Lectures on the Elements of Comparative Anatomy

XQL805H98 12

ON CLASSIFICATION. tion of the clnss JJfammalia is a statement of a ~aw of corrrlati·m', or coex1·s te"' 11co , of "IlJ.IlHtl structures frmn wlneh the 1nost 1m- H ( , portant conclusions are deducible. . . . . For example: if a fragmentary foRsil be (hseovero(l, ronsJst1ng of no more than a ramus of n mandible an<l that part of tho sh:nll with which it articulated, a knowledge of this law may <>nnhl (' the palmontologist to affinn, with great co~1fidC'n , thai tlw animal of which it fonned a part suckled Its young and ha<l non-nucleated rod bloocl-corpimcles ; and to pr diet that should the back part of that skull be eli. covoro(l, it will xhibit two occipital con<lyles and a well-o sifl 'd ln1.si-occipital bon<>. Decluctious of this kind, such as that 1nacle by On vier in tho famous case of the fossil opossun1 of J\Iontinartre, have often been verified, and are well caleulatccl to in1pr . s the vulgar imagination; so that they have tak n rank as the tri1.nnph of the anatomist. But it should carefully be l>Orne in 1nincl, that, like all merely empirical laws, which re t upon n COinpara1ivdy narrow observational basis, tho reasoning fro1n th 1u 1nay at any tin1e break down. If Ouvier, for oxainpl<>, had had to do with a fossil Thylacin~~s instead of a fossil Opo n1n, he \Vonld not havo found the marsupial bones, though tho inflect d angl of the jaw would have been obvious enough. And . o, though, practically, any one who n1et with a clu1.ract >ri tic ally 1nam mulinu jaw would be justified in expecting to find the <·harncte1i tieall)r mammalian occiput associated with it; yet, hr would Le a bold man indeed, who should strictly a ·crt the bcli f which is im­plied in this expectation, viz., that at no p riod of th0 world's history did animals exist which c01nbinecl a man11nalian o ·cipnt with a reptilian jaw, or vice vers£l. Not that it is to be supposed that tho colT lntimt.' of trnc­ture expressed by these empirical laws arc in any ::;ensc acci­dental, or other than links in the general chain of eanHc · and effects. Doubtless there is so111c very good r ason why the characteristic occiput of a 1\lanunal should be found in a sociation wi~h mammro .and non-nucleated blood-corpuseles ; but it i · one thmg to adm1t the causal couuexion of these pheno1nmut with one another, or with some third; and another thing to affirm that we have any knowledge of that causal connexion, or thnt GHEGARINIDA, RHIZOPODA, SPONGIDA, AND INFUSORIA. 5 physiological science, in its present state, furnishes us with any means of reasoning frmn tho one to tho other. Cnvier, the 111oro servile of whoso imitators arc fond of citino· his mistaken doctrines as to the nature of tho methods of palroor~ tology against the conclusions of logic and of co1nn1on souse, lms put this so strongly that I cannot refrain frorn quoting: his wonls.* "But I donut if any ono woul1l have divined, if untaught by ob~ervation, that all nuniuants have the foot cleft, and that th y alone have it. I doubt if any one would have clivine(l that there arc frontal horns only in this class : that thoso among then1 which have sharp canines for tho 1nost part lack horns. "However, since these relations arc constant, they 1nnt:lt have son1e sufficient cause ; but since we nro ignorant of it, we 1nust make good tho defect of tho theory by 1ncuns of observation: it enables ns to CBtablish empirical laws, which Lecome almost as certain as ratioual laws when they rest on sufficiently re­peated observations; so that now, whoso sees merely the print of a cleft foot 1nny conclude that tho animal which left this im­pression ruminated, and this conelusion is as certain as any other in physics or morals. This footprint alone, then, yields to hitn who observes it, the form of the teeth, tho form of the jaws, tho fonu of the vortobrm, the fonu of all tho bonos of the legs, of the thighs, of the shoulders, and of the pelvis of tho nnin1al which has passed by : it is a surer 1nark than all those of ~adig." Morphological classification, then, acquires its highest im­portance as a statmnent of the empirical laws of tho correlation of structures; and its value is in proportion to tho precision and the comprehensiveness with which tho~:->e laws, the definitions of the groups adopted in the classification, are stated. So that, in atte1npting to arrive at clear notions concerning classification, tho first point is ~o ascertain whether any, and if so, what groups of animals can be established, tho Jncmbors of whieh shall be at once united together aud separated from those of all other groups, by woll-clefinocl structural characters. And it will be 1nost convenient to C0111111CIWe tho inquiry with group~ of that order which arc cmn1nonly called Or.JASSES, and which arc onnnlo- * ' O~~OlllOllS fOl::il::iilu~,. e<l. ·lmc, tomu 1•·, p. 1()1.