Such a way of thinking may be considered as a simple and informal system of classification, and we usually do not even pay attention to this process.

    Formal classification, which is used in zoology, is a different matter. It conforms to strict rules, providing uniformity through an internationally understood nomenclature of all living organisms thereby simplifying cross-referencing and information retrieval. Formal classification, according to E. Mayr, may be restricted to taxonomy which "is the theory and practice of classifying organisms," or it may be practiced as systematics, which "is the science of the diversity of organisms." The latter definition suggests strong links to evolution, ecology, genetics and behavior which taxonomy need not have. These two branches of science––"taxonomy" and "systematics"––have a common basis: they must be scientific. This means that their prerequisite methods may be different, but their essence is measurement of phenomena and repeated observations. Such methods may differ for the various families of mollusks because of shell structure, but they must be repeatable and, preferably, quantitative. Scientific hypotheses and theories are usually developed after comprehensive material is collected based on the scientific method chosen.

    It is interesting and perhaps instructive to look at the history and fate of one such method.

    This method was apparently based on the following assumptions:
   1.) In each species, a correlation exists between the length of a Cowry shell and the number of its columellar and labial teeth. This relationship was established scientifically in the earlier, published works of F. A. Schilder.
   2.) The length of any individual shell of a given population follows the so-called law of normal statistical distribution, with smooth grading between the extremes of lengths (with the majority of individuals in the center, as can be seen in human height variation, for example). This assertion was later substantiated when the Schilders and other workers measured about 100,000 shells of different species.
   3.) Using this method, any given Cowry population (the race) may be characterized by four parameters expressed as a formula which usually has eight digits (ww.xx.yy.zz). The first two express the average length of shells in the population (in millimeters). The next two digits relate the average relative width of the shells as a percentage of length, and the last four digits, expressed as two numbers, indicate the number of normalized labial and columellar teeth, respectively (the values calculated to express the teeth on hypothetical shells of 25mm length).

    This is probably one of the first quantitative and scientific methods used in conchology. It may be especially useful for comparing similar populations of the same Cowry species from different localities, for comparing very similar Cowry species which inhabit the same region, and for predicting the range of shell length, width and dentition of a population when its statistical parameters are known. This method was used by its authors for zoögeographical studies of world-wide Cowry distribution and for comparing numerous Cowry populations. The Schilders felt that most living species of Cypraeidae with an area of distribution exceeding certain limits may be divided into two, or more, geographical races and, furthermore, each race usually has a center of population where it is relatively more frequent than at the periphery.

    The method's fate was rather complicated. From the beginning most collectors and professionals accepted it together with the zoögeographical concepts of its authors, and many subspecies were described and studied.

    However, in the sixties two new and different approaches were developed in the field of Cowry study. According to the first approach, it was recommended that identification of Cowry species be based on studying of the animal's soft parts (mantle and radula in particular), giving much less consideration to shell characteristics, and assuming that several different Cowry species can possess a very similar, or practically identical, shell.

    According to the second approach, all Cowry species were relegated to one Genus and their infraspecific gradation was abolished, making futile any study of conchological differences of Cowry populations.

    Many collectors immediately and without doubts accepted the new concepts and rejected the Schilders' method. It was almost forgotten in the last decades.

    The Schilders made critical revisions of their views several times during their prolonged work. In probably one of their last works (Schilder & Schilder, 1971), they revised the taxanomic status of numerous populations. Populations which were previously considered subspecies (sometimes only intuitively) were now treated as synonyms, apparently because there was still not enough convincing evidence to retain them as subspecies. But the Schilders' general approach remained the same: Cowry populations must be treated as subspecies in any case where there is enough conchological evidence. This definitive revision was published in a magazine of limited distribution and was almost unknown to most Cowry collectors (and still probably is).

    A list of some 2,000 works is "Cypraeacean and triviacean literature 1971–1993" (Groves, 1994). In very few of these works was the Schilders' method used or even mentioned, though there is not a single work in which its merits or shortcomings are analyzed.

    The Schilders' works often include abbreviations in order to make articles more compact (though seldom easier to read), explanations and discussions in order to assist in understanding the material and associated difficulties, including taxonomic ones (Schilder, 1966), and complicated relations (Schilder, 1963). When reading these works, the impression often arises that the Schilders have foreseen many possible problems (in the future) but the one possibility which completely escaped their attention was that their works might be ignored.

    Even L. Raybaudi Massilia, a supporter of the Schilders' concept, hardly used their quantitative method in his numerous works. In one of his articles, there is a note about more than a thousand Zoila venusta Sowerby, 1846 which passed through his hands. With so much conchological material, it could have been a great statistical asset to have had the data from these shells, resolving many "hot" taxonomic problems and preventing some futile discussion. But the quantitative data of that "'lot" were never published (maybe this will happen in the future?).

    The Schilders' method is simple, though some of its limitations and potential pitfalls must be remembered––foremost, that it is empiric and approximate. Only Cowry populations as a whole––and not individual shells––may be characterized by the formula and their taxonomic status may be considered. The most common objection from opponents of the method comes when small shells are sometimes found in a population normally characterized by large shells (or just the other way round, when large individuals are encountered in a population with typically small shells).

    For example, even today many collectors do not accept Mauiritia mappa geographica Schilder & Schilder, 1933 as a subspecies (typically characterized by small shells) because large shells are sometimes found in that same Thailand - Malaysia area. Apparently these collectors are forgetting that such shells are not only possible but even expected, according to population statistics (the law of normal distribution). If one has doubts about the validity of the subspecific status of the mentioned mappa population (related to the shell size), a good way is to study a large lot of shells from the region, then draw conclusions.

    This case leads us to another possible mistake regarding use of the Schilders' method––that an insufficient quantity of shells taken for measurement may lead to wrong conclusions. As many shells as possible should be measured (several dozens, at least) in order to obtain more or less reliable statistical results. All lots of specimens collected for measurement must be free of 'bias,' i. e., only randomly selected shells may be considered. The actual dispersion of values in the studied lots (compared to the law of normal distribution) must be checked before making any conclusions.

    It should be also remembered that deviations from statistical expectations are possible, due to the influence of various ecological conditions. Unusually small (or large), broad (or narrow), or differently colored shells may be sporadically encountered in almost any Cowry population. They are even given form names: e.g., oblonga, dilatata, saturata, deformis, ..., etc.

    F. A. Schilder considered Cowry species and subspecies as evolving populations which have been given taxonomic status. Discussing (Schilder, 1966) in detail the complicated problem of the infraspecific taxonomy, he considered that the official taxon 'subspecies' may contain several stages of evolution, and that it seems advisable to resolve it into five levels, which must be named. The proposed definitions of infraspecific taxa in the Schilders' work are listed below, in order, from the most genetically developed to the least developed:
    prospecies - this is a population on the border between being a 'subspecies' (as given above) and becoming a true species; this taxon should be used in all cases of uncertainty when deciding whether a taxon has reached the degree of a true species or not:
    subspecies (in the narrow sense) - this taxon represents distinguishable populations which are widely distributed geographically, and replace each other in those habitats, so that the border area where two subspecies merge is a relatively narrow zone;
    clines - these groups are characterized by a far more gradual mixing of adjacent taxa into each other, so that only specimens coming from opposite extremes of the inhabited areas show their notable characteristics, while the large area between these extreme locations contains a population of intermediate forms or a mixed population of forms from those extreme areas;
    infraspecies - differs from a true 'subspecies' in that the differences in the morphological characteristics being studied are still underdeveloped so that they can be hardly be distinguished––at most statistically, by means of studying the populations with regard to a number of small differences:
    morphes - these are sympatric groups of very closely allied specimens; they differ by some often unessential characteristic and, evidently, are highly stable mutants.

    In F.A. Schilder's opinion, there are two paths of evolution of Cowry species: one, the allopatric route, consists of isolated populations and selection of mutations by chances of the environment. A second, rarer path, involves the sympatric development of species living together in a restricted area.

    The allopatric way of evolution tends to develop a superspecies comprising all species- and lower-level taxa. These replace each other geographically, ecologically and stratigraphically, and are so closely allied that they evidently descended from a common ancestor.

    These views and the zoögeographical opinions mentioned above were mostly ignored, and the quantitative method was not used for decades.

    Nowadays, in almost every work on the subject of Cowries, one can find the Schilders' names. They sound, at times, like a mantra, mystical words of spiritual power. Unfortunately, sounds alone cannot be used as a scientific argument, a basis for taxonomic status, quantitative data or other characteristics of the many Cowry populations which are, unfortunately, still far from being understood––or even studied.