In recent years lively argument has arisen about such general problems of evolutionary paleontology as punctuated patterns of evolution (Eldredge and Gould, 1972; Gould and Eldredge, 1977), higher-level selection acting upon populations and species (Stanley, 1979; Gould, 1982), and the role of random effects in apparently directional morphological changes (Raup, Gould, Schopf and Simberloff, 1973; Raup, 1977). Punctuationists seem to hold that macroevolution as illustrated by fossil records is not explained by gradualistic morphological change within a lineage or by an easy expansion of the acknowledged mechanisms of microevolution.

In the old days the science of paleontology was developed largely by inductive and empirical methods. Graphy almost always preceded logy. Recent rapid development in various branches of biological and earth sciences as well as revolutionary advances in techniques, however, has given much stimulation to the methodology of evolutionary paleontology. Deductive and theoretical approaches have become more common. Good models and thoughtful extrapolations from accepted opinions in neontology are, of course, quite useful, but descriptive studies of fossil records on the basis of sound concepts are equally important. Without good data on real populations, statistical treatment of morphology would be carried out in vain, and models or hypotheses on the general pattern of evolutionary change would not be able to be tested. In paleontology, as in any other branch of natural history, induction and deduction (or graphy and logy) are like the two wheels of a cart; both are necessary if the field is to advance.

In Japan, Quaternary marine molluscan fossils, in spite of their abundance owing to rapid crustal uplift in many local areas, have rarely been treated as material for the study of evolutionary process. The reason is, I presume, chiefly because any morphological change that may be present within this period was considered to be too insignificant. More than 70 percent of Early Pleistocene molluscs are actually referable to living species (Stanley, Addicott and Chinzei, 1979). This very fact however gives Quaternary fossils a great advantage for evolutionary studies, because we may understand detailed phyletic evolution and, with good luck, the process of speciation on the basis of wellpreserved material and various results of neontological investigations on living populations. Even if genetic experiment is technically difficult, it is still possible to carry out considerable causal evaluation of morphological variation with various kinds of circumstantial evidence. These considerations have prompted me to study the evolutionary pattern and process of the pectinid genus Cryptopecten.

In 1970 I happened to find that very striking discontinuous variation of surface sculpture exists in some fossil and Recent samples of this scallop. I immediately remembered Kurtén's (1955) impressive paper on the dimorphism and evolution in the first upper molar of Pleistocene-Recent bears which I had read several years before. The industrial melanism of Biston betularia, a famous example of natural selection (Kettlewell, 1961; Ford, 1964), and Mayr's (1963, 1969) discussions on individual variation as well as Falconer's (1960) and Kimura's (1960) textbooks on population genetics were also important sources of my knowledge and interest. After two years of field and laboratory works, I prepared manuscripts on the evolutionary change of C. vesiculosus on the basis of 12 fossil and 5 Recent samples (Hayami, 1972, 1973). In these papers the change of phenotypic frequency after the Middle Pleistocene was tentatively interpreted as resulting from the accumulation of a mutant gene by natural selection, and it was emphasized that intensive studies on such discontinuous variations in lineages with living end members, if adequately combined in the future with genetic experiments, would contribute to the development of evolutionary theory.

The genetic background of this discontinuous variation is still obscure, but phenotypic substitution was recognized as one of the possible causes for the punctuational morphological changes often observed in fossil records, as was generalized by Hayami and Ozawa (1975). Subsequently some authors discussed general problems of evolutionary paleontology and biostratigraphy citing our discussion on C. vesiculosus (Reyment, 1975, 1980; Eldredge and Gould, 1977; Gould and Eldredge, 1977; Raup, 1977; Stanley in Fairbridge and Jablonski, 1979).

Cryptopecten mainly consists of lower sublittoral to bathyal species, and is not necessarily ideal material for the study of "paleogenetics", because direct observation of the ecology and embryology at its habitat and Mendelian experiments seem to be considerably difficult. Some intertidal, fresh-water, and land molluscs are more advantageous for ecological and genetic studies. As enumerated before (Hayami, 1972), materials that satisfy the following conditions are preferable for studies of this kind: (1) abundance both in living and fossil states, (2) readiness of culture, (3) readiness of observations on general biology (especially, mode of life, growth and reproduction, population structure and functional significance of morphology), (4) short generation (necessary condition for Mendelian experiments), (5) relatively simple evolutionary lineage(s), (6) presence of genetic polymorphism preferably recognized in the morphology of hard tissue, (7) nearly panmictic populations (without a wide range of geographic variation in phenotypic frequency) and (8) inferable absolute age for each fossil sample. My own observations of extant gastropods and bivalves, however, has shown that materials satisfying many of these conditions are rather rare.

Land snails are certainly advantageous with respect to readiness of culture and ecological observation. As exemplified by Gould's (1966) surprising work on Poecilozonites in the Quaternary of Bermuda, studies assuming "an evolutionary microcosm" may be possible on this material in some oceanic islands (e. g., some species of Mandarina in the Bonin Islands). Generally speaking, however, populations of land snails are highly localized, and the phenotypic frequency often changes drastically within a small area (Lamotte, 1951; Sheppard, 1952; Komai and Emura, 1955). Diver's (1929) pioneer study on genetic variation in fossil populations of Cepaea notwithstanding, evolutionary change is hard to detect owing to the wide range of geographic variation. Moreover, in Japan (except for some subtropical islands) good fossil samples of land snails are rather rare, probably because of soil and climate conditions unfavorable for preservation.

Gastropods and bivalves in embayments and intertidal waters may be more tolerant of conditions in aquaria than those in oceanic waters. Some are considerably well represented in fossils. As the result of Yoda's and my preliminary observations (unpublished data) on three ubiquitous species of Batillaria (potamidid gastropods) in Japan, however, the relative frequency of white-banded individuals was found to vary greatly from one embayment to another, suggesting relatively weak gene flow between local populations in spite of their free-swimming larval stage. According to Colton (1922), Moore (1936) and Berry and Crothers (1974), there is also a significantly wide geographic variation in Nucella lapillus, a muricid gastropod, common in the intertidal zone around Great Britain. Such non-panmictic species may be useful for examining the influence of environmental factors and differential selection pressure among different areas, but they are generally unsuitable for the study of evolutionary process because the net amount of chronological change may be difficult to detect.

Color polymorphism is a widespread phenomenon in various groups of gastropods and bivalves. Though the variability of ground color is generally difficult to recognize in fossils, color pattern (band, stripe, spot, etc.) is often observable not only in Quaternary but also in much older molluscs, particularly in gastropods with compact shells. However, the recognition of color polymorphism in fossils, even if more clearly observable under ultraviolet light (e. g., Wilson, 1975), depends largely upon a favorable state of preservation. Thus for practical reasons, polymorphism in the sculpture of shell, if it is controlled by genetic cause, is more informative.

Phenotypic discrimination in Recent and fossil samples of C. vesiculosus is so easy that one can quickly recognize the phenotypic frequency of a large population sample almost regardless of the state of preservation, differential sorting during post-mortem transportation, incompleteness and deformation of specimens, and age heterogeneity. Such cleancut discontinuous variation and abundant occurrence both in fossils and present seas are rarely found in other molluscan species. Recent great progress in the chronology and correlation of the Neogene and Quaternary sediments in Japan has made possible the estimation of the absolute ages of many fossil beds yielding Cryptopecten samples. These factors lead me to return to a more exhaustive study on the systematics and evolution of Cryptopecten.

The Pectinidae are a characteristic bivalve family which has flourished mainly on sublittoral sandy substrates since the Triassic. Their tests are generally tough and well preserved in sediments owing to the stable foliated shell structure. Strongly abraded valves are often found in fossil beds, but they are commonly free from fragmentation except for the auricular parts. Large living and fossil samples are frequently obtainable in some species. Various quantitative characters are easily counted or measured owing to the strong and regular surface ornamentation and easily defined orientation. The ecology, physiology, anatomy and reproduction of several living species have been intensively studied by a number of malacologists (e. g., Dakin, 1909) and workers concerned with commercial fisheries and cultivation (e. g., Yoshida, 1964). The functional significance of shell morphology in some pectinids has been successfully interpreted by several malacologists and paleobiologists (Yonge, 1936; Stanley, 1970; Gould, 1971; Thayer, 1972; Waller, 1972a, b). Though the genetic variation and geographic variation of Indo-Pacific pectinid species have hitherto been little studied at the population level, the abovementioned favorable conditions and accumulation of knowledge should be put to full use in evolutionary studies of this family.