|
,/ /0 /
NEW YORK ACADEMY OF SCIENCES
SCIENTIFIC SURVEY (
OF
5 .
A. "
^ *- A --
Porto Rico and the Virgin Islands
VOLUME VII-Part 3
Paleobotany of Porto Rico
BY
Arthur Hollick
__m
NEW YORK:
PUBLISHED BY THE ACADEMY
1928
[Issued October 31, 1928]
NEW YORK ACADEMY OF SCIENCES
SCIENTIFIC SURVEY
OF
Porto Rico and the Virgin Islands
VOLUME VII-Part 3
Paleobotany of Porto Rico
BY
Arthur Hollick
NEW YORK:
PUBLISHED BY THE ACADEMY
1928
[Issued October 31, 1928S
CONTENTS
PAGE
Introduction....... . . . ................................. ... 177
Descriptions of species. ........................ ... . . . 181
Thallophyta .. .. ... ...................................... 181
A lgae . . .................................. . . ...... 18 1
P teridophyta ..................................... ........... 182
F ilicin ae ................................... ............. 182
L ycopodinae ............................................... 183
Spermatophyta ................................................. 183
G ym nosperm ae .................. ......................... 183
A ngiosperm ae .............................................. 186
Monocotyledonae ....................................... 186
Dicotyledonae .......................................... 194
C horipetalae ....................................... 194
G am opetalae ...................................... 218
Plant remains of undetermined taxonomic relationship .............. 225
Discussion of the flora ............................................... 226
B otanical discussion ............................................ 226
G eological discussion ............................................ 232
Summary and conclusions ........................... . . . ... 238
Plates ....................... ..................................... 239
Index ... ........ .............................................. 391
PALEOBOTANY OF PORTO RICO
ARTHUR HOLLICK
INTRODUCTION
The earliest published references to fossil plant remains in
Porto Rico were by Hodge,* who mentioned (loc. cit. (a), p. 278)
the presence of a bed of limestone, and an adjacent bed of hematite
containing fossil leaves, in the northeastern part of the region
discussed. The beds were inferred to be lower Cretaceous, by
reason of a characteristic species of coral found in the limestone,
and from the general facies of the flora as indicated by a few plant
remains found in the hematite bed.
In the author's second paper the subject was discussed at
greater length and more in detail, in connection with a description
of an unconformity between the Barranquitas-Cayey series of
rocks and the Sierra de Cayey series, in the course of which dis-
cussion he remarked (loc. cit. (b), pp. 192-193): "A short distance
below this unconformable contact two calyxes of a coral were found
in a limestone, which have been identified as Cladophylla furcifera
... Its presence here should indicate that the rocks are at
least of Comanche age. Below the limestone occurs a thin bed of
bog iron ore and in it were some leaves. Dr. Edward W. Berry
and Dr. F. H. Knowlton identified these leaves as follows:
Nelsonia [= Nilssonia], an old Mesozoic order [= genus] of
cycads.
Protorhipis, a fern with the same range as above.
Another species of Mesozoic fern.
A dicotyledon.
The above authorities stated that these plants were not
critical for the Comanche, but that they tended to support the
evidence of the corals-at least they indicate Mesozoic age."t
Hodge, E. T. (a) Geology of the Coamo-Guayama region, Porto Rico [Secretary's
abstract], New York Acad. Sci., Annals vol. 27, pp. 277-278. 1917. (b) Geology
of the Coamo-Guayama district. New York Acad. Sci. scientific survey of Porto Rico
and the Virgin Islands, vol. 1, pt. 2, pp. 111-228. 1920.
t Incidentally it may here be pertinent to remark that I took occasion to refer to and
to discuss the identifications above mentioned, with both Professor Berry and Doctor
Knowlton, and that each merely recalled, rather vaguely, having seen the specimens,
but could furnish no additional information in regard to them. I also made diligent
effort to locate the specimens, in order to examine them personally, but without success.
177
SCIENTIFIC SURVEY OF PORTO RICO
The bed of iron ore, and the associated strata, were investigated
and critically examined by Dr. N. L. Britton and me in 1926,
and again by Doctor Britton in 1927 and 1928; but except for
a few small fragments and impressions of stems or branches and
disintegrated vegetable tissue, and some obscurely defined seed-
and fruit-like organisms, the rocks appeared to be barren of plant
remains. None of this material is identifiable further than as
above mentioned.
The matrix in which the remains are included consists in part
of thinly bedded black shale and 'in part of ferruginous shaly
sandstone, which are exposed on the side of the Aibonito-Barran-
quitas Road in the vicinity of Kilometer 3, and on the Aibonito-
Coamo Road between K. 86.5 and K. 90; and, more recently, an
investigation by Dr. H. N. Coryell, of fossiliferous limestone
from K. 89.8, on the road about half way between Aibonito and
Coamo, has revealed remains of coralline algae of the genus Litho-
thamnium in that rock; and in this connection it is of interest
to note that Hodge (loc. cit. (b), p. 155, fig. 15; p. 156, fig. 16;
p. 158, fig. 18), in his discussion of the Coamo Springs limestone,
described and figured remains of Lithothamnion (= Lithothamnium)
as constituents of the rock which he apparently regarded as Eocene
in age.
During the years 1915-16 Hubbard* was engaged in the geologi-
cal survey of the Lares district, during the prosecution of which
he made a considerable collection of fossil plants in the San Sebas-
tian or Collazo shales, on the Collazo River, and a few in the
Lares limestone, on the Guajataca River; but the only published
mention of these remains was a brief reference, in connection with
the discussion of the stratigraphy of the shales, where he remarked
(loc. cit., p. 39) that "many of the lignitic clays contain fossil
leaves and fresh or brackish water mollusca." The original
locality labels that accompany the specimens from the locality
last mentioned read: "Guajataca River, south bank, about one-
fourth mile north of town of Lares ('Collazo shales, uppermost of
the series')." Only three identifiable specimens were included in
the collection from this locality, which were described and figured
by met under the following names:
Hubbard, Bela. The geology of the Lares district, Porto Rico. New York
Acad. Sci. scientific survey of Porto Rico and the Virgin Islands, vol. 2, pt. 1, pp. 1-115.
1923.
t Hollick, Arthur. A review of the fossil flora of the West Indies, with descriptions
of new species. New York Bot. Gard., Bull. vol. 12. No. 45, pp. 259-323, pls. 1-15.
Sept. 13, 1924.
HOLLICK, PALEOBOTANY OF PORTO RICO
Coccoloba (?) sp. (loc. cit., p. 297).
Anona [= Annona] saraviana Berry? (idem, p. 298).
Aniba portoricensis Hollick (idem, p. 300).
These three species represent the only published record of the
flora of the particular geological horizon in which they were found.
The fossil plant locality on the Collazo River is designated by
Hubbard, on the labels, as "Collazo River, near (or "at") base of
second falls below Carretera bridge." The collection made by
Hubbard at this locality included twenty species that were
described and figured by me in the paper cited; and these will be
found included, in their proper systematic sequence, with the
descriptions of new species, in the present volume.
In the meantime, and subsequently, a considerable number
of specimens were collected at or near to Hubbard's locality on the
Collazo River, by Don Narciso Rabell, of San Sebastian, which
were transmitted to me for examination.
During the winter of 1926 Dr. N. L. Britton was in Porto
Rico, and took advantage of the opportunity to confer with Sefior
Rabell in regard to the material transmitted. The locality where
.the specimens were collected was visited, in company with Mr.
W. D. Noble of San Juan, and a brief period of examination and
collecting yielded such gratifying results that I was called upon to
come to Porto Rico and assist in the investigation. During the
month of March of that year the ravine of the Collazo River, from
the falls above the Lares-San Sebastian Road bridge to a distance
of about a kilometer below was explored, and fossil plant remains
were collected at four localities. At two of these localities, viz, at
the base of the first falls below the bridge (designated "station
B"), and at the base of the second falls (designated "Station A")
well defined fossil plant remains were collected in abundance.
The latter station is apparently identical with Hubbard's locality,
previously mentioned. Plant remains found at the other two
stations in the Collazo ravine consist mostly of twigs and branches,
disintegrated woody tissue, and fragments of bark and leaves,
none of which was identified. An exposure of lignitic clay was
also discovered in the ravine of Media Luna Creek, parallel with
and a short distance to the east of the Collazo River ravine. A
bed of finely comminuted lignitic debris is a prominent feature at
this locality. Specimens of this material were collected, but they
yielded nothing that was identifiable.
SCIENTIFIC SURVEY OF PORTO RICO
A brief account of the explorations made in 1926 was included
by Doctor Britton* in a report upon the general investigations
prosecuted in Porto Rico during that year, together with one by
me (idem, pp. 102-104) issued under the title Paleobotanical
exploration in Porto Rico.
During the year 1924, Sefior Adriano Gonzales, of Arecibo,
collected specimens of lignitic debris, including well defined nuts
of Juglans, in the valley of the Guajataca River near Quebradillas,
in deposits of recent (probably Post-Pliocene) age. This material
was subsequently submitted to me for examination and was made
the subject of a paper,t in which the nuts were described and figured
under the name Juglans archaeoantillana.
The species described and listed in the following pages were
all collected in the ravine of the Collazo River, either by Dr.
Bela Hubbard in 1915, by Don Narciso Rabell in succeeding years,
or personally and by other members of the expedition headed by
Dr. N. L. Britton in 1926.
The figures of leaves and other parts of existing species of
plants, introduced for comparison on the plates of this volume.
all represent specimens in the herbarium of the New York Botanical
Garden.
Britton, N. L. Further botanical investigation in Porto Rico. New York Bot.
Gard., Journ. vol. 27, No. 317, pp. 97-102. May, 1926.
t Hollick, Arthur. Fossil walnuts and lignite fi omPorto Rico. New York Bot. Gard.,
Journ. vol. 27, No. 322, pp. 223-227, text figs la, a', ib, bl, lc. c'. Oct. 1926.
DESCRIPTIONS OF SPECIES
Phylum THALLOPHYTA
Class ALGAE
Order DICTYOTALES
Family DICTYOTACEAE
Genus CHONDRITES Sternberg
Chondrites dictyotoides n. sp.
Plate 51, Figure 1.
Frond or thallus, irregularly branched; main branches about 1 millimeter
in maximum width, proximad; each successive series of branches narrower than
its supporting series; all ligulate in shape; ultimate ramifications truncate (?),
about .25 millimeter in width; costae not discernible.
Fragmentary remains of this species, including those figured, are scattered
over the surface of a shaly layer of matrix about 7.5 by 10 centimeters in extent.
They evidently represent dismembered parts of a thallophyte that bears a close
surficial resemblance to existing marine algae included in the family Dictyotaccae,
and may be especially compared with certain forms of Dictyota Bartayresii
Lamouroux-an existing species common in the West Indian region.
The fragmentary condition of our specimen renders impossible any satis-
factory concept either of its size or of its exact system of branching, in its entirety.
In places the branches are alternately disposed, and in others dichotomy is
apparent, and the spacing of the branches and their angles of divergence appear
to be more or less irregular. It is possible, however, that a specimen complete
in all its parts might show certain of these apparent characters to be due to
imperfection of preservation and to the fragmentary condition of the remains.
About twenty species and varieties, surficially similar to our specimens,
have been described and figured under the fossil algal genus Chondrites, especially
by Heer,* from the Eocene (Flysch) of Switzerland, among which Chondrites
Targioni arbuscula (Fischer-Ooster) Heer (op. cit., p. 155, pl. 61, fig. 9; pl. 62,
figs. 1-7; pl. 63, figs. 12-17), and Chondrites intricatus (Brongniart) Sternberg, in
its several forms (idem, p. 157, pl. 63, figs. 1-10), may be regarded as those
that are most closely comparable with ours. A tracing of one of the figures
of the former species, representing HIeer's figure 1, plate 62 (op. cit.), is reproduced
in our Figure 2, Plate 51, to facilitate comparison. The same author also de-
scribed similar remains from Alaska, designating them "Chondrites sp." (op.
cit., p. 21, pl. 10, fig. 5); but in as much as the specimen upon which the descrip-
tion and figure was based is stated to have been found "in lapide erratico" or
as "float rock," its geological age was not satisfactorily determined. Incidentally
also, comparison may be made with Sphaerococcus crispiformis (Schlotheim)
Heer, Oswald. Flora fossilis Helvetiae, pt. 5 (Die Eocene flora der Schweiz), pp.
147-182, pls. 59-69 in part. 1877.
t Heer, Oswald. Flora fossilis Arctica, vol. 2, pt. 2 (Flora fossilis Alaskana). 1871.
181
182 SCIENTIFIC SURVEY OF PORTO RICO
IIeer,* from the Miocene of the Old World, a tracing of which is reproduced in
our Figure 3, Plate 51.
Fragmentary remains of an alga from the Tertiary of Wyoming, that is
evidently of the same general type as those previously mentioned, were de-
scribed and figured by Lesquereuxt under the name Fucus lignilum, and it is
interesting to note that he compared it with Sphacrococcus crispiformis; and
Wardt described and figured similar remains from the same region and referred
them to Lesquercux' species, together with mention of its resemblance to S.
crispiformis and also to "some forms of Chondrites."
Locality: Station A.
Phylum PTERIDOPHYTA
Class FILIcNAE
Order POLYPODIALES
Family CYATHEACEAE
Genus HEMITELIA R. Brown (CNEMIDARIA Presl)
HEMITELIA BRANNERI Hollick & Berry?
Plate 51, Figure 4
Hemitelia Branneri Hollick & Berry. Johns Hopkins Univ. Studies in
Geology No. 5 (A late Tertiary flora from Bahia, Brazil), p. 46, pl. 1, figs. 3-6.
1924.
A single median fragment of a pinnule, or a pinnatifid division of a pinna,
is all that we have thus far found to represent ferns in the fossil flora of Porto
Rico; otherwise the specimen here described and figured would probably have
received but little attention. The nervation is poorly defined, and it is impossible
to determine if the veinlets are forked or simple. The sori appear to be round
and to occupy a position about midway between the midvein and the margin.
The margin is minutely denticulate.
In its general characters, as far as these are discernible, the specimen is
highly suggestive of certain existing species in the genus Hemitelia, and it bears
so close a resemblance to the fossil species H. Branncrii (op. cit.), from the
Tertiary of Brazil, that I have referred it, tentatively, to that species-the
only apparent difference between them being that the specimen from Porto
Rico is narrower and the sori are somewhat larger, as may be seen by comparison
with Figures 5 and 6, Plate 51, which represent figures 5 and 4, respectively, of
the work cited.
The species, based upon the specimen from Brazil, was compared by the
authors with the existing Hemitelia grandifolia (Wildenow) Sprengel; and illustra-
tions of pinna lobes of the existing species H. subglobosa (Underwood) Maxon
(Figure 7, Plate 51) and H. Imrayana Hooker (Figure 8, Plate 51) are introduced
for generic comparison with our specimen.
Locality: Station A.
Heer, Oswald. Flora Tertiaria Helvetiae, vol. 1, p. 23, pl. 4, fig. 1. 1S55.
f Lesquereux, Leo. U. S. Geol. Survey Terr., Rept. vol. 7 (The Tertiary flora), p.
42, pl. 61, figs. 24, 24a. 1878.
1 Ward, L. F. Synopsis of the flora of the Laramie group. U. S. Geol. Survey,
Sixth Ann. Rept. 1884-85, p. 549, pl. 31, figs. 1, 2. 1885. Types of the Lararnie flora.
Idem, Bull. No. 37, p. 13, pl. 1, figs. 1, 2. 1887.
HOLLICK, PALEOBOTANY OF PORTO RICO 183
Class LYCOPODINAE
Order ISOETALES
Family ISOETACEAE
Genus ISOETES Linnaeus
Isoetes (?) incerta n. sp.
Plate 52, Figures 1, 2
Organism apparently consisting of a bundle of linear leaves, united or
closely compacted at their bases; leaves about 1 to 1.5 millimeter in width
by 6 or more centimeters in length.
The two figures that serve to illustrate this organism represent counterparts
of one and the same specimen; but the impression of only the upper portion of
the specimen is represented by Figure 2. In the specimen represented by
Figure 1 the base consists of a relatively thick, more or less disintegrated mass
of carbonaceous material, impossible to differentiate, visually, into any definite
form or structure. It was examined and studied from the viewpoint of its
possibility as a sheath, as in Pinus, and as an aggregation of ligulae, as in Isoetes,
but without arriving at any satisfactory conclusions. Another possibility
appeared to be indicated by its general surficial resemblance to a partly dis-
integrated monocotyledonous leaf that had split and separated longitudinally
along the lines of the nervation, thus presenting an appearance simulating
several narrow, linear leaves. However problematic its taxonomic position
might be, the specimen, nevertheless, could not be ignored or disregarded in
any illustrated, descriptive list of the flora in which it was an element; but its
reference to the genus 1soetes is to be regarded as tentative.
For purposes of generic' comparison I have introduced (Figure 3, Plate 52)
a drawing of a herbarium specimen of the existing Isoetes riparia Englemann,
the resemblance of which, to the figures of our specimen, is as remarkable as it
is interesting.
Only one fossil species, Isoetes brevifolia Lesquereux,* from the mid-Tertiary
of Florissant, Colorado, has been heretofore recorded from anywhere in the
New World, as far as I am aware; but unfortunately it was not figured. The
author, however, mentioned its resemblance to lsoRtes Braunii (Unger) Ileert;
and comparison between our figures and certain of those by IIcer, especially
his figure 6 (op. cit.), reproduced in our Figure 4, Plate 52, may be seen to be
quite striking.
Locality: Station B.
Type specimen: Figure 1.
Phylum SPERMATOPHYTA
Class GYMNOSPERMAE
Order CYCADALES
Family CYCADACEAE
Lesquereux, Leo. U. S. Geol. Survey Terr., Rept. vol. 8 (The Cretaceous and
Tertiary floras), p. 136. 1883.
t Heer, Oswald. Flora Tertiaria Helvetiac, vol. 1, p. 44, pl. 14, figs. 1-7. 1855.
SCIENTIFIC SURVEY OF PORTO RICO
Genus ZAMIA Linnaeus
Zamia collazoansis n. sp.
Plate 53, Figures 1, 3, 5, (7?)
Leaflets broadly spatulate-obovate, truncate and more or less constricted
at the base, 3.5 to 4 centimeters in length by 2 to 3.5 centimeters in maximum
width; margins entire laterally, obscurely finely notched-dentate apically;
nervation sub-parallel, somewhat flabellate, closely approximated or confluent
at the base, more widely spaced above, uniform and apparently simple through-
out, about fifteen per centimeter, measured across the widest parts of the leaflets.
These specimens apparently represent proximal leaflets of a cycad that may
be compared with certain existing species of Zamia native in the West Indies and
adjacent regions, such as Z. integrifolia Jacquin (= Z. latifoliolata Preneloup),
represented by Figure 2, Plate 53, and Z. salicina Britton, represented by Figures
4 and 6, Plate 53, introduced for comparison.
Figure 1 may be regarded as representing a normal, average sized specimen,
comparable with the leaflet of Z. integrifolia depicted in the adjacent Figure 2;
Figure 3 with the leaflet of Z. salicina shown in Figure 4; Figure 5 with a dis-
torted leaflet of the latter species shown in Figure 6; and Figure 7 with the leaflet
of an unnamed species from Cuba, represented by Figure 8. I am somewhat in
doubt as to whether or not the specimen represented by Figure 7 should be in-
cluded in the same species with the others, as it is obviously more flabelliform in
shape, and has a system of nervation that is spaced somewhat more widely than
is apparent in the other specimens; but the shape of the leaflet differs but little
more from the average normal type than do many of the leaflets that might be
selected from among those of the existing species utilized for comparison.
It may be noted that in all of our specimens the bases are broader, and not so
conspicuously constricted as they appear to be in connection with the leaflets of
species of Zamia in general, thus simulating, in this particular, a feature that is
more characteristic of certain other genera-Microcycas, Dioon, etc. Whether
or not this feature has any generic or morphologic significance, however, is not
apparent.
Only two New World species of Tertiary cycads have heretofore been describ-
ed and figured, viz. Zamia tertiaria Engelhardt* from Chile, and Zamia (?) Wilcox-
ensis Berryf, from Louisiana. The figure of the former is reproduced, for com-
parison, in our Figure 12, Plate 53, and the figure of the latter in our Figure 13,
Plate 53.
Engelhardt's discussion of his species indicates that he regarded it as allied
to the same foliar type of cycad as that with which ours appears to be most
closely comparable. Thus he remarked that "es ist nur ein Fiederbliittchen
vorhanden, das jedoch so sehr mit denen von Zamia integrifolia Ait. [ = Z. floridana
De Candolle ?], tibereinstimmt, das ich nicht z6gere, diese Pflanze als ndchste
Verwandte anzusehen, wenigstens so lange, bis uns vollstandig erhaltenes Material
eines Besseren belehrt." His figure is more or less conventionalized, with
parallel nervation; but it appears as if meant to represent a leaflet very similar to
our specimens.
*Engelhardt, Hermann. Ueber Tertiarpflanzen von Chile. Senckenb. Naturf.
Gesellsch., Abl. vol. 16, No. 4, p. 646, pl. 2; fig. 16. 1891.
1tBerry, E.W. The lower Eocene floras of southeastern North America. U.S.
Geol. Survey, Prof. Paper 91, p. 169, pl. 114, fig. 2. 1916.
HOLLICK, PALEOBOTANY OF PORTO RICO
The specimen described and figured by Berry (loc. cit.) obviously represents
a cycad leaflet, but it is too fragmentary for any but tentative generic identifica-
tion. It appears to represent the basal portion of a longer and narrower leaflet
than any that I have included under Zamia collazoensis, and it may, perhaps, be
more nearly comparable with leaflets of the species next described.
In the existing flora the genus Zamia includes about thirty recognized species,
all of them natives of tropical and sub-tropical regions in the New World, extend-
ing northward into Florida.
Locality: Station B.
Type specimen: Figure 1.
Zamia Noblei n. sp.
Plate 53, Figures 9, 10; Plate 54, Figures 1, 3a; Plate 55, Figures 1-3, 4a, 5a.
Leaflets linear-obovate or clavate in shape, 5 to 9 centimeters in length by
1.25 to 2.25 centimeters in maximum width; margin entire throughout and, for
the most part, apparently, involute; nervation sub-parallel; nerves closely
approximated or confluent at the base, more distinctly separated and spaced
above, uniform and, apparently, simple throughout, approximately fifteen per
centimeter, measured across the widest parts of the leaflets.
The relationship of the specimens above described to the Cycadaceae appears
to be obvious, and some of them bear such a close resemblance to leaflets of
certain existing species of Zamia that this generic relationship appears to be
reasonably certain.
For purposes of comparison I have introduced a drawing of a leaflet of the
existing Zamia pumila Linnaeus (Figure 11, Plate 53), which may be compared
with our Figure 9 on Plate 53; and one each of Z. integrifolia Jacquin (=Z.
latifoliolata Preneloup) (Figure 2, Plate 54), and Z. umbrosa Small (Figure 5, Plate
54), which may be compared, respectively, with our Figures 1 and 3a on Plate 54.
I am in some doubt as to whether or not the specimen represented by Figure 1 on
Plate 54 should be included in the same species with the others, as it appears to
possess a somewhat coarser nervation; but otherwise their respective features
appear to be identical.
Other fragmentary remains, represented by Figure 10 on Plate 53, and by
Figures 1, 2, 3, 4a, and 5a, on Plate 55, may or may not all belong to one and the
same species; but they are not sufficiently complete for satisfactory differentiation.
As previously remarked only two species of cycads of Tertiary age have been
heretofore recorded from the New World; and of these Zamia (?) Wilcoxensis
Berry,* appears to be the only one that might, possibly, be inferred to be identical
with Z. Noblei. Berry's figure (loc. cit.) is reproduced in Figure 13, Plate 53, and
it may be seen that it might, perhaps, be regarded as representing the proximal
part of a leaflet, comparable in its general characters with a similar foliaceous
fragment of Z. Noblei.
It may also be noted that, as mentioned in connection with Zamia collazoensis,
in all of our specimens the bases are broader and are not as constricted as they are
in existing species of Zamia; but this feature may be more apparent than real,
and attention is called to the fact merely in order that it may not be overlooked In
connection with the study of any further material that may be brought to light in
the future.
*Berry, E. W. U. S. Geol. Survey, Prof. Paper 91, p. 169, pl. 114, fig. 2. 1916.
SCIENTIFIC SURVEY OF PORTO RICO
Incidentally, also, it may be of interest to compare certain figures of our
specimens, such as Figure 4a, Plate 55, with a similar one of a specimen from the
Tertiary of Chile, figured by Engelhardt,* designated "ttIonokotyler Blattrest"
(loc. cit., p. 687), and described as follows (loc. cil., p. 686): "Taf. I, Fig. 4 bilde
Ich einen Pflanzenrest ab, dcr auf eine monokotyle Pflanze hinweist; doch diirfte
es kaum m6glich sein, ihn in eine bestimmte Familie einzureihen, weshalb ich
ohne Namen lasse. Vielleicht dass splttere Funde Klarheit Uber ihu zu bringen
vermSgen. And in this connection it is interesting to note that Berryt described
and figured specimens of cycad-like leaflets from the Tertiary of Chile, which he
identified as Zamia tertiaria Engelhardt,$ and under that name included also
Engelhardt's "Monokotyler Blattrest" above mentioned. In no instance,
however, Is a leaflet depicted intact; and in those figures in which the missing
extremities are restored these are indicated as tapering and acute, which characters
alone would serve to differentiate them from our specimens from Porto Rico.
In order to illustrate the preceding citations and remarks one of Berry's figures
is reproduced in our Figure 1, Plate 88. It represents the initial figure (not
numbered) on his plate 2.
Lesquereux and Ward also described and figured somewhat similar remains,
from the Eocene of Montana, and referred them to Phragmiles alaskana Ileer.
Certain of these figures compare quite closely with certain of ours; but the de-
scriptions, and enlarged details of the nervation, show the nerves to consist of a
major and a minor series, whereas in ours all the nerves appear to be of equal rank.
The specific name is in honor of MSr. W.D. Noble, to whom we are indebted
for valuable assistance in connection with the work of collecting the specimens.
Locality: Station B (Figures 9, 10. Plate 53; Figures 1,3a, Plate 54; Figures 3,
4a, 5a, Plate 55). Station A (Figures 1, 2, Plate 55).
Type specimen = Figure 3a, Plate 54.
Class ANGlOSPERMAE
Sub-Class \MONOCOTYLEDONAE
Order ARECALES
Family ARECACEAE [PALMACEAE]
Genus BACTRIS Jacquin
Bactris Pseudocuesco n. sp.
Plate 56, Figures 18, 19
Fruits spheroidal or slightly oblate-spheroidal in shape, about 1.75 centi-
meter in diameter; nut apparently marked with obscurely defined, longitudinal
striae; exocarp relatively thick, apparently originally coriaceous or chitonous in
texture.
*Engelhardt, Hermann. Ueber Tcrtiarpflanzen von Chile. Senckenb. Naturf.
Gesellsch., Abh. vol. 16, No. 4, pl. 1, fig. 4. 1891.
-Hlerry, E. W. Contributions to the paleobotany of Peru, Bolivia and Chile. Johns
Hopkins Univ. Studies in Geol. No. 43 (The flora of the Concepcion-Arauco coal
measures of Chile. George Huntington Williams Memorial Publication No, 17.), p. 120,
pl. 1, fig. 4; pl. 2, figs. 1-3 [not numbered on plate]. 1922.
:Engelhardt, Hermann. Ueber Tertitrpilanzen von Chile. Senckenb. Naturf.
Gesellsch., Abh. vol. 16, No. 4, p. 646, pl. 2, fig. 16. 1891.
ILesquereux, Leo. U. S. Geol. Survey. Terr., Rept. vol. 7 (The Tertiary flora), p.90,
pl. 8, figs. 10-12. 1878.
Ward, L. F. Synopsis of the flora of the Laramie group. U. S. Geol. Survey,
Sixth Ann. Rept. 1884-85, p. 550, pl. 32, figs. 1-3. 1885. Types of the Laramie flora.
Idem, Bull. No. 37, p. 17, pl. 3, figs. 1-3. 1887.
HOLLICK, PALEOBOTANY OF PORTO RICO
These two specimens-one (Figure 18) with a portion of the exocarp
attached, and the other (Figure 19) with the exocarp entirely gone-resemble so
closely the fruit of certain species of the existing genus Bactris that generic
relationship appears to be definitely indicated; and specifically they can hardly
be distinguished from the fruit of Bactris Ciesco Crueger, as may be seen by com-
paring Figures 18 and 19, respectively, with Figures 20 and 21 on the same plate,
the latter representing specimens of B. Cuesco from the island of Trinidad.
They may also be compared with Palmocarpon (?) globosum Lesquereux,* from
the Tertiary (Oligocene) of Colorado (See our Figure 22, Plate 56, reproduced
from Lesquereux' figure), and with certain of the specimens from the Tertiary
(Eocene) of Wyoming, referred by the same author to Carpiles lineatus?,
Newby." (= Carpolithes lineatus Newberry)t from the Eocene of Dakota-a
reference, however, which does not appear to be justified by comparison between
the figures.
It may also be of interest to refer, in connection with the several species
above mentioned, to one from the Tertiary of the island of Trinidad, originally
described and figured by me' under the name Palmocarpon bactrioides, and
subsequently found in the same locality by Berry. Reference is here made
to this species merely in order to facilitate comparison by those who may be
interested in the subject, especially in view of the remark by the latter author
(loc. cit.) that "it might well be suggested that the fossil should be referred
directly to this genus [Bactris]."
Locality: Station B (Figure 18); Collazo River. Collected by Don Narciso
Rabell (Figure 19).
Type specimen = Figure 18.
Genus IR1ARTEA Ruiz & Pavon
Iriartea collazoensis n. sp.
Plate 56, Figure 1.
Nut rounded, unsymmetrically oblong-spatulate or balloon shaped, about 4
centimeters in length by 1.5 centimeter in maximum width; surface of the endo-
carp marked by relatively coarse, longitudinally extended reticulations.
This specimen possesses characters that closely simulate those of certain
species in the existing genera Arenga and Iriartca, as may be seen by comparing
the figure of our fossil with Figure 2, on Plate 56, representing a nut of Arenga
saccharifera Labillardiere, and with Figure 3 on the same plate, representing a
nut of Iriartea setigera Martius. Such comparison shows so close a resemblance,
in the shape and external markings of the nuts, between the fossil and these two
existing species that generic relationship between it and one or the other of them
would appear to be satisfactorily indicated. The species first mentioned is more
*Lesquereux, Leo. U. S. Geol. Survey Terr., Rept. vol. 8 (The Cretaceous and
Tertiary floras), p. 144, pl. 24, fig. 3. 1883.
tLesquereux, Leo. U. S. Geol. Survey Terr., Rept. vol. 7 (The Tertiary flora), p. 302,
pl. 60, figs. 1 (in part), lb, 1c, Id. 1878.
+Ncwherry, J. S. U. S. Geol. Survey, Mon. vol. 35 (The later extinct floras of North
America), p. 138, pl. 40, fig. 1. 1808.
Ilollick, Arthur. A review of the fossil flora of the West Indies, with descriptions
of new species. New York Blot. Gard., Bull. vol. 12, No. 45, p. 286, pl. 1, fig. 4. Sept. 13,
1924.
Berry, E. WV. Johns Hopkins Univ. Studies in Geology, No. 63. (The Tertiary flora
of the island of Trinidad, B. W. I. George Huntington Williams Memorial Publication No.
22), p. 79, pl. 1, fig. 1. 1925.
SCIENTIFIC SURVEY OF PORTO RICO
nearly comparable with it in size, but this species, as well as the genus to which it
belongs, is exclusively of Old World distribution, whereas Iriartea is a New
World genus. The factor of geographic distribution would seem to demand
more serious consideration and inference than that of mere difference in size, and
therefore I have thought it more logical to regard our fossil as belonging to
Iriartea rather than to Arenga. If any trace of a germ pore were discernible its
location might serve to identify the generic relationship of the fossil more
definitely.
A fossil palm fruit, from the Tertiary of the Canal Zone, was described and
figured by Berry* under the name Iriartites Vaughanii. The reference of the
fruit to the fossil genus Iriartites was stated by the author to be on account of its
resemblance to the fruit of certain existing genera in the tribe Iriarteae; but,
unfortunately, no specific examples were cited, although the genera Iriartea and
Astrocaryum were mentioned. Incidentally, also, it may be pertinent to remark
that the original generic description of the genus Iriartites was based by the
author exclusively upon fragmentary remains of palm leaves, described and
figured under the name Iriartites tumbezensis. In any event, however, no generic
relationship is apparent between our specimen and Iriartites Vaughanii Berry, as
may be seen by comparison between Figure 1, Plate 56, representing our specimen,
and Figure 23 on the same plate, representing a reproduction of Berry's figure 2
(loc. cit.), which is described as one half the natural size.
Locality: Collazo River. Collected by Don Narciso Rabell.
Genus MANICARIA Gaertner
Manicaria portoricensis n. sp.
Plate 56, Figures 24a, 24b.
Fruits globose and 2.5 centimeters in diameter, or oblong-globose and 3
centimeters in length by 2 centimeters in width; nut marked with fine, thread-
like striae on the surface of the endocarp; exocarp apparently thick and rugose.
These two specimens may have been, originally, more nearly alike in shape
than their actual appearance might seem to indicate. The one represented by
Figure 24b, for example, may be seen to be fractured and, apparently, is more or
less distorted, and may have been, in its entirety, more globose-oblong than
oblong-globose in shape. The surficial striae on the endocarps may be assumed
to have a proximad-distad direction; but such assumption, in the absence of any
visible indication of a pore or hilum, does not assist in locating either extremity.
Comparison of our specimens with the fruit of the existing genus Manicaria
shows such a close resemblance between them that generic identity appears to be
satisfactorily indicated. Figure 25, Plate 56, represents a fruit of the West
Indian species Manicaria sacchifera Gaertner, with the exocarp partly removed,
exposing the smooth, finely striated surface of one side of the nut, with the rough
broken edge of the exocarp surrounding it-the figure as a whole, except for its
larger size, showing a striking resemblance to the figures of the fossils.
Locality: Station A.
Type specimen = Figure 24a.
*Berry, E. AW. A palm nut from the Miocene of the Canal Zone. U. S. Nat. S1us.,
Proc. vol. 59 (No. 2356), pp. 21-22, figs. 1-3. 1921.
tBerry, E. W. Aliocene fossil plants from Northern Peru. UT. S. Nat. MIus., Proc.
vol. 55 (No. 2270), p. 285, pl. 14. 1919.
IIOLLICK, PALEOBOTANY OF PORTO RICO
Genus PALMOCARPON Lesquereux
PALMOCARPON ACROCOMIOIDES Hollick
Plate 56, Figure 4.
Palmocarpon acrocomioides Hollick. A review of the fossil flora of the West
Indies, with descriptions of new species. New York Bot. Gard., Bull. vol. 12,
No. 45, p. 287, pl. 13, fig. 4. Sept. 13, 1924.
This specimen appears, surficially, to be generically identical with the one
described and figured by me (loc. cit.) under the above name, from the Tertiary of
Santo Domingo, island of Hispaniola; and no characters are discernible that
might serve to differentiate them specifically, one from the other. They resemble
compressed, globose fruits of certain species of the existing genus Acrocomia,
with the exocarp more or less fractured, as may be seen by comparison with
Figure 5, Plate 50, which represents a fractured specimen of Acrocomia crispa
(Humboldt, Bonpland, and Kunth) C. F. Baker.
Somewhat similar fossils, but more elongated in shape, from the Tertiary of
Switzerland, were described and figured by Heer* under the name Palmacites
[Antholithes] Martii (see our Figure 7, Plate 56, representing Heer's figure 3, loc.
cit.): but beyond the surficial mutual resemblance to palm fruits with thin, brittle
exocarps, the apparently near relationship between the specimens from Switzer-
land and those from the West Indies need not be further discussed.
It may also be pertinent here to refer, incidentally, to certain fruits from the
Tertiary of Chile, described and figured by Engelhardtt under the name Carpoli-
thes cycaeformis, which he regarded as suggestive of the genus Cycas. The largest
one, described as "rundlich, etwas zuzammengepresst, may be seen to bear more
or less of a resemblance to our figures of Palmocarpon acrocomioides, and to
simulate, perhaps a little more closely, Heer's Palmacites Martii previously cited.
Locality: Station A.
Palmocarpon cetera n. sp.
Plate 56, Figure 10.
Fruit rounded, oblong-ovate in shape, somewhat compressed or flattened,
2.5 centimeters in length by 1.75 centimeter in maximum width, at about the
middle; endocarp obscurely and finely roughened or wrinkled latitudinally and
marked longitudinally with faintly defined striae.
It is possible that this specimen might be regarded as specifically Identical
with Palmocarpon exemplary, the species next described (see Figures 8, 9, Plate 56).
It may be seen that they present a similar mutual resemblance; but minor char-
acters serve to differentiate the specimen under discussion from the two upon which
the species mentioned is based.
Relationship is surficially indicated with either one or the other of the
existing genera Astrocaryum or Cocos, but the absence of any trace of a germ pore
In our specimen renders accurate or satisfactory comparison impossible. In
connection with the former genus the pores are distad, as shown in Figure 16,
Plate 56, which represents a decorticated fruit of the Brazilian species Astrocaryum
Jauarii Martius, while in Cocos they are proximad, as shown in Figure 17, Plate
*Heer, Oswald. Flora Tertiaria Helvetiae, vol. 1, p. 97, pl. 41, figs. 1-4. 1855.
tEngelhardt, Hermann. Ueber Tertiarpflanzen von Chile. Senckenb. Naturf.
Gesellsch., Abh. vol. 16, p. 685, pl. 13, fig. 1. 1891.
190 SCIENTIFIC SURVEY OF PORTO RICO
56, which represents a specimen of Cocos eriospatha Martius, with its exocarp
partly removed.
In the same general category with the fossil and existing species above
discussed we may be justified in including a specimen from the lower Tertiary of
New Mexico, described and figured by Lesquereux* under the name Palmo-
carpon mexicanum, reproduced in our Figure 14, Plate 56, which he apparently
represented with the distal extremity downward, as may be seen by comparison
with Figure 15, Plate 50, representing a fruit of Cocos Datil Grisebach & Drude
with the distal extremity upward. From the foregoing discussion it may be
appreciated that surface markings which might be indicative of pores in any
specimens of fossil palm nuts might well be of considerable importance as generic
indices.
Locality: Station A.
Palmocarpon exemplare n. sp.
Plate 56, Figures 8, 9.
Fruits rounded, ovoid in shape, about 2.5 centimeters in length by 2 centi-
meters in maximum width; exocarp apparently relatively thin and smooth; nut
apparently smooth.
These fruits, as in the case of the species last described, might be more or less
satisfactorily compared with those of certain existing genera, such for example, as
Astrocaryum and Cocos. It is not possible, however, to determine which is the
distal and which is the proximal extremity in either one of our specimens, nor is
there any indication discernible of the location of any germ .pores; and whether
the consistency of the exocarp was originally either coriaceous, or chitonous, or
fibrous is not apparent. In view of these conditions I have deemed it best to
include our specimens, for the time being, in the comprehensive genus Palmo-
carpon, pending the possible future discovery of specimens with better preserved
and more clearly defined characters which might serve to identify them, generi-
cally, in a more satisfactory manner.
For purposes of comparison 1 have introduced (Figure 6, Plate .56) a drawing
representing a fruit of the Brazilian species Cocos Inarjai Trail, with exocarp
intact and the proximal extremity downward, which, except for its somewhat
larger size, may be seen to bear a suggestive similarity to our specimens. Also,
further reference may not be here out of place to Figure 7, Plate 56, representing
the fossil species Palmacites [Antholithes] Mcartii Heer,t from the Tertiary of
Switzerland, introduced primarily for purposes of comparison with Palmocarpon
acrocomioides Hollick, described on page 189.
Locality: Station B (Figure 8); Station A (Figure 9).
Type specimen = Figure 9.
Palmocarpon opinabile n. sp.
Plate 56, Figure 24c
Impression of a broadly spatulate or balloon shaped fruit, about 3 centimeters
in length by 2.5 centimeters in maximum width, apparently with a fibrous or
shreddy exocarp.
*Lesquercux, Leo. U. S. Geol. Survey Terr., Iept. vol. 7 (The Tertiary flora), p. 119,
pl. 11, fig. 5. 1878.
tleer, Oswald. Flora Tertiaria Ilelvetiae, vol. 1, p. 97, pl. 41, fig. 3. 1855.
HOLLICK, PALEOBOTANY OF PORTO RICO
It is with a feeling of more or less uncertainty that I have included this
specimen under the generic name that implies relationship with the palms; but
that it was originally represented by some thick, rounded organism, such as a
nut-like fruit, appears to be certain, as indicated by the pronounced concavity of
the impression; and in this connection the fact that it occurs in the same piece of
matrix, and in close proximity to the fruits previously described on page 188 under
the name Manicaria portoriccnsis, may possess a certain significance.
It is also suggestive of a specimen from the Eocene Tertiary of Mississippi,
figured by Berry* and referred to Nipadites umbonatus Bowerbank-an Old
World Tertiary species, supposed to be allied to the existing Old World genus
Nipa. Reference to this species, however:, should be here regarded merely as
suggesting surficial resemblance, and in order to facilitate comparison for those
who might be interested in the further study of similar fossil remains.
Locality: Station A.
Palmocarpon Rabellii n. sp.
Plate 56, Figure 11
Nut unsymmetrically ovate in shape, 1.8 centimeter in length by 1.4 centi-
meter in maximum width, marked by widely spaced, longitudinal striae.
I have been unable to arrive at any satisfactory conclusion in regard to the
probable generic relationship of. this specimen. Resemblance to the genus
Elacis may be noted, but this genus, in our existing flora, is of Old World distribu-
tion. In order, however, that the interesting resemblance may be visualized I
have introduced (Figure 12, Plate 56) a figure of a nut of Elaeis guinensis
Jacquin. In this genus the pores are distad, and in the figure the distal extremity
of the nut is represented as downward. A New World genus with which compar-
ison may be made is Copernicia, and for this purpose I have introduced (Figure 13,
Plate 56) a figure that represents a nut of Copcrnicia cerifera Martius, with the
distal extremity upward. In this genus the nuts are without pores. Any
indication of pores are not apparent in our specimen, nor is there any trace of a
hilum, hence it is impossible to determine or even to infer, satisfactorily, which is
the distal and which is its proximal extremity. In the circumstances it would not
seem to be justifiable, therefore, to suggest, through the medium of a name, an
assumed relationship of our specimen with any existing genus or species. The
generic appellation Palmocarpon is indicative of its general taxonomic and
morphologic status, and the collector may properly be recognized in connection
with the specific name.
Locality: Collazo River. Collected by Don Narciso Rabell,
Genus PALMACITES Brongniart
Palmacites alius, n. sp.
Plate 57, Figure 1.
Fragment of a finely striated woody part of a monocotyledonous plant,
surficially resembling a flattened piece of a petiole of a palm leaf.
This fragmentary specimen may not appear to be worthy of description and
illustration; but similar remains of Monocotyledonae are so few, in the col-
*Berry, E. W. U. S. Geol. Survey, Prof. Paper 91 (The lower Eocene floras of south-
eastern North America), p. 176, pi. 112, fig. 13. 1916.
192 SCIENTIFIC SURVEY OF PORTO RICO
elections of fossil plants from Porto Rico, that every fragment is of interest, and
may be of value in eventually piecing out and building up more or less complete
specimens of the plants of which they are parts.
A somewhat similar fragment, from the same region, was described and
figured by me* under the name "Palmophyllum sp. (fragment of petiole)?"; but
It would be hazardous to infer specific or generic identity between them.
Locality: Collazo River. Collected by Don Narciso Rabell.
Palmacites conformis n. sp.
Plate 7, Figure 2
Fragment of a relatively coarsely striated culm or leaf of a monocotyledonous
plant; dimensions not known.
This specimen may represent merely a smaller part of a plant of the same
species as the one next described (P. sparsistriatus), in as much as they are both
from the same station, and the only apparent difference between them, as may be
seen by comparison of the two figures, is that of size.
Possibly future discoveries may bring to light additional material that will
enable us to determine, more satisfactorily, the taxonomic position and the
morphologic characters of the remains.
Locality: Station A.
Palmacites sparsistriatus n. sp.
Plate 58
Fragment of a trunk, stem, or other part of a monocotyledonous plant,
apparently palmaceous, coarsely striated longitudinally, the interspaces varying
in width from 2 to 3 millimeters.
Beyond the fact that this fragment evidently represents remains of the woody
part of a monocotyledonous plant its taxonomic status is problematic, and the
generic name under which it is placed may be regarded merely as indicative of
its assumed family relationship, and as indicating close similarity in appearance
to certain other fossil plant remains that have been described and figured under
the generic name Palmacites, such as P. canaliculatus Heer,+ from the Miocene
of Switzerland, in connection with which the author discusses the general re-
semblance of the striated surface to that of the Paleozoic genera Sigillaria and
Calamites, but also remarks upon the impossibility of generic or family relation-
ship with either, by reason of the absence, in tire Tertiary species, of both leaf
scars and nodes.
Locality: Station A.
Genus PALMOPHYLLUM Conwcntz
PALMOPHYLLUM sp, Hollick'
Palmophyllum sp. Hollick. A review of the fossil flora of the West Indies,
with descriptions of new species. New York Bot. Gard., Bull. vol. 12, No. 45,
p. 285, pl. 9, fig. 1. Sept. 13, 1924.
*Hollick, Arthur. A review of the fossil flora of the West Indies, with the descriptions
of new species. New York Bot. Gard., Bull. vol. 12, No. 45, p. 286, pl. 9, fig. 2. Sept.
13, 1924.
fileer, Oswald. Flora Tertiaria Ilelvetiae, vol. 1, p. 95, pl. 40, figs. 2a-2d, 3a, 3b.
1855.
HOLLICK, PALEOBOTANY OF PORTO RICO
Locality: Collazo River, near base of second falls below Carretera bridge.
Collected by Bela Hubbard.
PALMOPHYLLUM sp. (fragment of petiole)? Hollick
Palmophyllum sp. (fragment of petiole)? Hollick. A review of the fossil
flora of the West Indies, with descriptions of new species. New York Bot. Gard.,
Bull. vol. 12, No. 45, p. 386, pl. 9, fig. 2. Sept. 13, 1924.
Locality: Collazo River, near base of second falls below Carretera bridge.
Collected by Bela Hubbard.
Order SCITAMINALES
Family MUSACEAE
Genus MIUSOPHYLLUM Goeppert
MUSOPHYLLUM Sp.
Plate 61, Figure 1.
Leaf of unknown shape and dimensions, with a thick midrib and fine,
parallel secondary nerves, about 5 millimeters distant from each other, that
subtend angles of about 650 with the midrib.
This leaf fragment almost certainly represents the genus Musophyllum,
although only a few of the secondary nerves are discernible. The probability is
that the leaf, in its entirety, possessed a major and a minor series of secondary
nerves and that no traces of the latter were preserved. The specimen apparently
represents the median or lower part of a leaf, where the nerves subtend more
obtuse angles with the midrib than do those toward the apex. It is comparable
with Mausophyllum trinitense Hollick,* from the Tertiary of the island Trinidad;
Ml. elegans Engelhardt,t from the Tertiary of Colombia; and a specimen of the
latter species from the Tertiary of Venezuela, described and figured by Berry:
under the name Heliconia elegans (Engelhardt) Berry, in connection with which
he remarked: "Aside from the actual resemblance between these fossil American
forms and the existing Heliconias, it seems to me that general considerations
point to the conclusion that the genus Musa was never present in the Western
Hemisphere [in Tertiary time], despite the fact that it flourishes so greatly under
cultivation in the American Tropics at the present time. Engelhardt (loc. cit.,
pp. 25-26) also expressed doubt in regard to the actual relationship of his species
with the genus Musa in the following words: "Die Stticke entstammen jedenfals
einer Art der Gattung Heliconia L., welche Sildamerika eigenthiimlich ist,
wenigstens liefs sich aus der Nervatur trotz vieler Vergleichungen nicht nach-
weisen, ob sie zu dieser oder zu Musa L. zu rechnen zei, weshalb ich mich der
provisorischen Bezeichnung [Musophyllum] bediente. "
In my discussion of the probable generic relationship of AMusophyllum trini-
tense (loc. cit., p. 288) I commented as follows: "This specimen . has so close
a superficial resemblance to leaves of the existing genus Mesa that I would have
*Hollick, Arthur. A review of the fossil flora of the West Indies, with descriptions of
new species. New lork Bot. Gard., Bull. vol. 12, No. 45, p. 288, pl. 1, fig. 2. Sept. 13,
1924.
-Engethardt, Hermann. Ueber neue Tertidrpflanzen Sild-Amerikas. Senckenb.
Naturf. Gesellsch., Abh. vol. 19, p. 25, pl. 4, figs. 1-3; pl. 5, fig. 1. 1895.
tBerr., E. W. Tertiary fossil plants from Venezuela. U. S. Nat. Mus., Proc. vol.
59 (No. 2388), p. 560, text fig. 1 (p. 561). 1921.
194 SCIENTIFIC SURVEY OF PORTO RICO
but little hesitation in so referring it were it not for the doubt that is generally
entertained, and the uncertainty that obtains, in regard to the New World
nativity of the genus. "
It may be recalled, however, that Alexander von Humboldt argued for the
probable American nativity of Mesa; and, more recently, Cook* discussed
the matter in a favorable vein. Definite proof was lacking, however, until the
discovery of undoubted banana seeds in Tertiary (Oligocene?) coal beds in
Colombia, which were described by Berryt and commented upon as follows:
"The present discovery shows that Musa was an undoubted member of American
Tertiary floras; and although it does not prove that the native banana was brought
into cultivation by the aborigines in the Western Hemisphere, it lends probability
to such a belief, and in a measure serves to substantiate the statement of the Peru-
vian author Garcellasso de la Vega (1530-1568) that the banana was one of the
staples of the aborigines before the discovery of America; and that of Montesinos
(1527), as quoted by the historian Prescott, that the natives of Tumbez brought
bananas as an offering to Pizarro when he disembarked. "
Finally, Bassler4 described a seed-bearing species of Musa, found growing
in eastern Peru, which may represent a heretofore unrecorded species, native to the
region.
In any event we no longer need hesitate to consider the possibility, if not the
probability, of the generic relationship between Jlusa and M3sophyllum, and the
probable presence of the former as an element in the Tertiary flora of America.
Locality: Station A.
Sub-Class DICOTYLEDONAE
(CHORIPETALAE)
Order URTICALES
Family MORACEAE
Genus Ficus [Tournefort] Linnaeus
Ficus hyphodroma n. sp.
Plate 57, Figure 3.
Leaf apparently elliptical-lanceolate in shape and coriaceous in texture, about
14 centimeters in length, including the short, thick petiole, by 5.5 centimeters in
maximum width, at about the middle, tapering below to a narrow, cnneate base,
and apparently tapering or curved to the apex; margin entire, undulate, or in-
curved in places, probably inflexed; midrib strong, terminating below in a short,
thick petiole about .5 centimeter in length: nervation, other than the midrib, not
discernible.
This specimen evidently represents a leaf that was of thick, coriaceous
texture, in which all traces of the secondary and finer nervation is lost to view.
In shape and marginal characters it is strikingly suggestive of a leaf from the
*Cook, 0. F. The American origin of agriculture. Pop. Sci. Monthly, vol. 61, pp.
492-505. Oct. 1902. Food plants of ancient America. Smithsonian Inst., Rept. 1903
(No. 1515), pp. 481-497. 1904.
tBerry, E. W. A species of Musa in the Tertiary of South America. Nat. Acad.
Sci. U. S. A., Proc. vol. 11, No. 6, pp. 298-299. Jtme 1925.
Blassler, Harvey. Musa in tropical America. New York Bot. Gard., Journ. vol.
27, N o. 315, pp. 49-54, pls. 301, 302. March 1926
HlOLLICK, PALEOBOTANY OF PORTO RICO 195
Tertiary (Eocene) of Arkansas described and figured by Berry* under the name
Ficus eolignitica; but the absence of any trace of nervation in our specimen
prohibits satisfactory identification.
Locality: Station B.
FIcus SCHIMPERII Lesquereux
Plate 57, Figm'e 4.
Ficus Schimperi Lesquereux. On species of fossil plants from the Tertiary
of the State of Mississippi. Amer. Philos. Soc., Trans. vol. 13, art. 14, p. 417, pl.
18, figs. 1-3. 1867.
The fragmentary condition and imperfectly defined minor characters of our
specimen render accurate comparison somewhat unsatisfactory; but its general
shape, and the characters of the secondary nerves, as far as they are discernible,
are so closely similar to those of Lesquereux' figures that mutual specific identity
appears to be reasonably certain. Lesquereux' descriptions and illustrations were
based upon specimens of Tertiary (Eocene) age collected in Mississippi; and sub-
sequently Berryt described and figured other specimens from the same region and
horizon, which he identified with Lesquereux' species. The specific identity of
our specimen with those figured by Berry, especially in comparison with his
figure 3 (loc. cit.), can also hardly be doubted.
Locality: Station B.
Ficus vexativus n. sp.
Plate 59, Figures 1-4
Leaves varying greatly in size, apparently obovate in shape, narrowed and
abruptly constricted to a bluntly apiculate or somewhat decurrent base; margin
entire; nervation simply pinnate, camptodrome; midrib and main secondary
nerves robust; secondary nerves irregularly disposed and spaced, subtending
angles of various degrees with the midrib, more or less flcexuous, ultimately
ascending and curving upward, following the general contour of the adjacent
margin, and thinning out into a series of irregular loops that constantly decrease
In size and ultimately coalesce; occasional minor intermediate secondary nerves
extend from the midrib and merge into the tertiary cross nervation, where they
are connected by nervilles of minor rank, forming a fine reticulated network of
oblong and polygonal areolae.
These specimens are all so fragmentary that it is impossible to visualize the
exact appearance of a perfect leaf, but it may be assumed that the shape was
pyriform-obovate, and that the characters of the nervation indicate close relation-
ship with the foliar type represented by the Tertiary species Ficus comparabilis
Hollick,t from the island of Trinidad; and more perfect specimens of each might
indicate that the several variable forms now regarded as representing two species
should be included under a single specific name.
Locality: Station A (Figures 1, 2); Station B. (Figures 3, 4).
Type specimen = Figure 2.
*Berry, E. W. The lower Eocene floras of southeastern North America. U. S. Geol.
Survey, Prof. Paper 91, p. 203, pl. 31, fig. 4. 1916.
tBerry, E. W. The lower Eocene floras of southeastern North America. U. S. Geol.
Survey, Prof. Paper 91, p. 204, pl. 31, figs. 1-3. 1916.
t lollick, Arthur. A review of the fossil flora of the West Indies, with descriptions of
new species. New York Bot. Gard., Bull. vol. 12, No. 45, p. 293, pl. 3, figs. la, ib, 2; pl. 4,
figs. 2a, 2b; pl. 5, figs. 2-4. 1924.
SCIENTIFIC SURVEY OF PORTO RICO
Ficus sp.
Plate 87, Figures 1, 2.
The specimens represented by these two figures are, obviously, too fragmen-
tary for satisfactory, identification, or even comparison; but the probability
appears to be that they might be tentatively referred to Ficus vexatinus, the species
last described.
Locality: Collazo River. Collected by Don Narciso Rabell (Figure 1);
Station A (Figure 2).
Order RANALES
Family ANNONACEAE
Genus ANNONA Linnaeus
Annona cetera n. sp.
Plate 60, Figure 1; Plate 76, Figures 2a, 2b; Plate 85, Figure ib; Plate 86,
Figure c.
Leaves ovate-oblong in shape, about 14 centimeters in length by 5 centi-
meters in maximum width, broadest about 2 centimeters below the middle,
tapering above, rounded below; margin entire; nervation simply pinnate, campto-
drome; midrib relatively thick; secondary nerves irregularly spaced and disposed,
subtending angles of 45o-75 with the midrib, ascending, curving upward and
continuing close to the margin where they ultimately coalesce in a series of
successively smaller and smaller loops and fine connecting cross nervilles.
These specimens appear to be comparable with the foliar type represented by
the existing species Annona squamosa Linnaeus, which is native in the West
Indies and adjacent regions. A figure of a leaf of the latter (Figure 2, Plate 60) is
introduced for comparison.
A number of Tertiary species of the genus Annona have been described and
figured, but they all, apparently, represent the foliar type with more regular and
less ascending secondary nerves than in ours; although, in several instances
individual figures in certain species are difficult to differentiate from certain ones
of ours, as in connection with Annona lWilcoxiana Berry,* from the Tertiary of
Tennessee, and Annona lignitum Unger,t from the Tertiary of the Old World.
Locality: Station A (Plate 60, Figure 1); Station B (Plate 76, Figures 2a, 2b;
Plate 85, Figure 1b; Plate 86, Figure c).
Type specimen = Figure 1, Plate 60.
Annona pseudoglabra n. sp.
Plate 54, Figure 3b.
Leaf ellipsoidal in shape, slightly curved to one side, tapering to an acutely
cuneate base and apex, 7 centimeters in length by 3.2 centimeters in maximum
width, entire; nervation simply pinnate, camptodrome; secondary nerves sub-
tending obtuse angles with the midrib, consisting of a major series widely and
irregularly spaced, and a minor series irregularly disposed between and connected
with and merging into the tertiary cross nervation.
*Berry, E. W. U. S. Geol. Survey, Prof. Paper 91 (The lower Eocene floras of south-
eastern North America), p. 216, pl. 41, figs. 1, 2. 1916.
tUnger, Franz. K. Akad. Wissensch. [Wien], math.-naturwiss. Cl., Denkschr. vol.
19, (Sylloge plantarum fossilium, pt. 1), p. 25, pl. 10, figs. 1-7. 1860.
HOLLICK, PALEOBOTANY OF PORTO RICO
These leaves compare so closely with those of the existing species Annona
glabra Linnaeus that their generic relationship appears to be assured and their
near specific relationship to be indicated. A tracing of a leaf of the latter species,
from Cuba (Figure 4, Plate 54), is introduced for comparison.
Incidentally it may be of interest to note that Berry* included Annona
glabra in his descriptive list of Pleistocene plant remains from Florida, based upon
the identification of a single seed.
Locality: Station B.
Order THYMELEALES
Family LAURACEAE
Genus ACRODICLIDIUM Nees
Acrodiclidium pseudosalicifolium n. sp.
Plate 61, Figure 2
Leaf oblong lanceolate-ovate in shape, 9.5 centimeters in length by 3.5
centimeters in maximum width, broadest below the middle, tapering above to an
apiculate apex, rounded below to a convex-cuneate base; margin entire; nerva-
tion simply pinnate, camptodrome; midrib slightly curved to one side at base and
apex; secondary nerves irregularly spaced and disposed, subtending angles of 450
to 750 with the midrib, curving upward and becoming camptodrome in a series of
successively finer and finer connecting loops in the marginal region.
This specimen is so closely similar, in all of its surficial features, to certain
leaves of the existing West Indian species Acrodiclidium salicifolium (Swartz)
Grisebach that it is practically impossible to differentiate one from another, as
may be seen by comparison between Figure 2, Plate 61, representing the fossil,
and Figure 3, Plate 61, representing a leaf of Acrodiclidiuhm salicifolium, intro-
duced for the purpose.
The genus is represented by about a dozen species in the existing flora of the
West Indies and tropical South America, but has not heretofore, as far as I am
aware, been recorded in paleobotanical literature.
Locality: Station B.
Acrodiclidium Pseudocanelo n. sp.
Plate 62
Leaf oblong-elliptical in shape, apparently about 19 centimeters in length by
6 centimeters in maximum width, at about the middle; margin entire; nervation
simply pinnate, camptodrome; secondary nerves irregularly spaced and disposed,
leaving the midrib at obtuse angles, bending or curving rather abruptly upward
and then ascending and continuing upward in the marginal region, where they
thin out, approach one another, and ultimately coalesce in a series of progressively
lessening loops and finer and finer connecting nervilles.
This specimen is so closely comparable with certain leaves of the existing
South American species Acrodiclidium Canelo Rose that they appear to be ab-
solutely identical in all discernible features. It is unfortunate that both base and
apex are missing in our specinlen; but the marginal curves, where they are broken,
indicate tapering both proximad and distad, as in leaves of the existing species.
*Berry, E. V. The fossil plants from Vero, Florida. Fla. State Geol. Survey,
Ninth Ann. Rept., p. 26. 1917.
198 SCIENTIFIC SURVEY OF PORTO RICO
For purposes of comparison I have introduced a tracing of a leaf of Arrodiclidiam
Canelo (Plate 63).
It may also be compared with Ocotca pseudomartinicensis Hollick,* from the
Tertiary of the island of Trinidad, from which it appears to differ only in its more
oblong shape. Incidentally it may here be remarked that surficial differences
between leaves of many species included in the lauraceous genera Aniba, Acro-
diclidium, Ocotea, Nectandra, Persea, etc., are often so slight that accurate or
satisfactory generic identification, based upon foliar characters alone, is impossible.
Locality: Station B.
Genus ANIBA Aublet
Aniba collazodnsis n. sp.
Plates 64, 65
Leaves entire, broadly falcate and inequilaterally oblong-ellipsoidal in shape,
about 18 to 19 centimeters in length by 6.2 centimeters in maximum width,
tapering and ultimately contracted to an acute cuneate base; nervation simply
pinnate, camptodrome; midrib thick and conspicuously curved below, becoming
slenderer and straighter above; secondary nerves irregularly disposed and spaced,
subtending angles of various degrees with the midrib, the basilar ones more
acutely divergent than those above, all curving gently upward, ultimately
extending in a series of irregular loops, and thinning out near the margin.
The two specimens that represent this species were selected from a number
in our collection, all of them closely similar in general surficial appearance, but
varying in minor details of nervation. They might be compared more or less
satisfactorily with leaves of existing lauraceous species in the genera Aniba,
Ocotea, Nectandra, Persca, etc. They are also comparable, except for their
larger size, with Nectandra corvatifolia Engelhardt,t as identified by Berry4 from
the Tertiary of Trinidad, which species was compared by Engelhardt with the
existing South American species Nectandra amazonum Nees. It may also be
found of interest to compare them with Nectandra lWoodringii Berry, from the
Tertiary of Costa Rica, which he compared with the existing West Indian species
N. antillana 1Meissner. Finally, for still further comparison, I have introduced
(Plate 66) a tracing of a leaf of the existing species Aniba riparia (Necs) Mez,
which appears to match our specimens about as satisfactorily as do any of the
numerous ones, existing and fossil, with which they have been compared.
Locality: Station B.
Type specimen = Plate 65.
Genus IItFELANDIA Nees
HIUFELANDIA PORTOR1CENHIM (Ilollick) 11. comb.
Plate 67, Figures 1-3, 5, 6
Aniba portoriccnsis Hollick. A review of the fossil flora of the West Indies,
with descriptions of new species. New York Bot. Gard., Bull. vol. 12, No. 45,
p. 300, pl. 9, figs. 4, 5. Sept. 13, 1924.
*IIollick, Arthur. A review of the fossil flora of the West Indies, with descriptions
of new species. New York Bot. Gard., Bull. vol. 12, No. 45, p. 299, p9 7. 1924.
tEngelhardt, Hermann. Jleber neue Tertidrpflanzen Sild-Amerikas. Senckenb.
Naturf. Gesellsch., Abh. vol. 19, p. 28, pl. 9, fig. 3. 1895.
.Berry, E. W. The Tertiary flora of the Island of Trinidad, I. W. I. Johns Hopkins
Univ. Studies in Geology No. 622, p. 116, pl. 9, fig. 6. 1925.
HBerry, E. W. Tertiary fossil plants from Costa Rica. UT. S. Nat. Mus., Proc. vol.
59 (No. 2367), p. 177, pl. 26, fig. 1. 1921.
IIOLLICK, PALEOBOTANY OF PORTO RICO
The specimens here figured are apparently specifically identical with certain
fragmentary foliar remains from Porto Rico previously described and figured by
me (loe. cit.) under the name Aniba portoricensis, in connection with the discussion
of which species I remarked that The remains might be more or less satisfactorily
compared with any one of a dozen or more existing lauraceous species in the
genera Mcspilodaphnc, Oreodaphnc, Ocotea, Ampclodaphne, Aniba, etc. The
reference to the genus Aniba is, therefore, to be regarded as representing their
family rather than their definite generic relationship."
This additional material has now made possible the following more extended
description and illustration of the species:
Leaves oblong-elliptical to oblong-falcate in shape, 6 to 8.5 centimeters in
length by 2.5 to .3.5 centimeters in maximum width, abruptly narrowed above
to a blunt acuminate apex, and below, to a narrow, more or less inequilateral
base; margin entire, sinuous or wavy; nervation pinnate, camptodrome; midrib
curved; secondary nerves irregularly spaced and disposed, subtending various
angles with the midrib, mostly flexed, curved upward, the extremities conforming
to the contours of the adjacent margin and ultimately becoming camptodrome.
A more satisfactory comparison with existing genera and species than was
originally feasible was also rendered possible, and in consequence I am now
inclined to regard the specimens as comparable with the genus Hufelandia
(= Beilschmicdia) Nees rather than with Aniba; and as an example of one of the
numerous specific comparisons made I have introduced tracings of two leaves of
the existing species Hufelandia pendula (Swartz) Nees (Figures 4, 7, Plate 67).
Leaves of this species, even on an individual branch, may, and often do show as
great a variation in shape and dimensions as ma~r be seen in connection with the
fossil specimens.
Comparison may also be made with certain fossil specimens, from the
Tertiary of Ecuador, described and figured by Engelhardt* under the name
Trigonia varians. Whatever may be thought of the specimens represented by
his figures 4-6, in plate 7, the one represented by his figure 9, on plate 9, is so
strikingly similar to our Figures 1 and 2, on Plate 67, that they all might be
regarded as specifically identical, as far as fragmentary specimens may be com-
pared.
Locality = Station B (Figures 1-3); Station A (Figures 5, 6).
Genus MIISANTECA Chamisso & Schlechtendahl
MISANTECA DUBIOSA (Hollick) n. comb.
Plate 68, Figures 1-6; (Plate 70, Figure 6b?)
Lavrophyllum dubiosum Hollick. A review of the fossil flora of the West
Indies, with descriptions of new species. New York Bot. Gard., Bull. vol. 12,
No. 45, p. 300, pl. 12, figs. 1-3. Sept. 13, 1924.
The specimens upon which this species was originally based were recognized
as belonging to the Lauraceae, but were relegated to the comprehensive fossil
genus Laurophyllum. Recent critical examination of additional material, how-
ever, has resulted in more or less satisfactory comparison with leaves of certain
existing species in the genus Misanteca, and particularly with those of Misanteca
*Engelhardt, Hermann. Ueber neue Tertiitrpflanzen Stid-Amerikas. Senckenb.,
Naturf. Gesellsch., Abh. vol. 19, p. 35, pl. 7, figs. 4-0; pl. 9, fig. 9. 1895.
SCIENTIFIC SURVEY OF PORTO RICO
triandra (Swartz) Mez, tracings of which are represented by Figures 7 and 8 on
Plate 68.
Locality: Station A (Figure 1, Plate 68): Collazo River. Collected by Don
Narciso Rabell (Figures 2, 4, 5, Plate 68); Station B (Figures 3, 6, Plate 68, and
Figure 6b, Plate 70).
Genus OREODAPHNE Nees
OREODAPIINE MISSISSIPPIENSIS Berry?
Plate 67, Figures 8-10
Orcodaphne mississippicnsis Berry. U. S. Geol. Survey, Prof. Paper 91 (the
lower Eocene floras of southeastern North America), p. 303, pl. 82, figs. 3-5. 1916.
Laurus primigenia Unger. Hollick, Arthur. A report on a collection of
fossil plants from northwestern Louisiana. Special Rept. No. 5, [in] Harris,
G. D., and Veatch, A. C., Geol. Survey La., Prelim. Rept., 1899, p. 284, pl. 41,
fig. 1 lexcl. fig. 2]. 1900.
These three specimens are too fragmentary to serve as the basis for a specific
description; but they may be satisfactorily identified as belonging in the Laurace-
ae, and may be tentatively referred to Oreodaphne mississippiensis Berry (loc.
cit.), from the lower Tertiary of the southern United States. Berry's figure 4
(= Hollick's fig. 1, lor. cit.) appears to be specifically identical with our Figure 9,
Plate 67; and our Figure 10 on the same plate appears to be merely a smaller
specimen of the same species. Our Figure 8 may or may not be referable to it.
The difficulty of identifying,'generically, leaves of the general type represented
by these specimens was commented upon by Berry (loc. cit., p. 304) as follows:
"Numerous existing species of the American Tropics and sub-tropics in this and
allied genera approach closely to this type. In fact, though there may be differ-
ences among students of fossil floras, as there are among students of the existing
flora regarding the proper generic limits of the genera of the Lauraceae, no one
can dispute the correctness of the family reference of these . species. The
present species [Oreodaphnc mississippiensis] is very similar to . the existing
Persea pubescens (Pursh) Sargent . I have also seen unnamed specimens of
Ocotea (Oreodaphne) from New Grenada identical with it."
In the circumstances, therefore, I have deemed it best to regard our frag-
mentary specimens as tentatively referable to the species described and figured
by Berry, pending the possible future discovery of more perfect specimens
that might or might not confirm the identification.
Locality: Station A (Figures 8, 9); Station B (Figure 10).
Order ROSALES
Family LEGUMINOSAE (AMIMOSACEAE)
Genus INGA [Plumier] Scopoli
Inga curta n. sp.
Plate 69, Figure 1.
Leaflet oblong (or oblong-lanceolate?) in shape, apparently about 6 centi-
meters in length by 1.8 centimeter in maximum width; base inequilateral, broadly
cuneate on one side, apparently narrowed on the other; margin entire; midrib
HOLLICK, PALEUOUTANY U' FUI'TU Miu
flexuous, bent in opposite directions at base and apex; nervation simply pinnate;
secondary nerves rather widely spaced.
The imperfect condition of this specimen renders satisfactory description or
comparison impossible. It apparently represents a leaflet of a compound leaf,
and the outline and basilar contours are so suggestive of certain species of Inga,
both existing and fossil, that I have ventured to refer it to that genus; and for
comparison I have introduced a drawing of a leaflet of the existing species Inga
affinis De Candolle (Figure 2, Plate 69). Comparison may also be made with
Inga mississippiensis Berry*, from the Tertiary of Mississippi, and with Leg-
vminosites cclastroides Ileer,t from the Tertiary of Switzerland. If the specimens
that represent these last two species and ours had been found associated together
all three would probably have been regarded as specifically identical.
Locality: Station B.
Inga pseudinsignis n. sp.
Plate 79, Figures lb, 2b
Leaflets oblong or oblong-lanceolate, inequilateral, entire, apparently about
8 centimeters in length by 3 centimeters in maximum width; base rounded on one
side, cuneate on the other; nervation pinnate; secondary nerves irregularly
spaced and disposed, subtending angles of approximately 450 with the midrib,
curving upward and ultimately extending close along the margin; tertiary nervilles
fine, obscurely defined.
These two figures represent counterparts of one and the same specimen, and
although neither of them is complete they bear such a close resemblance to leaflets
of certain existing species of Inga that their reference to that genus appears to be
justified; and for comparison I have introduced (Figure 3, Plate 88) a tracing of
a leaflet of Inga insignis Kunth, with which our Figure Ib, Plate 79 may be seen
t.o compare very satisfactorily.
Locality: Station B.
Type specimen: Figure lb.
INGA PSEUDONOBILIS HIollick
Plate 70, Figure 6a; Plate 76, Figure la
Inga pscudonobilis Hollick. A review of the fossil flora of the West Indies,
with descriptions of new species. New York Bot. Gard., Bull. vol. 12, No. 45, p.
301. pl. 5, fig. 6; pl. 10, fig. 7. Sept. 13, 1924.
It is unfortunate that the original specimens on which this species was based,
and also those subsequently collected, are all imperfect and in a similar manner-
only the basal parts of the leaflets being preserved.
Existing species with which these fragmentary remains may be compared
include Inga Rourzoni De Candolle, and Inga nobilis Willdenovw; and a tracing of
a leaflet of the latter species, represented by Figure 3, Plate 76, is introduced
for this purpose.
Locality: Station B.
*Berry, E. W. U. S. Geol. Survey, Prof, Paper 91 (The lower Eocene floras of south-
eastern North America), p. 222, pl. 45, fig. 1. 1916.
tlieer, Oswald. Flora Tertiaria Helvetiae, vol. 3, p. 125, pl. 139, fig. 43c. 1859.
ZUI
202 SCIENTIFIC SURVEY OF PORTO RICO
Inga pseudospuria n. sp.
Plate 69, Figure 3
Leaflet oblong-lanceolate in shape, slightly unsymmetrical, one side rounded
or obscurely convex, the other almost flat, 4.75 centimeters in length by 1.5
centimeter in maximum width, entire, rather abruptly narrowed to an acuminate
apex and rounded to a blunt base; midrib slightly curved distad and bent in the
direction of the flat side of the leaflet; nervation simply pinnate; secondary
nerves irregularly spaced and disposed, those on the rounded side of the leaflet
diverging from the midrib at angles more obtuse than those oi the opposite side.
This specimen resembles so closely certain of the leaflets of the existing Inga
spuria HIumbolt & Bonpland that they might well be regarded as specifically
identical, one with the other; and a drawing of a leaflet of this species (Figure 4,
Plate 69) is introduced for comparison.
It may also be suggested that this leaflet merely appears like a small form of
Inga pseudinsignis, as represented by our Figure Ib, Plate 79; but unless
they should be found actually associated together on one and the same leaf stalk
they may be regarded as specifically different.
Locality: Collazo River. Collected by Don Narciso Rabell.
INGA (?) sp. Hollick
Inga (?) sp. Ilollick. A review of the fossil flora of the West Indies, with
descriptions of new species. New York Bot. Gard., Bull. vol. 12, No. 45, p. 301,
pl. 12, fig. 5. Sept. 13, 1924.
Locality: Collazo River, near base of second falls below Carretera bridge.
Collected by Bela Hubbard.
Genus PITHECELLOBIUM [= PITHECOLOBIUM] Martius
Pithecellobium (?) imperfectum n. sp.
Plate 69, Figure 5
Leaflet, apparently about 3 centimeters in length by 1.1 centimeter in maxi-
mum width, inequilateral, oblong (?) in shape; base rounded-cuneate on one side,
concave-cuneate on the other; margin entire; midrib fine; secondary nervation
not discernible.
This imperfect leaflet is tentatively referred to the genus Pithccellobium, and
It may be compared with the existing P. gracilliflorum Blake (Figure 6, Plate 69),
as far as the shape of the lower part of the leaflet is concerned; but, the absence of
the upper part of the fossil and also of any indication of secondary nervation,
render further comparison useless.
Somewhat similar leaflets, from the Eocene of the southern United States,
were described and figured by Berry*, under the name Caesalpinia Wl'ilcoxiana,
which he compared with Cassia longifolia Engelhardt,t from the Tertiary of
Ecuador, and with several Old World Tertiary species described under the genera
Caesalpinites, Legumninosites, etc. In view, however, of the fragmentary con-
dition of our specimen and the lack of essential diagnostic characters, any further
*Berry, E. W. U. S. Geol. Survey, Prof. Paper 91 (The lower Eocene Floras of south-
eastern North America), p. 235, pl. 50, figs. 9-12. 1916.
tEngethardt, Hermann. Ueber ncue TertiArpflanzen SUd-Amerikas. Senckenb.
Naturf. Gesellsch., Abh. vol. 19, p. 19, pl. 2, figs. 15, 16. 1895.
HOLLICK, PALEOBOTANY OF PORTO RICO
discussion of its probable generic relationship would not be likely to lead to
conclusions of any value.
Locality: Station B.
PITHECELLOBIUM PSEUDOTRAPEZ1FOLIUM Hollick
Pithecolobiuom 1 seudotrapezifolium Hollick. A review of the fossil flora of the
West Indies, with descriptions of new species. New York Bot. Gard., Bull. vol.
12, No. 45, p. 302, pl. 10, fig. 8. Sept. 13, 1924.
Locality: Collazo River, near base of second falls below Carretera bridge.
Collected by Bela Hubbard.
PITHECELLODIUM VEXATIVUM Hollick
Plate 71, Figure 3
Pithecolobium vexativum Hollick. A review of the fossil flora of the West
Indies, with descriptions of new species. New York Bot. Gard., Bull. vol. 12,
No. 45, p.303, pl. 11, fig. 4. Sept. 13, 1924.
This specimen appears to be specifically identical with the one upon which
the species was originally based. Unfortunately each one lacks the distal
portion, hence a full specific description is impossible.
In many respects these specimens resemble leaflets of certain other fossil
species in the Leguminosae, such as Cynomctra Rabellii (Figures la, 2, Plate
71), collected in the same locality; but that species is narrower, and may
be further distinguished by its smaller size and more rounded cuneate base.
Whether or not these characters may properly be considered sufficient to con-
stitute specific difference, however, should be regarded as a matter of personal
opinion. Pithecellobium vexativam IIollick may also be compared with Legumi-
nosites copaiferaeoidesEngelhardt,* from the Tertiary of Chile, and with the ex-
isting species P. ligustrinum (Jacquin) Klotzsch.
Locality: Collazo River. Collected by Don Narciso Rabell.
Family LEGUMINOSAE (CAESALPINIACEAE)
Genus CASSIA [Tournefort] Linnaeus
CASSIA (?) DUBIOSA Hollick
Cassia (?) dubiosa HIollick. A review of the fossil flora of the West Indies,
with descriptions of new species. New York Bot. Gard., Bull. vol. 12, No. 45, p.
304, pl. 10, fig. 4. Sept. 13, 1924.
In the original discussion of this species (loc. cit.) I remarked that "the
systematic position of this specimen has not been satisfactorily determined.
The apex is missing and the nervation is obscurely defined and incomplete." It
was compared, incidentally, with Cassia Glennii Berry,f from the Tertiary of the
southern United States. It may also be compared with Pithccellobium vexativum
Hollick, the species last described (Figure 3, Plate 71); but in connection with
either species the specimens are too fragmentary for satisfactory comparison.
*Engelhardt, Hermann. Ueber Tertiirpflanzen von Chile. Senckenb. Naturf.
Gesellsch., Ali. vol. 16, No. 4, p. 682, pl. 9, fig. 11. 1891.
+Berry, E. W. U. S. Geol. Survey, Prof. Paper 91 (The lower Eocene floras of south-
eastern North America), p. 233, pl. 45, figs. 15, 16, 17a, 18; pl. 52, fig. 6. 1916.
SCIENTIFIC SURVEY OF PORTO RICO
Locality: Collazo River, near base of second falls below Carretera bridge.
Collected by Bela Hubbard.
Cassia evidens n. sp.
Plate 69, Figure 8
Leaflet unsymmetrically oblong-obovate in shape, entire, about 3.25 centi-
meters in length by 1 centimeter in maximum width, rounded-cuneate at the base
on one side, narrowly-cuneate on the other; midrib straight; nervation simply
pinnate; secondary nervation fine.
There can hardly be any doubt in regard to the generic relationship of this
specimen, and for comparison I have introduced a drawing of a leaflet of the
existing species Peiranisia [Cassia] biflora (Linnaeus) Britton & Rose (Figure 9,
Plate 69). Leaflets of several other species of the genus might, however, have
been utilized for the same purpose, and it would be hazardous to infer closer
specific relationship to any particular one than to any of the others.
The genus is well represented in Tertiary deposits of both the Old World and
the New, and several of the species that have been described and figured are more
or less suggestive of the general type of leaflet represented by our specimen, as
for example Cassia dimidiato-linearis Engelhardt,* from the Tertiary of Ecuador.
Locality: Station B.
Cassia imitativa n. sp.
Plate 70, Figures 1, 2a, 2b
Leaflets ovate-lanceolate to ovate-elliptical in shape, slightly unsymmetrical,
about 5 to 6 centimeters in length by 2.25 to 2.8 centimeters in maximum width,
rounded or curved below to a cuneate base, narrowed more or less abruptly above
to an acuminate apex; margin entire; nervation pinnate; secondary nerves
irregularly spaced and disposed, subtending obtuse angles with the midrib.
It is difficult'to decide if these leaflets should be described as representing a
new species, or if they should be regarded merely as foliar variants of a hetero-
phyllus species, such as Cassia Glennii Berry,f from tile Tertiary of the southern
United States, in connection with which the author remarked (loc. cit.): "It may
be matched by a number of the abundant existing species of Cassia from the
American tropics. Among fossil forms it shows great similarity to certain
European Tertiary species, especially to the abundant and wide spread Cassia
BerenicesUnger (Unger, Franz. Die fossile flora vonSotzka, p. 58, p1. 43, figs. 4-10,
1850), and Cassia hyperborea Unger (idem, pl. 43, figs. 1-3), both so common in
thr Oligocene of Southern Europe."
In the circumstances, therefore, it may be regarded as an open question
whether or not specific differentiation of our specimens from one or another of the
species cited may be justified. Incidentally it may also be pertinent to here
remark that equal difficulty would be experienced in connection with the deter-
mination and differentiation of existing species of the genus if scattered leaflets
were the only material available for study.
Locality: Station B.
Type specimen: Figure 1.
*lEngelhardt, Hermann. Ueber neue Tertitrpflanzen Stid-Amerikas. Senckenb.
Naturf. Gesellseli., Ablh. vol. 19, p. 18, pl. 1, fig. 25. 1895.
fBerry, E. W. U. S. Geol. Survey, Prof. Paper 91 (The lower Eocene floras of south-
eastern North America), p. 233, pl. 45, figs. 15, 16, 17a, 18; pl. 52, fig. 6. 1916.
HOLLICK, PALEOBOTANY OF PORTO RICO
Cassia imparilis n. sp.
Plate 70, Figure 5
Leaflet oblong or oblong-lanceolate in shape, entire, about 5 to 5.5 centimeters
in length by 2 centimeters in maximum width, rounded below to a convex cuneate
base, narrowed distad; nervation pinnate; secondary nerves widely spaced,
irregularly disposed, subtending angles of about 450 with the midrib.
This leaflet may be compared more or less satisfactorily with any one of
several different existing species of Cassia, and also with certain individual
figures of Cassia emarginata Berry (not C. emarginata Linnaeus) and C. Wilcoxiana
Berry,* from the Tertiary of the southern United States. In our specimen,
however, the apex is missing and hence any satisfactory comparison is not feasible.
In connection with a genus such as Cassia it is as easy to discern differences as
it is to recognize resemblances between leaflets of many of the species, and satis-
factory conclusions in regard to specific relationships, based entirely upon leaflets
are well nigh impossible.
Locality: Station B.
Cassia ordinaria n. sp.
Plate 69, Figure 10
Leaflet ovate-lanceolate in shape, entire, about 5 centimeters in length by 2.2
centimeters in maximum width, tapering above to an acuminate apex and rounded
below to a convex-cuneate base; midrib straight; nervation simply pinnate;
secondary nerves fine, irregularly spaced and disposed, leaving the midrib at
angles of approximately 45.
This specimen represents a type of leaflet that is common to several different
species of the genus Cassia in our existing flora, and also to certain fossil species
that have been referred to the genus, For general generic comparison I have
introduced a drawing of a leaflet of the existing Cassia leptocarpa hirsuta Bentham
(Figure 11, Plate 69), which may serve to represent the type of leaflet men-
tioned; and as an example of a closely related fossil species, comparison may be
made with Cassia Glennii Berry,t from the Tertiary of Tennessee and Mississippi,
especially with his figure 15, plate 45, and figure 6, plate 52 (op. cit.). Our
specimen might also be regarded merely as a small leaflet of Cassia imitativa,
previously described (page 204, Figures 1, 2a, 2b, Plate 70), from the same
locality and station.
Locality: Station B.
CASSIA PURYEARENSIS Berry
Plate 70, Figures 3, 4
Cassia puryearensis Berry. U. S. Geol. Survey, Prof. Paper 91 (The lower
Eocene floras of southeastern North America), p. 230, p). 51, figs. 13, 14. 1916.
There do not appear to be any discernible characters in our specimens that
might serve to differentiate between them and those figured by Berry (op. cit.)
from the Tertiary of Tennessee, which he compared with the existing species
*Berry, E. W. U. S. Geol. Survey, Prof. Paper 91 (The lower Eocene floras of south-
eastern North America), pl. 45, flg. 17b; pl. 48, fig. 5 (= C. emarginata); pl. 50, fig. 2
(= C. Wilcoxiana). 1916.
fBerry, E. W. U. S. Geol. Survey, Prof. Paper 91 (The lower Eocene floras of south-
eastern North America), p. 233, pl. 45, figs. 15, 16, 17a, 18; pl. 52, fig. 6. 1916.
SCIENTIFIC SURVEY OF PORTO RICO
Adipera [Cassia] laevigata (Willdenow) Britton & Rose and Cassia corymbosa
Lamarck.
Locality: Station B (Figure 3); Station A (Figure 4).
Cassia visibilis n. sp.
Plate 69, Figure 12
Leaflet ovate-acuminate in shape, entire, 2.75 centimeters in length by 1.5
centimeter in maximum width; apex slightly flexed or bent toward one side;
midrib slightly bent distad; nervation simply pinnate; secondary nerves fine,
leaving the midrib at angles of approximately 60.
This specimen may possibly represent one of the lower leaflets of Cassia
ordinaria, previously described (Figure 10, Plate 69). Leaflets on individual
leaves of any one of several different existing species of Cassia differ between
themselves, in shape and size, far more than do the two fossil specimens under
consideration; but in the absence of any definite evidence on either side I have
deemed it best to describe the leaflet in question under a distinct specific name,
and to compare it with the existing species Ditremexa [Cassia] occidentalis (Lin-
naeus) Britton & Rose (Figure 13, Plate 69). In this connection it is of interest
to note the great diversity of leaf forms included by Heer* under the Old World
Tertiary species Cassia Berenices Unger, and incidentally to compare his figure 42
(op. cit.) with the figure of our specimen. As far as this particular figure of
Heer's is concerned it might well be regarded as representing a specimen very
difficult of specific differentiation from ours.
Locality: Station B.
Genus COPAIVA Jacquin
Copaiva oligocenica n. sp.
Plate 71, Figure 5.
Leaflet inequilateral, the broader side expanded distad, the opposite,
narrower side expanded proximad, 3.75 centimeters in length by 1.8 centimeter
In maximum width, entire, abruptly narrowed to an obtuse apiculate apex,
rounded below to an inequilateral curved-cuneate base; midrib curved, convex
toward the broader side of the leaf; nervation simply pinnate; secondary nerva-
tion fine, consisting of irregularly disposed and spaced major and minor nerves,
those on the convex side of the midrib subtending angles more acute than those
on the opposite, concave side, all curving upward, approaching and merging
together in the marginal region through the connecting tertiary nervilles.
This leaflet is so closely comparable with lateral leaflets of Copaiva chiriquensis
Pittier, that from selected specimens it would be difficult to differentiate the fossil
form. A tracing of an average leaflet of the above species (Figure 6, Plate 71),
is introduced for comparison.
Locality: Station A.
Genus CYNOMETRA Linnaeus
Cynometra Rabellii n. sp.
Plate 71, Figures la, 2.
Leaflets inequilateral, 3.5 to 4.5 centimeters in length by 1.5 to 1.8 centi-
metei in maximum width, broadest distad, abruptly narrowed to an apiculate
*Heer, Oswald. Flora Tertiaria Helvetiae, vol. 3, p. 118, pl. 87, figs. 42-56. 1859.
HOLLICK, PALEOBOTANY OF PORTO RICO 207
apex, rounded to the base on the broader side, cuneate on the narrower side,
entire; midrib curved, convex toward the broader side; nervation pinnate-
reticulate, delicate, apparently consisting of a major and a minor series that are
irregularly disposed and spaced, leaving the midrib at various angles of divergence,
bent, flexed, curving upward and connected by tertiary nervilles that form a
series of irregular reticulations.
These specimens certainly represent lateral leaflets of a compound leaf,
apparently belonging to the Leguminosae. Comparison with leaflets of a number
of different genera having similar foliage finally resulted in matching our specimens
most satisfactorily with leaflets of certain species of the genus Cynometra, and, as
an example, I have introduced (Figure 4, Plate 71) a tracing of a lateral leaflet
of the existing species Cynometra trinitensis Oliver. Comparison may also be
made with Pithecellobium vexativum Hollick, discussed on page 203, and illustrated
by Figure 3, Plate 71, which differs from the former, however, in its larger size
and also, apparently, in the more uniform convexity of its sides. Possibly their
mutual generic identity, at least, might be regarded as a matter of individual
opinion.
Identifications of the leaflets of leguminous plants are frequently not entirely
satisfactory. In many instances they differ but little in surficial characters, and
certain species are exceedingly heterophyllous, as far as individual leaflets are
concerned; hence isolated fossil leaflets in which important diagnostic characters
are lacking, or only obscurely defined, may be identified as representing any one
of several genera; and subsequent discovery of specimens, in a better state of
preservation, might result in indicating that the original identification was
erroneous.
Locality: Collazo River. Collected by Don Narciso Rabell (Figure la);
Station B (Figure 2).
Type specimen: Figure 2.
Family LEGUMINOSAE PAPILIONACEAEE)
Genus LONCHOCARPUS Humboldt, Bonpland, and Kunth
Lonchocarpus praelatifolius n. sp.
Plate 71, Figure 7a
Leaflet lanceolate-ellipsoidal in shape, inequilateral, entire, 6 centimeters in
length by 2.5 centimeters in maximum width, tapering rather abruptly to a
constricted apiculate apex, rounded to an inequilateral cuneate base; midrib
slightly flexuous; nervation simply pinnate; secondary nerves irregularly disposed
and spaced, leaving the midrib at acute angles of divergence on the narrower side
of the leaf and at more obtuse angles on the opposite, broader side, all curving
upward, thinning out and approaching closely in the marginal region and connect-
ed by tertiary cross nervation.
The general appearance of this specimen at once suggests that it represents a
lateral leaflet of a compound leaf; and critical examination suggests generic
relationship with the genus Lonchocarpus. For comparison I have introduced
(Figure 8, Plate 71) a tracing of a leaflet of Lonchocarpus latifolius (Willdenow)
Humboldt, Bonpland, and Kunth, which may be seen to resemble our specimen
very closely-so closely, in fact, that specific relationship might almost be implied.
Locality: Station B.
208 SCIENTIFIC SURVEY OF PORTO RICO
Genus SOPHORA Linnaeus
SOPHORA (?) SUSPECT IHollick
Plate 69, Figure 7
Sophora (?) suspccta IIollick. A review of the fossil flora of the West Indies.
with descriptions of new species. New York Bot. Gard., Bull. vol. 12, No. 45, p.
304, p1. 10, fig. 3. Sept. 13, 1924.
The identity of our specimen with this species can hardly he questioned,
despite the absence of any trace of secondary nervation. In shape the agree-
ment is perfect; also our specimen and the one above cited, on which the specific
type was based, both came from the same locality. It may be compared with the
existing species Sophora affinis Torrey & Gray.
Locality: Station A.
Order GERANIALES
Family TMELIACEAE
Genus GuiREA Allamand; Linnaeus
Guarea opinabilis n. sp.
Plate 86, Figure d
Leaf apparently oblong lanceolate in shape, about 15 centimeters in length
by 4 centimeters in maximum width; margin entire, sinuous; nervation pinnate,
camptodrome; secondary nerves irregularly spaced and disposed, subtending
obtuse angles with the midrib, curving upward and thinning out in the marginal
region, where they approach and join in a series of successively smaller loops.
This may not be regarded as a satisfactory specimen upon which to base a
new specific description; but it may be compared so closely with certain existing
species of the genus Guarca that I have ventured to so refer it, and to compare it
with an unnamed species from South America, a tracing of a leaf of which is
represented by Figure 2, Plate 88.
It may also be compared with a fossil specimen, from the Tertiary of Ecuador,
described and figured by Engelhardt* under the name Tapiria [= TapiriraJ lanceo-
lata, which he compared with the existing species T. guianensis Aublet.
Locality: Station B.
Genus TRICHILIA P. Browne; Linnaeus
Trichilia evidens n. sp.
Plate 72, Figure 7
Organism consisting of a tripartite, loculicidal capsule; divisions ovate-
lanceolate in shape, about 1.8 centimeter in length by about .9 centimeter in
maximum width, varying somewhat in size between themselves, thick and coriace-
ous in texture, dorsal surfaces channeled medianly from base to apex.
This specimen evidently represents three confluent valves or divisions of a
loculicidal capsule that was split open, flattened out and compacted. Apparently
it is the exterior that is exposed, as indicated by the clearly defined linear depres-
sion in the median region of each division, extending from the proximal to the
distal extremity.
*Engelhardt, Ilermann. Uebr noniie Tertiarpflanzen Siid-Amerikas. Senckenb.
Naturf. Gesellsch., Abhh. vol. 19, 1). 15, pl. 9, fig. 4. 1895.
HOLLICK, PALEOBOTANY OF PORTO RICO
It is so closely comparable with capsules of certain existing species of Trichi-
lia that I have no hesitation in referring it to that genus; and for purposes of
comparison I have introduced two drawings of a dried capsule of Trichilia odorata
Andrews, from Jamaica,-one representing the interior (Figure 8, Plate 72), and
one the exterior (Figure 9, Plate 72). In the former tie median lines represent,
ridges or fins. In the latter they represent dorsal depressions.
Locality: Collazo River. Collected by Don Narciso Rabell.
Trichilia Pseudobakerii n. sp.
Plate 72, Figure 5
Leaflet obovate-elliptical in shape, entire, about 12 centimeters in length by
5 centimeters in maximum width, tapering to a narrow, elongated base and,
apparently, to a more or less abruptly constricted apex; nervation pinnate,
camptodrome; secondary nerves irregularly spaced and disposed, subtending
obtuse angles with the midrib, mostly flexed and more or less angled and curved,
or bent abruptly upward toward their extremities, where they join in a series of
angled loops; tertiary nervation all more or less flexed, forming irregular polygonal
areolac.
There can hardly be any doubt that our fragmentary specimen represents a
species very close to the existing Trichilia Bakerii C. De Candollc (Figure 6, Plate
72), and we may infer that our specimen, in its entirety, would have possessed
basal and apical characters comparable with those of the specimen of T. Bakerii
represented by Figure 6, in as much as the parts that are available may be
matched so satisfactorily.
A fossil species, from the Tertiary of the island of Trinidad, that resembles
our species closely, was described and figured by me* under the name Geissanthus
Brittonii. The resemblance of this species to Trichilia Pseudobakerii is so marked
that the differences between them appear to be almost negligible; but the former
may be seen to be smaller in size and of a more obovate or spatulate shape. If
they should ever be proved to be specifically identical, however, it might yet be a
difficult matter to determine the genus to which they should be referred. It
would not be inconsistent to regard the figures of the two species as generically
identical, and as possibly representing leaflets of a compound leaf of one and the
same species.
Locality: Station B.
Trichilia pseudohirta n. sp.
Plate 72, Figure 1
Leaflet lanceolate in shape, entire, about 4.5 centimeters in length by 2
centimeters in maximum width, rounded below, constricted above to an acuminate
apex; nervation pinnate, camptodrome; secondary nerves rather widely and
uniformly spaced, subtending approximately uniform angles of 450 to 550 with the
midrib, extending almost straight, or curving slightly upward, and joining in the
marginal region in a series of angled loops.
Leaflet lanceolate in shape, entire, about 4.5 centimeters in length by 2
centimeters in maximum width, rounded below, constricted above to an acuminate
apex; nervation pinnate, camptodrome; secondary nerves rather widely and
*Hollick, Arthur. A review of the fossil flora of the West Indies, with descriptions of
new species. New York Bot. Gard., Bull. vol. 12, No. 45, p. 312, pl. 2, figs. 2, 3. Sept.
13, 1924.
SCIENTIFIC SURVEY OF PORTO RICO
uniformly spaced, subtending approximately uniform angles of 450 to 550 with the
midrib, extending almost straight, or curving slightly upward, and joining in the
marginal region in a series of angled loops.
This specimen is so closely comparable with certain leaflets of the existing
species Trichilia hirta Linnaeus that it would be difficult to differentiate between
equivalent fragmentary remains, as may be seen by comparing Figure 1, Plate
72, representing the'fossil, with Figure 2, on the same plate, representing a
leaflet of the species mentioned. The figure of the fossil matches the upper
part of the leaflet so perfectly that we may assume the basal part of the former
to have been identical in its contour with the equivalent part of the latter, as
indicated in Figure 2.
Our figure may also be seen to closely resemble, except in its smaller size, that
of a leaf from the Tertiary (Pliocene) of Bolivia, described and figured by Berry*
under the name Anona [= A nnona] cochambcnsis, and compared with the existing
species A. acutiflora Martins all of which indicates the difficulties experienced
In endeavoring to satisfactorily compare and identify Tertiary fossil floras of
tropical regions with the flora now in existence in the same regions.
Locality: Station B.
Trichilia spatulata n. sp.
Plate 72, Figure 3
Leaflet apparently spatulate in shape, slightly inequilateral, entire, about
3.75 centimeters in length by 1.25 centimeter in maximum width; nervation
pinnate; secondary nerves numerous, fine, leaving the midrib at angles of approxi-
mately 45, curving upward and becoming camptodrome in the marginal region.
This leaflet appears to be comparable with those of the existing species
Trichilia trifolia Linnaeus, of which a tracing of a specimen, represented by
Figure 4, Plate 72, is introduced in order to facilitate comparison. Although
the apical portion of the fossil, represented by Figure 3, is lacking, the basal
portion matches the equivalent part of Figure 4 so closely that it would seem
justifiable to infer that they were also alike in their apical parts.
Locality: Station B.
Order SAP1NDALES
Family SAPINDACEAE
Genus MELICOCCA Linnaeus
Melicocca immutata, n. sp.
Plate 73, Figures 1-4
Leaflets inequilaterally obovate, oblong or oblong-lanceolate in shape,
6 to 7.5 centimeters in length by 3 to 3.5 centimeters in maximum width, entire,
tapering to the apex or rounded and abruptly acuminate, turned to one side,
unsymmetrically rounded or curved-cuneate at the base; midrib curved, turned
to one side distad, conforming in general to the trend of the adjacent marginal
contour; nervation pinnate, reticulate-camptodrome; secondary nerves numerous,
consisting of a major series, with a minor series interspersed between, all fine,
closely approximated, irregularly spaced and disposed, subtending mostly obtuse
*Berry, E. W. Johns Hopkins Univ. Studies in Geol. No. 43 (Pliocene fossil plants
from eastern Bolivia. George Huntington Williams Memorial Publication No. 18), p. 172,
pl. 8, fig. 4. 1922.
HOLLICK, PALEOBOTANY OF PORTO RICO 211
angles with the midrib, connected by numerous fine tertiary nervilles that form an
irregular reticulated network of elongated and polygonal areolae throughout,
with rounded and angled loops in the marginal region.
The figures representing these specimens were compared with leaflets of
several different existing species in the Sapindaceae-the family to which they
appear to be referable. A species that matches them very closely in general
surficial characters, as well as in minor variations of size and shape, is Melicocca
bijuga Linnaeus. Tracings of two typical leaflets of this species are represented
by Figures 5 and 6, Plate 73. Figure 5 may be most satisfactorily compared with
Figures 1 and 3, and Figure 6 with Figures 2 and 4.
It is possible that a fragmentary specimen originally described and figured by
me* under the name Apocynophyllum pseudowillughbya, may properly belong in
the same generic, and possibly in the same specific category with them. They
were all collected in the same locality; but the specimen in question is too frag-
mentary and too imperfectly preserved for accurate comparison, and the resem-
blance had best be regarded merely as suggestive.
Locality: Station B (Figures 1-3); station A (Figure 4).
Type specimen: Figure 1.
\MELICOCCA sp.
Plate 87, Figure 4
This imperfectly preserved foliar fragment is, apparently, closely related
to the specimens previously described and figured (page 210, Plate 73, Figures
1-4) under the name Melicocca immulata: and the similarity between them might
be regarded as sulficient for specific recognition. The particular one of the above
figures with which it may be most nearly matched is Figure 3, from which it
appears to differ merely in its somewhat larger size.
Locality: Station B.
Genus SAPINDUS [Tournefort] Linnaeus
Sapindus Brittonii n. sp.
Plate 55, Figure 5d; Plate 74, Figures 1, 2; Plate 75, Figure 5; Plate 76,
Figure ib; Plate 78, Figures 3a, 3b; Plate 79, Figures la, 2a; Plate 86, Figure e;
(Plate 87, Figure 3?)
Leaflets curved, unsymmetrically obovate-elliptical in shape, short petiolate,
9 to 12 centimeters in length by 4 to 5 centimeters in maximum width; margin
entire, more or less sinuous or flexuous; midrib curved, bent to one side and
flexuous in conformity with the contour of the adjacent margin; nervation
pinnate; secondary nerves irregularly disposed amd spaced, consisting of a major
series with others of minor rank interspersed between, subtending obtuse angles
with the midrib on the convex side and angles more acute on the concave side,
mostly curved or bent abruptly upward and becoming camptodrome in the
marginal region, where they join in a series of loops successively diminishing in
size.
Most of our specimens compare closely with certain leaflets of the existing
species Sapindus Saponaria Linnaeus. This species is one in which the leaflets
vary in size and shape to an unusual extent, and the general type with which our
specimens may be most satisfactorily matched is indicated by Figure 3, Plate
*Hollick, Arthur. A review of the fossil flora of the West Indies, with descriptions of
new species. New York Bot. Gard., Bull. vol. 12, No. 45, p. 318, pl. 11, fig. 3. Sept. 13,
1924.
SCIENTIFIC SURVEY OF PORTO RICO
74, representing a tracing of a leaflet of Sapindus Saponaria which may be
regarded as an example of an average broad-type leaflet of the species.
A considerable variation may be noted among the specimens included under
Sapindus Brittonii, but the differences in size and shape of the leaflets are negli-
gible in comparison with the differences that may be seen among leaflets of
Sapindus Saponaria.
A fossil species from the Tertiary of Brazil, which might be regarded as
merely a small form of Sapindus Brittonii, was described and figured by Hollick
& Berry* under the name Sapindus praesaponaria, by reason of its close resem-
blance to S. Saponaria. All of our specimens, however, appear to be represen-
tative of a larger type of leaf.
Locality: Station B (Figure 5d, Plate 55; Figure 5, Plate 75; Figure lb,
Plate 76; Figures 3a, 3b, Plate 78; Figures la, 2a, Plate 79; Figure e, Plate 86;
(Figure 3, Plate 87?). Station A (Figure 1, Plate 74). Collazo River. Collected
by Don Narciso Rabell (Figure 2, Plate 74).
Type specimen: Figure 1, Plate 74.
Sapindus gracilentus n. sp.
Plate 75, Figure 3
Leaflet curved, unsymmetrically linear-lanceolate, sub-falcate in shape,
entire, apparently about 8.5 centimeters in length by 1.75 centimeter in maxi-
mum width, tapering to a narrow, inequilateral, cuneate base; midrib curved in
conformity with the curvature of the leaflet; nervation pinnate; secondary
nerves irregularly disposed and spaced, numerous, fine, including obscurely
defined series of major and minor rank, subtending angles of approximately 45'
with the midrib, becoming camptodrome-reticulate in the marginal region.
This leaflet is apparently comparable with the ordinary narrow type of
leaflets of the existing species Sapindus Saponaria Linnaeus, a tracing of one
of which is represented by Figure 4, Plate 75. The specimen of Sapindus
represented by S. gracilentus, and those last described, under the name S. Brittonii
differ so greatly in size and shape that, surficially, they may not at first appear
to be generically related; but, as a matter of fact they do not differ, one from the
other, any more strikingly than do the leaflets on different individual trees of S.
Saponaria, with which species both of the two fossil ones might be compared.
A number of narrow leaved Tertiary species of the genus have been described,
from both tne Old World and the New; but in no instance does there appear to
be a resemblance to ours suflicientlystriking to definitelyindicate specific identity.
Interesting comparisons, however, may be made with Sapindus falcifolius
Alex. Braun as depicted by Heer,t from the Tertiary of Switzerland, and Sapindus
angustifolius Lesquereux,t [not S. angustifolius Blume] from the Tertiary of the
western United States, which was subsequently split up into several different
species (S. coloradensis Cockerell, S. lconis Cockerell, S. mississippiensis Berry,
etc.)-an example of the confusion that obtains in connection with efforts to
identify and to differentiate between leaf forms referred to this genus.
Locality: Station A.
*Hollick, Arthur, & Berry, E. W. Johns Hopkins Univ. Studies in Geology No. 5
(A late Tertiary flora from Bahia, Brazil), p. 82, pl. 8, figs. 6, 7. 1924.
titeer, Oswald. Flora Tertiaria Htelvetiae, vol. 3, pl. 119; pl. 120, figs. 2-8; pl. 121,
figs. 1, 2. 1859.
TLesquereux, Leo. U. S. Geol. Survey Terr., Rept. vol. 7 (The Tertiary flora), pl. 49
figs. 2-7. 1878.
HOLLICK, PALEOBOTANY OF PORTO RICO
SAPINDUS OBESUS Hlollick
Sapindus obesus iHollick. A review of the fossil flora of the West Indies,
with descriptions of new species. New York Bot. Gard., Bull. vol. 12, No. 45, p.
307, pl. 10, figs. 1, 2. Sept. 13, 1924.
This species may be compared with certain short lower leaflets of the existing
species Sapindus Saponaria Linnaeus, and with Sapindus obtusifolius Lesquereux*
from the Eocene of the western United States.
Locality: Collazo River, near base of second falls below Carretera bridge.
Collected by Bela Hubbard.
Sapindus pseudomarginatus n. sp.
Plate 73, Figure 7
Leaflet slightly unsymmetrically oblong-ovate and broadly falcate in shape,
5.25 centimeters in length by 2.3 centimeters in maximum width, rounded at the
obscurely inequilateral base, abruptly constricted to a blunt acuminate apex;
margin entire, wavy or sinuous; midrib curved or bent in conformity with the
contour of the adjacent margin; nervation pinnate; secondary nerves widely
spaced, irregularly disposed, subtending acute angles with the midrib on the
concave side and obtuse angles on the convex side, all curved and extending up-
ward in the marginal region.
This specimen closely resembles certain lower leaflets of leaves of the existing
species Sapindus marginatus Willdenow, as may be seen by comparison with
Figure 8, Plate 73, which represents a tracing of a leaflet of that species.
Locality: Station B.
Order RHAMNALES
Family RHAMNACEAE
Genus ZTZYPHus [Tournefort] Miller
Zizyphus Pseudochloroxylon n. sp.
Plate 77, Figures 1-4.
Leaves apparently oblong-ovate in shape, entire, 3-nerved from the base;
midrib straight; lateral primaries curved outward, apparently aerodrome;
tertiary nervation fine, obscurely defined, the inner ones subtending obtuse angles
with the midrib and bent or flexed in extending to the inner sides of the lateral
primaries, the outer ones subtending somewhat more acute angles with the lateral
primaries and joining in the marginal region in a series of flattened and angled
curves, with obscurely defined nervilles extending from the angles outward and
upward.
It is unfortunate that ih no instance has a perfect specimen of this species
been found. The basal outline is revealed, however, and the general shape of the
leaf is clearly indicated, as are also the essential characters of the nervation, and I
have but little hesitation in referring it to the genus Zizyphus and in comparing it
with the existing West Indian species Z. Chloroxylon (Linnaeus) Oliver, a drawing
of a leaf of which is introduced for comparison (Figure 5, Plate 77.). The close
similarity between the fossil specimens and the leaf of the existing species is so
striking that it may not be ignored in any comparison or discussion of the extinct
and existing floras.
*Lesquereux, Leo. U. S. Geol. Survey Terr., Rept. vol. 7 (The Tertiary flora), p. 266,
pl. 40, fig. 10 [excl. figs. 8, 9, 11]. 1878.
214 SCIENTIFIC SURVEY OF PORTO RICO
A leaf that approaches it in appearance, as far as general characters are
concerned, is Miconia Ettingshausenii Hollick & Berry,* from the Tertiary
(Pliocene?) of Brazil; but critical examination of the secondary nervation,
especially in the marginal region, clearly indicates that this species possesses a
true marginal vein, characteristic of the genus MAiconia and the Melastomataceae
in general, which appears to be lacking in our specimens; but specimens in which
critical minor characters may be more clearly defined, if discovered at any time in
the future, might result in change of opinion as to the generic and family relation-
ship of the specimens under consideration.
Several other fossil species, under the genera Miconia and Melastomites,
have been described and figured from America and from the Old World, but it
would be of no value to cite or to discuss them here, in connection with our
specimens, in view of the fragmentary and unsatisfactory condition of the latter.
Also, the same may be said in regard to the many fossil species of Zizyphus that
have been published; and the best we can do in the circumstances is to compare
our specimens with the existing species of the region that appears to resemble
them most closely.
Locality: Station B.
Type specimen: Figure 1.
Order MALVALES
Family MALVACEAE
Genus Malvocarpon n. gen.
Malvocarpon clarum n. sp.
Plate 75, Figure 6
Fruit globose-oblong in shape, about 1.5 centimeter in length by 1.25 centi-
meter in width, consisting of an aggregate of what appear like elongated, apiculate
carpels arranged around a common center.
Relationship of this specimen with the 1Malvaceae can hardly be questioned,
and generic relationship with Abutilon is strongly indicated by the shape and
arrangement of the carpels. The fruit as a whole, and also the individual capsules,
are more or less broken and distorted, so that original minor features have prob-
ably been destroyed; but comparison with a mature fruit of the existing species
Abutilon texense Torrey & Gray-an illustration of which is introduced for com-
parison (Figure 7, Plate 75)-may serve to show their close mutual resemblance.
Locality: Collazo River. Collected by Don Narciso Rabell.
Order MYRTALES
Family RHIZOPHORACEAE
Genus RHIZOPHORA Linnaeus
Rhizophora (?) doctrinalis n. sp.
Plate 70, Figure 6c; Plate 71, Figure 7b; Plate 82, Figure 5b.
Leaf elliptical-oblong, entire, about 8 centimeters in length by 3 centimeters
in maximum width, with rounded or curved-cuneate base and convex-cuneate
*Hollick, Arthur, & Berry, E. W. Johns Hopkins Univ. Studies in Geol., No. 5
(A late Tertiary flora from Bahia, Brazil), p. 95, pl. 12, fig. 1. 1924.
HOLLICK, PALEOBOTANY OF PORTO RICO 215
apex; texture coriaceous; midrib well defined, somewhat abruptly thickened
proximad; nervation hyphodrome.
These are very unsatisfactory specimens upon which to base a generic
identification and a specific description, and their reference to the genus Rhizophora
is therefore questioned. For comparison, however, I have introduced a tracing
of a leaf of the common Mangrove of the tropics, Rhizophora Mangle Linnaeus
(Figure 4, Plate 78), which may be seen to resemble our specimens very closely,
as far as shape and marginal contour are concerned. In neither is there
any visible trace of secondary nervation. Berry* described and figured a species
from the Eocene Tertiary of Georgia under the name Rhizophora eocenica, and
two specimens from the Pleistocene of the island of Trinidad which het referred
to'Rhizophora Mangle. These are both suggestive of the general type of leaf to
which our specimens belong; but the figures are as unsatisfactory for definite
generic identification as are those representing our specimens. Apparently these
are the only specimens of the genus Rhizophora as a fossil thus far recorded from
America; but it is possible that certain species of fossil leaves, in which secondary
nervation is lacking or obscurely defined, that have been identified under other
generic names, might all be included under one and the same generic category.
This matter was discussed by Berry (Prof. Paper 84, loc. cit., p. 145), especially in
connection with specimens from the Eocene of \Mississippi, described and figured
by Lesquereuxt and identified as Quercus chlorophylla Unger. Also it may be
pertinent to refer in this connection to a specimen from Cuba which was described
and figured by me1 under the name Mimusops Leonii. A comparison of our
Figure 5b with these several figures cited may serve to demonstrate the difficulty
of arriving at satisfactory conclusions in regard to the generic identity of fossil
specimens in which essential or critical characters are obscure or lacking.
Locality: Station B.
Type specimen: Figure 5b, Plate 82.
Family COMBRETACEAE
Genus COMBRETUM Linnaeus
Combretum Pseudojacquinii n. sp,
Plate 55, Figure 4b; Plate 78, Figure I; Plate 79, Figure 3a.
Leaves elliptical-lanceolate in shape, entire, about 8.5 centimeters in length
by 4 centimeters in maximum width, broadest at about the middle, contracted
to an apiculate apex and tapered to an acute cuneate base; midrib straight;
nervation pinnate, camptodrome; secondary nerves rather widely spaced,
irregularly alternately arranged, subtending approximately uniform angles of
about 50 with the midrib, connected in the marginal region by a series of more or
less angled loops.
As far as may be inferred from the fragmentary specimens upon which the
above description is based the general resemblance to leaves of certain existing
*Berry, E. W. U. S. Geol. Survey, Prof. Paper 84 (The upper Cretaceous and Eocene
floras of South Carolina and Georgia), p. 144, pl. 20, figs. 1, 2. 1914.
tBerry, E. W. A Pleistocene flora from the island of Trinidad. U. S. Nat. Mus.,
Proc. vol. 66, Art. 21, p. 5, pl. 2, figs. 2, 4. 1925.
Leosquereux, Leo. On species of fossil plants from the Tertiary of the State of
Mississippi. Amer. Philos. Soc., Trans. vol. 13, art. 14, p. 416, pl. 17, figs. 5,6, 7. 1867.
Hollick, Arthur. A review of the fossil flora of the West Indies, with descriptions of
new species. New York Bet. Gard., Bull. vol. 12, No. 45, p. 314, pl. 14, fig. 10. Sept. 13,
1924.
216 SCIENTIFIC SURVEY OF PORTO RICO
species of the genus Combretum is too close to be ignored. Unfortunately,
however, any of the finer nervation is lacking in our specimens; but a tracing
of a leaf of Combretum Jacquinii Grisebach (Figure 2, Plate 78) is introduced
for comparison. Very slight variations in minor characters, however, are fre-
quently essential for satisfactory generic differentiation between leaves possessing
the general shape and nervation represented by the specimens under consideration,
as may be seen by comparison with the specimen described and figured under the
name Trichilia Psendobakerii (Figure 5, Plate 72) and compared with the
existing Trichilia Bakerii De Candolle (Figure 6, Plate 72). The discovery ot
a perfect leaf of such general type, with well-preserved minor characters, mighf
result in its reference to either of the above genera, or even to some other genus.
Locality: Station B.
Type specimen: Figure 1, Plate 78.
Family MYRTACEAE
Genus EUGENIA Linnaeus
EUGENIA COMPARABILIs Hollick
Plate 55, Figures 5b, 5c; Plate 79, Figures 3c, 3d; Plate SO, Figure 1.
Eugenia comparabilis Hollick. A review of the fossil flora of the West Indies,
with descriptions of new species. New York Bot. Gard., Bull. vol. 12, No. 45, p.
310, pI. 8, figs. 2-5, 6b. Sept. 13, 1924.
The original specimens upon which this species was based came from the
Tertiary of the island of Trinidad; and Berry* also referred other specimens from
the island to the same species. A considerable range of variation may be noted
between the figures cited, but it would be dilffcult to recognize any differences
that might be satisfactorily recognized as of specific value.
Locality: Station B.
Eugenia pseudaeruginea n. sp.
Plate 80, Figure 3
Leaf apparently slightly inequilateral, oblong-elliptical or slightly obovate in
shape, entire, 7 centimeters in length by 2.5 centimeters in maximum width,
rather abruptly narrowed or constricted to an acuminate apex and apparently
tapered to an acute cuneate base; midrib slightly curved; nervation pinnate;
secondary nerves numerous, fine, irregularly spaced and disposed, consisting of a
major series -ith others of minor rank interspersed between, subtending various
angles of divergence with the midrib, all more or less curved or flexed.
This leaf is closely comparable with those of the existing species Eugenia
aeruginea De Candolle, a tracing of a leaf of which (Figure 4, Plate 80) is
introduced for comparison. Our specimen, and the one with which it is compared,
represent, in their general characters, a common foliar type that is difficult to
differentiate and identify satisfactorily. As far as discernible, however, the
generic identity of our specimen appears to be reasonably assured.
Locality: Station B.
*Berry, E. W. Johns Hopkins Univ. Studies in Geol.No. 6s (The Tertiary flora of
the island of Trinidad. George Huntington Williams Memorial Publication No. 22), p.
119, pl. 9, figs. 1-5. 1925.
HOLLICK; PALEOBOTANY OF PORTO RICO
Genus MYRCIA Do Candolle
Myrcia denuntiativa n. sp.
Plate 80, Figure 5
Leaf apparently inequilaterally ellipsoidal in shape, entire, about 10 centi-
meters in length by 4 centimeters in maximum width, curved on one side and
tapering on the other to an inequilateral cuneate base; midrib curved in con-
formity with the curvature of the leaf; nervation pinnate-marginal; secondary
nerves sub-parallel, irregularly disposed, widely and irregularly spaced, subtend-
ing angles of 750 to 850 with the midrib, extending almost straight until close to the
margin, where they connect with a slightly wavy marginal nerve.
This fragmentary specimen is suggestive of certain leaves of the Myrtaceae,
and, as far as may be seen, it resembles those of the existing genus Myrcia more
or less closely. For purposes of comparison I have introduced a tracing of a leaf
of an unnamed species of the genus, from the island of Antigua (Figure 6, Plate
80), the lower part of which may be seen to simulate our specimen quite close-
ly. Apparently there is an obscurely defined minor series of secondary nerves in
both specimens, and if this series had been better preserved in the fossil its
similarity to the other would, probably, have been more pronounced.
Locality: Station A.
StYRCIA EUGENIOIDES Hollick
Myrcia evgenioides Hollick. A review of the fossil flora of the West Indies,
with descriptions of new species. New York Bot. Gard., Bull. vol. 12, No. 45, p.
310, p1. 9, figs. 6, 7. Sept. 13, 1924.
This species was based upon leaf form alone-no trace of nervation, other
than the midrib, being discernible. It was compared with the existing species
Myrcia sylvatica De Candolle.
Locality: Collazo River, near base of second falls below Carretera bridge.
Collected by Bela Hubbard.
Genus PSIDIUM Linnaeus
Psidium dissimile n. sp.
Plate 81, Figure 1
Leaf elliptical in shape, entire, about 7.5 centimeters in length by 3.5 centi-
meters in maximum width, tapering or narrowed to the apex, curved and tapering
below into an obscurely inequilateral cuncate base; nervation pinnate; midrib
almost straight: secondary nerves relatively numerous, rather evenly spaced at
distances of 5 to 7 millimeters, sub-parallel, subtending angles of approximately
300 with the midrib, extending almost straight until reaching the marginal
region where they bend upward and become camptodrome in connecting series of
upward diminishing loops.
This leaf may be compared more or less satisfactorily with those of certain
existing species of Psidium, such as P. Guajava Linnaeus, a tracing of one of
which is represented by Figure 2, Plate 81. The simple, comparatively uniform
character of the secondary nervation-a somewhat unusual type-may be seen
to be common to both.
Locality: Station A.
SCIENTIFIC SURVEY OF PORTO RICO
PSIDIUM (?) sp. Hollick?
Plate 81, Figures 3, 4, 5a
Psidium (?) sp. Hollick. A review of the fossil flora of the West Indies,
with descriptions of new species. New York Bot. Gard., Bull. vol. 12, No. 45, p.
311, pl. 12, fig. 4. Sept. 13, 1924.
As far as may be determined from comparison between these fragmentary
specimens and the similar one (loc. cit.) to which they are referred, reproduced
in Figure 4, Plate 81, they apparently belong at least in the same generic
category, and might be regarded as specifically identical. The original generic
identification was based upon the general resemblance of the fragment to the
basal portions of certain leaves in the existing species Psidium Guajava Linnaeus
and P. pomiferum Linnaeus.
Locality:. Station B (Figure 3); collected by Bela Hubbard (Figure 4); station
A (Figure 5a).
(GAMOPETALAE)
Order PRIMULALES
Family MYRSINACEAE
Genus MYRSINE Linnaeus
Myrsine pseudoferruginea n. sp.
Plate 82, Figures 1-5a; Plate 85, Figure la
Leaves linear-oblong in shape, slightly curved to one side, about 9 centi-
meters in length by 2 to 2.4 centimeters in maximum width, narrowed to an
acute apex and tapering to the base; margin entire; midrib more or less curved
or slightly flexuous; nervation simply pinnate; secondary nerves numerous,
delicate, consisting of a major series with finer nerves between, all more or less
flexed and angled, irregularly disposed and spaced, leaving the midrib at various
angles of divergence; tertiary nervation forming with the secondaries a series of
irregular reticulations that apparently become camptodrome close to the margin.
These leaves are apparently referable to the Myrsinaceae, and they simulate
very closely, in general shape and in nervation, those of certain existing species of
3Myrsine, such as M. ferruginea Sprengel and M. myricoides Schlechtendahl,
illustrations of which are represented, respectively, by Figures 6 and 7, Plate
82.
A fossil species which may not properly be ignored in connection with any
discussion of our specimens is Hancornia minor IHollick,* a species based upon an
imperfect fragmentary leaf, with nervation very similar to that of Myrsine
pseudoferruginea, and collected at the same locality. That all of these specimens
mentioned belong to one and the same species appears to be possible, and with
Myrsine rather than Hancornia as the mutual generic relationship.
Locality: Station B (Figures 1, 3-5a, Plate 82; Figure la, Plate 85); station
A (Figure 2, Plato 82).
Type specimen: Figure 2, Plate 82.
*Hollick, Arthur. A review of the fossil flora of the West Indies, with descriptions of
new species. New York Bot. Gard., Bull. vol. 12, No. 45, p. 316, pl. 11, fig. 6. Sept. 13,
1924.
HOLLICK, PALEOBOTANY OF PORTO RICO
Genus ICACOREA Aublet
ICACOREA PRISCA Hollick
Icacorca prisca Hollick. A review of the fossil flora of the West Indies,
with descriptions of new species. New York Bot. Gard., Bull. vol. 12, No. 45, p.
312, pl. 11, fig. 5. Sept. 13, 1924.
This species was originally compared (ltoe. cit.) with the existing species
Icacorea (Ardisia) latipes Martius.
Locality: Collazo River, near base of second falls below Carretera bridge.
Collected by Bela Hubbard.
ICACOREA (?) sp.
Plate 87, Figure 5
A fragmentary leaf specimen, of unknown shape and dimensions, with
entire margin and weak, pinnately arranged secondary nervation, apparently
referable to the Myrsinaceae and possibly to the genus Icacorea.
It is possible that this specimen may represent a leaf of Icacorea prisca, the
species next preceding, but this suggestion is to be regarded as merelyincidental.
Comparison may also be made with Chrysophyllum ficifolium Berry,* from the
Tertiary of Tennessee, in connection with the discussion of which (loc. cit.) he
mentions its resemblance to leaves of Brosmium, Ardisia and Icacorea, and Ficus.
In the circumstances the problem of the generic identity of our specimen
may best be regarded as an open question.
Locality: Station A.
Genus STYLOGYNE A. De Candolle
STYLOGYNE (?) FRAGMENTA Hollick?
Plate 85, Figure 2
Stylogyne- (?) fragment Hollick. A review of the fossil flora of the West
Indies, with descriptions of new species. New York Bot. Gard., Bull. vol. 12,
No. 45, p. 313, pl. 10, fig. 6. Sept. 13, 1924.
It is unfortunate that the original specimen upon which this species was
based (loc. cit.), and the specimen represented by Figure 2, Plate 85, should
each be so fragmentary that they can not be satisfactorily compared, one with
the other, or with other species. Even their common reference to the genus
Stylogyne should be regarded as tentative only. As a possible measure of assist-
ance, however, I have introduced (Figure 3, Plate 85) a tracing of a leaf of
the existing species Stylogyne lateriflora (Swartz) Mez, which may serve as a sug-
gestion of the possible specific relationship of our specimens.
Locality: Station B.
Order EBENALES
Family SAPOTACEAE
Genus CHRYSOPHYLLUM Linnaeus
*Berry, E. W. U. S. Geol. Survey, Prof. Paper 91 (The lower Eocene flora of south-
eastern North America), p. 335, pl. 100, fig. 7. 1916.
SCIENTIFIC SURVEY OF PORTO RICO
Chrysophyllum comparable n. sp.
Plate 84, Figure 1
Leaf of unknown size and shape, tapering distad, terminating in a blunt
apiculate apex; margin entire, sinuous; rervation pinnate; secondary nerves
numerous, fine, sub-parallel, subtending angles of about 450 with the midrib.
It is possible that this foliar fragment may belong with Chrysophyllum
pseudargenteum, the species next described. It appears, however, to be narrower
distad and to have somewhat finer, more uniform secondary nervation. I am
also inclined to give it distinct specific rank by reason of its close similarity to the
existing species Chrysophyllum Cainito Linnaeus, a tracing of an apical portion of
a leaf of which species (Figure 2, Plate 84) is introduced for comparison.
Locality: Collazo River. Collected by Don Narciso Rabell.
Chrysophyllum pseudargenteum n. sp.
Plate 83, Figure 1
Leaf apparently obovate in shape ard about 10 centimeters in length by 5.4
centimeters in maximum width, entire, terminating above in a short, blunt, con-
stricted apiculate apex; midrib straight; nervation pinnate, camptodrome;
secondary nerves numerous, subtending mostly obtuse angles with the midrib,
consisting of a major series with a minor series in between that merge together
in the marginal region into a network of angled and curved areolae of various
shapes and dimensions, ultimately forming an irregular, wavy, marginal nerve.
It is unfortunate that we have only the upper part of this leaf from which to
try to visualize what it was probably like in its entirety. It may, however, be
satisfactorily compared with similar parts of broad leaf forms of the existing
species Chrysophyllum argenteum Jacquin, a tracing of a specimen of which (Figure
2, Plate 83) is introduced for the purpose.
Locality: Collazo River. Collected by Don Narciso Rabell.
Chrysophyllum pseudargenteum oblongum n. var.
Plate 83, Figures 3, 4, 5
Leaves that differ from the specific type merely in their, apparently, oblong
rather than obovate shape.
It is with some hesitation that I have decided to recognize these three frag-
mentary specimens as varietally distinct from Chrysophyllum pseudargenteum,
the species last described. They might well be included in one and the same
specific category, especially as the leaves of Chrysophyllum argenteum-the existing
species with which the fossil species was compared-differ between themselves
far more than do the fossil species and variety.
The latter might be regarded as a relatively narrow, and oblong form, and
the former as a relatively broad, and obovate form of one and the same species.
Recognition of the varietal difference may, however, be more convenient at the
present time.
Locality: Station B.
Type specimen: Figures 3 and 4 (composite).
HOLLICK, PALEOBOTANY OF PORTO RICO
Genus DIPHOLIS A. De Candolle
Dipholis pseudoleiantha n. sp.
Plate 84, Figure 3
Leaf narrowly oblong-elliptical in shape, slightly curved, entire, somewhat
Inequilateral, apparently about 4.5 to 5 centimeters in length by 1.5 centimeter
In maximum width, tapering proximad to a narrow, slightly inequilateral, cuneate
base; midrib curved in conformity with the curvature of the leaf; nervation
pinnate; secondary nerves numerous, fine, subtending angles of about 450 with
the midrib.
It is with some hesitation that this fragmentary leaf is referred to the genus
Dipholis, although it may be seen to compare quite satisfactorily with leaves of
the existing species Dipholis leiantha Standley, a tracing of a specimen of which
(Figure 4, Plate 84) is introduced for comparison. My hesitation is due to
the fact that species in several other genera might have been utilized for the same
purpose, as it represents a foliar type which, in general characters, is common to
several families as well as genera, and the preservation of every detail of outline
and nervation would be necessary for critical comparison and thoroughly satis-
factory generic determination.
Locality: Station B.
Genus SAPOTA [Plumier] Miller
Sapota agnitionalis n. sp.
Plate 75, Figure 1
Leaf oblong-ovate in shape, entire, rounded proximad to a broad convex-
cuneate base; nervation pinnate; midrib curved or bent to one side, proximad;
secondary nerves numerous, fine, subtending mostly obtuse angles with the mid-
rib.
The secondary nervation in this specimen, upon the details of which its
generic identification is dependant, is very obscurely defined. As far as discern-
ible, however, the foliar characters in general agree best with those of certain
leaves of Sapota Achras Miller, a tracing of a specimen of which (Figure 2, Plate
75) is introduced for comparison.
Comparison may also be made with a fragmentary specimen collected at the
same locality, which was described and figured by me* under the name Hancornia
pseudopubescens, and with one from a supposed Tertiary deposit in Cuba, de-
scribed and figured by me (idem, p. 314, pl. 14, fig. 10) under the name Mimusops
Leonii. The weight accorded to resemblances or to differences would, probably,
represent the personal factor in determining the generic reference of each of these
fossil specimens.
Locality: Collazo River. Collected by Don Narciso Rabell.
Genus SIDEROXYLON [Dillenius] Linnaeus
Sideroxylon aequale n. sp.
Plato 84, Figures 5, (6?)
Leaf irregularly oblong in shape, entire, apparently about 8 to 9 centimeters
in length by 3.5 centimeters in maximum width; nervation pinnate; secondary
*Hollick, Arthur. A review of the fossil flora of the West Indies, with descriptions
of new species. New York Bot. Gard., Bull. vol. 12, No. 45, p. 316, pl. 11, fig 7. Sept.
13, 1924.
SCIENTIFIC SURVEY OF PORTO RICO
nerves numerous, consisting of a major series with others of minor rank inter-
spersed between, all subtending obtuse angles with the midrib, curving upward
and becoming reticulate-camptodrome in the marginal region.
I have compared these leaves with those of the existing species Sideroxylon
foetidissima Jacquin more or less satisfactorily, at least as to their generic relation-
ship. A tracing of a leaf of that species (Figure 7, Plate 84) is introduced to
serve as an example of their close mutual similarity in shape and nervation.
The specimen represented by Figure 6 is so imperfect and the nervation is
so obscurely defined that Its specific identity is somewhat doubtful and is, there-
fore, questioned.
Locality: Station B.
Type specimen: Figure 5.
Order GENTIANALES
Family APOCYNACEAE
Genus APOCYNOPHYLLUM Unger
APOCYNOPHYLLUM PSEUDOWILLUGHBYA IIollick
Apocynophyllum pseudowillughbya Hollick. A review of the fossil flora of
the West Indies, with descriptions of new species. New York Bot. Gard., Bull.
vol. 12, No. 45, p. 318, pl. 11, fig. 3. Sept. 13, 1924.
This species was originally (loc. cit.) compared with leaves of the existing
apocynaceous species Willughbya [Willughbeia] scandens WVildenow.* The only
specimen figured is fragmentary, however, and the nervation is scanty and ob-
scurely defined. In its entirety it might have presented a different appearance,
one more nearly comparable with certain forms of Mclicocca immutata, described
and figured in this volume (page 210, Plate 73, Figures 1-4), especially If compared
with Figures 2 and 4.
Locality: Collazo River, near base of second falls below Carretera bridge.
Collected by Bela Hubbard.
APOCYNOPHYLLUM WILCOXENSE Berry?
Plate 85, Figure 4
Apocynophyllum Wlilcoxense Berry. U. S. Geol. Survey, Prof. Paper 91
(The lower Eocene flora of southeastern North America), p. 342, pl. 103, figs. 2,
3; pl. 108, fig. 4. 1916.
It would be of little use, and might be misleading, if any effort were made to
definitely identify this leaf fragment with any existing or fossil species. In its
entirety the leaf was, apparently, ligulate in shape with hyphodrome, or perhaps
very delicate secondary nervation.
It may be compared with certain fossil species that are generally regarded as
referable to the Apocynaceae, such as Apocynophyllum Wilcoxense Berry (op. cit.)
from the Tertiary of the southern United States, which he mentioned as being
similar to Nerium bilinicum Ettingshausen,f from the Tertiary of the Old World.
*The specific name adopted was an unfortunate one, as it might be thought to imply
relationship with Willugbaeya scandens (Linnaeus) Kunze (= Mikania scandens (Lin-
naeus) Willdenow), a species belonging to the Compositae.
tEttingshausen, Constantin von. K. Akad. Wissensch. (Wien], math.-naturw. Cl.,
Denkschr. vol. 28 (Die fossile flora des Tertilr-Beckens von BMlin, pt. 2), p. 218 (30), pl. 36,
fla. 20: ol. 37. fig. 2. 1868.
JIOLLICK, PALEOBOTANY OF PORTO RICO
In the circumstances the provisional reference of our specimen to Berry's species
would appear to be all that is advisable in the matter of Identification and dis-
cussion.
Locality: Station B.
Genus ASPIDOSPERMA Maitius & Zuccarini
ASPIDOSPERMA COLLAZOENS]S Hollick
Aspidosperma collazoinsis Hollick. A review of the fossil flora of the West
Indies, with descriptions of new species. New York Bot. Gard., Bull. vol. 12.
No. 45, p. 317, pl. 11, figs. 1, 2. Sept. 13, 1924.
This species was originally compared with leaves of the existing species
Aspidosperma polyneuron Mueller.
Locality: Collazo River, at base of second falls below Carretera bridge.
Collected by Bela Hubbard.
Genus ECHITES P. Browne
Echites pseudostellaris n. sp.
Plate 85, Figure 5
Leaf roughly ovate-lanceolate in shape, about 6 to 6.5 centimeters in length
by 3.5 centimeters in maximum width, abruptly contracted distad to a blunt,
slightly curved or bent apiculate apex, rounded proximad; margin entire, irregular
and wavy; midrib curved or bent distad, in conformity with the curvature of the
apex; nervation pinnate, camptodrome; secondary nerves few, widely and ir-
regularly spaced, subtending mostly obtuse angles with the midrib, connected in
the marginal region in a series of angled loops of diverse dimensions.
This specimen represents a leaf that is strikingly similar in appearance to
those of the existing species Echites stellaris Lindley, a tracing of one of which
(Figure 6, Plate 85) is introduced for comparison.
Leaves of this genus and of the allied fossil genus Echitonium have been
described and figured in paleobotanical literature, but none, apparently, with
which our species is likely to be confused.
Locality: Station B.
Genus ECHITONIUM Unger
ECHITONIUM (?) sp.
Plate 79, Figure 3b
This fragmentary leaf is manifestly too imperfect to serve as a basis for
either specific or generic description, or for satisfactory identification in connec-
tion with any described or figured specimen. It may, however, be referred
provisionally to the fossil genus Echitonium, and be compared with E. lanceolatam
Ettingshausen, as identified and figured by Berry,* from the Tertiary of Tennessee,
and with Apocynophyllum chilense Engelhardt,f from the Tertiary of Chile-
especially the specimen represented by his figure 11 (loc. cit.).
Locality: Station B.
*Berry E. V. U. S. Geol. Survey, Prof. Paper 91 (The lower Eocene floras of south-
eastern North America), p. 345, pl. 103, fig. 1. 1916.
tEnglehardt, Hermann. Ueber Tertiiirpflanzen von Chile. Senckenb. Naturf.
Gesellsch., Abh. vol. 16, No. 4, p. 659, pl. 5, figs. 9, 11. 1891.
SCIENTIFIC SURVEY OF PORTO RICO
Genus HANCORNIA Gomez
HANCORNIA MINOR Hollick
Hancornia minor Hollick. A review of the fossil flo-a of the West Indies,
with descriptions of new species. New York Bot. Gard., Bull. vol. 12, No. 45,
p. 316, pl. 11, fig. 6. Sept. 13, 1924.
This species was based upon a fragment of a median part of a leaf, which was
described and figured under the above name by reason of its apparent surficial
similarity to narrow leaf forms of the existing species Hancornia pubescens
Nees & Martius, and H. speciosa Gomez; and incidentally it may now be compared
with certain of the leaves described and figured by me (page 218, Plate 82, Figures
1-5a; Plate 85, Figure la) under the name Myrsine pseudoferruginea-especially
Figure 5a, Plate 82.
The possibility of the specific or generic identity of Hancornia minor with
Myrsine pseudoferruginea may, perhaps, be regarded as an open question. The
specimen upon which the former species was based is too fragmentary to serve as
a basis for any positive conclusion.
Locality: Collazo River, at base of second falls below Carretera bridge.
Collected by Bela Hubbard.
IIANCORNIA PSEUDOPUBESCENS IHollick
Hancornia pseudopubescens Hollick. A review of the fossil flora of the West
Indies, with descriptions of new species. New York Bot. Gard., Bull. vol. 12,
No. 45, p. 316, pl. 11, fig. 7. Sept. 13, 1924.
This species, as in connection with Hancornia minor, the one last discussed,
was based upon a median fragment of a leaf that was compared with
broad leaf forms of the existing species Hancornia pubescens Nees & 1lartius,
and H. speciosa Gomez. Whether or not this generic reference was justified may,
perhaps, be regarded as an open question, especially if the figure of Hancornia
pseudopubescens is compared with the leaf described and figured in the present
work (page 221, Plate 75, Figure 1) under the name Sapota agnitionalis. It is
always unfortunate when generic identifications are attempted, and specific descrip-
tions are published, on fragmentary material that fails to convey an adequate idea
of the species in its entirety; and yet such material may represent all that may
ever be known about the species, and a description and illustration, however in-
adequate they may be, accompanied by an appropriate name, would always seem
to be warranted.
Locality: Collazo River, at base of second falls below Carretera bridge.
Collected by Bela Hubbard.
Genus PLUMIERA Linnaeus
PLUMIERA EVIDENS Hollick
Plumiera evidens Hollick. A review of the fossil flora of the West Indies,
with descriptions of new species, New York Bot. Gard., Bull. vol. 12, No. 45, p.
315, pl. 10, fig. 5. Sept. 13, 1924.
This species was based upon a foliar fragment that was compared with leaves
of the existing species Plumiera bracteata A. De Candolle, P. rubra Linnaeus, and
P. Sucuuba Spruce.
HOLLICK, PALEOBOTANY OF PORTO RICO
Locality: Collazo River, near base of second falls below Carretera bridge.
Collected by Bela Hubbard.
Order RUBIALES
Family RUBJACEAE
Genus GUETTARDA Linnaeus
Guettarda intercalaris n. sp.
Plate 81, Figure 5b
Leaf oblong medianly, tapering proximad to a narrow cuneate base, entire,
apparently about 12 to 13 centimeters in length by 3.5 to 4 centimeters in maxi-
mum width; midrib straight; nervation simply pinnate; secondary nerves relative-
ly numerous, irregularly spaced and disposed, sub-parallel, subtending approxi-
mately uniform angles of about 450 to 500 with the midrib, extending almost
straight, or gently curved, to the marginal region, where they bend rather abruptly
upward.
This foliar fragment is referred to the genus Guettarda by reason of its
resemblance to basal parts of certain leaves that belong to existing species of the
genus. For comparison I have introduced (Figure 4, Plate 88) a tracing of a
lower part of a leaf of the genus, from Cuba.
A figure of a specimen that is somewhat suggestive of ours is Tapirira lance-
olata Engelhardt, as identified and figured by Berry,* from the Tertiary of Peru.
The resemblance to Engelhardt'sI figure that represents a specimen from the
Tertiary of Ecuador is not, however, particularly striking, and anything more
than casual reference to the species, in connection with our specimen from Porto
Rico, would not have occurred to me as pertinent had it not been for Berry's
identification.
Locality: Station A.
PLANT REMAINS OF UNDETERMINED TAXONOMIC
RELATIONSHIP
RAMULUS gen. et sp.?
Plate 71, Figure lb
Specimen consisting of a branch or twig, flattened, broken across at intervals,
thus presenting an appearance of a series of nodes and internodes, the surface
cracked and fractured latitudinally and longitudinally, forming a mesh of ir-
regular quadrilaterals.
The specimen possesses no apparent features of any diagnostic value, and it
is included merely because of its connection with the leaf with which it is intimate-
ly associated in the matrix. This association, however, does not appear to be of
any special taxonomic significance or interest.
Locality: Collazo River. Collected by Don Narciso Rabell.
*Berry, E. W. Miocene fossil plants from Northern Peru. U. S. Nat. Mus., Proc.
vol. 55 (No. 2270;, p. 291, pl. 15, fig. 1. 1919.
tEngelhardt, Hermann. Ueber Tertibrpflanzen SOd-Amerikas. Senckenb. Naturf.
Gesellsch., Abh. vol. 19, p. 15, pl. 9, fig. 4. 1895.
SCIENTIFIC SURVEY OF PORTO RICO
DISCUSSION OF THE FLORA
BOTANICAL DISCUSSION
The flora listed in the introductory part of this volume may be
seen to consist of ten elements, all of which, with two exceptions,
are of unknown or doubtful specific or generic identity, viz:
Juglans archaeoantillana Hollick.
Aniba portoricensis Hollick.
Annona saraviana Berry?
Coccoloba?
A dicotyledon (gen. et sp.?).
Nilssonia (sp.?).
Protorhipis (sp.?)
A Mesozoic fern (gen. et sp.?).
Lithothamnium (sp.?).
Lithothamnium (sp.?).
It is manifest that the tentatively identified elements are of
little value in determining the exact stratigraphic position or age of
the rocks in which they occur; and any attempt to include them in a
critical discussion of either the botanical or the geological aspects
or significance of the extinct flora of Porto Rico might lead to
erroneous conclusions. They include, however, certain generic
elements (Protorhipis and Nilssonia) which, if correctly identified,
indicate the presence of a Mesozoic (apparently Lower Cretaceous)
flora on the island, and as such they should serve as an incentive
toward further exploration and search for additional equivalent
material.
The two species definitely identified (Aniba portoricensis, and
Juglans archaeoantillana) are of particular interest for the reason
that the former is the only specific element identified without
question from the upper or youngest member of the Collazo shale
series of deposits, and because the latter represents the only
described species from a horizon very much younger or more recent
than any member of the Collazo shales. As suggested in connection
with the Cretaceous specimens, the two species above mentioned
should act as a stimulus to further exploration and search in the
localities where they were, respectively, found.
The present discussion, therefore, in the circumstances, will
be restricted in its scope to the flora of the Collazo shales; and, to
facilitate comparison, the specific elements of this flora, together
HOLLICK, PALEOBOTANY OF PORTO RICO
with surficially similar elements in the existing flora of the world,
are arranged in taxonomic sequence in the following parallel
columns:
EXISTING SPECIES OF CLOSE SURFICIAL
FLORA OF THE COLLAZO SHALES. RESEMBLANCE.
Chondriles dictyotoides n. sp........ fDictyota Bartayresii Lamoroux.
Hemitelia Brannerii Hollick & Berry.*Hiemitelia subglobosa (Underwood) Maxon
Hemitelia Imrayana Hooker.
Isoetes (?) incerta n. sp.............. Isootes riparia Engelmann.
Zamia collazotnsis n. sp............ .Zamia integrjfolia Jacquin.
(= Z. latifoliolata Prencloup).
Zamia salicina Britton.
Zamia Noblei n. sp................ Zamia integrifolia Jacquin.
Zamia pumila Linnaeus.
Zamia umbrosa Small.
Bactris Pseudocuesco n. sp.......... *Bactris Cuesco Crueger.
Iriartea collazoensis n. sp........... Iriartea setigera Martius.
Manicaria portoricensis n. sp ........ M. anicaria sacchifera Gaertner.
Palmocarpon acrocomioides Hollick.. \Acrocomia aculeata Lodd.
Acrocomia crispa (H. B. K.) C. F. Baker
Palmocarpon cetera n. sp........... Astrocaryum Jauarii Martius.
Cocos Datil Grisebach & Drude.
Cocos eriospatha Martius.
Palmocarpon excmplare n. sp ........ Cocos Inarjoi Trail.
Palmocarpon opinabile n. sp........ Manicaria sp.?
Palmocarpon Rabcllii n. sp......... Copernicia cerifera Martius.
Palmacites alius n. sp .............. ?
Palmacites conformis n. sp.......... ?
Palmacites sparsistriatus n. sp....... ?
Palmophyllum sp. Hollick .......... ?
Palmophyllun sp. (fragment of pet-
iole) ? H ollick............. .... ?
Musophyllumn sp.................. Musa sp?
Ficus hyphodroma n. sp............ *Ficus elastic Roxburgh.
Ficus Schimperii Lesquereux. ...... Ficus ferruginea Desfontaines.
Ficus populiformis Schott.
Ficus velu tina Willdenow.
Ficus vexativus n. sp............... Ficus sp.?
Ficus sp .................. ... Ficus sp.?
Annona cetera n. sp................ tAnnona squanmosa Linnaeus.
Annona pseudoglabra n. sp.......... tAnnona glabra Linnaeus.
Acrodiclidium pseudosalicifolium n. sp. tA crodiclidium salicifolium (Swartz)
Grisebach.
Acrodiclidium Pseudocanelo n. sp.. . Acrodiclidium Canelo Rose.
Aniba collazoinsis n. sp............ *Aniba riparia (Nees) Mez.
*Indicates genera represented in the native flora of Porto Rico.
ftndicates species represented in the native flora of Porto Rico.
SCIENTIFIC SURVEY OF PORTO RICO
Hufelandia portoricensis (ITollick)
n. comb........................ fHufelandia pendula (Swartz) Nees.
Misanteca dubiosa (Hollick) n. comb. t Misanteca triandra (Swartz) Mez.
Oreodaphne mississippiensis Berry?.. *Persea pubescens (Pursh) Sargent.
Oreodaphne (?) sp.
Inga curta n. sp................... *Inga affinis De Candolle.
Inga pseudinsignis n. sp............ Inga insignis Kunth.
Inga pseudonobilis Hollick ......... Inga nobilis Willdenow.
Inga pseudospuria n. sp............ Inga spuria Humboldt & Bonpland.
Inga (?) sp. Hollick ............... Inga maritima Bentham.
Pithecellobium (?) imperfectum n. sp. *Pithecellobium gracilliflorum Blake.
Pithecellobium pseudotrapezifolium
Hollick........................ Pithecellobiwn trapezifolium Bentham.
Pithecellobium vexativou Hollick. . Pithecellobium ligustrinum (Jacquin)
Klotz.
Cassia (?) dubiosa Hollick ......... *Cassia sp.?
Pithecellobium sp.?
Cassia evidens n. sp................ tPeiranisia [Cassia] biflora (Linnaeus)
Britton & Rose.
Cassia imitativa n. sp.............. Cassia sp.?
Cassia imparilts n. sp.............. Cassia sp.?
Cassia ordinaria n. sp.............. Cassia leptocarpa hirsula Bentham.
Cassia puryearensis Berry.......... tAdipera [Cassia] laevigata (Willdenow)
Britton & Rose.
Cassia corymbosa Lamarck.
Cassia visibilis n. sp............... tDitremexa [Cassia] occidentalis (Linnaeus)
Britton & Rose
Copaiva oligocenica n. sp........... Copaiva chiriquensis Pittier.
Cynometra Rabellii n. sp........... *Cynometra trinitensis Oliver.
Lonchocarpus praelatifolius n. sp.. . tLonchocarpuslatifolius (Willdenow) H.B K
Sophora (?) suspect Hollick....... .*Sophora sp.?
Guarea opinabilis n. sp.......... .. *Guarea sp.?
Trichilia evidens n. sp.............. Trichilia odorala Andrews.
Trichilia Pseudobakerii n. sp........ Trichilia Bakerii C. De Candolle.
Trichilia pseudohirta n. sp.......... .Trichilia hirta Linnaeus.
Trichilia spatulata n. sp............ Trichilia trifolia Linnaeus.
Melicocca immutata n. sp........... tMelicocca bijuga Linnaeus.
M elicocca sp................. . . M elicocca sp.?
Sapindus Britionii n. sp.......... tSapindus Saponaria Linnacus.
Sapindus gracilenius n. sp.......... Sapindus Saponaria Linnaeus.
Sapindus obesus Hollick............ Sapindus Saponaria Linnaeus.
Sapindus pseudomarginatus n. sp.. . Sapindus marginatus Willdenow.
Zizyphus Psewlochloroxylon n. sp.. . *Zizyphu.s Chloroxylon (Linnaeus) Oliver.
Malvocarpon claruam n. gen. et sp.. .. *Abutilon texense Torrey & Gray.
Rhizophora (?) doctrinalis n. sp.. ... Rhizophora Mangle Linnaeus.
Combretum Pscudojacquinii n. sp., . Combretum Jacquinii Grisebach.
Eugenia comparabilis Hollick....... Eugenia coffeifolia De Candolle.
Eugenia forameoides Richards.
HOLLICK, PALEOBOTANY OF PORTO RICO 229
Eugenia pseudaeruginea n. sp ....... tEugenia aeruginea De Candolle.
Myrcia denuntiativa n. sp.......... *Myrcia sp.?
Myrcia eugenioides Hollick......... Myrcia sylvatica De Candolle.
Psidium dissimile n. sp............ tPsidium Guajava Linnaeus.
Psidium (?) sp. Iollick............ Psidium. pomiiferum Linnaeus.
Myrsine pseudoferruginea n. sp...... Myrsine ferruginea Sprengel.
Myrsine myricoides Schlechtendahl.
Icacorea prisca Hollick............. *Icacorea [Ardisia] latipes Martius.
Icacorea (?) sp ................ ... Icacorea sp.?
Stylogyne (?) fragmenta Hollick..... tStylogyne [Ardisia] laterijlora (Swartz)
Mez.
I. i.i..." ,; ..... comparabile n. sp.. . tChrysophyllumn Cainito Linnaeus.
Chrysophyllum pseudargenteum n. sp. fChrysophyllumn argenteum Jacquin.
Chrysophyllum pseudargenteum oblon-
gum n. var..................... Chrysophyllum argenteum Jacquin.
Dipholis pseudoleiantha n. sp..... *Dipholis leiantha Standley.
Sapota agnitionalis n. sp............ tSapota Achras Miller.
Sideroxylon aequale n. sp........... f Sideroxylon foetidissima Jacquin.
Apocynophyllumn pseudowillughbya
Hollick........................ Willughbya scandens W illdenow.
Apocynophyllum Wilcoxense Berry?. Nerium sp.?
'Aspidospermna collazoensis Hollick. . Aspidosperma polyneuron Mueller.
Echiles pseudostellaris n. sp......... *Echiles stellaris Lindley.
Echitonium (?) sp................. Echites sp.?
Hancornia minor Hollick........... Hancornia pubescens Nees & Martius.
Hancornia speciosa Gomez.
Hancornia pseudopubescens Hollick. Ifancornia pubescens Nees & Martins.
Plumiera evidens Hollick........... Plumiera bracteata A. De Candolle.
f Plumiera ruba Linnaeus.
Plumiera sucuuba Spruce.
Guettarda intercalaris n. sp ......... *Guettarda sp.?
Ramulus gen. et sp.? ........ . .... ?
A numerical analysis of the above listed flora of the Collazo
shales shows it to consist of 91 specific elements or entities-
using these terms in the broadest sense-included in 51 genera,
24 families, and 20 orders, grouped, taxonomically, as follows:
Species Genera Families Orders
Thallophyta.................. 1 1 1 1
Pteridophyta................. 2 2 2 2
Cymnospermae.............. 2 1 1 1
Monocotyledonae.............. 14 7 2 2
Dicotyledonae ................ 71 39 17 13
Undetermined................. 1 1 1 1
91 51 24 20
SCIENTIFIC SURVEY OF PORTO RICO
The single representative of the Thallophyta (Chondrites dic-
tyotoides) in the flora, which apparently represents a marine or
brackish-water alga, is of ecological interest as indicating lagoon or
estuary conditions under which the sediments were deposited in
which it and its associated land plant remains were entombed and
preserved; and such conditions are further indicated by the presence
of fresh- and brackish-water molluscs, crustaceans, and other animal
remains in the same series of sediments. Chrondites dictyotoides
may be compared with certain species of the existing cosmopolitan
genus Dictyota, especially with D. Bartayresii Lamoroux, a marine
alga native in Porto Rican waters and in tropical and subtropical
waters elsewhere along the Atlantic coast. The generic name
Chondrites, representing a fossil algal genus, has been rather loosely
applied; but it is largely associated with species of Tertiary age.
from both the Old World and the New.
The only fern in the flora, Hemitelia Brannerii Hollick &
Berry, is, apparently, specifically identical with specimens previous-
ly described from the Tertiary of Brazil, and not recorded from
elsewhere until found in Porto Rico. The genus Hemitelia, in our
existing flora, is cosmopolitan and tropical in its geographical
distribution; and three of its approximately forty recognized species
are native in Porto Rico.
The genus Isoetes includes several fossil species, of Tertiary
age, and some fifty existing species of tropical and temperate dis-
tribution, in the Old World and in North and South America,
Mexico and Cuba. It is not, however, an element in the existing
flora of Porto Rico. The species from the Callazo shales (Isoites
(?) incerta), if correctly identified generically, is, therefore, of inter-
est by reason of its geographical location, in addition to its purely
botanical interest as the second one of the only two known represen-
tatives of the Pteridophyta in the flora of the shales, and, incidental-
ly, as one of the only two recorded fossil species of Isoetes from the
New World.
The Gymnospermae are represented by two more or less well
defined species of Zamia (Z. collazoensis and Z. Noblei) and, possi-
bly, by one or more additional species or varietal forms, which might
be differentiated from them. Their remains are, relatively,
abundant, and it is apparent that the genus was an important
element in the flora. Cycads in general, although well represented
in rocks of Mesozoic age throughout the world, are very rare as
HOLLICK, PALEOBOTANY OF PORTO RICO
Tertiary fossils, either in the Old World or the New, and the dis-
covery of these abundant and well defined remains in the Collazo
shales is of considerable interest. In the existing flora the genus
Zamia is represented by about thirty species, all of which are con-
fined to the tropical and subtropical regions of the New World.
Three of these are native in Porto Rico, one of which (Z. integrifolia
Jacquin) is so clearly comparable with Z. collazoensis as to be almost
indistinguishable from it, as far as surficial foliar characters are
discernible.
The Monocotyledonae are well represented by a number of
species referable to the Arecales. Stem and leaf fragments and
fruits of palms are abundant, both relatively and actually. The
number of genera included is six, and the number of described
species is thirteen; but certain of these may or may not represent
distinct genera or species, in as much as some are based wholly
upon fragments of woody parts, others on foliar remains, and others
on fruits. The comprehensive fossil genera Palmocarpon, Palma-
cites and Palmophyllum, and the species included under them can-
not, therefore, be properly utilized in any enumeration in connec-
tion with either category. They indicate however, that the palms
were an abundant and important element in the flora. The genus
Bactris is represented in the existing native flora of Porto Rico by
one species, and Iriartea and Manicaria are tropical American
genera. How many other existing genera may be represented in
the other fruits and fragmentary remains could be inferred, approxi-
mately; but any such inference would be of no statistical value.
The Scitaminales are represented by a single genus, Musophyl-
lum, the identification being based upon a single leaf fragment
which might almost equally well be referred to the existing genus
Musa. Definite and satisfactory identification of any fossil remains
of Musa in the flora of the Collazo shales would be of great interest,
in view of the doubt that yet obtains in regard to the nativity of
the genus in the New World, even though the identity of banana
seeds, found in Middle Tertiary deposits in South America, has not
been questioned, as far as I am aware.
An analysis of the list of the Dicotyledonae shows that the
twenty-five genera of the Choripetalae include fifty-two species, or
over fifty-seven per cent of the entire flora. Nine orders and
eleven families are represented in the group, of which the Rosales,
with one family (Leguminosac), seven genera, and nineteen species
SCIENTIFIC SURVEY OF PORTO RICO
is the most extensive of the entire flora, in both generic and specific
elements; and the genus Cassia, with seven species is the largest of
the genera. Furthermore, all of the genera, with three exceptions
(Oreodaphne, Malvocarpon, and Combretuin) are represented by one
or more species each in the existing flora of Porto Rico.
The Gamopetalae are represented by fourteen genera and
nineteen species, included in four orders and four families, of which
the Gentianales, with one family (Apocynaceae), six genera, and
eight species, contains the largest number of generic and specific
elements. Two of the fourteen genera (Apocynophyllum and
Echitonium) are fossil genera. Three (Myrsine, Aspidospermum,
and Hancornia) are not represented in the native flora of Porto
Rico, although all are represented in the flora of the adjacent
mainland. The remaining nine genera are elements in the native
flora of the island.
If the flora as a whole is analysed, in connection with its ex-
isting generic elements, certain salient facts are obvious. Every
genus is tropical, or tropical and subtropical in its distribution.
Some, such as Ficus, Cassia, Eugenia, Myrsine, etc., are cosmopol-
itan; others, such as Zamia, Bactris, Iriartea, Manicaria, Acrodiclid-
ium, Inga, Hancornia, etc., are exclusively American genera.
Thirty-five of the genera are elements in the present flora of Porto
Rico; and a number of the species such as Zamia collazoensis, Z.
Noblei, Annona cetera, Acrodiclidium pseudosalicifolium, Loncho-
carpus praelatifolius, Trichilia pseudohirta, etc., are so closely
comparable with certain existing species in the flora of the island
that any differences between them are not obvious, and it would be
practically impossible to differentiate between them in any de-
scriptions based upon surficial foliar characters alone.
GEOLOGICAL DISCUSSION
The flora of the Collazo shales is of limited value as an index of
the exact geological age of the shales, in as much as it is, for the
most part, composed of heretofore undescribed species, nearly all
of them closely similar to certain ones now in existence. Of the
ninety-one recognized floral entities, sixty-four, or about seventy
per cent., represent new species, These, therefore, are not available
for the purpose of critical stratigraphic identification; and, of the
remaining twenty-seven, which represent previously described
species, twenty-two are from the Collazo shales, and consequently
HOLLICK, PALEOBOTANY OF PORTO RICO
they also are negligible for the same purpose. The genera, however,
are clearly indicative of the major stratigraphic position of the
flora, and from these we may be justified in assuming that its
Tertiary age will not be questioned; but whether it may properly be
regarded as Eocene, Oligocene, Miocene, or even Pliocene represents
a legitimate subject for discussion.
Those who have studied the stratigraphic relations of the
shales in the field, and those who have studied the included fauna,
are not entirely in agreement in their inferences and conclusions as
to the exact age of the shales; but the range of difference is so slight
as to be of little consequence-the extremes being upper Eocene
to lower Miocene.
Only five extra-territorial elements have been recorded in the
flora, and it would be hazardous to deduce any critical conclusions,
or even inferences, based upon species that represent such a small
percentage of the whole. It may be pertinent, however, to list
them, together with data relating to their stratigraphic status and
geographic distribution, as far as known, viz:
Hemitelia Brannerii Hollick & Berry. Upper Miocene or Pliocene? Brazil.
Ficus Schimperii Lesquereux........ Eocene (Wilcox formation), southern
United States; (Raton formation), Colo.,
N. Mex.
Oreodaphne mississipiensis Berry. Eocene (Wilcox formation), southern
United States.
Cassia puryearensis Berry.......... Eocene (Wilcox formation), southern
United States.
Apocynophyllum Wilcoxensis Berry.. Eocene (Wilcox formation), southern
United States; (Raton formation), Colo
From the above data it may be seen that the species listed are
in substantial agreement with the stratigraphy of the shales as
interpreted by other observers.
In the tabulation that follows an effort has been made to com-
pare certain of the elements of the Collazo shales flora with de-
scribed fossil species that appear to simulate them most closely:
FOSSIL SPECIES OF SURFICIAL RESEM-
FLORA OF THE COLLAZO SALES BALANCE
Chondrites dictyotoides n. sp. ......... Chondrites Targione arbuscula (Fischer-
Ooster) Heer.
Chondrites inlricalus (Brongniart) Stern-
berg.
Hemitelia Brannerii Hollick & Berry?
Isoetes (?) incerta n. sp............. Isoates Braunii (Unger) Heer.
Zamia collazoinsis n. sp............ Zamia tertiaria Engelhardt.
SCIENTIFIC SURVEY OF PORTO RICO
Zamia Noblei n. sp................
Bactris Pseudocuesco n. sp..........
Iriartca collazoensis n. sp...........
Manicaria portoricensis n. sp.......
Palmocarpon acrocomioides Hollick..
Palmocarpon cetera n. sp...........
Palmocarpon exemplare n. sp.
Palmocarpon opinabile n. sp........
Palmocarpon Rabellii n. sp.
Palmacites alius n. sp.
Palmacites conformis n. sp.
Palmacites sparsistriatus n. sp.......
Palmophyllum sp. Hollick.
Palmophyllum sp. (fragment of pet-
iole)? Hollick.
M usophyllum sp..................
Ficus hyphodroma n. sp............
Ficus Schimperii Lesquereux .......
Ficus vexativus n. sp............. .
Ficus sp.
Annona cetera n. sp.............. .
Annona pseudoglabra n. sp.
Acrodiclidium pseudosalicifolium n.
sp.
Acrodiclidium Pseudocanclo n. sp.. ...
Aniba collazoansis n. sp.. ........
IIufelandia porloricensis (Hollick) n.
com b ........................ .
Misanteca dubiosa (Hollick) n. comb.
Oreodaphne mississippicnsis Berry?..
Inga curta n. sp ................. .
Inga pseudinsignis n. sp.
Inga pseudonobilis Hollick.
Inga pseudospuria n. sp.
Inga (?) sp. Hollick ...............
Pithecellobium (?) imperfcctum n. sp..
Pithecellobium pscudotrapezifoliumi
H ollick ...................... .
Pithcccllobiun vcxativum Hollick . ..
Zamia (?) Wilcoxensis Berry.
Palmocarpon (?) globosum Lesquereux.
Palmacites [Antholithes] Martii Heer.
Palmocarpon truncatum major Lesquereux.
Pal mocarpon mexicanum Lesquereux.
Nipadites unbonatus Bowerbank.
Pal macitcs canaliculatus Heer.
Musophyllum trinitense Hollick.
Musophyllum elegans Engelhardt.
Ficus e6lignitica Berry.
Ficus monodon (Lesquereux) Berry.
Ficus cuspidata Watelet.
Ficus comparabilis Hollick.
Anona [Annona] Wilcoxiana Berry.
Anona [Annona] lignitum Unger.
Ocotca pseudomartinicensis Hollick.
Nectandra curvatifolia Engelhardt.
Nectandra Woodringii Berry.
Trigonia varians Engelhardt.
Nectandra Glennii Berry.
Laurus attcnuata Watelet. Litsaca ex-
pansa Saporta & Marion.
Inga mississippicnsis Berry.
Leguminosites cclastroides Heer.
Inga sanchezensis Berry.
Caesalpinia Wilcoxiana Berry.
Cassia longifolia Engelhardt.
Pithccolobium tenuifolium Engelhardt.
Laguminosites copaiferoides Engelhardt.
HOLLICK, PALEOBOTANY OF PORTO RICO
Cassia (?) dubiosa Hollick .........
Cassia evidens n. sp............. . .
Cassia imitativa n. sp........... . .
Cassia imparilis n. sp............ .
Cassia ordinaria n. sp..............
Cassia puryearensis Berry ..........
Cassia visibilis n. sp...............
Copaifera oligocenica n. sp.
Cynometra Rabellii n. sp.
Lonchocarpus praelatifolius n. sp.
Sophora (?) suspect Hollick ........
Guarea opinabilis n. sp.............
Trichilia evidens n. sp.
Trichilia Pseudobakerii n. sp........
Trichilia pseudohirta n. sp..........
Trichilia spatulata n. sp.
Melicocca immutata n. sp...........
M elicocca sp......................
Sapindus Brillonii n. sp............
Sapindus gracilentus n. sp..........
Sapindus obesus Hollick ...........
Cassia Glennii Berry.
Cassia dimidiato-linearis Engelhardt.
Cassia Glennii Berry.
Cassia Berenices Unger. C. hyperborea
Unger.
Cassia emarginata Berry.
Cassia Wilcoxiana Berry.
Cassia Glennii Berry.
Cassia Berenices Unger
Cassia Berenices Unger.
Sophora paleolobifolia Berry.
Tapiria [Tapirira] lanceolala Engelhardt.
Geissanthus Brittonii Hollick.
Anona [Annona] cochambensis Berry.
Apocynophyllum pseudowill ughbya Hollick.
Melicocca immutata n. sp.
Sapindus praesaponaria Hollick & Berry.
Sapindus falcifolius Alex. Braun.
Sapindus angustifolius Lesquereux.
Supindus obtlsifolius Lesquereux.
Sapindus caudatus Lesquereux. S. acumin-
atls Engelhardt.
Sapindus pseudomarginatus n. sp.
Zlzyphus Pseudochloroxylon n. sp.. Miconia Ettingshavsenii Hollick & Berry.
Malvocarpon clarum n. sp.
Rhizophora (?) doctrinalis n. sp ...... Rhizophora eocenica Berry.
Mimusops Leonii Hollick.
Combretum Pseudojacquinii n. sp.. . Trichilia Pseudobakerii n. sp.
Eugenia coinparabilis Hollick....... Myrcia (Cryptomyrcia) nitens Engelhardt.
Eugenia pseudaeruginea n. sp.
Myrcia denuntiativa n. sp.
Myrcia eugenioides Hollick ......... Sapindus hispaniolana Berry.
Psidium dissimile n. sp.
Psidium (?) sp. Hollick
Myrsine pseudoferruginea n. sp...... Hancornia minor Hollick.
Icacorea prisca Hollick.
Icacorea (?) sp .................... Icacorea prisca Hollick.
Chrysophyllum ficifolium Berry.
Stylogyne (?) fragment Hollick?
( I. ",.,f.i... comparable n. sp.. **i' .;l ... pseudargenteum n. sp.
Chrysophyllum pseudargenteum n. sp.
Chrysophyllum pseudargenteum oblon-
gum n. var.
SCIENTIFIC SURVEY OF PORTO RICO
Dipholis pseudoleiantha n. sp.
Sapota agnitionalis n. sp........... Hancornia pseudopubescens Hollick.
Mimuisops Leonii Hollick.
Sideroxylon aequale n. sp.
Apocynophyllum psendowillughbya
Hollick ................... . ... elicocca immutala n. sp.
Apocynophyllum Wilcox.rense Berry?. Nerium bilinicum Ettingshausen.
Nerium sarthacense Saporta.
Aspidosperma collazoensis Hollick.
Echites pseudoslellaris n. sp.
Echitonium (?) sp ................. Echilonium lanceolatmn Ettingshausen.
Apocynophylluit chilense Engelhardt.
Hancornia minor Hollick .......... Myrsine pseudoferruginea n. sp.
Hancornia psendopubescens Hollick. Sapola agnitionalis n. sp.
Plumiera evidens Hollick ........... Apocynophyllum mexicanum Berry.
Guettarda intercalaris n. sp ......... Tapiria [Tapirira] lanceolata Engelhardt.
Ramulus gen. et sp.?
An analysis of the data included in the above tabulation might
be utilized to deduce various inferences or conclusions; but they
would not be likely to result in anything that could alter the point
of view previously outlined. In other words the general facies of
the flora of the Collazo shales, compared with that of a flora com-
posed of fossil species most closely similar, would be indicative of
lower-middle Tertiary age.
Incidentally it may also be of interest to compare this Tertiary
flora of a tropical region with approximately contemporaneous floras
northward, and to infer from their elements the climatic conditions
that probably prevailed in the region where each flora was growing.
Specific comparisons would be too few to be of any value; but com-
parisons based upon the generic elements in each may be utilized to
advantage.
As previously outlined, the forty-two existing genera included
in the flora of the Collazo shales are all representative of tropical
and subtropical environment. The flora, in its general facies, is
essentially 100 per cent. tropical.
The Eocene (Wilcox and Lagrange) flora of the southern part,
or Gulf region of the North American Continent, is composed of
about 130 generic elements, of which ninety-four are identified as
existing genera; and of these about eighty are of tropical distribu-
tion, and the others might be classed a subtropical and warm
temperate. Among the former may be noted Zamia, Ficus,
Annona, Oreodaphne, Inga, Pithecellobium, Cassia, Sophora, Sapin-
dus, Zizyphus, Combretum, Eugenia, Myrcia, Icacorea, Chryso-
HOLLICK, PALEOBOTANY OF PORTO RICO
phyllum, Sideroxylon, and Guettarda-all of which are represented
in the flora of the Collazo shales-and Pistia, Chamiaedorea,
Sabalites [= Sabal?], Canna,Artocarpus, Coccolobis, Pisonia, Chryso-
balanus, Dalbergia, Canavalia, Simaruba, Carapa, Cedrela, Vantanea,
Banisteria, Sterculia, etc. About a dozen genera, however, are
indicative of cooler climatic conditions, among which may be
noted Hicoria, Juglans, Quercus, Planera, Rhamnus, Cornus, and
Fraxinus. It may be regarded as a flora that is about eighty-five
per cent. tropical-certain of the genera indicating climatic
conditions warmer than now obtain except in the extreme southern
part of the region.
The Eocene (Lance and Fort Union) flora of the Rocky
Mountain region includes, approximately, 108 generic elements, of
which eighty-five are identified as existing genera. If these genera
are analyzed it may be seen that only about ten, or approximately
eleven per cent., would be classed as strictly tropical, viz, Ottelia,
Sabal, Artocarpus, Laurns, Cabomba, Parrolia, Bauhinia, Acacia,
Sapindus, and Zizyphus; and about ten others as subtropical or
warm temperate, such as Taxodium, Pistia, Ficus, Nelumbo,
Magnolia, Grewia, etc. The flora is one that would be representa-
tive of the North American temperate zone today, with about
twenty-two per cent. tropical and subtropical generic elements
included in it, which indicate warmer climatic conditions than those
that now prevail in the same latitude.
If the Eocene floras of the Arctic region are analyzed and com-
pared in a similar manner the fact will be apparent that although
they include a few tropical and subtropical genera, such as Taxo-
dium, several cycads and palms, Artocarpus, Ficus, Laurus, Magnolia,
etc., their general facies suggests a temperate zone flora, evidencing
mild instead of the arctic climatic conditions that now prevail
throughout almost the entire region.
The essential fact of geologic interest that the identification
and analysis of the floral elements of the Collazo shales have
emphasized is that, although mild climatic conditions evidently
prevailed throughout the northern hemisphere during early Tertiary
time, zonal differentiation was already established to a limited
extent. There was a distinctly southern or tropical flora in the
equatorial region, and a northern flora in which the tropical elements
became fewer and fewer in regions successively further and further
north, as temperate zone elements supplanted them.
238 SCIENTIFIC SURVEY OF PORTO RICO
Old generic types were eliminated and new generic types were
evolved in the northern regions as a result of changing climatic con-
ditions; but the flora of the southern regions underwent a barely per-
ceptible change during the same period and subsequently. What-
ever new types may have been evolved were the result of factors
other than those of climatic change or fluctuation.
SUMMARY AND CONCLUSIONS
The flora of the Collazo shales represents one of tropical en-
vironment; its habitat was in the vicinity and on the borders of
lagoons or estuaries, in which brackish water was present; it is
typically New World in its general facies; it is almost identical,
generically, with the existing flora of Porto Rico and adjacent re-
gions; it is Tertiary in age and is referable to the lower-middle part
of that period.
PLATES
239
PLATE 51
Page
Figure 1-. Chondrites dictyotoides n. sp. Station A....................... 181
Figure 2. Chondrites Targione arbuscula (Fischer-Ooster) Heer. Introduced
for com parison......................................... 181
Figure 3. Sphaerococcus crispiforinis (Schlotheim) Heer. Introduced for
com parison............................. .............. 181
Figure 4. Hemilelia Brannerii Hollick & Berry? Station A.............. 182
Figures 5, 6. Hemitelia Brannerii Hollick & Berry. Introduced for com-
parison ............................ .................... 182
Figure 7. Hemitelia (Cnemidaria) subglobosa (Underwood) Maxon. Costa
Rica. W. A. Maxon, 1906. No. 382. Introduced for compari-
son ..................................... ........... 182
Figure 8. Hemitelia Imrayana Hooker. Grenada, W. I. W. E. Broadway,
1896. Introduced for comparison ..................... ... 182
SCIENTIFIC SURVEY OF PORTO RICO AND THE VIRGIN ISLANDS
I, I
Kr
I r
',\/ 5 6
1
A:# p
/- 1 '
-< i ).,^
4b '
K "*< ''
r ^ "s {J6
n /
/T\]i=^
/ \// c
THALLOPHYTA. PTERIDOPHYTA.
VOL. VII, PL, 51
PLATE 52
Page
Figures 1, 2. Iso~les (?) incerta n. sp. Station B ..................... 183
Figure 3. Isoeles riparia Engelmann. Wilmington, Del. W. M. Canby,
1866. Introduced for comparison ......................... 183
Figure 4. Isoates Braunii (Unger) Heer. Introduced for comparison. 183
SCIENTIFIC SURVEY OF PORTO RICO AND THE VIRGIN ISLANDS
'> I?>
F
,1
N>
2
PTERIDOPHYTA
VoL. VII, Pi.. 52
PLATE 53
Page
Figure 1. Zanmia collazoensis n. sp. Station B .......................... 184
Figure 2. Zatnia integrifolia Jacquin (= Z. latifoliolata Preneloup). Porto
Rico. N. L. Britton and J. F. Cowell, 1906. No. 1422. Intro-
duced for comparison ............................... . ... 184
Figure 3. Zamia collazolnsisn. sp. Station B ...................... . 184
Figure 4. Zamia salicina Britton. Cuba. N. L. Britton, E. G. Britton,
and Percy Wilson, 1916. No. 14166. Introduced for compari-
son . . ................................ ......... .... 184
Figure 5. Zamia collazonsis n. sp. (distorted). Station B ............... 184
Figure 6. Zamiia salicina Britton (distorted). Cuba. N. L. Britton, E. G.
Britton, and Percy Wilson, 1916. No. 14166. Introduced
for comparison ....... ................................ 184
Figure 7. Zamia collazofnsis n. sp.? Station B ......................... 184
Figure 8. Zamia sp. Cuba. Ant. Luna, 1920. No. 585. Introduced for
com parison.............. ............. ........... 184
Figure 9. ZamniaNoblein. sp. Station B.............................. 185
Figure 10. Zamia Noblei n. sp. Station B .............................. 185
Figure 11. Zainia pumrila Linnaeus. Cuba. J. A. Shafer, 1909. No. 1402.
Introduced for comparison .............................. 185
Figure 12. Zamia tertiaria Engelhardt. Introduced for comparison. ....... 184
Figure 13. Zamia (?) WVilco.rensisBerry. Introduced for comparison ....... 184
SCIENTIFIC SURVEY OF PORT RICO AND THE VIRGIN ISLANDS
A
11
3
5
12
13
GYMNOSPERMAE.
\I
9
VoL. VII, PL. 53
PLATE 54
Page
Figure 1. Zaimia Noblei 1n. sp. Station B. . . .................. 185
Figure 2. Zamia integrifolia Jacquin (= Z. latifoliolata Preneloup). Jamai-
ca. N. L. Britton and Arthur ITollick, 1908. No. 2012. Intro-
duced for com prison ........................... .. 185
Figure 3a. Zamia Noblei n. sp. Station B ....................... ... 185
Figure 3b. Annona psevdoglabra n. sp. Station B ....................... 196
Figure 4. Annona glabra Linnaeus. Isle of Pines, Cuba. A. H. Curtiss,
1904. No. 502. Introduced for comparison ......... ...... 197
Figure 5. Zamia umbrosa Small. Florida. J. K. Small and J. J. Carter,
1903. No. 1313. Introduced for comparison .............. 185
SCIENTIFIC SURVEY OF PORTO RICO AND THE VIRGIN ISLANDS
GYMNOSPERMAE. ANGIOSPERMAE (DICOTYLEDONAE).
VOL VII, PL. 54
PLATE 55
Page
Figures 1-4a. Zamia Noblei n. sp ..... .......................... 185
Figs. 1. 2. Station A.
Figs. 3, 4a. Station B.
Figure 4b. Combretun Pseudojocquiniin. sp. Station B ................. 215
Figure 5a. Zamia Noblei n. sp. Station B .......................... 185
Figures 5b, 5c. Eugenia comparabilis Hollick. Station B ................. 216
Figure 5d. Sapindus Britltonii n. sp. Station B...................... ... 211
SCIENTIFIC SURVEY OF PORTO RICO AND THE VIRGIN ISLANDS
VOL. VII, PL. 55
GYMNOSPERMAE. ANGIOSPERMAE (DICOTYLEDONAE).
PLATE 56
Page
Figure 1. Iriartea collazoensis n. sp. Collected by Don Narciso Rabell .... 187
Figure 2. Arenga saccharifera Labillardiere. Engler and Prantl. Die
natfirlichen Pflanzenfamilien, theil 2, abth. 3, p. 54, fig. 43C1.
1889. Introduced for comparison .................... . ... 187
Figure 3. Iriarlea setigera Martius. Flora Braziliensis, vol. 3, pt. 2, pl. 127
in part. 1882. Introduced for comparison .................. 187
Figure 4. Palmocarpon acrocomioides Hollick. Station A................ 189
Figure 5. Acrocomia crispa (Humboldt, Bonpland, and Kunth) C. F. Baker.
Cuba. N. L. Britton, E. G. Britton, and Percy Wilson, 1916.
No. 15454. Introduced for comparison ............... . ... 189
Figure 6. Cocos Inarjai Trail. Martius. Flora Braziliensis, vol. 3, pt. 2,
pl. 88, fig. 2. 1882. Introduced for comparison .............. 190
Figure 7. Palmnacites [Antholithes] Martii Hecr. Introduced for comparison 189
Figures 8, 9. Palmocarponexemplare n. sp... .......... .......... 190
Fig. 8. Station B.
Fig. 9. Station A.
Figure 10. Palmocarpon cetera n. sp. Station A ......................... 189
Figure 11. Palmocarpon Rabelliin. sp. Collected by Don Narciso Rabell. .. 191
Figure 12. Eleais guincdnsis Jacquin. Hawaii (cultivated). Whitman,
1909. Introduced for comparison ..................... . 191
Figure 13. Copernicia cerifera Martius. Flora Braziliensis, vol. 3, pt. 2, pl.
128 in part. 1882. Introduced for comparison .............. 191
Figure 14. Palmnocarpon mexicanum (Lesquereux) Lesquereux. Introduced
for comparison .............. ...... .................... 190
Figure 15. Cocos Datil Grisebach and Drude. Martius. Flora Brazilien-
sis, vol. 3, pt. 2, pl. 93 in part. 1882. Introduced for comparison. 190
Figure 16. Astrocaryum Jauorii Martius. Histoire Naturalis Palmarum,
vol. 12, pl. 65', fig. 16. 1820. Introduced for comparison... 189
Figure 17. Cocos eriospatha Martius. Cultivated specimen. Herb. New
York Bot. Gard. No. 19669. Introduced for comparison... 190
Figures 18, 19. Bactris Pseudocuesco n. sp ............ .......... . 186
Fig. 18. Station B.
Fig. 19. Collected by Don Narciso Rabell.
Figures 20, 21. Bactris Cuesco Crueger. Trinidad. N. L. Britton, E. G. Brit-
ton, and T. E. Hazen, 1920. No. 413. Introduced for
com prison .................................... . 187
Figure 22. Palmocarpon (?) globosum Lesquereux. Introduced for compari-
son .. . ............ ........................ . 187
Figure 23. Iriartiles Vaughanii Berry. Introduced for comparison. ..... 188
Figures 24a, 24b. Manicaria portoricensis n. sp. Station A ........ . .. 188
Figure 24c. Palmocarponopinabile n. sp. Station A ..................... 190
Figure 25. Manicaria sacchifera Gaertner. Jamaica. Henshaw, 1901.
Introduced for comparison ........................... 88
SCIENTIFIC SURVEY OF PORTO RICO AND THE VIRGIN ISLANDS
ANGIOSPERMAE (MONOCOTYLEDONAE).
VOL. VII, PL. 56
PLATE 57
Page
Figure 1. Palmaciles alius n. sp. Collected by Don Narciso Rabell....... 191
Figure 2. Palmacites conformis n. sp. Station A ....................... 192
Figure 3. F :... T.J .T.,p.r..!...' r 1.. Station B .......................... 194
Figure 4. Ficus Schimperii Lesquereux. Station B..................... 195
SCIENTIFIC SURVEY OF PORT RICO AND THE VIRGIN ISLANDS
K
I
ANGIOSPERMAE (MONOCOTYLEDONAE AND DICOTYLEDONAE).
VOL. VII, PL. 57
PLATE 58
Page
Palmacites sparsistriatus n. sp. Station A.............................. 192
SCIENTIFIC SURVEY OF PORTO RICO AND THE VIRGIN ISLANDS
ANGIOSPERMAE (MONOCOTYLEDONAE).
VoL, VII, PL. 58
|
RSS
HIDE
