Eridostraca a new suborder of ostracods and its phylogenetic significance

A C T A Vol. VI

P AL A E O N T 0 L O G I C A 1961

FRANCISZEK ADAMCZAK

P O L O N I C A No. 1

ERIDOSTRACA A NEW SUBORDER OF OSTRACODS AND ITS PHYLOGENETIC SIGNIFICANCE

Abst r ac t . - This paper reports the results of investigations on the st r uc t u re and text u re of multilam ellar ostra cod cara paces an d of com pa rative stud ies on the micro structu r e of the she ll in Leperd itiida e Jones an d Heal d ii da e Harlton. The wri te r's 'inf erences are that the multilamell ar ostracod carapac e has been derived from Conchostra ca (P hy llopod a) with conc entric growth lin es. The con sequ en t result is the sep aration of the grou p Eridoconchinae Hennin gsmo en into a disti n ct su bor de r - the Er idostrac a . A new genus , Abe r roconch a, has been establis he d in this subord er and two spec ies fr om the Mid dle Devonian of Pola n d are described as new : Aberroconcha plicata an d A. devonica. Moreov er, one new species of Eri doconch a Ul ric h & Ba ssler, equally of Devonian age, is descr ibed : Eridocon ch a gran ulifera. A con ce pt of the diphyl etic origin of ostracods is postulat ed, based on

the morphology and microstu ct ure of thei r carapa ce.

INTRODUCTION

Th e here r ep orted resu lts of studies on the ori gm and st r ucture of th e ostracod carapace do not clear up a ll the problems conce rning this m atter. Th e qu estion, given h er e th e m ost att ention , i s tha t of the occ u r rence of multilamellar she lls of certain ostracods w hich ha ve, since long, b een r ecorded in p al aeontological literature and known as E r ido- conc hi nae He n ningsmoen, 1953. Det ail ed stud ies on the morphology of the carapace of Ordovi ci an and Devoni an r epr esen t at ives of t h ese ostracods reasonably su ggest their separation in t o a new subor de r , for w hich t he na me Eridostr aca is introduced.

The she lls of these ost r acods are multilamellar, being composed of

up to 11 lam ell ae underlying one anot her an d expressing t he su ccessi v e

growth stages. In m or phology and structure t hey resemble the carap ac e

of Conchost r aca (P hyllopoda). In t hese , ho wev er, the carapac e is mad e

up mostly of chitin , or of chitin p ermeat ed wit h calcium carbo na te, less

ofte n of calcium carbonat e only. To the ne w suborder Er idostraca th e

writer h as r eferr ed , b esid es the genus Erid oconch a, in 1923 establish ed

by Ulrich and Bassler , also his n ew genus A berroconcha des cribed in th e

present pa pe r , as w ell as gen er a with t h e carapace up to 6 lamellae,

some times even onl y unil amell ar , bu t di splay ing cl cs e morpho logical and

30 ^FRA N CISZ E K AD A MC ZA K

genetic r elat ions hips with the just mentioned ge ne r a . Such are th e genera Cryptophyllus Levinson, 1951, Milleratia Swartz , 1936 , Schmid t ella Ul rich, 1892 , and probably Paraschmidtella Swartz, 1936. The above en u m er at ed genera correspond to a morphological series permitting to und er st and the origin of the unilamellar carapace of typical ostracods.

The supposition h ere made as to the derivation of the ostracod car ap ace f rom a conchostracoid typ e is based on a purely the oretical co n cept ion, ad vanced in 1892 by the zoologist Grobben.

is confirmed by fossil material with regard to the Eridoconchidae and t o the genera Crypt ophyllus Levinson , Millera t ia Swartz and Schmidtell a Ulrich, probably al so t o th e family Conchoprimitiidae H enningsmoen (emend.

Pokorny , 1958). On the other hand, the Leperditiidae Jon es, 1856, ow in g to fundamental structural differences of the carapace, have most pro bably different ancestors, not closely allied with Eridostraca n. subordo.

The ancestors of the multilamellar ostracods may possibly be certain Conchostraca with concentric growth lines, such as the Lower Cambrian genera FordiHa Walcott and Lepidit ta Matthew, known alr eady from the Lower Cambrian and united by Kobayashi (1952) into the family L epidittidae. On the base of ornamentation their carapace may be presumed t o have consisted, similarly as in representatives of recent Conchostraca, of numerous lamellae underlying on e another.

is not impossible that these forms may have been the ancestors n ot on ly of the Palaeozoic, but also of the living representatives of this group of animals. Eridostraca may possibly also have evolved from this assemblage of forms. Th e evol u t io n of this group was ex p resse d by the gradual increase of the ability to moult, which finally led to the development of a strong, -u ni lam ella r carapace of the ostracoid type . The above con sid er at io ns have suggested a revision of Levinson's (1951) hypothesis, know n as the "retention of moults" which, it is shown, do es not ad eq u at ely in terpret the occurrence of multilamellar shells.

The h ypothesis here advanced, as r egards the carapace of ostracods

consisting of successive moultings, is based, in the first place, on an

analysis of the structure of multilamellar valves of living Conchostraca

whose carapace, form ed of successive instal's, in creases throu ghout the

animal's life-time. Analogous conditions must have occurred in the group

of Eridostraca. Their carapace consists of calcareo-chitinous lamellae

di splaying a cryptocrystalline microscopic structure. In this it does not

differ from representatives of the gro u p of P alacc copa, the Leperditiidae

Jones excepted. The latter ostracod group is known to hav e the carapace

built of distinct calcite prisms. A similar structur e of the shell is likewise

noted in on e of the most primitive Podocopa (sensu G. W. Muller, 1894) ,

namely the Healdiidae Harlton, 1933. Studies on the microstructure of

th e carapace of Eridostraca and many repres entatives of the Beyrichiacea

ERIDOSTRACA - A NEW SUBORDER OF OSTRACODS 31

J ones, 1854, have r evealed many features in common, while fundamental structural differences .h av e been ascertained in this respect in L eper- di tiidae and Healdiidae. Detail ed comparative studies of the shells i n t hese groups suggest the or igin of the ostracoid carapace to h ave been associated w ith many group s of Cambrian crustaceans referred to Archaeo- straca Claus. This problem, h ow ever, is far f rom being definitely cleared up in the present paper. S om e light only is being thrown on phylogen etic rel ations, su ggested by shell structure, but these call for additional m ore extensive comparative st ud ies in other groups.

The basic and com par at ive materials used in preparing this paper include: Cambri an representatives of the Archaeostraca (collected by late Prof. J. Samsonowicz), fossil and living Conchostraca and many groups of ostracods from various periods and regions. Th e basic material has been collected from Middle Devonian strata in the Holy Cross Mountains - Gory S wi etokrzyskie (the Lysogora region) which represents a classical section of that age in Poland. Abundant ostracods of this section indicate close connections with the Devonian ostracod fauna in the Eifel Mountains. The occurrence is here noted of ostracod species and genera, lately recorded from Western Germany (Krornmelb ein, 1950 , 1952, 1953, 1954, 1955), namely: Polyzygia symmetrica Giirich (Adam- czak, 1956), Poloniella tertia Krommelbein (Adamczak, 1959), Baird ia seideradensis Krommelbein (Pribyl, 1953), representatives of Ko zlow- skiella (Pribyl) and many species belonging to the genera Bairdiocypris (Kegel), Pachydomella Ulrich, Condracypris Roth, and others.

is not excluded that a number of them may be used in stratigraphic correlation of these far distant areas.

Faunal similarities are likewise noted with analogous deposits in the Sovi et Union , in the first place with the "Main Devonian Area" where th e presence has been ascertained of the same species as th ose occurring in the H oly Cross Mountains, partly also in Western Germany, namely:

Polyzygia symmetrica Gurich (Adamczak, 1956), Pol onieHa

Dizygo- pl eura) curta (Polenova) and Bairdia seideradensis Krommelbein, Ther e is a number of genera common to all these areas, such as Bairdiocypri s (Ke gel) , Euglyph ella Wa rthin (Polenova, 1960c), Bufina Coryell

Malkin and others occasionally r epresented by extremely numerous species. Some faunistic elements of the Lysogora region are likewise present in North- -American f aun as , such are: Poloniella cingulata Warthin, Ponderodictya punctulifera (Hall), Eridoconcha rugosa Ulrich & Bassler, E. arsiniata (Stover), Ctenoloculina cicatricosa (Warthin) and the genera Aberroconcha n. gen., Holl in ella Cory ell, Abditol oculina Kesling, and others.

The excellent st at e of preservation of ostracods from the De vonian

of the Holy Cross Mountains, their diversity and abundance, mak e this

region one of marked palaeo geographical interest.

is a cross-road of

32 FRAN CI SZ EK AD AMC ZA K

faunal elements from the east and the we st, as is among others also indicated by the presence of representatives of Eridoc onchidae. Genera of this grou p (Aberroconcha n. gen. and Eridoconcha Ulrich & Bassler ) ha v e a ve ry wid e geographic and stratigraphic range. Beginning with the Ordovician they occur throughout the United States of America, as well as in the Siberian Platform (USSR) and in later times (Devonian) in Europe, too.

Use has been m ade in the preparation of th is paper of fossil material colle ct ed fr om four Middle D evonian profiles , involving Couvinian b eds at Wyd r ysz6w a nd Grzegorzowice, a nd Givetian - at Skaly and Swie- t om ar z-Sniadka. D evonian outcrops at the just mention ed lo cali t ie s hav e y ielde d, in addition t o ostracods, r epres entati v es of Tabul ata (Stasinsk a , 1958), T et r acor alla (Rozkowska, 1954, 1956), Brachiop od a (Biernat, 1954, 1959) and Trilobita (Kielan, 1954 ; Osmolska , 1957). The os t r acod faun a is div er sified a n d many species are ve ry ab u n dant, h owe ver su ch groups as the Erid ocon ch id ae a re ex t re mely rare. On the w h ole, on ly some sco res of s pecimen s bel on ging to species of the gener a Aberrocon cha n. ge n . and Eridoc on cha Ul r ich

Bassler have been di sc ov ered in the m aterial containing many tens thousands of v ar io us ostracod ca rapaces.

Thes e have be en obt ained by washing marly and marly-argill ac eo us r ocks m or e th an h alf a ton in weig h t . With a few ex ce pt ions, the speci mens a re in a very satisfactory stat e of preserv ation. The shells ar e ofte n f ill ed with calcite, thanks to which very delicat e int ernal ch itinou s st r uc t u res have not been damaged.

B esi d es fo rms described at some le ng t h in th e presen t paper, the writer h ad the a dva n tage of h aving at his di sposal comparative m aterial co ll ected outside of P oland , mos t use ful in hi s studies on the struc t ur e and texture of the carapace . Ver y valuable spe ci m ens of A berroconc ha from the Ordovician of t he Siberian P latform hav e be en most kindly su pplied by Dr V . A. Iv anova of the P al aeon t ologi cal In sti t u te in th e Acad emy of Sciences of U.S.S .R. in Mo scow. These have been lik ewise described an d incl ude d in t he chapter on sy st ematics . Mo reover, s ev eral sp eci me ns of r epresen tatives of th e genera .Eruioconcha" and Cry pt o- ph y Hu s L evinson h av e bee n graciously sent by Dr S. A. Levinson of the Humble Oil and Ref ining, Co. of Ho uston , T exas. Speci m ens o f C cn ch oprim it ia Opik w ere received throug h the courtesy of D r A.

N etzkaja from the VNIGR I Institute of Leningrad , whil e man y Devonia n spe cies from the Soviet Un ion hav e been kin d ly con tributed by Dr E. N . P ol enova of the sa me In st itu t e .

For t he sake of compar ison th e writer h as m ad e u se of living

Adriatic ostracods turned over t o him b y Dr J. Malecki of the P al aeoz oo-

logical Laborato ry at t he Academy of Mi ni ng and Metall urgy of Cracow ,

als o of Co nch ost r ac a (Esth eria sp .) k indly su pplied b y th e Zool o gical

ERIDOSTRACA - A NEW SUBORDER OF OSTRACODS 33

Institute of the Warsaw University. Moreover the writer has found some interesting material of Conchostraca among Prof. R. Kozlowski's material of Upper Silurian rocks fr om Podolia. Many valuable sp ecimens have also been coll ected by the writer from Ordovician and Silurian erratic blocks in Poland.

The present paper h as been written at the Palaeozoological Laboratory of the Warsaw Unive rsity, under the guidance of Prof. R. Kozlowski, t o whom the writer here expresses the w armest thanks for the unrestricted assistanc e tendered to him. A cknowledgements are also due to all th e persons mention ed hereabove, for their friendly attitude in m aking avail able to him so m an y , often very valuabl e, sp ecimen s.

Th e drawings and figures in this paper have been done by Mrs K . Budzynsk a and Mrs D. Platajs, the microscopic sections have be en ph otographed b y Miss L . Luszczewska , the English translation of this pa p er has b een mad e by Mrs J. Humnicka, and the writer thanks them cordially, too.

METHODS

Problems he re dis cus sed concerning the origin and phylogeny of ostracods hav e been studied in the first place on an analysis of the sk elet al parts of fossil fo rms. Particular attention was focussed on th e struct ure and microscopic texture of sh ells . P al a eontological literature contains· but very meagre data connected with such problems. Compara- tive systematic studies of this type have not , thus far, been m ade of sh ells pertaining to different groups of ostracods, eve n though they hav e proved of importanc e in w h at phylogeny is concern ed. This is all the more so considering th at cr it eria for re cent forms cannot always be used in th e cas e of many fossil groups. These studies n ecessitate very special and time-consuming m ethods, such as thin precisely oriented sections, serial sections, fluoridization (Sohn , 1956; Upshaw, Todd

Allen, 1957), treatment of shells in acids, etc.

The complicated structure of the carapace of Eridoconchidae Henningsmoen has, in the first place, been investigated by means of serial sections and fluoridization . The latter method is a very h elpful one since it permits the observation of in t ernal morphology without d estro ying the shell. The sections through carapaces of Eridoconchidae, as well as of L eperditiida e Jones and Healdiidae Harlton , have l ed to a thorough knowl edge of their structural elements having taxonomic sig n ific ance . B y those m ethod s it has been possible to clarify th e com plica ted internal structure of the carapace of Aberroconcha n . gen.

Th e dorso- central parts of ca r apace lamellae invaginated into the interior of the val v e (fig. 1), occupy a space hardly 0.1 mm in width. In order t o ascertain t he exa ct sequ ence of the gro wth stages a nd their mutual

3Acla Pal a e on l ol o g ica Nr 1

34 FRANCISZEK ADAMCZAK

relation, the section must cut these elements centrally, for outside of the area of maximum convexity of lamellae they are intimately confluent and resemble the structure of valves in the genus Eridoconcha Ulrich

&

Bassler. Observations of valves in transversal plane were supplemented

by serial sections (text-pl. I), cut parallel to the longitudinal axis of the carapace. These serial sections have made it possible at least partly to reconstruct the spatial structure of the shell and the mode of development of the dorsal part (fig. 2 and 6).

On thin slides the presence has been ascertained in valves of Bairdiocypris (Kegel), PachydomeLla Ulrich and Condracypris

sp. of an internal, extremely thin and strongly chitinized lamella (fig. 14

. As has already been mentioned above, fluoridization of shells is markedly helpful in the morphological studies of the carapace. Moreover, this method has been used when investigating the structure of the carapace. Recent ostracod valves are known to consist of three layers, the outer and the inner ones being chitinous, the middle one - calcareous.

Among fossil forms a three-layered carapace has been ascertained in Leperditiidae only (Levinson, 1951). In other groups of fossil ostracods the carapace is not so well differentiated. Though a kind of inner layer has been noted. in Bairdiocypris iixheimensis (Kegel), particularly so within the free marginal area (Krornmelbein, 1952, pl. 4, fig. 1 a-c), Kr6mmelbein's interpretation (I.c., p . .322) of that occurrence arouses some doubts (comp. p. 65-66 of the present work). No such structures have been observed on thin slides of the carapace of Palaeocopa, whose shell is colourless and with a cryptocrystalline structure.

Fluoridization and thin slides cut through shells of various groups of Palaeozoic ostracods have revealed the presence in Leperditiidae and Bairdiocypris sp., as well as in other Podocopa (Condra cypris

sp.), of a distinctly three-layered structure (fig. 13

while in representa- tives of Palaeocopa the carapace is most commonly two-layered (fig. 12 C;

pl. II, fig. 1). In the course of fluoridization of shell, e.g. in Leperditiidae, the prismatic structure is not obscured, but quite contrary it becomes more stressed (pl. II, fig. 7). On the other hand, in Podocopa which have the shell with extremely fine calcite prisms, the primary microstructure is sometimes completely effaced (pl. II, fig. 5).

In the group of Palaeocopa (the Leperditiidae excepted) presence of

muscle scars is still an open question. In spite of the

of diverse methods

such as fluoridization, polishing of the outer shell layer and cutting of thin

slides, etching of shells by weak acid solutions, immersion in liquids with

strong light refraction, particularly so in monobromnaphtaline (Triehel,

1941) , no adequate results have been obtained. These methods have only

revealed the position of the muscle attachment which, as a rule, is

indicated as a rounded area with stronger light refraction. In what the

ERIDOSTRACA - A NEW SUBORDER OF OSTRACOD S 35

muscle scars are concerned , excell ent r esults hav e b een obtained in Bairdiocypris sp. b y treatment in a weak solution of hydrochloric acid, a nd for Bairdia sp. by fluo ridization of shells.

In treating r ocks yie ld ing ostracod shells, m arly and ar gill ace ous deposits have been was he d in water after being red uced to fragm ents by m eans of Glauber salt or lixivi ation . In the case of limest on e rocks r oast in g at high temperatu r e (up to 900

and r apid coo lin g in water h ave been used. For this pu rpose th e sed ime n t was brok en up into fra g- ments of ca. 3 em, w h ich cru mble d up aft er b eing su b jected to re pe ated heating and coolin g. A disad van t age of th is m ethod is t he large per- cen t age of damaged sh ells (particularly so in the case of l ar ger ostracods and of strongly ornamen t ed forms). Sudden temp er ature change did not cause such strong disintegration of the sh ell in small and smo oth er forms resulting in a small er number of damaged specimens. Neither doe s their microscopic structure alter, but they becom e whiter than the uncalcinated specim ens , and covered with a minute n etwork of cracks.

By using the ab ov e described method it has be en p ossible to isolate from the rock numerous shells of v arious groups of ostracods which have served as comparative material in the preparation of the pr esent paper.

TERMINOLOGY

Th e t erminology here used to define th e morpholo gical elements of the carapace of Eridoconchidae H enningsmoen is taken partly from earlier palaeontological and zoological works concerning Conchostraca (Grube, 1865; Raymond , 1946; Novoshilov , 1954, 1960). Terms accepted for and used in papers on ostracods (Schmidt, 1941 ; H essland, 1

Kesling , 1951 c; Jaanusson, 1957 ; Triebel, 1958 ' b) and other groups of animals hav e been introdu ced by the w r it er .

b elo w are given the morphological definitions for the main st r uctu r al eleme n ts of the carapace in Eridoconchidae. Definitions of t er m s concerning the stru ct u re of th e carapace in other ostracod s cited in this paper are not given since they are commonly known a n d used. Bri ef exp lan at ions only acc om p any th eir illustrations.

Nau,pLi oconch = shell of th e first (post-embryonal) grow th stage , forming the umbo on a multil amellar ca rapace (fig. 1).

Multilam ellar carapace = a carapace consisting of moults not sh ed off during gro w t h. The m ost common occurrence of this t ype of carapace is encountered in t he Conchostraca.

Lamella = an in sta r or a m oult stage corresponding t o a gi v en growt h ph ase of t h e multilamell ar carapace wit h w h ich it is closely m orphologic all y con n ect ed . In a sing le lamella, not yet q uite calcified, the followin g parts m ay be dist in guished: a proximal cal cifi ed portion, by t h e present w r it er common ly r ef erred to as t he dorsal or h inge

3'

36 FRANCISZ E K ADAMC Z A K

lamella, the central one consisting of chitin - hence the chitin l amella, and finall y th e distal lamella, calcified as the dorsal one, and including t h e growth band (Ra ymond , 1946; No voshilov, 1954) whose peripheral en ds form the free edge of th e she ll during successive growth stages.

At the base of th e invagination of the l amella occurs a semicircular list.

91 .

2

A ·--- -- ""'---:=--

B

C ---

D- - E

F-- G--

H -...

II I

7

---4 ---3 --5

Fig. 1. - Aberroconc ha plicata n. sp., comb ine d diagram vie we d from the anterior part of the carapac e

-'\ na uplio con ch , B-J successive growt h stages , 1 hin ge , 2 dorsal la m ella e , 3 growth bands (distat part of lamella), 4 inter lamellar cham ber . 5 chitinous lamella , 6 con ce n tric gro ov e , 7 free edge, 8 growth lines (concen tric groove), 9 dor sal margin. 10 semici rc u la r lis t.

Gro w th band = ^{t he out er di stal} part of the carapacial lamella,

co r respo nd ing ito ad vent r al Structures (Jaanusson, 1957) in ostracod

ca r a paces from t he P al a eocopa group. On a multil amellar carapace

th ese el emen ts are called c on centric r in gs (Ulrich & Bassler ,

ri dges (Levins on , 1951), wh ich indicate the number of lam ellae com-

pos in g t he cara pace. The particular grow t h bands are separated by

ERIDOSTRA CA- A NEW SUBORD ER OF OSTRACODS 37

grooves, the so-called growth lines (Raymond,

Novoshilov,

Interlamellar chambers = free spaces produced by ch itin lamellae, invaginated into the interior of the valve. Neither their or igin n or fun ction is fully understood (comp. p.

Bisecting line (J aanusson,

= a dark line (Adamcza k ,

cutting - parall el to the surface of shell - the adventral structur es.

It

cons tit utes the fi r st stage (sinus) in the f ormati on of a sma ll f r ee space within th e growth band, in t he present paper r ef erred t o as th e lacuna.

GENERAL P ART

ON THE ORIG IN OF OSTRACO DS

The origin and ph ylogeny of ost ra cod s have not as ye t been definitely cl arified . Many of th e problems connect ed ther ewith a r e still hi ghl y h ypothetical. Zoologic al and palaeontological concepts advanc ed with rega r d to t h e phyloge ny of these animals often di sagree. A distinct la ck is felt in this fi eld 'of a const ructive theory supported by b oth - palaeonto logical an d recent m aterials. Neither will a solution of these p ro b le ms be f ound in the vi ews of Pokorny

based on investigations of e arlier authors (Muller ,

R aymond ,

or in the cla ssification and new systematic gr oups in t r od uced by him.

Many of the ost racod groups differentiated by Pokorny are not mutu all y relat ed (Leperditiida Pokorny , Bey ri chiida P okorny , Podocopida P o- k orny) .

In spite of the great abundance of th e av ail abl e foss il m aterial it has not been poss ible more clo sel y t o co r rela te man y of th e ex tinct ostracod groups w ith the living forms. E ssential differ en ces occur eve n amo ng fairl y we ll kn own groups of primitive os t r acods, t o wh ich

the first place b elong L e p erditellidae Ulrich

Bassl er,

and Leperditiida e J on es,

Even though some authors (Bassl er

K ellett ,

H enningsmoen,

unit e t hese gr ou ps in to o ne supe rf amily , t he L ep erditacea, yet this standpoint has not be en commonly acc epted".

In what re gards th e L ep erditellidae and the di vi si on of this famil y by Henningsmoen

into sm aller t ax on omic uni ts: L ep erditellin ae Ulrich

Bassl er ,

Eridoconchinae H enningsm oen ,

Con ch o- primitiinae Henningsmoen ,

this con ception wa s t h en quite corr ect an d expressed th e opinions h eld and th e state of knowl ed ge at th at

I Quite recently the genus Apar chites Jones, 1889, has by Polenova (l960b) been included into the Lep erd it ellida e. Eo ipso, this family has au to mat ically become a synonym of Apa r ch it idae Jon es, 1901 (P okorny, 1958). Moreover, Polen ov a has erected the superfam il y of Aparchitacea to conta in the families: Aparch itidae, Graviida e Polenova, Aechminida e Boucek, This group, togethe r wit h the Lep er di- tac ea , has bee n included into the Leperd itiida Pokorny.

38 ^{FRANCISZEK ADAMCZA K}

time concerning the m or ph olog ical struc t u re of sh ells of this grou p.

R ecent in v esti gati ons, an d part icula rly so data r elating to t h e internal structu re of Eridoco nc h idae h av e revealed some fund amen t al struct u r al and mor phological diff eren ces (fig. 2). Sinc e they invol ve not on ly th e re maining subfa milies of L eperdit ellid ae, but the Pal a eocopa as a whole, too, t hey h av e - on t he base of their multilamell ar cara pa ce - been

EE

,..,

C;

Fig. 2. - Aberroconcha plicata n. sp., schematic drawing of the carapace, oriente d sli ghtl y obliquely towards the anterior part. Left valve shows in cross sect io n

a reconstructed spat ia l picture of chitinous la mellae.

se p ar ated by the present writer into a d istinct fam ily, the Eridocon- ch idae , and placed in a new subor der - t he Eridostraca. Th is new

has som e featur es in common n ot .only with th e Conch ost r a ca (multila- m ellar carapace), but with the Ostracoda, too (ability to produ ce a un ilamellar calcite carapace). Among t he ostracods the Erid ostraca occu pies an exceptiona l phylogenetic posit ion.

Moul t stag es are of particular sig nificance in the st u dy of ostraco d

ph yloge ny . Th ey are an impor t an t factor in the st u dy of the pas t history

ERIDOSTRACA - A NEW SUBORDE R OF OSTRACOD S 39

of thes e animal s. Moreov er, the y permit to ascertain the modes of changes through which their evolu tion was realized (Adamczak, 1959).

P alingenetic cha r acters ar e fr eq uently preserved on young she ll s; these provid e very sound suggestions for the determination of mutual r el ationships between the lower taxonomic groups. The youn gest

w ou ld indeed be here of gr eatest significance, but u n fortu n at ely, owing t o unfavourable con dition s, the y are but seldom preserv ed in the fossil sta te.

During the latest y ear s, wi th the purpose of clearing up taxonomic and phylogenetic problems, much attention has been fo cussed on the st udy of the microscopic structure of the carapace by m eans of thin sli des (Hessland, 1949 ; Triebel, 1950 ; Spjeldnaes, 1951; Levinson, 1951;

K rommelbein,1952 ; Jaanusson, 1957; Kesling,

Adamczak, 1958).

Application of this method permits to take into account charact ers which, thus far, have scarcely received due attention.

se ems that , besides ontogeny, investigations carried on from this standpoint permit to trace changes of phyletic importance independently of those in the ornamentation of shell. In the course of research work on the Eridocon- ch idae, it was found that structural studies of fossil ostracods may help to solve many problems other than taxonomic.

has been ascertained t ha t the cryptocrystalline structure of the carapace of Eridoconchidae is common to many groups of Palaeocopa, the Leperditiidae excepted.

Th ese, on the other hand , display an aston ish ing resemblance with th e P odocopa (in the s ense of Muller, 1894). Observations made during t he stu dy of this problem likewise shed some light on the origin of th e ostracod carapace and on ostracods in gen eral. They permit to understand bet t er the ev olutionary t r ends of these an im als from a different aspect t ha n has be en don e h eretofore.

seems that considering the origin of th e cara pace , t he ostracods co r respond t o a grou p rather di phyl etic than monophyletic.

In w h at r egards th e opinions held b y palaeontologists concerning t he phylogen y of ostracods, probably th e most popular h ypothesis i s tha t fi r st advanced by Ulrich · and Bassler (1931) postulating th at th e bivalved Cambrian Archaeostraca from the group of Bradoriid a e Matt he w and B eyrichionid ae Ulrich & Bassler may correspond to th e ancestors of ost ra cods. Th ese authors w r it e : "For the r easons mentioned we are confid ent th at t he Bradoriidae and allied Cambrian Crustacea are not true Ost r acoda. How eve r , as they pr eced e the Ost r ac od a in time, it is quit e probable that t he latte r were deri ved from the ear lier Brado- ri idae and Bey ri chionidae" (I.c., p. 3). This stand poin t w as su bseq ue ntly accept ed by R aymon d (1946) and m any ot her palaeo ntologists.

Th e Cambrian A rch ae ostr aca h ave no d efinit e taxono m ic position;

t hey are consid er ed as Arthropod a incertae ordinis (Sh r ock

Twenhofel ,

40 FRAN CISZEK ADAMCZAK

1953). This does not diminish the interest they arouse since it is among them that we should most likely search for ancestors not of ostracods only, but of other groups of Entomostraca, too.

Without now developing in detail these ideas, the writer wishes to recall here the opinions of zoologists concerning the problem of ostracod phylogeny (Grobben, 1892; Giesbrecht, 1921) which, as a rule, are neglected by the palaeontologists. Muller's (1894) conception will not be here considered, as not bearing much significance on questions studied in this paper. That author does not, indeed, concern himself with the concrete group of the ancestors of these animals, but presents only the hypothetical "prae-ostracod" form (Protoostracoda Skogsberg).

Far more interesting are the opinions advanced by Grobben whose chief assertments have been confirmed by palaeontological evidence.

In a short report "Zur Kenntnis des Stammbaumes und des Systems del' Crustaceen" Grobben (1892) discusses three morphological types of the living Phyllopoda (Branehipus, Apus and Estheria) which simul- taneously reflect the evolutionary trends in this group of animals.

According to that author, the type Estheria has certain features in common with ostracods, as well" as with Cladocera, and this does not certainly arouse the slightest doubts. As regards the ostracods, Grobben thinks (l.c., p . 247) that " ... lassen sich auch hier genugende Anhaltspunkte gewinnen, welche die Herleitung del' Ostracoden von einer Estheria- -Typus angehorigen Urphyllopoden zu stutzen vermogen". These

"Anhaltspunkte" are foremost the laterally compressed body and the

"bivalved" carapace protecting the body. The latter is, according to Grobben, an element of a great significance in the interpretation of ostracod phylogeny. Grobben's considerations were based on Myodocopa, representing today the most primitive body type, and providing suitable material for comparative studies. Disregarding, however, this side of the problem, the present writer focusses his attention chiefly on palaeonto- logical materials (Eridoconchidae), which support

hypothesis that forms of the conchostracoid type with concentric growth lines may have been the ancestors of ostracods (though of not all of them).

Giesbrecht (1921) disputes the main concepts of Grobben and rejects

his leading idea. He discusses but the aspect of the problem of phyletic

significance, namely the comparative anatomy of the body, disregarding

elements studied palaeontologically. In Giesbrecht's opinion (Lc., p. 227)

the shells are convergent structures, consequently of small significance

in phylogenetic speculations. Hence that author does not, contrary to

Grobben, trace the ostracods back to the Conchostraca, but to Proto-

straca, i.e. to a hypothetical ancestral group from which may, indeed,

descend all the Crustacea after they had passed the stage differentiating

them from the Copepoda. Giesbrecht's views are not confirmed by fossil

ERIDOSTRACA - A NEW SUBORDER OF OSTRA CODS 41

evidence and it seems that his inferences are one-sided, since they are based on evaluation of r ec ent materials only.

STRUCTURE OF CARAPACE OF ERIDOCONCHIDAE

Outer morphology of shell

The carapaces of Aberroconcha n. gen. and Erido concha Ulrich

Bassler are distin guished by an extremely characteristic type of orna- mentation du e t o th eir pe culiar mode of growth. Th e outer surfac e of

E E

C'2_

o

- --f ---G --- H - ---1

____e. J

Fig. 3. - Aberroconc ha plicat a n. sp., combined diagram as seen from the dorsal pa rt of the carapa ce ; righ t valve polished up to midhe ight

F-J lamellae of las t gr ow t h stages. 1 int e rl a m ella r cham ber, 2 ch i tin ous lam ella .

th e car ap ace, consisting of calci te, is mark ed by numerous (up to 10) concentric bands and grooves (fig. 1, 2), in literature frequently re fe r re d t o as ridges (Ulrich & Bassler,

Levinson, 1951) or concentric rings (Harris, 1931), som etime s as concentric bands (Matern, 1929). This typ e of orname n tat ion displays certain analogies with pelecypod ornamentation , as is mentioned by Coryell and Williamson (1936).

As regards co m par ison with the last named group, it must be stress ed

that the growth pattern is here entirely different. In p elecypods the

42 ^{FRA~CISZEK ADAMCZAK}

growth of shell is essentially continuous , while in Eridcconchidae distinctly periodical. In the latter group each band may be interpreted as a definite growth or moult stage.

In Eridoconchidae the carapace outline is ovate, posteriorly cordate (Harris, 1931; Keenan, 1951). The particular grooves and bands parallel

Fig. 4. - Aberroconcha plicata n. sp., longitudinal section of ca ra pa ce, cut belo w the invagination of the la st gr ow t h st age

F-I layers of last growth stages

the fr ee edge and con ve r ge below the umbo (fig . 3), which is form ed by the first growth stage, here r eferred to as nauplioconch. This is directed slightly to the front and hence is h elpful in det ermining the or ien t at io n of carapa ce. Th e umbo is variously shown . In Aberroconcha it is conspicu ous, in Eridoconcha (though this is no rule) the ap ex of the cara pac e is slightly flatt en ed.

Sh ap e of th e c ar ap ace of Eridoconchidae varies. Most likely it was

strongl y affected by en viron m en tal cond it ions and d ep en ds also on th e

development of the animal's body.

may be considered to b e a r esultant

of th e action of these two factors . D orsally and ventrally they are

ERIDOSTRACA - A NEW SUBORDER OF OSTRACODS 43

distinctly stream line in shape, with a smooth surface, occasionall y only marked by minute punctae on the growth bands.

The valves of the carapace are symmetric. No overlapping or shifting of valves has been noted (fig. 1, 4). The hinge-line is straight and short, placed in a depression , with both its ends meeting the free edge at an obtuse angle. The hinge is simple, freely confluent with one another on either valve, without forming differentiated structures. The free edge of the valve is formed by the last band. The number of the bands in Eridoconchidae varies from 6 to 10 (without the nauplioconch). Thi s number reflects the primitive character of the carapacial structure and is of marked taxonomic and phylogenetic significance. The maximum number of growth (moult) stages, ascertained in Beyrichiidae Matthew, is eleven (Spjeldnaes, 1951). These agree with the number of growth stages in Eridoconchidae. We may infer that this character (11 growth stages) is the upper limit attainable by ostracods of the Beyrichiidae group and that it corresponds to the number of growth stages in primitive Eridostraca.

The individual variability of shell is likewise connected with its individual development. Ontogeny of the carapace may be traced on one specimen in which all the growth stages have been preserved. In th e fossil state, however, young shells are known with a small number of lamellae (growth bands). They are markedly smaller than the adult individuals . Moreover, they differ from adult forms in the valve outline which varies strongly during the growth process (fig. 18, 21). Usually th e posterior ar ea of the carapace is more strongly d eveloped than th e anterior, so that the mutual ratio of these parts is 1:1.83. With growth these proportions gradually alter and in adult forms they are 1:1.24. Th e more inten se growth of valve in th e caudal direction during t he initial growth st ages may .probabl y be accounted for by the addition t o th e bod y of n ew segments. In connection with problems concerning the young forms we should take into the account that fr equ ently such individuals ma y be describ ed as ad ult forms with a lesser number of lamell ae on the carapace.

is not ex cl ude d that Eridoconch a minutissima Ivanova, described from the Ordovician of the Siberian Platform, represents indeed such a yo u ng form. This is indicated by the sm all size of she ll and the few growth bands (camp. p. 83).

Th e size of sh ell in Eridocon ch id ae varies. Th e la r gest shells in

the Dev oni an of t he Holy Cro ss Mount ains ar e not ed in A berr oconcha

plic ata n. sp.: approx. 1.08 mm, w h ile the smallest ones a re those in

E . granuLif era n. sp. - 0.62 mm, an d E. arsiniata (Stove r) - 0.50 mm .

The length/h eight ratio within the studied species of th e genus

Aberr oconcha and Er idoconcha varies, too . In A. pLicata n. sp . the mean

l ength/hei gh t ra tio is 1:1.26, in A. devonica n. sp. 1:1.16. Shell s of th e

44 FRAN CISZEK ADAMCZAK

...

l'l :tI

8

:tI

»

o

»

K L M

N ¥ _{O ~} ^»

l'l ::;J rnC tll 0:tI Cl l'l:tI 0"l rn0...,

»

o0 rnCl

Aberroconcha plicata n. sp., serIa l polish ed surfaces of the carapac e, pa r allel to the longitudinal axis; A- J dor sal view, U-K ventral view (sections placed in reverse order as compar ed with fig. A-J).

""

t1l

46 FRAN CISZEK ADAMCZAK

last named species are more circular, while in the former one they are slightly elongate. Analogous changes have been observed in E. rugosa Ulrich

Bassler (mean LlH being 1:1.39) and in the related E. granu- lifera n. sp., where it is 1:1.33. On the whole, it may be observed that during the process of phylogeny these extremely conservative r epresen- tatives of Eridoconchidae tend to produce a rounded carapace.

A peculiar problem is presented by the microornamentation of the carapace noted on detached valves of E. granulifera n. sp. in transmitted light and under strong magnification (X 450), (see fig. 22).

It

consists of a network with polygonal meshes. Moreover, in one of the species (E. rugosa), on partly decalcified valves, minute pits (canalicules) strongly resembling the pores (fig. 8) in living Conchostraca (Grube, 1865) have been noted to occur in the free margin area, as w ell as on the last growth band.

Detail ed morphology and architecture of the carapace

A shell of the Eridoconchidae is a structure consisting of nu- merous lamellae, underlying one another, each successive lamella being larger that the preceding one by a fixed value, approximately corresponding to the numerical values of Brooks' law (vide Kesling;

1951b,

1952c,

Observations described in the present chapter are based on thin and serial sections (text-pl . I) and fluoridized shells belonging to Aberroconcha and Eridoconcha. As a rule within these forms two architectural patterns of the carapace may be differentiated:

1) aberroconchoidal and 2) eridoconchoidaI.

The aberroconchoidal type is distinguished by a highly complex structure, readily observable only in sections normal to the longitudinal axis of the shell, or in longitudinal seetions (fig. 1, 3).

The cross section of the shell reveals a number of lamellae, each

one being a structure independent from the preceding one . This

structural pattern has been ascertained in Aberroconcha magna (Harris),

A . plicata n. sp . and A. devonica n . sp. In the first named species (fig : 9

17) lamellae of the early growth stages (I - III) consist of calcite

and are closely adherent. The first lamella is referred to as the

nauplioconch . Beginning with stage IV the growth stage displays

a different structure . In the proximal part sloping into the dorsum

and in the free edge they are calcareous, while centrally they are

chitinous. This structural pattern is consistently followed to the last

stage of growth. Chitinous lamellae, forming each moult stage of which

the carapace is composed, call for special attention. Th ey are domed

into the interior of the shell and do not lie close together , but produce

a sort of chamb ers. So far this pattern of structure has not been n oted

ERIDOSTRACA - A NEW SUBORDER OF OSTRACODS 47

A B · c D

Fig. 5 A. - Diagrammatic sections of the distal part of carapace in outer view:

A Aberroconcha magna (Harris), B A. plicata n. sp. , C Eridoconcha rugosa Ulrich ^&

Bassler, D CryptophyHus oboloides (Ulrich & Ba ssl er)

0-1 su ccessiv e ph ases of the last growth stages , 1 lac una, 2 bis e cting line (s h allow ba y sun us ), ~ conce ntric (U-sh aped) gro ov e, 4 broad ba y or fold, 5 semicir cular list .

"'F

"'G

A B c

Fig. 5 B. - Diagram s of inner su rfa ces of valves; for explanations - see fig. 5 ^A.

in anyone ostracod group and it seems that among Eridostraca they are

typical solely in representatives of Aberroconcha n. gen. The species

A . magna repr esents the most primitive type of carapacial structure

within the group of Eridoconchidae. Its lamellae are but slightly

calcified, while their chitinous parts occupy a fairly large area of the

sh ell. In thin sections of these forms the chitinous lamellae are distinct

as dark lines (pl. I, fig. 3). During the process of fossilization, secondary

impregnation of calciu m carbonate has been noted on these dark lines,

48 ^{FRANCI S ZE K ADAMCZAK}

displaying a different microstructure than the remammg ele m en t s of the ca ra pace w h ich have a cr yptocrystalline structure.

Stron ge r morphological differ entiation is d isplay ed by the aberro- conchoida l type of carapace in Aberroconcha pLicata n. sp. (fig. 1, 2;

pl. I , fig. 1, 2 a,

seems very pro ba ble t hat t his species is a successive mo rpholo gical link of the same evolu tion ary trend. Both these species altho ug h congeneric ar e not clo sely r elated t o one another, bel on ging to diff eren t phylogenetic t r ends. Independently of t hat f act they probably represent the gradu al t ransformation s of forms bel ongin g to t h e group of A . magna (Har ri s) in the ea r ly evo lutiona ry stage durin g t he Or d ovici an , or m aybe at the boundary of th e Camb ri an and Ordov ic ia n . In the speci es A . pLicata n. sp. the calcification of lamellae is furth er adva n ced . Their ch itin ou s su rface has decrea sed in favour of th e calc a reo us ones. These differ en ce s ar e particul arl y striking when compari n g th e final growth stag es in forms A. magna and A . pLicata n . sp. (fig. 9

In the latter sp eci es the di st al calcareou s parts of lamella e have grown proximally. The surface of calca reou s lamellae c onverg ing to t he dorsal area has likewise increased. These changes h a v e r esulted in incr eased capacity of the dorracilium, w hile th e ch itino us lam ella e form sm all er air chambers. Th e pa rticu lar growth b ands in A . plicata ar e characterized by gr eat uniformity. Morphologi- call y t hey are but little differentiated , w h ile in A. magna they form a kind of "podiu m " - in the last bands particularly so (fig. 5

The concentric gr ooves separat ing them are U-shaped in section.

A ccording t o L evinson (1951), the shape of the groove is taxonomically important. However, it m ay be actually asserted that on th e same individual of A berroconcha magna the section of th e sulcus v ar ies with t he gr ow t h of the sh ell. An analogous phenomenon has b een noted also in repr esen ta t ives of the genus Eridoconcha (E. granulifera n . sp.) , (comp. p. 84). During the ea r ly gr owth stages the groove is V-shaped , being U-sha ped in the last growth stages. In A. plica ta n. sp. or A . d ev onica n. sp. the groove is invariably V-shaped in section. Th e realization of eith er of thes e two types of sulcus is controll ed in the first place by the structure of the growth band. Th e m ore this resembles the " velate ridge" , the greater probability 'to produce a V-shaped sulcus.

As h as alr eady be en mentioned, the chitinous eleme n ts invaginated

into th e int erior of the v alve do not touch one another. In longitudinal

se ction (fig. 3) it is se en tha t lamellae of th e old er gr owth stages overlap

the yo unger stages . The maximum number of chambers formed b y

ch itinous l amell a e (t ext-pl. I , K) , not ed on on e section, is th ree . On

the out er surface of fl uo r idized -car a p aces t he chambers r es emble

a string of vesicles d ecr easing dorsally. On the inner su r fa ce of shell

ERIDOSTRACA - A NEW SUBORDER OF OSTRACODS 49

they form surfac es (fig . 6) diverging at the base of the dorsal area. The process of calcification involves a larger area dorsally and this gradually shifts to the ventral side. In A. devonica n. sp. this process is fairly well advanced. Dorsally nearly all the lamellae have been calcified and this process is accompanied by the disappearance of chambers in this zone, while in the central portion of shell changes are but small (fig. 9 C).

This species ends up the morphological series of the aberroconchoidal t ype. The last two species of Aberroconcha

pLicata n. sp. and

devonica n. sp .) are closely allied with one another and at the same time they correspond to the same phylogenetic line.

The eridoconchoidal type of carapace morphology is, in the first place, encountered in representatives of the genus Erid oconcha Ulrich

Bassler, and in some Conchostraca species found by the writer in Upper Silurian beds of Podolia . Species E. rugosa Ulrich

Bassler, E. granuLifera n. sp. and E. arsiniata (Stover) are representatives of the eridoconchoidal type of the carapace. In these forms the lame ll a e of the carapace have been completely calcified and mutually adhere throughout the surface.

Hence, the shell is massive and heavy. The number of lamellae continues to be considerable, ranging from seven to eleven. This morphological type of the carapace

linked by transitions with the preceding one and there is n o sharp boundary between them. The fundamental differenc e lies in that the latter forms lack air chambers, though some carapaces bear traces of them preserved as slightly corrugated calcite lamellae.

In representatives of Eridoconcha the carapace morphology is simpl e. On th e whole they are poorly differentiated forms with mono- tonous ornamentation. Size and contour of shell are, in the first place, subject to some variations. The growth bands, i.e. the distal parts of lamellae are still more conservative characters, subject to onl y un im p or t an t changes. Practically speaking, the shells of Eridoconcha are hardly distinguishable from those of Aberroconcha n. gen. on outer morphology only . This question is 'f ar simpler in the case of representa- tives of the genus Cr y pt oph y llu s Levinson (Levinson, 1951) relat ed with them. The latter have distinctly fewer lamellae on the carapac e and their valves are subject to far more important variations. The decreased number of lamellae is correlated with a process of marked biological significance, i.e. the ability to shed moults which, in their ancestors (Aberroconcha n. gen . and Eridoconcha), formed a carapace morphologically greatly complicated. The species referred to the genus Cryptophyllus and the allied representatives of , the genera Milleratia Swartz and Schmidtella Ulrich are, taxonomically speaking, a separ ate gro up , but nevertheless related to the Eridoconchidae.

In what concerns the structure of carapace in Eridoconcha, it displays st rong resemblance to shells of certain more closely not

4Ac t a Pa l a e o n t o lo g ic a Nr I

50 FRANCISZEK AD A MC Z A K

Fig. 6. - Aberroconch a plicata n. sp.; sche matic drawing, showing a fragm en t of the righ t valve from whi ch the ear liest growth st a ges

(A- C) hav e been removed

D-l growth stages, 1 dorsal lamella, 2 hinge, 3 inter lamella r cham ber, 4 chitino us lamella , 5 dista l part of la- mella, 6 concentric groove, 7 free ed g e.

E EN_

O

- ---1 ---2

----5

,

,6

,

7

--)""

...

,

'

....

..~,','/,'

.""

,'II'

.1'11 Ifll,

,,",

,II"

J :: : ,

til ' llt l

, ," ,,'

, II '

I II '

.

' "

.

. ^,.

: \ 3

--t-~ --_....

,': _{r:: ':} 4

,,

H--- D ---

F- -- E ---- -

O ,I5mm

I I

Fig. 7. - Concho strac a gen. & sp . in de t ., tra nsverse section of left val ve, partl y reconst ruct ed (b roken lines) ; Upper Silurian of Podoli a .

ERIDOSTRACA - A NEW SUBORDER OF OSTRACODS 51

det er minate Co nc host r aca from the Up per Silurian of P od olia, collected by Ko zlowsk i from the Cz or t k 6w fo r m ation, which h as yielded th e w ell known brachiopod fauna with Mutationella podolica Kozl. (Ko- zlowski, 1929). These shells, unfortunately strongly damaged, were built exclusively of calcium carbonate. In cross secti on (fig. 7) w e m ay observe their lamellae and thickened growth bands. The dorsal m ar gin area forms a strongly extended surface, . resembling analogous structures in living Conchostraca. The number of lamellae was up to 13. Th e most noteworthy fact is that so far Conchostraca with calcareous shell s and concentric growth bands have not been recorded from Silurian rocks. Devonian forms which, in the opinion of many authors (Raymond, 1946; Novo shilov, 1953, 1960), represent t h e earliest type of these crustaceans, have a c h itinous car apace. R epresen tatives of this group of animal s found in Siluri an bed s r ea sonably suggest th at it s hi st ory lasts ov er a far greater sp an of time than h as h eretofore be en suppos ed.

is not excl ude d that they c ons titute a link b et w een th e Cam br ian repre sen tat ives of Lepidittidae Kobayashi and t h e D evoni an f or m s. Nei t her are t hey referable to the Eridoconchidae, f or th ey disp lay a nu mber of fundamen t al di ff er ences from that grou p, such as: lar ge nu mbe r of lam ell ae (13), development of the hinge m argin and d imensio ns greater t han those in t ypical Eridoconchidae.

Observations made during the study on Eridoconchidae reveal th at:

1) their carapace was a multilamellar (up to 11) structure reflecting the growth stages of the valve, and 2) that the calc ifi cation of l amell ae was gradually realized during the evolutionary process of this group of animals, while in their ancestors the lamellae were most likely built of chitin.

Differen ces and affinities of shell in Eridoconchidae and livin g Conc h ostraca

Sc h m idt (1941) w as the fi rst to call attention to the fact that t he carap ace of the genus Eridocon ch a greatly r esem bles th at of Con cho- st r ac a. He n ce, he r eferred them to on e of the famili es (Cycl estheriidae Sar s) of that gr oup of animals. Novoshilov (1953), an outst andin g author it y on fossil P h y ll op oda, thinks the genus E rid oconcha t o be a r epr esentative of the ostracods, postulating that the concen t ric or nam en tation is not a diagnostic taxonomic criterion of fos sil Con- chost r aca (1953, p. 6). Moreover, in all the works available to the writer, the genus Eri docon cha is invariably assigned to the ostracods. Actually, h ow ev e r , in what the structure of carapace is concerned, this group shows mar k ed sim ilarities with Conchostraca. With a view of cl earing up th ese rel ations, th e write r has investigated the carapace of certain living Con cho st r aca . In the light thrown by these studi es the multila-

4*

52 FRANCISZEK ADAMCZAK

mellar st r u cture of Eridoconchidae and the prob lem of the "retentio n of mou lts" (Levinson, 1951) seems to be soundly clarified.

The car ap ace of th e Conchostraca is chitinous and is but to a small ex ten t sub ject to calcification. Neverthel ess, as it has been afore men- tioned , in cer tain fossil f orms from the marine Silurian d eposits of P odolia, th e carapace was complete ly calcified (fig. 7) . Probably, in conne ct ion wit h t he ir gr adu al mi gration from salt to brackish an d fresh water basins, these fo r m s lost the ability to accumulate carbonate salts, whose concen t ra tion p ercentag e among organisms living in this enviro n me n t is markedly low . Th eir shells are made up of many chitinous lamell ae und erlaying one another. On th e out er surface of the car apace th ey form gr ow t h bands separated by concentric groo v es. Grube (1865) has studied the car ap aci al m o rphology of th ese anim als in gre at detail an d the com para t iv e studies here d escr ib ed have bee n based , amon g others , on that author's data.

The shell of the living genus Estheria R upp el is easily etched after being for som e hou rs immersed in KOH (Grube, 1865), wher e it becom es swollen . Subsequ ently, therefore, the successive lam ellae of the carapace m ay be easily separ at ed on e from an other. Under the microscope the se la m ell ae r ev eal a ch aract eristic m esh-like ornamentation, by Grube re fe r re d to as " Net zwerk" .

also occurs on the bands of growth, but t he r e

is obscured by additional orname nt ation of this part. In fossil fo rms, in w h ich th e ornamentation pattern on concentric bands is subject t o strong variations, tHis reticulafi on has, among others, been enc ou n tered in Gl y ptoasmussid ae Novo shilov (Novoshil ov, 1960) . Tr aces of it hav e by the writ er been obse rve d also in Eridoconcha granulifera n . sp. (comp . p. 46). Moreover, this re tic u la ted pattern likewise occurs in living ostracods of the s pecies Cythereis pra va Bai rd (Muller, 1894, p. 97-98). In the la st named species the network consists of ch it ino u s rods, often in vading the calcareous lamella, while in o t her forms they adhere to them on the inner side (Bairdia McC oy) .

s eems that so me ves t iges of a na logo us struct ure s may be obse r ve d in ot her fossil forms of the gro up Palaeoco pa , e .g. in Ko zlow. skieHa (Pribyl) (Adamczak , 1958), H ibbardia Ke sling (Ke sling , 1957a) and in Tra ch yleberis tu bulosa Triebel (Cyth eridae Baird), (Tri eb el

Klinger, 1959).

Th e re ticu lated ornamentation on the surface of chitinous carapace

has like wi se b een not ed in Cambrian Archaeostraca (Hipponicharion

Matth ew).

r esembles th e orname ntation of Glyptoasmussidae . In

C ambrian for m s th e st r uct u re of this t ype of ornam entation su gg ests

that it re - in fo rced t he cara pace, contributing to its elasticity and

res istance to deforma tion . H ow far these structures are homologous,

still r emains an open qu estion . Is the "N etzwerk" a vestige of the

pri m ar y ch itinous cara pac e, or is it a convergent structure produc ed

ERIDOSTRACA - A NEW SU BORDE R OF OSTRA COD S 53

independently by the various crustacean groups? The solution of this problem may be of great phylogenetic significance, sinc e certain C ambrian crustaceans (Bradoriidae Matthew and Beyrichionid ae Ulrich

Bassler) have been su pposed to be th e ancesto r s of ost r acods (Ulrich

Bassler, 1931).

Comparative studies on the Er idoc on ch id ae an d Con chostraca, and still more so m orphological observations of th e r epresentatives of

A

/ / /y / / / / / / /

~

B //////////

c 11~

Fig. 8. - A Eridoconcha rugosa Ulrich ^& Bassler; sche mat ic drawing, show ing pores on partly decalcified sh ells at the boundary of the gro wt h band s; B Estheri a sp.,schemat ic drawing, show ing presen ce of pores at the boundary of gro wt h bands

in a recent form; C Estheria sp., fragme nt of ch it ino us mantle with se t ae.

Aberroconcha and Eridoconcha , lead to the conclusion that these t w o groups are related. This is indicated by the follo wing ch aracters : 1) multilamellar c arapace, 2) minute pits of the gro wth bands (E. rugosa) (fig. 8), and 3) r eticul ation-i.Net zw erk" , w hich may b e traced in Eridoconcha (E. granulata n . sp .).

The multilamellar ca ra p ace, common to both th ese grou ps, is the

essent ial feature su ggesting the relationship between Eridoconchida e

an d Conchostraca. Though in the opinion of most author s the Concho-

54 ^{FRANCISZEK ADAMCZAK}

straca did not occur bef or e the De vonian (comp . Raymond, 1946;

Novoshilov , 1954), forms with con cent r ic growt h lines of th e con ch o- stracoid type occur al r eady in the Cambri an (FordiUa Walc ot t and Lepidita Mat t hew) . Acc ording to Raymond (1946, p. 304), " ... bot h sh ow grow t h - li n es, an d, technically I d o not see any way of excluding them fro m t he L ioesth eriidae. Possibly th ey are ancest or s of the Con ch ost r ac a, but until Up per Cambrian, O rdovician, Silurian and Lo w er D evo ni an con n ectin g links are found, it is probably wise to leave t h em in their pr esent uncer tain position." Th e prese nt writer does not beli eve their position so un cer t ain as is thought by R aymond, since the hia tus bet w een the U pper Camb ri an an d th e L ow er Devonian is partly fill ed up b y a specim en of Conchostraca, foun d by the writer in the Siluri an (comp.

p. 51), an d, to a ce r t a in ex tent, by the Ordovician E ri docon chidae.

R et ent ion or in itiation of m ou lt s?

Mu lt il am ell a r carapaces must be con sidered as quite exceptional amon g ostracods, for as a rule the sh ell of these animal s consists of one lamell a on ly . Levinson (1951) was t h e fi r st thoroug hly to investigate t he m u ltilam ellar carapace. Independen t ly of t hat author , this problem has also been studied by Egorov (1954).

Levin son, w ho was t he fi r st to take up the problem of multilamell ar carapaces , asser ts th at : "Bot h of the ge ne r a Eridocon ch a an d the new gen us Cryp t oph yllu s exh ibit a man y-l ayered shell which a ppears t o be for med by t h e r et ention of molts .

is pr oposed that th e cohesion of these m olt s is not depen dent on t he car a p ace b u t it is a speciali sed ad apt ation of ge ne r ic impo r t ance" (Levinson, 1951, p . 553). As is stated by L evinson himself, t he h ypothesis of the re ten tion of m oults had bee n adva nced in earlier papers b y Swain an d Bolli n, and by Drake, in 1940 (Levin son , 1951). H enningsm oen

also con cern ed himself w ith that pro b le m and succeeded to discov er analogous occur r en ces in representatives of th e genus C on eh oprimit ia Opik and Eet odemit es plummeri Coop er. Of some int er est is al so a report by Egorov (1954) on the multilamellar she lls in for ms of t h e "Astarte " sociali s Eichwal d gro up . These used t o be assigned a mo ng the pel ecypod s, but Egorov proved that they actua lly belong to Erid oconcha. He stat es that: "In sections, thin sections and samples it may be obse rved th at the shells of Astarte socialis Eich w . consist of several l aye r s whose n umbers cor respond to the num be rs of growt h lin es" (I.c., p. 25).

In t h e light of L evin son 's stu dies (1951) th e " r et ent ion " is a distinc t ly adaptative fea ture, rest ricted t o a small gr ou p of forms.

Wh en viewed from this standpoi nt, th is process acq u ires p eculiar

characters, sin ce the p urpose of the ad apt at ion it self is not clear. We

do no t k now w hat factors i nflu en ced the ani mals to ca r r y t h e burden

ERIDOSTRACA - A NEW SUBORDER OF OSTRACOD S 55

of the moulti n gs an d su bs eq u ent ly, with the lapse of time , to lose this ability, for , as is me ntioned by L evinson on p. 557: "The evolution of t h is group is tow ard a d ecrease in th e number of specimens consisting of only on e or two m olts".

Ac cordin g to Levin son , the problem of the r et ention of moults suggests, t oo, that t hese multilamellar forms descended from ancestors with a unilam ell ar c ar apa c e. Such a conception explains w ell that author's st atement that the carapace of the genera E ridoconcha and CryptophyHus " ... is a specialized adaptation of generic importance".

Our knowledge of these ancestors is very inadequate. The collected materials and observations, however, permit the problem of " t h e reten t ion of moults" to be pu t in a different light. In Eridoconcha and 'A berrocon cha the sh ells consisted of many lamellae (7-11) and it is reasonable t o suppose th at this w as the original condition of the s t ru ct u r e of the car apace , inherited fr om ancestors of the Conchostraca group. The r em ain in g forms with fe wer lamellae (6-1), to which Cr y pt oph y llu s belongs fore most , referred by the present w ri t er to Eridostraca, too (comp. p. 71 - 72), r eali ze a new structural t y pe of the carapace.

seems h ig hly probable that sh ell s in r ep r esentatives of this gen us do not, as has b ee n suggested by Levinson (1951), illu strate

" re tent ion" , but t he initiation of m oults (fig . 9). In this connection we m ay infe r that the r ed uced number ' of lamellae in sh ells of these a nimals (the maximum numbe r of lamell a e in Cryptophyllus being 6) is to be regarded as a proterogene ti c character (Schindewolf, 1927, 1950).

Speci es of CryptophyHus m ay be consid er ed as a link in the morphol o-

gical chain uniting forms of th e conc hostracoid a l t ype (Erid oconc hidae)

with the typical representativ es of ostracods such as MiHerat ia and

Sc hm idt ella . Ob serv at ions thus far made suggest that represen t ati v es

of Con ch oprimit ia Opik , ' w h ich had made their appearance as early as

in t he Ordovician, evolved acc or ding to a sim il ar pa ttern as Cr y pt ophyl-

Iu s. This is indicated b y traces of ea r lier moult st ag es, th e simpl e

structu r e of the carapa ce, the m edial sulc us and a kind of velar stru ct u re

w it hin t h e f ree edge of valve (fig. 11

A number of species of this

gen us (Conc hoprimitia gammae Opik , C. deminuat a Opik, C . t allin ne n sis

Opik , et alii) hav e coh erent valves of moult stages (Henningsmo en,

J aanu sson , 1957;Sarv, 1959). In this ca se, too , w e may speak

a bout m oult s ta ges fixed on the carap ace of mature forms as of

a prote rogenetic p he nome no n . Th ese animals had acquired the ability for

nearly comple te m oulting which in our considerations h ere is regar de d

as a t ax ono m ic and phylogeneti c factor of . great impor tanc e. Th ey

probably descend from fo rms of the Eridoconchidae-type. This group

though provided with many primitive features , e .g. partly calc ifie d

l am ellae (A. magn a, A. plic ata n. sp .), displays m an y pro gr essiv e fea t ures