1889 dall on the hinge of pelecypods and its development

8/6/2019 1889 Dall on the Hinge of Pelecypods and Its Development

http://slidepdf.com/reader/full/1889-dall-on-the-hinge-of-pelecypods-and-its-development 1/18

Dall-lIinge of Pelecypods and its Development. 445

. each. Omitting mention of the Permian, or uppermost portion of the Carhoniferous system, which seems to be absenthere, the hiatus between the Lower Cretaceous, and the Car

boniferous strata of the Chinate Moulltains amOH nts to at leastthe whole of the .JUl'a and Trias.

This hiatus is no greater than is exhibited in others of the

mountain uplifts in the reg-ion under discussion, and not sogreat as it is in some eases. Indeed, so far as I am now aware,the hiatus between the Lower Cretaceous and the next underlying rocks is nowhere in all that region less than it is at thelocality in the Uhinate Mountains just mentioned. That is,both the Jnra and Trias are helieved tu be absent there. That

neither the .Jura, Trias, 01' Lower Cretaceous occur betweenthe Upper Cretaceous and Oarboniferous in central New Mexico, accords with observations tllat I made there during- the

fast season. So far as both the .Jura and Trias are concerned,have not yet been able to obtain any satisfactor,Y proof that

either of these g-eological divisions are represented by any

North American stl'ata south of the 34th parallel of latitudc,*

ART.LV. -On

the flillge ttl Peleeypods ({udits IJcc,zop-11U:lIt, witll ((11 attempt tm/)l(l'd a better 8ubdivision Cff tilegroup,. by W ~ I . II. DALL. Palcoutologist. U, S. Geol. Sllrvey,-and Curator Dept. of Mollusks, U. S. Xational Museum.

THE attempt to divide the elai's Pelecypoda or Lamdlibran

okiat(l into orders has so fa I ' been unsuccessful, 01' at least thesubdivisions adopted have from time to time been found ullsatisfactory, 011 account of the discovery of forms which combinein their org-anization charactel's which had pl'Cviously been

reg-arded as diagnostic of important subdivisions, such as orders.This has resulted from the selection of characters as diagnos

tic which arc reaIIv not fundamental in tIle evolutionary history of the minor groups. As we gtoaduall'y becomc acquaintedwith the mutability of the adductol' m u ~ c l e s , the gills, thearrangemcnts for retraeting the siphons and other factol's inthe mechanics of these organisms. the classification based upontheir mutations has g-radually ceased to satisfy students thuugh

one phase or another of it lllay still retain a place in ordinary

text books.• Reference is \Jot hen> '"IHle to tl'e .. Dino.aur Sand"" which lie at the baSIl

of the Comunc'he Cretaeeous upon [,oth sides of the Paleozoic. area ill Texas.

These beris I have pro\'isionallr included in the Lower Cretllceous, but it is prob-

able that they represent Ihe i 'olomac formation of the atlalltie coast region. andit is regarded as possible that they represent the uppermost Jurassic of ~ : u r o p e .

DALL, W. H. 1889. On the hinge of pelecypods and its

development, with an attempt toward a better subdivision of the

group. American Journal of Science, Ser. 3, 38(228):445-462.



...

446 D a l ~ I l i l i g e of Pelecypods and itB Development.

To cite a few examples it will be remembered that tile most

per!>istent of the earl,Y systems for classifyinp; these animals wasbased on the number of adductor muscles or the 8ears upon the

shell by which they might be traced. At first the p;roups ofMonomyarians or forms with one adductor, like the oyster, andDimyarians with two adductors, like the ordinary edible clam,seemed sufficiently well distinp;uished, Later when transitional forms like the mussel and its allies were carefully studied

a new group, l/etero1n!Jl1rlft was erected for those which would

not fit into either of the others.But when it is considel'ed that there are forms like J)imya,

in which with a monomyarian organization two distinct adduct-

orR are found, one at each end of the shell; that in Uhlamydo-cOl/cha we have a specially modified animal with no adductors

at all; that in J.l'llllleria we have the young (not larval) animal

typically Dimyal'ian, and becoming in its adult stage as typi-

cally monomyarian in its muscular apparatus as an oyster; then

it is sufficiently evident that bettm' and more fundamental

diagnostic characters should bc found 01' the so-called orders

given up.

Again, an attempt has been made to use the characters of

one of the most mutable parts of the whole organism, namely

the gill, as a baiis fOl' primary divisions of the group. I haveshown elsewhere,* I venture to think eonclusivel)', that this

selection is ill-ad \'ised and calmot successfully solve the problem

The simplicity or sinuation of the pallial line has been reogarded as a dIamcter of high importance and has been used asdiagnostic of divisions of primary importance. I have recently

shown that, in certain gI'OUJlS, long siphons may exist with asimple pallial line, as in C l l . ~ p i d a r i a " that in species without

long siphons. members of the same family (Pol'omyidw), and

perhaps of the same genus, lllay show a simple or a stronp;lysinuated pallial line according to the modifications of certain

muscular elements whieh cert.'linlJ cannot be claimed to have

any high systematic importance.

The question is further complicated by the fact that certain

chamt'ters, which in general are indicative of very early evolutionary divergencies. may be simulated or assumed as very

nlodel'n speeial modifications brought about in animals of

dh'erse groups by natnral selection under the influence of spe

cial circumstances. Species th us lately modified will verynaturally he classed with those which hear the same or similar

c h a l ' a c t ~ r s as the early result of \'el'Y aneient ancestral diver-

~ e n e i e s , and. as a eonsequence. other characters not harmoniz-

mg', the systems are thrown into confusion. These are difficlll-

• Bull. Mus. Compo Zoo1., xviii, pp. 43:1-438, JUDe, 1889.



Dall-Hinge of Pelecypods alld its Development. 447

ties a m o n ~ which the sum total of the o r ~ a n i c characters mustbe our guide in attempting to decide. Only too often we mayfinti, as knowledge increases, that our first judgment was more

or less in error.In reflecting upon the origin of the complicated mechanical

arrangements in bivalves which we call-the hinge, I have cometo the conclusion that here, as in the cases of the mammalianfoot and tooth, elaborated so clearly by Cope and Ryder, wehave the result of influences of a mechanical nature operatingupon an organ or apparatus in the process of development.

The hinge of a bivalve, reduced to its trltimate terms, con-sists of two more or less rigid edges of shell united by a flexi·ble membrane or ligament.

The ligament may be wholly external or may be supplemented by an intel'llal addendum (called the cartilage), whichexerts a stress in the same direction within certain limits. The

movements of the hinge are dependent upon the elasticity ofthe ligament and cartilage and upon force exerted by one or

more adductor muscles uniting to the valvesThe rigid e d ~ e s or cardinal margins of the valves may be

simple or modified by the presence of interlocking processesknuwn as teeth, whose purpose is to regulate the direction of

the valves in opening and closing.There are three fundamental types of hinge: 1, the simpleedentulous margin closing by simple apposition of the edges ofthe two valves; 2, the hinge in which the teeth are developedin a direction transverse to the cardinal margin; 3, the hingein which the direction of the teeth is parallel to tlJe margin.The mechanical features of the second and third types may bemore or less combined in a single hinge, but the affinities ofthe particular form in which this may occur are llsually n It

difficult to determin.e on a general sUl'vey of all its organic

characters, .I am disposed to think that the time relations of the differ·

ent types are those of the order in which I have cited them;

the most perfed hinge, morphologically speaking, would beone which should combine the most eftectivc features of the

second and third types.The archetypal form of bivalve may be imagined as "mall,

with nearly equilateral, symmetrical, sub-circular valves withedentulolls cardinal margin and a short external ligament

nearly central between the umbones. This is the character ofmany larval bivalves at the present day, though it is probablethat many of the forms now edentulous in the adult state, havepassed through an evolutionary stage in which they had a moreor less denticulate hinge-margin, while their present conditionis one in which the hinge has diminished in complexity or, inother words, undergone degeneration.



448 J)all-Ilinge of P e l e c y p o d ~ and itB j)evelopment.

Very few of the earliest known bivalves appear to have

hinge teeth. yet this may be on account of our imperfect

knowledge of many of them since they are often represented

by fossils in which no evidence of the hinge structure is discernible It is highly probable that the evolution of hingeteeth closely followed the differentiation of the Pelecypod

class and that the ;;egregation of the muscular apparatus for

closing the valves into two bunches or adductors was accom·plished very early in its history.

'rhe first bivalves are all small, as far as known, when COIll-

pared with a majority of their descendants. It is highly

probable that they possessed a developed foot and that their

gills were either larnelliform on either side of an arterial

stem, as in jYucula, 801enomya and many Gastropods, or filiform, as in j)£mya and certain Pectens. The siphons were

probably little developed and the lobes of the mantle rather

widely separated or perhaps entirely free.As long as the shell remained small and subglobular, the

l igament short and wholly external, the imperfect character of

the hinge was of less importance. With the essential difference between the anterior and the posterior halves of the

animal, and especially wit.h any material increase in the magnitude of the

adult, moreor

less discrepancy would developitself between the two ends of the shell, the subglobular formwould disappear, and certain other consequences would follow.Eithel' the ligament must increase with the size of the shelland become longer or its powel' would become inadequate forthe proper performance of its functions.

Here I will turn aside for a moment from the direct line of

argument to describe the mechanical relations of ligament and

shell, a proper understanding of which is vel'y neceSsary to thecomprehension of the whole question. •

. With a wholly external ligament the operation of the valvesis that of two appendages to the free ends of a O-shapedspring. The action of the muscles in pulling the valvestogether includes the bringing nearer to each other of the two

extremities of the ligament which the latter hy its elasticityresists, consequently the operation of the ligament is in the

direction of opening the valves to a certain distance. Beyondthis distance the separation of the valves tends to compress the

ligament, which again resistH, and therefore beyond the normaldistance of separation the action of the ligament tends to pre-

vent the valves from opening. This very simple matter may

be observed by anyone who will examine an ordinary clamwith the ligament in fresh condition and whose adductor

muscles have been severed.When the ligament, in harmony with the elongation of the

cardinal margin, becomes elongated it must be either straight



Doll-Hinge of Pelecypods and its Development. 449

or angulated. For obvious reasons a ligament forming a curve

or the arc of a circle is mechanically impoBBible .This any

one may prove to his own satisfaction by p u t t i I J ~ two light

wooden saucer:; edge to edge, convexity outward, and attach-ing a leather or paper l i ~ a m e n t by cement. A curved l i ~ a ment when the valves open will tear 01 ' break at once, either

itself or the e d ~ e to which it is fixed. In other words, the

axis of motion of the h i n ~ e must be in a straight line. I f anypart of the l i ~ a m e n t diverges from the axial line, it must ceaseto take part in the axial motion and must be capable of stretch-ing to an extent which will neutralize its angulation, or it willbe broken or torn away. But if the thickness of the l i ~ a m e n tincreases ventrally, as may be the case when it is situated

between the valves rather than as an arch above them, acertain portion may extend to and beyond the axial plane in adownward direction. The portion thus projecting will then

partake of the axial motion in an opposite sense to that portionwhich remains above the axial line. I t will be compreBBedwhen t he l at te r is stretched by the c l o s i n ~ of the valves andwill expand as the opening of the valves allows the externalportion to contract. This change may be brought about by adownward angulation of one end of the ligament (as in Soleno-

mya) or as a simple downward growth; which may be central(as in Neiloilella or Galeomma). The former may be the

result of an a n ~ u l a t i o n of the h i n ~ e - r n a r g i n consequent onelongation or ventml extension. Its result is to separate aterminal segment of the original ligament, which segment maybe totally detached or remain physically connected; while ineither case its mechanical function has u n d e r ~ o n e a reversal ofdirection.

The second mode likewise removes a segment but in avertical direction. This segment may be physically continu-

ous tllroughout its upper portion with the lower portion of thesuperjacent l i ~ a m e n t , it may be wholly detached, or it may beattached by one extremity while t he o th er is sepamted; in the

last case its direction will be oblique or at an acute angle withthat of the original l i ~ a m e n t . This detached s e ~ m e n t what-ever its position has always similar mechanical relations to the

movement of the hinge and is called the cartilage. The sepa-ration of the cartilage from the ligament is generally eithercentral or toward the shortest end of the hinge, which is usu-

ally the anterior, owing to the fact that when the size of alamellibranch increases, the siphons, the ovaries, the visceralmaBB or the gills are the organs where proportionally increasedgrowth is most likely to occur, and these are usually central or

p o ~ t e r i o r to the umbones. In Solen01nya, which is exceptionalin having a posterior cartilage, the posterior portion is shortest.



DaU-Hinqe of Pelecypods alld its Development. 451

In the thin-shelled Ullspidal'iidaJ a special buttress i oftendeveloped to support the shell at this weak point. In the

Isocal'diidaJ an independent cartilage was possibly never devel

oped, but the infolding of the anterior part of the ligamentwent far enough to leave permanent traces on the shell. Thatit did not result in a cartilage if this was the case may possiblybe due to the fact that, owing to the great size and spiral character of the umbones, the antcrior part of the ligament wasturned up instead of downward, and therefore did not tend toshift toward the interior.

I f it is not clear how the thickening or vertical extension of

the ligament below the cardinal axis should cause its separationinto two parts, I need only recall the familial' experience of

everyone in breaking off a wire or piece of tin by bending itbackward and forward on the line of the desir'ed fracture.The mechanical principles and results in the two cases are precisely similar.

When finally developed in the same individual the ligamentand cartilage work in identically the same manner hut in different directions The resistance of the ligament to compression prevents any straining of the adductors by a too wideopening of the valves, The same resistance in the cartilage

prevents the ventral margins from crushing each other by sud·den and violent contractions of the adductors when the animalis alarmed, and closes its valves.

The nYIll phre, or processes to which the ligament is attached, and the fossette, or socket of the cartilage, have beenstrengthened and regulated by the development of various buttresses and other devices, varying in different groups. The

cartilage in turn has its rigidity and strength increased inmany species by the special development of shell substallceknown as the ossiculum.

'I'o return to the development of the cardinal margin. Theasymmetry of the shell and ligamcnt relative to a vcrtiealtransverse plane passing through the Um h U l l e ~ , would he prumoted not only by the natnml discrepancies between the anterior and posterior halves of the body, but by the mechanicaleffect of the projecting umbones. "'here a shell opens laterallY, in the strict sense of the word, unless the beaks are vervi n ~ o n s p i c u o l l s , or are separated by a wide projection of the ca;dinal border (as ill Area 1IOlE), they will strike against and wearout one another. This abnormal or accidental result is veryconstantl.y observable in many AnatinidlE, such as 0111' ownTkraci<t Co-nradi. But it lllust be a suurce of weaklless anddanger to the animal. I f the ligament is shifted posteriorlythe valves must open more obliquely, with a result that thisdangerous friction will be avoided in 11Iost cases.



452 Doll-Hinge of Pelecypods and its Development.

In a protective armor like the valves of bivalves, otherthings being equal, it will be obviously beneficial if not absolutely essential that it should offer as few weak joints or open

spaces as pOf:sible. Burrowing animals, whu themselves serveas a supplementary defence of their burrow, may be able toperpetuate gaping shells and exposed siphons without serionsdanger from their enemies. Those animals which burrow butslightly or live in material which enemies may also easily penetrate in their forays, will unquestionably benefit greatly by anaccurate and exact closure of the valves. The intrusion ofsolid bodies can be to some extent guarded against by theaction of the cilia or processes of the mantle margin, but suchintl'llsion would he greatly facilitated by any organization of

the hinge which would permit an independent rocking motionof the valves with respect to each other. The sudden closingwhich danger incites leave", no time for clearing out obstructions and the gap is especially liahle to the incursion of gravel,etc., in species which live with the plane of junction of thevalves in a vertical direction. In certain brachiopods such asGl{)ttulia and fhscina such a semi-rotan' motion of the valvesexists, but is less dangemus to them s i n c ~ the plane of junctionwith them appears to be generally hOl·izontal.

To avoid these dangers and to guide the motion of the valvesin closing, and to prevent their sliding upon one another afterclosing, Nature, through natural selection and physicalstresses, has developed tllese cardinal processes which areknown as b'leth.

Attention has already been called to the fact that there canbe but three fundamental types of hinge, which may be calledthe anodont, prionodont and orthodont, the latter term beingused to indicate the forms in which the cardinal m a r ~ n hasbecome 10ngitlldinall,Y plicate. Actually the pure orthodont

type hardly exists; in nearly all forms traces of the prionodontcharacters are mingled with it. For those forms ill which thearchaic anodontislll still persists as the characteristic of chiefimportance, though freejuently modified by special mechanicalcontrivances which to a certain exteut lIlask the type, I haveproposed the term jL;o7llalor!e8IJw('ea. The fossette, cuiller'onor spoon-shaped process for the c a r t i l a ~ e is a separate development serving a special pUl'pose; t h o u ~ h influencing the teeth,if any exist, in its vicinity, it must 1Iot be confounded withthem. The weakne;;;s of the anodont type has left an openingfor the specialization and perfection of this process which, to aconsiderahle extent in this group, assnmes tlw functions whichin groups without a cartilage are the special office of theteeth.For those fOl'ms in which transverse plication of the hinge isthe chief characteristic, though rarely wholly exclusive of



Dafl-Hinge of PelecypodB and its Development. 453

orthodont influence, I have used the term Prionodeamacea.In some cases what may seem to be the chief features of the

hinge as regards size and strength are orthodont, yet these I

believe to be comparatively modern specializations illustratingthe general tendency of evolutionary processes toward a teleodont hinge. In cases of doubt the sum of the characters willenable us to decide on the proper place for a given genus. I t

must not be supposed that, because the name8 suggested by asingle set of characters are used to denominate the proposedorders, that therefore that set of characters is to be our sole criterion. Such too hasty assumptions are a relic of the dayswhen the immutability of species was an orthodox dogma in

biology, and doom to failure any system founded upon them.For those forms in which the various types of hinge have

become harmoniously combined, though ill varying proportioncontributing to the final mechanism, I have Eelected the designation of T e l e o d e . ~ m a c e a . These may be regarded as the

highest and evolutionally the most perfect in type of hinge,though this perfection Ehows itself in a variety of forms.Prionodont traces remain with most of them but are nevercharacteristic of the type.

The three groups 1 propose to call Orders. It is difficult to

say whether they can be compared in systematic value withorders in other clatises. All that can be said is that these threedivisions are discernible in the very compact and homogeneoueclass which includes them, and it contains no other groups ofequal value or significance.- Each Order as it now exists contains archaic and modern specialized types. Each indicates a tendency toward an ideal of

fitness to the environment, which results in a certain parallelismof minor characters common to minor groups in each of the

three orders. In each (we are coming to regard it as inevitable),

certain members show affiliations \ \ ~ t h members of the otherorders. In each there are certain groups which represent arelatively modern specialization carried 8 0 far as to be quitepeculiar.

Pearliness or a trnly nacreous character of shell substance isa !:'ource of weakness. This kind of shell is more fully permeated with animal matter, is more liable to decay and exfoliation and is more readily drilled by enemies than the arragonitictype of shell s u b 8 t a f l ( ~ e which conchologists call porcellanous.

The tendency of evolution is to promote the porcellanoustype. The older groups (Priollode8macea and Anom alodes·macea) contain all the pearly Pelecypods, among the Teleodea-

macea there is not a singh.. one. Furthermore, in the twoformer orders the most specialized and, developmentally, the

most modem forms are preferably porcellanous; those whichAll. JOOR. SOL-THIRD SKRIF;8, VOL. XXXVIII, No. 2 ~ 8 . - D E O . , 1889.

29



454 lJall-Hinge of Peleaypods and its Development.

we may reasonably regard as of more ancient type tend topearliness. For example in the Allomalodes-maeea the moststriking instances of srecialization are the Pholads, TubicoUBand certain fffyaeea, al are earthy or at least not pearly. The

Anatinacea which paleontologically are very ancient, are

largely pearly. The Prionodesmaeea have few porcellanousgroups, but those which show this character, such as Ostrea and

Pecten, generally stand at the nearer end of the long line of

progressive modification. There are exceptions to this. such asTi1ldaria in the Nuculacea which is obtrusively porcellanous,while Leda and I . . ~ ' o l e n o m y a , which retain so many archaic fea-tures in their soft parts, have almost lost the pearly layer while

stlll falling short of the porcellanous character conspicuous in

most of the Teleodesmacea. The Arcas conspicuously earthyin their shells are mudern in their total characters comparedwith the pearly Nuculas. Turning to Gastropoda for a mo·ment we find that PleuroWrnai'ia, one of the very earliest types

of that class which can be recognized in the now existing fauna

is extremely pearly. On the whole the relation between the

two types of shell substance if not constant enough to be

called a rule is sufficiently so to be extremely suggestive.I have already suggested the mechanism of the infolding

which resulted in the cartilage and its supporting soeket. I t

is a very difficult ta8k to account for the initiation of all thetypes of teeth. A few suggestions may be ventured upon.

The radiating or transverse corrugations which we see in

ribbed shells are not merely ornamental. They serve to add

strength while they do not increase the weight as would a cor-responding thickening of the shell. A familiar example of

the same principle is affOl'ded by the corrugated sheet metal 80

frequently used by builders. The ends of these ribs impinge

on the margin of the shell and crenulate it when the shell is

thin. Crenella is a notable example. Many .Mytilacea ex-hibit a similar 8tructure. These crenulations of the hinge line

and margin are not to be distinguished from nascent teeth and

have frequently been described as such by naturalists. Nucu{ocardia of Orbigny is a well known instance. The crenula-tions of the margin are useful in securing a close tit between

the closed valves, whether at the cardinal or the basal margin.But they would he more useful at the cardinal margin becausethere they would prevent sliding of the valves upon one an-other before they were completely closed, as do the long teeth

of the .Nuculacea. Hence. it is probable that they would beperpetuated and specialized there even if the ribbing disap-peared from the exterior of the val ves. Greater stre88 arising

froll! friction and pl'essure resisted, would tend towards the

thickening, widening and even buttressing of the cardinal



456 Dall-Hinge i f PeleC'ypod8 and itB lJevelopment.

to the highly developed and specialized cardinal teeth inMadra. Thus it will be obsm'ved the teeth called" cardinals"in Pele-::ypods are by no means all necessarily homologous;and it is even conceivable that cardinals of both types mightcome to be united in the hinge of a single species.

The development of lateral'teeth from transverse teeth is avery easy process of which a filII exhibit might be made by

arranging in a continuous series the \'alves of selected Areacea

and Nuculacea. It is probable however, that not all Orthodont

dentition originated in this way. The thickening of the cardinal margin rendered necessary by the stresses in volved in the

mechanical operation of cardinal teetll or strong external liga

ments, would render parallel plication of the thickened areaalong the margin not only easy but almost inevitable in somecases. The infolding of the edge of the mantle necessarilyaccompanying the propuction or a strong specialized socketfor an internal cartilage would lead incidentally to occasionaldeposition of shelly matter in ridges parallel with the longer

edges of such sock.·ts. The greater efficiency in guiding the

valves to effective clmmre, in proportion to the increased distance from the umbonal region, of such interlocking plicationswould teud through natural selection to the perpetuation (}f

favorable variations and to their gradual removal farther and

farther from the beaks until the most useful distance wasattained.

When we consider the remarkable uniformity in hinge characters attained by the species with more perfected forms ofhinge, through long series of individuals, It Reems almost incredible that these results should he brought about by the

action of a thin soft film of secretive tissue, which, unaided,could not hold itself erect. It is only when we remember that

the result, in the main, is brougllt "about through the actionand reaction of certain dpfinite mechanical stressel", propagatedthrough the hard valvular skeleton and constantly imposedupon the softer tissues, that any adefJuate reason for the marvellous uniformity presents itself. There are certain groups

such as the lsocarrbirlm in which the hinge seems still to be in

what may be termed a transition state. With these lIO suehstrict unifurmity prevails. While the differences are not excessive, yet the hinge of each individual specimen comparedwith others of the sallle age will show individual characteris

tics and the changes which the hinge undergoes in the sameindividual between adolescence and old age are greater than

one would ordinarily find in the whole membership of aspecies, say of the Venerid(1!, taking all ages, above the larvalstage into account.



458 IJall-Hinge of Pelecypods and its ])evelopment.

l it tle processes, the feeble remnants of the original fossettes.An appendage analogous to and possiblJT homologous with an

original ossiculum has (that view being taken) revolved aroundthe cartilage, taken its place outside of the axis of motion ofthe hinge, and instead of keeping the valves from crushingeach other by checking the closiug stress of the adductorsas in Verticordia or Buslu'a and other Anatinacea, it accom-plishes the same end by locking over the retlected edges of the

shell on the dorsal sul'face acting, like the anterior adductors,on the short instead of the long arm of the lever, and as beforein a sense opposed to the action of the adductors. Though

greatly specialized and modified this appendage retains some-

thing of the butterfly shape of a broad ossiculum.An appendage, sometimes called the styliform process orapophy61s, with its proximal end attached in the hollow of the

beaks, has been homologized by Deshayes with the cardinalteeth. In Plwlas costata it supports the posterior oral palpuswhich is very massive, and some of the internal viscera. I fone of the umhonal laminffi of Oallocm'dia were detached from

its connection with the cardinal margin and allowed to project

into the cavity of the valve, it would 80mewhat resemble the

apophysis of PllOlas. But on this view I am at a loss to

explain the present connections of this process about thedevelopment of which little or nothing is known. How acardinal tooth should come to he situated inside the mass of

the body would seem to be hard to explain. The environment

of the Pholads is of a very special character and the modifica-tions of the organization march with the peculiar circumstancesunder which it exists. To enter into their llIutual reactionswould takp much space and obscure the more general questionsto which this papC!' is addressed.It may be added that in this order as well as the others the

particular constituenp.y of each of the suborders, even the

number and scope of the families, must be regarded as tingedwith uncertainty from the magnitude of our ignorance. To

properly ascertain and correlate the data in regard to the differ-ent genera and the families of which thev are the memberll isa labor worthy of devotion, but whieh will yet require a largeamount of original research.

In the Pl'ionodesmacea the .lYuclllacea represent an archaictype in many of their features. So far as the hinge is con-

cerned Area. (Nom am] related species) is perhaps the mostfully and typically developed instance of Prionodont dentition.The Naiades declare in Spat/ta and Iridina their Priollodont

origin, traces of which are to be seen in the transverse stria-tion of the teeth of many species of [fnio, even when lateralteeth have become well developed and preeminent. The same



Dall-Hin(Je of Pelecypods and its Development. 459

is true of Trigonia which has many points in common with

Naiades. To the latter immediately XUlleria bears such a

relation in its adult state as do the Monomyarian Pecten and08tl'ea to the rest of the Prioftodesmacea as a whole. The

Prionodollt character of the Mytilacea will not be questioned.Through them we pass to the Pectinacea, in which in 8polllly-lU8 we have the finest instance of a Prionodont h i n ~ e with fewteeth, as Area is of one With many teeth. The original trans-verse grooving of the hinge is visible on the very young valvesof many species of Pecten, Janira, etc. The Ostl'acea are thelast term of specialization in this line; the Anomiacea, arebrought in by the total of their characters, though so far modi-

fied as to indicate little, by the hinge, of what I suppose to betheir origin. Above all it must be admitted that the Mono-rnyaria and IIeterornyaria represent not fundamental typesof structure but special modifications though geologically an-cient. The presence of a prismatic layer of cretaceous otellsnbstance, outside of the pearly layer, is also characteristic of

most of the forms of this order.The remaining forms representing the march of p r o ~ r e s s

toward a mechanical perfection in hinge characters. though re-taining traces (as in the striated teeth of some Mactras) of

Prionodont ancestry which once dominated the dentition, con-stitute the order Teleodesmacea.

In the main, in the ~ o m b i n a t i o n of hinge characters whichthey represent, the most striking features are the effectivemanner in which the orthodont laterals and prionodont cardi-nal teeth are subordinated to and supplement each othersaction, the occasional introduction of the internal cartilage inhappy combination with the others and the general absence ofnacre in the shell structure and archaic characters in the soft parts.

It is a qnestion whether the Rudistes are to be considered agroup apart, or, like the Plwladacea among the Anomalodes-7YUtcea, merely an erratic special development, of forms relatedto the Charnacea. Leaving the question to be settled by the

special stndies its difficulties call for, I conclude this paperwith a tabular view of the orders and suborders into which the

class is divided. Olle group, the Leptonacea, stands much inneed of thorough study without which its component familiesand even its permanent standing must remain doubtful. 'Vithour present knowledge it is yet impossible to determine the

number of families of which each suborder should be com-posed, or even how many gronps are entitled to rank as fami-lies. Rut in the major groups I feel a certain amonnt of con-fidence that the present arrangement is in most respects moreharmonious and in accord with the balance ()f characters thanany of the systematic arrangements of the class which havebeen hitherto proposed.



460 lJall--IIinge i f PelecypodB and its lJevelopment.

CLASS PELECYPODA.

I. ORDER AN01U.LODE8MACEA.

Suborders.1. Solenomyacea.2. Anatinacea.

3. Myacea.4. Ensiphonacea.5. Adesmacea.

II. ORDER PRIO:YODKSMACEA.

Sub01·ders.

I. Nuculacea.2. Arcacea.3. Naiadacea.4. Trigoniacca.

III.

5. l\lytilacea.6. Pectinacea.

7. Anomiacea.8. Ostracea.

ORDER TELEODESMACEA.

Suborders.

1. Tellinacea.2. Solenacea.3. Mactracea.4. Carditacen.5. Cardiacl:'ll.6. Chamacea.7. Tridacnacea.

8. Leptonacca?

9. Lucinacea.10. bocaI'd iacea ?I 1. Veneracea.

... .? Rudista.

SuppU177U'ntary iVote.-When I first began to consider the

relations of the teeth and other parts of the hinge, I naturallyremembered the brief abstract of the important paper on the

hinge of Rivalves by :M. Neumayr which I had seen in the

Zoological Record for 1883. I intentionally deferred a carefulpemsal of Neumayr's eBEay until I had entirely completed myown. Then a careful examination of his original afiorded megreat pleasure. I t showed that in the matter of the influenceof ribbing in promoting the nascence of teeth; in the discrimi-nation of lateral plications, arisinv in connection with the fos-sette of the cartilage, from the true cardinal t('eth; in the in·fluence of the environment on the degeneration of hingecharacters: in the estimate of the characters of the primitivebivalves; and some minor points we had arrived independently

at the same conclusions and ('ven illustrated them by identicalor nearly identical examples. This is certainly strong presump-tive evidence of the correetness of those inferences. In thepoints in which we differ. it seems to me that the differencesarise from the fact that Neumayr has approached the subjectmore from the paleontological standpoint and has less consid·



Dall-Hinge of Pelecypods and its Development. 461

ered, or has given less weight to biological considerations not

imprinted on the shell; while in my own case from the nature

of my previous studies I have been led to attack the problem

from the other side. Recent investigations, available onlysince the date of Neumayr's paper, have thrown much light

on the inosculation of characters not before known to interlace.Neumayr also, from my standpoint has insufficiently graspedthe importance of the different processes involved in the production of the internal cartilage and its shelly coefficients onthe one hand and the denticulation of the hinge margin on the

other. These two pracesses, though they must often have proceeded simultaneously in the same genus, were not necessarilyconnected except in so far as by resulting stresses each might

react on the hinge-product of the other. So instead of havinga Desmodont type of hinge as opposed to a Prionodont, and, asNeumayr would Ba.Y, a Heterodont (Teleodont) type, we ma'yhave either an Anodont (Paleoconch), a Prionodont (Taxodont),or a 'l.'eleodont (Heterodont) type of hinge, either with or

without an internal cartilage and its accessories.By the elaboration of this view, as attempted in the fore

going diseussion, it seems to me the discrepancies so evident inNeumayr's system have been avoided; the types of hinge

assigned their proper wcight in the system; while thosebiological relations which are not fully reflected in the shellyparts, have not been slighted; though inevitably numerous improvements in detail will suggest themselves to students, or beeffected b'y a future expansion of our knowledge.

As regards the Rudistes, if, as claimed by Woodward andothers, they possessed an internal cartilage, it is probable that

they must form a specially modified and extraordinary ramification of the Uhamacea. If, however, as is claimed by some

authors, there was no internal cartilage or external ligamentand the smaller valve simply rose and fell under the control ofadductor muscles guided by interlocking processes, it is evident, that this would establish an interrelation between the

valves unlike anything among the Pelecypods, and only comparable, perhaps, with that of certain operculated corals. Inthe latter case the Ruclistes would have to be regarded as ranking at least among the subclasses, if mollusca at all. My ownimpressions are that the first mentioned view is the more prob

ably correct one.The opinion is occasionally expressed in scientific literature

that the shell is a .• mere secretion of the mantle." This

usually proceeds from some person who has not thoroughlyst.udied the molluscan shell, or appreciated its relations to the

animal. Such a statement is one of those half-truths which are

n



462 Trowbridge and Sheldon-Magnetism qf

more dangerQus than pure error since the ballast of truth theycontain will enable 'the error to navigate some distance, whilethe unfreighted error would capsize at once.

The shell is in one sense the product of secretion from themantle, as the mammalian tooth is deri ved from the ectodermof the jaw or the skeleton from the pm'iosteum and cartilages.Both are that and much more. It would be as reasonable tl>

sar that a steam boiler in process of construction is the productof the boy inside who holds the rivet-heads, as to claim that theshell has no more significance than is implied in the term.. secretion of the mantle."The original theoretic protoconch may have been so, but, as

soon as it came into being, its development was governed bythe physical forces impinging upon it from all sides andthrough it influencing the growth and structure of the softparts beneath. The Gastropod shell is the result of the actionand reaction between the physical forces of the en \ 'ironmentand the evolutionary tendencies of the organic individuar. Inthe Pelecypod we have the mechanical stresses and reactions ofone valve upon the other added to the category of influences.To some extent it is doubtless as true that the animal ismoulded by its shell as it is that the shell is shaped by the soft

parts of the animal. This results in that correlation of structure which has enabled students to, in the main, correctlyjudge of the relations of mollusks by their shell-characters, whenthe latter were intelligently studied and properly appreciated.

ART. LVI. - The Magnetism of Nickel and TungstenAlloys;* by JOHN TROWBRIDGE and SAMUEL SHELDON.

INTRODUCTORY.

THE fact that diffm'ent kinds of steel, alloyed in small proportions with tungsten or wolfram, and magnetized to saturation, increase in specific magnetil'tn,t has long been known.·Whether the same effect would result from the use of nickelalloyed with t u n ~ t e n has never been investigated. Thispaper has for its o1)ject a partial answer to the query. I t wasinstigated by Mr. ·Wharton, proprietor of the American Nickel·Work>!, whose chemist, Mr. Hiddle, kindly prepared the alloyswhich have been employed. These alloys were in two groups.The first, recei ved in Novembel', 1888, consisted of three barsof the same shape, one being of pure nickel and the other two

• From the Proceedings of the American Academy of Arts and ScienlX's.t Jour. Chern. Soc., 18G8, xxi, 284, says 300 per cent.

1889 dall on the hinge of pelecypods and its development

Documents