hito the canal formed by the right sulcal list. The presence of this canal has not been established as yet, but appears very probable on account of the peculiar structure of the right 618 THE DINOPHYSOIDAE. sulcal list (see Plate 20, fig. 7, of Histioneis elongnta). In some of the specimens figured by us (H. carinata, H. biremis, and H.josephinae; Figure 95: 1, 11; Figure 96: 7) the posterior cingular list has a ventral split, but this is presumably but a temporary condition, possibly attending fission, since specimens with and with- out such a split have been found in the same species {H. hiremis). In the most primitive species {H. costata, H. paulseni, H. inclinata; Figure 95:4, 8, 13) this list forms q cylindrical, upright collar of moderate height, about as high ventrally as dorsally or but slightly lower ventrally than dorsally, with nearly straight or but slightly sigmoid dorsal, ventral, and lateral outlines, and with the anterior opening subequal to the basal circumference. The minimum height on the dorsal side is about 0.47 the greatest depth of the body (H. paulseni, Figure 95 : 8). From this fundamental tj^pe a fairly great number of variants have evoh'ed. In some species this list has maintained its original shape but has increased very strikingly in height ; as examples of this kind we may mention H. longicollis and H. mitchel- lana (Figure 95: 7; Figure 96: 4), in which the dorsal height is 0.91 and 1.10 the greatest depth of the body, respectively. In other species the main change con- sists in the decided difference between the dorsal and ventral heights; for instance, in H. highkyi (Figure 95: 14) the average dorsal height is 0.62, the average ven- tral height 0.37 the greatest depth of the body. In H. dongata (Figure 96: 3) this list is not only higher than in the primitive species, but it is characterized also by a pronounced difTerence between the dorsal and ventral heights, the dorsal height being about 0.95, the ventral 0.72 the greatest depth of the body. The greatest dorsal height recorded as yet, viz., 1.40-1.42 the greatest depth of the bod}^, was found in H. dolon (Figiu-e 96 : 6), a species in which this list is character- ized also by other changes (see below). In H. carinata (Figure 95: 1) this hst grows narrower anteriorly; and this peculiarity reaches its most striking development in H. panda (Figure 95:9), which is characterized by a very narrow anterior opening and a very large basal circumference. In other forms (e.g., H. navicula. Figure 95: 3) it is more or less strongly constricted just behind the anterior margin, which thus becomes flaring. H. inornala (Figure 95 : 10) is characterized by having the anterior opening of this list wider than the basal circumference. In some species the dorsal margin of this list is evenly or irregularly concave {H. josephinae and H. hippoper aides; Figure 96: 7, 5); in others it is more or less bulging {H. navicula. Figure 95: 3). Perhaps the most important change in the shape of this list is due to the develop- ment of lateral pouches, one on each side. The pouches are fairly small in some species {e.g., in H. pulchra and H. dolon; Plate 21, fig. 4; Murray and Whitting, SYSTEMATIC ACCOUNT. 619 1899, pi. 33, fig. 5b); in others {e.g., H. hippoperoides, Plate 23, fig. 1) they are of medium size; and they reach their greatest size in H. panda, H. pnnaria, and H. reginella (Figure 95: 9; Figure 85: 8; Plate 20, fig. 11). In H. panaria the pouches are broadly rounded in dorsoventral \'iew, and each of them is slightly wider than the body; in H. reginella they are fairly narrowly rounded, and each of them is slightly narrower than the body. The structural differentiation of this list is simple and uniform in the primi- tive species. These have only a pair of simple, dorsal ribs, one rib on either side of the sagittal suture, a similar pair ventrally, and a simple cross-rib running sub- marginally to the anterior edge of the list, turning posteriorly on the ventral side of it, and connecting with the dorsal ribs. The portion of the list in front of the cross-rib is called the frill. From this fundamental type, which can be recognized even in the most advanced members of the genus, some variants have evolved. First, there is a tendency among the advanced species for the frill to become pro- gressively higher. For example, in H. mitchellana (Figure 96:4) it is decidedly larger than the rest of the list. The enlargement of the frill, which is more pro- nounced on the dorsal side of the list than in the middle, and which progresses until the cross-rib issues from near the base of the dorsal ribs (Figure 96: 6), is in some species (e.g., H. pacifica) the only change to be recorded. Second, the cross- rib tends to be branched. Branches may be developed only on its anterior side, in other words in the frill (e.g., H. panaria, Figure 95: 6), or on both its anterior and its posterior sides (e.g., H. reginella, Figure 95: 2). The branches of this rib are first simple and nearly straight, but they tend to become irregular and branched, and to form a wide-meshed or a fine-meshed reticulum (H. pulchra, Figure 96: 2; //. mitchellana, Plate 21, fig. 2). Third, accessory ribbing usually is confined to the frill (H. mitchellana, Plate 21, fig. 2) and is always better devel- oped anteriorly than posteriorly. Irregular ribs may develop posteriorly; for in- stance, on the lateral pouches of H. reginella (Plate 20, fig. 9, 11) and from the postmargin of this list in H. hippoperoides (Plate 23, fig. 1). Fourth, in some of the species with a high frill a secondary frill is present (H. panda, Figure 95: 9; and the species of the Dolon group. Figure 96:6). The cross-rib that borders this secondary frill may be either simple or branched, and the branches may be present only on the anterior side (H. dolon, Figure 96: 6), or anteriorly as well as posteriorly [H. hippoperoides, Figure 96: 5). The right sidcal list is known in two species only, viz., in Histioneis elongata (Plate 20, fig. 7) and in H. josephinac (Figure 96: 7), and in these it is verj' small and difficult to detect. In the former species, which may present the condition 620 THE DINOPHYSOIDAE. FiGDRE 96.— Histioneis, right lateral view. X 700. 1, H. striata Kofoid and Michoner, Station 4720 (Salpa); 2, //. pulchra Kofoid, Station 4730 (300-0 fathoms); 3, H. elongata Kofoid and Michener, SYSTEMATIC ACCOUNT. 621 primitively typical of this genus, the ventral margin of this list is attached to the base of the left sulcal list. Thus a canal is formed, which opens posteriorly at the fission rib of the left sulcal list and presumably also into the girdle. The water caused by the movements of the transverse flagellum to enter the girdle probably passes out through this canal. In H. josephinae the right sulcal list continues be- yond the margin of the theca to a distance slightly greater than from that point to the flagellar pore ; it is very narrow at this pore, lanceolate posteriorly, and has a rib along the dorsal margin. The left sulcal list (I. side. I., Figure 94), which usually is rather large, has undergone very profound structural modifications and affords ^'ery important taxonomic features; indeed, the systematic subdivisions of this genus suggested by us are to a very large extent based on differences in this list. In only two of the known species {Histioneis costata and H. mitchellana; Figure 95: 4; Figure 96: 4) does this list extend slightly beyond the antapex of the bod}'; in the remaining species it ends somewhere between the antapex and the point midway between the antapex and the flagellar pore. In H. inclinata (Figure 95: 13), which is characterized by having this list smaller than in any of the other members of this genus, this list ends about midway between the antapex and the flagellar pore, and its maximum width is about 0.30 the greatest depth of the body. In H. dolon (Figure 96: 6), which has this list of an extraordinary size, it ends at about the same place as in H. inclinata, but its width at the fission rib and at the posterior main rib equals not less than 1.45-1.61 and 1.70-1.80 the greatest depth of the body. Finally, in H. elongata (Figure 96:3), which has this list unusually long and narrow, the posterior main rib of this list measures not less than about 2.14 the greatest depth of the body (in the case of the fission rib of this species, the corresponding value is not more than about 0.57). In one species (H. reginella, Figure 95 : 2) the posterior portion of this list turns anteriorly on the right side of the body and ends near the posterior end of the longitudinal furrow. In all the known species except one (H. inclinata, Figure 95: 13) the portion of this list which is in front of the fission rib is decidedly smaller than the posterior; and it is this posterior portion which has been especially subject to structural differentia- tion. The shape of this list is exceedingly variable, as will be seen from the follow- ing representative examples. In H. inornata (Figure 95: 13) this list is very short, not extending even to the level of the antapex of the body, of subuniform width Station 4722 (300-0 fathoms); 4, H. mitchellana Murray and Whitting, Station 4699 (300-0 fathoms); 5, H.hipimpcroiiJts Kofoitl and Miohener, Station 4590 (300-0 fathoms) ; 6, H.Joloii Murray and Whitting, Station 4097 (300-0 fathoms); 7, H. josephiiuie Kofoid, Station 4699 (300-0 fathoms). 622 THE DINOPHYSOIDAE. throughout and very narrow, its maximum width being only about 0.30 the greatest depth of the body; its ventral margin is gently sigmoid, concave anteri- orly, and convex posteriorly ; and its postmargin is subcircular. In H. indinata (Figure 95: 10) this list differs from that of the latter species mainly in being slightly longer, extending somewhat beyond the level of the antapexof the bodj^ and in having the postmargin sigmoid instead of subcircular. The left sulcal list of H. navicula (Figure 95:3) Is also comparatively small, ending somewhat ventrally to the antapex and ha\'ing a maximum width, at the posterior main rib, only 0.3G the greatest depth of the body; its ventral margin is moderately and evenly convex, its dorsal margin gently concave, and these margins form together an angle of 60°-70° at the posterior main rib. In H. cosiata (Figure 95: 4) this list is fairly large, extending slightly beyond the antapex and with a maximum width (at the posterior main rib) about 0.67 the greatest depth of the body, and squarish in shape, its margin being angular somewhat behind the fission rib as well as at the posterior main rib. This is the only example of biangularity known in this genus. In many species (e.g., H. carinata, H. hyalina, H. paulseni, H. pacifica, Figure 95: 1, 5, 8, 12; H. pulchra, H. elongala, Figure 90: 2, 3) this list is large and of a narrow tongue-shape ; its ventral margin is gently convex or sigmoid ; its dor- sal margin, gently concave, straight, gently convex, or sigmoid; and posteriorly, i.e., at the posterior main rib, it is acute to subacute, or narrowly to rather broadly rounded. Finally, in some species {e.g., H. hippoperoides and H. dolon, Figure 96 : 5, 6) this list is very large (see above) and broadly tongue-shaped ; and in H. josephinae (Figure 96: 7) it is very large, broadly rounded posteriorly, and furnished with a narrowly rounded dorsal projection, the length of which is ap- proximately 0.6 the greatest depth of the body. Originally this list was l)ut slightly or gently curved, and this condition is still characteristic of most of the known species (see, for instance, H. paulseni and H. costata, Plate 20, fig. 2, 10; H. carinata and H. naiicula, Plate 21, fig. 1, 3; H. elongata, Plate 22, fig. 4). Sometimes, however, its entire dorsal margin is strongly curled to the right (//. hyalina. Figure 95:5; H. striata, Figure 96: 1). This peculiarity is further ad- vanced in H. pulchra (Figure 96: 2), in which the middle portion of this margin forms a fairly large, narrowly rounded lobe. It reaches its highest development in H. mitchellana (Figure 96 : 4) ; in this species the lobe is represented by a very circumscribed, tongue-shaped to subcircular flap, the dorsoventral width of which is 0.34-0.38 the greatest depth of the body. In H. navicula (Plate 21, fig. 0) the ventral margin is turned to the right ; and in H. reginella (Plate 20, fig. 9) the en- tire list is strongly curled to the right. SYSTEMATIC ACCOUNT. 623 In the species of the Dolon group this list is characterized by one to three accessory fins. When only one accessory fin is developed {H. dolon and H. hippo- peroides; Figure 96: 6, 5), this is always posterodorsal and transverse in position. Its shape usually is rounded (H. dolon, Murray and Whitting, 1899, pi. 33, fig. 5b) but may be pear-like {H. milneri, Murray and Whitting, 1899, pi. 33, fig. lb). It is either (H. dolon) very large, with a diameter not less than 1.64 the depth of the body; or {H. milneri) of moderate size, its diameter being only about 0.70 the depth of the body. More than one accessory fin is present in only one of the known species, viz., H. josephinae (Figure 96: 7). One of these fins arises along the origuially free ventral edge of the list in front of the fission rib, one from the left side of the fission rib, and one, or perhaps better a pair, transversely from the posterior main rib; the first two fins are large and squarish, the last is large and shaped like a pair of butterfly wings. In no species of this genus does this list have more than two main ribs, viz., the fission rib and the posterior main rib. The fission rib (/. r., Figure 94) origi- nally was straight and subhorizontal, and this condition still prevails in some of the primitive species {H. inornata and H. inclinata; Figure 95: 10, 13). Usually, however, it is more or less deflected posteriorly, and this progressive deflection has reached its maximum in H. mitchellana (Figure 96: 4), in which the anterior member of this rib is du'ected straight posteriorly. Originally the two members of this rib were lying close together along their entire extension, but in the species of the LoNGicoLLis and Pulchra groups the posterior member is recurved dorsally and joins the posterior main rib near the middle. In the primitive and in some of the advanced species the fission rib is very short, its minimum length being about 0.12 the greatest depth of the body (H. )un>icula, Figure 95: 3); but it shows a tendency to become progressively longer and in H. dolon (Figure 96 : 6) it reaches its maximum length, viz., 1.61 the greatest depth of the body. The posterior main rib (post. m. r., Figure 94), which is absent in only one of the known species [H. inclinata, Figure 95: 13), usually is marginal or submarginal, enforcing the dorsal edge of the list. Only in one species {H. josephinae, Figure 96: 7) does it run far from the dorsal margm, and even in this species it arises near the dorsal end of the Ust. As a rule, this rib extends to the posterior end of the list, but it may be only about half as long as the posterior width of the list (H. josephinae, Figure 96: 7). Its length and strength are very variable ; it may be weak and only about 0.22 the greatest depth of the body {H. inornata. Figure 95: 10), or strong and not less than 2.14 the greatest depth of the body. In only one of the known species (H. costata, Figure 95: 4) is it deflected posterodorsally (at about 15°); in the remain- 624 THE DINOPHYSOIDAE. ing ones it is more or less deflected posteroventrally. Its shape is variable ; it may be almost straight {H. costata, Figure 95: 4), gently sigmoid {H. carinata, Figure 95: 1), or gently to fairly strongly concave dorsally {H. elongata and H. striata; Figure 96: 3, 1). Besides the main ribs this list may have the following structural differentia- tions. First, a few short (incomplete) ribs between the fission rib and the pos- terior main rib {H. patilseni, Figure 95: 8). Second, one to four submarginal ribs along the ventral margin in front of the fission rib {H. hipjwperoides and H. dolon; FigiH'e 96:5, 6); and one marginal or one submarginal rib along the ventral margin behind the fission rib {H. pavlsnn and H. elongata; Figure 95: 8; P'igiu'e 96: 3). Third, a number of radial ribs behind the fission rib {H. inclinata, Figure 95: 13). Fourth, a wide-meshed or a fine-meshed reticulum. The reticulum varies in extension ; sometimes it is confined to the posterior end of the list, where it may be very heavy {H. elongata, Figin-e 96: 3) ; sometimes it covers the larger portion of it (//. hippoperoides, Figure 96: 5; //. mitchellana, Plate 21, fig. 2). Fifth, the posterior transverse fin of most of the species of the Dolon group is characterized by a concentric striation. Sixth, in H. josephinae (Figure 96: 7) this list is char- acterized by peculiar coral-like thickenings. Parasagittal lists have thus far not been found in this genus. The flagellar pore is a rounded or slightly elongated opening of moderate to rather small size, located to the right in the transverse furrow, i.e., on the right valve, at or slightly behind the posterior cingular list; its maximum diameter is about 3 IX. There are no records of pores of the kind found in Phalacroma (Plate 3, fig. 1, 2, of P. pulchrum and P. giganteum) on the ventral side of the left valve, near the sagittal suture, and just in front of the anterior cingular list. The longitudinal flagellum has not been recordetl as j-et. The transverse flagellum has been observed in one species only, viz., in Histioneis mitchellana (Plate 21, fig. 2). In this species it is short, undulating, arises in the flagellar pore, and encircles the narrow base of the anterior cingular list from the right around to the left. The last peculiarity is unquestionably due to a dislocation brought about when the specimen was killed in the preserving fluid (formalin), since in all other known cases this flagellum encircles the body frcjm the left around to the right and since it, on fixation, contracts into a close spiral whicli is thrown into various positions out of the normal one in the girdle. In most of the primitive species and in the members of the Remora and BiKEMis groups, the thecal wall is areolate or faintly reticulate as well as porulate. The meshes of the reticulation always are fairly small, subec[ual, and roinided SYSTEMATIC ACCOUNT. 625 to angular. In some of these forms nearly e\'ery mesh or areole has a central pore {Histioneis carinata, Plate 21, fig. 8); but in others {H. paulseni, Plate 20, fig. 1; H. inclinata, Plate 22, fig. 5) the number of pores is decidedly less than the number of meshes or areoles. In H. inornota, H. reginella, and H. panaria, and in all the members of the highly differentiated species of the Longicollis, Pulchra, and DoLON groups, on the other hand, the thecal wall is porulate but lacks every indication of reticulation and areolation. In these species, too, the number of pores is decidedly variable. In some (e.g., H. striata, Plate 20, fig. 6) it is very small, in others (e.g., H. josephinae, Plate 22, fig. 1, 2) large, ^\^len a great num- ber is present, there is usually a close-set row on either side of the posterior cingu- lar list and of the sagittal suture, as well as scattered pores. Originally the theca probably was areolate and porulate. In some species pores have not been ob- served, but they probably are always present. Very little is known about the protoplasmic contents. The nucleus, which has a moniUform chromatin reticulum, may be small, rounded, and located near the center of the body (Plate 22, fig. G) ; or it may be fairly large, ovoidal, and located posterodorsally (Schiitt, 1895, pi. 5, fig. 23). At least one large pusulc opens into the flagellar pore (Plate 22, fig. 6). Chromatophores appear always to be absent. Ovoidal or round phaeosomes, of small or medium sizes, are frequently found in the gu-dle. Sometimes they occur in \'ery large numbers (H. highleyi, Figure 98: 1). Their color is green, pale olivaceous, or yellowish green. The en- larged girdle serves throughout as a generalized phaeosome chamber in this genus. The length of the body ranges within the genus from 13.3 m (Histioneis inor- 7iata) to 88.7 n (H. hiremis) ; in our material the maximum length is G6.5 y. (H. highleyi) . Distribution: — The species of Histioneis may be compared in some particu- lars with the orchids and the hummingbirds; they are very multiform, alserrant, and beautiful in shape, brilliant in colors, exceedingly rare, and reach their highest development in the tropics. Like the species of Ornithocercus, Parahistioneis, Amphisolenia, and Triposolenia, they contribute to the striking contrast between the plankton of the warm equatorial regions and that of the cold polar seas. The genus is marine, of circumequatorial distribution, and appears to avoid coastal waters even more than does any other genus of Dinophysoidae. Most of its species seem to be restricted to tropical and subtropical seas; only a few of them occur in the Mediterranean. In the Atlantic Ocean as well as in the Pacific, its northern limit appears to be about lat. 35° N.; the limits of its southern extension are unknown not only in the Atlantic and Indian Oceans but also in the Pacific, 626 THE DINOPHYSOIDAE. since it was found by us at our southernmost station. The general distribution of the genus as a whole as well as of the individual species is very uncertain, due to the scarcity of distributional data available at the present tune. This paucity of data undoubtedly arises from the extreme rarity of these highly specialized organisms, and in part from their small size. Our knowledge of the vertical distribution of the genus is also very incom- plete. Nearly all of the records of this genus published at the present writing are from surface hauls. No records based on catches made ^^ith closing nets have been pubUshed as yet. Every one of the species is very rare. Representatives of Histioneis were found at thirty-eight (29.9%) of the 127 stations of the Expedition from which dinoflagellates were recorded. These thirty-eight stations are distributed over the whole area covered by the Expedi- tion in the following manner (Plate 30) : — 0 (0.0%) out of the 4 stations in the Cahiornia Current 2 (15.4%) out of the 13 stations in the Mexican Current 2 (11.8%) out of the 17 stations in the Panamic Area 0 (0.0%) out of the 27 stations in the Peruvian Current 5 (50.0%) out of the 10 stations in the Easter Island Eddy 0 (0.0%) out of the 4 stations in the Galapagos Eddy 27 (60.0%) out of the 45 stations in the South Equatorial Drift 2 (66.7%) out of the 3 stations in the South Equatorial Current 0 (0.0%) out of the 2 stations in the Equatorial Counter Current 0 (0.0%) out of the 2 stations in the North Equatorial Current The numbers of these thirty-eight stations are as follows : — 4590, 4604 (Mexican Current); 4619, 4623 (Panamic Area); 4689, 4691, 4695, 4697, 4699 (Easter Island Eddy); 4679, 4680, 4681, 4683, 4685, 4701, 4705, 4707, 4709, 4711, 4717, 4719, 4720, 4721, 4722, 4724, 4725, 4728, 4730, 4732, 4733, 4734, 4736, 4737, 4739, 4740, 4741 (South Equatorial Drift) ; 4742,4743 (South Equatorial Current) . At twenty-eight of these thirty-eight stations Histioneis was taken in vertical hauls and at ten in surface hauls. Most of the vertical hauls are from 300-0 fathoms, some are from 800 and 100-0 fathoms. Vertical hauls were made at only sixty-eight out of the 127 stations. The genus was thus found at 41.2%, of the stations at which vertical hauls were made. Disregarding the stations at which surface catches only were made, the record stations are distributed over the area covered by the Expedition in the following manner : — 0 (0.0%) out of the 4 stations in the California Current 1 (20.0%) out of the 5 stations in the Mexican Current 0 (0.0%) out of the 6 stations in the Panamic Area SYSTEMATIC ACCOUNT. 627 0 (0.0%) out of the 22 stations in the Peruvian Current 5 (100.0%) out of the o stations in the Easter Island Eddy 0 (0.0%) out of the 2 stations in the Galapagos Eddy 21 (91.3%,) out of the 23 stations in the South Equatorial Drift 1 (100.0%) out of the 1 station in the South Equatorial Current There were no vertical hauls made in the Equatorial Counter Current and in the North Equatorial Current. The numbers of these twenty-eight vertical stations are as follows : — 4590 (Mexican Current); 4689, 4691, 4695, 4697, 4699 (Easter Island Eddy); 4679, 4681, 4683, 4685, 4701, 4705, 4707, 4709, 4711, 4717, 4719, 4721, 4722, 4724, 4730, 4732, 4734, 4736, 4737, 4739, 4740 (South Equatorial Drift) ; 4742 (South Equa- torial Current). Surface catches and stomachs of Salpa taken in surface waters (81 sm-face catches and 24 samples of Salpa stomachs) were examined from eighty-two sta- tions. As previously mentioned, the genus was taken at ten surface stations, i.e., at 12.2% of all the surface stations. Taking into consideration the surface sta- tions only, these ten record stations are distributed in the following manner: — 0 (0.0%) out of the 4 stations in the California Current 1 (11.1%) out of the 9 stations in the Mexican Current 2 (15.4%) out of the 13 stations in the Panamic Area 0 (0.0%) out of the 20 stations in the Peruvian Current 0 (0.0%) out of the o stations in the Easter Island Eddy 0 (0.0%) out of the 2 stations in the Galapagos Eddy 6 (26.1%) out of the 23 stations in the South Equatorial Drift 1 (50.0%) out of the 2 stations in the South Equatorial Current 0 (0.0%) out of the 2 stations in the Equatorial Counter Current 0 (0.0%) out of the 2 stations in the North Equatorial Current The numbers of these ten surface stations are as follows : — 4604 (Mexican Current) ; 4619, 4623 (Panamic Area) ; 4680, 4720, 4725, 4728, 4733, 4741 (South Equatorial Drift); 4743 (South Equatorial Current). As will be seen from these tables, the distribution of Histioneis in the area examined by the Expedition is charactized by the following outstanding facts : — il) the genus is completely absent from the California and Peruvian Currents, two currents of temperate origin; (3) it is also absent from the Galapagos Eddy, which is under the direct influence of the Peruvian Current, and from the Equa- torial Counter Current and the North Equatorial Current ; (3) it is very rare in the Mexican Current and the Panamic Area, for in these regions only four out of the twenty-four species of this genus were found, and there are only four record stations out of thirty stations; (i) in the Easter Island Eddy and the South 628 THE DINOPHYSOIDAE. Equatorial Current, on the other hand, the genus occurred at all the vertical stations, and in the South Eciuatorial Drift it was found at no less than twentj-- one (91.3%) out of the twenty-three vertical stations. These facts indicate that the genus Histioneis is limited in its distribution largely by the temperature, and that its species are stenothermal, warm-water forms. The fact that most of these species avoid the coasts suggests that we are dealing with stenohaline organisms limited to high and uniform salinities. The causes of the rarity of this genus in the Mexican Current and in the Panamic Area are obscure. It may be that they should be sought in the unusually tiu'bu- lent current conditions of these regions (compare the distributions of Ornitho- cercus heteroporus, 0. splendidus, and 0. carolinae). There are ninety-four records of species of Hi-stioneis from vertical catches. Out of these ninety-four records, five (5.5%; Stations 4590, 4730 [two records], 4732 [two records]) showed a frequency of 1 %; the remaining eighty-nine (94.5%,) showed a freciuency of less than 1 %,. There are seventeen records of species of this genus from surface catches. Out of these seventeen records, one (5.9%; Station 4619) showed a frequency of 4%; the remaining sixteen (94.1%) showed less than 1% or did not have the frequency established. The average frequency of the individual species of this genus is decidedly less than in any other genus of Dinophysoidae except in Parahistioneis. Coincident occurrence of different species of Histioneis in catches from 300 (800, 100)-0 fathoms is recorded at the following of the twenty-eight stations mentioned above: — eight species occurred coincidently at one station (3.0%; Station 4724) ; seven species at one station (3.6 % ; Station 4730) ; six species at four stations (21.4%; Stations 4699, 4701, 4722, 4734); five species at one station (3.6%c; Station 4737); four species at three stations (10.7%,; Stations 4711, 4732, 4740) ; three species at six stations (21.4%>; Stations 4681, 4691, 4695, 4097, 4705, 4742); two species at eight stations (28.6 %o; Stations 4590, 4679, 4685, 4689, 4709, 4721, 4736, 4739). Coincident occurrence of different species of this genus in surface catches is recorded at the following of the ten surface stations men- tioned above: — four species occurred coincidently at one station (10.0%; Sta- tion 4720) ; three species at one station (10.0%; Station 4733) ; two species at two stations (20.0%; Stations 4725, 4741). Although the species of Histioneis appear to be well adapted to flotation, they are very rare in surface hauls. During the Expedition only two out of our twenty-four record species were found free-living in surface waters; these two species were H. costatn and H. dolon. The remaining surface records refer to SYSTEMATIC ACCOUNT. 629 specimens found in the stomachs of Salpa taken near the surface. In Salpa stomachs the following species were found: — H. biremis, H. costata, H. dolon, H. gubernans, H. hyalina, H. inclinata, H. inornaia, H. longicollis, H. navicula, H. pacifica, and H. striata. These facts suggest that Histioneis is not adapted to the highly illumined water near the surface of the sea but may be brought there by Salpa feeding at lower levels. Out of our twenty-four record species not less than seventeen have never been found outside of the area investigated by the Expedition. It is not probable that these species are limited to this area, nor that the eastern tropical Pacific is richer in species of this genus than other tropical seas. This fact indicates how incomplete is our knowledge in this field. It should be remembered that most writers who have treated Histioneis, such as Murray and Whitting (1899), have had only surface catches at their disposal, and that most of the species of this genus appear to avoid surface waters. Historical Discussion and Systematics Stein (1883, p. 25), who established Histioneis, gave no diagnosis but con- fined himself to some general remarks about the organization and systematic position of this genus and to the representation of five species, viz., H. biremis, H. crater if ormis, H. cymbalaria, H. megalocopa, and H. remora. Of the three generic diagnoses later published byBiitschli (1885, p. 1010),Delage and Herouard (1896, p. 386), and Schutt (1896, p. 29), the one by Schutt is by far the best and most nearly complete. No extensive description and discussion of the morphology of this genus have appeared hitherto. The amount of original observations avail- able is very small, due to the extreme scarcity of the species and to the nearly complete limitation of the genus to tropical and subtropical seas. Besides the five specific names of Histioneis introduced by Stem (1883), the following specific and subspecific names of this genus occur in the literature : — //. gubernans Schutt (189.5) H. dentala Murray & Whitting (1899) H. dolon Murray & WTiitting (1899) H.francescae Murray & Whitting (1899) H. hcknae Murray & Whitting (1899) //. highleyi Murray & Whitting (1899) //. magnifica Murray & AATiitting (1899) H. milneri Murray & AMiitting (1899) H. mitchellana Murray & \Miitting (1899) H. para Murray , 12, no. 1, p. 1-151. 1903. Phytoplankton from the sea around the Faeroes. //( Botany of the Faeroes, (Copen- h.agen, Deii Nordske Vorlag), 2, p. .5.')8-()l 1 , fig. 1 19-144 in text. 1906. Catalogue des especes de plantes et d'animaux observees dans le plankton recueilli pendant les expeditions periodiques depuis le mois d'aoiit 1902 jusqu'au mois de mai 1905. Publ. de Circ. Conseil Perm. Intern. Expl. Mer, no. 33, 122 pp. 1909. Catalogue des especes de plantes et d'animaux observees dans le plankton recueilli pendant les expeditions periodiques depuis le moisd'aout jusqu'au mois de mai 1908. Publ. de Circ. Conseil Perm. Intern. E.xpl. Mer, no. 48, 151 pp. 754 THE DINOPHYSOID.\E. 1913a. De danske farvandes plankton i aarene 1898-1901. Phytoplankton og protozoer. ], Phytoplanktonets livskaar og hiologi, samt dei vore farvande iagttagne phytoplank- tonters optraeden og forekomst. Kongl. Danske Vid. Selsk. Skr., ser. 7, 9, p. 1 17-478, IS tab., 8 figs, in text. 1915. A list of the phytoplankton from the Boston .Strait, Celebes. Dansk. Bot. Ark., 2, no. 4, IS pp., 10 figs, in text. 1916b. Catalogue des especes de plantes et d'animaux observees dans le plankton recueilli pendant les expeditions depuis le mois de juillet 1908 jusf|u'aii mois de decembre 1911. Publ. de Circ. Conseil Perm. Intern. Expl. Mer, no. 70, viii + 87 pp. OSTENFELD, C. H., .\ND PaULSEN, O. 1904. Planktonprovcr fra Nord-Atlanterhavet (c. 58°-60° N. Br.\ samlede i 1899 af Dr. K. J. B. Steenstrup. Medd. om Griinland, 26, p. 139-210. 1911. Marine plankton from East Greenland Sea (W. of 6° W. long, and N. of 73°-30' N. lat.), collected during the " Danmark Expedition" 190() 1908. IV. Remarks on the microplankton. Medd. om Gronland, 43, p. 319-336, 1 map. Ostenfeld, C. H., and Schmidt, J. 1901. Plankton fra det Rode Hav og Adenbugten. Vidensk. Medd. Naturh. Foren., Kjobenhavn, 1901, p. 141-182. Ostenfeld, C. H., and Wesenberg-Lund, C. 1903a. Plankton tables for Denmark, November, 1902. Bull. Conseil Perm. Intern. Expl. Mer, 3, pt. D, p. 100-10.5. 1903b. Plankton tables For Denmark, February. Bull. Conseil Perm. Intern. Expl. Mer, 3, p. 154-159. 1903c. Plankton tables for Denmark, May 1903. Bull. Conseil Perm. Intern. Expl. Mer, 4, p. 241-245. Ove, p. van 1921. Note sur les micro-organismes de I'eau saumatre du ^'ieux Port de Batavia. Ann. Biol. Lacustre, Bru.xelles, 10, p. 207-216. Paulsen, O. 1903. Plankton tables for Denmark, August and November, 1903. Bull. Conseil Perm. Ir.tern. Expl. Mer, 3, pt. D, p. 16-19, 88-92, 312-316. 1904. Plankton investigations in the waters around Iceland in 1903. Medd. Komm. Havundersog., Kjobenhavn, 1, no. 1, 39 pp., 2 maps, 12 figs, in text. 1907. The peridiniales of the Danish waters. Medd. Komm. Havundersog., Kjoben- havn, 1, no. 5, 26 pp., 33 figs, in text. 1908. Peridiniales. In Brandt, K., and Apstein, C, Nordisches Plankton (Kiel und Leipzig, Lipsius und Tischer), no. 18, 124 pp., 155 figs, in text. 1909. Plankton investigations in the waters around Iceland and in the North Atlantic in 1904. Medd. Komm. Havundersog., Kjobenhavn, Ser. Plankton, 1, no. 8, 57 pp., 9 figs, in text. 1911a. The plankton on a submarine bank, hi Biol. Arbejder til Eug. Warming (Kjoben- havn, Hagerup), p. 231-239, 2 figs, in text. 1911b. Marine plankton from the East-Greenland Sea (W. of 6° W. long, and N. of 73° 30' N. lat.) collected during the "Danmark i:xpedition" 1906-1908. III. Peridiniales. Medd. om Griinland, 43, p. 301-318, 17 figs, in text. 1912. Peridiniales ceterae. Bull. Planktonique, 1908-1911. Conseil Perm. Intern. E.xpl. Mer, Resume Planktonique, pt. 3, p. 251-290, pi. 38-52. 1918. Plankton and other biological investigations in the sea around the Faeroes in 1913. Medd. Komm. Havundersog., Kjobenhavn, Ser. Plankton, 1, no. 13, 27 pp., 6 figs. BIBLIOGRAPHY. 755 Pavillard, J. 1905. Recherches sur la flore pelagique (phytoplankton) de I'Etang de Thau. Trav. Inst. Bot. Montpellier et Stat. Zool. de Cette, Montpellier, serie mixte, mem. 2, 116 pp., 1 chart, 3 pis. 1909. Sur les peridiniens du Golfe du Lion. Bull. Soc. Bot. de France, ser. 4, 9, p. 277- 284, 5 figs, in text. 1912. L'evolution periodique du plankton vegetal dans la Mediterranee occidentale. C. R. Ass. Fran?. Av. Sci., 1912, p. 395-397. 1915a. Accroissement et scissiparite chez les peridiniens. C. R. .\cad. Sci., Paris, 160, p. 372-375, 2 figs, in text. 1915b. Peridiniens nouveaux du Golfe du Lion. C. R. Soc. Biol, de France, Paris, 78, p. 120-122, 1 fig. in text. 1916. Recherches sur les peridiniens du Golfe du Lion. Trav. Inst. Bot. Univ. Montpellier, serie mixte, num. 4, p. 9-70, 3 pis., 15 figs, in text. 1923a. A propos de la systematique des peridiniens. Bull. Soc. Bot. de France, 70, p. 876-882, 2 figs, in text. Petersen, C. G. J. 1897. Plankton studies in the Limfiord. Rep. Danish Biol. Sta., 7, 23 pp., 1 map, 4 tab. POCHE, F. 1913. Das System der Protozoa. Arch. f. Protistenk., 30, p. 125-321 , 1 fig. in text. POUCHET, G. 1882. Sur revolution des peridiniens et les particularites d'organisation qui les rappro- chent des noctiluques. C. R. Acad. Sci., Paris, 95, p. 794-796. 1883. Contribution a I'histoire des cilioflagelles. Journ. Anat. et Physiol., 19, p. 399-455, pi. 18-22, .50 figs, in text. lS85a. Nouvelle contribution a rhistoire des peridiniens marins. Journ. Anat. et Physiol., 21, p. 22-88, pi. 2-4. 1885b. Troisieme contribution a I'histoire des peridiniens. Journ. .\nat. et Physiol., 21, p. 525-.534, pi. 26. 1894. Histoire naturelle. In Voyage de " La Manche" a I'ile Jan Mayen et an Spitzberg. Juillet-Aoiit 1892. Nouv. Arch, des Miss. .Sci., 5, p. 154-217, pi. 22. PoucHET, G., A.ND De Guerxa, J. 1885. Sur la fauna pelagique de la Mer Baltique et du Golfe de Finland. C. R. Acad. Sci., Paris, 100, p. 919-921. Pritchard, a. 1845. A history of Infusoria, living and fossil, arranged according to " Die Infusions- thieren" of C. G. Ehrenberg. (London, Whittaker & Co.), vii + 439 pp., 12 pis. 1861. A history of Infusoria, including the Desmidiaceae and Diatomaceae, British and foreign. Ed. 4, enlarged and revised by J. T. Arlidge, W. Archer, J. Ralfs, W. C. William- son, and the author. (London, Whittaker & Co.), xii + 968 pp., 40 pis. Rauschenplat, E. 1900. Ueber die Nahrung von Thieren aus der Kieler Bucht. Wiss. Meeresunters., .\bth. Kiel, new ser., 5, p. 83-151. Redeke, H. C, and Van Breemen, P. J. 1903. Plankton en bodendieren in de Noordzee verzameld van 1-6 Augustus 1901 met di' "Nelly" Y. M. 9. Tijdschr. Nederland. Dierk. Ver., Leiden, 8, p. 1 18-147. 756 THE DINOPHYSOIDAE. Reinhard, L. 1910. Vorliiufige Mitteilung iiber das Phytoplankton des Schwarzen Meeres, der Meeres- strasse von Kertsch, des Bosphorus imd des Marniarameeres. Arb. Ges. Naturf., Mos- cow, 43, p. 1-31. RuDMosE Brown, R. X. 1903. Plankton and botany. //( Tlie Scotia's voyage to the Falkland Islands. Scottish Geogr. Mag., 19, p. 17.') -17((. Schiller, .1. 1911a. Hotanische Beobachtiingen .Uigiist-Septeniiier. Perm. Int. Konuu. Eri'orsch. der Adria, p. 51-52. 1911b. Botanische Beobachtungen, November-December. Perm. Int. Konnn. Erforsch. der Adria, p. 89-90. 1911c. Untersuchung des Phytoplanktons des Adriatischen Meeres. .\nz. Kais. Akad. Wiss., Wien, 48, p. 137-138. 1912. Bericht iiber die botanischen Untersuchungen und dercn vorliiufige Ergebnisse der III Kreuzung S. M. S. "Najade" im Sommer 1911. Oestr. Bot. Zeitschr., 62, p. 359- 368, 41 1-410, 477-495, 2 figs, in text. 1912a. Botanische Beobachtungen, August-September. Perm. Int. Komm. Erforsch. der .\dria, 1 p. 1914. Oesterreichisclie Adriaforsclnmg. Bericht iiber die allgemeinen biologischen \ er- haltnisse der Flora des Adriatischen Meeres. Intern. Revue Hydrob. u. Ilydrog., 6, biol. suppl., 15 pp., 9 figs, in text. Schmidt, A. 1874-1902. Atlas der Diatomaceen-kunde in Verbindung mit den Herrn Griindler, Gru- now, Janisch, Weissflog und Witt. Herausgegeben von .\dolf Schmidt, .Archdiaconus in Aschersleben. (Aschersleben, Leipzig, Schlegel), 240 pis. 1890. Verzeichnis der in A. Schmidt's Atlas der Diatomaceenkunde (vol. 3), Heft 1-30, (Serie I-III) abgebildeten .\rten und benannten Varietiiten nebst den mit angefiihrten Synonymen. (Leipzig, Reisland), [2] + 29 pp. 1901. Preliminary report on the botanical results of the Danish expedition to ,Siam (1S99- 1900). Flora of Koh Chang. Contributions to the knowledge of the vegetation in the Gulf of .Siiim. Part IV. Pcridinlales. Bot. Tidsskrift, Kjiibcnhavn, 24, p. 12-22, 8 figs. in text. SCHRCIDER, B. 1900a. Das Phytoplankton des Golfes von Neapel nebst vergleichenden Ausblicken auf das des Atlantischen Oceans. ISIitth. Zoiil. Sta. Neapel, 14, p. 1-158, pi. 1. 1900b. Das Phytoplankt(m des GoltVs von Neapel. Schlcsi.sche Ges. Vaterland. Kultur, 78, p. 1-4. 190()a. I5eitritge zur Kemitniss der Phytoplanktons warmer Mecre. Vierleljahrschr. Naturf. (jcs. Ziirich, 51, p. 319-377, 46 figs, in text. 190()b. Zur Charakteristik des Phytoplanktons temperirter Meere. Ber. Deutsch. Bot. (Jes., 24, p. 260-2(>;3. 1909. Phytoplankton von Westindien. Her. Deutsch. Bot. Ges., 27, p. 210-214, 2 figs, in text. 1911. .\(lriatis(hes Phytoplankton. Sitzber. Kais. .Vkad. Wiss., Wien, math.-nat. Kl., 120, pt. 1, p. 601 6.57, 16 figs, in text. ScHirr, F. 1890. Ueber Peridineenfarbstoffe. Ber. Deutsch. Bot. Ges., 8, p. 9-32, pi. 1-2. 1893. Das PHanzenleben der Hochsee. Ergcbn. Plankton Exped., 1, A, p. 243-314, fig. 57-92 in text. HIBLIOGRAPH^' lOi 1S95. Die Peridineeii der Plankt')n Expedition. I Tiieil. Studien iiber die Zellen der Peridineen. Ergebn. Plankton Expeil., 4, M, a., 170 pp., 27 pis. 1S9<). Peridiniales (Peridineae, Din:>fla<;ellata, Ciliofla^ellata, arthrodeh Flagellaten) : Gymnodiniaceae, Prorocentraceae. In Engler and Prantl, Die natiirlichen Pflanzen- faniilien, sect. 1, pt. 1, b., 30 pp., 43 figs, in text. 1899. Centrifugales Dickenwadisthuni der Memljian und extiamembranoses Plasma. Jahrb. f. Wiss. Bot., 33, p. 594-690, pi. 6-S. 1900a. Die Erkliirung des centrifiigalen Diekenwachsthunis der Menibran. Bot. Zeitung, 58, p. 24.5-273, 1 pi. SCHUSSNIG, B. 1914. Aus der Biologie des adriatischen Phytoplanktons. Verhand. K. K. Zool.-bot. Ges., Wien, 1914, p. 299-309. Smyth, L. B. 1912. Plankton tables for Ireland, May, 1910. Conseil Perm. Intern. Expl. Mer, Bull. Planktonique, 1908-12, p. 50-51. Stein, F. R. von 1878. Der Organismus der Infusionsthiere naeh eigenen Forscliungen in systematischer Reihenfolge bearbeitet. III. Abth. I.IIalfte. Die Naturgeschichte der Flagellaten oder Geisselinfusorien. (Leipzig, Engclmann), 154 pp., 24 pis. 1883. Der Organismus der Infusionsthiere nach eigenen Forscliungen in systematischer Reihenfolge bearbeitet. III. Abth. II. Hiilfte. Die Naturgeschichte der arthrodelen Flagellaten. (Leipzig, Engclmann), 30 pp., 25 pis. Steuer, a. 1910. Planktonkunde. (Leipzig und Berlin, Teubner), xvi + 723 pp., 1 pi., 365 figs, in text. 1911. Leitfaden der Planktonkunde. (Leipzig und Berlin, Teubner), 382 pp., 1 pi., 279 figs, in text. Stiasny, G. 1913. Das Plankton des Meeres. (Berlin und Leipzig, Goschensche Verlagshandlung), 160 pp., 83 figs, in text. Storrow, B. 1913. Faunistic notes. Rep. Dove Marine Lab., new ser., 2, p. 99-101. Stuwe, W. 1909. Phytoplankton aus dem Nord-Atlantik im Jahre 1898 und 1899. Bot. Jahrb. f. Syst., Pflanzengesch. und Pflanzengeogr., 43, p. 225-302, pi. 1-2. Tempere, J. 1898. Infusoires cilio-flagelles des recoltes pelagiciues. Le Micrographe Preparateur, 6, p. 123-128, pi. 15-16. Theel, H. 1909. Om plankton och jimnesomsattningen i hafvet. K. Svenska V'et.-Akad. Arsbok, 1909, p. 221-249, 39 figs, in text. TORREY, H. B. 1902. An unusual occurrence of Dinoflagellata on the California coast. Am. Nat., 36, p. 187-192, 3 figs, in text. Van Goor, A. C. J. 1923. Beitrage zur Kenntnis des Nannoplanktons der siidlichen Nordsee. Verhandl. Rap- porten Rijksinstituten voor Visscherijonderzoek, 1, p. 139-182, 2 tab. 758 THE DINOPHYSOIDAE. Vanhoffen, E. 1897ii. Botanische Ergebnisse der von der Gesellsdiat't fi'ir Erdkiincle zu Berlin, unter Leitung Dr. von Drygalski ausgesandten Gronlands Expedition nach Dr. Vanholi'en's Sammlungen bearbeitet. A.Kryptogamen. lll.Peridineen mid Dinobryeen. Bibb Bot., 8, p. 25-27, pi. 1. 1897b. Eauna und Elora Gronlands. Gronland Expedition der Gesellschaft fiir Erdkunde zu Berlin, 2, p. 254-320, pi. 3-6, map 10. Wallengren, H., and Hennig, a. 1911. Encelliga organismer. In Liirbok i biologi, 2, 60 pp., 55 figs, in text. Walther, J. 1S93. Biononiie fles Meeres. Beobachtungen iiber die niarinen Lebensbezirke und Ex- istenzbedingungen. (Jena, Eischer), xxx + 96 pp. AVeber, a. VAX Bosse 1901. Etudes sur les algues de I'archipel Malaisien. III. Note preliminaire sur les resul- tats algologi(|ues de I'expedition du Siboga. Ann. Jardin Bot. Buitenzorg, 2, p. Tie- Mi, pi. 17-19. Wesenberg-Lund, C. 1908. Plankton investigations of the Danish lakes. General part. The Baltic fresh-water plankton, its origin and variation. Part I. Text, xii + 389 pp., 10 figs, in text. Part II. Appendix, 6 pp., 46 pis. West, G. S. 1916. Algae, Myxopliyceae, Peridinieae, Bacillarieae, Chlorophyceae, together with a brief summary of the occurrence and distribution of freshwater algae. (London, Cam- bridge Univ. Press), x -|- 475 pp., 271 figs, in text. Whitelegge, T. 1891. On the recent discolouration of the waters of Port Jackson. Records Australian Mus., Sydney, 1, p. 179-192, pi. 28. WiLLEY, A., AND HlCKSOX, S. J. 1909. The IMastigophora. //( Lankester, Treatise on zoology (London, Macmillan), 1, pt. 1, p. 154-191, fig. 1-15 in text. Wright, R. R. 1907. The plankton of eastern Nova Scotia waters. An account of floating organisms upon which young food fishes mainly subsist. Eurther contributions to Canadian biol- ogy, being studies from the Marine Biological Station of Canada 1902-1905. Ann. Rep. Dept. of ]\Iarine and Fisheries, Ei.sheries Branch, Ottawa, 39, p. 1-lS, pi. 1-7. /achakias, O. 1906. Leber Periodizitiit, Variation und \'eri)reitung verschiedener Planktonwesen in siidlichen IMeeren. Arch. f. Hydrobiol. u. Planktonkunde, 1, p. 498-575, 23 figs, in text. 190()a. Das Plankton als Gegenstand eines zeitgemiissen biologischen Schuluntcrrichts. Arch. f. Hydrobiol. u. Planktonkunde, 1, p. 245-344, 17 figs, in text. 1907. Das Plankton als Gegenstand der naturkundlichen Lntcrweisung in der Schule. (Leipzig, Theod. Thomas), 213 pp., 1 map, 28 figs, in text. 1911. Das Siisswasser-Plankton. Einfiihrung in die freischwebende Organismenwelt unserer Teiche, Fliisse, und Seebecken. Ed. 2 (Leipzig, Teubner), iv + 132 pp., 1 pi., 57 figs, in text. Zern'ow, S. A. 1904. Annual changes in the plankton of the Black Sea at Sebastopol. Bull. Acad. Imp. Sci. St. Petersbourg, 20, p. 119-134, 1 pi., 2 figs, in text (in Russian). INDEX I. Names of Systematic Units Higher than Species. Synonyms are in italics; the main page references are in bold-face. Acuta group (of Dinophysis) emend., 23.5, 236, 239. Dimophysis van Oye (1921), 207, 2.33. Acuta section (of Dinojihysis) Pavillard (1916), Ethmophalacroma subgenus (of Plialacroma),sub- 231,237. gen.nov.,188. Acutae section (of Dinophysis) Jiirgensen (1923), Euphalacroma subgenus (of Phalacroma) (Jorgen- 231,237. " .sen)Pavillard(192.3a),62, 82, 83. Amphidinium Claparede and Lachmann (18.58), Euphalacroma section (of Phalacroma) Jorgensen 33,233. (1923), 62, 73, 77. Amphisolenia Stein (1883), 336, 339, 340, .508. Expulsum group (of Phalacroma) n., 68, 74, 77, 82, Amphisolenidae fam. nov., 31, 335. 155. Argus group (of Phalacroma) emend., 68, 71, 82,98. Extensa group (of Amphisolenia) n., 360, .362, .364, Argus section (of Phalacroma) Pavillard (1923a), 382. 62, 82. Formosus grouji (of Ornifhocercus) n., 515, 516, Argus section (of Phalacroma) Jorgensen (1923), 577. 62, 73. Garretti group (of Parahistioneis) n., .590, 591, 592, Bifurcata group (of Amphisolenia), n., 300, 363, 596. 364, 365, 432. Globifera group (of Amphisolenia) n., 360, 362, 364, Bispinosa group (of Amphisolenia) n., 360, 362, 388. 364, 395. Hastata group (of Dinophysis), 236, 237, 261. Biremis group (of Histioneis) n., 635, 640, 641, 644, Hastata section (of Dinophysis) Pavillard (1916), 669. 231, 238. Caudata group (of Dinophysis) noni. nov., 218, //as(a(ae section (of Dinophysis) Jorgensen (1923), 236, 238, 314. 231. Ceratocorys Stein (1883), 59, 226. Heteroschisma gen. nov., 25, 26, 33, 35, 36. Citharistes Stein (1883), 31, 707. Histioneis Stein (1883), 494, 495, 496, 508, 512, Citharistidae fam. nov., 707. 516, 580, 591, 611. Contractum group (of Phalacroma) n., 67, 68, 81, Histioneis Ostenfeld and Schmidt (1901), 611, 631. 83. Histioneis Steuer (1910), 631. Cuneus group (of Phalacroma) emend., 08, 73, 82, Histiophysis gen. nov., 34, 35, 333. 124. Ca/iejjs section (of Phalacroma) Pavillard (1923a), 62, 82. Cuneus section (of Phalacroma) Jorgensen (1923), 62, 74, 77. Cytharistes Hensen (1911), 707, 709. Dinofurcula, gen. nov., 34, 35, 201. Dinophysaceae Olimanns (1922), Family, 30. Dmopfcj/seneSchiitt (1896), Family, 30. Dmop^j/siocene Pavillard (1916), Family, 30. Dinophysida Btitschli (1885), Family, 30. Dinophysidae Stein (1883) emend.. Family, 32, 33. Dinophysidae Doflein (1909), Family, 30. " Aex Dinophysiden" Stein (1883), Family, 30. Dinophysis Ehrenberg (1840a), 34, 35, 207, 211, rV/tt/Zioco-n.s Graf (1909), 496, 512. 508. ( )rnithocercidae fam. nov., 31, 494. Dinophysoidae Kofoid (1926), Tribe, 30. ( Irnithocercus Stein (1883), 494, 495, 496, 508, 512, Dolon group (of Histioneis) n., 635, 643, 645, 696. 516, 580, 591. Doryphorum group (of Phalacroma) n., 68, 78, 79, OrmthocerusSchuii (1893), 496, 512. 80, 82, 172. Oram section (of Dinophysis) Jorgensen (1923), Dixophysis Kofoid and Micheuer (1911), 233. 231, 237. Honiunculus section (of Dinophysis) Pavillard (1916), 231, 239. HoniuHcidus section (of Dinophysis) Jorgensen (1923), 231. Incisae group (of Amphisolenia) Schroder (1900a), 354, 360. Limbatum group (of Phalacroma) n., 68, 77, 78, 82, 161. Longioollis group (of Histioneis) n., 635, 640, 641, 644, 645, 676. Magnificus group (of Ornithocercus) n., 515, 516, 528. (Jblongata section (of Phalacroma) Pavillard (1923a), 62, 82. 760 THE DINOPHYSOIDAE. Ox.vi)h,v.sis Kofoid (1926), 26,31, 33, 35. Retephalacroma subgenus (of Pludacroina) sub- OxytoxumStoin (1883), 26. gen. nov., 192. Palmata group (of .Vniphisolenia) n.,36n, 362, 363, Reticulata group (of Parahistioneis) n., 590, .591, 3fJ4, 365, 409. 592, 603. Paradiniiphv.'^is .M'ctiou (of Phalai-roiua) .lorgeuseii Heticulatuin group (of Phalaeroma) n., 68, 80, 81, (19231,62,70. S2, 192. Paradinophy.sis subgenus (of Phalaeroma) (.lor- Hotundati section (of Phalaeroma) Pavillard gensen) Pavillard (1923a), 62, S2. Parahistionei.s gen. nov., 494, 496, 582. ParaschutUa .subgenus (of Ornithocercus) Murray and Whitling (1899), 51 1, 515, .526. I'arrlUm Schmidt (1888), 496, 546. Phalaeroma Stein (1883), 34, 35, 40, 241, 508. I'halncrnmo Nathansohn (1908), 40, 63. T'haUiroroma Daday (1888), 40, 63. (1923a), 62, 82. Rotundatum group (of Phalaeroma) n., 68, 69, 70, 81,85. Rudgei line (of Uinojjhysis), 238. .SV/ccm/k.s .section (of Dinophysis) Pavillard (1916), 231,237. .Schroderi group (of L)inophysi.s) n., 360, .362, 364, 382, 400^ for/o/j/uj/ocTO wo section (of Phalaeroma) Pavillard .SceM/w .section (of Dinojihysis) Jorgcnsen (1923), (192.3a), 62, 82. 231. Podophalacroma section (of Phalaeroma) Jorgen- Simpliecs group (of .^mphi.solenia) Schroder sen (1923), 62, 76. (1900a), 3.54, 360, 364. Ponteroainiix subgenus (of Triposolenia) Kofoid .s'/'/iocriVa .section (of Dinophysis) Pavillard (1916), (190(je), 4.59, 462. 231, 237. Praetextum group (of Phalaeroma) ii., ()8, SO, 81, .S'/i/iofriforseetion (of Dinophysis) J6rgensen(1923), 188. 231,2.37. Pseudophal.Mcroma Jrirgen.sen (1923), 25, 31, 33, 35. Splendidus group (of Ornithoeercus) n., 515, 516, Pulehra group (of lli.stioneis) n., 635, 638, 639, 641, 517. 645, 683, 686. Theeadinium gen. nov., 31, 32, 33, 35. Pusilla line (of Dinophysis), 238. Trijiosolcnia Kofoid (1906c), 336,339, 354, 446,508. Ramiciformia subgenus (of Triiiosoleniii I Kofoid Triijosoleniasuligeiuis (of Triposolenia) nobis, nom. (1906c), 4.59, 461, 462, 489. subgen. nov., 461, 463. RdmiciJormtH subgenus (of Triposolenia) Kofoid Uraeantha line (of Dinophysis), 238. (1906c), 462. Urophalaeroma section (of Phalaeroma) Jorgensen Rapa group (of Phalaeroma) n., 68, 74, 82, 139. (1923), 62, 78, 80. Remora grouji (of Histioneis) n., 635, 639, 641, 644, rrophalacroma .section (of Phalaeroma) Pavillard 661. (1923a), 62, 83. II. ."specific and Sub.specific Names. Synonyms are in italics, the main page references are in bold-face. ticwicula Bergh, Balbiani (1884b), Diiio/jlii/iiis, 224, 226. acuminata Chiparede and Lachmann (1858), Dino- physis, 59. 219, 224, 228, 230, 234, 235, 237, 256. acuminata Jorgensen (1899), Dinophysis, 230. acuminata .Jorgensen (1912), Dinophysis, 229. acHminala Claparede and Lachmann var. (iranidata ((."levp) Jcirgenscn (1901), Dinophysis, 225, 228. acuminata Claparede and Lachmann f. rcnijorinis Pavillard (1905), Dinophysis, 225, 228. acuminatum (ClaparMe and Lachmann) Zaeharias (1906), Phalaeroma, 58. acuta Ehrenbcrg (1840a), Dinophysis, 208, 219, 220, 224, 226, 227, 229, 230, 234, 236, 2.56. acuta Jorgensen (1899), Dinophi/sis, 228, 259. acuta Ehrenberg var. gcminala Pouchet (1883), Dinophysis, 224, 226, 314, 322. acuta Khrenberg var. sculpla Jorgensen (1923), Dinophysis, 225, 228. acuia Ehrenberg var. steini Lemmermann (1902a), Dinophysis 225, 228. aculum (Schiitt) Pavillard (1916), Phalaeroma, 58, 61,68,75,76, 113, 143, 149. aequale sj). nov., Heteroschisma, 36. '>W'<''J"I<' \\ ebcr-van Bosse (1901 ), Dinophysis, 224, 227. alata Jorgensen (1923), Dinophysis, 225, 230, 236, 237, 268, 285. nllieri Gourret (1883), Dinophysi.% 224, 226, 314, 322, 325. ambulatrix Kofoid (1907a), Triposolenia, 4.59. 462, 468,471,485,486. ambulatrix Jorgensen (1923), Triposolenia, 486, 488. INDEX. 761 apicatum sp. nov., Phalacroraa, 50, 01, 6S, 71, 102, 108, 111, 116, 124. apiculata Mcunier (1910), Dnt«/)/i;/.si.s-, 225, 228. apsteini Schtitt (1805), Citharistes, 708, 711 , 712. arctica Mereschowsky (1878), Dinophy.sis, 210, 224, 228, 230, 2:34, 235, 236, 256. arcuatum Hensen (1011), Fhalacroiiiri, 5S, 60. argo Hensen (1011), Phalacroma, 58, 60, 104, 100. argus Stein (1883), Phalacroma, 58, 61, 68, 71, 82, 102, 104,113,121,124,133. armnta Daday (1888), DinnphiDiis, 224, 226. armatiim Hensen (1805), riialaeroma, .58, 60. nssinitlis Jorgensen (1023), (Jniithoccrcti^;, 510, 561, 569, 570. astragalus Kofoid and Mifhener (1011 1, .^nipliiso- lenia, 353, 360, 361, 365, 380. asymnietrica Kofoid (1007a), Aniiiliisolcnia, 348, 353, 354, 360, 363, 366, 413, 424, 428, 429. atlantica Ehrenberg (1854), Dinophysis, 224, 230, 231, 253, 255. baltica (Paulsen, 1908) nobis, Dinophysis, 210, 225, 220, 230, 235, 236. bicornis Kofoid (1006c), Tripo.solenia, 450, 462, 463,471,473,470. bidentala Schroder (lOUOa), Am|ilusolcnia, 318, 353, 360, 363, 366, 382. 409, 424, 428, 431 . bidentala Okamura (1907), Aiiiphisolenia, 414, 419, 420. bidentata Pavillard (1016), partim, Aiiipkisolcnia, 422. bifureata Murray and Whitting (ISOO), Aniphiso- lenia., 353, 3.54, 360, 364, 366, 432, 438, 445, 446. bihastata Hensen (1011), Dinophysis, 225, 226. bipartitum sp. nov., Phalacroma, .59, 61, 68, 78, 166, 205. bipes Hensen (1911), Dinophysis, 225, 226. biremis Stein (1883), Histioneis, 629, 630, 635, 646, 665, 669, 676. bLspinosa Kofoid (1907a), Am])hisolenia, 353, 360, 365, 395, 300. blackmani Murray and Whitting (1809), Phala- croma, 58, 61, 68, 73, 127, 133, 138. brevicauda Kofoid (1907a), -Vmphisok'nia, 353, 356, 360, 361,. 365, 372. carinata Kofoid (1907a), Histioneis, 630, 632, (J35, 640, 644, 646, 652, 654, 663, 668, 669, (572. carolinae Kofoid (1907a), Ornilhocercus, 509, 510, 511, 515, 516, 517, 5.34, 546, 572, .500. caudata Saville-Kent (1881), Dinophysis, 218, 220, 224, 226, 227, 230, 234, 236, 238,314, 331 , 330. caiidata Saville-Kent var. abbrcriata Jorgensen (1023), Dinophysis, 225, 226, 314, .325. caudata Saville-Kent f. acutiforniis noni. f. nov., Dinophysis, 225, 227, 231, .322, 330. caudata Saville-Kent var. allien (Gourret ) Jorgen- sen (1923), Dinophysis, 315, 325. cavdata SaviUe-Kent var. allien (Gourret) f. speci- osa Jorgensen (1023), Dinophysis, 225, 226, 315, 320, 322. cfi uda/a Saville-Kent f. marmarne Jorgensen (1023), Dinophysis, 225, 226, 315, 321, 325. cuii^ata Saville-Kent var. pedunmlata (Schmidt) Jorgensen (1023), Dinophysis, 315, 321. aiiulata Saville-Kent f. pontica Jorgensen (1023), Dinophysis, 225, 226, 315, 321, 324, 325. caudata Saville-Kent f. siibdiegensis Jorgensen (1923), Dinophysis, 225, 226, 231, 319, 332. ctindataviiT. rentricosa (Pavillard) J6rgen.sen (1023), Dinophysis, 333. ccratocorys Entz (1902a), I'hulacroma, 58, 59. cernlocorys var. iridentata (Daday) Entz (1902a), Phalncronw, .58, .59. circumcinctum Kofoid and Micliciier (1911), Phal- acroma, 58, 61, 68, 72, 113,114, 121. circumsutiun Karsten (1007), Phalacroma, 58, 61, 68, 78, 70, 170, 182, 188. circumsulutii Jorgensen (l9'2'.i),I'liiil(tcroni(i,7S, 168, 171, 185. clavipes Kofoid (1007a), .Am|)hisolenia, 353, 354, 360, 363, 3(54, 366, 401, 402, 406, 408. collaris Kofoid and Michener (1011), Dinophysis, 165, 168, 225, 230, 236, 238, 283, 288, 292. compactus Hensen (1911), Ornithocercus, 510. comiilanata sj). nov., Amphi.solenia, 354, 360, 363, 3(56,401, 402,404,408. contractum sp. nov., Phalacroma, 50, 61, 67, 60, 83. costata Kofoid and Michener (1011), Histioneis, ()29, 630, 632, 635, 636, 646, 647, 651, 654. crateriformis Stein (1883), Histioneis, 590, 620, 630, 640. crateriformis (Stein, 1883), Parahistioneis, 590, 501 , 502. cuncus Sluwe (1000), Phalacroma, .58, 60, 124. cuneiformis Meunier (1010), Dinophysis, 160, 219, 225, 229, 230, 234, 236. cuneolus sp. nov., Phalacroma, 50, 61, 68, 70, 171, 170, 186. cuneus Schtitt (1805), Phalacroma, 58, 61, 64, 68, 74, 124, 133, 138. curvata Kofoid (1007a), .Am|ihisolenia, 353, 360, 362,36(5,371,307,398. cymhalaria Stein (1883), Histioneis, 620, 630, 635, 642,646,081,688,692. dens Pavillard (1915b), Dinophy.sis, 225, 230, 235, 237. dentata Murray and Whitting (1899), Histioneis, 590, 610, 629, 630. dentata (Murray and Whitting, 1800), Parahis- tioneis, 500, 502. depres.sa Kofoid (1006c), Triposolenia, 4.50, 402, 463, 468. diegensis Kofoid (1907b), Dinophysis, 225, 226, 231,332. diegensis Kofoid var. caudata Pavillard (1916), Dinophysis, 225, 227, 231, 332. diegensis Kofoid var. curvata Kofoid (1907b), Dinophysis, 225, 227, 231, 332. 762 THE DINOPHYSOIDAE. diomedeae Kofoid and Michener (1911), HistioneU, 590,608,610,629,630. diomedeae (Kofoid and Michener, 1911), Para- histioneis, 5S3, 590, 592, 605, 608. dolichocephalica Kofoid (1907a), Amphholenia, 353, 354,429,431. dolichopterygium Murray and WTiitting (1899), Phalacroma, 58, 61, 68, 75, 76, 143, 149, 153. dolichopterygium Jorgcn.sen, Phalacroma, 75, 76. dolon Murray andWhitting (1899), Histioneis, 629, 630, 632, 633, 635, 643, 646, 696, 698, 703, 707, doryphorum Stein (1883), Phalacroma, 52, 58, 61, 68, 79, 121, 174, 175, 185, 188. ehriola Herdman (1924), Phalacroma, 32, 58, 59. ebriolum (Herdman) nobis, Thccadinium, 32. ellipsoides Kofoid (1907b), Dinophysis, 225, 230, 235, 237. elongatasp. nov., Amphisolenia, 354, 360, 365, 385, 386. elongata Kofoid and Miclicncr (1911), Histioneis, 629, 630, 632, 635, 640, 640, 661, 665. elongatum Jorgensen (1923), Phalacroma, 58, 61, 68, 72. euter Hensen (1911), Phalacroma, 58, 60. exigua sp. nov., Dinophysis, 88, 225, 230, 235, 236, 239. exilis Kofoid (1906c), Triposolenia, 459, 462, 463. expulsa Kofoid and Michener (1911), Dinophysis, 59, 188, 225. expidaa Kofoid and Michener (1911), Dixophysis, 157, 161. expulsum (Kofoid and Michener, 1911), Phala- croma, 61, 68, 74, 77, 82, 157, 2.52. exten.sa Kofoid (1907a), Amphisolenia, 353, 354, 360, 365, 380, 383. fatula Kofoid (1907a), Triposolenia, 459, 462, 463, 482, 483. favus Kofoid and Michener (1911), Phalacroma, 58, 61, 68, 75, 76, 143, 146, 153. fimbriatum Kofoid and Michener (1911), Phala- croma, 34, 58, 61, ()4, 65, 68, 78, 81, 191, 192, 197, 201. formosus Kofoid and Michener (1911), Ornitho- cercus, 495, 510, 511, 515, 517, 577, 649. fortii Pavillard (1923a), Dinophysis, 225, 228, 230, 235, 236, 237, 253, 260. /rancescoe Murray and Whitting (1899), Histioneis, 576, 590, 629, 630. francescae (Murray and Whitting, 1899), Para- histioneis, 494, 51 1, 590, 592. furca Hensen (1911), Amphisolenia, 354. galea Pouchet (1883), Dinophysis, 59, 179, 182, 185, 224, 226, 510, 540, 548, 561, 570. galea (Pouchet, 1883), Phalacroma, 60, 61. !7«rreHi Kofoid ( 1907a) , //is(J07ieis, 590, 596, 598, 630. garrctti (Kofoid, 1907a), ParahLstioneis, 590, 592, 596, 600, 603. geminata (Pouchet) Kofoid and Rigden (1912a), Dinophysis, 225, 226, 314, 323. giganteum Kofoid and Michener (1911), Phala- croma, 58, 61, 64, 67, 68, 74, 127, 128, 133, 136. globifera Stein (1883), Amphisolenia, 340, 348, 353, 354, 356, 360, 365, 380, 388, 414. globosa Gran (1912a), Amphisolenia, 388, 391. globulus Schutt (1895), Phalacroma, 58, 61, 67, 98. granulata Cleve (1899c), Dinophy.sis, 219, 224, 228, 229, 230, 234, 235, 236. granulata Jorgensen (1912), Dinophysis, 229, 230. gubernans Schiitt (1895), Histioneis, 622, 630, 635, 638, 642, 645, 646, 658, 684, 686. hastata Stein (1883), Dinophysis, 60, 219, 224, 228, 230, 234, 236, 237, 261, 278, 285, 302. hastata ateinviiT. farrula Lindemann (1923), Dino- physis, 225, 228, 261, 269. hastata Stein var. phalacroiyiides Jorgensen (1923), Dinophysis, 225, 228, 261, 268, 269. hastata Stein var. uracanthides Jorgensen (1023), Dinophysis, 225, 228, 261, 268, 269, 278. hastatum Hensen (1895), Phalacroma, 58, 60. hastatitm Pavillard (1909), Phalacroma, 58, 59, 261, 269. helenae Murray and Whitting (1899), Histioneis, 629, 630, 632, 635, 644, 646, 696, 700, 703, 707. heteroporus Kofoid (1907a), Ornithocercus, 509, 511,515,516,517,526,569. highleyi Murray and Wliitting (1899), Histioneis, 629, 630, 635, 646, 672, 673, 675. hindmarclii Murray and Wliitting (1899), Phala- croma, 58, 61, 68, 76, 143, 149, 156. Amc/marc/m Jorgensen (1923), Phalacroma, 153. hindmarchii Pavillard (1916), Phalacroma, 146, 149, 150. hiijpoperoides Kofoid and Michener (1911), Hi.s- tioneis, 629, 630, 633, 635, 643, 644, 646, 700, 701, 70(i. homumncalus Reinhard (1910), Dinophysis, 315, 325. homuncidus Stein (1883), partim, Dinophysis, 224, 226, 315, 323. hoinunculus Stein var. allieri (Gourret) Lemmer- mann (1899a), Dinophysis, 224, 226. homuncidus Stein var. appendiculata Zacharias (1906), Ditiophysis, 225, 226, 324. homunculus Stein var. carinala Zacharias (1906), Dinophysis, 225, 226, 315, 325. homunculus Stein var. gracilis Schroder (1911), Di7iophysis, 225, 226, 315, 325. homuncidus Stein var. incurva Lemmermann (1899a), Dinophysis, 224, 226, 324. homunculus .'^tein var. latuni Lindemann (1923), Dinophysis, 225, 226, 315, 325. homunculus Stein var. pedunculata Schmidt (1901), Dinophysis, 225, 226, 315, 321, 325. homuncidujs Stein var. tripos (Gourret) Lemmer- mann (1899a), Dinophysis, 224, 226. homunculus Stein var. ventricosa Pavillard (1916), Dinophysis, 225, 227, 319, 330, 333. INDEX. 763 hormunculus Hensen (1911), £)i?to/j/(ys(,s, 225, 22(5, 315,326. hornunculus Hensen (1911), Dinophijsis, 225, 226, 315, 326. horridus Hensen (1911), Ornithocercus, 510. hyalina Kofoid and Michener (1911), Histioneis, 630, 635, 641, 646, 678, 679, 683, 688, 693. inaequale sp. nov., Heteroschisma, 36, 38-40. inaequalis Gourret (1883), Dinophysis, 224, 226, 315, 323. inclinata Kofoid and Michener (1911), Histioneis, 630, 632, 635, 636, 637, 646, 649, 652, 656. inflata Murray and Whitting (1899), Amiihisolenia , 338, 353, 356, 358, 360, 365, 366. inornata Kofoid and Michener (1911), Histioneis, 630, 632, 635, 637, 638, 645, 646, 654. intermedia Cleve (1902b), Dinophysis, 225,226, 228, 253, 255, 630, 676. intermedia Pavillard (1916), Dinophysis, 225, 228, 253, 255. intermedia Pavillard f . pachyderma Jorgensen ( 1923) , Dinophysis, 225, 228, 253, 255. intermedia sp. nov., Triposolenia, 4.59, 462, 463, 477, 482. intermedium. Jorgensen (1923), Dinophysis, 253, 255. irregulare Lebour (1925), Phalacroma, 58, 60, 61, 67, 70. jorgenseni sp. nov., Dinophysis, 65, 225, 230, 236, 238, 302, .306, 307, 314. josephinae Kofoid (1907a), HLstioiieis, 630, 632, 633, 635, 644, 646, 704. jourdani Gourret (1883), Dinophysis, 59, 224, 226. jourdani (Gourret) Schiitt (1895), Phalacroma, 58. karsleni Kofoid and Michener (1911), Histioneis, 590, 603, 605, 630. karsteni (Kofoid and Michener, 1911), Parahis- tioneis, 590, 591, 592, ,594, 603, 610. kofoidi var. petasalum Herdman (1922), Amphi- dinium, 32, 59. kofoidi Jorgensen (1923), Dinophysis, 225, 227, 332. kofoidi Forti and Is.sel (1925), Histioneis, 630, 635, 641, 645, 646, 678, 681, 688, 693. kofoidi Herdman (1923), Phalacroma, 32, 58, .59. Zaects Claparede and Lachmann (1858), D!)«)/)/i,v.s!s, 53, 59,88,91,224, 225. laevis Daday (1888), Dinophysis, 224, 226. laems Pouchet (1883), Dinophysis, 253, 255. laevis Stein, Dinopyxis, 226. laticincta Kofoid (1907a), Amphisolenia, 353, 360, 361, 365, 369. lativelatumsp. nov., Phahicroma, 59, 61, 67, 70, 88, 89. lemmermanni Kofoid (1907a), Amphisolenia, 348, 353, 360, 363, 366, 412, 414, 419, 424, 428, 431. lenssp. nov., Phalacroma, 59, 61, 67, 70, 91, 96, 124. lenticula Daday (1888), Dinophysis, 224, 226, 228. lenticula Pavillard (1916), Dinophysis, 225, 228, 241, 245. lenticula Bergh, Diplopsalis, 226. lenticula Kofoid (1907a), Phalacroma, 58, 61, 66, 67, 70, 96. ?u»6ota Ehrcnberg (1840a), Dinophysis, 230. limbatum Kofoid and Michener (1911), Phala- croma, 58, 61, 64, 66, 68, 78, 117, 162, 168, 171. longicollis Kofoid (1907a), Histioneis, 630, 635, 641, 646, 676, 677, 681, 683, 688, 693. longicornis Kofoid (1907a), Triposolenia, 459, 462, 479, 485. magnifica Hjort (1911), Histioneis, 535. miign-ifica Murray and Whitting (1899), Histioneis, 529, 535, 538, 547, 629, 630. mayniftca Schroder (1900a), Histioneis, .529, 535, .561, 569. magnificn var. qiuulrala Entz (1902b), Histioneis, 570, 629. inngnijica var. steini Entz (1902b), Histioneis, 630. magrnficus Lemmermann (1899a), Histioneis, 529, 534, 547. magnificus Biitschli (1885), OrnHhoecrcus, 529, 540. magnificiis Doflein (1909), Ornithocercus, 521, 526, 536. magnificus France (1823), Ornithocercus, 529, 535. magnificus Fricke (1902), Ornithocercus, 529, 535. magnificus Hjort (1911), Orniihocereus, 529. m((g?«_/!ri(601tmanns (1922), OrMiZ/iofnTMS, 521, 526, 536. magnificus Stein (1883), s. .str., Ornithocercus, 496, .503, 509, 510, 511, 515, 517, 529, 540, 546, 547. ntagiiificits Stiiwe (1909), Ornithocercus, 535. magnificus ^^^litelegge (1891), Ornithocercus, 529, 535. magnificus Willey and Hickson (1909), Ornitho- cercus, 529, 535. magnificus var. a, Stiiwe (1909), Orniiluiccrcus, 529. magnificus var. b, Sttivve (1909), Ornithocercus, 551, 556. magnificus var. c and d, Stiiwe (1909), Ornithocer- cus, 540. magnificus var. e, Stiiwe (1909), Ornithocercus, 561. megalocopa Stein (1883), Histioneis, 629, 630, 635, 644, 646, 700, 703, 707. michaclis Ehrenberg (lS40a), Dinojihysis, 88, 224, 230,231,241. miles Cleve (1900b), Dinophysis,218, 220,224, 227, 230, 234, 236, 239, 322, 339. miles miles (Cleve, 1900b) nobis, Dinophysis, 227, 322. miles Cleve var. aggregata (Weber-van Bosse) Lemmermann (1901a), Dinophysis, 225, 227. miles Cleve f. indica Ostenfeld and Schmidt (1901), Dinophysis, 225, 227. miles Cleve f. maris jonii Schroder (1906a), Dino- physis, 22.5, 227, 231, 322. miles Cleve f. maris rubri Ostenfeld and Schmidt (1901), Dinophysis 225, 227. miles Cleve f. schroteri (Forti) Ostenfeld (1915), Dinophysis, 225, 227, 230, 322. '64 THE DIXOrilYSOTDAE. inilnei'i Murray and Whitliiis (1899), Histioncis, (529, 630, 632, f)3r), 614, 646, 697, 700, 703, 707. minor Jorsensen (1923), Ornitlwcercus, 510, 529, 536. mimituin Clevc (1900c), I'lKilanoina, 53, 58, 61, 61,68,75,76. initclicllana Miirniy and Whittinf;; (1899), Ili.«- lioncis, 629, 630, 632, 635, 642, 646, 68S, 690. iiiilchdliaiiii Schr(Kk-i' (1906a), IlUlioiicis, 630, 691), 693. miVro Okaniuia (1907), I'luilaciuiiiu, 139, 142, 144. HK7ra, Olt.manns (1922), Phalacroma, 124, 128. mitra Schutt (1895), Phalacroma, 58, 61, 68, 75, 76, 143, 144, 149, 1.53. ni'inacantha. s|). nov., I)inii|iliy,-, 238,268,283,288,314. niiicriinafuni ,s]). nov., I'lialacronia, 59, 61, 68, 79, 172, 179, 188. HciKuluin Stein (1883), Phulacroino, .58, 60. na.sutum (Stein) Jorgenscn (1923), Pseudophala- croma, 69, 84. navicula Kofoid (1907a), Histionei.s, 630, 632, 635, 640, 646, 665, 667, 668, 669. iiomyid/ui Mo.'is (1877), I)iiiliysi.s, 219, 224, 22(), 227, 230, 234, 236, 256, 2(i0. iwri'egica Clap.arede and Lachmann var. crassi- cornis Smyth (1912), Diuophysis, 225, 226. tborvcgica Claparede and Lachmann var. crassiar Paulson (1907), Dinophysis, 225, 226, 228. norregica Claparede and Lachmann f. debilior Paulsen (1907), Dinophys-U, 225, 228. nias Kar.sten (1907), Dinophysis, 225, 230, 236, 238,302,303,310,314. ohliquum llcn.sen (1911), Plialacroma, 58, 60. (iniJii Hjorl and Gran (1899), Dinophysis, 224,226. okamuraisp. nov., Dinophysis, 161, 225, 2.30, 235, 237, 260. opeir.ulotum Forti (1922), partim, Pluilacroiim, 118. opercuUilum Henscn (191 1), jiartim, l'halacroiiia,W, 102. opercidalum .Jorficnscn (1923), I'holdcroiiin, 118, 121 . operciikttum Schutt (1895), I'hcilarroiiia, 98, 102, 118. o|ierculatuin Stein (1883), s. str., Plialacroma, .58, 61, 68, 71, 109, 113, 118, 121, 1,53. o/jc/ru/mrfc.s Cleve (1899c), PhaUicroiim, 118. (iperciihi(kitVovV\ (1922), Phalacronm , 118. operculoidcs imp.'\\iin\ (1923), Phalacronid, 71, 118. o/«'m('"o/(/iy.s('.s-, 224, 225. croma, 58, 61, 64, 08, 78, 165, 168. punctata Jiirgensen (1923), Dinopliy.sis, 225, 230, 235, 236, 245. pusilla Jorgensen (1923), Dinojjliy.si.s, 225, 230, 23t'), 237, 268, 278. pyriforme sp. nov., Phalacroma, 59, 61, 68, 72, 122. quadrata Lemmermann (1901a), Hislioneis, 629. quadratusSchiitt (1900a), Ornithoeercns, 509, 511, 515, 517, 561, .'581. rotundatum (Claparede and Lachmann) Kofoid and Michener (1911), Phalacroma, 53, .58, .59, 61,64,67,87,93,90. rtituiidiitnin Hensen (1911), I'tuddcniniu, 58, 00. nit inidatum (Claparede and Lachmann) var. laevis (Claparede and Lachmann) Jorgensen (1923), Phalacroma, 53, 54, 58, 01, 63, 67, 91. rudgei Murray and Whitting (1899), Dinoiihysi.s, 224, 231, 230, 238. quadratus Schiitt f. a.ssiinilis (.hirgensen) nobis, ruilgei Murray and Whitting (18991, Phalacnima, Ornithocercus, .565. quadratus Sclititt f. intermedia n., Ornithocercus, 510, 526, 567, 569. quadratus Schiitt f. quadrata n., Ornithocercus. 510, 562. quadratus Schiitt f. schiitti n., Ornithocercus, 510, 564. quadratus ,Scluitf f. simplex n., Ornithocercus, 510, 565. quadricauchi Kofoid and Michener (1911). .Xniphi- solenia, 353, 3(i(). :!(;4, 300, 435, 444. .53, .58, 00, 01,04, 07, 93, 96. rngosa Kofoid and Michener (19111, Diiioidii/sis, 225, 226, 334, 335. rugosa (Kofoid and Michener, 1911), Histioi)hysis, 333, 334. .sacculus Stein (18S3), Diiioiihysis, 224, 231, 23.5, 237. scliauinslandiLemnu'rniaun ( lS99a),,\mi)liisolenia, 353, 356, 300, 301 , 305, 374, 379. schroderi Kofoid (1907a), .\mphisolenia, 352, 354, 35(i, 300, 303, 3t)4, 306, 400. 404, 406, 408. quadrisi)ina Kofoid (1907a), .Vmphisolcnia, 353, xchrudcri Pavillard (1909), Dinoplnisis, 225, 231, 360,305,379,391,393. 230,2.55,257. quinquecauda Kofoid (1907a), Amphisolenia, 353, ■■ichruteri (Forti, 1901), DiHophysin, 224. 300, 364, 366, 435, 445. bichrdtcri Forti (1901), Heterocenis, 227. ramiciformis Kofoid (1906c), Triposolenia, 459, schiitti Murray and Whitting (1899), Dinophysis, 462, 463, 489. 219, 224, 231, 2.30, 238, 241, 292, 296. roya Pavillard (191.5aJ, /■'/m;»rra7H(7, 139. schiitti Murray and Whitting var. uracanthoides r.apa Stein (1,S83), Phalacroma, .58, 01, 08, 75, 70, Korti and Issel (1924), Dinoi)hysis, 225, 227, 231, 139, 149, 1.53. 290, 301. rectangulata Kofoid (1907aj, ,\mphisolcnia, 3.53, semen Meunicr (1910), Dinophysis, 219, 225, 229, 360, 301, 365, 378. ^ 231, 234, 235, 236. recurva nom. sp. nov., Dinophvsis, 225, 228, 231, nfinicarinata Grenfell (1887), hinopln/sis, 224, 226, 235, 236. 31.5, .321, 324. reginella Kofoid and Michener (1911), Ilislioncis, nrrnitus Kofoid (1907a), Oniithnccraix, 510, 547, 630, 632, 035, 638, 646, 656. regius Stein (1883), Citharistes, 70S, 709, 714. remora Stein (1883), Histioneis, 629, 030, 035, 639, 646, 602, 663, 065, 0li9. renifonnis (Pavillard) nobis, Dinophysis, 225, 228, 231, 235, 237. retindata Kofoid (1907a), HUtii)iieis, .590, 605, ti07, 630. reticulata (Kofoid, 1907a), Parahistioneis, .590, 592, 605, 610. reticulatum Kof.iid (1907a), Phalacroma, 58, 61, 68,80, 191,195,201. rotundata Claparede and Lachmann (1858), Dinii- p/ii/sis, .59, 224,225. rotundata Lohmanii ( 1908a), Dinophysis, 230. 551, .5.56. siniilis sp. nov., Dinophysis, 101, 225, 231, 235, 237, 245, 247. .s('wM/o/i.v.Icirgensen (1923), /'/i(i/(KTr)»/(/, 58, 00, 146, 149. sp. Chun (1903), Amphisolenia, 409, 414. sp., Phalacroma, 60. sp. Okamura (1907), Phalacrmna, 111. sphaerinm Entz (1902b), Dinoplnisis, 225, 226, 241, 245. sjdiaerica Jorgensen (1923), Dinophi/xis, 242, 245, 247, 249. spliaerica Okamura (1912), Din)}pliysis, 242, 245. sphaerica Pavillard (1916), Dinophysis, 242, 245, 247, 2.50. 766 THE DINOPHYSOIDAE. sphaerica Schiitt (1895), Dinophysis, 241, 245, 247, 250. sphaorioa Stein (1883), partim, Dinophysis, 59, 219, 224, 226, 231, 234, 235, 237, 241, 249, 250. .iphaericum Zacharias (1906), Phalacrnma, 58, 242, 245. sphaeroidea Ilon.son (1911), Dinophysis, 225, 226. .spinulosa Kofoid (1907b), Amjihisolenia, 3.53, 360, 362, 365, 379. splendens Schiitt (1896) , Ornilhocercus, 509, 52 1 , 526. splendida Murray and Whitting (1899), Histioneis, 521, 629. splendidus Schiitt (1893), Ornithocercus, 509, 511, 515,516,519,521,569. uteini Loninicnnann (1901a), Histioneis, 629. steini .Schiitt (1900a), s. str., Ornithocercus, 509, 510, 511, 515, 516, 517, 534, 540, 546, 547, 551, 559, 581, 582. slenoptiri/gium Jorgensen (1923), Phalacroina, 58, 60, 74, 157, 160, 161. striata Kofoid and Michener (1911), Histioneis, 630, 635, 642, 645, 646, 684. striatum Kofoid (1907a), Phalacroma, 58, 61, 68, 73, 82, 108, 127, 128, 131, 138. subdiegensis Jorgensen (1923), Dinophysis, 332. swezyi sp.nov., Dinophysis, 225,231,234, 236, 238, 241,288,289,301. ieuella Gran (1912b), Atnphisolenin, 354, 388, 391. thrinax Schiitt (1893), Amphisnlenia, 353, 356, 360, 364, 366, 438. thrinax Zacharias (1906), Amphisolenia, 432, 435, 439, 442. thurni (Sclimidt, 1888), Ornithocercus, 511, 515, 517, .534, 540, 556, 569, 581, 582. Ihurnii Schmidt (1888), Parelion, 510, 540, 546. trapczhim sp. nov., Dinophysis, 225, 231, 236, 238, 286,292,295,314. triacantha Jorgensen (1923), partim, Dinophysis, 307, 312, 314. triacantha Karsten (1907), Dinophysis,303, 312, 314. triacantha Kofoid (1907a), Dinophysis, 225, 231, 236,238,301,306,310,312. tridens Graf (1909), Amphisolenia, 354. tridmdatu Daday (1888), Ccratocorys, 59. trinaxHensen (1911), Ainphisolenia, 439. tripes Henscn (1911), Dinophysis, 225, 226. tripos Schiitt (1893), Amphisolenia, 353, 354, 459. tripos Gourret (1883), Dinophysis, 218, 224, 226, 231, 236, 239, 321, 322, 323. tripos Goiurct f. brevicauda Jorgensen (1923), Dinophysis, 225, 226, 324. truncata Kofoid and Michener (1911), Amjihi-su- lenia, 338, 353, 354, 360, 363, 366, 368, 402, 406. truncata Cleve (1900f), Dinophysis, 224, 231, 235, 237. truncata Kofoid (1906e), Triposolenia, 338, 361, 447, 459, 462, 463. turbineum Kofoid and Michener (1911), Phala- croma, 58, 01, 68, 81, 191, 195, 197, 198. ultima (Kofoid, 1906c), Duiofurcula, 31, 34, 202, 203, 206. ultima Kofoid (1906c), Phalacroma, 58, 60, 203. uracantha Schiitt (1895), Dinophysis, 273, 278, 296, 301. uracantha Stein (1883), Dinophysis, 224, 231, 234, 236, 238, 241, 268, 273, 383, 385, 301, 314. uracantha Stein var. meditcrranca Jorgensen (1923), Dinophysis, 225, 228, 261, 269^ 273, 278. tirceolus sp. nov., Dinophysis, 225, 231, 230, 238, 27.8,281,286,295,314. vanhoffeni Ostenfcid (1899), Dinophysis, 224, 228. tmihoffenii Okamura (1907), partim, Dinophysis, 250, 252. vastum Schiitt (1895), Phalacroma, 58, 61, 68, 71, 82, 144. vastum Schiitt var. acuta .Schiitt (1895), Phalacro7na, 58, 60, 113. ventralis sp. nov., Dinofurcula, 204, 205. ventricosa Chiparcdc and Lachmann (1858), Dino- physis, 219, 224, 227, 228, 230, 231, 234, 235, 236. vermiculata Pouehet (1894), Dinophysis, 224, 230, 231. vertex Meunier (1910), Dinophysis, 219, 225, 229, 231, 234, 2.35, 236. EXPLANATION OF THE PLATES. PLATE 1. PLATE 1. Fig. 1. HeteroscMsma inaequale, sp. nov., right lateral view, somewhat tilted ventrally; type specimen. Station 4CG5 (300-0 fathoms). X 1090. Fig. 2. The same, ventral view. Note that left sulcal list crosses over to right valve at level of posterior cingular list. X 1090. Fig. 3. Phalacroma porosum Kofoid and Michener, oblique ventral view, somewhat tilted posteriorly; type specimen. Station 4721 (300-0 fathoms). X 525. Fig. 4. Phalacroma cimimcinctum Kofoid and Michener, ventral view; type specimen. Station 4671 (300-0 fathoms). The irregular, hyaline appendage, issuing from near the pos- terior main rib of left sulcal list is probably a temporary structure. X t')*)5. Fig. 5. The same, right lateral view. See explanation of Figure 4. X 665. Fig. 6. Phalacroma porosum Kofoid and Michener, right lateral view; type specimen. Station 4721 (300-0 fathoms). X 1690. Fig. 7. HeteroscMsma acquale, sp. nov., ventral view; type specimen. Station 4671 (300-0 fathoms). Note that left sulcal list crosses over to right valve at level of posterior cingular list X 1090. Fig. 8. Thesame, right lateral view, tilted ventrally. X 1090. MEM. MUS COMP, ZOOL, DINOFLAGELLATA. PLATE I 1 LH'ayiaii A. B. Stfeedain, del. M6LIOTYPE CO. BOSTON PLATE 2. PLATE 2. Fig. 1. I'halarruma fimbriatuiii Kofoid and Michener, ventral view; type specimen. Station 4613 (.300-0 fathoms). Note tiie heavy rib that indicates the presence of a postcinguhir plate in the left hypotheca. X 580. Fig. 2. Phalacroma giganteum Kofoid and Michener, right lateral view of right detached valve; type specimen. Station 4734 (300-0 fathoms). X 525. Fig. 3. Phalacroma turbincum Kofoid and Michener, right lateral view; type specimen. Station 4681 (300-0 fathoms). X 1090. Fig. 4. Phalacroma giganteum Kofoid and Michener, structure of thecal wall near sagittal suture, which is serrated. Station 4699 (300-0 fathoms). X 1090. Fig. 5. Phalacroma striatum Kofoid, ventral view, somewhat tilted posteriorly; type specimen. Station 4705 (300-0 fathoms). X 700. Pig. 6. Phalacroma turbineum Kofoid and Michener, protoplasmic contents, apparently shrunken, showing the large, round nucleus with moniliform chromatin and three nucleoli. Station 4715 (300-0 fathoms). X 525. F'ig. 7. Phalacroma farus Kofoid and Michener, right lateral view; type specimen. Station 4737 (300-0 fathoms). X 1090. Fig. 8. Phalacroma striatum Kofoid, right lateral view; type specimen. Station 4705 (300-0 fathoms). X 700. MEM MUS COMP ZOOL DINOFLAGELLATA. PLATE 2 A. B. Streedain, dei \ , HEllOTYPE CO. BOSTON PLATE 3. PLATE 3. P^ig. 1. Phalacroma jnilrhrum Kofoid and jMichener, ventral view; type specimen. Sta- tion 4H99 (30n-0 fathoms). X 1090. Fig. 2. Phalacroma gigantcum Kofoid and Michener, seen from above. Station 4(399 (300-0 fathoms). Note the two pores on ventral side of epitheca, near anterior cingular list (on upper side of figure). X 7S5. Fig. 3. Phnlncroma liinhatum Kofoid and Michener, ventral view; type specimen. Station 4667 (300-0 fathoms). X 1090. Fig. 4. Phalacroma limhafiim Kofoid and Michener (?), right lateral view of specimen of uncertain allocatiop ; parasagittal lists apparently not yet fully developed. Station 4687 (300-0 fathoms). X 1090. Fig. 5. Phalacroma limbahim Kofoid and Michener, right lateral view; type specimen. Station 4667 (300-0 fathoms). X 925. Fig. 6. Phalacroma -pulchrum Kofoid and Michener, right lateral view; type specimen. Station 4699 (300-0 fathoms). X 1090. MEM. MUS, COMP. ZOOL. DINOFLAGELLATA, PLATE 3 T^'^^^^^^^W etJ, 'Se^.« »«*«*«* ..iWf ■ . « ®«ee' r^ • • •" 77 -^ '^s A. B. Streedain, del. HfcLiOTVPE CO. BOSTON PLATE 4. PLATE 4. Fig. 1. Phalarroma turbineum Kofoid and Michener, ventral view; type specimen. Sta- tion 4681 (300-0 fathoms). X 1090. Fig. 2. The same, dorsal view, somewhat tilted to the left. X 1090. Fig. ;5. Phalttcroma reticulafum Kofoid, right lateral view, somewhat tilted ventrally; type specimen. Station 4740 (300-0 fathoms). X 525. Fig. 4. Phalacroma fimbriatum Kofoid and Michener, right lateral view; type specimen. Station 4f;i3 (300-0 fathoms). X 1090. P'ig. 5. Phalacronia riiiculatum Kofoid, dorsal view; type specimen. Station 4740 (300-0 fathoms). X 525. Fig. (J. Plialacroma prwivxtum Kofoid and Michener, right lateral view; type specimen. Station 4742 (300-0 fathoms). X 1090. Fig. 7. The same, ventral view. Note the large cribriform plate, a unique feature in this genus. X 1090. MEM, MUS. COMP. ZOOL. DINOFLAGELLATA. PLATE 4 A. B. Streedain, do. *-.£l.iOT¥PE CO. BOSTON PLATE 5. PLATE 5. Fig. 1. Phalacruma crpuhum (Kofoid and Micliener), right lateral \ lew ; type .spffiiiicii. Station 4717 (300-0 fathoms). X 1090. Fig. 2. Dinophys-is collarisKoiokl and Micliciifr, right lateral view; type speeiinen. Sta- tion 4671 (300-0 fathoms). X 665. Fig. 3. Dinophysis jorgnismi, sp. nov., ventral view of type speeimen, the two halves of which have broken apart, showing that the larger portion of the longitudinal furrow and the entire flagellar pore belong to the right valve and that the two members of the fission rib belong to different valves. The valves were separated by the pressure of the eover-glass. Station 4727 (surface). X 1090. Fig. 4. Dinofurcula ultima (Kofoid). right lateral view; type specimen. Station 4711 (300-0 fathoms). X 1090. Fig. ':>. Ilistiophysis nigimi (Kofoid anil Miehener), right lateral view; tyjjc speeiinen. Station 4705 (300-0 fathoms). X 1090. Fig. (i. Dhiojurnilu ultima (Kofoid), \entral view, showing that the two posterior processes do not lie in the same plane but form an angle of about 20°; type specimen. Station 4711 (300-0 fathoms). X 1090. Fig. 7. Dinophy.six jorgensini, sp. no\-., right lateral \iew; type specimen. Station 4727 (surface). X 1090. Fig. S. Dinuphysis coUarig Kofoid and Miehener, ventral \iew; type specimen. Station 4671 (300-{) fathoms). X ()65. Fig. 9. Dinophysisuwczyi, sp. nov., right lateral view; type specimen. La Jolla, California. X 975. MEM. MUS, COMP. ZOOL. DINOFLAGELLATA- PLATE 5 A. B Stieedain de HfeHOTYPE CO. BOSTON PLATE 6. PLATE 6. I'"ij;'. I. A III jiliLttilciiid /in riciiiiilii l\(il'ciiil, I'ij^lil hilci:il \ lew of :iiilcniir |icirliiin nF IkmIv; type specimen. Sl;ilii)n 4740 (300-0 I'ailidiii.s). X 1100. V\ff. 2. Aiiijililsiilctila injlnfa Murray and V^ liitting, right lateral \ie\v of anterior portion of liody showing structural differentiation of thecal wall, reticulation of right sujcal list, and peculiar dilfcreiitiation of left sulcal list near flagellar pore. Station 4(iSt (300-0 fathoms). X 1100. Fig. 3. A III jiliisitlniid rxfni,s-a Kofoid, right lateral \iew of anterior ])ortion of Kody show- ing irregular rihs ;interiorly on sulcal lists; type si)ecimen. Station 4(100 (300-0 fathoms). X 1100. Fig. 4. Am iihiKDhiiid linricdiiilii Kofoid, dorsal \ iew of ])osterior portion of liody showing short antapical, not inflated posteriorly, and the i)ointed antapex lacking spinules; tyjie s])eci- nien. Station 4740 (300-0 fathoms). X 1100. Fig. ."). Aiiipliinolniin r.rfni.'td Kofoid, dorsal \iew of the inflated, cluh-shaped, and suh- truncate posterior portion of antapical lacking si)inules; type specimen. Station 4(i!)0 (3000 fathoms). X 2200. I''ig. (). Aiiijihisiilniid Idiiciiictd Kofoid, lateral \ iew of jxistei-ior portion of antapical show- ing the rounded hut not inflated antapex with a minute spinule (in typical specimens probably one mimite spinule on each valve) ; type specimen. Station 4740 (300-0 fathoms). X 2200. Fig. 7. A)iiplii.mlrnia iiiflaia Murray and Whitting, right lateral view showing protoplas- mic contents with rather large ovoidal nucleus, numerous small chromatophores, and meta- plasmic inclusions. Station 4081 (300-0 fathoms). X •''30. |-'ig. S. Aiiijililxi)lniia hiiicincid Kofoid, right latei'al \ iew of anterior portion of body showing " acce.s.sorx sulcal list " on left sulcal list ; tyi)c sjiccimen. Station 4740 (300-0 fathoms). X 2200. ■■'ig. 0. The same, lateral \ iew of antcricir jjroccss and midbody showing \ acuolatt'd pro- toplasm with large, obhing nucleus. X 1100. MEM MUS COMP ZOOL. DINOFLAGELLATA. PLATE 6 ;^''?;': •\-;V:;\ W A. B. Slreedam, del. HEL OTYPE CO. BOSTON PLATE 7. PIRATE 7. Am jihisiih ii'id s(hiiiiii}.sliniili 1 ,ciiiiiiciiii;iiiii. Shitmii 47M(i (iJIKI 1) hil lioiri.s). Fig. 1. liiglil ImIci'hI \ lew 111' iinli-ridr' |Kiriii)n (if iio(l\ . X I 1 00. Fig. 2. liiglit hitcnil \'w\\ of middle |iiiiiiiiii nf liodv slitiwiiig |iriilii|)lMsiiiic ciinlcnts soiiiewliat shniiiki'ii away from tlicfal wail. Xiicli'U.s of modi-rate size, elongate, somewiiat irregular, and with \'entral depression, in wliieli a eentrosome ('') is loeated. Chromatophores are rod-shaped, moniliforin, and longitu-^i I / i: A. B. Streedain. del. HtLtOTYPE CO. BOSTON PLATE 8. PLATE 8. Fi^. 1. Aiiijiliisiilriiid i/I(ihlf{ ra Stfin, \ciitral \ic\v of poslciioi' portion (if aiiliipicMl, sIidw- iiiK terminal knol), iifck-likc constric-tion witii protoplasinic sheatli in front of knol), iiiflalion in front of constriction, porulation, and two antapical sjjinnles, one on each valve. Station 4613 (300-0 fathoms). X 2200. Fig. 2. The same, right lateral view of anterior portion of hotly. X 1100. Fig. 3. Aiiiphl.iolniia nxiunguhita Kofoid, right lateral view of anterior portion of body; type specimen. Station 4740 (300-0 fathoms). X 1100. Fig. 4. AmphhoJcniu glohifcra Stein, right lateral \ie\v of posterior portion of antapical of specimen represented by figure 1. X 2200. Fig. ."). Aiiijiliisolfiria rectangulata Kofoid, right lateral view of posterior portion of ant- apical showing sliglit inflation and porulation. .Vntapex is truncate with four spinules arranged syuunetrically, two on each \alve, two dorsal and two ventral; only the two si)inules of right \alve are visible; type specimen. Station 4740 (,300-0 fathoms). X 2200. Fig. (). The same specimen, dorsal \iew of ])osterior portion of antapical showing iiilateral compre.ssion. X 2200. Fig. 7. The same specimen, right lateral view showing protoplasmic contents with large, elongate nucleus and pusule opening at flagellar pore. X 23o. Fig. 8. Aiiiplii.solcnia glohifira Stein, right lateral view of middle portion of ])rotoplasniir body showing ovoidal nucleus of moderate size, pu.sule opening at flagellar pore, and se\cr,il rather small, o\oidal chromatophores. Station 4fil3 (300-0 fathoms). X 5.30. Fig. 9. Aiiijjhi.ioli'iila rrrtaiiguldtd Kofoid, sublateral view of posterior portion of ant- apical; type specimen. Station 4740 (300-0 fathoms). X 2200. MEM. MUS COMP ZOOL DINOFLAGELLATA. PLATE 8 ^^ . ' ( A. B. Strepdam, cle HfellOTVPE CO. BOSTON PLATE 9. PLATE 9. Fig. 1. Amphisolcnia hispinosa Kofoid, right lateral view of anterior portion of body showing longitudinal ridge on neck; type specimen. Station 4605 (300-0 fathoms). X 1 100. Fig. 2. Amphisolcnia quadrispina Kofoid, ventral view of posterior portion of neck and anterior portion of anterior process showing flagellar pore, anterior portion of efferent duct of pusule, and cytopharyngeal depression; type specimen. Station 47 11 (300-O fathoms). X 785. Fig. 3. The same specimen, ventral view of anterior portion of body in front of portion shown in figure 2. X 785. Fig. 4. The same specimen, lateral view of posterior portion of antapical showing terminal knob, neck-like constriction in front of knob, inflation in front of constriction, porulation, and four antapical spinules placed symmetrically on the two valves. X 785. Fig. 5. Amphisolenia curvafa Kofoid, right lateral view of anterior portion of body show- ing longitudinal ridge on neck and "accessory sulcal list" on left sulcal list; type specimen. Station 4tJ05 (300-0 fathoms). X 1100. Fig. 6. Amphisolenia hispinosa Kofoid, lateral view of middle portion of body showing protoplasmic contents with rather large, elongate, irregular nucleus and, anteriorly and pos- teriorly, some small rounded or ovoidal chromatophores; type specimen. Station 4605 (300-0 fathoms). X 530. Fig. 7. The same specimen, dorsal view of posterior portion of antapical showing the trun- cate antapex with two long lateral spinules, one on right and one on left valve, and porulation. X 1100. Y\g. 8. The same specimen, right lateral view of posterior portion of antapical showing slight inflation and rounded antapex. X 1100. Fig. 9. Amphisolenia curvata Kofoid, ventral \ icw of po.sterior portion of antapical show- ing short, strong antapical spinule on left valve, irregular ridges, and porulation; in typical specimens probably one spinule on each valve; type specimen. Station 4605 (300-0 fathoms). X 1100. Fig. 10. The same specimen, right lateral view of posterior portion of antapical showing slight ii-flation and rounded antapex. X 1100. Fig. 11. Amphisolenia Icmmermanni Kofoid, right lateral view of posterior portion of antapical of atypic specimen showing abrupt expansion of posterior end, truncate antapex with only one spinule, and protoplasmic sheath in front of expanded posterior end. Station 4734 (oCO-O fathoms). X 1060. Fig. 12. 'Ihe sane, right lateral view of posterior portion of antapical of type specimen showing abiupt expansion of porulose posterior end and in front of this expansion slight infla- tion and piotoplasmic sheath; antapex is truncate with two rather strong sagittal spin- ules, both belonging to left \ alve. Station 4730 (300-0 fathoms). X 1400. MEM MUS. COMP. ZOOL. DINOFLAGELLATA. PLATE 9 '/■ • %: A. B. Streedam, del. HELIOTVPE CO. BOSTON PLATE 10. I'l.ATK 10. Fig. 1. Amphitiuhniu astrayalus Koi'oid, imkIi' latci-al \ip\v of ;uiterif>r portion of Ixxiy showing longitudinal ridge on neck and longitudinal ril) on right sulcal list; type specimen. Station 4713 (300-0 fathoms). X 1100. Fig. 2. Ampliisolcnia schwdcri Kofoid, right lateral view of posterior portion of antapical; type specimen. Station 4737 (300-0 fathoms). XllOO. Fig. 3. The same specimen, right lateral view of anterior portion of body. X 1 100. Fig. 4. The same specimen, ventral view of posterior portion of antapical showing slight inflation, porulation, and, on antapex, two small sagittal spinules both probably belonging to left valve. X 1100. Fig. 5. Ampliisolcnia com pi an at a, sp. nov., right lateral view of anterior portion of body. Station 4739 (300-0 fathoms). X 1100. Fig. 6. Ampliisolcnia astragalus Kofoid, right lateral view of posterior foot-like portion of antapical showing porulation, rounded heel without any " heel-spinule," the rather strong, pointed spinule on antapex, and the rounded, wart-like protuberance somewhat in front of antapex; type specimen. Station 4713 (300-0 fathoms). X 2200. Fig. 7. Ampliisolcnia complanata, sp. nov., dorsal view of posterior portion of antapical showing slight inflation, porulation, and, on antapex, two rather strong sagittal spinules both probably belonging to left valve. Station 4739 (.300-0 fathoms). X 1100. F'ig. S. Ampliisolcnia astragalus Kofoid, right lateral view of midbody showing proto- plasmic contents with rather large, elongate nucleus, some rather small, rounded or ovoidal chroniatophores, and metaplasmic inclusions; type specimen. Station 4713 (.300-0 fathoms). X 530. Fig. 9. Aniphisolrnia complanata, sp. no\-., right lateral view of posterior portion of ant- apical of specimen represented by figure 7 showing very slight inflation. Station 4737 (300-0 fathoms). X 1100. Fig. 10. AmpMsolenia astragalus Kofoid, ventral \ i«w of posterior portion of antapical; type specimen. Station 4713 (.300-0 fathoms). X 2200. MEM. MUS. COMP. ZOOL. DINOFLAGELLATA. PLATE lO A. B. Streedain, de MtllOTYPE CO. BOSTON PLATE 11 PLATE 11. Fig. 1. Ampliisolrnia trKncnfa Kofoid and Michener, ri{;ht lateral view of anterior portion of body ; right sulcal list omitted ; type specimen. Station 4733 (surface). X 1 100. Fig. 2. Amphisolcnia palaeotheroides Kofoid, dorsal view of porulose posterior portion of antapical. Distal end foot-like and bent to the right. Foot-like portion about twice as long as wide. Left valve with three strong spinules: one, the "heel-spinule," placed at point of bending, the two others distal. Right valve with one spinule only which is placed between the distal spinules of left valve, and about as strong as or somewhat stronger than distal spinides of left valve; type specimen. Station 4732 (300-0 fathoms). XllOO. Fig. 3. The same specimen, right sublateral view of posterior portion of antapical show- ing inflation. X UOO. Fig. 4. The same specimen, right lateral view of anterior portion of body. Left sulcal list with fission rib. With numerous pores. X 1100. Fig. 5. Aniphigolenia nsyitiniefrim Kofoid, right lateral view of posterior portion of ant- apical of type specimen showing inflation. Station 4732 (300-0 fathoms). X 1100. Fig. 6. Amphisolenia asymmetrica Kofoid, \entral \iew of anterior portion of body of specimen described by Kofoid (1907a) as A. dolichoccphalicu. Cingular lists not yet developed following binary fission. Station 4728 (300-O fathoms). X 530. Fig. 7. Amphisolenia asymmetrica Kofoid, posterior portion of antapical of type specimen in dorsal view. It differs from .1. palaeotheroides (fig. 2) mainly in having the foot-like portion about seven to eight times longer than wide. Station 4732 (300-0 fathoms). X 1 100. Fig. S. Amphisolenia clavipes Kofoid, right lateral \-iew of anterior portion of body, with distal portion of efferent canal of pusule; type specimen. Station 4736 (300-0 fathoms). X 1 100. Fig. 9. The same specimen, left lateral view of posterior portion of antapical. Distal end somewhat inflated and gently curved toward the ventral side. Antapex is truncate, with two rather small sagittal spinules, one dorsal and one ventral, both probably belonging to left valve. X 1100. Fig. 10. The same specimen, ventral view of posterior portion of antapical. Distal end strongly bent to the right; with rounded antapex. X 1100. Fig. 11. The same specimen, lateral view of the middle portion of body showing proto- plasmic contents. Behind the rather large, elongate nucleus some rod-shaped ehromatophores are seen. X 1100. Fig. 12. Amphisolenia truncata Kofoid and Michener, lateral view of posterior portion of body; antapex truncate, not inflated, and without spinules; type specimen. Station 4733 (surface). X 1100. Fig. 13. Amphisolenia asymmetrica Kofoid, right lateral view of anterior portion of type specimen showing the extremely elongated head. Station 4732 (300-0 fathoms). X 530. MEM. MUS, COMP, ZOOL. DINOFLAGELLATA. PLATE 11 >^K ■> 70 V 11 V* IS A- 3. Streedain, del. HELIOTYPE CO. BOSTON PLATE 12. PLATE 12. Fig. 1. Amphisolenia bifurcata Murray and Whitting, subveiitral \ iow of posterior portion of body showing bilateral compression and porulation. Distal end of both antapical stem and antapieal branch has the same structure as distal end of antapical in .1. pulmuia; see explana- tion of figure 7. Distal spinules of left valve unintentionally omitted. Station 4099 (HflO-O fathoms). X 530. Fig. 2. Aiiijihi.sdli'iiia ihrinax Schiitt, lateral view of midbody and anterior portion of antapical showing protoplasmic contents with large dumljbell-shaped nucleus, fairly innnerous large, ovoidal or spheroidal chromatophores, and numerous groups of three to five very small, spheroidal bodies of unknown nature. In anterior portion of midbody a pusule is seen. Shape of nucleus probably indicates approaching di\ision. Station 4732 (.'500-0 fathoms). X 530. Fig. 3. Amphisolenia bifurcata Murray and Whitting, lateral view of posterior portion of body. Station 4(')99 (.300-0 fathoms). X 530. Fig. 4. Atiiphisohiiid paliiiata Stein, right lateral view of posterior portion of antapieal showing inflation, porulation, and three strong distal spinules. Station 4737 (300-0 fathoms). X 2200. Fig. 5. Amphisolenia bifurcata Murray and Whitting, left lateral view of middle portion of body showing protoplasmic contents with ovoidal nucelus, many spheroidal metaplasmic inclusions of various sizes, and anterior portion of efferent canal of pusule. Station 4r)99 (300-0 fathoms). X 530. Fig. 6. Amphisolenia thrinax Schiitt, right lateral view of atypic specimen with antapical stem in front of first antapical branch very short, and with antapical stem behind second branch very inflated and of about the same shape and size as midl)ody and anterior process. Station 4.594 (300-0 fathoms). X 120. Fig. 7. Amphisidcuia palmafa Stein, ventral view of porulose posterior portion of ant- apical. Distal end is foot-like and bent to the right. Foot-like portion is two to three times longer than wide. Left valve has three strong spinules: one, the "heel-spinule," placed at jjoint of bending, the two others distal. Right valve has one spinule only which is placed between and sometimes larger than distal spinules (jf left valve. Station 4737 (300 0 fathoms). X 220. MEM. MUS. COMP. ZOOL. DINOFLAGELLATA. PLATE 12 ■\ ^ '^. . . J > ?3 «* A- B. Streedain, del. HELlOiyPE CO. BOSTON PLATE 13. PLATE 13. Fig. 1. Amphisolenia quadricauda Kot'oid, ventral view of anterior portion of right valve showing that the line of fission runs along middle of head, along left side of sulcus, crossing left sulcal list at fission rib, which is seen just in front of flagellar pore, to the left of flagellar pore and cytopharyngeal depression, and along the middle of anterior process; type specimen. Station 4095 (300-0 fathoms). X 1100. Fig. 2. The same specimen, left vahe of posterior portion of antapical stem seen from within showing elongate inner opening of cavity of " heel-spinule," cross-ridge near antapex, and porulation. X 1 100. Fig. 2a-d. Four optical cross-sections of thecal wall of portion represented by figure 2 at places corresponding to positions of the four figures, showing shape and thickness of thecal wall. X 1100. Fig. 3-5. The same specimen, distal portions of left valve of third, second, and first ant- apical branches, respectively, seen from within showing that the first branch (fig. 5) has " heel- spinule" and is porulate, while the second and third branches lack "heel-spinule" and poru- lation. X 1100. Fig. 6. The same specimen, right lateral view of anterior portion of right valve showing the part of left sulcal list, behind fission rib, that belongs to right valve. X 800. Fig. 7. Amphisulcnia qiiinquccauda Kofoid, somewhat twisted; anterior portion seen in ventral view, posterior in right lateral view; type specimen. Station 4739 (300-0 fathoms). X 120. Fig. 8. Amphisolenia quadricauda Kofoid, left lateral \iew of anterior portion of left valve showing the part of left sulcal list, in front of fission rib, that belongs to left valve, ribs of the cingular and sulcal lists, and ridges and porulation of thecal wall; type specimen. Sta- tion 4695 (300-0 fathoms). X 825. Fig. 9-13. Amphisolenia quinquecavda Kofoid, distal portion of antapical stem and of fourth, third, second, and first antapical branches. Antapical stem (fig. 9) and first antapical branch (fig. 13) are porulose and of pcihiuita type (PI. 12, fig. 7) with long and slender "foot." Second, third, and fourth antapical branches (figs. 12, 11, 10) lack "heel-spinule" and porula- tion; type specimen. Station 4739 (300-0 fathoms). X 1100. Fig. 14. The same specimen, right lateral view of anterior portion of body showing struc- tural diff'erentiation of lists and porulation of thecal wall, anterior portion of efferent canal of pusule, protoplasm of anterior process, and cytopharyngeal depression. X 1 100. MEM, MUS, COMP. ZOOL. DINOFLAGELLATA, PLATE 13 14 A. B. Slr^«.)a,n. da^. HELIOTVPE CO. BOSTON PLATE 14. PLATE 14. Triposolenia inincata Kofoid. Fig. 1. Right latenil \ie\v of mitfrior ijortioii of l)0(ly showing surface differentiation of thecal wall. Of the three ribs of left sulcal list, the anterior is the fission rib. Station 4742 (300-0 fathoms). X 1100. Figs. 2, 3. Right lateral view of posterior portions of dorsal and ventral antapicals of specimen represented by figure 1. X 1100. Triposolenia dcpri'ssa Kofoid. Fig. 4. Right lateral view. Station 4580 (300-0 fathoms). X 530. Fig. 5. Right lateral view showing protoplasmic contents with large ovoidal nucleus and pusule opening at flagellar pore. Station 4580 (300-0 fathoms). X 530. Fig. 6. Right lateral view of anterior portion of body. Station 4580 (300-0 fathoms). X 1100. Figs. 7, 8. Right lateral view of posterior portions of dorsal and ventral antapicals showing thickness of thecal wall. Station 4580 (300-0 fathoms). X 1100. Fig. 9. Dorsoventral view of body. Station 4580 (300-0 fathoms). X 550. Triposolenia intermedia, sp. nov. Fig. 10. Left lateral view of type specimen showing fine striation of thecal, wall. Station 4580 (300-0 fathoms). X 530. Triposolenia ambvlatrix Kofoid. Fig. 11. Right lateral view of type specimen. Station 4711 (300-0 fathoms). X 530. Fig. 12. Right lateral view showing protoplasmic contents with large rounded nucleus, many rounded or ovoidal, highly refractive bodies, and distal end of efferent canal of pusule. Station 47/1 (300-0 fathoms). X 530. Fig. 13. Dorsoventral view of body. Station 4711 (300-0 fathoms). X 530. Fig. 14. Right lateral view of anterior portion of body. Station 4711 (300-0 fathoms). X 1100. MEM. MUS. COMP. ZOOL DINOFLAGELLATA. PLATE 14 ^'^ A. B. Streedain, d-;!. HELIOIYPE CO. BOSTON PLATE 16. PLATE 16. Triposolenia longicornw Kofoid. P'ig. 1. Right lateral view. Station 4583 (300-0 fathoms). X 530. Fig. 2. Right lateral view showing protoplasmic contents with large rounded nucleus and pusule opening at flagellar pore. Station 4.583 (300-0 fathoms). X 530. Fig. 3. Dorsoventral view of body. Station 45S3 (300-0 fathoms). X 530. Fig. 4. Right lateral view of anterior portion of body. Station 4583 (300-0 fathoms). X 1100. Figs. 5, 6. Right lateral view of posterior portions of dorsal and ventral antapicals showing thickness of thecal wall. Station 4583 (300-0 fathoms). X 1100. MEM, MUS COMP. ZOOL. DINOFLAGELLATA. PLATE 15 A. B. Streedain, del. HELI0TYP6 CO. BOSTON PLATE 16. PLATE 16. Fig. 1. Ornithoccrcus stelni Schiitt, s. str., right lateral view. Under the pressure of the cover-glass, the anterior cingular list has split dorsally and ventrally and the posterior cingular list dorsally. La JoUa, California. X 625. Fig. 2. Ornithocercus uplcndidus Schiitt, seen from above, showing the structural differen- tiation of the anterior cingular list (structure of theca omitted). The sinuation on the left side of the figure is on the ventral side of the organism. La Jolla, California. X 975. Fig. 3. Ornillwccrcus magnlficus Stein, ,l. str., right lateral view. La Jolla, California. X975. Fig. 4. Ornithoccrcus splendidus Schiitt, right lateral view. La Jolla, California. X 975. MEM. MUS, COMP. ZOOL. DINOFLAGELLATA. PLATE 16 •. • • . .. \ \ (• r- *. I. pCc I / TX^^s^ / / / ^y s r t X^^ '' -s' ^ O ro ^ -c- ■*" Hstern Pacific during the two winter quarters. It is based on the combined data contained in the " Quarterly Current Charts for the Pacific Ocean" published by the Hydrographic Department, Admiralty, London, and in the Pilot Charts issued quarterly by the Hydrographic Office of the United States Navy. V Miff ^tV', '' mmm ^ w^ ^ /h ■■% H 1 F #.iff Date Due Harvard MCZ Library lllll 11 II III I II I " • '" •'^" I I I I 3 2044 066 302 233