^ LIBRARY

I II K

UNIVERSITY OF CALIFORNIA.

01 FT OF-

Received
Accessions No. ** '

Shelf No.

A

SMITHSONIAN CONTRIBUTIONS TO KNOWLEDGE.

489

ON THE

CONTENTS OF A BONE CAVE

ISLAND OF ANGUILLA

( WEST INDIES).

BY

EDWA11D D. COPE.

WASHINGTON:

SMITHSONIAN INSTITUTION.

1883.

COMMISSION

TO WHICH THIS PAPER HAS BEEN REFERRED.

JOSEPH LEIDY.

ADVERTISEMENT.

THE following memoir describes the fossil vertebrates, shells, and indications of
human occupation discovered during the excavation of a cave in the West Indian
island of Anguilla.

The remains were first obtained in 1868, and brief notices of them have been
made at various times since, but the publication of the full account was delayed in
the hope that other objects might be added to the collection. The death of the
gentleman who procured the specimens, and other causes having shown that no
further exploration was practicable, the memoir was prepared and submitted to
the Smithsonian Institution, for publication, in 1878.

The other works in progress prevented the publication until the present time,
but the interval has been taken advantage of by the author to revise the manu-
script and superintend the preparation of the plates.

The importance of the subject presented is shown by the following considera-
tions : —

First — It is the first investigation of the life of the cave age in the West Indies.

Second — It gives the first reliable indication of the period of submergence, and
hence of separation of the West Indian islands.

Third — It furnishes the first evidence as to the antiquity of man in the West
Indies.

Fourth— It describes some very peculiar forms of animal life not previously
known.

The illustrations have been made particularly full on account of the archaeolog-
ical interest attaching to those animals, which were probably the contemporaries
of the earliest men of tropical America, and in order to avoid the necessity of any
subsequent presentation of the same subject.

SPENCER F. BAIRD,
Secretary Smithsonian Institution.

WASHINGTON, April, 1883.

(iii)

ON THE

CONTENTS OF A BONE CAVE.

IN the year 1868, a quantity of cave earth, limestone fragments, and bone
breccia was brought to the port of Philadelphia by a vessel and deposited on the
lot of Henry Waters & 13rO., manufacturers of phosphatic manures, on the Schuyl-
kill. The material was imported for the purpose of ascertaining its value as a
fertilizer, especially by the determination of its richness in calcium phosphate. It
was obtained from a cave in the small Antillean island of Anguilla, which belongs
to Denmark. Through the attention of Mr. Waters, 1 learned of the existence of
fossil bones in the cargo, and proceeded to examine them. Remains of long bones
lying irregularly in a rather hard but cavernous red cave deposit of limestone were
found mingled with fragments of lighter limestone from the walls of the cave in
irregular masses, the whole being penetrated and mixed with a yellow stalagmitic
•deposit of arragonite.

From a block of the breccia I dressed three molar teeth, two partially complete,
and two much broken incisors, fragments of maxillary and pelvic bones, shafts of
various long bones, and the distal extremity of a femur with a patella. These
were the first evidences of the existence of the large rodent Amblyrliiza inundata,
which was described in the proceedings of the American Philosophical Society for
1868. Other bones were found in other breccia masses, which I could not clearly
refer to any other animal. With them occurred a shell of Turbo pica, Linn.

Having learned that Dr. E. van Hijgersma, colonial physician of the Danish
Island of Saint Martins, was interested in all departments of the natural sciences,
I wrote asking him to make an examination of the deposit in question, and to
secure, if possible, all fossils discovered in excavating it. He accordingly very
kindly went to Anguilla, and was rewarded by the possession of numerous addi-
tional teeth and bones of AmblyrJiiza. Subsequent visits added two species of this
genus, together with the bones of a species of ruminant near the genus Capra;
bones of a probable rodent of smaller size, of two species of birds, of a lizard, and
a shell chisel of human manufacture. These remains are described in the following
pages.

Unfortunately no notes were taken as to the relations of the parts of the cave
deposit, or whether any stratification is observable. We are, therefore, left to a
consideration of the appearance of the fossils themselves in estimating their proba-

1 April, 1883. ( 1 )

2 CONTENT SOFA BONE CAVE.

ble relative positions when found. The bones of the Ambtyrhiza inunclata first
found were of an earthy-brown color. Those contained in the next sending were,
some of them, of a brilliant vermilion color, and not encrusted. Others were of
the usual earthy-brown color, and more or less encrusted; others of this lot were
of a rusty red and not encrusted. The specimens of AmMyrhiza q-uad raits and
A. latidens, subsequently discovered, are of a brownish-red, and are all encrusted
with a thin calcareous layer. When this is removed their color is white. The
bones of the ruminant and the shell-chisel are of the same superficial color, but
were not encrusted.

As regards the mineral character of the bones, there is nothing to distinguish
them from each other. They absorb moisture, so that broken surfaces adhere
when brought into contact with the lips. A few specimens which do not display
this character are not to be distinguished as to species from those that do. So
there is no superficial character by which any distinction of age among these fossils
can be made.

In the following pages the species are described in detail.

EEPTILIA.

IGUANA, Linn.

The femur of a species of this genus was found at the time of the discovery of
the Amblyrhiza latidens. In form and character it very closely resembles existing
species of the genus.

The femur, which is that of the left side, has a gentle fore and aft sigmoid
flexure. The head is laterally compressed, and is directed upwards and inwards.
The trochanter is large, and forms a stout ala whose inferior extremity approaches
and then sinks abruptly into the shaft. It is not curved, but its superior angle is
thickened. It does not reach to the level of the angular margin which encircles
the head. A low swelling marks the place of the little trochanter. The shaft is
cylindric. The inferior extremity is transversely expanded, but there are no
epicondyles. The condyles are but little contracted at the middle, but the external
is much the larger. Its rotular face is entire, while that of the inner condyle has
a shallow fossa on its posterior face. They are not separated by a popliteal notch,
and have a very shallow trochlear groove anteriorly. There is a popliteal fossa.
There is a slight epicondylar tuberosity on the superior external side of each
condyle.

MEASUREMENTS.

If.

Totallength . . . . . . . . . .015

Long diameter of head

Diameter at apex of great trochanter . . • -015

of shaft .... • -007

, (fore and aft .... .006

of condylep «

(.transverse , . • .

CONTENTS OF A BONE CAVE.

AVES.

Two bones of birds were discovered by Dr. van Rijgersma, one the coracoid of
a larger and a humerus of a smaller species. The first is robust, but not shortened.
Both its extremities are wanting, so that its determination is rendered difficult.
It presents average proportions, not displaying the lateral expansion seen in the
petrels nor the slender form of many waders. It resembles the corresponding
portion of a Grus, and also the same part in some birds of prey. There is a well-
defined inward-looking superior groove which is pierced by a large pneumatic
foramen at a good distance from the margin. There is a corresponding superior
foramen just below the scapular fossa. There is a rather large pneumatic fissure
on the flat supero-internal face within the glenoid face.

The humerus above mentioned is that of a species of Procellaria in a restricted
sense. The bird was of rather small size, and the characters of its humerus are well
marked. Among these may be enumerated the prominent incurved posterior tube-
rosity of the head of the humerus, inclosing a deep fossa below it; the spur-shaped,
anteriorly directed, acute epicondylar process of the inner side of the distal extre-
mity. Within this spur is a strong coronoid fossa. The posterior border of the
external condyle is produced well backwards.

MEASUREMENTS.

M.

Length of humerus . . . ' . . . . .066

Width of proximal end . . . . . . .014

" of shaft ........ .004

" of distal end ....... .008

Antcroposterior diameter of inner condyle .... .005

" of external condyle . . . .007

Length of epicondylar spur ...... .004

MAMMALIA.

As already observed, the mammalian bones are those of Rodentia and Artiodac-
tyla. The determinable species are considered separately, but another of unknown
reference will be first noticed.

This animal is represented by proximal and distal portions of opposite humeri,
and both extremities of the femora, one of the latter being nearly entire. These
appear to have belonged to a rodent of the size of an aguti, and perhaps allied to
the group which that genus represents.

The head of the humerus is subround, and, with the proximal part of the shaft,
a good deal incurved. The greater tuberosity is large, its rounded apex elevated
above the head and rather prominent, but not incurved, in front. Its anterior
margin soon joins the gently convex exterior margin of the shaft without irregu-
larity, and the extero-posterior side of the shaft at this point is flat. The lesser
tuberosity is quite small and separated from the outer by a wide bicipital groove.

4 CONTENTS OP A BONE CAVE.

The distal end of the humerus is not expanded, having small epicondyles. The
trochlear face is strongly concave, ending in a deep olecnmon, but no coronoid
fossa. The bridge over the elongate arterial foramen is narrow.

The head of the femur is well bounded by a neck. The great trochanter rises
as higli as the head, and is deeply concave behind. The little trochanter is a
prominent tuberosity on the posterior face of the bone. The shaft is slender and
subcylindric. The condyles are somewhat compressed, and their surface is continu-
ous with that of the patellar groove. The latter is longer than wide, moderately
concave, with equal borders, which rise higher than the level of the shaft.

MEASUREMENTS.

If.

Diameter of head of humerus ...... .007

" of shaft of humerus ...... .008

of epicondyles of humerus ..... .016

" of head of femur ...... .009

" at little trochanter of femur . . . . .012

" of shaft of femur . .... .008

" of condyles of femur . . . . . . .017

AMBLYRHIZA, Cope.

Proceedings of the Philadelphia Academy, 1868, p. 313. Proceed. Amer. Philosoph. Society, 1869,
p. 183. Loxomylus, Cope, Proceed. Amer. Philos. Sou., 1869, p. 186.

The characters -of this genus are derived from the study of the remains of ten
or more individuals of three species. These include portions of all parts of the
skeleton with some important exceptions. There are wanting the superior, and parts
of the inferior walls of the skull ; the anterior foot, the distal parts of the pelvis,
the calcaneum, and the ungual phalanges.

Skull and Teeth. — A fragment preserved includes the occipital condylo, with
the mastoid and petrous bones. The condyloid foramen is pierced through the
rather thin basi-occipital. The mastoid region is produced downwards and back-
wards, but the specimen being broken does not indicate its length. The petrous
bone is in contact with the basi-occipital, closing the foramen lacerum posterius.
Its inferior wall is not expanded bulla-like, but is flat and thick. The mcatus
auditorius externus is continued by the thick tympanic bone, in tubular form
backwards and outwards. The apex of the petrous bone is truncate, and is deeply
notched for the foramen. The premaxillary bones and the symphysis mandibuli
are much produced and narrowed, and were probably inclosed in fur-bearing in-
tegument, as in the existing chinchillas. The mandibular rami are completely
coossified. They are united at their lower borders, posterior to the divergence of
their dental ridges, by the expansion of the stout rib of the inner face of each,
which incloses the incisive alveolus. This ridge extends posteriorly very far ; in
the A. latidens to a point behind the last molar and even posterior to the base of the
coronoid process. The latter is very short, and forms a flat process extending
obliquely outwards from the ramus. The condyle of the mandible is narrow, and

CONTENT SOFA BONE CAVE. 5

convex in the antero-posterior direction. There are some unattached hooked pro-
cesses, which I suspect to be angles of the mandible.

The dental formula is i. J-; c. -§• ; m. |. The incisors have a moderately thick
enamel layer, which is wrapped round the external angle a short distance.
Their sculpture is not deeply cut. The molars are composed of Vertical columns
of dentine inclosed in and separated by a sheath of enamel. The columns are
more or less transverse, and are neither confluent nor divided in any of the teeth.
They number four in the superior teeth, excepting in the last molar where there
are five. The entire tooth is inclosed in a thin layer of cement. The superior
molars are curved bow-shaped, the convexity being directed forwards at the middle
of their length. The enamel plates are then directed backwards on the grinding
faces. The extremities of the roots are simple and contracted to an obtuse termi-
nation. The inferior molars differ in their form, being straight and directed ob-
liquely forwards in the jaw. From this it results that their triturating surfaces
are oblique to the axis of the teeth, while those of the superior molars are trans-
verse to the axis of the middle portion of the shaft. There are but three columns
in all of the inferior molars.

Vertebra;. — The atlas has widely expanded transverse processes whose base is
about equally divided behind by the vertebral foramen. The vertebral canal soon
issues on the inferior face, and again entering the transverse process near its middle,
divides and issues in two foramina. One of these is on the superior aspect of
the base of the transverse process, and the one opposite to it on the side of the
neural canal. There is no hypapophysis nor tuberculum atlantis. The facets for
the axis are directed obliquely inwards.

The axis is rather short and the neural spine long and directed upwards. Its
section is triangular ; the odontoid process is short and stout, with subcylindric
section. Opposite articular extremity flat.

The lumbar centra are about as wide as long, and with nearly flat articular faces.
The sacral vertebra?, although adherent in several individuals, preserve the suture
in all of them. They are more elongate than the lumbars, and contract rapidly
in diameter ; the difference between the anterior and posterior dimensions of the
first being considerable. The centra of all behind the first are much depressed,
and the intervertebral foramina are large. The sacral diapophyses are greatly
expanded, especially anteriorly ; their iliac suture is a plane sloping inwards and
upwards. The only caudal vertebra preserved is short and wide, has stout diapo-
physes, and no facets for chevron bones. As there is no trace of neurapophyses on
this centrum, I infer that the tail is short.

Anterior Limb, — The scapula is represented by portions of those of several indi-
viduals. In these, the spine is well developed, rising rather abruptly from near
the neck. The coracoid process is a short, stout hook. I have not found a clavicle
among the bones.

The fore-limb is of smaller proportions than the posterior one. The greater
tuberosity of the humerus is crest-like, extending along one side of the head, but
not rising much above it. The lesser tuberosity is subround and very protuberant,
inclosing a deep and wide bicipital groove, with the extremity of the greater

6 CONTENTS OF A BONE CAVE.

tuberosity. The distal end of the humerus is moderately expanded in the trans-
verse direction. The external epicondyle is protuberant; the internal wanting.
The condyles have the ulnar and radial portions about equal, and the intertrochlear
ridge is represented by an obscure angle. The external or ulnar keel is prominent.
On the posterior face a rather short, median, concave, trochlear face represents the
two anterior faces.

Both ulna and radius are rather slender ; at the middle of their length the former
is "compressed, and the latter is depressed, both moderately. The olecranon is
prominent and subcylindric. The head of the radius is a rather wide oval, directed
downward and outward, while the coronoid process of the ulna is directed upward
and inward. The fore-foot is unknown.

Posterior Limb. — The section of the neck of the ilium is an equilateral triangle,
and is quite long ; the crest is not preserved, but from the form of the sacrum may
have had more extent than in many rodents. There is a strong tuberosity in the
position of the anterior inferior spine. The fossa ligamenti ten's and the incisura
acetaluli are strongly developed, the latter not perforate. The posterior bones of
the pelvis are unknown.

The femur has several marked peculiarities. One of these is the great develop-
ment of the great trochanter, which is really an undiminished continuation of the
shaft for some distance beyond the head. The head is relatively small, and is more
than half a sphere ; its ligamentous pit is distinct and isolated ; the neck is abruptly
contracted within the head, and projects at right angles to the shaft. The posterior
base of the great trochanter is deeply excavated. The little trochanter is a promi-
nent tuberosity. The condyles have a predominating inferior exposure. The
patellar groove is wide and not elevated. There is a deep fossa in the external
epicondylar region.

The head of the tibia is expanded laterally, and the upper part of the crest is
replaced by a plane surface, as in the corresponding bone in man. The spine is
prominent, a portion belonging to each of the articular faces, which are separated by
a deep groove. The upper portion of the shaft is deeply grooved on the external
face. The distal portion of the shaft is relatively slender. The distal extremity is
expanded inwards. The astragalar facets are oblique, the external is larger than
the internal, and they are well separated by an obtuse ridge. There are two pro-
cesses on the internal border, which are separated by a deep tendinous groove,
which is in most of the specimens bridged over by a lamina connecting the pro-
cesses. The posterior of the two processes is the most elongate : it corresponds to
a process of the astragalus, which extends backwards and inwards from the internal
trochlear face. When extension of the foot is attempted, these processes come in
contact, and prevent further movement. The amount of extension from the hori-
zontal which this arrangement permits, is 45°. When at this point the processes
constitute a support to the Aveight of the animal in addition to that furnished by
the usual astragalar facets. The fibular plane is triangular, and has a posterior as
well as exterior exposure.

The astragalus is quite depressed, and the convexity of its rotular surfaces is not
great. The latter are of unequal size, the external having four times the extent of

CONTENTS OF A BONE CAVE. 7

the internal, and they are well separated by the rotular groove, which is here formed
by the meeting of planes of different angles. There is little indication of fibular
surface on the external side of the astragalus, but the internal side of the inner
rotula bears an oblique .facet. This forms an angle of 91° with the rotula proper, /fc?'
which angle is directed upwards. Outwards and backwards from the base of this!
rotula extends the process already described. The head of the astragalus is short,
narrow, and defined by a neck. Its navicular face is subround, slightly convex, and
in the transverse plane. The head bears on its external side a facet for a bone,
perhaps the cuboid, as in the aguti. On the inferior face the external calcaneal
face is concave, and much larger than the internal, which is plane. They are com-
pletely separated by a groove which runs out on the postero-interior process
already mentioned.

Of the hinder foot there is preserved a mass which includes the navicular, ecto-
and mesocuneiform, and the second and third metatarsals. There is a loose fourth
metatarsal, which presents a proximal face for the fifth. The metatarsals are stout,
indicating a plantigrade foot. The navicular is not elongate, and has considerable
horizontal extent. A part only of the proximal surface is occupied by the concave
facet for the astragalus ; the remaining portion extends outwards and backwards,
and supports a couple of facets at its anterior border, one superior, the other ante-
rior. Corresponding faces on the navicular bone of various rodents are in contact
with the entocuneiform distally, and with an additional internal tarsal bone proxi-
mally, the internal navicular. These facets, therefore, indicate the existence of a
hallux, but perhaps of a rudimental one. The mesocuneiform is smaller than the
ectocuneiform, having much less transverse and little less longitudinal diameters.
Neither are produced posterior to the proximal ends of their corresponding meta-
tarsals. The third and fourth phalanges are about equal in length, and a little
exceed the second. Their distal extremities have a prominent inferior median
keel. At this articulation there is an ovate sesamoid bone on each side. A single
proximal phalange, probably belonging to the A. inundata, is very much depressed
at the distal articular extremity indicating a rudimental digit, doubtless the hallux,
similar to that found in several existing allied genera.

The hinder foot of Amblyrliiza then is rather short, furnished with four developed
digits and a rudimental hallux, and was probably plantigrade. The lack of tibial
crest indicates that the knee was not constantly maintained in a flexed position.
The immense trochanter indicates great power of extension of the femur, but whether
this extension was effective in running or kicking is uncertain. The absence of
tibial crest and the shortness of the foot militate against the supposition that these
animals possessed powers of leaping, and their swimming powers would be impaired
by the same structural characters.

Affinities. — This genus clearly enters Prof. Brandt's division of the Rodentia
which he terms the Hystricomorpha. The evidence is seen primarily in the free
fibula and in the development of the angular portion of the mandible on the
external side of the incisive alveolus. The small coronoid process and the generic
characters add to the weight of the evidence. Mr. E. R. Alston has recently
published a very valuable resume of the characters of the subdivisions of the

8 CONTENTS OF A BONE CAVE.

Rodentia, including the genera. He divides the Hystricomorplia into numerous
families, some of which at least appear to the writer to rest on rather slender bases.
In the comparison with Amblyrhiza, the Ilystricidce and Dasyproctidce may be
dismissed from the fact that their molars are not divided transversely by laminae
of enamel. The comparison is with the CJi inch ill idee and Caviida:. The molar
dentition is that of the former family, and the absence of a masseteric ridge sepa-
rates it from the genera arranged by Mr. Alston under the Caciidce, although I
cannot perceive that such a character should define a family group. The incisors
of both these groups are called by Mr. Alston "short." I have shown those of
AmblyrJiiza to be very long, as in the Dasyproctidce ; nevertheless, their transverse
section and sculpture are much as in the genus Lag-idium. The affinities of this
form are then near to types now existing on the South American Continent, but it
presents characters which show that it cannot be referred to any existing genus.
The extinct Arcliceomys, Laiz., Par., of the French Miocene, resembles it in the
constitution of its molar teeth, but whether in other respects or not I cannot
ascertain. There is one less dentinal column in the posterior superior molar of
ArchcKomys, so that the formula for those teeth reads -| in AmblyrJiiza and -*- in
Archceomys.

The only other extinct form with which it is necessary to make comparison is
the one called by D'Orbigny, Meyamys patagoniensis. This species rests on a tibia
found in Patagonia, and it is described and figured in the Voyage en Amerique
Meridionale, Vol. III. PI. VIII. figs. 4-8. Although the specimen was injured,
enough remains to show that it belongs to a different species and genus from any
of those herein enumerated. The differences are to be seen in the distal extremity.
They are, first, the absence of the divergent internal malleolar process, which is so
striking in AmblyrJiiza ; second, the large size of the internal cotylus of the tibia,
indicating that the internal rotula of the astragalus is as large as the external one,
instead of being much smaller. The deep fissure which separates the articular
faces of the head of the tibia is similar to that seen in AmUyrhista.

My friend, J. A. Allen, the distinguished zoologist, remarks in his article on the
Castormdidoe :' "To the same group (i. e., the Castoroididce) are, however, probably
referable the genera AmblyrJiiza and Loxomylus, described from the bone caverns
of Anguilla Island, West Indies." . . . "The molars as described and figured
by Professor Cope greatly resemble those of Cas/oroidcs, having, in fact, the same
structure, differing mainly in being somewhat smaller and in possessing a greater
number of laminaB. ... As the lower jaw and skull are thus far unknown
in these genera, it is impossible to say whether their affinities are strictly with the
CJiincJiillidce or whether they are not most closely allied to Gas/oroides." The
description of the lower jaw and other parts of the skull found in the preceding
pages show conclusively that AmblyrJiiza has little affinity with the Castoroididce,
although, as Mr. Allen remarks, there is close resemblance between the molar teeth
of the two forms. Castoroides differs from AmblyrJiiza in its generally Sciuromor-

1 Final Report of the U. S. Geological Survey of Territories, vol. xi. p. 421.

CONTENTS OF A BONE CAVE. 9

phous characters. The inferior incisive alveolus does not project on the inner side
of the ramus, and the angle is a continuation of the general plane which incloses
the alveolus. The ascending ramus is much more elevated, rising abruptly from
the horizontal ramus. The symphysis is short, and, like the beaver, was not proba-
bly invested by the hairy skin as in the Chinchilla and AmllyrJiiza.

Hintory. — The first notice of this genus was published by the writer in the Pro-
ceedings of the Philadelphia Academy for 1868, p. 313, in the following language.
Prof. Cope . . . also exhibited bones and teeth of a large rodent from the cave
deposits of Anguilla, one of the Virgin West India Islands. The characters ob-
served were those of the genus Chinchilla, but the roots of the teeth were contracted
and not so open as in many rodents, as though having a more limited period of
growth, or perhaps like deciduous teeth, which are much reduced in number in
most rodents. The species was nearly as large as the Castoroides ohioensis of
North America, but had relatively smaller incisor teeth. The body was probably
as large as that of the Virginia deer, and the limb bones as stout, as seen in por-
tions of femora and other pieces preserved. He called the animal Amblyrhiza in-
iiudata, and thought that its discovery on so small an island, with others of like
character, indicated that the Caribbean continent had not been submerged prior
to the close of the Post-pliocene, and that its connection was with other Antilles,
while a wide strait separated it from the then comparatively remote shores of
North America.

Subsequently in the Proceedings of the American Philosophical Society-for 1869,
p. 183, I more fully defined the genus from more complete material received from
Dr. van llijgersma. I at the same time described the inferior molars, but on ac-
count of the great difference which they display as compared with the superior
molars, I did not suspect that they belonged to the same genus and species as the
latter. I then gave them other names, generic and specific, which are now with-
drawn. In the same periodical for 1870, p. 608, I added the species A. latidens,
and in the same of the following year (p. 102) I added some observations on the
geographical relations of the genus, and described the species A. quadrans.

AMBLYRHIZA INUNDATA, Cope.

Proceedings of the American Philosophical Society, 18C9, p. 183. Loxomylus longidens, Cope,
loc. cit. 1869, p. 18G. Plates IV. and V., figs. 1-3.

The remains of this large rodent were found in a mass of breccia, which was
thrown out in the excavations made in a cavern in the Virgin Island of St.
Martins, W. I. The remains, occurring in that most eastern region of the West
Indian Zoological district, might be anticipated to have a special interest in connec-
tion with the history of the submergence of a once great continent. With this
impression, the writer examined a quantity of the above breccia and cave deposit,
which was brought to Philadelphia as a probably available phosphatic manure.
It was found to be valueless for this purpose, and the only result of the outlay was
the discovery of the subjects of the present memoir. Most of the fragments first
described were dressed from a single block. There were in this, the extremity of

2 April, 1883.

10 CONTENTS OF A BONE CAVE.

a right femur with patella, shafts of various long bones, fragments of pelvis and
maxillary bones, with three molars, and two partially complete, and other much
broken incisors. The teeth were scattered among the bones, and are so related in
size to most of them as to induce the belief that they all belong to the same ani-
mal. This is strengthened by the entire absence of bones or fragments which
could be referred to any other animal.

Having requested Dr. van Kijgersma to make further search in the cave excava-
tions, that gentleman made an especial expedition to Anguilla for that purpose.
His efforts were rewarded by the recovery of ten or more teeth of the inferior
series, and a number of the superior molars of two individuals with various bones.
Teeth of five individuals of the species have probably been obtained in all. Dr.
van Kijgersma on subsequent expeditions obtained the other species, A. quadrans
and A. latidens.

Of the first individual, three superior molars are preserved, which present four
dentinal columns. These columns are transverse, the first, which I assume to be
anterior, transverse ; the second, the longest ; the third, shortened inwardly and
slightly curved round the very small fourth, which occupies a postero-external
angle of the crown. All are separated by rather thick enamel laminae. The form
of the crown of the largest presents two sides of a square anteriorly and externally ;
the inner side bilobed in correspondence with the two anterior columns; the poste-
rior strongly convex backwards and outwards. The other, similar molar, differs in
the posterior outline being more nearly transverse, and the anterior and interior
outlines being united by a continuous curve. The large portion of the third tooth
preserved is perhaps the external ; it is part of a nearly regular transverse oval.

The first described molar is strongly curved posteriorly, and its diameter narrows
regularly to the contracted base ; there is a shallow groove at the junction of the
anterior enamel lamina with the inner wall. This groove is much more strongly
marked in the second described, but ceases before attaining the contracted extrem-
ity. The shank of the tooth is less curved than in the other. The contraction is
less graded than in the first, but is strongly marked at the base, where the pulp
cavity is not wider than one of the columns.

Lines.

Length of anterior face, No. 1 (on curve) . . .14.3

Diameter of crown (longitudinal) ... .6

" " (transverse) . . . • .5.7

" of root (longitudinal) .... 4

Length of anterior face, No. 2 ...

Diameter of shank (longitudinal)

" " (transverse) . . 5

A portion of one of the inferior incisors of some forty-six lines in length, and
another shorter piece, furnish characters of the species and genus. The inner face
of the tooth is plane, and at right angles to the anterior ; the outer is rounded
obliquely inwards ; the inner face is broad and not prolonged ; the curve of the
tooth is in one plane, and the depth is about equal to the width. A narrow fold
of the enamel embraces the anterior border of the inner and outer faces; it is folded
back at a right angle within.

CONTENTS OF A BONE CAVE. 11

A section of all except the terminal teeth is an oblique rhomboid, the longitudi-
nal diameter being but little greater than the transverse. A single terminal tooth
(either superior posterior or anterior inferior) is narrowed in the terminal column.
All the teeth possess one longitudinal groove on one side, and two on the other,
which are covered but not obliterated by the cement layer. The teeth, though
much straighter and more slender than those of the superior series, yet possess a
light lateral, though no anteroposterior curvature ; those of the upper and lower
series curving in opposite directions.

Indies.
Length of a median molar . . . . . .1.7

Anteroposterior diameter (oblique) ..... .52

Transverse " (both of crown) . . . .43

" " (terminal molar) . . . .39

Longitudinal diameter of terminal molar .... .56

The inner face of the incisor tooth is plane and at right angles to the anterior;
the outer is rounded obliquely inwards ; the inner face is broad and not prolonged,
the curve of the tooth is in one plane, and the depth is about equal to the width.
A narrow fold of the enamel embraces the anterior border of the inner and
outer faces; it is folded back at a right angle within, and with a truncate angle
without. The enamel is sculptured into numerous close, fine longitudinal grooves,
which do not inosculcate. The separating ridges number fourteen near the middle
of the tooth, those near the borders being the strongest. One, strongest of all, is
on the external turn of the enamel, and near it numerous interrupted ridges have a
slightly oblique direction.

The incisors are, as in the modern representatives of the AmHyrliiza, of more
slender proportions than in the beavers, Arctomys, and other rodents, and their
extinct predecessors. They are therefore relatively less stout than those of the
Castoroides. Their sculpture is quite similar to that seen in the Lagidium and
other chinchillas.

Lines.

Width anteriorly -L • • • . • -6

Depth ......... 5.7

The inferior molars differ much from those of the superior series. They are
straight, prismatic, and composed of three dentinal columns, one of which is in-
curved, but none closed at the base. The triturating surface is very oblique in the
vertical direction, indicating the greater elevation of the teeth at one extremity of
the series than the other. A horizontal obliquity of the dentinal columns is pro-
duced by their lateral displacement. Enamel plates but slightly curved.

Extremities. — I describe under the head of this species those bones of adult
individuals of smaller dimensions which the different lots contain. Some of these
came associated with teeth of A. inundata without admixture of larger bones ; and
most of the larger ones came to hand with the jaws and teeth of the A. latidens and
of the A. quadrans. In other lots both sizes were mixed.

A femur distinguished by its bright red color came with teeth of A. inundata
of the same color. The triangular form of the section of the shaft disappears a
short distance below the little trochanter, and is replaced by a suboval one. The
measurements are as follows : —

12 CONTENTS OP A BONE CAVE.

M.
Diameter of base of great trochanter ..... .042

Vertical diameter of neck ...... .023

of head . . . .032

Length of head and neck ...... .033

-031

Diameter of shaft at middle

(antcro-posterior . . . .023

A proximal phalange of the inner digit of the posterior foot, or the hallux, accom-
panied this femur, and presents the same color. The obliquity of its proximal
articular cotylus demonstrates that its position in life is subhomontal. There are
two prominent proximal inferior tuberosities. The distal extremity is depressed,
so that the planes of the superior and inferior surfaces form an acute angle with eacli
other. In the middle of this extremity there is a rudiment of trochlcar keel.
The lateral ligamentous fossae are distinct but small, and are bounded by a low
tuberosity proximally.

Length of phalange ....... .026

Vertical proximal diameter Itotal '

(of cotylus . . . .008

Transverse proxiu.al diameter . . . . . . .014

" distal "...... .008

This phalange I originally described as distal, an error 1 now correct. The
teeth which accompanied it are figured on Plate II. fig. 6.

In a second individual the bones are earth-colored, and were not associated with
teeth. The middle of the shaft of the humcrus is extended in antero-postcrior
diameter, in consequence of the existence of an anterior median ridge. The inner
epicondyle is more distinctly developed than in specimens described under A. Jati-
dens. The crest or anterior ridge of the accompanying tibia is less prominent than
in those of the larger animals. The external glenoid surface overhangs very far,
and is supported by a lateral rib of no great prominence. The middle of the shaft
is strongly compressed so that the transverse diameter is small. The distal ex-
tremity, perhaps of the same tibia, is characterized by the great internal projection
of the processes, and absence of the ridges which rise from them. The tendinous
groove is very slightly roofed. The measurements of these bones are as follows: —

M.

Diameter of shaft of humerus jforeandaft ' • '023

(transverse .... .021

'• of distal extremitv jfore and aft • '021

" (transverse .... .045

of head of tibia and aft .044

(transverse ..... .Of>2

r i ti f i-i • (fore and aft . . .025

of shaft of tibia -{

(transverse . . . . . .019

, ... . ffore and aft . . .023

of extremity of tibia -

(transverse . . . .034

Transverse extent of astragalar facets ..... .028

Associated with these fragments and having the same brown color, is a portion
of the hind foot, which has furnished characters for the generic description. This
specimen is large enough to accommodate the astragalus which accompanies the

CONTENTS OF A BONP] CAVE. 13

bones referred to A. latidens, and I, therefore, am not sure to which species it
belongs. The exposed front face of the navicular bone is nearly twice as wide as
long, while that of the cctocuneiform is subquadrate. The tarsal facets of the
metatarsal bones are very oblique, indicating that the axes of the latter are directed
more anteriorly than that of the tarsus, causing a constant flexure of the foot.
The posterior angle of the second metatarsus is produced proximally behind the
mesocunciform, while that of the third metatarsal projects- behind the ectocunei-
form in an obtuse tuberosity. The shafts of the second, third, and fourth meta-
tarsals are proximally compressed, and medially depressed. The fourth metatar-
sal articulates with the cuboid proximal to the articulation of the second with the
ectocuneiform, a character to be seen in a less degree in the articulation of the
second metatarsal with the mesocuneiform. It, therefore, follows that the line of
tarsal junction of the mctatarsals is a convex semicircle.

DIMENSIONS OF HIND FOOT.

H.
f longest (oblique) .... .032

Diameters of navicular \ *ntoropoBterior • -024

I transverse f front) . . . . .021

*• longitudinal (front). . . . .014

f anteroposterior . . . .017

of mesocuueiform •< transverse .... .010

(longitudinal . . . . .010

i anteroposterior .... .020

" of ectocuneiform •< transverse .... .015

( longitudinal .... .015

Transverse diameter of I., II., and III. metatarsals . . . .037

of II. metatarsals .... .011

of III. " .... .011

Length of metatarsal, No. I. . . . . . . .037

No. II. . . . .043

No. III. . . . ... .044

The proximal portions of three scapulae accompanied the head of a humerus,
and the proximal halves of ulnae and radius of other adult individuals of the
smaller general proportions. These display well excavated supra- and infra-spinous
fossa?, the latter the most pronounced. The plate of the scapula expands rather
widely immediately beyond the short neck ; the glenoid cavity is a transverse oval,
narrowed and produced at the coracoid extremity. The head of the humerus
proper is subequal in its diameters. A section of the ulna just beyond the coro-
noid process presents five angles, one superior, one inferior, two interior, and one
exterior. The upper and lower interior planes thus formed are the narrowest;
the two superior and the exterior soon fuse into a convex face. The coronoid
humeral face is much narrower than the olecranon, whose rim extends downwards
on the inner side to the middle of the shaft. The olecranon is flat below, and as
deep as wide at the end. The head of the radius expands laterally from the neck.
Its articular face narrows outwards, being reduced by a bevel of the inner part of
the superior edge. Its face is a little concave, and is not sigmoid in section as in
some mammals.

14 CONTENTS OF A BONE CAVE.

MEASUREMENTS OF FORE LIMB.

II.

Diameter of glenoid face of scapula j vertical • • .020

(.transverse . . . .035

Transverse diameter of neck of scapula 035

Length of coracoid process (horizontal) . . 015

Elevation of spine of scapula . oiO

Diameter of head of humerus | transverse • • - .033

(fore and aft . . . . .028

of ulna at middle J^nsverse . . , . . .014

(vertical .... .024

Depth of ulna at coronoid process . . . .033

" of olecranon .... 022

Length "... . .032

Width ... ... .019

Diameter of radius at middle I vertlcal • -010

(transverse . . . . .018

" of head of radius \ vertical • -016

(transverse .... .028

AMBLYRHIZA QUADBANS, Cope.
Loxomylus quadrans, Cope, Proceedings Amer. Philosoph. Soc 1871, 102.

This species is represented by teeth intermediate between the A. innndnla and
A. lalidens in size. It was originally established on a first superior molar which
is worn down so far as to have lost the characteristic curvature, and to be nearly
straight in the shaft. Its section is nearly a quarter of a circle, a little longer
anteroposteriorly than transversely, not so elongate as the corresponding tooth of
A. inundata, but more as in A. latidens. The dentinal columns are oblique in the
transverse direction, the second being most extended, the third next, the first a
little shorter, and the fourth shortest, and transversely triangular in form. The
internal and posterior borders are nearly straight, and form a little more than a
right angle with each other. The sides of the shaft are not grooved, except a
shallow interruption between the first and second plates on the external aspect.

M.
Length of shaft preserved ...... .023

Diameter of grinding face | anter°P°sterior ' ' -015

(transverse .... .014

The posterior portions of both mandibular rami of one individual support teeth
which differ from the corresponding ones of A. latidens, and which agree in pro-
portions with the tooth above described. I refer them provisionally to the same
species.

The left ramus exhibits the alveoli of all the teeth. As in the other species they
are straight, and directed obliquely upwards and forwards, and are not oblique
laterally. The third and fourth are a little longer than wide, and their alveolar
septa are a little oblique outwards and backwards. The second alveolus is smaller
and relatively wider than the others.

CONTENTS OF A BONE CAVE. 15

MEASUREMENTS OP TEETH.

M.

Long, diameter of fourth molar ..... .017

Transverse " ..... .014

Long, diameter of third alveolus . . . . .016

Transverse " . . . . . .014

Long, diameter of second alveolus ..... .013

Transverse " " ..... .014

The form of the crown of the last molar differs from the corresponding tooth
of the A. latidens in its greater anteroposterior than transverse diameter; measure-
ments which are equal in the latter species. This is due to the greater anterior
projection of the middle dentinal column on the outer side, and the consequent
obliquity and transverse shortening of the anterior column in A. quadrans. The
section of the tooth in this species is 8-shaped; in A. latidens nearly quadrant-
shaped. The lateral groove in A. quadrans is on the inner side at the front of
the middle column; on the outer side, behind it.

The coronoid process begins opposite the third column of the last molar, and
its base terminates opposite to the dental foramen. The incisive alveolus reaches
to this foramen, which is at the base of the flat front of the ascending condylar
portion of the ramus. The external fossa of the ramus is an inch below the
alveolar border, and is opposite the septum between the second and third alveoli.
There is a distinct masseteric fossa, as indicated by a portion of its anterior part.
The anterior border extends to a point one-half inch behind that below the pos-
terior edge of the fourth molar.

MEASUREMENTS.

M.

Length of inferior molar series (anterior part of m. 1 estimated) . .060
" from m. 4 to mental foramen ..... .025

" of coronoid process from base ..... .014

" of base of coronoid process . . .030

Width of base of condyloid branch at foramen . . . .020

Diameter of dental foramen ...... .020

" of external fossa ...... .013

I mention here a superior incisor tooth, which is in dimensions intermediate
between the largest and smallest of the genus. I do not know whether it repre-
sents a distinct species from those possessing the usual form of incisor, or whether
its peculiar character is abnormal. With the surface of other incisors it is openly
grooved along the middle in a perfectly regular manner. The surface on the inner
side of the groove is plain ; on the outer side, the front is convex. Width of front
m. 014; depth of shaft .015.

AMBLYRHIZA LATIDENS, Cope.

Loxomylus latidens, Proceed. Amer. Pliilos. Soc. 1870, p. 608.

A number of bones and teeth from Simson's Bay came associated in one pack-
age, and agree in having a paler color than the specimens contained in other pack-
ages. Among them is a portion o,f the right maxillary bone of an AmbfyrJtlza,

16 CONTENTS OF A BONE CAVE.

containing the last two maxillary teeth, and having attached the adjacent portions
of the palatine bone. The maxillary bone projects between the last molar tooth a
distance equal to more than half the diameter of the latter. The palatine follows,
presenting some surface continuous with the external face of the maxillary ; its
maxillary suture pursues a nearly straight course forwards a short distance inside
the crowns of the teeth, It begins to curve inwards opposite the middle of the
penultimate molar, but continues some distance further forwards. The teeth
nearly resemble those already referred to the A. inundata, but are considerably
larger. The first molar of the left maxillary which accompanies the other pieces,
is also materially different in form from the corresponding tooth of that species.

The first superior molar includes four dentinal columns, of which the posterior
is much the smallest, and is situated on the inner angle of the posterior face of the
tooth. The enamel bands are directed a little backwards as well as outwards, and
the one between the first and second columns has an inflexion near its inner ex-
tremity. The anterior external enamel plate is regularly convex in section and
oblique in position, striking the dividing plate which follows it at a point on the
outer side of the crown which marks the middle of the long diameter. In the
A. inundata the corresponding tooth is not only very much smaller but is relatively
narrower in transverse diameter, and the anterior column does not extend so far
backwards on the outer side. This tooth is, in the A. latidens, as large in all its
diameters as the third or fourth molars.

The second molar is wanting from the specimen, but its alveolus shows that its
posterior wide columns project further inwards than in either of the teeth which
follow. The grinding surface of the third superior molar is straight on the inner
side and convex on the outer through the prominence of the second and third col-
umns ; the fourth is small and internal. The anterior column of the fourth molar
is more transverse than the others or than those of the third, which are oblique.
The inner side of the grinding face of the tooth is longer than the outer, owing to
the presence of the small fifth column.

A pair of incisors with both extremities wanting, which are held together by the
matrix, accompanied the above specimens. They belong to the upper jaw, and as
they are of appropriate size to the superior molars just described, I suspect that
they belong to the same animal. Their section is triangular, with the external
anterior margin projecting and curved with an incurved border of the enamel sheet.
The surface of the latter is marked with numerons irregularly spaced longitudinal
grooves, excepting for a narrow space along the inner margin, which is smooth.

MEASUREMENTS.

If.

Length of first molar ..... • -033

~. r • T r <• r ( fore and aft . • .018

Diameter of grinding face of first molar

(.transverse . . .(Jl<>

(fore and aft . .015

of third molar J.

(.transverse . . .015

, (fore and aft . . .017

of fourth molar <

(transverse . . .014

,,.,.. ... (fore and aft . . . .017

of shaft of superior incisor •< ,..„

(.transverse . . . .017

CONTENTS OF A BONE CAVE. 17

The larger number of specimens received from Dr. van Rijgersma are of a dull
rusty-brown hue, and were doubtless derived from one and the same locality, and
were found at or near the same time. They include the bones and teeth of the
largest specimens of the genus Amllyrluza, while most of the smaller individuals
were received at other times, and from differently colored deposits. The mandibles
and teeth of two different species are included in this lot, and one of them is pro-
bably the A. laiidens, but which one is to be collated with the maxillary bone above
described is not certain. As the maxillary teeth of A. latidem differ in the
specific character of a greater width from those of A. inundata, it is probable that
the inferior molars which bear the same relation to those of A. immdata, are the
ones to be described here. Those which present other differences, which are less
marked, are referred to another species under the name of Amblyrhiza quadrans.

Of mandibles referable to the A. latidens, there are both sides of one individual
iu a more or less mutilated condition, one only presenting the molar series complete.
Seen from the outer side, the summits of all the inferior molars rise an equal dis-
tance above the alveolar border, the insignificant excess in height of the first tooth
being only individual. But on the inner side the alveolar edge descends forwards
so as to leave the first molar projecting to four times the elevation of the fourth.
Viewed from above, the molars successively narrow forwards, the first being only
half as wide as the fourth. The direction of the transverse enamel bands becomes
more oblique forwards externally, so that the anterior column of the first molar
stands at an angle of 45° to the long axis of the series. Those of the fourth
molar are moderately oblique, and the middle column in the third and fourth
molars has the greatest transverse extent. The inner posterior angle marks the
middle of the basis of the small coronoid process. The apex of the latter is
recurved, and the direction of its plane is 45° from that of the mandible. The
inner alveolar border is continued as a ridge for a short distance behind the last
molar, and then descends and disappears at the distal foramen. This foramen
opens on a horizontal surface or plane which is bounded without by the external
border of the ascending limb, and within by the strong ridge which connects that
portion with the prominent alveolar sheath of the inferior incisor, The inner face
of the jaw is gently concave to the incisor alveolar ridge ; the external face is
marked by a fossa, about as wide as the diameter of a molar tooth 25mm below
the alveolar border, and opposite the third and part of the second teeth.

MEASUREMENTS.

M.

Length of series of molars (No. 1) . . . . • -058

" of posterior molar ...... .047

Projection of posterior above alveolar edge .... .005

Diameter of fourth molar |fore and aft ' -0,16

(.transverse .... .016

« of third molar J fore and aft . -013

(.transverse. .... .015

e , , (fore and aft . . . .012

" of second molar «

(transverse ..... .014

. c (fore and aft ... . .015

of first molar 4

(transverse .... -Oil

Elevation of first molar on inner side ..... -015

3 April, 1883.

18 CONTENT SO PA BONE CAVE.

The premaxillary bones with incisor teeth of several individuals accompany
the mandibles just described, and some of them probably belong to this species,
and others to the A. quadrans. I cannot detect any difference in the two which
are best preserved, and which correspond to the two individuals represented
by molar series, and symphyses with inferior incisors. The section of the pro-
duced pair of premaxillary bones is triangular near the middle. That of the
lateral walls is greatly convex, and the median line below is keeled for a short dis-
tance. Immediately behind the keel is a narrow fossa. The section near the base
of the incisors is subquadrate, and there is a deep longitudinal median fossa just
behind the base of the teeth, whose fundus appears to receive a foramen.

The symphyses above mentioned are long and acuminate. The union of the
inferior surface continues posterior to the point of divergence of the dental ridges,
forming the bottom of a large fossa. The section of the anterior or beak-like
portion of the symphysis is subtriangular, the sides approaching each other upwards,
but separated by a broadly obtuse median ridge. Near the middle of the length a
slight enlargement is visible in both the specimens. The inner face of the incisor
teeth is more strongly convex than the external.

MEASUREMENTS.

M. •
Length of beak of symphysis . . . . . .090

Diameter of beak at middle |vertical • -025

(.transverse .... .023

„ . ,. . . . (vertical .014

" of inferior incisors <

(. transverse . . . . .014

The most complete mandible includes the right ramus with all the alveoli, the
entire symphysis with the right incisor tooth in place. The molar alveoli are empty.
The posterior alveolus is as wide as long, as in the typical specimen, and is in form
about a quarter of a circle. The third and second alveoli are similar in form, but
a little more quadrate, and the second is materially the smallest of all. The third
is much longer and a little wider than the second. The symphyseal beak is more
attenuated than in the specimens just described, and its superior middle line is
more angular, giving a roof-shaped section. The incisor tooth is exposed on its
external side for almost the entire length of this rostrum, to a point 17m"u
anterior to the divarication of the alveolar ridges. The cutting edge of the incisor
is transverse. The lateral fossa is below the second and part of the first molar.
This is anterior to the position which it occupies in the mandibles described under
A. latidens and A. quadrans. This fact, in connection with the slight difference
in the form of the symphyseal rostrum, has led me to question the propriety of
the reference of the animal to the A. latidens. The fourth molar evidently has
the form characteristic of that species, rather than that of the A. quadrans.

CONTENTS OF A BONE CAVE. 19

MEASUREMENTS.

M.

Length from apex of symphysis to posterior base of coronoid process .160

" of symphyscal rostrum ..... .080

" from base of rostrum to the first molar . . . .012

" of free inner face of incisor ..... .038

" " outer " . . .110

" of alveolus of first molar ..... .018

" " of second molar . . . . . .011

" " of fourth molar . . . . .016

Width " " " . . .016

" between rami at fourth molars .... .050

The mandibles, with teeth of the two larger individuals last described above,
were sent by Dr. van Ilijgersma in connection with numerous bones of the skeleton,
which, in color, mineral character, and dimensions, are so appropriate to them, that
I do not hesitate to believe that they belong to the same individuals, excepting in
so far as they include portions of the skeleton of the A. quadrans. Of these the
vertebra; first claim attention.

The atlas is depressed in form, and the condylar fossa? are divergent upwards.
The odontoid fossa viewed from the front has only half the diameter of the neural
arch. The transverse processes expand backwards and outwards, their postero-ex-
ternal angle being behind the plane of the facets for the axis, and the border being
a regular curve from a short base behind the margin of the condyloid fossa. The
axial facets are wide oblique ovals separated below and bounded by a deep groove
at the base of the neural arch. The neural spine is lost.

The neural canal of the axis is subquadrate, and a little wider than deep. The
neural spine is strongly keeled in front, the inferior extremity of the keel project-
ing beyond the arch. The posterior face of the spine is subconcave. The poste-
rior zygopophyses are well developed. The posterior articular face of the cerebrum
is a hexagon, a little wider than deep, and its vertical plane is directed a little for-
wards above. The condyles for the atlas descend to the plane of the inferior sur-
face.

DIMENSIONS OF CERVICAL VERTEBRAE.

M.

Length of atlas below < "east (median) . .019

(greatest . . . . .025

Width of atlas 5antcl'iOT|y • • -059

c posteriorly . . . . . .108

Length of transverse process behind ..... .032

Depth of " " at base . . . . .020

Width of neural canal .... . . .025

Vertical diameter of facet for axis . . . . .016

Length of centrum of axis (without odontoid) . • . .024

of odontoid process .... .014

Diameter of posterior face of centrum \ ver

(transverse . . .023

Greatest width of centrum . . ... .040

Interior elevation of neural arch ... . .018

Diameter of neural spine { tral>s™rse • -020

( antero-postenor . . . .018

20 CONTENT SO FA BONE CAVE.

The centra of the lumbar vertebrae are strongly constricted at the middle, and
they display an arterial foramen near the middle of each side; their inferior surface
is regularly convex, and without keel. One of the articular faces has a slight ten-
dency to convexity. Neural canal wide. The proximal extremity of the first sa-
cral centrum is plane, and not depressed nor concave. The inferior faces of the
sacral vertebrae are slightly concave in the anteroposterior direction. The cen-
trum of one of the canals is oblique in the vertical plane of one of its extremities.
Its inferior surface is strongly convex, the superior moderately so. The base of
the diapophysis is oblique.

DIMENSIONS OF LUMBAR AND SACRAL VERTEBRAE, ETC.

M.
Length centrum No. 1 ... ... .036

Diameter anterior articular face 5vertlcal • °38

(.transverse .... .038

" posterior face jvertical • '°41

(transverse . . . . .046

Length centrum No. 2 ...... .045

anterior articular face

posterior articular face

Diameter anterior articular face fvertlcal • • -039

(transverse . . .044

(vertical . . . . .040

(transverse .... .054

Length three sacral centra . . . . . . .113

" of first sacral centrum ..... .052

Diameter of articular face of first sacral { vertlcal

(transverss . . .048

Length of third sacral centrum . . . .026

Diameter of third sacral centrum in front I vcrtlcal

(transverse . . .026

Length of centrum of a caudal . . . . .022

Diameter of articular face of a caudal \ vcrtlcal

(transverse . . .024

No scapula certainly referable to this species has been preserved. The proximal
and distal extremities of the humerus differ from those of the A. inundata in their
superior size. The supracondylar fossa is not perforate. The fossa of the great
trochanter of the femur is only excavated in the inferior half, its lower border being
opposite the middle of the neck of the femur. Its superior portion is the deepest,
and the fundus corresponds to a projection of the opposite border of the external
face. The external face of the femur is flat to as far below the little trochanter
as the length of the great one, when it passes by. a regular curve into (he posterior
face. The angular ridge bounding the anterior face continues further down. The
anterior face of the femur is well defined by lateral parallel ridge-angles, and main-
tains a nearly equal width from its middle to near the proximal extremity of the
great trochanter. Near the middle (where my specimens are broken off) it is
gently convex, but it soon becomes flat, and is slightly concave at the base of the
great trochanter. The extremity of the latter is a wide continuation of the exter-
nal face of the shaft, rising obliquely inwards. Its borders are roughened and at
some points overhang, furnishing points of muscular insertion. It projects beyond

CONTENTS OF A BONE CAVE. 21

the neck a distance equal to 1.25 transverse diameters of its base. The fossa
ligamenti teris is small but deep, and far removed from the edge of the head,
which is not connected with it by a groove. The condyles of the femur are rather
divergent. The condylar surfaces are continuous with the patellar. The external
bounding angle of the latter is a little more elevated than the internal. The latter
bounds a trochlear bevel which extends band-like along its inner side. This looks
as though the patella possessed a corresponding bevel which would include an angle
with its remaining inferior surface.

The proximal portion of the tibia is characterized by the depth and width of the
longitudinal external fossa, which is bounded by the crest. The proximal articular
faces are not concave. The posterior face of this part of the shaft is narrow. The
distal portion of the shaft is marked on its external aspect by a narrow groove
which runs behind the fibular ridge, terminating where the latter begins to expand.
The angle rising from the anterior malleolar tuberosity becomes the internal edge
of the shaft, while the one which arises from the posterior one disappears on the
posterior face. The anterior and posterior edges of the external glenoid cavity
are equally produced, and the external and internal borders have the same relation.
The internal superior facet of the tibial articulation of the astragalus is half as
long and half as wide as the internal ; the area of the internal lateral face is still
smaller. There is a shallow fossa at the neck on the outer side. The navicular
facet is subround, and deeper than wide ; the facet for the additional proximal tarsal
bone is subtriangular with the apex posterior; its surface is moderately convex.

MEASUREMENTS.

M.

Diameter of distal extremity of humerus | fore alld aft ' '029

(transverse . . . .062

•' ofacetabulum \ vertical • • '°41

(.transverse .... .038

" of neck of ilium . . . . ... .038

" of shaft of femur near middle -{fore and aft ' -029

(transverse . . . .043

" of base of great trochanter 5 fore and aft . .034

(transverse . . . .045

Length of great trochanter ...... .065

" of head and neck ...... .036

Diameter of head of femur ...... .036

" of neck of femur ...... .026

" of condyles of femur (transverse) .... .085

Fore and aft diameter of femur at patella .... .052

Diameter of head of tibia l^e and aft . .060

(transverse .... .077

Width of external articular face ..... .030

" of internal " ..... .035

" of posterior plane of shaft ...... .030

Diameter of inferior part of shaft jf°™">daft • -020

(transverse . . . .024

» of distal extremity (anteroposterior . .035

(transverse .... .049
Transverse diameter of trochlear face . .031

22 CONTENTS OF A BONE CAVE.

M.

Width of astragalus ]PosteriorIy • -044

(medially ..... .030

" of external superior face . . . . . .017

" of internal " ..... .009

" of internal face. ...... .009

Depth at internal border . . . . . . .014

" of internal facet of head ..... .012

Diameter of navicular facet -j vertical •

(transverse . . . .018

Portions of a sacrum and a lumbar vertebra of another individual of the size of
this species present some additional characteristics. Thus the anterior margin of
the middle of the neural arch in the first sacral is produced into a pointed process
which separates two fossa?. The anterior zygapophyses in front of these are well
developed, directing their articular faces upwards and outwards at an angle of 45°.
The diapophysis of the lumbar arises from the summit of the neurapophysis, and
is directed forwards at an angle of 45°. The section of the shaft is triangular and
the apex is acuminate. Its length is .030 ; vertical diameter of neural canal .024.
The anterior zygapophysis is produced upwards and a little outwards ; its face is
nearly vertical.

ARTIODACTYLE.

The following characters are derived from the right maxillary bone of a young
animal which supports the temporary dentition, and from the bones of two indi-
viduals, one adult and the other young, probably the one to which the maxillary
bone pertained. I give the full characters for the benefit of those who have the
opportunity of studying the numerous species of the order found in the Old World.

Temporary superior dental series of the maxillary bone, Pm. 1 ; M. 3 ; other
teeth unknown. Last premolar with an external subacute column, and an inter-
nal short column, which are united posteriorly, forming a rather wide grinding
face, which has an external cutting edge; no single anterior column. True molars
with four usual crescents, whose anterior horns are the inferior extremities of ver-
tical ribs. The lakes are without cementum, and are united at their adjacent
angles for some distance into the crown. The crowns of the molars are moderately
elongate, and the roots of the anterior teeth are rather long.

The temporary premolar of this genus is more simple than the last of the per-
manent series in Anchcnia, biit presents a more complete inner crescent than does
the penultimate premolar in that genus. Its inner crescent is better developed than
that of the genus Camelus, since it is united with the external one at both ex-
tremities. The animal is apparently a member of the Bovidce.

Attached to the entire maxillary bone are the malar and palatine bones, in
complete preservation. These bones display some interesting relations which I
now describe. The premaxillary bone evidently reached the nasals in the complete
skull. The palate is deeply emarginate posteriorly. The emargination between
the maxillary and the procesms pyramidalis of the palatine is profound. The

CONTENTS OF A BONE CAVE. 23

palatine bone extends forwards on the middle line to opposite the middle of the
dental series. In front of the processus pyramidalis, and separated from it by a
deep fissure, the thin lamina perpcndicularis rises. Its anteroposterior diameter
is greater above than below. Its anterior margin articulates by squamosal suture
with the interior side of the superior alveolar wall of the maxillary bone. This
wall is so strongly convex as to divide the maxillary sinus into two fossae, an inter-
nal large, and external small. The latter is bounded internally by the wall of the
infraorbital canal, which extends along the summit of the alveolar arch. The
nasal roof of the antrum is thin, and only extends as far posteriorly as the middle
of the first true molar. The posterior extremity of the malar bone is deeply
notched, and was connected with the frontal, inclosing the orbit behind.

The palate is rather wide, and the molar series is curved or convex outwards.
The posterior external column of the last true molar is half as wide again, in antero-
posterior diameter, as the corresponding anterior column. The vertical ribs are
most prominent in the second and third true molars, the anterior being obsolete
on the first true molar, and wanting on the premolar. In all the molars the anterior
column has a median external rib, which is strong on the first true molar, less
marked on the other teeth, and wanting on the posterior column in all. The two
inner columns are separated by a vertical fissure to a considerable depth in the
second true molar, but not in the first. There is a slight contraction of the anterior
lake of the first true molar caused by a low rib on the posterior part of its internal
enamel wall.

The form of the face is short and high, the facial plate of the maxillary rising
above the level of the lachrymomalar suture. The external foramen of the infra-
orbital canal is large, and issues above the posterior part of the premolar. There
is no lateral facial fossa. The inferior orbital border is strongly flared outwards.
The posterior palatal notch enters as far as the anterior margin of the last molar,
and the external pterygo-maxillary notch is not quite so deep. The palato-maxil-
lary suture leaves the alveolar border at the front base of the last molar.

MEASUREMENTS.

M.
Length of entire molar scries . . .... .047

" of true molars ....... .039

Diameter of preraolar j «">teroposterior . . .0075

(transverse ..... .0055

" of first molar Anteroposterior . . .0120

(.transverse ..... .0055

« of second molar | anter°P°sterior • • '°13

(.transverse .... .006

Width of palate in front of premolar . . .014

" at second true molar .... .030

Associated with the cranial mass already mentioned as representing this species,
there are several bones of the skeleton of an immature mammal, which are appro-
priate to it in size and proportions. These are an axis and a dorsal vertebra,
humeri, radius, pubis and ischium, femur and cannon bones, all without epiphyses.

I now describe the other bones already mentioned. There is no sinus in the

24 CONTENTS OF A BONE CAVE.

anterior border of the obturator foramen. The ischium has a flat section, with
the long axis subtransverse. The section of the pectineal part of the pubis is a
wide oval. There is a prominent transverse ischiatic tuberosity. The axis is
moderately elongate, and the summit of its neural spine is horizontal. The ante-
rior border of the neural arch is pierced by a large foramen on each side. The
diapophyses are small and originate as a keel on the posterior half on each side,
which continues into a short acute posteriorly directed process. The dorsal vertebra
is short and has an opisthococlian centrum. The diapophyses are short and with
down-looking articular surfaces, and the neural spine is narrow and high.

The distal fossae of the humerus do not communicate ; the olecranar is deep.
The distal extremity is more transversely expanded than in Procamclus occidentalis.
The radius is flat, and with convex upper surface. The shaft is slightly arched
upwards, and the section of each extremity is a transverse oval of subequal long
diameters; the distal end has the greater vertical diameter. The shaft of the
femur is not very long and is subcylindric. There is a slight anteroposterior
curvature. The great trochanter is large, and is strongly recurved posteriorly.
The posterior cannon bones are rather stout, and the section at the middle is sub-
quadrate with rounded sides. Of the proximal facets the ectocuneiform is the
largest. The anterior cuboid is a quarter of a circle, and is separated from the
narrow posterior cuboid by a wide groove. The mesocuneiform is larger than the
posterior cuboid, and is rather wide; its posterior border forms a protuberance.
The inner extremity of the posterior cuboid is the most elevated part of the proxi-
mal end.

The distal fissure does not extend above the bases of the epiphyses.

MEASUREMENTS (WITHOUT EPIPHYSES).

M.

Length of axis on centrum (laterally) .... .023

Total elevation of axis .... . .030

Anterior diameter of centrum of axis J transv<;rse •

(vertical . . .008

Posterior •' " J transverse

(.vertical . . .011

Total elevation of anterior dorsal ..... .050

Elevation of neural arch .... • .012

of spine .... • -032

Length of centrum . . . . . . .010

of humerus ....... .081

Transverse diameter proximally ....

distally .... • -023

of shaft ... • -009
Length of radius (chord) . .
Transverse diameter proximally ....

medially. . . . .Oil

distally .

Length of femur ..... • -0"

Transverse diameter at trochanter ....

medially. . -009

" " distally 020

CONTENTS OF A BONE CAVE. 25

M.

Length of cannon bone ....... .065

Transverse diameter proximally ..... .016

" " medially. ..... .009

" " distally . . . . . . .019

Length of isehium ....... .047

Width medially . . . . . . . .010

Length of pubis to symphysis ..... .012

Greatest diameter of symphysis ..... .007

Several bones of a second individual accompanied those above described, which
are of larger size and present adult characters. They include mutilated scapula and
humerus, a nearly perfect ulno-radius, a metacarpal cannon bone, and a phalange.

As compared with those of a rather small Ovis aries, these bones are smaller,
and present the following differences : The glenoid cavity of the scapula presents a
straighter or less convex internal border. The coracoid process is a little longer ;
the base of the spine is nearer the anterior edge. The form of the humeral shaft
is similar, but the internal peduncle of the condyles has greater anteroposterior
width. The extremities of the bone are lost. The radius scarcely differs from
that of the sheep. I notice a little greater depression of the external part of the
head, and a rather longer proximal interosseous foramen ; the fossa behind the
scaphoid facet is deeper. The proximal end of the metacarpal cannon bone has
the same form as that of the sheep, but differs in the larger size of the facet for
the os tmpeziomafjnum, which is also without the large interior emargination seen
in the sheep. Distally the fossa? at the anterior bases of the condyles are less pro-
nounced, and the external trochlea? of both condyles are less produced outward.
In accordance with this form the proximal phalange is a little narrower proximally
than in either foot of the sheep; the shaft is also narrower than that of the sheep's
fore foot, while the entire bone is longer than that bone, or the corresponding pos-
terior phalange of the sheep. The ligamentous fossae are bounded by much more
prominent and rugose ridges of insertion. The inferior tendinous insertions are
about as in the sheep.

MEASUREMENTS.

M.

Diameter of middle of shaft of humerus j anteroposterior . . .018

(transverse . . .015

Width of neck of scapula . . . . . . .0175

Length of radius ....... .038

Proximal diameter of radius j vertical (greatest) • • -013

(transverse . . . . .027

Distal diameter of radius lvertical <of facets) ' -013

(transverse .... .023

Length of anterior cannon bones ..... .097

Diameter of proximal end of cannon bone |transvel<sc

(fore and aft (internally) .014

Length of external proximal phalange .... .037

Diameter ' proximallr jvert!cal • '°15

(transverse . .011

|vertical (chord) . .011
(transverse (least) . .007
Discovered by Dr. Rijgersma with the bones of Amblyrhiza, etc.

4 April, 18S3.

26 CONTENTS OF A BONE CAVE.

HOMO, Linn.

Associated with the remains described in the preceding pages, Dr. Rijgersma
discovered a highly interesting relic of the stone age of the human inhabitants
of this portion of the West Indian Islands. I use the term " stone age" in a
chronological sense only, since the region in question possesses chiefly coral rock,
and little or none that is adapted for conversion into cutting instruments, so that
the inhabitants resorted to the use of animal products, as teeth, bones, and shells.
The implement found by Dr. Rijgersma is a long-ovate spoon-shaped scraper or
chisel, cut by human hands from the lip of the large Strombus gigas. The ribs of
the external surface and the smooth internal surface are easily distinguished, and
the distinct natures of the lamellar and prismatic layers have been evidently well
understood by the artificer, who has ground away the latter in order to put a sharp
edge on the former at one end. This edge is sharp, and mainly well preserved.
The implement has a greater median width, and smoothly ground thick margin ;
the end of the plate is obtuse and with thick edge, almost entirely composed of
the prismatic layer. It has evidently been held in the hand, and been used after
the manner of the stone scrapers of the North American Indians.

This chisel was found by Rijgersma under circumstances precisely similar to those
attending the discovery of the gigantic rodents. Some portions of each of the
species described were imbedded in the breccia, and others occurred loose in a red
earth in cavities of the breccia. The chisel has the color and constitution of the
latter teeth and bones, and was found with them in this earth. Some of the teeth
are even more fresh-looking and less stained than the chisel.

A good figure may be found in the Proceedings of the American Philosophical
Society for 1868, Vol. XI. Plate V. fig. 4, which is repeated ou Plate 1., of the
present work.

CONCLUDING REMARKS.

So far as the appearance presented by the specimens described in the preceding
pages is an indication, all may have been buried in the cave deposit at one time,
and may represent animals which lived together on the Island of Anguilla. But
the iron oxide of the deposit evidently imparts so deep a stain to whatever is buried
in the cave, that it is impossible to base an estimate of the age of such objects by
their color alone. It is well known that no species of Bovidce is indigenous to any
part of the Neotropical realm, including the West Indian Islands, and that indi-
viduals of that genus now in the islands arc importations. But the bones of the
ruminant of Anguilla are as deeply colored as those of the Amblyrhiza, some of them
even more so. It is also evident that the implement made from the lip of Strombus
gigas has been for a long period exposed to the iron oxide, since its stain is deeply
fixed, totally replacing the natural color, even to the usual white of the dense
enamel of the inner surface seen in fossils of early tertiary age.

Whether the extinct genus Amhlyrliiza was contemporary with the ruminant and

CONTENTS OF A BONE CAVE. 27

man, is a question which must be left for future investigators. Of one point we
may be assured, namely, that the deposit inclosing the Amblyrhiza is of late geo-
logical age. This may be derived from the close affinity which exists between that
genus and existing ones, and also from some species of shells which I found im-
bedded in the matrix inclosed in the intercondylar fissure of the femur of the
Amblyrhiza latidens. I sent these to my friend Thomas Bland of New York, whose
knowledge of West Indian land-shells is exhaustive, and he informed me that they
belong to a species of Tudora so nearly allied to the existing species T. pupce-
formis as to make it difficult for him to characterize it as distinct. I would not
place the age of the Amblyrhiza earlier than the Pliocene.

The island of Anguilla, now embracing but thirty square miles, could not readily
have supported a fauna of which these huge rodents formed a part. Such large
animals have no doubt ranged over a more extended territory. This, and other
facts mentioned by Pomel, lend probability to the hypothesis of the latter author,
that the submergence of the ranges connecting many of the islands of the Antilles
has taken place subsequent to Pliocene times.

EXPLANATION OF PLATES.

PLATE I.

Reptiles, birds, and mammals, all of the natural size, excepting fig. 12.

Fig. 1. Femur of an Iguana, posterior view; a, exterior; 6, proximal, and c, distal views.

Fig. 2. Hunierus of a species of Procellaria, front view ; a, posterior ; 6, proximal, and c, distal
view.

Fig. 3. Coracoid bone of a large bird, superior view ; a, interior view.

Figs. 4-6. Long bones of a rodent.

Fig. 4. Left Inimerus without distal extremity, external view ; a, proximal view.

Fig 5. Distal portion of right humerus, anterior view ; a, posterior view.

Fig. 6. Left femur, front view ; a, posterior view.

Fig. 7. Proximal portion of a rib which accompanied the jaws of Amblyrhiza latidcns, posterior
view ; a, proximal end.

Fig. 8. Proximal phalange of rudimental first digit of Amblyrhiza inundala, superior view;
a, inferior, b, proximal, and e, distal views of the same.

Fig. 9. Unknown bone found with those of Amblyrhiza; a and b, extremital views.

Fig. 10. Symphysis mandibuli of Amblyrhiza from above.

Fig. 11. Maxillary, molar, and palatine bones of an Arliodaclyle, supporting the temporary
dentition, from the right side ; a, from below.

Fig. 12. Shell chisel of human manufacture, three-fourths natural size, viewed on the external
surface of the shell ; a, edge view.

PLATE II.

Figs. 1-3. Amblyrhiza quadrant, natural size.

Fig. 1. Portion of the left maudibular ramus from above ; a, the same, the external side ; b, the
inner side, the incisive alveolus broken open.

Fig. 2. Ilight ramus, probably of the same animal, from above ; b, the inner side.

Fig. 3. First superior molar from below; a, outer side; b, inner side.

Figs. 4 and 5. Left and right mandibular rami of Amblyrhiza latidens, natural size ; lettering
as above.

PLATE III.

Mandibular and premaxillary bones of Amblyrhiza latidens or A. qnadrans, natural size.

Fig. 1. Premaxillary bone with portions of both incisor teeth, right side; a, superior view.

Fig. 2. Inferior view of premaxillary bone of another individual.

Fig. 3. Right mandibular ramus and symphysis ; a, from above.

Fig. 4. Symphyseal region with incisor teeth from below; a, from above.

(29)

30 EXPLANATION OF PLATES.

PLATE IV.

Bones accompanying the cranial fragments of Amblyrhiza quadrans and A. latidens, one-half
natural size.

Fig. 1. Atlas from before ; a, from behind ; b, from below ; c, from above.

Fig. 2. Axis from left side ; a, from behind.

Fig 3. Lumbar vertebral centrum from side ; b, from front ; c, from back.

Fig. 4. Lumbar vertebral centrum ; b, from front; c, from behind.

Fig. 5. Three vertebra? of sacrum, profile ; a, from below ; b, from front ; c posterior extremity.

Fig. 6. Caudal vertebra from side ; a, front view.

Fig. 7. Proximal end of humerus from behind; a, external view ; b, proximal view.

Fig. 8. Condyles of humerus broken from the shaft from behind ; a, proximal view ; b, distal
view.

Fig 9. Femur without head, external view ; a, internal view ; c, section of shaft at fracture.

Fig. 10. Proximal portion of another femur, with the great trochanter distorted by pressure,
posterior view; a, internal view; c, broken extremity of shaft; b, proximal end of shaft.

Fig. 11. Distal extremity of a femur, right side ; a, anterior, and b, posterior views.

Fig. 12. Patellar surface of another femur.

PLATE V.

Bones of Amblyrhiza and ruminant, one-half natural size.

Fig. 1. Anterior part of sacrum of Amblyrhiza Ualidens, which accompanied the bones of
Amblyrhiza inundala, viewed from above • a, from below ; b, anterior extremity.

Figs. 2-6. Bones which accompanied the jaws and teeth of Amblyrhiza quadrans and A. latidens.

Fig. 2. Distal portion of right humerus from front ; a, posterior view.

Fig. 3. Right acetabulum, external view.

Fig. 4. Proximal portion of left tibia, interior view ; a, posterior, 6, exterior, and c, proximal
views.

Fig. 5. Distal portion of left tibia with astragalus in place, anterior view ; a, posterior view ;
b, distal view of the same tibia alone.

Fig. 6. Left astragalus from above ; a, from below ; b, from the inner side ; c, distal view ;
d, posterior view.

Fig. 7. Fourth metatarsal bone, which accompanied the sacrum represented in fig. 1, from the
front; a, from behind.

Figs. 8-10. Bones of an adult ruminant described on page 24.

Fig. 8. Right radius, with parts of the ulna adherent, anterior view ; a, posterior view.

Fig. 9. Metacarpal cannon bone, anterior view; a, posterior; b, proximal; c, distal views.

Fig. 10. Left proximal phalange of the same, from above j a, from below; 6, proximal ; c, distal
views.

PLATE

T. Sinclair &. Son, Lith PKil»

Smittisoman Publication N9489

r'CSSILS ETC. FROM THE ANGU1LLA CAVE

PLATE II.

Smithsonian Publication NV4-89.

T. Sinclair it Son, LitX . PWln

l-S.AMBLYRHLZA QUADRANS 4-5 AiATIDENS.M.

PLATE IV

Smithsonian Publication N9489.

T. oindaiT St. S<m, Lnh FKila.

AMBL.YRHLZA.M.

Smithsonian Publication N°489.

1-7 AMBL.YRHIZA 8-10 ARTIODACTYLE K

u:

RETURN TO the circulation desk of any
University of California Library

or to the

NORTHERN REGIONAL LIBRARY FACILITY
Bldg. 400, Richmond Field Station
University of California
Richmond, CA 94804-4698

ALL BOOKS MAY BE RECALLED AFTER 7 DAYS

• 2-month loans may be renewed by calling
(510)642-6753

• 1-year loans may be recharged by bringing
books to NRLF

• Renewals and recharges may be made 4
days prior to due date.

DUE AS STAMPED BELOW

MAY 1 6 2003

M.

LD 21A-50™-8 '&1
(C8481slO)476B