I
In 1977, I mentioned the existence of two general frames of mind in the study of a particular taxonomic group. In one, the intent is to understand, more or less idiographically, the taxonomic group. In the other, constituents of the taxonomic group may be treated as experimental animals, simply because they are more useful than other groups of organisms for the more or less nomothetical study of causal relations in biology. In reality, however, a biologist's mind seems to oscillate between the two above attitudes, perhaps because in order to successfully utilize a particular taxonomic group in experimental studies one must know it as well as those specializing in the taxonomy of that group. Balance between the two attitudes seems to be the only way to attain proper and self-controllable biological knowledge and avoid monstrous deviations.
When the checklist of Ostracoda from Japan and Southeast Asia was compiled, we recognized that our idiographic knowledge of Japanese Ostracoda was more scanty than we had thought and that the place where we should send the first expedition to investigate ostracod faunas was Japan itself. Therefore, utilization of Ostracoda as experimental animals was still far away.
The first step in a series of idiographic studies of Japanese Ostracoda was commenced by building upon taxonomic knowledge that had been summarized in the checklist and directing effort in the following two directions. One was to determine the spatial and temporal distribution of Ostracoda in relation to the surrounding environment with hope "to elucidate paleontologically significant historical events which may be complicated but which may be interpreted in terms of biological theories" (Hanai, 1977, p. 85). The other was to study individuals in terms of the causal relation between structure of calcareous exoskeleton, which remains in the fossil record, and internal cells, which are responsible for secretion of the hard tissue but are easily destroyed after the death of the animal. The underlying hope is that certain groups of Ostracoda may become "marine Drosophila" in the not-too-distant future. Some progress has been made in both directions, and a few definite steps may have been made toward the understanding of Japanese Ostracoda.
Among the studies of the time and space distribution of ostracod faunas, the following questions are of special interest. How and how far will the transgressions and regressions, though the events are small in scale and local in nature, affect the environment and hence the change of ostracod fauna? How imperfect are fossiliferous sediments and consequently fossil records in terms of representing time, and how fragmental is their preservation in space they once occupied-that is, the problem of "imperfection of the fossil record"? What is the rate and way in which faunal compositions change over geologically short time intervals-namely, the "tempo and mode" of species migration and of new element addition?
II
Before discussing our research, a few terms commonly used in our papers will be explained. Observation of the Petri dish culture of species of different genera, which occur together in one bottom sample, shows that one species ignores the presence of the other. Thus, the coexistence of ostracod species in one sample may not imply any direct biological interrelation between the species. Study of the distribution pattern of some shallow water Ostracoda shows that no two species share exactly the same area of distribution, though some species quite often have overlapping distribution ranges, suggesting similar but different environmental requirements of the species. However, these differences in environment are often very small and not at all detectable in evidence preserved in the sediment. Further, areas actually occupied by ostracod species do not always coincide with the area where those species can potentially live. Thus, a sedimentary unit which is occupied or potentially occupied by an assemblage of living Ostracoda characteristic of a particular environment becomes a practical unit for research applicable to paleontology, and the sum of bilogical features of that unit exhibited by Ostracoda may be called ostracod biofacies. Biofacies are named by words describing a particular environment, for example, bay mouth biofacies, brackish water inlet biofacies, etc. Here again the constituent species of one biofacies do no necessarily have any direct or even indirect relationship with any other, except for the accidental relation arising from the preference for a similar environment.
Paleoecologists have discussed ad nauseam the preburial and postburial relation between the blocoenosis and the thanatocoenosis, pointing out that a group of fossils that occur together in a particular lithologic unit of a certain stratigraphic level, called fossil assemblage, may not directly indicate the particular environment in which the constituent species lived. Therefore, fossil assemblages of Ostracoda are named not by using environmental terms but by using characteristic constituent species, for example, Spinileberis quadriaculeata-Nipponocythere bicarninata assemblage, etc. The assemblage is a combination of ostracod species, which lived in several adjacent environments. However, if fossil evidence is adequate and proper knowledge of correlation between the distribution of living Ostracoda and of dead carapaces, applicable to that particular case, is available, biofacies-that is, an assemblage of Ostracoda characteristic of a particular environment when alive-may be detectable with high certainty through investigation of fossil assemblage. It is evident that fossil assemblages range from those in which biofacies of the past are detectable to those in which allochthonous occurrence of ostracod carapaces completely masked the original biofacies. In the former extreme case, the term biofacies is even applicable directly to the biofacies of the past inferred from the fossil assemblage. However, Frydl, in his paper on Holocene Ostracoda, marked the biofacies of the past by prefixing the term biofacies with the first letter of the characteristic species in a fossil assemblage, for example, S biofacies for Holocene Spinileberis quadriaculeata biofacies, and he showed in detail the relation between inferred biofacies of the past and the present-day biofacies descended from them. In the latter case, Yajima employed a large-scale environmental term which includes several adjacent biofacies once probably characterized by distinct assemblages of living Ostracoda, but undistinguishable at present, to cover safely all the variations within a set of fossil assemblages, for example, subtidal sand assemblage, etc.
III
Even though the knowledge of change of population size and distribution of ostracod species due to transgression and regression is important from a paleontological point of view, there is a shortage of concrete data. Because the sediments of the Jomon transgression include material which can be dated by the 14C method, it is possible to obtain their age and thus trace lateral shifts of ostracod species on a minute time scale. However, Jomon transgression sediments are generally not exposed and are thus difficult to examine in detail. In the recently uplifted southern part of the Boso Peninsula, these sediments form marine terraces where they are accessible to direct observation. These terraces are the result of both eustatic sea level changes and tectonic uplift of the area, and they and the sediments forming them can be used to infer the mode and rate of relative sea level changes, and hence the changes of fauna due to transgression and regression.
Frydl's work shows that Ostracoda of the Spinileberis, Keijella and Keijella-Nip-ponocythere assemblages, which occupied the inner bay areas of deeply incised coves during the transgressive phase, died out during the regressive phase. Only Ostracoda of the Neonesidea-Schizsocythere assemblage and of further seaward located assemblages were able to retreat during the regressive phase. This demonstrates that even smallscale transgression and regression can have a profound influence on population size and the size of area occupied by ostracod species constituting the innermost bay assemblages. In the case of a large bay (i.e., Tateyama Bay) or slower regression, Keijella and Nipponocythere could retreat and continue to occupy the central bay mud biofacies. In general, the effect of a regression does not entirely depend on the scale of the regression itself, but also considerably on the scale of the preceding transgression.
A field survey of the Hamana-ko Bay made by Ikeya and Hanai illustrates the condition of a drowned valley after a long period of nearly constant sea level. At present the mouth of the valley is almost closed by a sand bar, and the enclosed bay is being filled from the entrance inwards with well-sorted coastal sand supplied by tidal current. Continuation of one condition may often emphasize or exaggerate some phenomena and minimize or efface others. In drowned valleys with a wide bay mouth and thus good marine water circulation, the area occupied by Spinileberis lies close to that occupied by Keijella or Nipponocythere, as shown by Frydl. In Hamana-ko Bay, however, the area with abundant Spinileberis is in direct contact with the Main channel biofaces, which are characterized by abundant Hemicytherura and Semicytherura living on the sandy bottom under direct influence of tidal current. Further, it seems probable that even if the present nearly standstill condition of the sea level continues, the area of Spinileberis may finally be effaced. Keijella and Nipponocythere have already disappeared from the present Hamana-ko Bay, though they are abundant in the Pleistocene sediments in the same area, which were probably deposited in a bay with wide bay mouth.
All of these facts suggest that the extreme change in selection pressure caused by transgression and regression exerts the strongest influence on ostracod species inhabiting the innermost parts of bays, which are often influenced by freshwater inflow, and the change may provide the trigger of a mechanism which develops unique elements of the subsequent transgressive fauna. For this reason Quaternary Spinileberis quadriaculeata and S. fumyaensis, and Cytheromorpha acupunctata represent suitable material for studies dealing with the relation between speciation and morphological variation as a function of time. Influence of selection pressure on offshore species, living in areas with good marine water circulation, seems relatively weak.
Further, when the change of population size and the size of the area occupied by a particular species of Ostracoda caused by a small-scale regression is extrapolated to a large-scale regression, such as Oligocene-Lower Miocene regression, the effect of the regression on some species appears to be drastic. Schizocythere with Paleogenetype ornamentation, for example, was completely replaced, on a worldwide scale, by the Neogene type during the period of the Oligocene-Lower Miocene regression (Hanai, 1970).
IV
The relation between phenomena occurring in the time span called ecological time and those occurring in geological time is one of the most important problems of paleontology. Upper Pleistocene sediments of northern Boso Peninsula under discussion were deposited between approximately 263,000 (marker tephra GoP) and 132,000 (marker tephra KlP) years B. P. and contain abundant fossils. They are thus particularly well suited for investigation of the faunal changes occurring in the time interval lying between 14C (1,000 years) and biostratigraphical (1,000,000 years) time scales.
The Upper Pleistocene sediments of the area mentioned above were deposited during a period characterized by two geologic phenomena. One is volcanic activity, which produces volcanic ash, thus inserting exact time planes across various types of sediments in several horizons and further providing material that can be dated by the fission track method. The other is glacial and interglacial eustatic sea level change, which produces cyclic change of the environment and thus repetitious faunal sequences.
Yajima's study illustrates faunal changes in offshore areas of a large bay (i.e., Paleo-Tokyo Bay), complementing studies of faunal changes occurring in the nearshore environment at the transgression-regression front. Five sedimentary cycles (even though incomplete) separated by diastems can be recognized in the sediments deposited over a period of approximately 130,000 years.
In the last of the five cycles, which was formed by the Shimosueyoshi transgression, sedimentary sequence (Toyonari Member) is represented by a nearly complete cycle in the southern area, with nonmarine sediments in its lower and upper sequences and marine sediments of high sea level phase in its middle sequence. In the northern area (Kioroshi Member), depositional sequence starts with nearshore marine sediment of the high sea level phase, which directly overlies the lower cycle. Periods of nondeposition or erosion represented by the diastems generally get longer landwards and probably extended over a considerable time span, comparable in length to the periods of rising and stable high sea level. The evidence of events which occurred during the lowest sea level is not preserved in sediments of the Paleo-Tokyo Bay. The nearshore sediments of the regressive sea may be hidden somewhere between the Bay and the site of the deep sea cores.
The imperfection of the fossil record in sediments of a regressive sea of this magnitude both in time and area has so far been the "fatal imperfection" which can only be overcome by inference based on extrapolations from the result of small regressions and experimental simulation.
V
Composition of ostracod fauna is controlled mainly by the nature of local substratum and the salinity, oxygen content and other characteristics of sea water. On a 10,000-year scale these factors were influenced by the transgression, regression and sedimentary filling-up during sea level standstill. Thus the transgressive and regressive shift of sediments, which form the substratum, seems to prescribe the mode of the shift of ostracod fauna as a function of time, and it may be the basic controlling mechanism.
When considering events on a 100,000-year scale, in addition to the above changes, large-scale modification of topography and resulting alteration of sea water circulation must also be taken into account. This modification will, through change of paleogeography and of current circulation, give populations opportunities to interbreed and may aid some ostraced species, aided by their unchanging ecological requirements and homeostatic mechanisms, to maintain phenotypic status quo for a long time on a biostratigraphical time scale.
Paleo-Tokyo Bay was initially open to the east. Each of the three sedimentary cycles represented by the Yabu and Kamiizumi Members of the Yabu Formation and the Kiyokawa Formation has a middle horizon which consists of sediments deposited at the time of the highest sea level. The presence of warm water pteropods and bivalves indicates that higher sea level resulted in the opening of the Paleo-Tokyo Bay to the south as well as to the east, though only temporarily. Ostracod assemblages are, however, dominated by warm temperate species, suggesting bottom water staying in the Bay. The Shimosueyoshi transgression again connected the Paleo-Tokyo Bay to the south during the period of highest sea level, and the sea retreated toward the east as well as toward the south to form the present Tokyo Bay. It is highly possible that many ostracod species which inhabited the shallow water of the Paleo-Tokyo Bay followed the retreating sea to the inner part of the present Tokyo Bay.
VI
Ostracod faunas are also affected by changes in temperature. During the interglacial period of rising sea level, shallow thermophilic species extended their geographical range of the area of reproduction along the coast toward the north. The extension was controlled by a stepping-stonelike distribution pattern of substratum favorable to a given species. Ostracod migration over these 'stepping-stones" covering distances of several hundred kilometers along the coast may have occurred in a relatively short period of time of about 100 years or less. In the case of the Jomon transgression, at least one species, Ambocythere japonica, common in subtropical areas, appeared in shallow water sediments of the southern part of the Boso Peninsula deposited during the period of high sea level. It is interesting to note that thermophilic species, for example, Trachyleberis niitsumai, Neocytheretta sp., are represented in the area of their northernmost extent by abundant instars, but adult carapaces have been found only rarely or not at all, suggesting that the reproductive area of these species in the northern marginal area of the species range may be relatively small in comparison with the area of instar distribution.
Analogously, during the glacial period, cryophilic species might have extended their distribution toward the south along the coast of the retreating sea. It is, however, quite rare to find evidence of their migration, as it probably lies somewhere between the present shore and the site of the offshore drillings. The rareness may be accentuated by the fact that the nearshore environment predominated along the transgressive sea became a minor element along the shore of the retreating or retreated regressive sea. Cryophilic species, which migrated from the north and lived in the retreated sea of the glacial stage, entered the Paleo-Tokyo Bay with the rising sea of the interglacial transgression. Examples are given by Finmarchinella (Finmarchinella) uranipponica, Howeina camptocytheroidea, H. higashimeyaensis and Robertsonites reticuliforma. Cryophilic species are relatively abundant in the sedimentary cycle represented by the Kamiiwahashi Formation, in which warm water influence is not detectable even at the time of the highest sea level. In the sediments constituting the remaining four cycles, presence of warm water pteropods and bivalves indicates that the increased sea level resulted in the opening of the Paleo-Tokyo Bay also to the south. However, the increased sea level also caused deepening of the bay, and no marked influence of the warm water current is apparent in relatively deep ostracod faunas.
Elements may originate either as unique elements, which developed in response to extreme change in selection pressure during transgression and regression, or as marginal isolates of cryophilic species. In either case they are added to shallow water fauna during transgressive migration. It is interesting to note that the cryophilic species, which migrated from the north, are without exception smaller than the closely related northern forms.
The metaphysical domain lying between the paleontological facts and the biological mechanisms will not necessarily be overcome either by waiting passively for the new advancement of favorable biological theories or by speculations based on reallocation of insufficient evidence. In order to reduce the domain, tracing the factual tracks offaunal change and perceiving the missing details of the imperfect fossil records may be one approach, and a study of living animals viewed from the paleontological angle and finding out actively the biological mechanisms explaining paleontologically interesting phenomena may be another approach.
VII
Prior to the selection of experimental animals for study at the individual level in terms of the relation between the structure of calcareous exoskeleton and internal cell, ten years elapsed in collecting materials to find the habitat of species suitable for the study and, in examining materials for their viability, determining how long one can keep specimens alive and developing techniques to control culturing conditions. Hanai (1977) pointed out four basic conditions of Ostracoda specimens suitable for a culturing experiment: 1) ubiquitous and obtainable both in any season and in any place, 2) tolerant and standing up to environmental fluctuations produced unintentionally by culturing, 3) easy to feed, and 4) easy to reproduce. Because long-term culturing was not necessary for this study, conditions 3 and 4 were ruled out. Among the species which meet conditions 1 and 2 and are thus easy to culture by the simple Petri dish technique and easy to keep for a time in laboratory, Keijella bisanensis, a large species with simple but definite surface ornamentation, was selected. The movements of individuals are observable in a Petri dish with the naked eye, and thus are easy to manipulate.
Many of the methods of foraminiferal culture, which were explained briefly in Ikeya et al. (1976), were an application of the methods developed actually for ostracod culture in our laboratory. For collecting ostracod specimens, Ockelmann-type sledge sampler modified for semiquantitative sampling is used. From a specific station, mud samples which contain approximately 500 to 2,000 individuals of K. bisanensis are collected by a single sampling covering a bottom surface of approximately 15 × 1,000 cm2.
Mud samples are placed in a shallow pail and left in the laboratory, keeping the temperature of the sea water as it was on the sea bottom. Within about ten minutes, small organisms, including K. bisanensis, crawl out from the disturbed mud toward the mud surface. After clay and mud particles settle on the bottom of the pail and the sea water becomes clear, surface layer mud 1 mm in thickness is sucked up and moved into small Petri dishes together with the soft bottom suspension and sea water using a pipette with an opening of about 2.5 mm in diameter, K. bisanensis crawls just below the sediment water interface, but its way of movement is detectable through the characteristic movement of the mud surface. Phototaxis of K, bisanensis can be utilized to concentrate individuals. Annelids, small Crustacea other than K. bisanensis, molluscan larvae, etc., which consume oxygen while alive and produce poisonous substances after death, are removed. Thus K. bisanensis are kept alive for a maximum of 4 months or so in Petri dishes placed in an incubator at about 15°C. Replacement of half of the water in the Petri dishes with fresh sea water every .4 days facilitates culturing.
Observation of the life cycle of this species showed that experiments on development of eggs and on ecdysis ought to be made during the period from October to December and from November to April of the next year, respectively. In the other periods of a year, both the appearance of eggs in the female specimens and the occurrence of young instars become very rare. Details of the observations on the local variation and seasonal fluctuation of population structure in a deme are summarized by Abe (unpublished thesis, University of Tokyo, 1981).
VIII
Studies on the reticulation pattern of ostracod carapace were actually started with Pokorn ý's demonstration (1964, 1969) of individual meshes and ridges which grouped meshes into areas. The purpose of his study was simply to classify forms which differ from each other only in minor sculptural features. Liebau (1969, 1971, 1975a and b), following the method used by Pokorn ý, utilized the constancy in number and arrangement of the mesh pattern appearing on the carapace surface to trace phylogeny and to establish classification of Trachyleberididae s. 1. Independent of the studies on reticulation pattern, the distribution pattern of normal pore canals has been noted as similar in all specimens of one species, following a definite pattern of distribution (Triebel, 1941; Morkhoven, 1962; Plusquellec and Sandberg, 1969). Hanai (1970) illustrated a remarkably conservative and almost identical pattern of pore distribution within one species, by superimposing the distribution pattern of one specimen on that of another.
These two lines of studies were accelerated by the development of the technique of scanning electron microscopy. It provided us with extreme details of the sculpture and structure of the ostracod carapace and stimulated us to introduce a diversity of minute terminology into the description of the sculpture of ostracod carapace. To describe and to name details of carapace morphology without seeking their biological meaning was largely a matter of refinement of laboratory technique, but it was an unavoidable step of study and soon combined the idea of the constancy of reticulation pattern with the idea of the constancy of distribution of pore canals-this time, especially of pore cones by giving reference points of pore cones on the reticulation walls.
The second step of study seems to have developed in two directions. Benson (1972, 1974, 1975) approached the surface sculpture and form of carapace from its mechanically functional skeletal structure, through constructing abstract models of carapace sculpture and form. Liebau (1977, 1978) tried to distinguish evolutionary levels in an evolutional trend of cytheracean carapace ornamentations. In the primitive stage of ornamental genetics, ornamental elements vary within a species in number and arrangement, and genetic changes affect the entire ornamentation or the inexactly defined part of it. In the most advanced stage of ornamental genetics, ornamental elements are constant in number and arrangement within a species, and genetic changes may even affect a single element. Liebau also recognized an intermediate stage between the two stages. Liebau's work seemed to have proved that practically everything on the ostracod carapace is genetically controlled (Sohn, 1975).
There is, however, another approach from the viewpoint of developmental biology. The short-circuit connection between precisely placed ornamentation, including meshes and pores, and their genetic control may result in the oversight of the importance of the study of developmental mechanisms. Thus Hanai (1977) emphasized the necessity of developmental study of the spacial pattern of specifically placed pore canal openings on the ostracod carapace, adopting the term "organules" coined by Lawrence (1966) originally for specialized structures such as tactile bristles and secretory glands in integument of insects in the latter's paper on the pattern formation.
Studies on the carapace surface of Japanese Paijenborchellini (Hanai, 1970) suggest that the diversity of ornamentation of the tribe is explainable through modification of mesh pattern into two directions: one from reticulate to ridged (ex. Hanaiborchella miurensis to H. triangularis) and the other from reticulate to spinose (ex. H. miurensis to H. spinosa). The fact suggests that the mesh pattern of a certain size, excluding secondorder reticulation, may form the basis of carapace ornamentation, and ridges and spines may be modifications of walls.
The smooth area of the carapace surface of genus Neomonocertatina, which is surrounded by strong ridges, reveals the lining of mesh structure under transmitted light. Further scanning electron microscopy demonstrates that the mesh pattern of ornamentation is observable on the adequately etched surface of a certain species, which have been described as having a smooth surface. Although the boundary between meshes does not always turn into the ridges of carapace surface, the ridges of surface ornamentation, if present, always coincide with the boundary between the meshes. This further suggests that the mesh pattern always underlies the cytheracean carapaces, homologously with the cuticular polygones of certain insects and other arthropods, which correspond to the underlying epidermal cells.
Pore canals of trachyleberidids usually open on the ridges of the mesh, either at the junction of the ridges or on the ridge between the junctions. When a pore canal opens on the floor of the mesh, it is located close to the ridge and the ridge quite often extends and is connected to the pore (Liebau, 1978). It may present an interesting problem to relate the precise locations of pore canal openings to the precise location of meshes and further to the lining of epidermal cells, approaching them from the pattern formation of meshes and pores.
The promise of the study depends, in general, on the selection of experimental species. Among the few species that I have already examined for the ease of keeping them for a while in the laboratory, a trachyleberidid, Keijella Usanensis, was selected for close examination, because the species is large in size and easy to manipulate and is ornamented by simple uniform meshes of similar size surrounded by ridges with pore canal openings.
Okada's work explained the constancy of mesh pattern. Cell junctions between outer epidermal cells, which adjoin each other, are strengthened by development of apical desmosomes and septate junctions. Fibrille de soutien (Rome, 1947) or supporting fibers (Kesling, 1951) consisting of clustered columns of microtubles connected to basement membranes of both outer and inner lamella cuticules by conical hemidesmosomes, and running perpendicularly to shell lamella through the outer and inner epidermal cells. On the boundary between inner and outer epidermal cells, where the columns meet, microtubles are connected by means ofintermediate junctions. Thus, in addition to the cell junction of the apical desmosomes and septate junction, supporting fibers probably serve as fixed points maintaining the position and form of outer and inner epidermal cells. Further, just after the ecdysis, ridges being folded into a T-shaped cuticule in cross-section and forming the mesh are always underlain by apical desmosomes and septate junctions, suggesting that ridges correspond to the boundaries between outer epidermal cells, and therefore each reticule quite likely corresponds to a cell in this species. The area on the carapace where the number of reticule increases by division of a reticule after ecdysis from A-2 to A-l is already known through Okada's work. The effects of actual cell division on the formation of new ridges are being investigated in our laboratory by inhibiting cell division using colchicine.
IX
Knowledge of the structure and function of the pores with setae on the cytheracean carapace has depended mostly on the classical works of G. W. Müller (1894) and Rome (1944, 1947) until comparatively recently. Their observations on the pore canals and setae were made solely by the light microscope, and inference of their function was based on the limited knowledge at that time of the physiology of the sensory mechanism. Yet Müller already noted the coexistence of two types of setae that are different in their function on a single valve: one is long and thick and is interpreted as receptor for direct touch by the solid object, and the other is fine and short and may be sensitive to the delicate touch-like movement of the surrounding water or even of sound oscillations (Hartmann, 1966). Rome (1944, 1947) in his classical works pointed out the similarity of ostracod sensilla to those of the insect and showed the presence of scolopidia in Ostracoda.
Simple pores without setae or of unknown nature have also been found on the outer surface of the cytheracean carapace. Van Morkhoven (1962) predicted that they are openings for hypodermal glands. Sylvester-Bradley and Benson (1971) called pore canals of this category tegumental ducts, a term employed to describe those of the other orders of crustacea. Scanning electron microscope observation, made by Hanai,Ikeya, Okada and Nishida, on Keijella bisanensis revealed the presence of probable tegumental ducts, each of which has a peculiarly shaped cap. The duct itself is a simple cylindrical tube 2 &um;m in diameter which opens to a large hollow space under the cap. The cap is surrounded by a circular furrow and consists of approximately 30 long, often bifurcate arms which project from the periphery of the pore opening centripetally towards the center of the hole, making a slightly swollen dome. Adjacent arms are often fused distally. Because the general shape of this cap reminded us of the function of the device to protect against the backward flow of dirt seen in the hole of a Japanesestyle water closet (Benjo in Japanese) in a coach of the Shinkansen Line, we have called this type of pore Ben-type. Ben-type pores open always at precisely the same position in both male and female-on the floor close to the ridge of the mesh.
Certainly, pore canals were ideal subjects for scanning electron microscopy (Sylvester-Bradley and Benson, 1971). Studies on pore with setae have advanced rapidly since the invention of the scanning electron microscope (SEM). In illustrating SEM photographs, Sandberg and Hay (1967) materialized Triebel's (1941, 1950, 1956) prediction that "blind pores" of the sieve plate actually open to the exterior. The range of variations which have been found in the normal pore structure and in the nature of setae is so wide that it seemed difficult to put this diversity in order. However, a few generalizations seem to emerge in the course of SEM illustration and description of these variations.
In the early stage of SEM illustration, Sandberg and Plusquellec (1969) showed, in certain species of the Thaerocytherinae and Campylocytherinae, coexistence of the two types of pores located separately but side by side on one valve. One is a sieve plate with short, thin, furcate seta, or even without an opening for seta. Sieve pores are more or less regularly disposed and open on the level of the carapace surface or its extension. The other is a simple pore with a long, stout seta, or a deeply sunken sieve plate with irregularly perforate sieve pores and radiation of buttress-like supports for the long stout seta from a large central opening. Coexistence of two types of setal pore on one valve is now popularly known from the species of many other subfamilies (i.e. Cytherinae, Hemicytherinae).
The sieve plate without a setal pore may be merely an extreme case of the reduction of a short, thin sensory seta of the sieve plate. Recently, Liebau (1978) summarized the pore canal morphologies according to their historical development: the prototype of sieve pores moved from the muri into the floor of the mesh during the Lower Cretaceous, and the reductions in large central pore canals are traceable back to the Upper Cretaceous.
However, a question still remains on the relation between the simple pore and the deeply sunken sieve plate with a large central opening for the sensory seta. Puri (1974) illustrates, in the simple pore of Reticulocythereis sp., a structure called a circular reinforcement of the basal part of the seta, located between the simple pore and a structure similar to the deeply sunken sieve plate. Our reconnaissance SEM observation of the simple pores of Keijella bisanensis found that the circular reinforcement appears as a bellows-like structure at the base of the seta and may serve to preserve the roundness of the transverse section of the seta when it bends sharply. Below this is an anchoring structure, the deeply sunken sieve plate. It consists actually of a cluster of test-tube shaped tubes of various length which open distally but are blind proximally. In Cythere omotenipponica as well as in many species of cytherine and hemicytherine Ostracoda, the bellows-like structure is found at the base of the seta emerging from the typical sieve plate, suggesting its common occurrence among cytheracean Ostracoda.
When one follows Rome's idea of comparing sensory organs of the ostracod carapace with the sensilla of the insect, e.g. sensillum trichodeum, it may be predictable that the simple pore with a narrow lip is surrounded by the product of the epidermis, while the deeply sunken sieve plates and the setae are products of the outer enveloping (tormogen) and intermediate enveloping (trichogen) cells respectively. Further, the sheath may be generated by the internal and the glia cells.
Since the pore canals and setae which emerge from them are the external cuticle apparatus of the sensilla, it is quite natural to study the external features of pore canals, always taking the nature of the setae into consideration. The recent development of methods of fixation and sublimation for SEM observation have supplied us with details of the setae. Sandberg (1970) illustrated the sensory seta on the sieve pore of Aurila conradi. It is interesting to note that the seta of this species is dendritic with a stout upright stem which terminates like a tube. Thus, it may also be interesting to find out how far the dendrite with ciliary structure is generated distally into the tube-like seta by a bipolar sense cell, and to compare the sensilla with a certain kind of insect contact chemoreceptor because the receptor always makes contact with the sea water. Cythere omotenipponica seems to receive sensory information on delicate changes in the chemical condition of sea water even with its carapace tightly closed. It is well known among Japanese collectors that the trap for collecting Cypridina hilgendorifii is designed utilizing the animals' response to "olfactory" stimuli, though the location of the sensilla is not know yet.
Variations in the canal and seta related to position on the carapace have also been encountered. Our SEM observation of Cythere omotenipponica revealed that at least four types of setae are distinguishable on the carapace of this species (Hanai, Abe, Tabuki and Kamiya, in preparation). The first type is distributed in the marginal area along the free margin. Seta is stout in its lower half, tapering rapidly and easily twisted in its upper half, and terminating somewhat like a tube. The pore of this type of setae has irregular decoration around its opening. The second type is distributed also along the free margin but more or less on the inside area of the zone of the first type of pore canals. Seta of this type is bifurcate near its base, having branches of similar size and extending widely apart parallel to the free margin. Branched setae taper gradually into a sharp point. Pores of this type are simple and small with no decoration. The third and fourth types are distributed widely in the central and dorsal areas and corresponded to the two types of setae or two types of pore canals which coexist on one carapace and have been described elsewhere. The third type is a long staut seta tapering gradually and terminating with a pointed end. The pore has a wide and clearly rimmed lip. Pore canals of the fourth type correspond to the sieve-type pore canals. The fourth type of seta seems to include two forms. One is a stout seta without branches, and the other is a stout seta with one slender branch near its base. The setal pore of the former form seems to occupy a margin of the sieve plate, whereas that of the latter form emerges from the central area of the sieve. The nature of the stout seta of both forms seems similar to the first type seta.
The pores with the third type seta are constant in number, totaling 26 before and after ecdysis from the later stage instars to the adult, keeping a precise pattern of distribution on the carapace, while the other types of setal pore seem to increase in number with every ecdysis. In Keijella bisanensis, the Ben-type pores remain constant in number and location on the carpace, while the pores with setae increase in number through ecdysis. Thus it may be presumable that there exist two categories of pores on the cytheracean carapaces; one remains constant in number and the other increases in number through ecdysis. Since the third type seta of Cythere omotenipponica is a good example of the sensillum trichodeum, Ben-type pores of K. bisanensis, which are quite similar in number and distribution to those of the third type setal pores of C. omotenipponica, are likely to be a neuro-secretory modification of the sensillum.
Diversities of the pore canal structure which have been encountered in relation to position on the carapace are concordant with the continuous nature of the ostracod carapace across the hinge margin. Further, when one adopts the idea underlying the term "marginal infold" instead of using the term "calcified portion of the inner lamella," a homologous relation may be expected even between the structural elements of normal pore and radial pore canals. Work begun in our laboratory with these programs may provide a good starting point for arguments on these problems.
X
In the systematic description, only the synonymies after the publication of checklist (Hanai, et al., 1977) are listed. As stated by Neale (1965, p. 258), accurate synonymy is a sine qua non as the basis for biological study. Species are thus identified prudently, and there is a general agreement among authors in regard to the result of the species identification. However, a slight difference of opinion exists on the generic assignment of certain species, but this problem will not be considered here. This is simply because an element of arbitariness is involved in the generic assignment. Where emotive views can be used, as is exemplified by the case where a trivial name is chosen for a new species, is rare in the branches of natural science. The name of the new species without any explanation on etymology in Yajima's description is derived after the heroines of Genji Monogatari (The Tale of Genji), a classic novel completed by Murasaki Shikibu in ca. 1020 A.D.
The following abbreviations are used in the section of systematic description in all articles:
UMUT: University Museum, University of Tokyo.
IGSU: Institute of Geoscience, Shizuoka University.
CA: Cenozoic Arthropoda. O: Ostracoda.
S: Sample. Sa: Sample number. Ho: Sampling horizon. F: Formation.
M: Member.
Sp: Specimen measured. C: Carapace. LV: Left valve, RV: Right valve.
A-l: Instar of adult minus one stage.
Me: Measurements. L: Length. H: Height. W: Width. N: Number of specimens measured. 
: Arithmetric mean (mm). Sd: Standard deviation (mm). V: Coefficient of variability. OR: Observed range (mm).
The following expressions are used to show the abundance of each species.
Abundant: The species occurs in more than 50% of sample collected from a given member of formation and occupies more than 10% of total individuals.
Common: The species occurs in more than 50% of samples and occupies less than 10% of total individuals.
Rare: The species occurs in less than 50% of samples and occupies less than 10% of total individuals.
All the types and illustrated specimens are deposited in the collection of either the University Museum, University of Tokyo or Institute of Geosciences, Shizuoka University.
Three articles were prepared by P. Frydl, M. Yajima and Y. Okada as a part of their doctoral dissertations based on studies made in our laboratory.
This introductory note is a result of a study supported in part by the grant-in-aid for co-operative research (project no. 434042) and the grant-in-aid for special project research (project no. 56117004) of the Ministry of Education, Science and Culture, the Government of Japan.