Coelacanth Fishes From Madagascar - 1935 Woodward

Coelacanth Fishes from Madagascar. 213

1932. WRIGHT, J. " The Scottish Species of Allagecrinus," GEOL. MAO.,LXIX, 337.

1933. " Two New Crinoids from the Scottish Carboniferous Limestones,with Notes on the AHagecrinidae," GEOL. MAG., LXX, 193.

1934. " New Scottish and Irish Fossil Crinoids," GEOL. MAG., LXXI, 241.1934. " Note on the Occurrence of Blastoids with brachioles at Hook

Head, Co. Wexford, Ireland," GEOL. MAG., LXXI, 267.1913. ZITTEL-EASTMAN. Text Book of Palaeontology, 2nd ed.CORRECTION TO WRIGHT, 1934, p. 253. In Diagnosis of Synerocrinus (?) smithi

sp. nov., for " IBr one to three ", read " IBr two, iBr one to three ".

VII. EXPLANATION OF PLATES.

PLATE VII.

FIGS. 1-8.—Edapocrinus rugosus gen. et sp. nov. Figs. 1 and 5, ventralsurface of cup, showing orals and " roof " plates of anal tube. Figs.2 and 7, side views of the cup from posterior, showing anal tuberesting on posterior B. Figs. 3 and 8, side views of cup from anterior.Fig. 4, another side view showing surface o; namentation, left posteriorRon right. Fig. 6, view from below. Figs. 9, 10, 11, 13, 14, and 15,Cyathocrinus patulosus sp. nov. Figs. 9 and 11, anterior and posteriorviews of cups. Fig. 10, cup from below. Fig. 13, posterior viewshowing part of anal tube in position. Figs. 14 and 15, ventralviews of cup shown on Fig. 13. Fig. 12, a typical cup of Cyathocrinusconicus Phillips. Figs. 1, 2, 3, 4, and 15 are x 2, the others are naturalsize. All from Coplow.

PLATE VIII.Cyathocrinus patulosus sp. nov. J. Wright Collection No 2261 from Coplow.

Natural size.PLATE IX.

Pachyhcrinus aff. kmgidactylus (Austin). J. Wright Collection No. 2260,from Coplow. Natural size.

The Coelacanth Fishes from Madagascar.By J. A. MOY-THOMAS, M.A., University Museum, Oxford.

INTRODUCTION.

T7ERY little has been written about the Coelacanths from^ Madagascar, although those from similar formations in East

Greenland and Spitsbergen are about the best known members ofthis group. Smith Woodward (1910) described and figured a singlespecimen as Coelacanthus madagascariensis, distinguished from othermembers of the genus by the ornament of closely set tubercleson the operculum. Further specimens were figured by Priem (1924)and attributed to C. madagascariensis, but no attempt at anydetailed description was made. Dr. E. I. White, of the BritishMuseum (Natural History), very kindly placed at my disposal acollection of these fish, which he had obtained himself. The materialexamined during this investigation consisted of these specimens,twenty-eight in number, and the original specimen described bySmith Woodward. The specimens were nearly all external moulds(although in some a certain amount of bone still remained) in

214 J. A. Moy-Thomas—

ferruginous clay nodules. Very little preparation was necessaryexcept washing. This treatment produced wonderful results insome cases and almost perfect impressions of the head were obtained.So perfect were these casts in many cases that the pores of the lateralline canal were left as small hummocks making parts of this systemeasy to identify.

The fairly complete anatomy which it has been possible to describeat once reveals that the specimens collected by Dr. White belongto a new genus of Coelacanth which I have named Whiteia in hishonour. This genus is readily divisible into two easily distinguishable

ds stfr

fr

Pr

pop

sp

Flo. 1.—Whiteia. Restoration of the head and lateral line canal. Lateralview. The lateral line canal is restored from the position of the pores. X 2.

species. The original specimen of Smith Woodward is, however,quite distinct and clearly belongs to the same genus as Coelacanthuselegans, Newb. from the Carboniferous. There is some doubtwhether the latter really belongs to the genus Coelacanthus, but inthe absence of a better description of the type species C. granulatus,Ag., it is provisionally placed in this genus.

Before proceeding further I would like to take this opportunityof expressing my gratitude to Dr. White for lending me this materialand for valuable help during this work, and also to Professor Good-rich for reading this manuscript.


Genus WHITEIA, gen. nov. (Text-figs. 1 and 2).Genotype. Whiteia woodwardi, gen. et sp. nov.

Diagnosis.—-Medium sized or small slender Coelacanths. Headrelatively large. Skull with triangular operculars and preoperculars,quadrato-jugals, squamosals, postorbitals, lacrimo-jugals, supra-temporals, parietals, five extra scapular (post-parietal) plates andthe angular ornamented with numerous tubercles. The coronoidsare dumb-bell shaped and unornamented. The dermosphenotics(post frontals) are separate from the frontals, and the supraorbitalseries are pierced by large lateral line pores. Circum-orbital ring

FIG. 2.— Whiteia. Restoration of the skull. Dorsal view, x 2.

well developed. Parasphenoid rather broad. Pterygoids with wideanterior limb and apparantly no notch in their dorsal margin. Allthe lepidotrichia of the anterior dorsal fin, and at least the mostanterior of the dorsal and ventral lobes of the caudal fin ornamentedwith two rows of denticles. Pelvic fins situated behind the anteriordorsal fin about half-way between the two dorsal fins. Anal finbehind the posterior dorsal fin. Scales ornamented with tubercles,elongated tubercles, short ridges, or ridges stretching the whole


length of the exposed area of the scales. This ornament runsapproximately antero-posteriorly.

WHITEIA WOODWARDI, gen. et sp. nov. (Text-figs. 3-6).Coelacanthus madagascariensis Priem (1924). x

Holotype.—P172OO and its counterpart P17201 in the BritishMuseum (Natural History) (Text-fig. 3).

Paratypes.—P17204-P17209, P17212-P17213, P17167-P17169,P16236, Nl, N47, N82, N113, K269, K271, B121, B122, B125,B126, B143.

Fio. 3.—Whiteia woodwardi. Outline drawing of the type specimen P17200(B.M.H.N.). The scales have been omitted from the drawing, f natural size.

Formation and Locality.—Lower Trias ; Anaborano, Bobasatrana,Ambarakaraka.Measurements of the Type.—

c m s .4 - 36 1

4 14 - 94 - 2

L e n g t h o f h e a d . . . . . . . . . .L e n g t h f r o m r o s t r a l e n d o f h e a d t o first l e p i d o t r i c h o f a n t e r i o r d o r s a l finL e n g t h f r o m first l e p i d o t r i c h o f a n t e r i o r d o r s a l fin t o first l e p i d o t r i c h

o f p o s t e r i o r d o r s a l fin .B r e a d t h a t first l e p i d o t r i c h o f a n t e r i a l d o r s a l fin . . . .B r e a d t h a t first l e p i d o t r i c h o f p o s t e r i o r d o r s a l fin . . . .

Description.The Skull (Text-figs. 1 and 2).—The endocranium was never

preserved, but in some specimens moulds of the basisphenoid were1 Priem's figures leave no doubt that he is describing this species.


present and by means of casts it was possible to show that thisregion is similar to that of other Triassic forms, and has a largeantotic process only, the basipterygoid process being absent. Theparasphenoid is broad anteriorly and covered by numerous smallteeth. Two small prevomers lie anteriorly to the parasphenoid,bearing two or three teeth, which are larger than those on theparasphenoid. The palatines are oval and covered with minuteteeth, as also are the pterygoids. The latter have relatively broadanterior and posterior limbs, but no notch on the dorsal marginwas observed. The metapterygoid and quadrate are both stronglyossified, neither of them probably being very much fused to thepterygoid (Text-fig. 4).

mt

4.—Whiteia woodwardi. The pterygoquadrate. Median view. X 2.

The supratemporals are separate from the parietals, havingwell-marked lateral line canals with small pores, and an antero-ventrally directed process connecting them with the prootics.There are five extrascapulars which probably were not very stronglyattached to one another laterally. Between the parietals andthe premaxillae are four paired ossifications. I believe that the twoposterior pairs represent subdivided frontals. Their sutures differmarkedly from those separating the more anterior pairs of bones,which are more or less straight from side to side. Those separatingthe posterior pair, on the other hand, end pointedly in the pairbehind. The anterior ends of the most posterior pair, therefore,end in a median point between the backwardly projecting part ofthe bones in front, and the general effect is of a single bone havingbeen subdivided secondarily. A similar subdivision of the frontalsprobably occurs in Macropoma (Woodward, 1909). There are twopairs of postrostrals anterior to the frontals. The dermosphenoticsare separate from the frontals, and the supraorbital series of bonesexpands ventrally in front of the orbit to form a series of antorbitals.


These antorbitals and supraorbitals are all pierced dorsally by largelateral line pores. The " premaxilla " is apparently toothless andconnected to the lacrimo-jugal by a small bone with three largelateral line pores. A similar bone occurs in Macropoma (Watson,1921). The lacrimo-jugal is well developed and extends upwardsbehind the orbit to meet the postorbital, both bones being piercedby the sensory canal. The sensory canal is continued backwardsfrom the lacrimo-jugal into the squamosal in which it passes postero-ventrally presumably into the preopercular, but no trace of it hasbeen identified in this bone. Between the preopercular and thelacrimo-jugal lies a well-developed quadrato-jugal. The opercularis a triangular bone and between it and the postorbital lies a smallbone, the suprasquamosal. A similar bone is present in Sassenia(Stensio, 1921), in the Carboniferous Coelacanthus elegans (Text-fig. 10), and also in Coelacanthus madagascariensis (Text-fig. 9).All the bones so far described are covered with an ornament ofclosely set tubercles similar to those figured by Woodward (pi. 1,fig. 5a) but slightly smaller. The bones of the cheek formed a com-plete covering, and were not loose in the skin, as in many of the formsfrom Spitsbergen.

Only four bones of the lower jaw, the angular, splenial, dentary,and coronoid have been identified. The angular and splenial havelarge lateral line pores, and the former are ornamented in a mannersimilar to the other head bones. The coronoid is very character-istically dumb-bell shaped, and it is unornamented. I have beenunable to demonstrate any teeth except a few very small ones onthe dentary in a single specimen.

There are between twenty and twenty-five well ossified platesaround the eye, which are without any ornament. These are usuallytermed sclerotic plates, but must be distinguished from the latteras circum-orbital plates (see below).

Although there are traces of the branchial arches and urohyal,nothing can be made of their structure. The ornament of thegular plates is not shown in any of the material at my disposal.

Paired Fins and Girdles.—The pectoral fins lie very slightly infront of or below the anterior dorsal fin, and have between nineteenand twenty lepidotrichia, which are jointed distally for about two-thirds of their length and without denticles. The pectoral girdleconsists of a clavicle ornamented with tubercles as on the headbones, a cleithrum with a backwardly directed process for articula-tion of the internal skeleton of the fin, and a small supracleithrum.

The pelvic fins are situated about half-way between the anteriorand posterior dorsal fins. They are slightly larger than the pectoral,have about eighteen lepidotrichia jointed distally for about halftheir length, and bear no denticles. The pelvic girdle (Text-fig. 5)is very similar to that of Carboniferous Coelacanths, and consistsof paired plates flanked by two ridges, and a lateral process forarticulation of the internal skeleton of the fin, and a median process


for articulation with a similar process from the girdle of theother side.

Unpaired Fins.—The anterior dorsal fin is the largest of themedian fins consisting of from eight to nine stout lepidotrichia.The first lepidotrich is short and unjointed, but the distal halvesof the remainder are much segmented. Each lepidotrich of thisfin bears two rows of small sharply pointed denticles. The basalplate is nowhere preserved, but the impression which is frequentlyto be seen at the base of the fin suggests that it is of the kind usualin Coelacanths. It can be seen not to be forked at its proximaledge.

The posterior dorsal fin has fifteen lepidotrichia less stronglydeveloped than those of the anterior dorsal fin. The lepidotrichiaare jointed distally for about half their length, except those of thetwo most anterior which are jointed. None of the lepidotrichia bearsdenticles. The internal skeleton lies about half-way between theanterior and posterior dorsal fins, and consists of a bifurcated basal

FIQ. 5.—Whiteia looodwardi. Pelvic girdle. X 2J.

plate, one limb of which is directed anteriorly, the other antero-ventrally between the neural spines.

The anal fin consists of from fifteen to sixteen lepidotrichia jointedthroughout. These lepidotrichia are unornamented, and there isno trace of any internal skeleton of this fin.

The caudal fin consists of the usual three lobes, although onlythe most anterior part of the supplementary caudal fin is preservedin any of the specimens. The dorsal lobe of the caudal fin has fromthirteen to fifteen lepidotrichia, and the ventral lobe has from elevento thirteen lepidotrichia. The ventral lobe starts slightly behindthe dorsal. The lepidotrichia, except the anterior dorsal andanterior ventral, are segmented distally for about half their length,and at least the first two anterior dorsal and ventral are ornamentedwith rows of denticles similar to those on the anterior dorsal fin.

Axial Skeleton.—Very little of the axial skeleton is preserved andthere is nothing which merits special attention. No ribs orhypocentra are preserved.

Swim-bladder.—The swim-bladder is not visible in any of thismaterial.


Squamation.—The scales are large and oval, the exposed portionwider than long, and ornamented with elongated tubercles mixedwith ridges running antero-posteriorly. Some of these tuberclesare hardly elongated at all, whilst others extend over the entireexposed areas (Text-fig. 6).

This species has been named in honour of Sir Arthur SmithWoodward.

Whiteia tuberculata, sp. nov (Text-figs. 7 and 8).Holotype.—P17214 and counterpart P17215 in the British Museum

(Natural History).Paratype.—N54.Formation and Locality.—Lower Triassic ; Anaborano.

FIG. 6.—Whiteia woodwardi. Scales, x 4.

Measurements of the Type.—cms.

Length from rost ra l end of head to beginning of supplementarycaudal fin . . . . . . . . . . 1 1 - 5

Leng th of head 3-5Leng th from rostral end of head to first lepidotrich of an ter ior

dorsal fin . . . . . . . . . 4 - 5Leng th from first lepidotr ich of anter ior dorsal fin to first lepidotr ich

of posterior dorsal fin . . . . . . . . 3Leng th from ros t rum to first lepidotr ich of dorsal caudal fin . . 9 -5Leng th from ros t rum to first lepidotr ich of vent ra l caudal fin . 10 • 1B r e a d t h a t front of an ter ior dorsal fin . . . . . 3B r e a d t h a t front of posterior dorsal fin . . . . . 2 - 7

Description.—This species differs only from Whiteia woodwardiin the ornament of the bones of the head and scales, and shape ofthe latter.


The ornament of the operculum and cheek bones consists ofsmall tubercles, which are relatively smaller and farther apartthan in the foregoing species. The ornament of the scales is verysimilar to that of the bones of the head, and consists of a fewscattered tubercles which are hardly elongated at all. The exposedportion of the scale is as broad as it is long (Text-fig. 8).

This species is named on account of the characteristic ornamenta-tion of the scales.

EEMARKS ON THE GENUS.

Whiteia is of particular interest as the dermal bones of the skullare so very completely known. A separate preopercular has never

pdf

anf

FIG. 7.— Whiteia tuberculata. Outline drawing of the type specimen P17214(B.M.N.H.). The scales have been omitted from the drawing, f natural size.

previously been described in Coelacanths, although it occurs in theCarboniferous form Codacanthus elegans, Newb. (Text-fig. 10) andin Coelacanthus madagascariensis (Text-fig. 9). The cheek posteriorto the orbit in this typical Carboniferous species resembles those ofWhiteia in having a postorbital, suprasquamosal, lacrimo-jugal,squamosal, and quadrato-jugal as well as a preopercular. Thechief difference is the relatively large size of the squamosal inC. elegans which excludes the quadrato-jugal from contact withthe lacrimo-jugal. Stensio (1921) has suggested that the supra-squamosal represents a piece of the squamosal; if this is so, thepresence of a preopercular produces an arrangement of bonesexactly similar to those of Osteolepis (Save-Soderbergh, 1934).Unfortunately the region anterior to the orbit is only very incom-pletely known in most other Coelacanths. In Macropoma (Watson,


1921), however, the general plan of structure is the same as inWhiteia, the supraorbital series coming in contact with the anteriorend of the lacrimo-jugal, and the latter connecting with the " pre-maxilla " by a small bone.

Unfortunately many of the Triassic genera are poorly describedor erected on very incomplete remains, and therefore it is verydifficult to be sure of the genus of any new material. Stensio (1932)has summarized those characters which are certain in the nine chiefTriassic genera Graphiurus, Diplurus, Heptanema, Wimania, Axelia,Sassenia, Scleracanthus, Mylacanthus, and Laugia. In addition tothe above genera specimens from the Trias have been attributed toCoelacanthus, Vndina, and Macropoma, but probably wrongly. Thespecies from Madagascar described above clearly do not belong tothe genus Coelacanthus, which has lepidotrichia without denticles.Of the remaining genera Graphiurus may be distinguished by theexpanded lepidotrichia and peculiar caudal fin. The position ofthe pelvic fins and bones of the head are very different from those of

FIG. 8.—Whiteia tuberculata. Scales, x 4.

Laugia. The dermal bones of the cheek are quite different fromthose of Wimania and Axelia. The lobate or spinous margin of theopercula of Mylacanthus readily distinguishes this genus. Dipluruscan be distinguished by possessing acute spines on the anteriorlepidotrichia of the paired fins as well as both dorsal fins.

Whiteia does, however, resemble Sassenia in many respects,particularly in the ornament of the bones of the head and presenceof a suprasquamosal. Stensio (1921) describes the bones of thecheek as being set loosely in the skin, apparently there is no pre-opercular and the coronoid appears to be triangular. Thesecharacters seem sufficient to distinguish this genus.

Scleracanthus and Heptanema are themselves very unsatisfactorilydistinguished from one another. However, the dorsal fin of boththese genera appears to have been more strongly developed and tohave had many more lepidotrichia. Scleracanthus has a forkedinternal skeleton to its first dorsal fin, a notch on the dorsal marginof the pterygoid, and the ornament of the head bones is strikinglydifferent. Although it may eventually be discovered that Whiteia


is really a synonym of one of these last three genera, which havebeen inadequately described, I have in the present state of know-ledge considered it necessary to erect a new genus.

Coelacanthus madagascariensis, Smith Woodward (1910) (Text-figs. 9, 11).

Type.—P10768 in counterpart in British Museum (NaturalHistory).

Formation and Locality.—There is some doubt as to the exacthorizon of this specimen. The fish resembles the Coal Measureforms so closely that I find it difficult to believe that it is from theLower Triassic, and not in reality Permo-carboniferous as it wasat first described. Andogozobe.

Measurements of Type.—•cms.

Length from rostral end of head to beginning of supplementary caudalfin 15

Length of head 3-7Length from rostral end of head to first lepidotrich of anterior dorsal

fin 7Length from first lepidotrich of anterior dorsal to first lepidotrich of

posterior dorsal fin . . . . . . . . . 4Length from rostral end of head to first lepidotrich of dorsal lobe of

the caudal fin . . . . . . . . . 13 ' 5Length from rostral end of head to first lepidotrich of ventral lobe of

caudal fin . . . . . . . . . . 1 4 - 2Breadth a t front of anterior dorsal fin . . . . . . 4 - 7Breadth a t front of posterior dorsal fin . . . . . . 3 - 6

Description.The Skull (Text-fig. 9).—The neurocranium is preserved only

as a cast of the basisphenoid which appears not to differ from otherCarboniferous forms. The parasphenoid and pterygoquadrate aretoo poorly preserved for description. The supratemporal is distinctfrom the parietals and there are no separate intertemporals. Thereare five extrascapulars which are in contact with one anotherlaterally. Posteriorly to these and anterior to the supracleithrumlies a single suprascapular on each side. The cheek region posteriorto the orbit is formed of six bones—a postorbital, suprasquamosal,squamosal, lacrimo-jugal, quadrato-jugal, and preoperculum. Thearrangement and shape of these bones resemble those of C. elegans(Text-fig. 10) so closely that there can be no doubt that theybelong to the same genus. The squamosal is large and lies betweenthe quadrato-jugal and lacrimo-jugal, and the quadrato-jugal doesnot extend forward to the lacrimo-jugal. The opercular, which istriangular, is ornamented with relatively large tubercles as figuredby Smith Woodward (pi. 1, fig. 5a). The ornament of the post-orbital, suprasquamosal, squamosal, and preopercular is similar


but the tubercles are smaller. The quadrato-jugal has in additionto tubercles many short ridges which run ventro-posteriorly parallelwith its anterior margin. The lacrimo-jugal runs forward in frontof the orbit, and there is a small more or less triangular plateornamented with tubercles directly above its anterior end.

The angular bone of the lower jaw is ornamented with elongatedtubercles running antero-ventrally parallel with the anterior upperborder. The coronoid is a large, somewhat oval bone. No ornamentis preserved on the gular plates.

P.a

po ex

dt

pop

FIG. 9.—Codacanihus madagascariensis. Skull of type specimen P10768(B.M.N.H.). x 2.

A few circum-orbital plates are preserved and they are ornamentedwith tubercles, as also are those of C. elegans. Since they areornamented they cannot be true sclerotic plates but must have beensuperficial and around the outside of the eye.

Paired Fins and Girdles.—The skeletons of the pectoral and pelvicfins are not well preserved and show nothing worth remarking;their lepidotrichia are too poorly shown to make counting possible.The pelvic girdle lies directly beneath the anterior dorsal fin, andthe lepidotrichia arise from a point which is beneath the mostposterior lepidotrich of this fin.


The Unpaired Fins.—Of the unpaired fins, only the anterior dorsaland part of the caudal are well preserved, the merest traces of theothers being present. The anterior dorsal fin has nine lepidotrichia,none of which bears denticles. A large part of these lepidotrichiais unjointed, but none of them is sufficiently complete to makecertain how much of its length is segmented. The internal skeletonis an oval basal plate, but is only fragmentarily preserved. Thecaudal fin has about twelve lepidotrichia both dorsally and ventrally,the latter starting slightly behind the former. The internal skeletonof the posterior dorsal fin is bifurcated and has an antero-ventrallyand a ventrally directed process, and lies about half-way betweenthe two dorsal fins.

P°

op

pop

ang

FIG. lO.—Coelacanthus elegans. Skull of P6286 (B.M.N.H.). N.B.—The pre-operculum is not complete, but is usually broader. X 2.

The Axial Skeleton.—The neural and caudal haemal arches areto a great extent preserved and are of the usual kind. No ossifiedribs or hypocentra are present.

The Svdm-bladder.—The swim-bladder is only preserved as animpression and it extends back to the beginning of the caudalhaemal arches.

Squamation.—The scales are very similar to Smith Woodward'sfigure (pi. 1, fig. 56), the ornament consisting of converging ridges.Further development of the specimen, however, revealed that theseconverging lines are always broken up posteriorly into tubercles,

VOL. LXXU.—NO. v. 15

226 Coelacanth Fishes from Madagascar.

whilst other scales have these striae themselves divided at intervals(Text-fig. 11). The exposed part of the scale is always longer thanbroad. These scales are very like those of many of the CarboniferousCoelacanths, both in shape and ornament.

Remarks.—The extraordinary resemblance of the head bones andscales of this form to the Coal Measure Coelacanths both in arrange-ment and ornamentation and the position of the pelvic fins makesit at once obvious that this species must belong to the same genusCoelacanthus, as originally suggested by Smith Woodward. Theirregularly arranged ornament of closely set tubercles on theopercular, which do not run parallel with the border, at once formsan easy distinction from other species at present attributed to thegenus Coelacanthus.

FIG. 11.—Coelacanthus madagascariensis. Scales from different regions ofthe body. The ridges are represented by the white, x 7.

EXPLANATION OP LETTERING OF TEXT-FIGURES.

adf = anterior dorsal fin.ang = angular.anf = anal fin.co = coronoid.cop = circum-orbital plates.ds = dermosphenotic.dt = dentary.ex = extrascapulars.fr = frontal.gu = gular.laju = lacrimo-jugal.11 = lateral line.mt = metapterygoid.op = opercular.pa = parietal.

pdf = posterior dorsal fin.po = postorbital.pop = preopercular.pr = posterior rostrals.pt = pterygoid.ptf = pectoral fin.pvf = pelvic fin.qj = quadrato-jugal.qu = quadrate.so = supraorbitals.sp == splenial.sq = squamosal.ss = suprascapular.ssq = suprasqiiamosal.st = supratemporal.

Basic Xenoliths in Plutonic Rocks. 227

BIBLIOGRAPHY.PRIEM, F., 1924. " Paleontologie de Madagascar. XII—Lea poissons fossiles,"

Annales de PaUontologie, xiii, 107.SAVE-SODERBERGH, G., 1933. " The dermal bones of the head and lateral line

system in Osteolepis macrolepidotus Ag.," Nova Ada Seg. Soc. Sci.Upsala, Ser. 4, ix, No. 2.

STENSIO, E. A., 1921. Triaasic Fishes from Spitsbergen, pt. 1, Vienna.1932. " Triassic Fishes from East Greenland," Medd. om Orenland,

lxxxiii, No. 3.WATSON, D. M. S., 1921. " On the Coelacanth fish," Ann. Mag. Nat. Hist.,

Ser. 9, viii, 320.WOODWARD, A. S., 1909. " The fossil fishes of the English Chalk," Palaeont.

Soc. Monog., pt. 5, London.1910 " On some Permo-Carboniferous fish from Madagascar," Ann.

Mag. Nat. Hist., Ser. 8, v, 1.

A Note on the Origin of Basic Xenoliths in PlutonicRocks, with special reference to their Grain-size.

By GEEMAINE A. JOPLIN, B.SC, Department of Mineralogy andPetrology, Cambridge.

1. INTRODUCTION.2. THE MECHANISM BY WHICH THE GBAIN-SIZE IS REDUCED.

(a) Recrystallization.(6) Hybridization.

(i) Formation of Poikilitic Crystals,(ii) Formation of Secondary Flakes and Granules.

3. THE ROLE OF HORNBLENDE.4. SUMMARY AND CONCLUSIONS.5. ACKNOWLEDGMENTS.6. REFERENCES.

1. INTRODUCTION.

T\UKLNG recent years much has been written on hybridization,•*^ and on the origin of basic xenoliths, but as far as the presentwriter is aware no comment is made upon the extremely fine grain-size of these inclusions as compared with their host.

Occasionally fringes of larger ferromagnesian crystals surroundthe xenolith, but these are not common.

Grantham (1928) has shown that the xenoliths in the Shap granitehave been derived from lavas, so there is nothing remarkable intheir finer grain-size. At Bibette Head, Nockolds (1932) has over-come the difficulty by postulating fragments of dolerite broughtup by the magma, but at Tregastel, Thomas and Campbell Smith(1932) have produced indisputable evidence of the xenoliths havingbeen derived from an earlier plutonic intrusion. They state thatthe grain-size of the basic plutonic rock is medium with flakes ofbiotite from 8 to 10 mm., and that the grain-size of the inclusionsaverages 0-25 mm.

Documents

Coelacanth Fishes From Madagascar - 1935 Woodward