r - so
* +tfF ‘4
.

pate hep ait imide ab
Ota Gee

a

a

ey
‘
~

oe

eZ

pa tr

sieht <a

é i> eae

, ot ae f
RAP

BULLETIN

' OF THE

ae Las wersity,
ee MUSEUM OF COMPARATIVE ZOOLOGY

AT. 3

a HARVARD COLLEGE, IN CAMBRIDGE.
es . pate SUNT 5S

(GEOLOGICAL SERIES, X.)

' \\ : P,
ANA NT gio /
. the At f SA?
: iy = Rae
, Ky in By
| “ PS
i, ASE
a
r
CAMBRIDGE, MASS., U.S. A.
1912-1916.
’ \

613358
ReSO Ap el tT: SS,
A

i.

CONTENTS.

- No. 1.— Geological Expedition to Brazil and Chile, 1908-1909. i

J.B. Woopworts. (37 plates). November, 1912

No. 2.— The Squantum Tillite. By Ropert W. Sayuzs. (12 plates).

January, 1914

No. 3.— Expedition to the Baltic Provinces of Russia and Scandinavia,
1914.
Part 1.— The Correlation of the Ordovician Strata of the Baltic Basin
with those of Eastern North America. By Percy E. Raymonp.
(8 plates). July, 1916.

No. 4.— Expedition to the Baltic Provinces of Russia and Scandinavia,

1914.

Part 2.— The Silurian and High Ordovician Strata of Esthonia, Russia,
and their Faunas

Part 3.— An Tistempsteraticny of tha etiam Bhction of Gotland’ ns
W. H. TwennHoreu. (5 plates). July, 1916 Vy Si Ar ee

139

177

287

341

ee ee a

Bulletin of the Museum of Comparative Zodlogy
AT HARVARD COLLEGE.

Vou EVIE Noe 15"

GEOLOGICAL SERIES, Vol. X. SHALER MEMORIAL SERIES, No. 1.

GEOLOGICAL EXPEDITION TO BRAZIL AND CHILE,
1908-1909.

By J. B. WoopwortTu.

Wits THIRTY-SEVEN PLATES.

CAMBRIDGE, MASS., U. S. A.:

PRINTED FOR THE MUSEUM.
NOVEMBER, 1912.

No. 1.— Geological Expedition to Brazil and Chile, 1908-1909.
By J. B. Woopworrtu.

TABLE OF CONTENTS.

PAGE

I. Prefatory note to the Shaler memorial series. } . . 3
II. Introduction . : ‘ ; ; 5
III. Itinerary . ; j : ; ; . Zz
IV. Outline of the eles ‘f sour Braail é j 41
VY. Permian glacial deposits of south Brazil ‘ ‘ ; ; 52
VI. The Triassic trap plateau ‘ : d ‘ ; , 91
VII. Geomorphology of south Brazil. : ‘ 99
VIII. Note on the changes of level of the coast of ee a Chile Sree cs
IX. Stone implements and pottery from Laguna. By R.B.Dixon. 132
X. Bibliography . i ; ; , > ‘ ; . 184

I. PREFATORY NOTE TO THE SHALER MEMORIAL
SERIES.

THE late Dean Shaler, at first Professor of Palaeontology and later
Professor of Geology at Harvard University, during his long career
as naturalist and teacher, both by his writings and his teaching,
displayed a very wide interest in the. various aspects of the earth’s
sciences. When in commemoration of his long services to the Uni-
versity a group of over seven hundred of its alumni raised an endow-
ment of more than $30,000. for the purpose of maintaining and
publishing the results of investigations suitable to his memory, the
Division of Geology, to which this gift was entrusted, found itself
provided with a fund for the advancement of knowledge in the entire
field of its scope.

The following are the terms governing the use of the Shaler Me-
morial Fund (see the letter of gift in the Annual Report of the Curator
of the Museum of Comparative Zodlogy for 1906-07, Cambridge,
1908, p. 19-20).

‘“‘The researches here contemplated are to be undertaken by persons nomi-
nated by the Committee of the Division of Geology and appointed by the
Corporation, whether officers or students of Harvard University or not.
The subject and the locality or field of research are to be approved by the

4 BULLETIN: MUSEUM OF COMPARATIVE ZOOLOGY.

Division Committee, preference being given to studies of an advanced and
original character. The sums of money allotted from the income for research

are to be determined by the: Division Committee with the approval of the ©

Corporation. The money appropriated for such work from the income of

the fund shall be in addition to the salary that would be otherwise paid to the

person or persons undertaking it; and any work or journey thus supported in

whole or in part shall be carried on under the name ‘‘Shaler Memorial Research”
r ‘‘Shaler Memorial Expedition.”

‘The publications here contemplated are to include the results of original
research carried on with the income of the fund, or independently of such aid;
but the results must in all cases receive the approval of the Division Committee
as to subject and presentation — though not necessarily as to the conclusions
stated — before they are accepted for publication.

‘All publications thus approved, whether appearing in independent vol-
umes or in some established journal, shall bear the general title, ‘Shaler
Memorial Series.”’ The allotment of money for publication shall be deter
mined in the same way as for research.

‘Beneficiaries under the fund, either as to research or publication, may be
invited by the Division Committee to give one or more public lectures in
Cambridge on the results of their studies, under the general title “Shaler
Memorial Lectures,’’ but no additional payment is to be made for these
lectures.

‘The income of the fund may be allowed to accumulate in case an investiga-
tion, expedition, or publication of considerable magnitude is contemplated
by the Division Committee, but it is not desired that such accumulation shall
continue beyond a reasonable period of time.”’

In geology, the action of védleanoes, the phenomena of the contact
of sea and land, and the evidences of past glaciation particularly
occupied Professor Shaler’s thoughts. This last subject was a direct
inheritance from his master Louis Agassiz. With James Croll,
Professor Shaler went further than Louis Agassiz did in perceiving
evidences of glacial periods in the geological record long anterior to
the great ice-age whose recognition was the lasting contribution of
Louis Agassiz to geological science. Professor Shaler anticipated
the discovery in the conglomeratic formations of the closing Palaeozoic

era of signs of glaciers, which only in recent years have been thoroughly -

scrutinized by others and found to be veritable products of glacial
action. With a view to contributing to the advancement of knowledge
in this field, the Division of Geology voted that a grant of money from
the Shaler Memorial Fund be expended by the author for the explora-
tion of the Permian conglomerates of the region south of Sao Paulo in
Brazil, the glacial origin of which had already been advanced by Dr.

WOODWORTH: GEOLOGICAL EXPEDITION TO BRAZIL AND CHILE. 5

Orville A. Derby. The report herewith submitted is the result of
that expedition. After the rainy season had begun in Brazil, I
devoted the time at my disposal to a brief examination of the changes
of level on the coast of southern Chile.

Other investigations will be undertaken from time to time as the
state of the fund may warrant expenditures.

II. INTRODUCTION.

In presenting the itinerary of portions of the region traversed, I
have taken the most convenient way of recording numerous observa-
tions not pertinent to the main object of the journey. Some of the
phenomena dealt with in this report have long been described in
other languages but without much discussion of causes or of geolog-
ical correlation. On this account I have been led into a free exercise
of the geologist’s privilege, if not his proper task,— to interpret his
observations and in the language of Robert Hooke “to raise a
chronology out of them.”’ The chapter on the Triassic trap plateau
presents the results of a rapid reconnaissance of a little known
geological field quite unfamiliar to North American students, and the
account of the topographic relief of south Brazil is a sketch en route
embodying observations in a more systematic order than as if left to
discrete and unrelated paragraphs in an account of scientific travel.

Through the courtesy of Dr. Orville A. Derby, the recently ap-
pointed Director of the Mineralogical and Geological Service of
Brazil it was arranged to conduct the Shaler Memorial party to the
glacial boulder-beds of Parané. To further facilitate the work of
the expedition Dr. Euzebio Paulo de Oliveira, Assistant geologist of
the Service, was detailed by the Director to act as “interpreter, guide,
and friend.” We were met by Dr. Oliveira on the confines of Parana
where I found him engaged in making a geological map of the state.
The generous conduct of this young geologist in placing freely at my
disposition the results of his observations upon the distribution of
the strata and in allowing me to examine his collections of rocks and
fossils makes me much indebted to him for many of the facts presented
in this report. With him and his pack-train I made the expedition
across the trap plateau from Rio Negro to Lages and later he made
with me the excursion up the valley of the Rio Tuberado in Santa
Catharina. Throughout these journeys the transportation was sup-
plied by the Brazilian Survey. Without this financial assistance the
work could not have been carried so far, and without the guidance

6 BULLETIN: MUSEUM OF COMPARATIVE ZOOLOGY.

afforded by Dr. Derby and his coadjutor many of the best localities
would not have been found in the short time that was at my disposal.

Dr. Joao Cardoso, the Director of the Geological and Geographical
Commission of Sao Paulo kindly detailed Dr. Pacheco, Geologist of
that survey to accompany me on the trip from Itaicy to Piracicaba
in that state and to Dr. Pacheco’s acquaintance with that region I
owe much.

In Chile, as a delegate of Harvard University to the Pan-American
Congress held in December, 1908, I was accorded free transportation
on the government railways through the courtesy of the Director of
railways, a privilege which I exercised in the journey from Concepcion
to Valdivia and return to Santiago and thence eventually to Val-
paraiso. Special rates were also given in the passage on the Chilean
steamer Limari from Valparaiso to Panama; both of these favors
reduced the expenses of the Shaler Memorial Expedition.

Prof. Robert DeC. Ward of Harvard University was appointed
a member of the Expedition to carry on studies in climatology and
to gather material for a course on the geography of South America.
He accompanied me as far as Ponta Grossa in Parana, whence he
journeyed to Paranagua, going by steamer to Santos, thence by rail
to Sdo Paulo, and so to Rio de Janeiro, from which port he took ship
for New York in August, 1908. The more important publications
resulting from his investigations are listed on p. 137.

Mr. Winthrop P. Haynes, an undergraduate student in the Uni-
versity, was appointed Assistant in geology and accompanied me at his
own expense as far as Ponta Grossa and Paranagua, whence he also
returned to the United States in August, 1908. He aided in the
collection of rocks and fossils in northern Parana.

I have retained the Portuguese spelling of the Brazilian names used
in the text. The pronunciation of these is similar to the Spanish but
the following peculiarities should be noted: — Ch is regularly and x
ordinarily equivalent to sh in English; g soft before e and 7 is like the
French j._ Words ending in am and Go have Portuguese nasal sounds
in which the nasal a is pronounced somewhat as ou in out with the
lips slightly closed at the end as if to give the letter m. Likewise
names ending in im are nasalized like ing in English but with the final
m sound slightly given.

WOODWORTH: GEOLOGICAL EXPEDITION TO BRAZIL AND CHILE. 7

Til. ITINERARY.

“The usual bane of such expeditions is hurry; because men seldom allot
themselves half the time they should do: but, fixing on a day for their return,
post from place to place, rather as if they were on a journey that required
dispatch, than as philosophers investigating the works of nature.”

Gitpert Wuire. The Natural History of Selborne. Letter

XXVI, December 8, 1769. London: 1789, p. 73.

The first Shaler Memorial Expedition following a generation after
the Thayer Expedition to Brazil sailed from New York for Rio de
Janeiro on the 20th of June, 1908, on the Steamship Voltaire of the
Lamport and Holt Line. As previously stated the party consisted
of Professor Ward, Mr. W. P. Haynes, and the author in charge. The
ship touched at Bahia on July 5th, and reached Rio de Janeiro on
the 8th of that month. |

On this lonesome tract straight out from the North American coast
at New York to Cape St. Roque the voyageur sights few vessels.
The minor changes of a June and enjoyable sea, the endless piles of
trade clouds, a solar annual eclipse — that of June 28th, — the first
view of the Southern Cross, the doldrums and their rains, the so-
called ‘green ray’’ of the setting sun,— these were the events of the
voyage of the Voltaire for those members of the party who made their
first crossing of the equatorial line.

At Rio de Janeiro we were met and taken care of by Dr. Orville
A. Derby, Director of the Mineralogical and Geological Service of
Brazil, and under his tutelage began preparations for the journey to
the planalto of south Brazil. Our stay in the Capitol was some-
what lengthened by the necessity of awaiting the discharge of Mr.
Haynes from the English Hospital, to which institution he had been
obliged to go for the treatment of an infected bruise received on
shipboard. At this time and indeed through my stay in Brazil, the
Capitol suffered greatly from an epidemic of small-pox. According
to reports given out on leaving the country as many as 6,722 deaths
were caused by this disease in Rio de Janeiro between January Ist and
November 22nd, 1908.

During this interval I visited Petropolis from which point under
the guidance of Dr. Miquel Arrojado Ribeiro Lisboa an excursion
was made to the valley of the Piabanha and the picturesque Valle do
Retiro (see Plate 1), a characteristic portion. of the eroded coastal

8 BULLETIN: MUSEUM OF COMPARATIVE ZOOLOGY.

opt sole

KToS

ALPARAI Gm fa!
“| * SA TIAGO

0 a. hee 30 40 30

Fria. 1.—— Route map of the Expedition from New York to South America
and return.

>,

WOODWORTH: GEOLOGICAL EXPEDITION TO BRAZIL AND CHILE. i)

border of the Brazilian tableland constituting the so-called Serra do
Mar.

Many maps of Brazil published in that country bear lines of longi-
tude reckoned from the position of the National Observatory in Rio
de Janeiro, which lies in 48° 10’ 21.15’’ West Long. from Greenwich.
The only high-grade maps, outside of certain municipal contoured
maps, have been published by the Comissao Geologica e Geographica
de Sao Paulo. Old-fashioned hachured maps of the topography exist
for some states but they are all inadequate for the purpose of geological
mapping. The territory of Brazil is vast and the interior so little
developed that it cannot be expected that the general government can
undertake, at present, the making of such maps of its domain as exist
for several of the European states, or even a map of the serviceable
character of the topographic map of the United States of America.
In the case of the state of Parana I was not able to procure in
published form, even an approximately accurate map though a very
useful manuscript map is in existence to which I had access. The
small-scale map of Santa Catharina published by the State is fairly
good for exploratory work. The best general map of Sao Paulo is
that of Williams.

Of the geological text-books which circulate mostly in Brazil,
owing to the higher education being based largely on the system of the
French, there are several editions of the elementary hand-book of the
late Professor de Lapparent. A Portuguese translation by Dr. B. F.
Ramiz Galvdo, of the third edition of the elementary text, entitled
Resumo de Geologia, with appendices relating to the geology of Brazil
by Dr. Derby, is much used in the schools where geology is taught.
Prof. John C. Branner published in 1906 an elementary geological
text written in Portuguese with special reference to Brazilian students
and embellished with illustrations from native sources including cuts
of South American fossils.

The National Exposition at Rio de Janeiro in 1908, held to com-
memorate the centenary of the opening of the ports of the country,
resulted in the bringing together of a collection of rocks and minerals
from many parts of this vast territory. Most of these exhibits were
intended to set forth the economic resources of the several states.
The best of these state collections was that from Sao Paulo formerly
under the directorship of Dr. Derby. From the new territory of Alto
Acre there was a small collection of rocks including a dark pebbly
sandstone locally used for whetstones, ferruginous sandstone, and a
fragment of an ironstone concretion, incorrectly labelled as an aerolite!,

10 BULLETIN: MUSEUM OF COMPARATIVE ZOOLOGY.

gypsum, and samples of clay. From Piauhy there were bottled
sulphurous mineral waters, talc, kaolin, graphite, concretionary
hematite, and fossil wood. From Sergipe, crystalline limestone and a
compact, light-colored argillite used for construction. From Ceara,
porphyritic granite, a red granite coarsely crystalline, fossil cetacean
bones from Cruxati, copper carbonate from Mildgres, concretions
with fossil fishes, and tile work. Alagoas sent pottery products,
particularly water-jars made from the clay of Penedo on the Rio
Francisco, noted for their porosity and consequent evaporating
capacity and cooling power, the most preferred carafes in Brazil.
The exhibit also included granites and marbles, among the latter a
dark and light-banded crystalline crumpled variety; garnets and
black tourmalines. From Rio Grande do Norte, there were soap-
stone, beryl, and aquamarines; Cretaceous and Tertiary limestones;
yellow bricks, salt from the evaporating pans of Caico and Macau, ~
and gypsum from Carambas. Rio Grande do Sul exhibited bituminous
coal from the Permian; agates from the Triassic trap sheets; wolfram-
ite from Rio Pardo; cuprite, indigolite from Bagé; molybdonite,
covellina, native copper from Colonia militar; tin ore; besides
artificial stone-ware and colored tiles. From the state of Parad, the
exhibits consisted chiefly of clay products, such as bricks and drain-
pipes from Cameté and Belem. From Matto Grosso there were gold
and diamonds from Coxipo mirim; diamonds and. sapphire gravel
from the Rio Coxim; gold and diamonds from a basal conglomerate
beneath a Devonian sandstone; besides ores of manganese, -hematite,
and exhibits of limestones. ‘The state of Amazonas was represented
by gneisses and schists, silicified wood, artificial stone-ware, tiles and
bricks manufactured from Tertiary clays. Goyaz furnished musco-
vite in merchantable plates, gold, galena, amethyst, rutile, diamonds,
yellow quartz (now exported), rose quartz, limonite, soapstones, and
millstones. Maranhao had an exhibit of gold. Bahia supplied a
collection of minerals, including manganese from Napareth, muscovite
from Conquista (said to exist in commercial quantities), monazite
sands from Prado, tabatinga (ochreous clays of a variety of colors),
copper carbonate from Bom Fim, manganese, limestone, a fine com-
pact brownish limestone with a tendency to a shelly fracture and often
horizontally banded from Campo Formoso; granites and gneisses,
including a red gneissoid granite from Jacaricy, tale blocks from
Caruanyba, lithographic (sic) limestone from Carinhanha and Bom
Jardim; clay products, including red porous water-jars. The state
of Santa Catharina furnished exhibits of the Permian coal, manganese

———

WOODWORTH: GEOLOGICAL EXPEDITION TO BRAZIL AND CHILE. I11

from Itajahy, molybdenite from Morro do Baht, magnetite, ochres,
slates, and the products of a newly established cement industry.
Minas Geraes (the state of “general mines”) exhibited of ordinary
minerals manganese, itabarite from the well-known peak, micaceous
hematite, limonite ores, and clay products from Bello Horizonte;
but the most striking exhibit was made by the Morro Velho Mine
which, in addition to a rich display of ore, put on exhibition a model
‘showing the working of the ore shutes, the shafts, and levels, with a
‘suitable explanation, in itself the most complete exhibit in the collec-
tion.

The newly established federal survey exhibited a relief map of
the vicinity of Rio de Janeiro and an excellent series of enlarged
photographs illustrative of types of Brazilian rocks and landscapes.

There is a great abundance of manganese in Brazil and from reliable
reports received then and since it is most probable that Brazil is
destined to be a producer of iron on a large scale and owing to the
belated resort to these deposits will enjoy prosperity from this re-
source when the workable ores of the United States have been ex-
hausted.

Our party left Rio de Janeiro on July 22nd, for Sado Paulo. The
points of geological interest on the line of the railway include Mt.
Tingua among the lower peaks of the Serra do Mar, which furnished
the dikes of phonolite consisting of an alkali feldspar, nepheline,
and aegirine to which rock Rosenbusch (Hunter and Rosenbusch,
1890) gave the name tinguaite. A good view of the black moun-
tain may be obtained from Ottoni Station. On passing the crest of
the Serra and attaining the valley of the Parahyba the Serra da
Mantiqueira comes into view, culminating in Mt. Itatiaia, a mass of
nepheline-augite syenite or phonolite, the loftiest point in Brazil,
2,994 meters (9,493 ft.), and the highest in South America outside of
the Andes. Its summit is a bare rock with weathered out joint-
structure forming needles on a large scale. It is plainly seen from
Rezende Station. H.H. Smith (1879, p. 441) states that snow occa-
‘sionally covers the summit.

The Tertiary deposits in the valley of the Parahyba are followed for
many leagues by the railway. In the dry season the train in traversing
these plains stirs up a fine reddish dust which penetrates the closest
ars. |

Near Pindamonhangaba Station I noticed two or three broad
shallow lakelets in the weathered surface of the rock, a type of basin
which abounds in the old weathered surface of Brazilian rocks.

12 BULLETIN: MUSEUM OF COMPARATIVE ZOOLOGY.

At Taubaté Station an oil shale is distilled from the Tertiary beds,
and Dr. Derby states that peat deposits occur near this place.

July 23rd.— The day was spent in Sao Paulo. At the office of the
Geological and Geographical Commission of the state there is to be
seen a unique collection of fossil silicified wood from the Permian
northwest of the city and some skeletons of Stereosternum, a reptile
occurring in the upper Permian. Among the sections of a diamond
drill recently obtained from a boring in the Permian strata I detected
a glaciated pebble with well-defined striae in a blue tillite bed, a
discovery which at once promised much for the objects of the expedi-
tion. An excursion was made to the vicinity of the American College
for the purpose of studying the terra roxa which in the state of Sao
Paulo plays so large a role in the cultivation of the coffee plant.

July 24th.— We left Sao Paulo at an early hour by the Sorocabana
railroad for the end of the line at Bury. At Sao Roque the line enters
a valley with outcrops of slates and limestones, an apparently infolded
member of the Pre-Devonian terrane of which the gneisses and schists
of the Serra do Mar region are the most ancient parts. The railway
follows along the contact of the limestone with granite as far as
Mayrink Station, beyond which town the slate belt is followed.
Beyond Pantoja Station a blue limestone crops out and kilns have
been built for making lime. Slates and limestones continue along the
route to Rodovahlo Station, where there is a cement factory. From
Pyragibu to Sorocaba the road leaves the belt of metamorphosed
sediments, and passes over granites and gneisses. The disintegration
of the porphyritic granite, as pointed out by Dr. Derby, produces the |
granitic surface sands known as swmardo, while weathering of a
deeper sort changes the granites and gneisses to a clay which from its
habit of balling under the shoe is called massa pé.

From Sorocaba onward to Bury the route lay over the Permian
area of sandstones, shales, and intercalated tillite beds. The line
passes within sight of the Ypanema stock some three miles in diam-
eter forming a slight elevation above the general level and mainly
composed of eruptive rocks in the Pre-Permian terrane. The igneous
mass is a nepheline syenite with segregations of magnetite famous for
its early use as an iron ore. Ferreira says the ore was discovered
in 1578. At Ypanema there are quarries in a grey sandstone of the
Permian, and old iron furnaces, perhaps the site of the earliest pro-
duction of iron in America.

Throughout the day the mainly open unwooded campos of the
upland, or planalto as the Brazilians term the slightly accidented

——————————— _ 9

WOODWORTH: GEOLOGICAL EXPEDITION TO BRAZIL AND CHILE. 13

surface of the plateau, was the striking feature in the landscape.
Long round swells or hills with graceful curves to the stream ways
betokened everywhere the long continued action of erosion. The
excellent topographic maps of this portion of the state by Mr. Horatio
Williams begun by the Geological and Geographical Commission of
Sao Paulo under the directorship of Dr. Derby made it possible to
follow the route intelligently and interpret the general outlines of the
erosive history of the district. In a following chapter on the geomor-
phology of the region, I have made free use of these maps.

July 25th— In 1908, the passenger service of the railway termi-
nated at Bury, but a line in process of construction extended the
communication by rail to Ponta Grossa in Parana, and so into connec-
tion with points on the south far towards the boundary of Uruguay.
As will be shown in the sequel the fresh cuttings of the rocks and
surface deposits along this new line afforded an exceptional opportunity
to study the geology not only of the underlying Permian tillite beds
but also of the superficial gravels and their cover of red earths vari-
ously known as terra roxa, ete. This day we proceeded on a flat car
some 18 kilometers along the line of construction, completing the
journey to I‘axina by a sort of carriage known as a “trolley.”” While
traversing the high campos a hailstorm came up from the west with a
well-defined horizontal vortical ring of black clouds, from which hail
associated with rain fell so as to coat the ground with hailstones.
For nearly an hour after the passage of the storm the small streams
now in flood carried a thick load of hailstones. This fall of hail
within the subtropical region at an elevation of 900 meters above the
sea suggested an inquiry as to the occurrence of hail at lower eleva-
tions nearer the equator as a possible factor in the Permian glacia-
tion since it appears that in this way ice may be precipitated in
regions where snow never falls.

July 26th.— Being unable to secure a conveyance to Itararé, the
day was spent in Faxina. This town is underlain by the Devonian
sandstone, a light colored to whitish rock, with cross-bedded layers,
and occasional bands of white quartz pebbles. Nodules of clay
occur also embedded in the sandstone. The strata are well exposed
in a ravine on the outskirts of the town.

July 27th.— Having procured the means for the conveyance of our
party to Sao Pedro de Itararé on the boundary of Parana, we journeyed
to that place. For most of the day the route lay over the Permian
basal sandstones which appeared in the stream bottoms, sometimes
with thin beds of white quartz pebbles but without trace of compound

14 BULLETIN: MUSEUM OF COMPARATIVE ZOOLOGY.

conglomerates such as constitute the tillite beds. ‘Some of the coarse
gritty beds contained clay particles. In all probability these were
once grains of feldspar indicating that the unaltered rock was a
granitic sandstone related to arkose. Dikes or sills of basic intrusives.
occasionally intersect the sandstone along this route.

July 28th— Through the courtesy of Dr. Cruz Lima we travelled
in a special car from Itararé to Jaguariahyva, stopping in the newly
opened railway cuts to examine the tillite beds. At the first stop in
a cut above the Rio Jaguaricatu in Parané a well-striated pebble was
found in a boulder-bed, establishing at once the identity of origin
of these deposits with those of India and South Africa.

July 29th— Continued the journey by rail to Ponta Grossa, the
headquarters of the party engaged in the geological survey of Parana.

The railway from Jaguariahyva to Ponta Grossa crosses the
Devonian sandstone cuesta affording a magnificent view of the
country.

Between Pirahy and Coxambii the Pre-Devonian rocks, exposed
in a lowland widened out along the course of the Rio Yapo, comprise
a tilted group of rocks of which I have seen no account in the descrip-
tions of the metamorphosed district of the Serra do Mar. At Pirahy
Station there is a monoclinal set of beds in a ridge west of the railroad.
The beds strike north by east and dip about 30° west. About a
mile south of Pirahy at a water-tank a felsitic breccia crops out.
Farther south the train passes through a cut in which slightly meta-
morphosed shales, sandstones, and a pebble bed with fragments of
red felsite, granite, etc., appear, having a reddish color and dipping
westward at an angle of about 30 degrees. These rocks from their
relatively unmetamorphosed condition appear to be younger than the
belt of slates and limestones described as occurring in Sao Paulo.
No fossils were seen in the section nor did time permit a satisfactory
search for further details of the stratigraphy. This formation, so
different from the members of the Pre-Devonian terrane on the east,
is the most western member of the highly inclined rocks seen in Parana
and suggests that some horizon between the Middle Devonian and the
slate and limestone terrane may yet be worked out and correlated
in this field.

Several days were spent at Ponta Grossa in the examination of the
surrounding country, including a visit to Conchas where a bed of |
tillite is to be seen. In the Carboniferous sandstones west of Ponta
Grossa I found some worm burrows of the type known as Monocra-
terion; these show a cup-shaped superior termination and a vertical

——————— es

WOODWORTH: GEOLOGICAL EXPEDITION TO BRAZIL AND CHILE. 15

tube penetrating to the bottom of the layer, some five inches in thick-
ness. The tubes were filled with a fine greenish shale. (Fig.42.)
Similar burrows occur in Pennsylvania in the Ordovician (J. P. Lesley,
’89, 1, p. 417-418) but seemingly are of no diagnostic value in determin-
ing the age of the strata in which they occur. It is possible that the cup
at the orifice of the tube in Monocraterion
is due to the caving in of the sands prior
to their covering by the superincumbent
layer, and that thus it is not to be taken
as showing the form of the anterior portion
of the animal which made the burrow.

From Ponta Grossa an excursion was
made under the guidance of Dr. Derby to
Curityba and thence to the coast at Pa-
ranagua. An excellent view of the profile
of the plateau and the Serra do Mar was py¢. 2— Monocraterion sp.
obtained and I have utilized the data in —A worm burrow occurring in
what follows on the geomorphology of tT meee
south Brazil. From Ponta Grossa also I near Ponta Grossa, Parand.
set out for a trip via Rio Negro, over the
trap plateau to Lages in Santa Catharina. As the itinerary of this
expedition includes some observations upon the general character of
the country not embodied in North American geographical writings, I
have transcribed this portion of my Journal with but slight con-
densation.

A Journey from Rio Negro to Lages and return to Porto da Unido
on the Iguasst. It having been decided to make a reconnaissance of
the section from Serrinha on the upper waters of the Iguassti to Rio
Negro and thence southwards to the base of the Triassic escarpment,
the traverse was continued southwards to Lages. The Triassic
formation was examined for any evidence which might have a bearing
on the transition period following that of Permian glaciation. Dr.
Euzebio Paulo de Oliveiro and myself set out from Ponta Grossa on
August 13 by rail for Rio Negro, the end of the railway, making stops
at ‘Tamandua, Serrinha, and Lapa. A synopsis of my notes on the
geology of this portion of the route is embodied in ensuing chapters
on the Permian deposits. See Plate 19 for map of route.

At Rio Negro we heard of the Bugres, wild aborigines who infest
the trap escarpment on the south and often ambush lonely travellers
on the pass over the Serra do Espigao. Southwest of this high point
along the Serra Geral there is a remnant, so I was informed, of the

16 BULLETIN: MUSEUM OF COMPARATIVE ZOOLOGY.

Botocudos, whose haunts are carefully avoided by Brazilian travellers.
These naked savages sometimes commit outrages on the new Euro-

;

7}

Fic. 3.— Sketch map of eastern Parana showing localities visited by J. B.
Woodworth.

pean settlers who have colonized the lands back from the old occupied
sites of the coast. It seems strange to the visitor from North America
to find here within a hundred miles of the Atlantic coast conditions

—_ ————

=

WOODWORTH: GEOLOGICAL EXPEDITION TO BRAZIL AND CHILE. 17

of European settlement such as characterised the period of the first
third of the 18th century in the New England states and New York.
The topography of this portion of Brazil, its inaccessibility, and the
inutility to the older inhabitants of European origin of a large stretch
of country along the ragged escarpment of the trap plateau accounts
in large part for this lingering of a hostile primitive people in close
proximity to the coast; encouraged and abetted by a more or less
easy communication for the untrammelled native by larger bodies of
indigenous folk in the hinterland of Parana, where, among the Coroa-
das savage life remains in yet greater exemption from the restraining
influences of advancing civilization. The extension of the railway
from Porta da Unido southward across the trap plateau divides these
people and promises to bring to an end a frontier struggle which has
endured too long.

Our pack-train having arrived at Rio Negro, we prepared for the
expedition to Lages. The equipment for geological field work in
this region is extremely simple. For the four members of our force,
one light tent was provided, poles for which were cut from night to
night in the forest surrounding our camp sites. The baggage was
carried in wicker baskets or panniers lashed on pack-saddles. Our
provisions consisted of prepared black beans, boiled rice, farinha
meal, broad thin slabs of drief beef known as xarque, and a supply of
powdered burnt coffee with which the proper quantity of sugar had
been mixed. A few small pots and cooking basins completed the
outfit. The business of camp life was equally simple. In the morning
before mounting, black coffee and bread were served. Breakfast

~ was eaten between 11 and 12 after a ride of a few hours, and a second

substantial meal was prepared at night after establishing camp.
Except for small supplies of corn (maize) obtained at long intervals
at some farm house, the mules subsisted on the leaves of the bamboo
and other tender foliage which they found along the route. Water
and wood for the camp fire were everywhere in abundance.

August 17th— There was a heavy white frost on the ground at
Rio Negro at an early hour this morning. We set out promptly for
the south, keeping track of the distance traversed by means of pedo-
meters. A mule pace in Brazil is reckoned at 0.72 meters (2.36 ft.).
The mule trail from Rio Negro to Lages traversing some thirty-six
rivers and streams was cut through the forest in the middle of the
18th century as a military necessity (V. da Rosa, 1905, p. 265-266).
As far as the Rio Laurengo, the route lies over shales and yellow
pebbly sandstones carrying an occasional erratic block. Beyond

18 BULLETIN: MUSEUM OF COMPARATIVE ZOOLOGY.

this point the road enters a heavily wooded district of moderate
dissection with flat-topped interstream areas. We camped for the
night near the road 2.3 miles south of a small settlement called Sepul-
tura.

August 18th.— It was so cold at midnight that one of the camara-
das made coffee with which we were served. At this hour a pan con-
taining water was coated with a thin sheet of ice. From this camp
to the Rio Contagem the route traverses a soft sandstone formation.
Just before halting for breakfast at 10 a. m. we passed a shaded mud-
puddle covered with a thin film of ice. These occurrences of frost and
ice I note particularly because they were so contrary to my precon-
ceived notions, based upon the imperfect accounts in North American .
geographies of the winter climate of even this elevated region in south
Brazil. Apparently horizontal beds of deeply weathered sandstone
continued to form the surface rock as far as camp No. 2 on the head-
waters of a small stream, the Rio Sao Joao, a north flowing branch
of the Rio Negro. The gentle descent from the plain to camp lay
through a forest of Araucaria, tree-ferns, and a graceful bamboo with
long curling tips.

August 19th— The night was cold again, with frost on the ground
and ice in the camp dishes at an early hour this morning. Shortly
after leaving camp we traversed a plain with large areas of dead brakes,
scattered palms, and tree-ferns. Araucaria is however the most
abundant forest tree; there are large tracts of it, and many young trees
are in evidence. For over a kilometer the path led through a dense
growth of tree-ferns, taquara or bamboo, Araucaria, and large imbuias
(Canella imbuia of Brazilian writers) with numerous mud-holes
crossed-ridged with the pildes made by mules, through which our
train of animals wallowed with extreme slowness. This jungle
suddenly gave way to a wagon road on good ground, the sign of
approach to some German or Polish settlement, of which however we
saw nothing. Along this route we came upon a fine example of the
tall Brazilian sassafras tree (Nectandra cymbarum), standing alone
in the forest. Descending through swampy places of the wet-wood
type we entered another forest of Araucaria. Tree-ferns partially
hidden on the edge of these forests or standing in the more or less open
growth of tall trees were in sight most of the day, but no rock ex-
posures were seen and no pebbles bestrewed the trail which seems to
have lain over decomposed shales. Here and there in the valley
bottoms two or three huts were encountered, around which were pigs,
a few sheep, cattle, and horses. We camped beside a stream on the
soft spongy ground of tree-fern growth in the forest.

—_—

WOODWORTH: GEOLOGICAL EXPEDITION TO BRAZIL AND CHILE. 19

August 20th— The night was cloudy with a little rain and con-
sequently warmer than on the two previous nights. We found that
we had gotten on to the road leading over the Serra do Espigao, the
very route we meant to avoid because of the Bugres, but quickly
resolved to keep on. After proceeding at a painfully slow pace with
many uncertain turns through the forest we halted at a distance of
not more than 3.2 kilometers for breakfast in a piece of open campo
surrounded by the now dripping forest, our way having led over vales
and ridges about 100 feet high and through tree-fern swamps. More
native huts appeared at about 5 kilometers from last night’s camp.
At about 6 kilometers from the camp we encountered a chert bed on
a hill south of a hamlet. Under this bed, at a river crossing immedi-
ately north, there outcrop green shales. We got into camp on the
bank of a small stream the valley of which is excavated in a yellowish
green shale. For an hour before reaching this camp, the flat sky-line
of the Triassic escarpment could be seen ahead on the south. Some
huts near camp were built of hewn boards and hand-made shingles,
with the usual open windows. This place is Chaxim. There was a
broken down cross and the remnants of a fence along side our camp,
enclosing the burial place of some one killed by the Bugres. All
travellers on this road we observed went armed, an example which we
followed.

August 21st.— We got off early from camp as the manuscript map
in our possession indicated that we were now near the base of the trap
escarpment and should make the pass over the Serra do Espigao
before noon. ‘The road led over some low hills of greenish shale near
the beginning of the ascent. Shortly before the climb began we came
to a few houses south of Chaxim, at one of which, a store kept by a
German, we found a bugreiro, or sort of special police, armed with a
cavalry sword and a double-barreled horse-pistol, whose evident
business it was to accompany parties over the pass. The ascent of
this pass, only some 1,200 feet above our base at the store, was so steep
as to oblige us to dismount, and was made in such a rush with all
revolvers drawn, that geological observations were, despite the fre-
quent exposure of ledges, neither advisable nor clearly made.

As is usual with South American mule paths the way led up the
steep spur in preference to following one of the adjacent creases of the
slope, since by so doing the minimum of mud and water is encountered
at all times of the year. The escarpment below the crowning trap
sheet is mainly sandstone of a reddish tinge which succeeds the green
shales at the base. The chert beds before mentioned apparently

20 BULLETIN: MUSEUM OF COMPARATIVE ZOOLOGY.

correspond with the Estrado nova shales of Dr. I. C. White, while the
beds of sandstone in the face of the escarpment reddish below and
light colored above correspond with his Rio do Rasto and Sao Bento
beds respectively.

The trap crowned pass of the Serra do Espigao is separated from the
main mass of trap on the south by ravines parallel with the escarp-
ment, somewhat as in the annexed sketch of the profile. (Fig. 24,
p. 93). The summit where crossed by the trail gave an aneroid
reading of 3,950 feet. A new cross by the roadside at the southern
side of the first ravine marked the spot where but a few weeks previous
a Brazilian had been killed by the Bugres. Of these savages, however,
we saw none nor were we molested. Our bugreiro left us at a point
near the cross and returned to his post at the base of the escarpment.
On passing the summit I noticed fresh ice crystal marks in a dried up
mud-puddle. We descended at once the south slope of the ridge,
passing the lower contact of the trap on the sandstones to an open
campo watered by the Lageado liso, a small stream so named, in
common with many in Brazil, from the flaggy beds in its channel and
banks. The tilted attitude of a band of beds in the valley of this
stream suggested faulting parallel with the escarpment, but I was not
able to determine by the elevation of the base of the trap on the
opposite side of the valley the occurrence of a displacement of this
plane of reference, though it was my impression that the trap in the
Serra do Espigao lay higher. From the Lageado liso the ascent is
gradual but steadily upward to the top of a broad tableland of trap
giving aneroid readings of 4,050 to 4,100 feet elevation some 6 kilo-
meters south of the Lageado liso. At 8.6 kilometers some farm
houses appeared about which were fields enclosed with stone fences.
We descended into the valley of Passa dois on the south and went into
camp. Numerous poles set upright in the ground showed that here
the pack-trains halt for the night.

August 22nd.— There was a heavy frost last night. Between one
and two kilometers south of camp red sandstones crop out with a
northeast dip. Reaching the Rio Correntes at the noon halt, the
route continued on trap to the camp for night on a small stream, the
Rio das Pedras, near a farm house where were pigs and cattle in fields
enclosed with rail fences.

The surface of the basalt traversed in this day’s journey varied
much in the degree of decomposition. For long distances there was
good hard rock with a thin brownish crust of weathered products.
Between these stretches along the road the rock was weathered down

or At

WOODWORTH: GEOLOGICAL EXPEDITION TO BRAZIL AND CHILE. 21

to a deep brown earthy mass, and in small patches I saw reddish, and
in one place, greenish clay. Bales of spherical separation appeared
here and there and often the road led the mules between large angular
blocks which bestrewed the hard trappean surface. The contrast
between this thin coat of weathered trap and the deep beds of de-
composition forming the terra rora in Sao Paulo is very striking.

The Brazilian pine, Araucaria, occurred in small patches here and
there and many young plants pointed to favorable conditions of growth
for the species. An occasional Maté tree appeared along the trail,
evidently due to the droppings from some passing caravan. [rom
this high plateau there was a good view of a high trap range of tabu-
lar outline in the distant northwest, the prolongation of the Serra do
Espigao extending towards Porto da Uniaéo. The wooded surface
of the range concealed all the rocks, but the triple terraces of the mass
presumably signify a three-fold division of the trap sheets of which it is
composed. Between our position at Corisco and this tabular moun-
tain there lay an extended lower surface, the deeply dissected basin
of the Rio Correntes.

Along the mule path, we passed several small grassy pools, occupy-
ing depressions partly enclosed in the trap. In at least one instance a
pool lay on the upper side of the road and the water was held in by a
barrier of mud and gravel accumulated in the road by wash from the
descending grade on either side to the sag by which the drainage
normally overflowed. The surrounding gramineous plants displayed
the brown color of winter. The fine green grass of one of the pools
had attracted to it a domestic horse which stood up to his knees
feeding, evidence that the bottom was floored with probably the same
stiff residual clay which later I saw in an excavation in one of these
basins. On this monotonous succession of trap uplands of nearly
uniform structure the trivial relations of the life which found a place
upon them became matters of more than passing interest. On the
muddy bottom of the rivulet which flowed past the evening camp I
found a small fresh-water mussel resembling Unio crawling along
with the open edges of the valve down so as to leave a deep narrow
groove in the mud. ‘The trail was sinuous and ended in a burrow
where the molluse pushed under a cover of mud. I should have
assumed, had I not seen this animal at work, that the trail, as have
been so many found in the fossil state, was to be ascribed to some
gasteropod.

August 23rd.— The morning broke cloudy with rain threatening,
for in this latitude on the trap plateau the distinction between a dry

pid BULLETIN: MUSEUM OF COMPARATIVE ZOOLOGY.

winter season and a rainy summer is not so clearly recognizable as
it is farther north in Sao Paulo. On proceeding some three kilo-
meters through a forest of Araucaria and tree-ferns we heard a loud
reverberating roar in the tall woods, a noise which my Brazilian com-
panions announced to be that of an onza — the jaguar. With drawn
revolvers they started at once into the woods in the direction of the
unusual caterwauling sound, which certainly seemed to come from a
large and powerful cat. From my post with the pack-train I heard
their shots and presently the crash of a body falling through the trees.
They had shot and killed a large howling monkey or alouatte (My-
cetes), a red-furred specimen measuring about four feet from the nose
to the outstretched end of the long prehensile tail. Our experience
with this well-known monkey recalls the Italian proverb: —

La sera lione,
La mattine babbione.

Travellers in Brazil speak of the nocturnal howls of these monkeys
but we heard their cries but twice and then only in the early forenoon.

We were soon led by a well-beaten track to make a detour from our
proper route through the forest. This led us to an isolated cattle
and mule ranch, whence we were directed to the main road to Coryti-
banos.

Our path lay through a more or less open growth of Araucaria
broken in a valley by numerous stumpy palms, known as Butia.
Cattle grazed on several open valley floors. On the open interstream
areas we passed many of the small pools or lakelets and in one of them
with the brilliant green grass before mentioned lay a dead horse
whose struggles were evident from the disposition of the vegetation
near his feet. I note this as an instance of how large herbivorous
mammals may be tempted into swamps and under favorable geological
conditions become fossilized. ‘The animal in this instance had not
become mired but probably had fallen ill from the unwonted diet.
About an hour after noon we regained the thoroughfare near a cluster
of houses with a monjolo or farinha mill, where we halted and cooked
breakfast, on the bank of the Rio das Pedras. The basalt traversed
this morning displayed many cavities lined with zeolites, fragments of
these minerals glistening in the mule path every few yards. The same
dull dark brown hue of the thin trappean soil appeared as farther
north but I noted in one section where we descended to a stream chan-
nel a reddish rusty zone of decayed rock underlying the superficial
brown coat of rock which was here about one meter thick.

WOODWORTH: GEOLOGICAL EXPEDITION TO BRAZIL AND CHILE. 23

The day came off clear and fairly warm, giving from one or two of
the trap elevations distant views to the north and south of level sky-
lines broken only by an occasional remnant of a still higher basalt
sheet in the series over which we were travelling. The surface past
over this morning was one of rather immature relief with no deep,
steep-sided gullies or ravines. Numerous small streams showed short
falls over trap ledges and rapids of no great length, manifestations of
streams actively at work and far from being well graded. While we
were halted on the Rio das Pedras a troup of forty mules came along
bound southward. After searching in the nests of quartz which
here beset the decomposed trap for other minerals, we set out and
went into camp for the night on a small rivulet (Rio Ponte alto) at the
base of a steep trap slope surrounded by the araucarian forest.

August 24th— A puma came into camp in the night at about 1
o'clock and drove our dog into the tent. Alfredo fired twice at the
glowing eyes of the animal but missed him. This is the sole instance
in which on this expedition we were disturbed by any large nocturnal
eat. During the day we saw nothing of the mammalian fauna of the
forest. The tapir must be abundant in the deep recesses of the
woods along the streams. We saw hanging on the wall of the store
the skin of one which had been shot at the foot of the Serra do Espi-
gao. After a ride of an hour and a half from camp we forded the river
Marombas whose valley floor with a floodplain of some width lies
fully 300 feet below the surrounding trap plateau. There is a small
settlement of houses here in the garden of one of which I noted a palin
tree and a large prickly pear cactus (Opuntia) about ten feet high.
The road from this point onward crosses a succession of trap ridges
and valleys, with a relief varying from 200 to 300 feet, as far as the
broad elevation of cleared ground on which stands the pink and white
village of Corytibanos. Here we halted for breakfast by a spring on
the outskirts of the place and having rested on the warm dry grass and
procured an additional supply of provisions including some bread and
butter, proceeded southward along a mule path with bridges over
small streams to a camp for the night. On the way we encountered
in the south bank of a stream valley about 100 feet in thickness of red
beds in a clayey state overlain by trap. Quartz in radiating nodules
abounded in the trap traversed today, but little of geological interest
could be noted in the monotonous ride over the basalt surface other
than the minor variations of the topographic relief. Araucaria
remained the dominant forest tree, while tree-ferns and bamboo could
be seen near or along the water courses. Beside our camp, in the low

24 BULLETIN: MUSEUM OF COMPARATIVE ZOOLOGY.

ground a few kilometers south of Corytibanos, frogs abounded in
small grassy swamps or lakelets. One frog had a weak peep, another
a rattling croak, and one a cry like that of ababy. The great number
of these small lakelets on the trap plateau with standing sweet water
even in the dry season of the year is evidence that the rock crevices are
well supplied with water. The occasional rains which we encountered
and the impervious nature of the deeper rock together with the
residual clays which form the bottoms of depressions unite to keep
much water in sight at the surface. Yet there is a great variation
in the amount of permanent moisture present in the soil in the several
habitats of plants, the quantity increasing from the hills towards the
narrow valley floors as is exemplified in the distribution of the tree-
ferns and the bamboos. We saw grass or forest fires yesterday and
today in distant broad valleys.

Another puma was reported in sight by the men just as we retired.
The first snake which I have so far seen in Brazil, a small bright
graceful green snake, was encountered on one of the little bridges
south of Corytibanos. Araucaria continues to be the dominant
forest tree. With it and rivalling it in size is the scraggly imbuia
whose bole attains a diameter of 3 feet or about a meter. What
impressed me most concerning the trees of south Brazil was the
small ovate leaves with entire margins which so many of them pos-
sess. The broad leaved oaks, maples, tulip-trees and other forms
familiar to the North American as existing species with precursors
occurring fossil as far back as the Cretaceous are here wanting. So
far as leaf evolution goes these simple outlines recall the forms which
are so characteristic of primitive types in all organisms.

On the north bank of the small stream on which we camped this.
night there was exposed a bed of red shale traversed by small
vertical faults with downthrow in each case on the west. This
stream, the Lageado penteado, is a branch of the Rio Canoas. The
name Lageado applied to streams like that of Lages given to the town
to which we were bound has reference to the slabs of sandstones.
which abound in this region and “pave” as it were the beds of the
small streams.

The Canoas river has shifted its course in the degradation of the
region down the dip of the formations so as to hug the southern edge
of the basalt.

I saw no fossils in the red shales but found some concentric con-
choidal fractures or joints.

August 26th.— Light showers during the night. Ant-hills and

WOODWORTH: GEOLOGICAL EXPEDITION TO BRAZIL AND CHILE. 25

red soils are not so common on this Triassic area as on the north in
Paran4 and Sao Paulo. The basalt has disappeared from a large tract
about Lages without leaving any noticeable trace.

At 9:25 a.M. we came to the Rio Canoas which is here a broad deep
stream over which our pack-train was ferried on a platform supported
by four dug-out canoes and held to its course by a wire cable. On the
upland we passed the hamlet of Canoas with oxen ploughing. Shortly
before noon we descended a steep slope to a stream crossing with red
shales dipping north in the bank. Farther on the mule path trav-
erses a dike about seventy-five feet wide cutting the sandstones.
This dike is heavily charged with fragments of several rocks and min-
erals evidently brought up from below. About a mile farther south
a narrow dike about one foot wide occurs along the trail near a small
stream. About a mile farther south there is a short low ridge on the
east of the road with small conical spurs and buttresses of inclined
beds on its north side. On approaching this point on the road with a
slight rain falling thousands of winged ants flew over the campo up to
fifteen feet in the air and for some reason chose to collect behind my
head in great numbers. On catching up with the party I found the
tent pitched in a clump of bushes at the western end of the ridge just
as a steady rain set in for the night.

The small lakelets so characteristic of the trap surface also occur
in the sedimentaries south of the escarpment. Here the depressions
appear to mark the site of springs. From the dike southward the dip
is southerly and probably S.W. There is a ee anticlinal fold near
this camp with axis NNW-SSE.

August 27th.— It rained nearly all night and until 7 a.m. We
rode southward across the bleak campo to a descent over sandstone
beds which brought us, after a journey of fourteen days from Ponta
Grossa, into a broad irregular valley in which Lages lies.

August 28th. The light brown Triassic sandstone under a red
shale bed north of Lages is quarried for building stone and flagstones.
No fossils other than wandering trails were to be seen in the sandstone.
Certain greenish shale beds pass laterally and obliquely to the stratifi-
cation into red beds, and well-defined green-walled joints in the red
beds show that the green color is locally a post-depositional alteration
of the deposits. In the afternoon we returned northward, to the
locality of the dike bearing inclusions, for a more detailed study of
that rock.

Ant-hills about fifteen inches high bestrewed the surface of the
Triassic sedimentary tract. As usual most of the hills had been

26 BULLETIN: MUSEUM OF COMPARATIVE ZOOLOGY.

broken into by some burrowing animal and many colonies were aban-
doned. This cycle of change must several times have worked over the
surface materials of the campos of Brazil and facilitated the work of
erosion by winds and rain.

When it is recalled that a large insectivorous fauna including the
anteaters, Myrmecophagidae (Vermilinguia), inhabit the ant-occupied
surface of Brazil and that this group is specialized with reference to
the ants and that traces of this organic adaptation go back to Pliocene
times, it is evident that ants and anteaters have exercised an important
function in the geologic processes which have worked together in the
evolution of the surface deposits of Brazil. Fossil ants are numerous
in the Oligocene beds of Florissant in Colorado (Scudder) and ant-
eaters appear in the Miocene Santa Cruz formation of South America,
dates which are as far back if not earlier than the beginnings of the
present surface deposits of the Brazilian highlands.

August 29th.— The road northward from the camp at the dike
passed over alternating beds of red shale and yellowish to reddish
sandstones, dipping gently to the northwest towards the southern
margin of the trap sheets, whose escarpment, as far as it could be seen,
extended in a northeast-southwest direction roughly parallel to the
strike of the underlying sediments. From the top of one of the
monoclinal ridges between the Rio Ponte alto and the Rio Canoas, this
escarpment could be traced to the eastern horizon where a conical
outlier stood out in clear relief as a monument to the erosion of the
trap cover which formerly extended over the Lages area.

Fic. 4.— Field sketch of the trap plateau northwest from Rio Marombas in
Santa Catharina.

The monoclinal structure of the sediments results in a series of ridges
with steeper faces on the southeast and longer gentle slopes to the
northwest. At about twelve kilometers from camp we crossed the
Canoas River, and continued northward to the north bank of the Rio
Ponte alto where we camped by a covered bridge. The stream at this
point flows in a channel about twenty feet deep over sandstones with
an active drainage. .

August 80th.— After a march of an hour and fifty minutes from
camp we passed over sandstones striking northwest and dipping about
15 degrees to the northeast. About forty-five minutes later we

WOODWORTH: GEOLOGICAL EXPEDITION TO BRAZIL AND CHILE. 27

surmounted the dissected trap plateau at an elevation of some 700

feet above the valley of the Rio Ponte alto. The broad valley of the
Rio Cachoeiras which next succeeds exposes a bed of sandstone in its
banks below which lies a trap sheet over which in turn it flows. About
ten miles north of the river we camped for the night near a frog pond
between two of the mule bridges which mark the approach to Coryti-
banos.

August 31st.— An hour’s ride from camp brought our cavalcade
again to Corytibanos, at which point we took the road northwest
towards Sao Joao en route to Porta da Uniao on the Iguasst. For
several miles along this route we traversed campos with a scattered
growth of the the araucarian pine and then of the Butia palm. One
deeply weathered trappean hill bore numerous lakelets or lagoas
bordered by tall tufts of grass. Blackened heads of trap rock
cropped out over the surface of the inosculating ridges which separate
these solution-basins. At noon we halted beside a small stream whose
milky waters were without apparent cause since it had not rained.
On resuming our march we shortly heard the roar of falls in the course
of the Rio Marombas on the left in the forest where the river tumbles
over the bedded traps. The river was crossed on a balsa or ferry, a
sort of raft supported by four or five wooden canoes. (Plate 17).
A forest fire was burning on the east bank of the river. After travers-
ing two deep valleys and a broad hill of dry campo we camped on
another small milky stream in a cattle country. The milkiness of
these small streams was possibly due to cattle wading in them. The
day was warm, and beetles, butterflies, and dragon-flies were out
fluttering over the muddy flats of a stream at noon.

September 1st.— About an hour and a half after leaving camp this
morning, we came to the Rio Correntes which was forded on a trap
bottom. After proceeding for an equal length of time we crossed
another broad shallow stream with rapids, the Rio dos Patos, also on
trap. We followed up the right bank of this river in a northwest
direction until nearly noon when we came to a coboclo hut in the
forest, where a woman gave us directions. Our halt for breakfast at
the noon hour was beside another small sluggish milky stream.
Several of the solution-basin lakelets were passed during the fore-
noon. As the trail was not distinctly marked from the paths leading
from one hut to another in this forested region, much of the after-
noon was spent in seeking directions. A large forest fire at one time
was raging to the west of our position. Fires in the standing clumps
of dry dead bamboo burn with great rapidity. At length we entered

28 BULLETIN: MUSEUM OF COMPARATIVE ZOOLOGY.

a burnt over tract of Butia palms, where beside one of the milky
streams we encamped for the night. My mule had given out during
the afternoon and was towed into camp tied to the tail of a sound
animal. In the forest along the trail I observed a small wild tobacco
plant from four to five inches high, which our head camarada termed
“fumo dos Bugres.’”’ A pair of papagaios flew over the camp at sun-
set and a few mosquitoes buzzed about our fire of araucarian knots.

The Brazilian pine (Araucaria brasiliana), when young, resembles
in its branching habit the other members of the Coniferae; but
it scarcely attains full height before its branches become crowded
towards the top by the dropping off of the low ones. An old tree thus
presents a clean bole with usually one but sometimes two great whirls
of branches at or near the top and recurved upwards. Upon the fall
of a dead tree frequently as much as ten feet of the upper extremity of
the prostrate trunk is a mass of highly resinous knots which remain
undecayed for years after the surrounding wood has disappeared.
In the wettest weather a fire of these knots can be quickly made.
Along the line of railway in southern Paran4 the knots are used for
fuel on the locemotives, and bins of these combustible, inflammable
stores of the araucarian forest are frequently seen at the railway
stations. The young trees when not over four to six feet high present
an appearance very unlike the adult form with its smooth bole and
palm-like apical whirl of branches with leaves growing in large clusters
at the ends of the branches. The young tree is thoroughly covered
with broad pine leaves resembling small Cordaites. On the branches
these leaves point outward and upward, but those on the bole which
are still larger bend sharply downward at the point of attachment
and present sharp needle-like points to stay the progress of any small
climbing mammal which would find equal difficulty in reaching the
main stem of the tree by descending one of the branches. This
apparent adaptation of the foliage to protection from arborescent
animals in the young stage of the plant is a devise the use of which I
did not observe, for on account of the nocturnal habits of most of the
animals of the Brazilian forest and the noise made by the approach
of our troop of mules we saw during the day no wild mammals
whatever.

September 2nd.— For over two hours after leaving last night’s
camp, our route lay through the forest on the south slope of the Serra
do Espigao. About 11 a. M. we emerged on a more travelled trail
along which we passed several huts. Late in the afternoon the sick
mule lay down and we were forced to abandon the animal, as our

WOODWORTH: GEOLOGICAL EXPEDITION TO BRAZIL AND CHILE. 29

slender stock of provisions obliged us always to make rapid transits of
the long stretches between settlements. We camped on the head-
waters of the Rio dos Patos, the bed of which abounds in agate pebbles
derived from the trap.

September 3rd.— The mule road ascended rapidly from camp to
elevations by aneroid of 4,000 feet on the crown of the northwestern
arm of the Serra where there is a small settlement and a store. Small
streams continued turbid. The interstream areas were weathered
into deep pits with numerous bogs.

September 4th The march to the northwestward today lay at a
high level on the northeast side of the crest of the Serra, a rolling
country partly open campo and partly occupied by pine. Inhabitants
became more frequent and pack-trains indicated the proximity of the
terminus of the railway. At half-past two we heard the whistle of a
locomotive and at 5 Pp. M. came out upon the line of the railway in
construction from Porta da Uniao southward over the trap -plateau
to the Rio Peixe. At 9 o’clock at night we found our way into Sao
Joao, from which place on the following day we returned by train to
Ponta Grossa leaving the pack-train to come along at its own gait.

September 15th.— With the view of studying more in detail the
tillite beds on the banks of the Rio Jaguaricatu in northeastern Parand,
Dr. Oliveira and myself with a small camp outfit and one camarada
went from Ponta Grossa to Sengéns Station on the newly constructed
railroad. While absent from our camp on the 16th, our tent and most
of the contents were destroyed by a fire. Fortunately none of my
instruments or notes were lost. ‘Through the courtesy of the railway
officials we slept in a railway storehouse that night and the following
day returned to Ponta Grossa.

September 19th—In company with Dr. Euzebio Oliveira I
returned to Rio Negro with the intention of exploring the tillite beds
along the road between that town and Sao Bento. It was on this
occasion that Dr. Oliveira found a bed of fossiliferous marine shales
between boulder-beds on the south side of the Rio Negro. About
two legoas above Rio Negro there is a water-fall where the Ribeira das
Rutes falls over a hard bed of tillite.

From Rio Negro I returned to Curityba en route to Paranagua,
whence by steamer I reached Rio de Janeiro. The delay in waiting
for the steamer was utilized at Curitybaand at Paranagua in examina-
tion of the superficial deposits and studying the topography, the
results of which studies are embodied in the account of the geomor-
phology of this part of Brazil.

30 BULLETIN: MUSEUM OF COMPARATIVE ZOOLOGY.

From Rio de Janeiro, again through the intermediation of Dr.
Derby, I set out on October 12th for the planalto of Sao Paulo to-
examine the tillite beds on the railway line between Itaicy 4nd Piraci-.
caba. Dr. Cardoso, chief of the Sao Paulo Geographical and Geologi--
cal Commission detailed Dr. Pacheco of that bureau to accompany me.
The excursion was made with a railway automobile, over the part of
the line traversing the sediments. This afforded every opportunity
for a rapid reconnaissance. On returning to Rio de Janeiro, I sailed
from that port on October 24th bound to Laguna for the purpose of |
examining the Permian section of the Tuberao Valley. Dr. Euzebio
Oliveira joined me at Paranagua. On this and the return voyage the
necessary delays in waiting for the small coasting steamers gave
opportunity of making observations upon shore-line changes and the
general features of the Serra do Mar at Sao Francisco, Itajahy,
Florianopolis, and Laguna.

While waiting for the steamer at Laguna a study was made of a
sambaqui or shell-mound forming a small terrace on the flank of the
granite hill at the south end of the town. The top of this deposit is
about 100 feet above sea-level and has been dug into as a local source
of lime. The deposit is composed principally of a small lamelli-.
branch, in parts of the mass somewhat cemented together. In the
upper part of the heap I found a stone-axe, fragments of a fine-grained
rock evidently used for opening shells, fish-bones, mammalian bones,
and part of a human skull, as well as portions of a large sea-urchin, all
indicating by their leeched condition considerable antiquity. At a
lower level a large Ostrea was abundant. The deposit has a rough
stratified structure but nothing like that induced by deposition
beneath moving and assorting currents of water. Neither in the
topography nor in the structure were there characters seeming to
demand other than surface accumulation for the origin of this shell-
heap. The dead shells of a recent large snail, Bulimus, were rather
abundant on the surface of the kitchen-midden, and the old shells.
were sliding down to the present beach. Professor Dixon has kindly
written a note on the collections I made at Laguno, (p. 132).

Voyage from Rio de Janeiro to Talcahuana. November 25th, 1908..
The journey from Brazil through the straits of Magellan to southern
Chile was begun today by sea on the steamship Oravia of the Pacific
Navigation Company, with stops at Monte Video, Punta Arenas,
and Coronel.

November 29th— At Monte Video. Two partly dismasted barks.
lay at anchor in the harbor, having fallen back to this port for repairs.

WOODWORTH: GEOLOGICAL EXPEDITION TO BRAZIL AND CHILE. 3]

after an encounter with ice off Cape Horn. Captain Wm. B. Oakley
of the Oravia informed me that after the Valparaiso earthquake of
1906 there had been an. unusual amount of ice in the sea about the
southern end of South America, the ice having been dislodged from
glaciers by the earthquake, so it was believed. This report agrees
with the effects reported by the late Prof. R. 5. Tarr arising from the
Alaskan earthquake of 1899, which caused much disturbance of the
glaciers on the Alaskan coast and warrants the belief that at least
local earthquakes may greatly accelerate the flow of glaciers.

December 4th We entered the straits of Magellan during the
night. To the voyageur entering the straits from the east the high
barren plains on the north and the treeless low-lying plain of eastern
Tierra del Fuego on the south alike recall the plains of glacial Cape Cod
veneered with glacial drift. Between Elizabeth Island and Punta
Arenas there are some irregular terraces of varying height, appearing
more like glacial contemporaneous terraces than the horizontally
levelled benches cut by waves. Back of the town of Punta Arenas
there is a deep gully at the mouth of which lies a prominent deposit
forming a rounded southward slanting ridge, on the seaward slope of
which the town is mainly built. There is a remnant of what appears
to be a stream delta north of the town, now forming a terrace. The
Pour-quoi-pas of M. Charcot’s French Antarctic expedition lay at
anchor off the town.

The passage from Punta Arenas to Cape Holland was made after
4 p. M. but permitted a general view of the profile of the Andes
rising above the plains of Tierra del Fuego. (Fig. 5.) It is evident

Fria. 5.— Generalized profile of the Andes on Tierra del Fuego, showing the
arching up of a once baselevelled but now deeply dissected mountain mass.
It is not here intended to interpret the steep western descent into the
Pacific Ocean.

that here as far to the north the peaks and valleys are carved out of
a theoretically smoothened surface of the deformed rocks composing
the folded chain. It is conjectured that this now warped, once
baselevelled, surface passes beneath the Tertiary and Pleistocene de-
posits of the plains of Tierra del Fuego, forming the platform on which
these less ancient deposits repose. Towards the western margin of

32 BULLETIN: MUSEUM OF COMPARATIVE ZOOLOGY.

the Andesian uplift, the dissection of the surface is more complete,
and the descent to the bed of the Pacific Ocean is steep.

December 8th.— The Oravia put into Coronel for a.supply of coal.
The mines at and near Coronel are the most important on the coast of
South America. The principal outcrop of coal is inland, some distance
from the shore but the workings follow the coal beneath the sea.
The beds in which the coal occurs are regarded as of Eocene age.
(Sundt, 1908, p. 37-44). The coal is bituminous and is described as
bright and clean but light. The Arauco Company has produced
from its mines as much as 200,000 tons per annum. The coal is ex-
tensively used by steamers plying the west coast and in the locomo-
tives of the Chilean railway (Alcock, 1907, p. 85).

Captain Oakley of the Pacific Steam Navigation Co. expressed the
opinion that the coast at Coronel has risen in recent years. An old
wreck partly buried in the beach skirting the south side of the point
on the north of the anchorage, according to his observation now lies
higher than when he first saw it. I note the opinion as a matter for
further investigation.

While the steamer was taking in coal I went in a sailboat to Lota,
a small port. about four miles south of Coronel, where coal is also
mined. A conglomeratic sandstone here crops out, the scattered
pebbles and massive bedding of which strongly suggests the trans-
portation of the pebbles by ice. In my hasty examination of the rock
I was unable to find striated pebbles. It should be noted in this
connection that Darwin (1846, p. 69) described “great boulders of
granite and other neighbouring rocks, embedded in fine sedimentary
layers” in Tertiary deposits along the coast of Peru. Back of
Coronel! there are at least three terraces in the bed rock but whether
due to differential weathering or marine erosion at different levels I
was unable to determine. Along this coast towards the Tumbres
Peninsula there is an uplifted baselevelled surface forming a narrow
bench cut back and cliffed by the sea at the present level, with sub-
dued remnants of higher rock masses rising above the terrace as in the
case of the Paps of Bio Bio. This bench must be early Pleistocene or
late Tertiary in date. Narrow steep ravines crease the cliff face;
but practically no stream-cut channels cross it with their mouths at
sea-level. All the small vales which traverse the bench are hung up
at the seaward edge by reason of the cutting back of the terrace by
the sea. This bench is probably to be correlated with a well-defined
bench of erosion at Corral on the south but is much higher above the
sea.

WOODWORTH: GEOLOGICAL EXPEDITION TO BRAZIL AND CHILE. 33

We entered Concepcion Bay on December 8th, and on the 9th I
went ashore at Talcahuano. I spent several days in the vicinity of
Concepcion examining evidences of change of level. As the results
of my observations are given in a subsequent chapter I make mention
here only of some unrelated geological details.

On December 11th, while yet in bed, at about ten minutes past
seven, I felt a slight quivering followed by a sharp lurch of the hotel.
The wooden framework of the ceiling cracked and creaked. The
daily press stated that at 7.10 a short but violent shock was felt.
Again on the 13th of December while in Concepcion, the daily press
reported during various hours of the night subterranean noises ac-
companied by a shock at 12.40 a. M., which caused some alarm. I
was awakened at about 1.35 a. mM. by what at the time I thought was
the fall of an object in the room below; meanwhile I heard a rumbling
noise sounding like that made by street cars. The destruction of
Valparaiso by the earthquake of August, 1906, has made the inhabi-
tants fearful of a repetition of such violent earthquakes, particularly
at Concepcion whose ancient site at Penco on the shores of the Bay
has been the seat of the most famous earthquakes in the annals of
Chile. The practice of the Spanish in South America of leaving a
site more than once damaged by earthquakes has much to recommend
it. In the case of Old Concepcion or Penco, twice destroyed by
earthquakes and inundations from the sea, the site was mainly on a
marsh behind a barrier beach filling out an indentation of the coastal
hills, a location, owing to the soft nature of the recent alluvial deposits,
likely to be severely shaken by earthquakes. The new city of Con-
cepcion stands on a plain of Pleistocene alluvium mantling much
disturbed sandstones which here and there rise as low hills through
the plain. That this city has escaped destruction so far from earth-
quakes is seemingly due rather to the failure of local violent shocks
than to its location.

On December 13th, I visited Penco, the site of ancient Concepcion.
Since Darwin’s time the construction of a railway along the shore of
the bay has led to the partial demolition of the old Spanish Fort, the
seaward portion of the walls only remaining. (Plate 4). Of what
I presume to be this building, Lyell states: “It has, however, been
ascertained that the foundation of the Castle of Penco was so low in
1835, or at so inconsiderable an elevation above the highest spring
tides, as to discountenance the idea of any permanent upheaval in
modern times, on the site of that ancient port; but no exact measure-
ments or levellings appear as yet to have been made to determine this

34 BULLETIN: MUSEUM OF COMPARATIVE ZOOLOGY.

point, which is the more worthy of investigation, because it may throw
some light on an opinion often promulgated of late years, that there
is a tendency in the Chilian coast, after each upheaval, to sink grad-
ually and return towards its former position.” (Lyell, Principles of
geology, 11th ed., New York, 1887, 2, p. 156). I found this view
still current at Talcahuano, and it is evident from the view referred
to that no marked permanent change of level has affected the ancient
ruin on the beach at Penco since it was constructed.

The railway from Concepcion to Penco traverses an outcrop of
coarse, waterworn conglomerates at Paradero de Santa Ana, evidently
a member of the Tertiary series which underlie the plain between the
Coastal Cordillera and the ridge of crystalline rocks which form the
Tumbres Peninsula on the seaward side of the Bay of Concepcion.
In the bank of the bay shore immediately south of Penco, coarse
gravels composed of pebbles of crystalline rocks and occasionally
large rounded stones crop out in the railway cut in marked uncon-
formity upon sandstones. The change from the blue color of the
gravels at the base to orange at the top of the bluff is evidently an
effect of weathering now in progress. Certain portions of the pebble
beds contain stones a foot or more in diameter. The deposit is
mainly stratified with flattish ovoid pebbles lying in the planes of
bedding. In the lower part of the exposure the paste of fine material
is less obvious than in the upper portion. Lenticular beds of sand
and finer gravel appear at intervals in the section, pointing to inter-
mittent or shifting currents or streams. From its general relations
and want of consolidation I supposed the deposit to be of Pleistocene
date and possibly not older than the bench which at about the same
elevation can be traced around the seaward face of the Tumbres
district at the Paps of Bio Bio. If this correlation be correct the
deposit may bé of marine origin, but no fossils were seen in any part
of the section.

On the exposed face of the gravel bluff loose materials were sliding
down in such a manner as to afford an instructive example of the
post-depositional striation of pebbles. A large rounded cobble
protruding from the section (Fig. 6) was well striated on its exposed
surface by pebbles sliding down over it in the wasting of the upper
part of the bluff. I observed this process in action, and it showed the
necessity of taking every precaution in accepting detached striated
pebbles as evidence of glaciation.

On December 18 I left Concepcion for Valdivia to examine the shore-
lines of that district for the reason that Darwin stated that here he

WOODWORTH: GEOLOGICAL EXPEDITION TO BRAZIL AND CHILE. 35

found no local indications of elevation. The results of this journey
are contained in my remarks on the coast of Chile. (p. 132).

December 23th.— Being due at Santiago on Christmas day [I left
Valdivia, journeying northward by rail through the Longitudinal Valley
of Chile, with a stop at San Rosendo to make a study of the Pleisto-
cene terrace deposits on the Rio Bio Bio.
In ascending the valley of the Calle Calle
through the gorge in the Coastal Cordillera,
the crystalline schists were observed to
have the same steep eastward dip as in the
Tumbres Peninsula in the latitude of Con-
cepcion. The rock-bench which is so pro-
nounced a feature about the shores of
Corral and the Valdivia River along the
Calle Calle above Valdivia becomes suf- ,, ,

: . 1a. 6.— Section showing how
fused with gravels, presumably Pleistocene. upended Gartucs G0'é cobiie:
Pebbles of nonschistose rocks abound, — stone became striated by
indicating the derivation of the materials {°" Nee heat ce aan
from the Tertiary and voleanicrocks within chile.
the Longitudinal Valley.

From December 25th, 1908, until January 5th, 1909, I remained in
Santiago in attendance on the sessions of the First Pan-American
Scientific Congress. I have published a brief note on the geological
papers read at this meeting. (Woodworth, 1909).

Under the guidance of Dr. Phillipi, a visit was made to the Museum
of natural history and to the Museum of the mining society. Among
the collections of this Society I was shown several remarkably intricate
examples of stones carved by the sand-blast. These were gathered
on the surface of the desert of Atacama by Mr. Carlos Sundt. In some
cases large holes had been eaten through irregularly carved fragments
of rocks by this insidious process. Through the courtesy of Major
Montessus de Ballore I was also permitted to examine the Seismological
Observatory then in process of installation in the hill of Santa Lucia
in Santiago. ‘This is the principal station of the seismological service
of Chile. From Santiago I proceeded to Valparaiso for embarkation
on a steamer bound for Panama.

The Valparaiso Earthquake of August 16, 1906. While waiting in
Valparaiso for the sailing of the steamship Limari, a day was
devoted to a casual examination of the effects then visible of the
disastrous earthquake of August 16, 1906. As has been so often
observed in the downthrow of maritime cities by earthquakes, the

36 BULLETIN: MUSEUM OF COMPARATIVE ZOOLOGY.

damage was most severe along the water-front where the aliuvial
deposit and made-ground imposed the thickest layer of loose uncon-
solidated material upon the bed rock. In this zone many buildings
were completely levelled. Nearer the base of the cliff which partly
separates the upper from the lower portion of the city, buildings stood
with their walls cracked and cornices broken away but frequently
otherwise safe for habitation. Upon the sloping ground above the line
of cliffs where a superficial layer of weathered rock on the inclined
surface had slipped down carrying buildings with it, and in the ceme-
tery where similar conditions existed, the destruction was most pro-
nounced. The reconstruction of the business houses along the
incoherent ground of the water-front insures a recurrence of the tale
of destruction when in the future the seismic movement affects in this
vicinity the line of displacement which skirts the coast of Chile.
From the studies of Dr. H. Steffen in the case of this earthquake the
disturbance appears to have had its origin off Coquimbo, a sea-port
198 marine miles north of Valparaiso.

I was impressed with the fact that in a large number of the houses
that were not completely destroyed the damage was at a maximum in
the peripheral parts of the buildings; that though one or more of the
outer walls were demolished or thrown outward, the internal walls
and the inner angles of the floors were left standing in place unen-
cumbered by fallen wreckage in such a manner that persons unable
to leave these buildings would have escaped with their lives had they
sought refuge during the height of the shocks in the internal corners
of the rooms. I had occasion later to note instances of the same sort
amid the ruins of Kingston, Jamaica, produced by the earthquake
of January 14, 1907. It would seem advisable from this observation
that houses in earthquake countries should be constructed with one or
more sets of rectangularly intersecting, internal walls well united of
materials whose period and amplitude of vibration as a mass is
identical so that one end of the building will vibrate as nearly as
possible in unison with the other. From the lack of this synchroneity
and equality of lateral swaying the outer walls will probably be thrown
off or toppled down but the central structure of buildings properly
balanced, as numerous examples have shown, often stands and by a
slight special construction adapted to the purpose, the internal angles
of such intersecting walls might in many cases prove places of refuge
and security.

This Chilean earthquake, coming at a time when the people of the
United States were preoccupied with the calamities of the similar

WOODWORTH: GEOLOGICAL EXPEDITION TO BRAZIL AND CHILE. 37

great disaster which had but four months previous led to the destruc-
tion of San Francisco, has received such scant attention in North
American scientific journals that the following notes taken largely
from local accounts and the paper of Dr. H. Steffen of the University
at Santiago are deemed worthy of record.

In Valparaiso the immediate loss of life was estimated at 3,000
persons and the wounded at 20,000, some of whom subsequently died
in consequence of their injuries. Several hundreds of lives were
lost in surrounding towns and villages. (Rozas y Cruzat, 1906.)

Owing to the lack of seismographs in Chile in 1906, the exact time
of the earthquake is a matter of some uncertainty. At Santiago where
there is an astronomical observatory the first sensible shock appears
to have taken place 7h. 58m. 36s. Pp. M., August 16th, local time.
Taking the longitude of Santiago as 4h. 42m. 46.4s. west from Green-
wich, the initial shock was felt there at Oh. 41m. 22s. Greenwich
mean time, midnight to midnight, August 17th. At Valparaiso
whose time is 3m. 50s. later, the first shock is placed at 7h. 55m and at
7h. 56s. by several different time-keepers. The mean of the times
at Valparaiso, 7h. 55m. 30s. makes the apparent time of the first
shock at Valparaiso 44 secs. earlier than that at Santiago. The
earthquake from various studies appears to have originated in a
fault plane off Coquimbo about 228 miles north of Valparaiso. The
seismographic indications as to the time of origin point to Oh. 40 m.
as the probable moment of the primal great shock.

From. the varied estimates of observers, it appears that in the
central tract along the coast extending north and south of Valparaiso
between the parallels of 28° and 39° S. L. there were two series of
shocks separated by an interval of relative quiet. The first strong
shock of exceptional duration, lasted from four to five min., while
the second equal to or perhaps stronger than the first one had a dura-
tion of 1 min. or less. Outside of the epicentral region but one
continuous series of shocks seems to have been noticed. (Steffen,
1907, p. 23). A vertical movement was distinctly recognized at
isolated points between 38° and 36° S. L. with greater distinctness and
precision on the north as far as the river Mante. Most observers
judged the primal movement to be upward. Dr. Steffen obtained
testimony to the effect that heavy objects in Illapel, Santiago, Talca,
etc., within the central tract were thrown upward to a certain height
above the base on which they stood, contradicting as he notes the
statement of Dutton (1904, p. 148) that there never has been observed
an acceleration sufficient to overcome the force of gravity. It may

38 BULLETIN: MUSEUM OF COMPARATIVE ZOOLOGY.

be noted also that in the Indian earthquake of 1895 (Oldham,
1899), many large boulders were inverted on their sites In a manner
demanding apparently their projection free from the base on which
they rested.

The most important data gathered by Dr. Steffen has a definite
bearing upon the controversy raised by the distinguished Austrian
geologist, Professor Edward Suess, over the question of the uplift of
the coast of South America accompanying earthquakes, a thesis set
forth by Darwin and Fitzroy. Dr. Steffen was aware of the impor-
tance of critical observations made at once upon evidences of change
of level of land and sea along the disturbed coast. From the in-
formation obtained he came to the conclusion that there can scarcely
be any doubt as to an elevation of the coast from the mouth of Rio
Mataquito to that of the Choapa along a segment of the sea-border
corresponding to the area of maximun perturbation in which the
seismic intensity rose to the degrees of VII and X in Mercalli’s scale.
This movement appears to have been greater on the north than on the
south. The measurements most worthy of confidence in Dr. Steffen’s
opinion are 40 em. at Llico, south of Valparaiso in about 34° 40’ S. L.
and 70 to 80 cm. at Zapillar, north of that city in about 32° 25’ S. L.
Sefior Lorenzo Sundt, an experienced geological observer whom I met
in Santiago, stated that in the bay of Valparaiso some 200 meters west
of the pier of the Matodero at Portales, there was to be seen upon the
rocks after the earthquake a white band composed of a small species
of barnacle and of Algae of the Corallinacea forming a natural mark
which at time of low tide was left uncovered for two feet above low-
water mark, although before the earthquake it was not visible.
Likewise a local officer of Portales who had observed the coast for
eighteen years noticed after the earthquake that a rock, which he had
not seen before appeared above the surface of the lowest tides. These
stations which are composed of the solid rock are free from the doubts
which affect the altered position of loose materials. The probable
correctness of the contention of Darwin and Fitzroy that at times of
great earthquakes on the coast of Chile there is an upward movement
of the land seems now to be established; but whether this uplift is
permanent is doubtful, since as in the celebrated case of Concepcion,
I was informed when in that vicinity that it was the opinion of the
naval officers stationed at Taleahuano that a slow subsidence is now
in progress. As for the uplift of faulted blocks in relation to sea-level
it is now well established and nowhere more pronouncedly than in
Alaska by Tarr in the case of the earthquake of 1899 in which an

WOODWORTH: GEOLIGICAL EXPEDITION TO BRAZIL AND CHILE. 39

elevation exceeding forty-two feet was ascertained by indubitable
evidence. But this case arose in a folded mountain-chain of recent
development where uplift is not denied by Suess. The Coastal
Cordillera of Chile is composed of an elongated horst lying outside of
the folded chain of the Andes and the evidence of uplifts at the time
of the Valparaiso earthquake of 1906 is therefore of especial interest in
confirming the conclusion of Darwin and Fitzroy that an elevation
of the coast may take place concomitant with an earthquake on this
coast, though it does not prove that the coast has been permanently
elevated by successive stages at long intervals in this manner.

From the information collected and published by Dr. Steffen, we
also learn that no noticeable seaquake wave or tsunami was set up
in that part of the coast which was the seat of the maximum seismic
activity and change of level. At Constitucion and particularly in
the bay of Talcahuano unusual movements of the sea, however,
appear to have taken place. At Tomé on the eastern shore of this
large shallow harbor at a time differently stated as a quarter of an
hour and as an hour after the earthquake, the sea retired for about
fifty meters, returning quietly to its place. This movement was
repeated three or four times, the last two incursions being the greatest,
covering a space of seventy meters. At Penco, the site of Old Con-
cepcion, made famous by the number of times it has been devasted
by earthquakes and sudden irruptions of the sea, similar phenomena
took place. A wave rose to the level of the railway (Plate 4) along
the beach and passed through the bridges and drains to the low
ground behind, causing the inhabitants in their alarm to begin to
take refuge in the neighboring hills; but the sea returned, so it is
stated, to its normal level in less than ten minutes. (Steffen, 1907,
p. 66-67).

The Valparaiso earthquake followed immediately upon a heavy
earthquake on the submarine border of the Aleutian Island platform
in 50 N. L. and somewhere between 175° and 180° of longitude E.
from Greenwich according seismometric determinations.! The mean
of the determinations of the time at origin of this shock by Zoeppritz,
Oh. 10m. 47s., and by the observatories at Florence, Oh. 10m. 35s.,
Laibach, Oh. 11m. 19s., and Tokyo, oh. 11m. 16s., is Oh. 10m. 59s.

1 Zoeppritz places the origin within 100 kilometers of 180 of Longitude from Gr.
and the time as Oh. 10m. 47s. 20s. See E. Rudolph und E. Tams Seismogramme
der nordpazifischen und sudamerikanischen erdbebens am 16 August 1906, Strassburg
i. E. 1907. Professor Omori gives the origin at 175 +° E. L. and the time as Oh.
1lm. 44s. G.M. T., midnight to midnight.

40 BULLETIN: MUSEUM OF COMPARATIVE ZOOLOGY.

which may be taken as Oh. 11m. or about 29m. 10s. previous to the
calculated time of the Valparaiso shock at origin.’

According to Benndorf the secondary preliminary or transverse
group of seismic waves arrive at distances of 14,000 kms. after an
interval of 30m.; according to Rizzo’s later work, we may expect them
to arrive as early as 29m. 30s. after the primal shock, or after a mean
interval of 29m. 45s.+ 15s. It thus seems highly probable that the
Valparaiso shock was set off by the passage of the vibrations emanating
from an earthquake in the Aleutian Islands.”

Voyage from Valparaiso to Panama, and thence to New York. On
the 7th of January 1909, I sailed from Valparaiso by the Chilean
steamship Jimari for Panama, with stops at various ports on the
intermediate coast. Along this coast as far north as the island of
San Gallan near Callao, from time to time one sees from the deck of a
passing vessel sea-caves somewhat above the present level of the sea,
indicating a modern uplift in relation to sea-level. Above these
recent indications of a change of level the embayments of the coast
are terraced as at Coquimbo and Ilo to a height of a few hundred feet.
Usually above the highest terrace which is somewhat more eroded and
creased by ravines than those successively lower, there rises a dissected
slope to the edge of the lofty plateau. That these terraces facing the
sea indicate changes of level one can hardly doubt. Were they due
to differential weathering the upper ones would be as sharply defined
as the lower terraces. In this respect the coast is.in sharp contrast to
much of the region south of Valparaiso. We reached Panama January
26, where our fellow-passenger Colonel Gorgas of the Isthmian Canal
Commission showed us many courtesies. On January 27 I sailed by
the Royal Mail Steamship Nile for New York with a stop at Kingston,
Jamaica. This enabled me to spend a day in the examination of the
destructive work of the earthquake of January 14, 1907, the effects
of which were visible on every hand in the unfortunate city. The
Nile arrived at New York February 4th, 1909.

' The initial Valparaiso shock according to the results obtained at Laibach took
place at Oh. 40m. 5s. Cf. Galdino Negri. Velocidad de propagacion de las Ondas
Sismicas. Observatorio astronomica de la Universidad nacional dela Plata. Memoria
presentada al IV Congreso cientifico internacional americano celebrado en Buenos
Aires del 10 al 25 de Julio de 1910. La Plata, 1911. p. 100.

* See in this connection, Quelques constantes sismiques trouvées par les macro-
sismes. Nota d'Emilio Oddone da Roma. Strassburg, Bureau Central, 1907.

WOODWORTH: GEOLOGICAL EXPEDITION TO BRAZIL AND CHILE. 41

IV. OUTLINE OF THE GEOLOGY OF SOUTH BRAZIL.

For the, understanding of the relations of the Permian glacial
deposits of south Brazil it is proper to give a résumé of the geological
structure of the country. Passing over the pioneer work of Lieut.
Colonel Wilhelm L. von Eschwege, whose writings are lithological
rather than geological, the main outlines of this structure are to be
found in the publications of Dr. Orville A. Derby and his associates,
Dr. J. C. Branner’s résumé in his Geologia elementar, and in Dr. I. C.
White’s Report on the coal area. In this résumé the observations of
the writer have been allowed to a limited extent to enter into the
interpretation of certain features of the region.

The formations which enter into the structure of this part of Brazil
may be grouped in the following terranes:—1. The Pre-Devonian
or igneous and metamorphic belt of the coast including the Serra do
Mar region, frequently classed as Archean. 2. The Devonian
including the sandstone cuesta of the Serra das Furnas and the over-
lying fossiliferous shales of Ponta Grossa in the state of Parand. 3.
The Permian beds, including conglomerates, tillite beds, as well as
sandstones and shales, the latter coal-bearing in the south. 4. The
Triassic sandstones and trap sheets; the latter making the escarpment
known as the Serra Geral and its topographical equivalents elsewhere.
5. The Tertiary fresh-water deposits of the upland and possibly along
the coast. 6. The Recent deposits along the coastal border now
slightly elevated. |

The Pre-Devonian Terrane-— The Pre-Devonian belt is here so-
called because it comprises a complex of igneous and metamorphosed
sedimentary rocks unconformably overlain by the westward dipping
Devonian sandstones and shales of the upland, as yet the oldest known
fossiliferous group in the region. The use of the term Archean or
Pre-Cambrian for this complex seems at present inadvisable because
of the possibility that certain of the metamorphosed clastic members
of the series may be of Lower Silurian (Ordovician) or Cambrian age
analogous in their structural relations to those of these ages in the
metamorphic belt of the Piedmont terrane of the Atlantic slope of
North America.

In the latitude of Rio de Janeiro this belt of complex rocks includes
the elevations known as the Serra do Mar and an inner line of moun-
tainous relief known as the Serra da Mantiqueira. Gneiss of unde-
termined origin appears to be the most ancient member of the region

42 BULLETIN: MUSEUM OF COMPARATIVE ZOOLOGY.

and may be fairly presumed to be of Pre-Cambrian age. In the
coastal border gneissoid granites with well-developed augen-structure
abound. This rock is nowhere better shown than about Rio de
Janeiro, as in the Pao de Assucar at the entrance to the harbor.
Apparently of later date than the granite-gneisses are intrusions of
phonolite and tinguaite which occur in the form of stocks, while more
basic dikes are not wanting.

In eastern Sao Paulo there is an evidently infolded belt of slates and
limestone, seemingly the newest member of the metamorphosed
series. The distribution of this formation has not been shown on
geological maps. In southeastern Santa Catharina the belt between
the Permian border and the coast appears to be entirely granitic,
though north of this district quartz-schists are involved in the complex
as in the vicinity of Itajahy.

North and south of the Permian-Triassic basin of south Brazil
these Pre-Devonian rocks have a vast extension, stretching far into
the interior in the state of Minas Geraes and forming the greater part
of Uruguay. In the region south of Rio de Janeiro it is evident from
the relations of this series to the overlying Devonian beds that one
or more periods of deposition, mountaining-building, and igneous
intrusion preceded the deep erosion of the deformed mass as the
prelude to the incursion of the Devonian sea. The once eastward
extension of this deformed and eroded Pre-Devonian terrane into
what is now the basin of the Atlantic Ocean has no assignable limits.
The basal beds of the Devonian rest on a westward dipping now
slightly warped surface of these older rocks in a manner to show that
the sea crept in over a region of little or no relief but how far this
peneplaned surface extended to the eastward there are no definite
facts to show.

The Devonian Terrane-— The Devonian of south Brazil occupies
a narrow belt of outcrop along the eastern margin of the Permian area,
disappearing on the north in Sao Paulo and on the south in Parana.
As strata of Devonian age reappear far to the northwest at Cuyuba
in Matto Grosso, the Devonian is thought to extend beneath the
Permian and Trias over a vast area. It is agreed that the Devonian
of South Brazil includes at its base the thick, light-colored sandstones
which form the Serrinha and the Serra das Furnas. This formation
is overlain by fossiliferous shales of Mid-Devonian age. The De-
vonian shales are intruded by numerous dikes and sills of diabase
whose outcrops may be seen in the vicinity of Ponta Grossa. Their
date of intrusion is probably Triassic. The absence of the Devonian

WOODWORTH: GEOLOGICAL EXPEDITION TO BRAZIL AND CHILE. 43

wa’.

Wl, Seale
« »
g

5. 9°
ee
a

Povsp Acacre

'g Villa Prete.

; AN
7% Ventana *
YT

Mit
uaort
yy Yi YY

i

Lf })

.

|

| oa bogs

| e Ye Faneei =
sf Wo NG;

le Ah Le] Noo PR

eared ge. oF

ga
‘Z Wi /
Vit)! ~d
Pivop, ‘ H
sth wf 6 ~
| i IO Serr Azali:
fewere } ,

|
A
i

! p= R Batumi 77>

ee yaa -s
| Za Ppiranga
fDi 5 <a
<-
Lae 3 “
EY gee f
Meas Sssasy
\ EY wt a .
-@ fimbituve
Ms +)
‘. Sg Passiaho 2 he ee Mi, Wie A
. i¥- Soares , Se Uf THA LL
X : Z hes x PALMEIRA, Uf WY, > Qe Udi A
” 3s Fernandes Pinheiro — f Cn iy y rf MYY 4f4 = CAMPO LARGO
. of f. LIS Lt SI EZ : Ee
AS 7 - TY whl Ge ->e
lraty Ss) ae ht! hy] Wi TE
; (BOE
y Tamandus > RS
© Ve St EN Balen
: WAI
Ant Rebouge : cS Cuajevera
. . , ase wil! WA
i ; . - SiN Vy
' a 5 > . : We iin "i ANNA hie
S036 daTiumahe * Y b Cipiverys oS Se bn *
eee , Sent
ees Sv" aN \
oy i LAPAS
et Syste \

== ee

Basts. tes Rio (30 Tnsee Oy WP Moraes

Fic. 7.— Map of the Devonian areain Paran4&. The lined area is Devonian;
that on the west is Permian; that on the east is the Pre-Devonian meta-
morphic and granitic terrane. (Published by permission of the Director
of the Geological and Mineralogical Service of Brazil).

44 BULLETIN: MUSEUM OF COMPARATIVE ZOOLOGY.

north and south of the present line of outcrop where the Permian
rests upon the Pre-Devonian terrane is evidently due to erosion
of the beds in Upper Devonian or Carboniferous times.

That the unfossiliferous sandstones referred to the Lower Devonian
represent the shoreward facies of marine sediments is clear and it is
possible that the upper part of the beds may represent shore deposits
laid down simultaneously with the lower off-shore portion of the
fossiliferous shales known to be of Mid-Devonian age. The thickness

a ae /
— ~ 6% | ee SS ree Se Go ETT

oN tee SEP / Tot toe Saxe
SS ETL ( Lote PES

Ct

Fic. 8.— Sandstone escarpment looking northeast from Lago, Parana.

of the outcrop of shales is however such as to indicate that they
originally extended much to the eastward of the present underlying
eastern limit of the sandstone just as the peneplaned surface on which
the terrane rests evidently extended to the eastward of its existing
traces.

The inference above stated that the Devonian shore in this region
lay to the eastward is not only suggested by the westward existing
dip of the beds, which attitude might be explained by a rotational tilt
from an original eastward dip, but the assumption is in consonance
with the basin-like form of the entire geological province of south
Brazil. In Triassic times over a vast area non-marine sediments were
poured on to this tract from outlying areas of land. In the preceding
Permian both marine and non-marine sediments accumulated in the
same or nearly the same area; and on the eastern border of the outcrop
of these sediments, as will be shown later, there is evidence of the
derivation of materials from an area of erosion on the east. The
Carboniferous period was here one evidently of uplift or withdrawal
of the sea and erosion under the atmosphere, a fact of little import
on the attitude of the Devonian stratigraphic plane other than to
indicate the probable proximity of land in this region in the later
Devonian.

The Permian Terrane.— It now seems most likely that the strata.
referred to the Carboniferous in the earlier reports on this district are
of Permian age. The fossil plants and few reptilian remains found

—=s>*

WOODWORTH: GEOLOGICAL EXPEDITION TO BRAZIL AND CHILE. 45

above the base of the series are without doubt referable to the Per-
mian. The Permian strata are known along a narrow belt from
southern Minas Geraes southward through the states of Sado Paulo
and Santa Catharina; they extend into Rio Grande do Sul where the
beds turn westward on the southern border of the basin-shaped area
of late Palaeozoic and early Mesozoic sediments which form the
central geological province of South America. Except where the
Devonian beds intervene these strata rest with marked uncon-
formity upon the Pre-Devonian rocks. According to Dr. Derby !
the Permian consists of the following beds: —

Rocinha limestones. 3 meters. Stereosternum bed.
Estrada Nova shales. 150 meters. Gray, mottled shales with flints;
known in Santa Catharina.

Iraty shales. 70 meters. Black shales with Mesosaurus.

Palermo shales. 90 meters. With fossil wood.

Coal measures. 150 meters. Sandstones and shales with 2
beds of coal. Glossopteris
flora.

Tillite beds.
Basal beds. Sandstones with boulders and pebbles.

Dr. I. C. White in his admirable Report? on the coal fields of south
Brazil groups the Permian beds in the following section: —

Passa Dois series.. Rocinha limestone. 3 M.
223 meters Estrada Nova shales with chert 150
Iraty black shale (with Mesosaurus) 3 70
Tuberao series. Palermo shales 90
180 meters Rio Bonito sandstones and shales with
coal and Glossopteris flora 158
Orleans conglomerate 5
Yellow sandstones and shales to granite
floor nk

Total thickness 403 meters or 1,322 feet.

Dr. White appears to regard the various formations of the Permian
as more or less persistently parallel throughout Sao Paulo, Parana,

1 Verbally communicated at Ponta Grossa in 1908.

2 Dr. I. C. White, Relatorio final apresentado a S. Ex. o Sr. Dr. Lauro Severiano
Miiller. Comissio de Estudias das Minas de Carvao de Pedra do Brazil. Rio de
Janeiro, 1908, p. 33.

’ Dr. White includes Stereosternum in this bed; but this is an error: it is found
in the Rocinha limestone.

46 BULLETIN: MUSEUM OF COMPARATIVE ZOOLOGY.

and Santa Catharina. The best defined horizon is described as that ‘
of the Estrada nova shales with flint concretions. So far as I could
ascertain in 1908 the Glossopteris flora and beds of coal are only well
known on the south in Santa Catharina while deposits with fossil
trees are found on the north in Sado Paulo. The typical tillite beds
are at present best known and apparently most extensively developed
in the state of Parana. On the north in Sao Paulo only scattered
stones occur in the beds. On the south in the Tuberado Valley no
tillite beds are found but waterworn conglomerates occur. Farther
south and west in Rio Grande do Sul some of the reports mention
boulders near the base of ‘the section which may be presumed to
indicate Permian ice-action, but tillite has not been described from
there. Heretofore the Permian has shown only non-marine organic
remains but the discovery of a sparse marine fauna in black shales
near Rio Negro intercalated in the boulder-bearing beds demonstrates
the at least temporary invasion of the district by the Permian sea.
The sections traversed by the writer in Sao Paulo between Itaicy and
Piracicaba, in Parané between Ponta Grossa and Conchas on the
Tibagy, between Lapa and the base of the Serra do Espigao, and in
southeastern Santa Catharina along the Tuberao Valley, seem to
bear out the above statement as to considerable variation in the
character of the sediments from point to point in the lower portion
of the series. The following letter shows the progress of the geological
survey up to 1912.

Ponta Grossa, 13 de Dezembro de 1911.
Meo caro Dr. Woodworth.
Saude.

...-.Hstou aqui prosequindo os meus trabalhos e como sei que as suas
observacoées no Brasil ainda nfo estéo publicadas aproveito esta para lhe dar
algumas indicagdes sobre a idade dos conglomeratos glaciaes.

A primeira vez que encontrei camadas com fosseis marinhas em conglomer-
atos glaciaes foi quando eu estava executando a sondagem do Passinho a
12 km. ao sul de Imbituvo. Estas camadas foram encontradas a partir de
120 metros de profundidade; tinham 45 metros de espessure e nellas achamos
uma pequena lingula, escamas de peixes, e restos indeterminados de brachio-
podoselamellibranchios. A partir de 160 metros de profundidade a sondagem
atravessou até 395 metros exclusivamente camadas argillosas, sem estrati-
ficagio, com pequenos seixos, e algumos boulders de granito. Em 1908, na
sua companhia descobrei a camada fossilifera do Rio Negro. E’ um schisto
negro, ardoriano, com Lingulas, Discina, peixes e restos de esponjas. Em
Outobro deste anno consequi identificar as camadas da sondagem do Passinho.
Ellas apparecem no grotio leste de T. Soares, 45 metros abaixo do arenito

WOODWORTH: GEOLOGICAL EXPEDITION TO BRAZIL AND CHILE. 47

amarello com seizos que ahi conteim uma delgada camada de plantas fosseis.
A seccio e’ a sequinte:

Arenito amarello com delgado leito de carvio . ; ; . 60. metros.
Argilla negra com abundantes restos de insectos . ; io OO
Schisto cinzento com rica flora de Glossopteris . , ae

Schisto cinzento um tanto ardoriano contendo Lingula,
Discina, escamas de peixes, azas de insectos, Chonetes em abun-
dancia, e outros brachiopodos . ‘ . 40.
Camada argillosa com seixos e boulders até o rio aaa Atria 7 G0:
.Parece-me nfo haver duvida de que o conglomerate e’ Carbonifero
medio ao Permiano. Os insectos apparecem nfo sO nas camadas marinhas
como as que se acham logo abaixo do carvdo.
Esta vista modificada podré figurar em seu relatorio.
Abragos, 0 amigo admirador,
Euzebio Paulo de Oliveira.

(Translation by J. B. Woodworth).

Ponta Grossa, 13th December, 1911.
My dear Mr. Woodworth,
Greeting:

A I am here prosecuting my labors and as it may
be that your observations on Brazil have not yet been published I avail myself of
this opportunity to give you some data upon the age of the glacial conglomerates.

The first time that I found beds with marine fossils in the glacial conglomerates was
when I was executing the boring at Passinho 12 kms. south of Imbituvo. These beds
were encountered on going below a depth of 120 meters: they were 45 meters thick
and in them we found a small Lingula, scales of fishes, and remains of undetermined
brachiopods and lamellibranchs. Below 160 meters in depth the boring traversed
down to 395 meters exclusively shaly beds, without stratification, with pebbles and
some boulders of granite. In 1908, in your company, I discovered the fossiliferous
bed on the Rio Negro. It is a black, combustible shale, with Lingula, Discina, fishes
and remains of sponges. In October of this year I found the equivalent of the beds
of the boring at Passinho. They appear in the ravine east of T. Soares, 45 meters
below the yellow sandstones, with pebbles, that here contain a thin bed of fossil
plants. The section is as follows: —

Sandstone, yellow, with thin layer of coal, 50.meters
Shale, black, with abundant remains of insects. 0.60
Shale, ashy, with rich flora of Glossopteris. 5.00

do do somewhat burnable, containing Lingula, Discina, scales of fishes,
wings of insects, Chonetes in abundance, and other brachiopods. 40.00
Argillaceous bed with pebbles and boulders down to the river das Almas 60.00
..It appears to me not to be doubted that the conglomerate is Carboniferous
intermediate to the Permian. The insects appear not only in the marine beds but in
those which occur immediately below the coal.
This view modified may figure in your report.
Cordially yours,
Euzebio Paulo de Oliveira.

Dr. Derby, in 1888, in a letter to Waagen, which was published by
that geologist (Derby, 1888) announced the occurrence in this series
of erratics and likened the deposits to those of the glacial beds of the

48 BULLETIN: MUSEUM OF COMPARATIVE ZOOLOGY.

Permian in India and Australia, a view which every subsequent
geologist who has attentively examined the evidence in the field has
concurred in. (I. C. White, 1908).

The identification by Mr. David White (in I. C. White, 1908, p.
281, etc.) of the Glossopteris flora in the shales of the coal measures of
Santa Catharina has completed the evidence as to the Permian age of
those beds. In the correlation made with the Permian of India and
elsewhere by Mr. White, there is a satisfactory agreement. In south
Brazil as elsewhere the tillite beds occupy a position inferior to the
main occurrences of the flora.

Though mainly of non-marine origin, the Permian sandstones and
shales as well as the glacial beds were deposited at or near sea-level.
Further consideration of the geographical conditions of the period is
deferred to the sections dealing with the conditions of Permian
glaciation.

The Triassic Terrane-— Surmounting the Permian strata of south
Brazil there comes a group of mainly red beds with great sheets of
trap forming its highest members. The basal beds of this series have
afforded the remains of the Triassic reptile Scaphonyx and of fossil
wood. The series is apparently in unconformable relation to the
underlying Permian. According to Dr. Derby the beds on their
northern limits overlap the Permian and rest upon the Pre-Devonian
terrane. The series in most respects recalls the Newark group of
Upper Triassic age in eastern North America.

According to Dr. White’s report (1908, p. 33) the Trias is comprised
of the following members: —

Sao Bento series. Serra Geral eruptives . ; ; 600 M,
900 meters Sao Bento sandstones, red, gray, and
cream colored beds . ; : 200
Rio do Rasto red beds with Scaphonyx
and fossil wood ; : , 100

The reports are not always explicit as to the nature of the trap sheets.
Though apparently generally regarded as lava-flows Dr. White speaks
of examples in southern Santa Catharina and Rio Grande do Sul in
terms indicating the existence of thick sills. In this region, he also
describes the underlying beds as intruded by great irregular dikes as
if they were feeders to some of the overlying trappean masses. In

general the Triassic area forms in south Brazil an elevated plateau |

attaining elevations of 4,000 feet faced where it overlooks the Permian
tract by an escarpment crowned with sheets of basalt. The eastern

—_—— ee

——

Se

WOODWORTH: GEOLOGICAL EXPEDITION TO BRAZIL AND CHILE. 49

border of the Triassic formation thus constituted is deeply notched by
the greater rivers, such as the Iguassi. The escarpments lying like
great loops on the eastern limits of the formation between these river
valleys receive various names such as the Serra Geral, the Serra da
Esperanga, etc. Siemiradzki in a section reproduced by Suess gives
a somewhat different interpretation to the trap mass of the Serra da
Esperan¢a, but the section of the same field credited to Derby is in
harmony with the structure here described as is also the section
traversed by the writer between Rio Negro and Lages in Santa
Catharina. Further consideration of the Triassic rocks and their
bearing on the climatic conditions succeeding the Permian glacial
epoch are reserved for a following chapter.

The Tertiary Deposits Whether or no the Cretaceous deposits
covering the border of the continent south of the Amazon have
representatives now in some of the sands and clays of the coastal
border south of Rio de Janeiro, there are rather recondite reasons for
believing that such deposits may once have flanked the coastal slope
of the Serra do Mar province; if so they were not long after their
deposition worn away. On the upland or planalto, to use the Brazilian
name of the plateau region, no known deposits occur between the
Trias and certain sediments in Sao Paulo which from their fossil
fishes are shown to be of fresh-water origin and Tertiary in age. These
deposits are most extensively developed in the valley of the Parahyba
between the Serra da Mantiqueira and the Serra do Mar and in a
smaller tract underlying the city of Sao Paulo. In these cases, the
beds occupy not well understood depressions in the Pre-Devonian
terrane. In what follows on the topography of the plateau, the
occurrence of these non-marine deposits will be advanced as evidence
of a change of attitude of the region in late Tertiary times.

General Structure of the South Brazilian Permian Area. It remains to
describe the general structure of the region embracing the states from
Sao Paulo southward to Rio Grande do Sul. A glance at the geologi-
cal sketch map, Plate 15, shows that this region has been warped
in Post-Triassic times. A broad synclinal structure with its axis in an
east and west direction occupies southern Santa Catharina and
northern Rio Grande do Sul, causing the Permian and Triassic in turn
to approach the Atlantic coast; and, because of the depression of the
beds towards the synclinal axis, they attain sea-level and in the
case of the Permian descend below that level. On the north of this
structure a complementary anticlinal axis is less well defined in central
eastern Parana. Its position is marked out by the arcuate trend of

50 BULLETIN: MUSEUM OF COMPARATIVE ZOOLOGY.

the Devonian sandstone cuesta in the Serra das Furnas and by the
Serra da Paranapiacaba.

That this gentle broad warping of the Post-Devonian strata is of
Post-Permian date is shown by the sinuous trend of the outcrops
of the Permian terrane. That it is Post-Triassic is also indicated
by the down bending of the trappean sheets in the syncline on the
south. Precise evidence of a local nature is apparently lacking to
demonstrate how long after the completion of the local Triassic
section the deformation was produced. ‘The dislocation of the Triassic
Newark beds of eastern North America in presumably early Jurassic
time would lead us to admit that the deformation of south Brazil
may have taken place at this time. For topographical reasons set
forth in the account of the physiography of this region (p. 99) it
seems that this type of deformation took place in Pre-Cretaceous times.
A Jurassic date for the deformation thus appears the most probable one.

Summary of Geological History of the south Brazilian Plateau.—
From this brief account of the geology of the south Brazilian plateau
it appears that, long prior to the Devonian period, the region passed
through a series of changes registered in the occurrence of igneous
and metamorphosed sedimentary rocks as yet little understood. At
least one series of sediments now slates and limestones involved in the
complex may be the equivalents of the Ordovician or Cambrian strata
but fossils are wanting to prove the age of these beds. The basal
Devonian sandstones resting on the eroded edges of these meta-
morphosed and intrusive rocks can only be interpreted as evidence
of land conditions in the epoch preceding the Devonian and hence
in the Silurian (Upper Silurian of Murchison). The Devonian sea
transgressed the area and continued depositing sediments through the
middle of that period. Whether deposition took place in the Upper
Devonian is not now clear. The Carboniferous period was so far as
the records go one of land denudation in which most of the Devonian
was swept away. The extent of this denudation is such as to render
it probable that higher Devonian beds originally overlay the Middle
Devonian shales of Parana.

True Carboniferous deposits appear to be wanting in the area, such
strata as may have been referred to this epoch belong rather to the
basal Permian.

The Permian is characterised by a series of mainly non-marine
formations laid down at or near sea-level with locally well-developed
beds of tillite indicative of a glacial period. Elements of the Gloss-
opteris flora have been found in Santa Catharina in shales which carry

:
7
A
}
=

WOODWORTH: GEOLOGICAL EXPEDITION TO BRAZIL AND CHILE. 51

thin beds of coal. The lower members of the Permian section display
no marked persistence north and south but towards the top of the
section more uniform conditions of deposition appear to have pre-
vailed. The horizon of the Iraty black shales near the top appears
to be fairly continuous.

The so-called Trias characterised by red beds and contemporaneous
basaltic outflows and intrusions marked a great change of geography
and climate. Arid conditions succeeded to the moist climate of the
Permian but stream action appears to be demanded for the deposition
of the thick sandstone beds devoid of marine fossils. Presumably at
the close of the Triassic episode of deposition and at some stage in the
Jurassic period the region underwent gentle warping producing the
present structure of the plateau.

The Jurassic and Cretaceous periods are in this region without
existing remnants of deposits. If the elevated Cretaceous marine
beds of the north originally extended so far south, it is to be presumed
that they were preceded by an epoch of baselevelling and in the
following Tertiary time by the uplift of the district in such a manner
as to give rise to the Serra do Mar slope to the sea-floor. Erosion of
the plateau surface and of the short slope to the sea has worked out
in accordance with the specific resistances of the several terranes the
existing topography of the region, with a late Tertiary interruption of
the cycle of erosion indicated by small basins of non-marine Tertiary
beds in Sao Paulo. Over this surface a mantle of unconsolidated
materials produced by weathering with more or less transportation
has accumulated during Pleistocene and Recent times.

The annexed cross-section of south Brazil, after Derby, exhibits the
essential features of the structure of the plateau in the latitudes where
the Devonian strata occur.

Serra da uaapann ¢

goers ee Ak

. Ct We
Ports Grosser et ASerra doMar
C ; €

Fia. 9.— Geological cross section of south Brazil (after Derby).

For a bibliography of Brazilian geology, the catalogue of publica-
tions prepared by Prof. J. C. Branner (1909) should be consulted.

52 BULLETIN: MUSEUM OF COMPARATIVE ZOOLOGY.

V. PERMIAN GLACIAL DEPOSITS OF SOUTH BRAZIL.

“Tw the next great section of the earth’s crust, the Permian period, we have
an almost world-wide extension of glacial waste * * * Even in the Southern
Hemisphere we have what seems to be conclusive proof that the glaciers
during this age operated in regions nearer the equator than they did during
the last glacial period.”’

SHALER in Shaler and Davis. Glaciers, Boston: 1881, p. 96.

‘‘ Bvidence is slowly accumulating which serves to show that glacial periods
of greater or less importance have been of frequent occurrence at all stages in
the history of the earth of which we have a distinct record.”’

N.S. Saater. Outlines of the Earth’s History, New York:
1899, p. 247.

The foregoing statements concerning the boulder-bearing Permian
beds of south Brazil show the state of the inquiry concerning the
origin of these deposits as late as 1908. Dr. Derby, Professor Branner,
and Dr. I. C. White were essentially in agreement in regarding as
highly probable the glacial origin of the boulders, but striated rock
surfaces either of the bed rock or as transported erratics were wanting.

The boulder-bearing beds of the Permian in south Brazil are far
from presenting a persistent parallelism of strata from point to point
along their outcrop. In the state of Sado Paulo the typical tillite is
seemingly wanting except near the southern border, the evidence of
ice-action being limited largely to argillaceous sandstones carrying
occasional stones and boulders. The typical tillites crop out on the
northern border of the state of Parana and are exposed at what appears
to be more than one horizon as far south as Santa Catharina. In
eastern Santa Catharina in the section along the Rio Tuberao there are
no surface exposures of tillite, but waterworn conglomerates occur
at a low horizon in the section apparently representing as at Ponta
Grossa in Parana the glacial episode.

The following account of the sections studied by myself begins with
the beds on the north in the state of Sao Paulo, and includes the
following sections: —

a. The section from Itaicy to Piracicaba in Sao Paulo.

b. The section on the Rio Jaguaricatu in northern Parana.

ce. The section from Ponta Grossa to Conchas, Paranda.

d. Exposures between Ponta Grossa and Serrinha, Paranda.

e. The section from Serrinha via Lapa to Rio Negro, Parana.

f. The Orleans-Minas section in the Tuberao Valley, Sta. Catharina.

‘orneg Ovg Jo 04%4s Jo ued asojsvoyjnos Jo dew yoJoHs —‘OT 3) a I

53

—_——— et ee

Se a ae

5 WV EBD) 220
VOMWILIGVLIS

AIIV LI

a

WOODWORTH: GEOLOGICAL EXPEDITION TO BRAZIL AND CHILE.

54 BULLETIN: MUSEUM OF COMPARATIVE ZOOLOGY.

The section from Itaicy
to Piracicaba:— A _ recon-
naissance was made of the
section from the granite base
at Itaicy along the railway
upwards in the series to
Piracicaba. The following
rough notes serve to bring
out the general character of
the strata below the reptile-
bearing Permian horizon at
the latter place and have
been made the basis of the
accompanying section and
sketch map. The station
numbers are those of the
kilometer posts along the
curving and locally recurved
line of the railway. Notes
and posts are given in the ©
order in which they came.
Itaicy has an altitude of
1,200 M. (1,870 ft.) accord-
ing to local contour of the
map.

The projection of the strata

8 Lo SHALES
so SANDY TILLITE clit lcnerty SHALE
7 snares

30 SHALES

Y
Y

4p

;
PY
<~
w
yw
4
«ao
<
a
a
wd
—
Ej
[=
-
Qo
5
=

S (sates
2A J sanoy tute

151. Granite boulders; in
place (?).

153. On sediments. No
boulders noted.
One stone in yellowish
topping.
Cut with pebbles;
shale over sandstone.

beyond the line of route is subject to correction for topographic relief.

154... Ras
Reddish massive sandy
shale (now clay).
Pebbly sandstone; no
striae seen on peb-
bles.

Fic. 11.— Sketch map of Permian along railway line between Itaicy and Piracicaba, Sio Paulo.

155.

WOODWORTH: GEOLOGICAL EXPEDITION TO BRAZIL AND CHILE. 99)

156. near
Indaiatuba (Indaia, wild fruit; tuba, many).

157. Canga (limonitic crusts).

Bridge.
PES, . Leis
Reddish sandstone.
REOe rx:
Shales.
Crew house. Boulder on surface.
Massive checkered clays (shales).
Surface covered with cobbles.
ie
Small overturned fold south side of track in strongly folded
shales. Beds dip N. W. Dark shale bed.
161.

Cut showing yellow sandstones above last.
162. Cut, with cobblestones in shales.
Cut with sandstones.
do.
fie 2217
Deep cut, sandstones without pebbles.
ee ae
Boulder on surface.
165. Red cut.
166. Long cut; red sandstones without pebbles.
167 to 169; no exposures.
Red beds with pebbles.
170-171. Folded yellow tillite (?) bed.
Shaly sands now clayey.

ees eau:
Red shallow cut.
173. do.
do.
Bees.
Bouldery shales; 2 ft. boulder of quartzite. Also diabase sill?
Small fault running east-west; downthrown to west.
Black band bed. Carbon or manganese?
Beas . Paks
Shale.
176.

Black shale.

56

BULLETIN: MUSEUM OF COMPARATIVE ZOOLOGY.

177 to 178; no exposures.

172.
180.
181.
182.

183.
184.

185.
186.

187.

188.

189.

190.

191.

193.

194.
195.
196.

Boulder near Elias Fausto; ca. 2 ft. size.

Shales.

Tillite or sandstone; boulders loose on surface.
Long cut in sandstone.
Yellow clayey sandstone to

Catan diabase sill.
Tillite sand on top of diabase.
do.

Diabase sill to Tuburcio Pacheco Station.
Yellow beds with stones.

Tillite?

Rio Capivary.

Dark shales with embedded stones.

Tillite section, in which I find large angular piece with faint
striae; rock a compact white sandstone; like the bed under
the Jaguaricatu beds.

Waterworn pebbles in a layer overlaid by reddish sandstone but
these beds are possibly Post-Palaeozoic. S. E. of Faz. Barreiro.

One or two boulders seen on surface.

Black shale; greenish black, containing in cut at least one granite
boulder 15 to 18 inches (38 to 45 em.) long; continue to cross-
roads; there overlying

Sandstones.

Yellow fine sandstones (loessite?)

Similar to last; faulted and disturbed; small overthrust of sandy
loessite upon shales.

Massive beds of sandstone.
Rio Capiavary near to

Near to Capivary Station.

WOODWORTH: GEOLOGICAL EXPEDITION TO BRAZIL AND CHILE. 57

197;

198.
199,

200.

201.

202.

203.

204.

205.

206.

207.

208.

209.

210.

Loessite with boulder; sandstones with small vertical fault;
downthrow to east.

Good cut with two small overthrust faults; overthrust eastward.

Yellowish sandy shale.

Tillite; fine grained.

Begitr factory at Villa Rafford.
In cut beyond station, tillite with boulder of conglomerate
(Plate 20). Pebbles in boulder mainly quartzite.

Trap near to

Trap.

Tillite.

Sandstones.

Cut.

Boulders and stones in sandy beds.

Shales; somewhat crumpled.

Dark shales to

Sandstones.

Shales; with concretions (?).

Tillite, with stones.

Long side cut here with sandstones like moulding sands; con-
taining boulders of granite and quartzite.

Till-like sandstone.

Red do.
Beds with pebbles and small blocks.

Tillite? without pebbles.
Beds of characteristic yellow, and with bale structure; stones
rare and small. Beds flat, to near

Same as last.
Chert pebbles appear in surface deposit.
Pebbly sandstones in cut.

Sandstones in long cut.

do.

58

211.

212.

213.
204.
215.
216.
217.

218.

219.

220.
221.

222.

223.
224.
225.

226.

227.

BULLETIN: MUSEUM OF COMPARATIVE ZOOLOGY.

Same to Mombuca Station.
Shaly bed.

Pebbly sandstone (tillite?) near
do.

do.

Two cuts of same.

do., in good cut. No stones seen.
Glacial sandstones.

do. fine grained.

Fine mealy sandstones.

do.

do.

Steep dip locally to northeast?
Shales; thin?

Red sandstones over last.
Boulders on surface.

Purple shales.

Boulders on surface.
Red cut. Trap?

Brownish red rock in cut; trap.
do.

Cut in sands red to brown in color.
Terra roxa; deep cut.

do.

Trap near

Terra roxa cut.

do.

Sandstones.

Stratified beds, glacial sandstones and shales.
Rio das Pedras Station.

Purplish to reddish alternating sandy layers in cut.
Shales with sand band.

Shales.

do. greyish.

do. dark colored.

WOODWORTH: GEOLOGICAL EXPEDITION TO BRAZIL AND CHILE. 959

228 ! Beatis

do.

do. with thin sandstone bands.
BAO «eee.

do. purple.

Sandstones?
ys |

Red cut; bright red; terra roxa?
yO

do.

Shales, purple, under red bed.

Cherty layer, vesicular, from 1 to 1.5 inches thick; in shales near
232.

233. do.
234. do. terra roxa? mixed with sand?
235. do.
236. do. near

Purple shale (diabase sill above?).
237-238.

Red cut.

Chert beds.

Stratified purple shales; dip northwest.
2

Sandstones and purple shales, thick, interlaminated.
240-241.

Piracicaba Station.

Without deep borings it: is hardly possible to obtain an accurate
measurement of the thickness of the strata in this section from Itaicy
to Piracicaba. Any estimate is increased a slight amount by the
small overthrust faults, for which a deduction should be made, though
in round numbers this is possibly a negligible quantity. There is also
the question of the uniformity of the dip of the beds. The small
contortions associated with the little faults demonstrate the existence
of some Post-Permian disturbance of the beds, mainly an eastward
overthrust. From kilometer post 189 to the cross-road between that
post and 190 kilometers, there is a bed of greenish black shale which
appears to overlie as in a shallow synclinal fold the glacial sandstones
on the respectively south-east and north-west of the exposure. This
point is 8.5 kilometers north-west down the supposed dip from the

60 BULLETIN: MUSEUM OF COMPARATIVE ZOOLOGY.

eastern limits of the Capivary sandstone group. If this interpretation
of the structure be correct and the shale bed does not dip into the sec-
tion immediately westward of this exposure, the question is raised
whether it is not a southeastern outlier of the shales which certainly
overlie the group westward from Rio das Pedras. In this case the
sandstone beds northwestward from Elias Fausto to this shale outcrop
instead of being 1,400 feet thick would have a thickness not greater
than that of the second belt above the base, 7. e. about 600 feet.
This would reduce the total thickness of the section measured from
4,000 ft. to 3,200 feet.

As for the maximum thickness, neglecting the above possible causes
of diminution in the estimated thickness, there is the presumption that
the sill of diabase which crops out between the 230 and 231 kilometer
posts in the form of a bright red terra roxa zone represents the trap
sheet which forms the falls in the river at Piracicaba. The upper
contact of this layer on the southeast is near the 575 M. contour; on
the northwest down the dip at the river it is somewhere near 525 M.,
giving a drop of 50 M. in 7.5 kilometers or thirty-five feet to the mile,
which is equivalent to a little more than a dip of 1° to the northwest,
and since the strata overlying the sill are presumably parallel to the
plane of the sill, we have in this part of the section an estimate of the
dip. Again the thin bed of shales at and southeast of Elias Fausto
Station have a breadth of outcrop of four kilometers (2.48 miles, say
2.5 miles). Their base on the east along the line of the railway is at
about the 600 M. line; their top at the northwest at 575 M. giving
a fall of the surface of twenty-five M. in four kilometers or eighty-two
feet in 2.5 miles, or thirty-two feet to the mile. Assuming the shale
bed to be at least ten feet thick, we should have dip of thirty-six feet
to the mile, which considering the approximations on which the esti-
mate is based is in close accordance with the estimate of the dip made
for the beds immediately southeast from Piracicaba.

On the assumption of continuity of bedding and dip from the base at
Itaicy to Piracicaba, a dip of 1° would give a minimum thickness of
1,145 M. or 3,700 feet for the total thickness from the base to the
Stereosternum beds.

Rough calculations of the mainly sandstone group at the base about
Imbaiatuba with a breadth of outcrop of 5.76 M. or 4.19 miles gives an
estimate of 436 feet for their thickness, with a maximum of 700 when
scaled to the maximum estimate.

The succeeding shale bed 153 feet as a minimum, 226 as a maximum.

The second group of sandstones may be estimated as having at the
minimum a thickness of 400 feet, a maximum of 600 feet.

WOODWORTH: GEOLOGICAL EXPEDITION TO BRAZIL AND CHILE.

The shales at Elias Fausto Station, by the minimum estimate

fifteen feet, by the maximum twenty-two.

The succeeding group composed mainly of sandstones including al]

the beds about Capivary, by the minimum estimate
based on the hypothesis of duplication of exposure
through gentle folding 600 feet, or by the maximum
1,400 feet.

From the top of these beds to the shales with
Stereosternum at Piracicaba 700 feet by the minimum,
1,000 feet by the maximum method.

These minimum estimates give a thickness of 2,343
feet, or about 714 meters.

The Bofete-well record in Sao Paulo cited by Dr.
I. C. White gave forty-eight M. for the Iraty Black
shale and 837 feet below that group to the bottom of
the well.

The accompanying section (Fig. 12) represents the
succession of strata between Itaicy and Piracicaba
interpreted on the probable solution that the shales
east of the Rio das Pedras are in a synclinal outlier
of the beds about Piracicaba.

In this section there appears no bed equivalent
in lithological character to the boulder tillites of the
Jaguaricatu section in northern Parand. In the
section along the line of the railway there are scattered
stones and small blocks embedded indifferently in the

peculiar sandstones and shales. This
indiscriminate distribution of fre-
quently striated stones in beds of
sandstone and shale is presumptive
evidence of the action of floating ice
particularly in the case of the shales.
The weathered state of the sandstones
in the railway cuts makes precise
comparisons with the deep} blue
tillite beds farther south difficult and
unsatisfactory.

Fig. 12.— Section of Permian from

granite floor at Itaicy to Pira-
cicaba. 1. Pre-Devonian granite.
2. Sandy tillite about Imbaiatuba.
3. Shales. 4. Sandy tillite. 5.
Shales at Estacaio Elias Fausto.
6. Sandy tillite east and west of
Capivary. 7. Shales west of Rio
das Pedras. C. Cherty layers.
8. Shales and Stereosternum beds
about Piracicaba. D. Diahase sills.

So far in this section of the Permian no marine fossils have been
found below the Stereosternum beds if indeed that horizon be marine.
On the contrary, notwithstanding the trunks of trees discovered by
Dr. Derby somewhere near the middle of the section, no fossils appear

62 BULLETIN: MUSEUM OF COMPARATIVE ZOOLOGY.

to have been found in place; but the tree trunks cannot be far out of
position, and the general character of the section points to the sub-
aerial accumulation of certain of the sandstones in the manner of
outwash plains such as now accumulate in front of glaciers or in
similar situations permitting the growth and fall of trees between
times of aggradation by débris-laden streams. Only a detailed survey
of the region would suffice to determine more definitely the relations
to sea-level and the existence of lenses of tillite in the section.

Tillite Beds on the Rio Jaguaricatu in Paranda. Sengéns is a station
on the newly constructed line of railway from Itararé to Jaguariahyva
in northeastern Parand, on the south bank of the Rio Jaguaricatu.
In the railway cuts which were fresh in 1908 good sections of the
glacial boulder-clays or tillite were exposed for several kilometers in
either direction. The Station is near the 228 kilometer post. The
following notes pertain to the exposures along the curving course of
the railroad which follows down the right bank of the river towards
Sao Pedro de Itararé. Near post 234 a well-striated sandstone pebble
was found dislodged from a sandy conglomerate or tillite at this
locality. This pebble is shown in Plate 21 to the left of the
hammers. The striated surface is well flattened and the edges
battered off. It displays no zigzag striae and may be a fragment of
the old glaciated floor. Somewhat farther to the northeast between
posts 234 and 235 the boulder-clay type of tillite is well shown, as in
Plate 21, there containing small boulders of granite.

Between posts 235 and 236 apparently near the top of the local
bed the view shown in Plate 22 was taken. At this point in a
fresh railway cut there is exposed an angular slab-like block of a fine
grained white sandstone resembling novaculite, nine feet (three
meters) long and about twenty-one inches thick, the third dimension
not being ascertainable, the largest angular block I myself saw in any
section in Brazil. This rock is practically identical, lithologically,
with a white sandstone layer which occurs lower in the section at a
locality along the railway where a fault brings the beds towards the
surface. (Fig. 13). .

This fault extends in a northwest direction and the downthrow
appears to be towards the west. The sandstone in question is here
overlaid by pebble beds. Between these eastern exposures and
Sengéns there are coarse sandstones containing at least one pebble
bed with a greater proportion of granitic pebbles than are found in the
more typical tillite of the district. From this pebble bed I collected
three stones carrying faint striae.

WOODWORTH: GEOLOGICAL EXPEDITION TO BRAZIL AND CHILE. 63

West from Sengéns on the north of the Rio Jaguaricatu in one of
the deep cuts along the ascending grade striated stones were first
found in Brazil. The bed exposed at this locality is a typical tillite,
containing large subrounded masses of a whitish rock resembling the

ja

White

Sand Stone

Fig. 13.— Section along the railway showing white novaculitic sandstone
bed under tillite bed.

white, fine, silicious sandstone from near the base of the Permian
section.

This occurrence of the fine white sandstone in the boulder-beds from
Sao Paulo southward in close association with the preglacial stratum
of the same character is strong evidence of the eastern derivation
of the erratics.

Conglomerates at Ponta Grossa: — Overlying the trap-invaded fossili-
ferous Devonian shales which underlie the “big point” between

Fic. 14.— Section of Permian conglomerate beds in vicinity of Ponta
Grossa, Parané.

streams on which the city of Ponta Grossa stands, there are to be
seen on the northwest of the town near the base of the Permo-Car-
boniferous series, at least two beds of waterworn conglomerate (Fig.
14) interpolated in a thick sandstone series. These waterworn
pebbles present in themselves no evidence of glacial origin or trans-
portation. The lower pebble bed, from four to five feet thick, is made
up of subangular to well rounded, waterworn, light-colored sandstone
pebbles from .5 inch to 6 inches in diameter. Rarely larger cobble-

64 BULLETIN: MUSEUM OF COMPARATIVE ZOOLOGY.

stones occur at the top of the bed embedded in the overlying sandstone.
No glacial striae or facets were seen on any of the exposed pebbles.
Besides the sandstone pebbles, I saw one of micaceous quartzite, a
crumpled phyllite, a whitish chert, a red argillaceous sandstone, a
decomposed granite with reddish feldspar, a silicious schist, im toto
a variety of rocks indicating the derivation of at least a small portion
of the débris from the Pre-Devonian terrane. The pebbles could
hardly have been transported far by water action alone, because the
sandstones pebbles are of a sort which do not wear well in any stream
journey. I estimate their journey as stream pebbles to have been
tens of miles rather than hundreds of miles. ‘The assemblage of these
pebbles in a well-stratified bed between layers of a coarse-grained
sandstone leaves no doubt of water action. A few feet of sandstone
separates this bed from the one above, in which the sandstones peb-
bles again formed the predominant constituents but were noticeably
more rounded than the crystalline pebbles. Dr. I. C. White men-
tions in his Report a bed of boulders at Ponta Grossa and a deposit
containing fossil wood, but I saw none, nor were such deposits known
to the Geological Survey staff at the time of my visit.

Tillite Bed at Conchas: — Conchas lies on the north side of the Rio
Tibagy about four leagues west from Ponta Grossa. On the south

_ of the village a small quarry
7 was opened some years ago
in a grayish somewhat in-
durated stony clay bed, a
boulder-clay phase of the
tillite beds. The scattered
pebbles consist of silicious
rocks and rarely a granitic
Fic. 15.— Fracture of the tillite bed at Conchas. Pebble. The bed fractures
with a giant ball structure
(Fig. 15). No striations were seen on the pebbles. The mode of
occurrence of the pebbles seems best explained by dropping from
floating ice and probably the clay with its sand grains of irregular size
originated in the same manner. The rock when exposed to the weather
breaks down by a process of checking and the opening of ragged
fractures into smaller and smaller blocks so that it is valueless for
building stone.

This bed overlies the sandstones with waterworn pebbles at Ponta
Grossa, and recalls in its lithological characters the beds on the south
of the Rio Jaguaricatu east of Sengéns and also the beds southeast
of Rio Negro.

WOODWORTH: GEOLOGICAL EXPEDITION TO BRAZIL AND CHILE. 695

Signor Cicero de Campos (1908, p. 3) of the Brazilian Geological
Service states that between the port of the Indias and Salto-Maua in
Parana there outcrops a little below the falls on the left bank of the
Rio Tibagy a vellow sandstone with enormous blocks of pink porphyry
granite and of porphyry. The fall is caused by basalt. Campos
refers the strata of the hilltops to the Carboniferous (Permian?).

Near the sixty kilometer post on the railway from Ponta Grossa to
Porto da Unido I noted an exposure of the compact yellowish beds
which appear to be the weathered phase of the tillite as at Conchas.
A large boulder was also seen in a cut a short distance north of this
locality.

Boulders and Pebble Beds near Palmeira.— On the divide south of
Palmeira yellowish compact beds with blackened joints exhibit a few
pebbles north of the 133 km. post. These beds continue in the railway
cuts to and somewhat beyond the 130 km. post, then laminated yellow
beds come in and a boulder about three feet (1 meter) in diameter
appears onthe surface west of the track near a house north of 128 km.
post, about eleven kms. north of Restinga Secca Station. Large
blocks appear in a cut between 122 and 121 km. posts, north of branch
road to Amazonas. Beyond 105 km. post, pink pebbly beds are
intersected by the railway in the long descent, with tilted pebble
bearing beds near 103 km. post. Tilted beds also occur near the
ninety-nine km. post, possibly faulted beds. Tamandué4 Station is
at the ninety-three km. post. Beyond this Station pebbles occur in
sandstone at eighty-five km. post; again at eighty-three km. post.
At seventy-eight km. post there is a good pebble bed and north of .
seventy-six kms. gravels appear in a cut.

The Glacial Conglomerate at Serrinha.— At Serrinha in the gorge of
the Iguassti, where that river has cut deeply into the sandstones at
the southern limits of the sandstone cuesta locally known as the
Serrinha (little serra), there is a good exposure of beds, the general
characters of which are shown in the annexed unmeasured section,
(Fig. 16). .

There is here east and west of the railroad station an exposure of
small pebbled conglomerate rising about twenty-five feet above the
level of the track. Quartz pebbles either angular or well rounded occur
in this bed, with occasional pebbles of gneiss and possibly granite,
together with quadrangular fragments of a now reddish friable sand-
stone, presumably a rock derived from underlying beds. East of the
station between the 69th and 70th kilometer posts, Dr. Euzebio
Oliveira found in the conglomerate beneath the sandstones a dis-

66 BULLETIN: MUSEUM OF COMPARATIVE ZOOLOGY.

tinctly striated pebble, a compound quartz-bearing rock, with one
well-rounded side and a surface of more recent fracture. The
rounded surface bore striae. The pebbles in this conglomerate are

Compact blue pess]
oe a :
Louassik Plats

Fria. 16.— Section of the conglomerate and overlying sandstones at Serrinha,
Paran4, in the gorge of the Iguasst.

rather closely scattered through a matrix of abundant sandstone of an
argillaceous character, suggesting the deposition of the mass by float-
ing ice.

The whitish sandstone overlying this conglomerate is so like that
forming the base of the Devonian northward in the Serrinha ridge
that the question is raised whether this conglomerate belongs to the
Permo-Carboniferous series or to the Devonian. ‘There is no appear-
ance of faulting in the neighboring district which I discerned. The
Devonian shales which separate the Devonian sandstone from the
basal Permian at Ponta Grossa on the north have at this distance on
the south disappeared, and it is possible that the Sandstone has also;
the Permian by overlap coming close down to if not in contact with
the Pre-Devonian east and north of Serrinha, the local resemblance
of the white sandstone to the typical Devonian being assignable to a
redeposition of those sandstones. The determination of the Devonian
age of the lower beds at Serrinha, should it be made, would introduce
another glacial epoch in south Brazil parallel to the case in South
Africa. In this report I have preferred to follow the tentative opinion
of Dr. Derby that the Serrinha section is Carboniferous (Permian).
This view involves at least two advances of the ice in Permian time.

The Section at Lapa.— Lapa is a station on the railway south of Ser-
rinha. Along the line of railway southward from Serrinha after
surmounting and crossing the sandstones on the south bank of the
Iguassti the tillite beds are traversed by the train. Near and north of
Pivary the campo surface is strewn with rounded stones. About

WOODWORTH: GEOLOGICAL EXPEDITION TO BRAZIL AND CHILE. 67

a mile east of Lapa the Morro do Monge (Monk’s Hill) rises, the
most northern of a series of sandstone ridges forming a line of westward
facing escarpment, broken down at intervals by cross-valleys pre-
sumably of structural origin. This sandstone overlies the tillite beds,
and is traceable in these ridges southward to the Rio Negro, south of
which it plunges beneath higher beds, in which, as will be shown,
isolated boulders are not wanting. At the base of the Morro do
Monge sandstone, there is a ferruginously spotted variety, which is
locally quarried. At the surface it weathers into fist-sized quartzitic
knobs. At the top of the tillite series and beneath this sandstone at
the northeast angle of the Morro do Monge, where the road crosses,
there is a conglomerate bed about five feet (13 meters) thick with
rounded pebbles in a sandy matrix. This is overlain by one foot of
sandstone, then by a bed with scattered pebbles and sandstones. In
this series I found a polished quartzose pebble with faint striae.
Further south along a private road on the west face of the Morro and
at about the same horizon as the conglomerate beneath the sandstones
I found a loose pebble of striated sandstone, a specimen of doubtful
indications. Striated pebbles in this locality at this horizon are rare.

Between Lapa and Rio Negro.— South of Lapa the train runs over
the yellow weathered surface of the tillite beds. In a cut between
kilometer posts 51 and 52 one or more thin sills of diabase appear.
South of the Rio Vargem the train ascends one of the sandstone
outliers which is deeply dissected by streams. Between the 73rd and
74th kilometer posts on the descent to the valley of the Rio Negro,
just under the hard sandstone bed, in a cut a well-defined boulder-bed
appears. Some of the rounded masses have a diameter of three feet
(1 meter). Occasional cobbles and yellow compact weathered beds
at lower levels farther south indicate the continuance of the tillite
sands in this direction. |

Localities near Rio Negro: — About two miles north of the city of
Rio Negro on the northwest bank of a stream which falls into the Rio
Negro there is a high bluff of laminated shales with a pronounced
concretionary structure with very small intersecting joints. These
shales carry scattered pebbles at various levels, the pebbles ranging
up to six inches in diameter. One of these larger pebbles showed a.
smoothened sole or side and faint striae. These shales lie above the
sandstone at the bridge over the Rio Negro at the city of that name.
Some of the pebbles are a hard sandstone; mica schist also occurs.
Most of the pebbles were partly rounded, resembling kame and esker
pebbles rather than typical stream pebbles. A flattish broken edged
piece of sandstone also showed weak striae of glacial character.

68 BULLETIN: MUSEUM OF COMPARATIVE ZOOLOGY.

Sandy beds with boulders succeed to the shales for a few feet of
thickness. Shales come in again above this level, rarely with striated
pebbles.

Fic. 17.— Section of Permian shales, boulder-beds, and sandstones at Rio
Negro.

The section, at the bridge (Fig. 17) above referred to, displays
sandstones at water-level somewhat cross-bedded and carrying
scattered pebbles and small blocks. I measured one granitic frag-
ment 16 inches long.

On the south side of the Rio Negro along the road going eastward
towards Sao Bento yellow tillite beds with intercalated dark shales
contain many granitic boulders. A cobble eight inches in diameter
showed good striation. The boulders range from two to three feet

ce = _ =—

Fic. 18.— Section of fossiliferous Permian marine shales between boulder-
beds near Rio Negro.

in diameter, all in the yellow sandy tillite, with very few small pebbles,
arguing for floating ice. A few kilometers along this road eastward
from the bridge where a small stream has cut a well-defined valley in
its descent to the Rio Negro, Dr. Oliveira found a marine fauna in
black shales intercalated between two boulder-beds of the tillite series
(Fig. 18). I refer to this bed and its apparent significance elsewhere.

A few kilometers farther east by south along this road a new cut
near the small Rio da Vida nova displayed the shales and tillite beds
(Plates 23 and 24), at horizons closely corresponding to the beds above

WOODWORTH: GEOLOGICAL EXPEDITION TO BRAZIL AND CHILE. 69

the marine fossiliferous zone. In this locality the shales also carry
boulders and stones. That in the view being a boulder of gneiss
about twenty inches in the diameter on the exposed section. Below
the middle of the part of the bed shown in the plate there is a thin,
well-defined layer one and a half inches thick of the yellow material
which is called tillite in this paper. The layer consists of sand grains
and earth particles with coarse angular grains ranging to the size of
very small pebbles, and by its mode of occurrence is strongly sug-
gestive of deposition, like that of the stones included in the shales,
by floating ice, probably the manner in which the thicker beds of the
same character were laid down in this section. The shales part with
a thin splintery fracture subparallel to the bedding.

The overlying tillite bed is composed mainly of fine earthy material
with coarse sand grains, and scattered pebbles; some of these latter
display glacial striae. Among the types of rock was noticed one
example of a coarse-grained diabase. This bed is about eight feet
(2.5 meters) thick, and is remarkable for the almost perfect develop-
ment of the bale structure so characteristic of some varieties of
igneous rock. No trace of stratification was detected within the
limits of the stratum.

About two Brazilian leagues from Rio Negro and further to the
south and east along this same road, the Ribeira das Rutes, a small
tributary of the south bank of the Rio Negro, has cut a gorge with a
fall at its head in the blue unweathered tillite. This rock contains
rather abundant stones in a matrix of clay beset with small subangular
fragments of rock and coarse sand grains.

North of the fossil locality above mentioned and lower down along
the course of the Rio da Vida Nova the river road intersects a water-
worn conglomerate from six inches to two feet thick with pebbles of
sandstone and rocks resembling the formation in which it is inter-
calated. It is overlain and underlain by sandstones of the same
yellowish hue as the tillite beds, resembling a solidified loess.

Section from Rio Negro Southward to the Top of the Series — South of
Rio Negro and upward in the series, sandstones and shales are crossed
by the Lages road as far as the Rio Laurengo, at 1,500 mule paces
north of which I saw a large granite block four feet long lying on the
outcrop of bluish shales. On the south bank of the Rio Laurenco a
thick sandstone formation comes in. Scattered pebbles and a few
boulders are seen along the same road farther south towards Sepultura.
Beyond this point southward and higher in the section no good
evidence of embedded pebbles was seen.

70 BULLETIN: MUSEUM OF COMPARATIVE ZOOLOGY.

The Tuberdo Valley Section— The Permian beds at Orleans and
near Minas in the upper valley of the Rio Tuberao are better known
than any section of the Palaeozoic north of Rio Grande do Sul. Not
only have the coal beds in that hydrographic basin been reported
upon at different times by a number of foreign mining engineers but

AZ? 1G 2ias” (Gree wieP f
roa “A as

CLES:

Triassic with Permian;with Pre-Devonian;
lava flows. thin coal beds, mainly granite,

Fic. 19.— Sketch map of the Minas-Orleans district in the Tuberao Valley,
Santa Catharina.

the region has been very thoroughly explored by Brazilian geologists
and recently the staff of the Coal Commission under the directorship
of Dr. I. C. White has made a still more detailed report upon the
geological structure and the nature of the strata of this part of the
field. According to the terminology of Dr. White’s Report this section
from the granite basement to the top of the Serra Geral trap plateau
comprises the following formations named in descending sequence: —

Triassic
Sao Bento series
Serra Geral eruptives
Sao Bento sandstone
Rio do Rasto red beds

WOODWORTH: GEOLOGICAL EXPEDITION TO BRAZIL AND CHILE. 71

Permian

Passa dois series
Rocinha limestone
Estrada nova beds
Iraty black shale

Rio Tuberao series
Rio Bonito beds
Orleans conglomerate

In this report we are concerned mainly with the beds grouped in
the above section under the head of the Orleans conglomerate at the
base of the Permian section in southeastern Santa Catharina.

Dr. White (1908, p. 51) describes the Orleans conglomerate as
follows: —

‘Resting upon these lower sandstones and shales often in apparent
conformity, we find a coarse conglomerate which is well exposed in
the town of Orleans, Santa Catharina, from which locality it has been
named. It contains boulders of granite, quartzite, and other hard
rocks, some of which are 20 to 25 centimeters in diameter. The same
formation is frequently visible along the Rio Tuberao between Minas
and 2 kilometers below. The bore hole put down near Minas station
began near the top of this rock and passed through the same at a
depth of 5.35 meters. In Rio Grande do Sul large granite blocks are
frequently found at this horizon, as well as at many points in Parand
and the adjoining region of Santa Catharina, where several localities,
near Rio Negro, 10 kilometers from any outcrop of granite, exhibit

‘granite boulders in vast numbers up to 3 meters in diameter, all
embedded in a fine and apparently unstratified gray muddy sediment.
A very coarse deposit with large rounded boulders of granite, quartzite,
sandstone, silicified wood, etc., may also be seen resting unconforma-
bly upon Devonian shales at Ponta Grossa, and other localities in
that region. This deposit appears to correspond closely to the Dwyka
conglomerate of South Africa, and most probably, like it, is of glacial
origin, although no scratches were observed upon the boulders in
question.” Between the conglomerate and the base there usually
intervenes a few meters of sandstones and shales. In the boring at
Minas, these beds together with the conglomerate are said to have a
thickness of twenty-seven meters. On the road from Lages to
Florianopolis farther north the total thickness is from 150 to, 160
meters (White, loc. cit., p. 49). At Xarqueadas, in Rio Grande do
Sul near the right bank of the Rio Jacuhy a boring was put down in

72 BULLETIN: MUSEUM OF COMPARATIVE ZOOLOGY.

which a conglomerate 1.80 meters thick is recorded as resting on the
granite floor (White, loc. cit., p. 43). Eighteen kilometers south-
ward from Xarqueadas in another boring the Orleans conglomerate is
reported with large blocks of granite and with a thickness of 16.16
meters, resting directly on the granite.

In the exposures reported between Florianopolis and Lages and

F Rio Brac a de Va
Paine Jorteans oNortel apivavy

Tubarao Kee ve

ns

MER samt ings

Fia. 20.— Section of the Minas-Orleans basin.

between Lages and Blumenau in Santa Catharina, there is described
among the beds beneath the conglomerate horizon “a very hard fine
grained grayish white whetstone grit” in layers eight to twenty
centimeters in thickness. These layers it is also stated gave the name
Navalha to the village on the Blumenau-Lages road near which
locality they were once quarried for whetstones. Similar beds are
likewise mentioned in the same report as resting with a few meters
thickness on the granite near Suspiro in Rio Grande do Sul and also
along the Rio Trombudo where crossed by the Blumenau-Lages road.
Attention may here be called to the fine-grained compact white sili-
ceous rock which underlies the boulder-beds on the right hand bank
of the Jaguaricatu along the railroad between Sengéns and Sao Pedro
de Itararé in northeastern Parana described on page 62; fragments
of this rock appear to be abundantly distributed in the shales and
sandstone of Sao Paulo often with glacial scratches.

I have little more than details of structure and the conclusions
based thereon to add to Dr. White’s account of the Permian con-
glomerates of the Orleans basin. This basin (Fig. 21) is a down-
faulted outlier of the Permian area. The boundary fault, on the
western margin of the basin, brings the sediments against a basic dike
whose shattered condition suggests that the faulting occurred after
the intrusion of the dike. A short distance west of this broad dike
is a narrow basic dike somewhat faulted within its mass. I observed
two nearly vertical slickensided surfaces striking nearly northwest
southeast on which the slickensides pitch to the southeast on the
eastern fault at angles between twenty-five and thirty degrees and

WOODWORTH: GEOLOGICAL EXPEDITION TO BRAZIL AND CHILE. 73

indicate, by the detailed structure of the surfaces, that the down-
throw was on the northeast side in accordance with the structure of
the Orleans basin. It is evident therefore that repetitive small faults
occasion the base of the Permian section in this district.

The conglomerates are exposed along the river banks below the

AAA AZ ” AG 7 a7
| a eae Aw
Ai AIT, 7

SPR se 7 ES RR Bey

ra

Fra. 21.— Section of beds in Orleans basin, south bank of the Rio Tuberao.
The rock on west is a broad trap dike.

town and are best shown on the north bank under the railroad track,
(Fig. 22). There is here a water laid conglomerate mainly of granite
pebbles with a few quartzite and quartz pebbles. The conglomerate
is overlain by cross-bedded grits, the cross-bedding dipping to the
north and northwest as if deposited by currents of water flowing at
least locally in that direction. The pebbles in the conglomerate bed,
mostly three inches in diameter, sometimes attain five inches, and are
embedded in a paste of granitic detritus. The subrounded shape of
the pebbles indicates no distant journey and their lithological character
betokens a derivation from the granitic terrane which immediately
underlies the local Permian section. (See Plate 25.)

Another exposure of the conglomerate about 90 meters down stream
from the preceding exposure presents the cross-section of a north-
south ridge of coarse pebbles
enveloped in cross-bedded
sandstones. In a layer of
conglomerate varying from 0
to 60 cm. in thickness the

cobbles attain a diameter of Fic. 22.— Esker-like ridge of conglomerate
9 S oe : in lower Permian beds; near Orleans, Sta.
20 cm. The ridge-like de Sita faia en

posit, so suggestive of a

buried esker, is apparently continued across the river on the south
bank of which there is a ridge-like exposure of conglomerate also
covered by sandstones. A few yards east of the ridge on the
north bank there is exposed a bed with rounded granite pebbles
scattered through a sandstone matrix as if dropped by floating ice.

74 BULLETIN: MUSEUM OF COMPARATIVE ZOOLOGY.

All of the strata below the railroad track are essentially devoid of
decomposition evidently because of the geologically recent lowering of
the bed of the river. While the assemblage of water deposited
conglomerates in this section are entirely consistent with neighboring
glacial conditions there is no true tillite in the lower part of the section
and while the ridge may be an esker the evidence is not conclusive.
It is possible to regard the narrow belt of conglomerates as deposited
in the channel of an aggrading stream which in its later stages of
deposition laid down the cross-bedded enveloping sandstones.

As the Orleans section by no means affords a typical exposure of the
tillite beds the name seems inappropriate for the glacial conglomerates

Fic. 23.— Basal boulder in Permian shales, resting on granite, near Minas,
Sta. Catharina.

and Dr. White (Woodworth, 1910, p. 779) having kindly coincided
in the matter the name Jaguaricatu has been substituted for this
horizon because of the splendid exposures of tillite along the banks of
this river in railway cuts in northeastern Parana.

In the exposed section of base of the main area at Minas no con-
glomerates were encountered. A few scattered granite and quartzite
pebbles occur in sandy beds but without striae or flattened sides or
crushed and snubbed ends indicative of glacial action. About 2.4
kms. below the railway station in a railway cut the basal Permian
shales may be seen resting on the Pre-Devonian granites. On this
ancient surface reposes a boulder of granite about three feet (1 meter)
in diameter covered by the shales. Some small grooves or channels
in the granite are filled with a gravelly sandstone. No traces of a
glaciated floor were discernible.

It remains to characterize the glacial features of the above described
localities as a whole and to draw from the evidence now in hand such
conclusions as appear tenable.

There can be no doubt as to the glacial origin of the massive tillite
beds in Parané and their likeness to the tillites of Permian age in
India, Australia, and South Africa. The massive aggregation in

;
q
4
5
:

ee

WOODWORTH: GEOLOGICAL EXPEDITION TO BRAZIL AND CHILE. 75

northeastern Parand is highly suggestive of moraines laid down by
land ice; but other sections in Sao Paulo and to the south and west
in Parand are equally clear as to their deposition by floating ice at
times at any rate in the sea as the presence of marine fossiliferous
marine shales in Parana plainly show.

More particular problems arising in the study of these glacial
deposits are discussed under separate headings.

The Striated Pebbles. — So far as present knowledge goes, striated
surfaces have not been found on the large boulders and block erratics
in south Brazil, but such ice-worn surfaces are frequently met with
on pebbles and fragments of rock ranging.from the size of a hen’s egg
to that of a man’s head. In almost every exposure of conglomerate
whose pebbles fail to display the well-rounded contour of water-
wear a few striated pebbles may be found after a few minutes search;
yet of these sections I saw none in which all the pebbles and fragments
were striated. On the whole the proportion of striated pebbles in the
tillite beds is about as large as it is in the granitic till of the glacial
deposits of New England, but glaciated stones are certainly not so
abundant in any given mass as they are in the stony blue clays of
many localities in the Pleistocene deposits of, for instance, portions
of the Wisconsin moraines of the state of New York.

The absence or apparent absence of striae from the larger blocks
and boulders is quite in keeping with the distribution of striae in the
Pleistocene of many districts and has little significance in the argu-
ments for or against glacial action, though in the case of the Brazilian
deposits striae were not earlier noted perhaps because attention was
given more to the blocks of striking size than to an examination of the
worn surfaces of small fragments of transported rock.

Broken up striated Rock-floors— As the glaciation and consequent
striation of the indurated terrane over which the ice moves proceeds,
the rock-floor particularly at the upstream edges of small declivities
in the rock-surface breaks away, so as to produce angular fragments
with one flat well-striated surface — that of the original floor. Such
pieces of rock may subsequently be striated over all their new fractured
surfaces but will for short journeys tend to preserve their features.
Thus the striated stone shown in Plate 27, the first striated fragment
feund in Parana, presents all the ear-marks on its well worn and
striated side of a piece of old rock-floor even to the chatter-marks in
the broader and deeper groove; while on the reverse, where there are
two intersecting warped surfaces one is slightly scratched, and the
other is a much more recent fracture, still of a date anterior to the
final embedding of the pebble.

76 BULLETIN: MUSEUM OF COMPARATIVE ZOOLOGY.

Although no glaciated rock-floor has yet been found in south Brazil,.

pebbles of the type above described point to the one-time existence of |
such a rock-floor though it is no longer to be seen. In this particular-

instance that of a pebble from the tillite beds on the Jaguaricatu the
rock is a reddish brown ! fine compact argillaceous sandstone, with a

perceptible clayey odor when breathed on, and carrying minute scales.

of muscovite.
Crushed and Blunted Rock Fragments.— Striated pebbles or rock
fragments with or without flattened sides or “soles”? are commonly

regarded as the most characteristic molar constituents of moraines.

directly due to glacial action; but there is another type of pebble in
ice-laid moraines which is equally peculiar to the process of erosion
and transportation; that is the pebble with blunted and snubbed

ends with or without striae. This kind of pebble is usually rather

elongate and oftener displays its fractured and splintered surface at
that extremity which has the smallest cross-section. Such fragments

are not uncommon in the glacial drift of North America and I have-
found them in the tillite of Parana in the beds along the Rio Jaguari--

catu. One example quite characteristic in every feature is illustrated
in Plate 27, fig. 3.

The small intersecting surfaces which give rise to the beveled appear--

ance of the subpointed portion of the periphery of this rock fragment

are surfaces of fractures produced by the riving off of spalls of the

rock through pressure applied at points along the major perimeter
of the pebble. In the case of this pebble, there are three such larger

fracture surfaces and each of the later fractures was followed by a.

repetition of the pressure at approximately the same point so as to

force a smaller spall with a subconchoidal fracture. At some stage-

in the same process after the intermediate surface was produced by
fracture, the wedge-shaped point of the pebble was broken squarely off.
All of this fracturing was accomplished previous to the final embedding
of the pebble in the tillite bed. Owing to the concavity of the fracture

surfaces, they escaped striation, yet some slight scratching took place

on the two larger surfaces.
Such fracturing is apparently due to the forcing of pebbles against
the bed rock or against other rock fragments in the ice or by their

being caught under boulders so as to have a great weight of ice con-

centrated upon them when they in turn are in contact with the bed
rock.

! The color of the dry isolated rock is close to Orange 130, Klincksieck et Valette>
©dde des coleurs (Paris, 1908).

Pi Son
ra]

WOODWORTH: GEOLOGICAL EXPEDITION TO BRAZIL AND CHILE. 7/7

Where this type of pebble occurs on the surface of the ground at the
present day as it does in the glaciated regions of eastern North America
it might possibly be interpreted as owing its conchoidal fractures to
the work of aboriginal man, but when firmly embedded in the glacial
till of that district or in the tillite beds of Brazil no one would presume
to connect it causally with human art. These crushed and bruised
rock fragments with their sides well striated are common in south
Brazil and along with other striated pebbles argue for the crushing,
bruising action of a thick body of ice such as a glacier would afford.
Even where stones of this type of contour occur isolated in fine shales
into which they have been dropped from floating ice, the evidence as
to their original handling by glacial action is equally good.

Classification of Lithified Glacial Deposits and Derived Sediments.—

‘The lithological classification of sediments takes no account of genesis,

its names, somewhat more carefully defined than in common usage,

express ideas concerning the size of constituent particles, as in the

terms conglomerate and sandstone; or designate vaguely a mode of
fracture, as in shale with an understood composition of particles too
small to be distinguished by the unaided eye. This simple primitive

classification embraces all lithified glacial deposits when used with

proper qualifying terms. The glacial deposits, so far identified, have
given rise to two independent terms, boulder-bed and tillite, the first

of which is conveniently vague, except for its reference to boulders,

while the second covers a wide range of commingled rock fragments

and particles having this in common that they were deposited by the

agency of glacial ice. The study of modern glacial deposits would
lead us to expect among ancient glacial deposits the lithified counter-
part of each product of glacial action and the glacio-natant waters.
Tillite, as consolidated till, would naturally be applied to all un-
stratified, unassorted deposits due to the direct agency of a glacier.
The term thus is applicable to the rock of fossil moraines whether
frontal, or ground-moraine, and to fossil drumlins. These varieties
of tillite, since they are distinguished by topographic form, will not
in the nature of the case take petrographic designations. The same
remark applies to the assorted glacial gravels and sands forming the
group of kames, eskers, and proglacial deltas, or gravel- and sand-
plains. For their petrographic designation there is no distinctive
term correlated with till and tillite. The coarser deposits are in-
cluded in the conglomerates and the finer among the sandstones.
The glacial rock-flours or clays, normally unweathered, finely divided
clastic materials of complex mineralogical composition, often feldspa-

78 BULLETIN: MUSEUM OF COMPARATIVE ZOOLOGY.

thic, should be sufficiently distinct from true clays to receive a specific
name in the system of classification; but none occurs for modern
deposits of this type in accepted usage. Such lithified glacial rock-
flours may be called pelodite from mndwdns, es (andos, €fdos), rte, clay-
like rock, in reference to the fact that these rocks often clay-like or
shaly in texture and structure, when first formed may differ widely
from typical clay in the very small amount of kaolin which they
contain. Pelodite is to be differentiated from pelite also by its in-
cluded glacial pebbles.

Closely associated with modern glacial deposits though distinct in
origin and often very differently distributed is the fine mainly eolian
deposit known as loess frequently derived from glacial moraines
through their deflation by winds. Some of the finer sediments of the
Brazilian Permian resembling clays seem to be of this origin. Such
lithified deposits of loess may in analogy with tillite be called loessite.

With these preliminary statements it is to be presumed that the
following tabulated classification of glacial deposits will need no
further explanation.

Modern unconsolidated Divisions depending Ancient lithified
deposits. on form. deposits.
ig lt aa Dh aOR OD ie MATOS SEMI AT ES
Houldersiay ROG iets, o> 2. fisiou cal odacek CC rsd MRMRIE son ce ee
Moraine, frontal, ground-.
drumlins
Glacia] gravels........ 97 .Glacial conglomerates.

ioc eee
glacial deltas.

Glacial sands Deltas and plains Sandstones.
Glacial’ clays eo 0) ues cc sce ne areal eae coe
TESS ye cit. peek casanS, «sles: pra Git ae ena ae

Classification of the Glacial Deposits of South Brazil— With the
above simple classification in mind, it appears that the Permian
moraines of south Brazil presents several facies of a glacial series,
both stratified and unstratified.

In the tillite group must be placed the typical blue boulder-clays
of the Rio Jaguaricatu section, the gravelly beds with large angular
blocks, beds presumably of the ground-moraine or subglacial type.
Unfortunately the limited exposures in the deeply forested river
valley do not permit of a determination of the topography of these
ancient beds of drift. In somewhat sharp contrast with the analogue

;
a
:

WOODWORTH: GEOLOGICAL EXPEDITION TO BRAZIL AND CHILE. 79

of true boulder-clays is the earthy type of tillite seen in the vicinity of
Rio Negro. This I believe was deposited by floating ice in a shallow
sea without stratification.

There are many beds composed of small rock particles embedded in
an argillaceous ground-mass carrying only occasional striated stones
which present some of the characters of flood deposits but which by
reason of their glacial features it is to be presumed are tillite. In
their yellow weathered state, their diagnosis is made with difficulty.
They grade into “mealy”’ yellow sandstones.

The group of stratified gravels appears to be represented in the
Orleans basin by the small ridge of conglomerate whose cross-section
and horizontal extension recalls the form of an esker. The other
occurrences of conglomerates as at Ponta Grossa do not show topo-
graphic features which enables one to distinguish them from gravels
of non-glacial origin.

The dark colored shales in the Permian series, on the whole much
less developed in thickness than the coarse sediments, have not been
studied with reference to their glacial derivation. Where they con-
tain marine fossils it is to be presumed that they have been worked
over and any original glacial characteristic has been lost. Some of
the fine clayey beds of sandy aspect which behave like loess in their
weathered condition I suspect were originally loess, but this determina-
tion is difficult to make.

The large content of clay in the tillite beds of Jaguaricatu and
Conchas on the Tibagy is not consonant with the derivation of these
beds from the granites and gneisses of the present Serra do Mar
region since such rocks under the direct attack of glaciation would
produce predominantly gravelly and sandy beds with a minimum of
clay or rock-flour of an argillaceous character. For this reason I am
disposed to regard these clayey tills as worked over from the under-
lying Devonian shales and from the slates of the Pre-Cambrian series.
This makes it possible to suppose that much of the material did not
come from any great distance to the east of the Serra do Mar.

The tillite beds which approach nearest in their lithological charac-
ters to a solidified boulder-clay appear when fresh of a bluish color
somewhat darker than the hue of the glacial brick clays of many parts.
of the United States of America. On exposure the rock weathers to a
light brown or yellowish brown color often with streaks of reddish
iron oxide.

This rock joints irregularly; frequently its fracture assumes a
curving plate-like structure tending toward the dome structure and

80 BULLETIN: MUSEUM OF COMPARATIVE ZOOLOGY.

actually attaining that of spherical separation, thus producing the
bales shown in the Rio Negro section, characteristic of many weather-
ing basalts, an effect coupled with the want of definite stratification
within the stratum.

Once the tillite beds underlying sandstones are reached by dissecting
streams or rivers, the overlying strata widen out with steep cliffs, the
small streams falling over these rocks have falls, and, owing to the
ready removal of the tillite and the growth of the main valley in this
section, the side streams present the appearance of hung up valleys.

Age of the Boulder-Beds.— As commonly stated on the authority of
Dr. Derby the boulder-bearing deposits are to be regarded as the
equivalent of the Upper Carbonic or Permian glacial period now
recognized in India, Australia, and South Africa. This Permian age
of the beds in Brazil was accepted by Professor Branner in his Geologia
elementar (Rio de Janeiro and Sao Paulo, 1906), and still more
recently by Dr. I. C. White in his Report on the Brazilian coal field.
In the same monograph, Mr. David White presents the most complete
description and analysis of the flora of the south Brazilian field yet
published. From this report it appears that the flora succeeding the
conglomerates on the south in Santa Catharina are of the Lower
Gondwana type which immediately succeeds the glacial beds in India
and elsewhere in the eastern hemisphere. Dr. David White corre-
lates the Jaguaricatu beds [Orleans conglomerate] and associated
sandstones and shales of Dr. I. C. White with the Talchir beds of
India, the Dwyka conglomerate of South Africa, and the basal
conglomerate of the Permian series of Argentina.

Divisibility of the Glacial Beds—— The question is naturally asked
whether there was one or more than one episode of glaciation or ice-
action in southern Brazil. The facts concerning the distribution of
conglomerates in this field are as yet too little known to enable one to
give a satisfactory statement in reply to such a question. The
localization of the typical tillite beds near the base of the series so
suggestive of local contributions of débris complicates the question.
If the conglomerate beneath the whitish sandstone at Serrinha Station
with its glaciated fragment be of Permian age and pass, as it appears to,
beneath the boulder-bearing beds exposed westward and southwards
towards Lapa, then in that section there are evidences of two epochs
of ice-action separated by the deposition of the sandstone beds which
form the walls of the gorge of the Iguasst.

The tendency of the boulders to display themselves in a marked
manner in certain zones often only a few feet apart is evidently a

WOODWORTH: GEOLOGICAL EXPEDITION TO BRAZIL AND CHILE. Sl

result of deposition from floating ice and such distribution can not be
depended upon as a basis for discussing advances and recessions of
glaciers. It can only be said here that the evidence seen points to the
existence of two horizons on which ice-action directly or indtrectly
arising from glaciation appears to be demanded.

It would seem from a comparison of the partial sections visited in
the course of this Expedition that the boulder-beds are not persistently
parallel formations, that the more typical tillite of one district may
pass by gradation or intercalation of deposits into contemporaneous
waterworn gravels or sands, now conglomerates and sandstones in
another district along the strike of the beds. Thus in the sections
along the Jaguaricatu in northeastern Parana, the tillite with large
angular blocks surmounts with few intervening feet of beds the white
fine sandstone which appears to be a fairly persistent basal and pre-
glacial member of the Permian series. In the latitude of Ponta
Grossa the tillite appears at a much higher horizon, the apparent
place of the Jaguaricatu tillite formation being taken by beds of
waterworn pebbles. Again in the gorge of the Iguassti at Serrinha
Station, if the beds be truly of Permo-Carbonic age at this horizon, a
conglomerate with striated pebbles occupies an inferior position near
the base of the series. ‘At Orleans, yet further south, the tillite is
replaced in the section by waterworn conglomerates possibly, though
not certainly, laid down in the presence of ice in the manner of eskers.

This alternation from point to point along the present roughly
meridional] line of exposures, thus described of deposits approximately
at the same level in the series, and probably more or less contem-
poraneous, finds a parallel in the existing deposits of glacial origin
within the glaciated areas of Europe and North America. In travers-

ing the glaciated region of the latter area we pass from north to south
over belts of till with alternating strips of waterworn gravels and
coarse sands and clays. In this particular case the deposits are
successively newer in the direction in which the ice retreated, 7. e.
towards the north. We encounter another mode of deposition of
alternating accumulations of ice-laid and water-laid drift, however, in
which the likeness to the Brazilian distribution, so far as it is at
present known, is equally close. That is where glaciers coming down
either as distinct valley glaciers or as outflowing tongues from a
central ice-cap reach the coastal plain or sea-floor at the base of a
highland region so as to deposit till in the vicinity of the paths by
which they reach the low grounds while the intervening areas receive
only the waterworn débris. In south Brazil, what seems to be evi-

82 BULLETIN: MUSEUM OF COMPARATIVE ZOOLOGY.

dence of an eastern origin for the glacial débris points to this latter
mode of deposition rather than to the first above outlined.

Our knowledge of the Permian of south Brazil is limited to the
narrow belt from which the overlying supposed Triassic beds have
disappeared by erosion. The Permian is not represented as reappear-
ing on the western flank of the broad shallow syncline which subtends
the structure as far west as Cuyuba in Matto Grosso but it is probable
that the Permian extends far beneath the Triassic cover.

Boulder-Beds of the Argentine Republic The brief references of
Bodenbender (1895) to the occurrence of boulder-bearing beds be-
neath the Gondwana flora in the provinces of San Luis and Mendoza
in central Argentina may be presumed, though this geologist does not
infer their glacial origin, to indicate an extension of the Brazilian
Permian geographic conditions to the south and west quite to the
base of the Andes, thus practically carrying the peculiar geographic
and climatic features of the Permian across the continent of the
southern hemisphere from one shore of the Pacific Ocean to the other.
Details concerning the agency concerned in the transportation of
these Argentine deposits is as yet lacking and until the precise facts
on this point are known it would be presumptuous to draw further
conclusions from the deposits concerning the glacial problems of
the South American Permian.

Boulder-Beds of the Falkland Islands— Mr. Thorre G. Halle (1911,
p. 115-229) reports the existence of a boulder-bed and typical
tillite with striated flat-faced pebbles at the base of a Permo-Carboni-
ferous series on East Falkland Island. Higher up in the section occur
Glossopteris, Gangamopteris, Phyllotheca and other members of the
Gondwana flora.

The occurrence of glacial deposits in south Brazil and on the Falk-
land Islands carries with it the presumption that the boulder-beds
beneath a Gondwana flora in Argentina are also glacial in origin
whether or not they now show striae, which latter may yet be found
on the small rock fragments.

Gondwana-land. Did it include Parand-land? — Palaeogeographers
relying largely upon the distribution of the fossil genera Glossopteris,
Gangamopteris, and their plant allies of Permo-Carboniferous times
with the associated glacial boulder-beds have gradually extended the
name Gondwana-land from its original application to the Indian
South African area so as to include the South American tracts in
which traces of the Gondwana flora occur. To a certain extent the
name has thus become designative of a set of geographical conditions

WOODWORTH: GEOLOGICAL EXPEDITION TO BRAZIL AND CHILE. 83

rather than a definitely determined continental tract of such vast
extent. Some geologist of the distant future viewing the embedded
traces of the Pleistocene glacial period of northwestern Europe and
northeastern North America might with almost equal assuredness
point out the existence of a vast continent embracing these areas and
the included basin of the existing North Atlantic Ocean. The geo-
graphical conditions peculiar to Gondwana-land are found in India,
Australia, and in South America from the Falkland Islands to the
northern borders of the state of SAo Paulo. Representatives of this
flora apparently of later date occur as far north as the Mexican state
of Oaxaca, there with plants of a Triassic facies, Williamsonia, Zamites,
Otozamites, etc. (Wieland, 1909, p. 441-442). For the purpose of
designating the larger tracts in which this Permian flora and the
glacial conditions occur, Gondwana-land has thus lost its original
limited meaning. The south Brazilian field with its boulder-beds and
later Triassic trap sheets constitute a well-defined geological province
to-day for which the name Paran4-land is quite appropriate. It is to
be presumed that Parand-land was conterminous with land southward
over Argentina and thence to the continental island group of the
Falklands. All three of these Gondwana areas lie within that of the
existing continental block.

The question of the connection of South America with Africa in
Permo-Carboniferous times may be stated in other terms in a more
general form to be that of the origin and history of the Atlantic Ocean
basin, which Suess has discussed in no uncertain way. The fact that
there is no recognizable trace of the Atlantic Ocean along its existing
borders in early Triassic times, except for a narrow sea marginal to
the Mediterranean tract, and the fact that the Atlantic is a narrow
and not deep sea are quite consistent with the hypothesis of the origin
of this depression’ since Lower Triassic times. Certainly the assump-
tion of an Atlantic trough in Pre-Triassic times having anything like
the present extent of that basin must be abandoned as being with-
out sufficient geological evidence. Viewed in the light of Suess’s
masterful generalization of the geology of the Atlantic shores, we find
no trace of the South Atlantic Ocean during the Carbonic period
until the possible marine episode of the Permian epoch in South
Africa + and in south Brazil if indeed these waters penetrated these
continents from the Atlantic basin. In Brazil it is more probable
that the sea invaded the state of Parana from the west. Without

1 Halle notes the report by Schroeder of marine fossils above the Dwyka con-
glomerate in German southwest Africa. Op. cit. p. 203.

84 BULLETIN: MUSEUM OF COMPARATIVE ZOOLOGY.

further details concerning the African Permian sea it is possible to
admit a shallow sea in the Atlantic area as far north in that district
as the southern tropic, but this at a later stage than the epoch of
principal glaciation in Africa. Both in Africa and South America,
the marine beds of the Permian demand land near sea-level immedi-
ately before and after the invasion by the sea. Little else can at
present be argued from their occurrence.

The volcanic islands which stake out the mesial line of the Atlantic
and its connection with the Arctic Ocean hug the shores of Europe
and Africa rather than those of the Americas. Since those voleanoes
whose substructure is known are associated with crustal displacement
involving either horizontal or vertical motion or both and particularly
the latter mode of derangement, it is probable that we see in these
islands the indirect evidences of a geologically recent movement of the
Atlantic bottom which we know from Iceland was well under way in
Miocene times. Along the shores of North and South America which
trend northeast and southwest we find traces of downsinking of the
land in Cretaceous times. In the North Atlantic region, the Pan-
Appalachian mountain-chain extending from the dislocated structures
of the Rocky Mountains across the southern United States to the
coast in Newfoundland can be traced north of the Alps in Europe
into Asia but is interrupted in the Mississippi embayment and by the
North Atlantic basin which cuts across the folded structures as if they
had sunk to form the present ocean floor. These and other indicated
changes of depth and outline of the Atlantic province make it in-
credible that in Permian times the basin had much of its present
length, breadth, and depth. We may conclude therefore that the
geologist is free to converge the coasts of Africa and South America in
Permian and earlier Carboniferous time as closely as any biological
facts and geological evidences of land may demand for their explana-
tion.

In south Brazil all the known facts from the Atlantic border demand
an extension of land beyond the present coast in late Palaeozoic time,
but how far towards the coast of Africa we can not say. The dis-
covery of the Gondwana flora in the Falkland Islands makes it possible
to effect the distribution of these plants from Australia into South
America by way of the Antarctic continent. Possibly Africa also
received its population by this route or more likely from India.
The revealing of the geology of the Antarctic continent receives from
this state of the problem of Gondwana-land a renewed interest.

The Permian Glacial Problem.— In the foregoing account of the

WOODWORTH: GEOLOGICAL EXPEDITION TO BRAZIL AND CHILE. 85

Permian boulder-beds of Brazil I have set forth the evidence for the
existence of glacial discharge of débris at or near sea-level. The
occurrence of boulder-beds, presumably likewise of glacial origin, in
central Argentina is supported by the finding of tillite in Permian beds
on the Falkland Islands. These occurrences practically complete
the evidence that the Permian glacial traces occur on all lands outside
of the Antarctic circle in the Southern Hemisphere from shore to shore
of the Pacific Ocean. In South America the evidences exist from the
tropic of Capricorn southward to 52° S. L. They occur in South
Africa coextensive with the sediments of the time. In Australia they
occur over a vast area in the eastern part of that continent and in
Tasmania. It is now apparent that no shift of the polar axis in
Permian time will bring these evidences of glaciation into a better
circumpolar distribution than they now display. Thus those hypo-
theses which attempt to account for Permian glaciation by a shift of
the earth’s axis of rotation have not been called for by any facts which
we now possess. The discovery of tillite in the latitude of 52° S.
diminishes the difficulty of the climatic problem by removing the
supposition that the glaciation was dominantly a subtropical affair.
The general absence of existing land in the Southern Hemisphere in the
latitude of the Falkland Islands probably accounts for the present
lack of signs of Permian glaciation in high latitudes. The Antarctic
continent is too little known as yet to premise what evidences future
explorations may bring forth. The facts upon which the Permian
glacial period rest still come largely from the Southern Hemisphere
where at present the ratio of area of land to sea is so small.

In the Northern Hemisphere the Permian glacial deposits remain
most typically developed and best known in the subtropical region of
India, in the Salt Range and Talchir districts. But traces of ice-
action in high latitudes are not wanting and are coming continually
to light with the more critical diagnosis of the conglomerates of the
late Palaeozoic terranes. The Permian breccias of England regarded
by Ramsay as of glacial origin as early as 1855 appear to be now
accepted as such by English geologists.

These traces occur in Latitude 53 N. A. Julien supposes the coarse
breccias of the Carboniferous in France to be of glacial origin, and —
Kalkowsky attributes to glacial action a pebbly shale among the
Carboniferous rocks of the Frankenwald. These occurrences in
Europe pointing to some kind of ice-action if not in every case to the

1 Halle draws the same obvious conclusion from his discovery of the tillite on the
Falkland Islands,

86 BULLETIN: MUSEUM OF COMPARATIVE ZOOLOGY.

existence of glaciers demonstrate the wide distribution of the climatic
conditions producing Permian glaciation and where the traces of ice-
action exist in horizons below the Permian show that the glaciation
of that period was induced by causes not suddenly brought into play.

In North America in latitude 42° S., Messrs. R. W. Sayles and L. A.
Laforge have called attention toa remarkable breccia at the top of the
Roxbury conglomerate south of Boston in which many of the peculiari-
ties of glacial till occur, including a few striated pebbles. The precise
position in the time scale of this bed is not known from contained
fossils in the local series but its relations to the Narragansett coal
field on the south make it presumable that the beds are Permian.
In Oklahoma in about latitude 35° S. erratic blocks occur in a shale
of Subcarboniferous age. The striae on these transported stones
are of post-depositional origin due to rock motion but the transporta-
tion of the boulders remains unexplained except by floating ice.
The evidence from North America, as yet meagre, points to the same
conclusion as that derived from the distribution of breccias and
conglomerates of Carboniferous and Permian age in Europe. So far
as the facts from North America go, local glaciers are the most that
at present can be postulated.

What combination of geographic with atmospheric conditions
brought about Permian or even Pleistocene glaciation we do not know

with any degree of certainty. All our hypotheses of glaciation —

postulate glaciers and ice-sheets engendered from the fall of snow.
It is possible, though not now seemingly probable, that under the
peculiar recurrence of hailstorms which are, in the existing regime
unusual forms of precipitation, masses of ice might accumulate in
tropical and subtropical areas where snow never falls.

Hail and allied forms of frozen rain-drops enter more. largely into
glaciers in existing mountains than is perhaps commonly recognized.
Kaemtz (1845, p. 380-382) gives several citations to show that hail
frequently falls on the Swiss Alps. It is probable that no small
amount of hail enters into the structure of Swiss glaciers. De Saus-
sure states that during a stay of thirteen days on the Col de Geant at
an elevation of 3,428 meters he was struck with the frequency of hail
and sleet which he observed eleven times. Balmont experienced a
shower of hail during the night that he passed on the summit of Mont
Blane and Paccard found much hail beneath the snow with which the
summit is covered.

Edward Whymper (1892, p. 164 et seg.) describes frequent falls of
hail within the region of glaciers on the mountains of Ecuador. When

~

]
|

)

7

WOODWORTH: GEOLOGICAL EXPEDITION TO BRAZIL AND CHILE. 87

nearly 16,000 feet high on the slope of Sincholagua on Feb. 23, 1869, a
furious hailstorm took place followed after a lull by a fall of snow
mixed with hail which gave way in turn to a thick fall of large flakes
of snow. There was a glacier at this time on the mountain. Again
in describing his ascent of Antisana on March 9, 1869, he states
that at the height of 15,984 feet at the top of a moraine on the right
bank of the glacier a fierce hailstorm occurred about 4 P. M. and snow
fell heavily afterward. :

Charles Darwin (1845, p. 115-116, or 1887, p. 115-116) describes
a remarkable fall of hail in Argentina near the foot of the Sierra
Tapalguem between Bahia Blanca and Buenos Aires on the night of
Sept. 15-16, 1833. The hailstones were said to have been as large
as apples. The same distinguished naturalist cites the account of
the Jesuit Brobritzhoffer who states that in the region much further
north hail fell of an enormous size and killed vast numbers of cattle,
hence on this account the Indians called the place Lalegraicavalca,
meaning the “little white things.”’ Darwin also quotes Dr. Malcolm-
son to the effect that this observer witnessed in India in 1831 a hail-
storm which killed large birds and injured cattle. These stones were
flat; one was ten inches in circumference, and another weighed two
ounces. They ploughed up a gravel walk like musket balls, and passed
through glass-windows, making round holes but not cracking them.

Sir Joseph D. Hooker (1854, 1, p. 405) gives a personal account of
a hailstorm which took place on the 20th of March, 1849, on the south
slopes of the Himalayas. “A violent storm from the southwest
occurred at noon, with hail half an inch across and upwards, formed
of cones with truncated apices and convex bases; these cones were
aggregated together with their bases outwards. ‘The large masses
were followed by a shower of the separate conical pieces, and that by a
heavy rain. On the mountain this storm was most severe, the stones
lay at Darjiling for seven days, congealed into masses of ice several
feet long and a foot thick in sheltered places; at Purneah, fifty miles
south, stones one and a half inches across fell, probably as whole
spheres.”’ |

According to Lieut.- Commander Gorringe, U. S. N., (1873, p. 331)
the city of Rio de Janeiro was visited by a storm with a heavy fall of
rain and hailstones on October 10, 1864. The hailstones were as large
as pigeon’s eggs. ‘“‘Branches of trees were broken and twisted off,
and the foliage destroyed by the hail, which poured down in such
quantities that piles of it remained at the corners of the streets until
the afternoon of the following day.”

88 BULLFTIN: MUSEUM OF COMPARATIVE ZOOLOGY.

Thomas Russell (1895, p. 98-100) gives a few statistics regarding
the falls of hail as follows: — Hail falls during thunderstorms in
summer and during the hottest part of the day. It rarely forms
layers six inches thick. On Aug. 13, 1851, hail fell in New Hampshire
to the depth of 4 inches. On July 24, 1818, it fell on the Orkney.
Islands to the depth of 16 inches, on August 17, 1830, in Mexico City,
to the depth of 16 inches. In the Yellowstone valley in Montana a
fall of 14 inches has been recorded... .Hail is more common at 15,000
feet than at sea-level, it forms at elevations from 5,000 to 16,000 feet;
the greatest size of hailstones is found below 5,000 feet. It falls most
frequently on the lee side of rising ground... . There are on the average
fifteen hail storms a year in France, five in Germany, and three in
Russia.

In the Antarctic region hail now falls but rarely. Commodore
Wilkes alone reports two instances. Fog there gives rise to some
crusts of ice. (Fricker, 1900, p. 244).

Alfred R. Wallace (1892, p. 299) states that he had good authority
for hail having once fallen on the upper Amazon at a place enly three
degrees south of the equator and about 200 feet above the level of the
sea. Humboldt (1852, 2, p. 217) instances a case of hail falling at
Paruruma in the Orinoco valley on a plain near sea-level. He states
that hail in the tropics generally falls only at an elevation of 300
toises (about 600 meters).

The Shaler Memorial Expedition encountered in July, 1908, at an
altitude of 900 meters on the campos of Sao Paulo between Bury
and Faxina (p. 13), a hailstorm which covered the ground as thickly
with hail as do many similar storms in New York and New England.

I cite these instances of hailstorms with their attendant circum-
stances because hail presents us now with a means of precipitating ice
at low altitudes in regions near the equator where snow never falls.
Hailfalls appear to increase toward the hot regions of the globe and
to diminish in frequency of occurrence towards the polar tracts. Thus
the probability of glaciers originating from hailfalls in subtropical
regions would seem to be as great as from snowfalls, provided the
hailstorms came frequently enough to overcome the effect of melting
due to the higher temperature of those parts of the earth’s surface.
Hailstorms like thunderstorms are secondary movements of at-
mospheric vapor normally engendered in the wake of cyclones of far
greater size and with a much longer path, and we do not at present
know what geographic conditions, if any, would cause in Permian
times a succession of hailstorms sweeping with the regularity of

WOODWORTH: GEOLOGICAL EXPEDITION TO BRAZIL AND CHILE. S89

cyclones of the existing atmospheric circulation over the glacial fields.
The distribution of land and sea in the Southern Hemisphere was then
so different from what it now is, particularly in the Atlantic and
Indian regions, that, while we may freely speculate upon a different
system of distribution of aqueous vapor, it is at present impossible to
construct an unassailable theory.

‘The “pendulationstheorie” of H. Simroth (1907) which is but a
modification of the idea of a shifted axis of rotation does not better
than the original conception explain the phenomena of distribution
of Permian glaciation for the reason that no shifting of polar climates
will bring the glacial deposits in the polar regions. Dr. Wilh. R.
Eckardt (1910, p. 125-127) has well observed in connection with the
above mechanical hypothesis that Sumatra, assumed to be one of the
fixed equatorial poles of pendulation in this speculation, is placed on
the borders of the principal area of Permian glaciation and hence no
essential shift is accomplished.

The recently elaborated hypothesis that glaciation may be brought
about through the temporary reduction of the amount of carbon in
the earth’s atmosphere appears to fail as an adequate explanation of
the phenomena of glacial periods since the view does not explain the
succession of epochs of glaciation and deglaciation in the Pleistocene
period. our times in this period the ice-sheets apparently came on
and went off. If the abstraction of carbon in the form of coal and
limestones in the preglacial period led to the first ice accumulation
and advance, the hypothesis leaves unexplained the shortly succeeding
ice advances between whose dates no corresponding appreciable
reduction in the carbon is registered by rock-making in the earth’s
crust.

As we do not know with any certainty the cause of the latest glacial
periods so near our own times, it is evident that the geographical
conditions of the Permian must be thoroughly ascertained before we
can construct a plausible explanation of a glacial period so remote and
taking place on lands whose contours are as yet drawn with too much
conjecture. Certain lights appear however to be burning as guides
to the path which shall lead to the discovery of the probable cause of
Permian glaciation. These may be briefly summarized as follows: —

The axis of the earth appears to have lain in Permian times where
it does now. ‘This excludes a favorite group of hypotheses.

The glaciated lands of south Brazil and German southwest Africa
were in Permian time at or near sea-level. This does not exclude the
extrusion of glaciers from highlands to the sea-border provided the
highlands lay over the site of the Atlantic basin.

90 BULLETIN: MUSEUM OF COMPARATIVE ZOOLOGY.

In the Northern Hemisphere on the existing shores of the Atlantic
Ocean glaciated stones are found entering into the contemporaneous
marine deposits much farther equatorward than those produced by
the action of coast-ice and the same extension is true for the occurrence
of glaciated ledges. Thus on the Atlantic coast of North America
glaciated ledges and deposits of true glacial till line the shore of the
United States as far south as New York Narrows in latitude 40° 30’
north; no modern observer has described stones or ledges bearing
distinct traces cf coast-ice action within this belt south of the British
possessions, though it is to be acknowledged that thin sheets of coast-
ice form in winter and may do some geological work of this character.
The reason for this greater equatorward extent of glacier-ice as con-
trasted with ordinary sea-ice is due to the extent to which glacier-ice
may be pushed equatorward beyond the gathering ground of the
glaciers. In a like manner, glaciers originating on high lands in low
latitudes may reach the sea-level and impress that region with marks
of glaciation where the normal sea-level temperatures at the time
preclude the existence of coast-ice. It can not be too much insisted
upon, therefore, that glaciation is of all forms of ice-action that most
likely to be met with in any marked degree at sea-level in low latitudes.
Hence it is the more reasonable to assume that the Permian ice-deposits
represent the existence of glaciers in the regions where these coarse
accumulations occur, invoking as we may where the geologic evidence
is permissible a favorable geographical relief such as now controls the
distribution of glacial ice at one place or another on the earth’s
surface.

The rédle which hail might play in producing glaciers in subtropical
regions as a complement to:snowfall in higher latitudes and high
altitudes is tentatively suggested as a factor in Permian glaciation,
but the feasibility of the thesis encounters some of the same objections
which meet the accepted origin of glaciers in snowfall.

There are accumulating evidences of the existence of glaciation in
the Northern Hemisphere in Permian time, and there are not wanting
signs of ice-action, probably floating ice, in the preceding Carbonifer-
ous epoch, facts which assist in the attempt to devise hypotheses.

The tendency of the geologic evidence is towards the recognition of
glacial conditions independent of latitude which points to a weak-
ness of the climatic zones, a feature characteristic of Palaeozoic
temperatures, in which non-glacial climates show no zones correspond-
ing to the present ones.

The hypothesis of internal heat controlling surface temperature

i
1

WOODWORTH: GEOLOGICAL EXPEDITION TO BRAZIL AND CHILE. 91

which has been advocated by Mr. Marsden Marson finds little support
in the geological evidence derived from the occurrence of rocks at a
high temperature over large areas beneath the surface. The great
Triassic trap outflows of Brazil were erupted but shortly after the
Permian glaciation in that field. During the epoch of glaciation the
magma was still confined within the crust but had no recognizable
effect in controlling surface temperatures or in preventing glaciation
of the Permian land surface.

VI. THE TRIASSIC TRAP PLATEAU.

Overlying the Palaeozoic strata of the southern Brazilian highlands
and either intruded in or interstratified with a group of red beds of
presumed Triassic age is a series of trap sheets, mainly lava-flows, of
vast extent, comparable in age and geological position with the basic
igneous rocks of the so-called Newark group of the eastern coast of
North America, and rivalling in their present surficial extension the
Cretaceous lava-fields of the Deccan of peninsular India if not also the
more recent lava-flows of the Columbia and Snake River basins of
western North America. From the northwestern part of the state
of Sao Paulo and the borders of Minas Geraes where the trap rests
upon the Pre-Devonian schists, these sheets of trap form high plateaus
broken through by the tributaries of the Rio Parana as far south as
central Rio Grande do Sul. From the sea-border near Porto Alegro
in the latter state the trap formation extends westward according to
the report of Dr. M. A. Lisboa as far as the Serra do Maracajii in
Matto Grosso at a point 460 kilometers east of the Rio Paraguay or
nearly 12 degrees of longitude west from Rio de Janeiro. The trap
formation has been recognized over a breadth of country five degrees
in longitude and some eight degrees in latitude between the parallels
of 20 and 28 degrees south. The area in which these rocks dominate
the surface is in round numbers about 100,000 square miles or ap-
proximately a region as great as that, of the state of Nevada in North
America.

In describing the geological features of the trap plateau in the
Rio Pelotas basin, I shall use for the igneous rocks the non-committal

term trap since it avoids the inexactness which might arise from the

general application of such terms as basalt, augite-porphyry, etce., to
which kinds of rocks certain parts of the trappean series have at one
time or another been referred.

As is stated more explicitly in what follows I include in the area of

9? BULLETIN: MUSEUM OF COMPARATIVE ZOOLOGY.

the Pelotas plateau a large tract about Lages in Santa Catharina in
the headwaters of the Rio Pelotas from which the trap has been
denuded so as to leave the Triassic red beds exposed at the surface.
This region is separated from the trappean plateau of Parand by
the Rio Iguassti as far west as Porto da Unido da Victoria, some
leagues west of which point the dissected trap plateaus of the south-
ern Brazilian coastal states merge in the longitude of the Rio Parana
into a broad sheet with little topographic differentiation. Through-
out Santa Catharina and Rio Grande do Sul the border of the trap
area gives rise to high escarpments capped by trap sheets overlooking
the Permian sedimentary tract. This escarpment with its varying
relief and declivity receives local names. South of Rio Negro it is
known as the Serra do Espigao which attains an elevation exceeding
4,000 feet. The main, almost unbroken, escarpment farther south and
east is known as the Serra Geral.

The Trap Escarpment.— This line of escarpments has been inter-
preted as a fault cliff and as a true retreatal escarpment due to differ-
ential erosion. Doubtless small faults intersect the trappean series
as elsewhere in the region but the descriptions of others and my own
observations upon the escarpment in the Serra do Espigao and at the
head of the Rio Tuberao leave no question in my mind that the
escarpment is the effect of differential erosion on the Permian and
Triassic beds capped by resistant sheets of trap.

At the head of the Rio Tuberao the Serra Geral is a typical steep-
walled trap escarpment below the base of which erosion spurs and
ravines of the sedimentary beds comprising the Trias and Permian
are developed in sharp relief under the active headwater attack of the
Rio Tuberao and its tributaries.

Inland in the region of the Serra do Espigao the alignment of the
escarpment is less simple in its curvature. Siemiradzki in a section
reproduced by Suess in “ La face de la terre’ assumes a normal fault
along this line as he does also in the case of the “cuesta” of the
Devonian sandstones but without structural evidence, and here
also the topographic development of the rocks is simply a matter of
secular denudation as shown in the geological section drawn by
Derby and published by Branner in his Geologia elementar.

Siemiradzki represents the Serra do Espigao as a ridge of lava piled
up above a dike. The northwestern spur of the Espigao forms a high
terraciform ridge standing out above the trap plateau south of the
escarpment but the table-like masses of which the ridge is composed
indicate that it is a portion of the trap sheets left standing out in

WOODWORTH: GEOLOGICAL EXPEDITION TO BRAZIL AND CHILE. 93

‘consequence of erosion. In the section which can be constructed
from the exposures along the road from Rio Negro to Lages at the
pass through Collectoria it is possible that a fault can be traced
cutting off the Espigao from the main mass of trap on the south
at the right of the section in Fig. 24. But this interpretation of the

Fic. 24.— Section across the Serra do Espigao at the Collectoria on the road
from Rio Negro to Lages.

tilted beds at the bottom of the valley and the apparently lower
level at which the trap lies south of the stream crossing demands
a throw of the normal fault on the south side in a manner nowise
supporting the hypothesis that the escarpment is due to a fault.
This northerly steep dip has by some been regarded as cross-bedding
and is stated to occur elsewhere along the trap escarpment in the
underlying sandstones.
Number of the trap sheets— It is stated that there are four trap
sheets or sets of sheets in the trap mass of the Serra Esperang¢a on the

s. LAGES CORYTIBANOS SERRA
— Oo
pa are , SK Sex IXKE REP ESPIGAO,

FBO 0.0.0 04'S Ye ~ ee
ee SS SS ns 9 OS as oS as Oo GT RY OH So

Mi PERMIAN Wn ee See
OEE i aaliec sy iar ae

Fia. 25.— Cross-section of the Trap plateau from the Rio Negro to Lages.
1 (dotted). Sao Bento beds. Triassic. 2 (right lined). The Collectoria
trap sheet. 3 (cross lined). The Corisco trap sheet. 4. The Third
(and Fourth?) trap sheets.

eae ate

north of the valley of the Rio Iguassti. In Dr. Derby’s section the
trap is diagrammatically represented as if composed of surface flows.
The number of flows or sheets entering into the plateau between the
Serra do Espigéo and the sandstone area about Lages was not de-
finitely determined by observation upon the ground but the plotting
of the profile of the route and the outcrops of sandstones encountered
in the trappean tract leads to the conviction that along this route
there are at least three great sheets of trap separated by sandstone
beds. The subjoined figure section (Fig. 25) gives the approximate
profile and geological section along a line paralleling the road from

Q4 BULLETIN: MUSEUM OF COMPARATIVE ZOOLOGY.

Rio Negro to Lages. The elevations are from aneroid readings and
are subject to such a correction that the altitude as given may be 100
feet out. The structure as drawn involves a gentle folding of the trap
sheets in accordance with the varying dip of the strata between the
southern edge of the trap and Lages. The inclined bedding seen in
certain sections exposed in stream bottoms is interpreted as due to
displacement indicating downthrow along faults or else sharp flexures.
An interpretation of the structure regarded by the writer as less
probable than that above given demands the presence of four trap
sheets and requires the existence of a fault of several hundred feet
downthrow along the southwestern border of the trap in the upper
waters of the Rio Canoas.

In either construction the trap in the pass over the Serra do Espigao
at the old Collectoria appears as the lowest in the series. The north-

See ea cg, SANE ey, Fee

Fic. 26.— Trap ridges of northwestern arm of Serro do Espigaio, seen from
heights near Corisco.

western branch of this trappean ridge as seen from the heights near
Corisco shows three distinct tabular masses rising above a common
level which may be that of an underlying trap sheet. A four-fold
division of the trap series is there strongly suggested. The fact that
one of the great trap sheets may consist of more than one flow without
intervening beds of sandstone or shale makes it impossible to rely
implicitly upon topographic profiles. The locality laid too far from
our route to be geologically examined.

The lowest bed of trap on this route which I have called the Collec-
toria sheet is relatively thin, according to a rough estimate given below
and based on a mule-back transit, about 300 feet. It appears to form
the surface along the line of the Lages-Rio Negro Road as far south
as the crossing of the Rio Correntes, where the second sheet or Corisco
flow is encountered. Along our line of route the upper surface of the
first flow or sill was not seen and though I am inclined to regard it as
an effusive sheet I am not able to point out local evidence for such a
conclusion.

The Corisco sheet is, according to my rough measurement, the
thickest of all and has the greatest surface exposure in this district.
This sheet forms the surface of the plateau along the road from Corisco
to Coritybanos. The trap is moderately dissected by numerous
streams. Amygdalar trap abounds on the higher parts of the surface

WOODWORTH: GEOLOGICAL EXPEDITION TO BRAZIL AND CHILE. 995

affording presumptive evidence that the sheet is an effusive flow.
The road from Coritybanos northwestward to Sao Jodo traverses
the deeply eroded surface of this sheet across the campos of Guarda
Mor. In fact the upper waters of the Rio Marombas and its tribu-
taries, the Rio das Pedras and the Correntes, mainly lie on this surface.

South of Coritybanos there appears to be good reason for consider-
ing the high trap mass there encountered as an outlier of a third sheet
limited on the south by the Rio das Cachoeiras. This mass seems on
account of the sandstones which crop out on its southern slope to be
divisible into two beds of basalt. If this interpretation is correct,
the top of the hill is a remnant of a fourth trap sheet in this area.

Thickness of the Trap Sheets.— Exact estimates of the thickness of
the several trap sheets are not possible from the notes I have; but the
following approximations are thought worth recording as giving the
order of magnitude of the thickness of the sheets. The estimate
postulates the supposition that the base of the Corisco sheet is to be
seen on the south bank of the Rio Correntes at an altitude of ca.
3,400 feet, and that its upper surface passes beneath higher beds in
the vicinity of Coritybanos at an elevation of ca. 3,200 feet. By
graphic construction the sheet is found to be ca. 600 feet thick (190 M.).
The overlying trap remnant south of Coritybanos consisting possibly
of two flows must be nearly 600 feet (190 M.) thick with the inter-
calated sedimentaries. The basal bed or Collectoria sheet on this
estimate would be about 300 feet (95 M.) thick. The total thickness
of the trap sheets would thus be about 1,500 feet (475 M.).

Origin of the Trap Sheets — No known voleanic necks or explosive
vents have been reported in connection with these sheets. In the
region which I visited the succession of sheets is suggestive of surface
flows and the Corisco sheet is very probably of that nature. Far to
the southeast in Rio Grande do Sul, Dr. I. C. White has described the
occurrence of sills and intrusive masses of trap as if breaking up
through the sediments into the horizons occupied by the sheets in a
manner to support the theory of fissure eruptions for the origin of the
true flows of this region. In our journey across the sandstone tract
north of Lages one noteworthy dike was encountered presenting
characters which bear upon the mode of origin of the trap sheets.
That occurrence will now be described.

The Dike with Inclusions North of Lages—— About 8 kilometers north
of Lages on the road to Coritybanos there is to be seen a remarkable
dike agreeing in petrographic character with the neighboring trap

sheets on the north and probably to be regarded as a feeder to one of

96 BULLETIN: MUSEUM OF COMPARATIVE ZOOLOGY.

them. This dike crosses the road in a direction N. 70° E. (magnetic)
having a width at the road of 23 meters (75 feet). The sandstones
on the north side of the dike strike north-south (magnetic) and dip
ca. 5° west. The sandstone on the south side of the dike strikes
N. 23° W. (magnetic) and dips 7° west. West of the road the dike is
traceable by means of small weather-boulders in a brook for about 1,098
meters (3,600 ft.) but the trace of the dike to the east of the road was
not followed. (See Plate 18.)

In the disintegrated rock of the dike ose are crystals as large as a
hen’s egg of a now rusty black augite and large flakes of a black mica
carrying holes from which some included mineral of earlier genesis
has been dissolved out. A soft, white, partly altered, mica is probably
bleached biotite. The feldspar constituent was not observed.

The most striking feature of this dike is the large number of in-
clusions of foreign rocks which it contains. These constitute at the
exposure in the road quite one half of the volume of the dike and
comprise at least the following varieties of older rocks, viz: —

Red sandy shale in fragments up to 51 em. (20 inches) diameter.

Red shale; also a black shale.

Coarse grained basalt with lathe-shaped feldspars.

Fine grained basic rock.

Amygdaloidal basalt.

Greyish amygdaloidal basalt, less vesicular lava than the preceding.

A fine-grained, dark, thin-laminated rock weathering white, origin

not determined.

The fragments of sedimentary rock may well be regarded as dis-
rupted from the walls of the fissure which appears to have been the
locus of a vertical displacement of the strata. Presumably the
fragments came from underlying strata though they may just as well
have fallen in from above. The vesicular lavas of different types,
unless there are flows buried beneath the Lages area of sandstones,
contrary to my own determinations of the geological structure and
those of Dr. I. C. White, must have been derived from overlying lavas
Though rarely, amygdales form vertical bands in dikes, they are
characteristic of lava-flows or of lava which has been raised to the
vent of a volcanic conduit. This dike apparently communicated in
its time with the surface and permitted lavas already extravasated
and cooled to yield fragments which sank in the still fluid magma of
the dike. So much of the history which appears in the nature of the
phenomena leads to the further conclusion that the dike communicated
at the time of the infalling fragments with the Corisco, or some yet

WOODWORTH: GEOLOGICAL EXPEDITION TO BRAZIL AND CHILE. 97

higher sheet. From these observations it is to be inferred that
fracturing of the underlying sandstones and the migration of the basic
magmas towards the surface went on during the period of great
trappean outbreaks presumably with a foundering of a vast area
which became flooded with successive sheets of lava from many
fissures. Between one and two kilometers south of the dike above
described another but much narrower basic dike about a foot wide
crosses the mule path near a small stream. An extended search would
probably reveal many other dikes once serving as feeders to the over-
lying trap sheets.

The Lages Area.— For several leagues around Lages the trap sheets
have been denuded leaving the Triassic sandstones and shales of the
Sao Bento beds of Dr. I. C. White’s report at the surface apparently
in an anticlinal dome. On the northern margin of this tract the trap
overlooks it with a well-defined, but much notched, escarpment rising
a few hundred feet above the general level. At the eastern limit of
vision from the Lages road, a conical outlier of the trap forms a
prominent hill, showing that along this line, as on the face of the escarp-
ment overlooking the Permian territory, erosion and not faulting has
produced the trap escarpment.

The Lake Basins of Santa Catharina— “La province de Santa
Catharina est couverte de petits lacs.” (Malte-Brun. Geographie
1857, 6, p. 687.) In the interstream areas of the trap surface large
tracts frequently depart widely in their slopes from the sculptured
forms produced by running water. The surface becomes undulating
with saucer-shaped pits always opening out on one side towards the
drainage way of the district. As the small basins become deeper and
more numerous, inosculating rounded ridges rise between them, giv-
ing such tracts the appearance of New England kame-kettles and their
winding kame-ridges. In the pits there are shallow lakes or pools.
These depressions and ridges are evidently the work of long continued
secular weathering of the basalt combined with the removal of the
products of disintegration and decomposition.

Most of these basins were at the time of my visit more or less
occupied by standing water, some of them forming shallow lakelets
in which grew a brilliant green grass. Other basins presented the
appearance of level meadows from the filling of presumably residual
clay and vegetable matter which they contained. At Sao Joao on the
trap plateau south of Porto da Unido, a pit some three or four feet
in depth had been dug in one of these floored depressions, showing
beneath a few inches of vegetable matter a light colored clay evi-
dently the product of decomposition of the trap rock.

98 BULLETIN: MUSEUM OF COMPARATIVE ZOOLOGY.

The exposures of trap on the rounded ridges which separate the
basins are usually crusted with superficially segregated oxides of iron,
and loose weathered blocks are not infrequently seen in positions to
indicate the almost complete absence, for a long geological period in
the past, of any transporting agency such as sliding snow, ice, or run-

A B

Fic. 27.— Basins of decomposition on the Triassic trap plateau. A, cross-
section of basin overhanging a stream valley. B, contour map of lakelet
converted into a swamp. Santa Catharina.

ning water. There is thus no reason for supposing that the basins are
due to other causes than deep secular decay and the slow wasting away
of the rock under a moderate rainfall. That these weathered basins
are of great antiquity is obvious from the consideration of the mode
of origin which thus may be ascribed to them. There is no clear local
indication of the geological date of beginning of the basins. Inasmuch
as they abound on the surface of the Corisco lava-flow above described
they, in this instance, are more recent than the erosion of the overlying
sheets of trap. I saw nothing in them by which to distinguish
Pleistocene from Tertiary processes unless it be the deposits of clay
which would argue for probably a Tertiary date as the time of begin-
ning of the corrosion, but they may be early rather than late Tertiary.

Such solution-basins are not limited to the trap plateau but are
to be seen here and there on the Permian area in Sao Paulo where
springs find their way to the surface.

Mr. T. A. Allen (Derby, 1906, p. 888) has described pot-holes often
of great size and containing water, in the gneisses to the east of the
Serra do Esperanco on the plains of Bahia. He regarded these pits
as due to a peculiarly localized action of disintegration.

Professor Pumpelly (1879, p. 136) has called attention to the manner
in which deep secular weathering followed by a period of active erosion
as by ice would result in the production of a topography quite unlike
that of normal land sculpture by streams. He notes that “as masses

(

'
}
4

WOODWORTH: GEOLOGICAL EXPEDITION TO BRAZIL AND CHILE. 99

of decomposed rock may be observed to a depth of over 100 feet, the
surface of the still solid rock underneath presents ridges and hollows,
succeeding each other according to varying durability under thé
influence of percolating carbonated water. In this kind of weathering,
where erosion does not come into play, it is evident that the resulting
topography must, in some important respects, differ from that of an
ordinary surface of superficial denudation. In particular, rock basins
may be gradually eaten out of the solid rock. These will remain full
of the decomposed material, but any subsequent action, such as that
of glacier ice, which could scoop out the detritus, would leave the
basins and their intervening ridges exposed.”

VII. GEQMORPHOLOGY OF SOUTH BRAZIL.

In the preceding chapters so much has been stated concerning the
form and relief of the tableland of south Brazil, in describing the
structure and position of the Permian glacial beds, that little remains
in treating specifically a sketch of the geomorphology of the region
than to summarize the matter in more systematic terms with the
added enumeration of certain details.

Regarded as a land form south Brazil is an elevated tableland with
a short steep slope descending to and below the Atlantic sea-level and
a long gentle slope towards the interior of the continent. The surface
of this warped mass appears to have been in Cretaceous times much
more nearly a plane. Since its elevation and warping it has been
dissected by streams which have etched out the structure of the
westward dipping beds of the long westward slope into lines of escarp-
ments formed of the edges of the harder more resistent beds over-
looking lowlands.

For convenience of treatment the region may be divided into two
districts which by their geological nature and relief at once impress

_ the visitor to Brazil. First, the steep coastal border of the Serra do

Mar, and second, the tableland or planalto proper.

The Serra do Mar, notched and pinnacled in the states of Rio de
Janeiro and Sao Paulo, declines to the southward, and in Santa Catha-
rina retains more of the character of the warped surface of mature
relief which appears to have been characteristic of the whole belt in
an early stage of its development following the warping above noted.
In eastern Parana the summits of the Serra do Mar form a line of
peaks and ridges rising well above the eastern portion of the tableland

100 BULLETIN: MUSEUM OF COMPARATIVE ZOOLOGY.

or of the locally baselevelled tracts developed betweer
the more resistent members of the tableland. The an-
nexed diagram of the mountain crests as seen rising
above the lowland east of Curityba is typical of the
region for many leagues northward into the state of Sao
Paulo.

Isolated summits in this region rise somewhat above the
level of the trap plateau. It is probable that in the late
Mesozoic baselevelling of the region the granitic bosses
and some of the gneissic areas were not reduced to the
general level. On thenorth the lofty Serra da Mantiqueira
culminating in Mt. Itatiaia nearly 10,000 feet in eleva-
tion, warrants this statement.

The slope from the crest of the Serra do Mar to the sea
is generally steep. It is deeply ravined by short streams.
The interstream areas form sharp spurs which in some
portions of the slope are deeply dissected, standing out as
isolated peaks and mountain blocks as on the south side
of the harbor of Rio de Janeiro. The immediate descent
to the sea is often so precipitous and the relief so high
that where the geological structure is permissive of the
hypothesis down-faulting on the ocean side has been advo- .
cated as by Dr. Derby as a factor in the production of
the topography.

This deeply dissected slope has been depressed beneath
sea-level since its dissection arrived at an advanced stage.
The submerged valleys form harbors and reentrants such
as those of Rio de Janeiro,Santos, Itajahy, Sao Francisco,
and Florianopolis.!. Since this depression in relation to
sea-level took place, a slight uplift of about ten feet (3
meters) has occurred, raising up in the form of a plat-
form about bay shores a recent deposit of littoral sands,

D
ZL

LLL

1 The frequent repetition of the circumlocutions one is obliged to
employ in expressing concisely the fact of our ignorance as to whether
the land has sunk or the sea-surface risen when reference is made to a
change of level of land and sea becomes intolerable in writing at length
of such matters. Suess’s terminology partially avoids the embarrass-
ment but does not provide a name for a change of level of land and sea.
The French term denivellement, a variation of level, suggests the use in
English of its natural equivalent delevelling in analogy with baselevelling.
A positive delevelling thus becomes a depression of the land in relation
to the sea-level, and a negative delevelling an apparent elevation of the
land in relation to the surface of the sea.

Fic. 28.— Crests of the Serra do Mar seen from the west on the tableland near Curityba, Parané.

WOODWORTH: GEOLOGICAL EXPEDITION TO BRAZIL AND CHILE. 10]

which also surround former rocky islets and unite them to the
mainland.

Off the’ coast from the entrance of Rio Harbor southwestward
to and beyond Laguna, low rocky islets of granite or gneiss reveal the
continuation of the dissected slope of the Serra beneath the sea.
These islands are of small size and mostly uninhabited except by
occasional fishermen or as they may be the site of a lighthouse. Navi-
gation between them is dangerous because of the fogs at certain
seasons of the year and for the reason that lesser islets, or rocks which
just reach the surface of the sea, are not wanting.

These off-shore islands are frequently partly surrounded by a plat-
form of rock rising a few feet above sea-level. These benches corre-
spond to the elevated strips about the bays and probably indicate the
extent to which the subaerial upper portion of the island was reduced
by weathering and the attack of the sea above the level of these
platforms during the episode of maximum depression.

Bordering the inner shore of the bay of Paranagua, there are rem-
nants of siliceous sands of presumably Tertiary date which appear to
have been deposited within the dissected slope of the Serra do Mar,
showing that the basal portion of the coastal slope was well dissected
before the close of the Tertiary periods.

Professor Hartt found evidence at various points along the coast
that an elevation or negative delevelling had recently taken place.
Whether the coast from Rio de Janeiro southward is now under-
going a slow delevelling or not I could not ascertain. Faint traces
of an old beach not now reached at high tide near hogy seein favors
the idea of a recent elevation there.

The hypothesis above advanced of the warping up of the south
Brazilian plateau with the axis of curvature along the Serra do Mar
belt and the reference of the now dissected surface of a former pene- ~
plain to a Cretaceous date is based upon the fact that in the region of
Bahia the Cretaceous strata extend far inland. It is thought that
similar conditions prevailed on the south during a stage in the evolu-
tion of the present topography, that with the negative delevelling and
the development of a steep slope to the sea and a gentle slope towards
the interior the Cretaceous strata were denuded in early Tertiary
time. The Tertiary basin in the valley of the Parahyba between the
Serra do Mar and the Serra da Mantiqueira is taken as evidence of
uplift and dissection of the Cretaceous peneplain prior to the deposi-
tion of the Tertiary fresh-water beds. The definite determination
of the date of the Tertiary deposits would fix the limits of time to be

102 BULLETIN: MUSEUM OF COMPATATIVE ZOOLOGY.

placed upon the epoch of dissection. (Branner, 1906, p. 283. A. S.
Woodward, 1898, p. 63-75). Some warping of the surface has appar-
ently taken place since the Tertiary beds were deposited.

At the northern limits of the region under discussion the Serra da
Mantiqueira rises as a long lofty monadnock range in the Pre-Devonian
terrane. The western slope of the crest of this gneissic mass subtends
the surface of contact of the Permian of Sao Paulo upon the same
ancient rocks. It therefore appears probable that in the Serra da
Mantiqueira we have a remnant of the Permian floor east of the
present line of outcrop of those strata.

Between the Serra do Mar and the escarpment formed by the
westward dipping Devonian sandstone cuesta! there is a high level
tract belonging to the planalto or tableland. On the South in the
headwater region of the Iguassi about Curytiba it is essentially a
peneplain and swamps of great extent exist locally. But in south-
eastern Sao Paulo an Atlantic stream, the Rio Ribeira de Iguape, has
breached the Serra do Mar and gnawed a deep ravine with its head-
waters pushed against the crystalline Serra da Paranapiacaba for a
watershed. The situation of this stream, the single example of any
size to push its headwaters past the Serra do Mar and drain the
planalto, in the great concave are formed by the Paranapiacaba and the
Devonian sandstone cuesta is evidently an effect of the Devonian
sandstone ridge. The almost level crest of this Serra indicates the
approximate level of the Cretaceous peneplain up to which level the
plateau was filled with rocks before the present valleys and widened
out lowlands of the planalto were excavated. Under these conditions
the Devonian sandstones must have extended much to the eastward,
possibly to the Serra do Mar crest. In the dissection of the country
east of the present retreatal escarpment of the Devonian sandstones
the Rio Ribeira de Iguape, protected from capture by the westward
flowing streams, has worked backward following the shifting of the
watershed formed by the retreating cuesta until its headwaters are in
a position almost to capture the Rio Yapo and the uppermost Iguasst.
(See map, Fig. 7, p. 48.) The short course and steep gradient of the
river have enabled it to cut its present profound ravine. A similar
history probably is true of the Rio Tuberao which in southeastern
Santa Catharina has excavated its valley across the granitic terrane,

‘The term cuesta used in a technical sense in North American writings on geo-
morphology for obvious reasons is not adoptable in Portuguese. ‘‘Costa do outeiro"
misses the point in the English use of the Spanish name cuesta. Serra monoclina
expresses in structural terms the essential characteristic of a cuesta.

+
f

WOODWORTH: GEOLOGICAL EXPEDITION TO BRAZIL AND CHILE. 103

corresponding there geologically to the Serra do Mar, and which has
its watershed in the lofty crest of the Triassic escarpment on the
west of the Permian belt of coal-bearing shales.1_ The Ypiranga, an
affluent of the Bay of Paranagua, is apparently in the act of tapping
the westward sloping surface of the planalto back of the Serra
do Mar.

According to Drs. Derby and Euzebio P. de Oliveira the Permian
sandstone overlaps the Devonian shales at places in Parana and rises
towards the crest of the Serra as at Vilha Vehla, a point of deeply
disintegrated rocks southeast of Ponta Grossa, the name Vilha Vehla
being applied in the same sense that “Rock City” is employed in
North America. This ridge dies out northward in Sado Paulo and
southwards in Parané so that in Santa Catharina the geologically
higher escarpment of the trap plateau on the south comes in as the
first range inside the coastal mountain belt.

The relatively even sky-line of the Devonian cuesta forms a strik-
ing feature in the treeless landscapes for many miles in Parana.
The summit attains elevations exceeding 3,000 feet and approaches
the level of the hypothetical Cretaceous peneplain out of the nearly
level surface of which the lowlands between the ridges and plateaus
of the planalto have been sculptured by Tertiary and more recent
erosion. In northern Parana and southern Sao Paulo the topographic
relief becomes complicated by the erosion of longitudinal valleys and
by the association of ridges developed on the lower Permian sandstones
which succeed on the west of the Devonian outcrop. The streams in
this district also have cut deep gorges across these ridges apparently
along lines of drainage inherited from the time when they flowed on
the Cretaceous peneplain. Such appears to be the origin of the
defile in the massive sandstones through which the Yapo flows in
the country northwest from Castro to Tibagy. At Joachim Murinho
Station the railway follows a broad gap in the sandstones which
appears to have been once occupied by a river much larger than the
present stream. The escarpment of the Devonian here becomes
very irregular by reason of dissection. The steep sandstone cliffs
form precipices overlooking lowlands excavated to the level of the
crystalline rock-floor as at Pirahy. A characteristic view is to be had
from Fabio Rego Station looking towards the cliffs of the Serra
Morumgaba surmounted by the Morro do Chapeo.

1 For some account of the Ribeira de Iguape with maps and plans see a report
entitled Exploracio do Rio Ribeira de Iguape. Comm. Geog. e Geol. de 8S. Paulo,
1908.

104 BULLETIN: MUSEUM OF COMPARATIVE ZOOLOGY.

The Devonian shales in central eastern Parana give rise to a longi-
tudinal valley but the non-resistent character of these rocks is locally
counteracted by intrusions of diabase. The western side of this
longitudinal valley, where it is developed, is formed by sandstones and
conglomerates of the basal Permian. Ponta Grossa stands on the
western side of this valley.

The sandstone and conglomeratic members of the Permian form an
irregular grouping of uplands separated by river valleys. The level
at which the hilltops stand, intermediate between that of the lowlands
and the trap plateau and the Devonian cuesta probably indicates
an intermediate stand of the land, the date of which is difficult
to determine. The valleys of such rivers as the Tieté in Sao Paulo
certainly have been excavated since the Tertiary deposits which
underlie the city of that name. It seems highly probable therefore
that the uplands in the Permian tracts antedate not only these Tertiary
deposits but also the erosion of the depressions in which the beds were
accumulated. The date of the evanescent peneplain with which the
summits of the upland areas accord must be placed therefore in early
Tertiary time.

The westward flowing drainage of the planalto gives rise to the
dissection of the Permian terrane by long westward aligned valleys
such as those of the Paranapanema and the Rio Negro, but the
tributaries of the latter river including such large streams as the
Tibagy flow obliquely across the trend of the Permian belt on courses
expressing the resolution of the double control of the westward dip of
the formation on the one hand and of the slope towards the master
stream developed by concentration of the drainage on the other. The
Rio Negro displays a rectangular adjustment of its course to the
strike and dip of the Permian strata.

In the latitude of Curityba there is traceable westward from the
vicinity of that city a divide between the waters which drain northward
into the Paranapanema and southward into the Rio Iguasst. This
watershed includes the Serrinha, passes south of the towns of Pal-
meira and Iraty and thence joins on the west the trap escarpment
known as the Serra da Esperanca. The great extent of the drainage
basin of the Rio Paranapanema as compared with that of the com-
bined Iguassii and Rio Negro in the Permian terrane is apparently a
consequence of the antilinal axis, which, normal to the are of the
outcrops in eastern Parandé and southern Sao Paulo, caused the
Palaeozoic beds along this east-west line to stand relatively high and
erode more rapidly. As a result of this distribution of the drainage

a

WOODWORTH: GEOLOGICAL EXPEDITION TO BRAZIL AND CHILE. 105

the headwaters of the Rio Paranapanema drain almost entirely the
western slopes of the Devonian beds, as does the Ribeira de Iguape the
crystalline terrane on the east of the Devonian sandstones.

The Permian tract is bounded on the west by the Triassic escarp-
ment of sandstones crowned by trap sheets. The westward dip of the
formation combined with the flow of the streams in that general
direction has caused the trap to retreat far to the west along the axis of
the main drainage lines, such as that of the Paranapanema and the
Iguasst. In the western part of the south Brazilian states of Sao
Paulo, Parana, and Santa Catharina the rivers flow over the trap
sheets whose resistance to erosion holds up to their local baselevel
the entire drainage area of the planalto. The Rio Parandé on the
confines of Brazil and Uruguay is gnawing back the southern edge of
the trap sheets. Below the cascades and falls the river joins the
drowned valleys of the La Plata system.

The rate at which the falls of the Parana and Iguassti are receding
has not I believe been determined. But it is evident that the rivers
have cut back from the southern edge of the trap sheets since the land
had something like its present elevation above the sea. If the land
at the confluence of the Iguassti and the Parana had been as long above
sea-level as it has in the upper valley of the Iguassti, where the trap
has been swept away over a large tract of the Permian, it is incon-
ceivable that the youthful characteristic of falls and cascades should
still persist. We are therefore compelled to conclude that the country
immediately adjacent to the Parana and Paraguay rivers has recently
been uplifted in relation to the sea. As on the south of the trappean
country in Banda Oriental there are marine Tertiary beds now above
sea-level (Darwin, 1846, p. 1-3) and as along the coast of Brazil
from north of Rio de Janeiro to the Amazon (Derby, 1907, p. 218-237)
there are evidences of uplift since the Tertiary beds were there laid
down, it seems a valid hypothesis that the excavation of the Parana
channel in the traps began in later Tertiary time through an uplift
of the whole planalto of Brazil.

The upper courses of small streams in eastern Sao Paulo and
northern Parana generally flow in narrow gorges so recently cut that
many side streams particularly of the wet-weather type enter by a fall
over the brink of the gorge. Without a thorough understanding of
the local baselevels of the Parana system it seems out of the question
to infer the cause of this revived stream-action.

. A remarkable example of one of these streams is the Rio Itararé
flowing north into the Paranapanema along the boundary between

106 BULLETIN: MUSEUM OF COMPARATIVE ZOOLOGY.

the states of Parand and Sao Paulo. Near the town of Sao Pedro de
Itararé at the railway bridge this river flows in a channel which at one
point in its cross-section is not more than three feet wide.

The channel lies in the white Devonian sandstones which present
no great variation from layer to layer offering opportunity for the

y

RO. pan

4 La

iH | Hii itt He Imo

il il i FN nd

Fia. 29.— Map of the Parahyba and Tieté rivers in Sio Paulo (After H. Williams).

selective solution which in limestone countries often produce similar
gorges. The bottom of this gorge is said to be between 62 and 63
meters below the railroad bridge.

At one point west of the railway bridge there is a natural bridge
of the sandstone which evidently points to the origin of this gorge
as an underground stream.

It remains to note the curious course of the Parahyba in relation to
the headwaters of the Rio Tieté in eastern Sao Paulo. The annexed
map, traced from that of Sao Paulo by Mr. Horatio Williams, late of
the Sao Paulo Geographical and Geological Commission, sets forth
the pattern of the streams. (Fig. 29.)

It will be noted that the upper course of the Rio Parahyba under
the name Parahytinga follows a southwest course to the great bend
at Guarerema whence the course is northeastward to the sea beyond
the limits of the map. These courses are in essential adjustment to
the structure of the underlying Pre-Devonian rocks but the basin of
the river below the great bend is largely formed by the Tertiary non-
marine beds before mentioned. The great bend is made by a trans-
verse gorge cut through the Pre-Devonian series which rise a few
hundred feet above the riverplain. It is therefore to be presumed
that the course of the river at this point is inherited from a former
course which lay at the level of the intervening hill-tops. This earlier

WOODWORTH: GEOLOGICAL EXPEDITION TO BRAZIL AND CHILE. 107

“stage of the river presumably was developed on the surface of the
Tertiary beds.

The divide between the Tieté at Mogy das Cruzes and the great
bend is occupied by rock-hills of low relief rising about 200 feet above
the weakly developed drainage lines of the district. The natural
course of the Parahytinga would appear to be westward into conflu-
ence with the Rio Tieté of which it may be regarded as a beheaded
portion, captured by the Rio Parahyba, which, pushing its head
southwestwards along the easily eroded Tertiary beds, diverted the
stream before erosion had swept away the Tertiary beds between the
Parahyba basin and that of the Tertiary beds at Sao Paulo.

The Pleistocene and Recent Formations.— The discrimination of the
Post-Tertiary changes in extra-glacial regions into Pleistocene and
Recent is attended with difficulties. In Brazil the surface deposits
are prevailingly residual clays or clays, sands, and pebble beds derived
from the secular washing and transportation of weathered Pre-
Pleistocene formations. Great differences exist as to the depth of the
decayed rock even on the same formation. Where the rainfall is
heavy at certain seasons of the year, the slope of the ground steep,
and the run-off effective, the decayed materials are removed nearly as
fast as their disintegration or decomposition is accomplished and thus
nearly fresh rock occurs at the surface. I was frequently surprised
in the valleys of the Rio Negro and the Rio Tuberao by the apparent
freshness of carbonaceous shales at a depth of a few centimeters be-
low the surface but in these situations erosion has been and still is
actively in progress.

The new cuts of the railway in construction from Bury in Sao
Paulo via Sao Pedro de Itararé and Jaguariahyva to Ponta Grossa
in Parand gave at the time of my visit an unusual opportunity to see
many excavations in the mantle rock. Along the banks of the Rio
Jaguaricatu in the Permian tillite beds these cuttings were often from
5 to 10 meters deep. At these depths most of the pebbles were still
undecomposed.

At numerous localities along the railway line across the mature
topography of southern Sao Paulo and Parand the rounded swells
between streams display tracts of ancient gravel beds usually with
concave lower limits as if occupying old stream channels long since
abandoned. The same phenomenon is observable in a pronounced
manner where the railway from Ponta Grossa to Serrinha Station in
Parana skirts the lower westward slopes of the sandstones of the
Devonian cuesta. In all these cases the history of the surface appears

108 BULLETIN: MUSEUM OF COMPARATIVE ZOOLOGY.

to be as follows: — The streams and wet-weather wash concentrate’

pebbles in their beds leaving the interstream swells or rounded ridges

comparatively free from coarse material. The finer decomposed ma--
terial of the interstream areas becomes more readily eroded than the

coarse débris in the stream channels and erosion proceeds more rapidly

Fic. 30.— Showing supposed stages, A. B. C., in the concentration of gravels
in creases and the deposits on ridges between creases on the dip-slope of
pebbly sandstone beds.

along them so that they become depressions between the old deposits.
Concentration of coarse débris begins again in the now new well-.
defined creases which carry off the rainfall. So far as my observations.

go there have been but two cycles of such gravel accumulation, an

ancient one and that now actually taking place. If the process.
depends solely upon the relative resistance to erosion of the gravel-
bearing creases and the gravel-free ridges between streams the change

may well be automatic under a constant rainfall during the period

of alteration. After one such shift the interstream areas become:

partly gravel-capped and an equilibrium is established which at first
did not exist. Thus the dual character of the phenomenon in Brazil

may be due to this limitation inherent in the nature of the process,

even with a variable rainfall.

In the case of the deposits of this nature on the Serrinha near

Tamandud in Parana the ancient gravels occur in abundance and
appear to exceed in thickness those of the present wet-weathered

channels. That these older gravels in this climate are as old as the
Tertiary period seems to me improbable since under the conditions:

of exposure to weathering they must have broken down. At best they

might be Pliocene, but if it is admitted that the glacial epochs of the

Pleistocene were signalized in south Brazil by a heavier rainfall than

that now prevailing it is probable that the older gravels represent one:

= sebbbi i> @if

WOODWORTH: GEOLOGICAL EXPEDITION TO BRAZIL AND CHILE. 109

of the older glacial episodes comparable to that of the Columbia

gravels south of the terminal moraine on the Atlantic slope of North
America.

A short distance west by south from Araucaria Station in Parand
near kilometer post 26 on the railroad from Serrinha to Curityba I

Lf S
;

f
- °
. _ ke oS é iy ae ~~ .
s . ~ : ~ . . ~
, . ”-~ . Pid .

Fig. 31.— a, cross-section of the road cut in which the upper dotted line
gives the present slope, the lower one, the slope on which the old gravels
were deposited; b, cross-section of the spur on right of railway; c, theo-
retical restoration of both sides of a gravel-bearing hill, the upper portion
being removed in the segregation of the gravels.

noted, in a railway cut through a spur, evidence that these old gravels
were deposited on a steeper slope than that of the modern neighboring
ereases. The relations of the ancient and modern profiles are shown
in the annexed diagrams. The gravel bed was evidently deposited in
a crease, the axis of which corresponds with that of the present spur
as above explained. From the consideration of such a case it be-
comes evident that the existing smoothened contour of the ground is
due to the removal of the crests of old spurs and ridges and the ac-
cumulation of the coarse detritus near the bases of the slopes, and thus
that a topography of sharper outlines preceded the present cycle.
This means a considerable lowering of the elevations of the region
since the old gravels were deposited and confirms the idea of an early
Pleistocene date at least for the age of these deposits.

On the baselevelled plain about Curityba where rounded ridges of
mature dissection modified by deep weathering are interspersed
between shallow water courses the cuts show many signs of the gentle
processes of solution by which the surficial rocks have been removed.
In one such cut near the Meteorological Station in the west bank of

110 BULLETIN: MUSEUM OF COMPARATIVE ZOOLOGY.

a small valley there was exposed in 1908 in the road the section
diagrammatically shown in Fig. 32.

A quartz vein about 2 inches (5 cm.) thick dipping E. 30° in a
section of decomposed schists had given rise to a sheet of quartz
fragments at the base of the residual structureless surface deposits

eS
ates

Fic. 32.— Train of residual quartz fragments derived from a vein during the
weathering and ablation of the crystalline schists. Near Curityba, Paran4.
a—b, a distance of 50 feet from the outcrop of the quartz vein to the limit
of fragments; c—d, the supposed surface at which the vein outcropped.

traceable fully 50 feet (15.2 M.) to the westward on the gently inclined
surface of the schists. The original aerial extension of the quartz vein
from the data here presented must have been at a height of 29.5 feet
(8.1 M.) above the present surface as may be readily determined by a
calculation of the right angle triangle. The two to three feet of
overlying structureless residual clay may or may not represent the
breaking down of about thirty feet (9.15 M.) of rock above the present
surface of the schists. In either case solution by percolating water has
been the chief agency in denudation. If the removal of this thickness
of rock went on at the average rate for such drainage areas as have
been studied — a rate as great as one foot in 3,000 years, the time
represented in this case for the lowering of the quartz fragments is
approximately 88,500 years, a period which takes us back according
to the newer! estimates to the close of the last glacial epoch in North
America. ‘To this estimate should be added the time for the accumu-
lation of the overlying clays whose superposition on the quartz

1 The most recent studies of the retreat of the Wisconsin ice-sheet and the glacial
lakes and marine phenomena which succeeded the glacial retreat demand from 5 to
10 times the 10,000 years of earlier estimates.

Tere

WOODWORTH: GEOLOGICAL EXPEDITION TO BRAZIL AND CHILE. 111

fragments seems to point to some shifting of the residual clays. The
definite termination of the band of quartz fragments at a distance of a
few yards from the outcrop of the vein means that this old surface
was at the beginning of the period of weathering swept clean of quartz
fragments, either because of a steeper slope than that now found at
the locality or by reason of a more powerful run off. This latter
possibility is consonant with the hypothesis of a heavier rainfall
during the Pleistocene, however much the above attempt to calculate
the time employed may vary from the true duration of Post-Tertiary
epoch.

River Terraces.— Along several of the larger rivers of south Brazil
there are terraces of sand and gravel evidently remnants of a former
aggraded floor of their valleys. These deposits date back presumably
to the Pleistocene with its greater rainfall.

For examples, a terrace occurs in the Rio Iguassii between kilometer
posts 47 and 48 along the railway between Araucaria and Balsa nova
Stations in Parana; a terrace also occurs north of Balsa nova at km.
post 59. Other fragments of this terrace occur along the river
further up the valley.

The Rio Capivary in Parané between Lago and Palmeira exhibits
a terraced plain.

In southern Sao Paulo a gravel terrace of old river cherts is crossed
by the railway between Herval Station and Engenheiro Hermilho
Station.

These gravelly terraces, apparently of the same epoch as the
Tamanduaé gravels on the hillsides, are probably in their later stages
derived from the washing down and gullying of these deposits.
Owing to the nature of river changes it is improbable that the forma-
tion of the terraces by reexcavation of the old valley floors should
have been synchronous throughout the area under discussion.

Numerous cuttings along the railways in Sao Paulo and Parana
show that the clayey deposits there, varying but little from the terra
roxa and the terra vermehlo, are not strictly residual but are rather
transported or shifted, however much they have decayed in their
present sites. Dr. Derby expressed the opinion that the red earth,
of which an excellent exposure was examined at the railway station in
Sao Pedro de Itararé on the confines of Sao Paulo and Parana, was
an equivalent of the loess of other regions.

The thickness of these deposits varies greatly. Many sections
were seen varying from six to ten feet, but in many the bottom was
not exposed. The material appéars to be developed particularly

i BULLETIN: MUSEUM OF COMPARATIVE ZOOLOGY.

along the lower slopes of the hills as if washed down by rains during
the wet season. Much dust is blown about by the winds in the dry
season and doubtless an eolian origin may be attributed to some of
the particles.

The relations of the red earths to the underlying pebble beds indi-
cate pretty clearly the order of magnitude of the powers of the rainfall
in the immediate past and the changing climatic conditions, a heavy
rainfall with a strong run-off followed by a marked weakening in this
agency. Wind action if registered in the loess-like red earth hardly
can be called upon in the case of the gravels in old creases. Lag-
gravels are typically developed upon wind-swept plains; besides
glyptoliths or sand-carved pebbles are not here forth-coming; farther
north in Brazil Dr. Lisboa has found them.

These deposits are related to each other in the range of dynamic
force concerned as are the Pleistocene glacial gravels to the Post-
glacial alluvial deposits of many North American sections. Hence the
probability that the gravels represent the Pleistocene. From an
excellent exposure in Parané at Tamandua Station I propose the
name Tamandua (anteater) beds for the gravels. As for the over-
lying shifted reddish earths whether terra roxa or not, they form a
group of surficial deposits blending in places with residual clays in
situ and do not so readily take a formation name.

Inrailway cuts in the white sandstones between Sao Pedro de Itararé
and Fabio Rego the red earth rests on the eroded surface of the white
sandstones. The sharpness of the contact between the two and the
absence of red coloring in the sandstones proves the shifting of the
superficial deposit with its coloring matter. The development of the
red oxide of iron would seem here to have antedated the transportation
of the material otherwise the red matter seemingly should have been
carried downward into the porous sandstones.

In the winter season of drought the red earth dries and cracks.
These cracks on the surface of newly cut banks by the railway stations
often form a hexagonal network. Similar cracks form over the surface
of the campo. As leaves, sticks, and insects peculiar to the existing
flora and fauna readily fall into these cracks to some depth in the
clays, it is obvious that by the closing and opening of these cracks
under the changing seasons any contemporaneous fossils they may
be found to contain must be carefully discriminated from post-
depositional entries.

Canga is a superficial segregation of oxide of iron or limonite in various
geological positions. On the road from Ponta Grossa to Conchas in

WOODWORTH: GEOLOGICAL EXPEDITION TO BRAZIL AND CHILE. 113

Parana it forms crusts in the surficial decomposed portion of the bed
rock. In a railway cut on the banks of the Iguassii near Serrinha
Station in Parand the outer portion of the Carboniferous rocks is
crusted with canga and this vein-like material there occurred in
joints. The segregation of canga by percolating water seems mainly
to be ancient; it may be older Pleistocene or still older, and probably
is not peculiar to any one episode of the modern geological history of
the region.

The canga in some localities appears to have been broken up and
transported, now occurring as rubble in the red surface deposits as
between Sao Pedro de Itararé and Fabio Rego. In this case the canga
must have been segregated prior to the transported red earth, and
if the red deposit be assigned a Pleistocene date the canga may be
referred to the Tertiary.

The decomposed state of the rocks in Brazil was early recognized
and correctly described by José Bonifacio de Andrada e Silva and
_Martim Francisco Reibeiro de Andrada in an account of a mineralogi-
cal journey from Santos to the tableland made in 1820. This article
is reprinted in Ferreira’s Diccinario geografico das Minas do Brazil.
Rio de Janeiro, 1885, p. 341, 342. |

W ceather-blocks — The weathering of the granites along the coastal
slope of the Serra do Mar has led to the production of thousands of
rounded weather-blocks which are particularly evident at the present
sea-level and just above within the zone of wave action, tuose above
the present sea-level having had the fine material between them
removed in part at a time when the land stood a few feet lower than
it now does. Abundant examples are to be seen about the shores of
Rio Harbor. The illustration, Plate 2, is from a photograph of a
group of blocks-on the shore of Sao Francisco Harbor in Santa Catha-
rina. In Madureira, a suburb of Rio de Janeiro, near Cascadura
Station there is fine large block said to be movable (pedra movedi¢a)
poised high up on a weathered tower of granite. The famous Furnas
de Agassiz at Tijuco in the Serra near Rio de Janeiro is another group
of weather-blocks, the most imposing to be seen anywhere in south
Brazil.

It remains to note certain rock benches and the uplifted fringing
coastal plain to be seen along the shores from Rio de Janeiro to near
Laguna on the south.

From observations made about Sao Francisco Bay in latitude 26° S.
I suspect that there exists along this coast an old bench of marine
erosion 150 or more feet above the present sea-level. Numerous rock

114 BULLETIN: MUSEUM OF COMPARATIVE ZOOLOGY.
'

islands along the coast at variable distances from the shore rise to
nearly this height and about Sao Francisco Bay well in back of the
town there are low foot hills separated by valleys from the base of the
Serra and from each other, and similar hills confront the somewhat
higher ones at the south side of the entrance. Presumably this

avaan

OL eee

. Fie. 33.— Terrace about islet in the sea north of Laguna, Sta. Catharina.

dissected rock bench is of very ancient date, early Pleistocene or
Tertiary.

A more recent set of rock benches form narrow platforms about
many rock islands along this same extent of coast apparently standing
from eight to ten feet above sea-level. They agree closely in level
with the alluvial plains bordering the bays. Figure 33 gives the out-
line of such a terrace skirting the base of a small rocky islet north of
Laguna as seen from a steamer. Hartt (1870) has given abundant
evidence of a recent uplift of the coast to this height. This change
of level is seemingly very recent.

Between the first described signs of a very ancient bench, and this
recent uplift must be interpolated an episode of subsidence carrying
the sea into the valleys at the base of the Serra do Mar. The harbors,
great and small, are due to this change of level. The filling up of the
harbors and river channels and the building of an underwater deposit
makes an estimate of the depth of this depression too small but as the
harbor at Rio de Janeiro has a maximum depth of 30 M. (Hartt,
1870, p. 7) the sinking must have been equal to this depth plus the
amount of the recent elevation.

On the flats east of Paranagua there is a well-defined low beach
ridge covered with dead shells of Ostrea and a smaller gibbous lamelli-
branch at an elevation of eight or ten feet above sea-level and separated
from the shore of the bay by two flats at slightly different levels.
The shells are not worn; some of them have both valves in position.
The situation of the deposit and the mode of occurrence of the shells
is very different from that of the accumulation of shells left by
aborigines on neighboring sandy deposits.

Geographic Control of Human Occupation.— The effect of the several
geographic features above outlined on the human occupation of

re ae

WOODWORTH: GEOLOGICAL EXPEDITION TO BRAZIL AND CHILE. 115

south Brazil are more or less patent to every visitor. The dissection
of the coastal slope and the depression of the resulting Serra do Mar
has given rise to commodious harbors so uniformly wanting because
of the unlike geological structure and form on the west coast of South
America. But the Serra do Mar renders ingress to the country
exceedingly difficult and possible for roads and railways only along
certain routes. Back of the Serra do Mar for the most part the lands
slope toward the interior of the continent, and the large rivers, naviga-
ble by small boats, serve only with ease to carry commerce into the
interior. The transportation to the coastal border of the plateau is
everywhere upgrade making the export of the products of plantations
and the forest more costly than the importation of foreign goods,
an item of cost which is offset on the inward journey by the necessity
of ascending the Serra do Mar, and, to reach the trap plateaus, of
surmounting the Triassic escarpment. Transportation is naturally
slow to develop, except where peculiar conditions, such as give rise
to the rich coffee-fields of Sio Paulo, have repaid the construction of
railways.

The recently uplifted plains of alluvium bordering the harbors afford
the sites for the first settlement of the sea-coast. The variable relief
of the dissected front of the Serra do Mar furnishes stations for resi-
dence at altitudes great enough for Europeans, whose affairs require
their daily presence in the federal capital, to escape the languishing
effects of a continuous abode in the hot zone at sea-level, but the way
to these retreats calls for special and costly methods of transportation,
as in the case of the route to Petropolis. Owing to the mountainous
relief of the coastal slope of the Serra do Mar and the luxurious growth
of tropical vegetation the inhabitants enjoy outlooks unsurpassed in
any land. As a scenic route for the traveller the railway journey
from Paranagua to the summit of the Serra through the defile of the
Ypiranga is surprisingly pleasant, and at many points exciting.

The contrast between this region and the surface of the plateau is
striking. The tableland is the seat of production. Variations in the
geological character of the surface modified by altitude and rainfall
come sharply into play. In northern Sao Paulo and adjoining parts
of Minas Geraes the soils known as terra roxa and terra vermehlo
developed by decomposition of the trap sheets which invade the
Permian terrane afford under the peculiar conditions of rainfall there
existing the richest coffee-fields in the world. Farther south the
Devonian and Carboniferous sandstones present less favorable
conditions. Open campos or prairies characterize much of the region
underlain by Palaeozoic strata.

116 BULLETIN: MUSEUM OF COMPARATIVE ZOOLOGY.

Much has been said as to the origin of the treeless condition of this
region. About Ponta Grossa in Parana the persistence of the elements
of the forest along the brooks and rivulets and near the water courses
would seem to point to the oft repeated supposition of a diminished
rainfall following the glacial period as the probable cause. In my
notes on the surface deposits I have presented some reason for thinking
that traces of such a change in the rainfall and run-off are observable.

In Parané and Santa Catharina farming and cattle raising find
suitable conditions and here the influx of European settlers from
Germany, Poland, and in Rio Grande do Sul from Italy, has under the
more temperate climate of the upland wrought commensurate changes
in the appearance of the country.

Mention has already been made of the harborage to lingering
remnants of hostile natives which the Triassic trap escarpment affords.
Farther in the interior larger bodies of aborigines favored by the
unnavigable rivers made inaccessible from their lower courses by
reason of the numerous falls over the trap sheets maintain to a large
degree the primitive state of the Brazilian highland.

VIII. NOTE ON THE CHANGES OF LEVEL OF THE COAST
OF SOUTHERN CHILE.

For more than seventy years Darwin’s raised beaches and terraces
of the west coast of South America have been generally regarded by
English-speaking geologists as typical examples of a relative change
of level of land and sea. Sir Charles Lyell by embodying the observa-
tions and conclusions of Darwin in his classic Principles of geology
gave wide distribution to the views of Darwin concerning the magnitude
and extent of the supposed recent elevation of the west coast of South
America. <A dissent from the views of the justly celebrated naturalist
of the “ Beagle” was scarcely heard in geological circles until in 1885,
when Edouard Suess brought out Das Antlitz der Erde, in which work
he sought to show by testimony mainly that of observers of the
Chilean coast since Darwin’s voyage, that no noteworthy elevation of
this coast in recent times had taken place. Upon the publication of
the French translation of Suess’s great work, La Face de la Terre, the
writer (J. B. Woodworth, 1898, p. 803-806) in a review expressed a
disbelief in the sufficiency of the evidence brought against Darwin’s
observations and conclusions. In fact, I went to Chile rather for the
purpose of studying the nature of the movements, whether by frac-

WOODWORTH: GEOLOGICAL EXPEDITION TO BRAZIL AND CHILE. 117

ture and faulting or otherwise, which must have attended this supposed
‘elevation, than to confirm or refute any conclusion as to the elevation
of the coast at a time so recent as that embraced within the period of
elevation of beaches, so well established in the Champlain or Hoche-
lagan marine district of northeastern North America. The following
notes and conclusions thus have the force of being made with a pre-

THE
CORDILLERA. { LONGITUDINAL as
CORDILLERA, {LONGN

Fic. 34.— Diagrammatic cross-section of southern Chile west of the Andes.
Based on Pissis and personal observation in the longitudinal valley and
Coastal Cordillera.

conceived opinion quite opposite in most respects to the judgment
which in the sequel I was obliged to make. But the geological struc-
ture bearing on changes of level should be first set forth.

The geological structure of Chile south of Valparaiso is in its broad
outlines readily grasped in such a rapid reconnaissance as I had the
opportunity to make. The main facts were clearly outlined by Pissis
(1875) from whose work and my own observations the following brief
introduction will serve to make clear the relations of the Coastal
Cordillera with its frequent earthquakes, and the chain of the Andes
lying back from the coast. |

In south Chile for a great distance north and south there are three
well-defined topographic and structural zones, viz.: — the Andesian
folded chain with volcanic vents, the Longitudinal Valley confronting
the chain on the west, and the Coastal Cordillera forming the coast.
These three differently constituted areas are shown diagrammatically
in the subjoined cross-section, Fig. 34.

The formations which enter into this section were roughly deter-
mined by Darwin (1891) as follows: —

118 BULLETIN: MUSEUM OF COMPARATIVE ZOOLOGY.

Darwin’s scheme of geological formations in Chile.
(Compiled from “Geological observations.’’)

Newer Tertiary, includ-
ing the basin-plains

of Chile.

Concepcion formation.

Superficial saliferous deposits of Iquique.
Salt and nitrite of soda, 3,000 ft. above
sea-level.

Beds of gravel, red sandy clay; lava,
pumice beds, overlying calcareous tufa.

Sandstones with lignites, silicified wood,
and “concretions”? carrying marine
shells. Eruptions.

Upper Cretaceous?

Cretaceo-Oolitic; Gyp-
seous formation.

Tuffs and submarine lava of Uspallata
section; conglomerates of valley of
Tenuyan.

Upheaval in central Chile forming
Cumbre and other ranges.

Coarse conglomerates, siliceous sand-
stones, dark mudstones, limestone,
pseudohornstones, tuffs, and vast beds
of gypsum, with marine fossils and sili-
cified trunks of fir-trees at Los Hornos.

Eruptions of purplish claystone and
greenstone porphyries, etc.

Metamorphic series.

Metamorphic schists, plutonic rocks, and
more or less altered clayslate.

The more recent work on the geology of Chile has been summarized
by Mr. Lorenzo Sundt (1909, p. 37-44). According to this geologist,
the sedimentary formations now recognized in Chile are as follows: —

WOODWORTH: GEOLOGICAL EXPEDITION TO BRAZIL AND CHILE. 119

Geological formations of Chile, mainly from Sundt.

Pleistocene. Eolian and other deposits of Atacama on
the north.
Glacial till and gravels in the south.

Eocene. Conglomerates and sandstones.
Lignitic group of Coronel.

Cretaceous. Sandstones of Quiriquifia Island, and
along the coast.

Conglomerates, sandstones, and_ black

calcareous, fossiliferous shales of Co-

quimbo and the Pequenes Range.

Jurassic.
Triassic Not recognized..
peesnview Devonian. Choapa.

Metamorphic series; mica schists of
Coastal Cordillera south of Caldera.

Sub-Carboniferous or Fossiliferous beds at mouth of the Rio
Devonian or older.

The Andes consist of deformed, squeezed up, and dissected beds of -
the Tertiary and older rocks, with eruptives. The Longitudinal
Valley comprises Tertiary and possibly older beds overlain by lavas
and Pleistocene gravels the preglacial members of which series appar-
ently dip gently westward until cut off by the eastern fault of the
Coastal Cordillera. The Coastal Cordillera, made up of the meta-
morphic series and granite intrusions, is fringed by Cretaceous and
Eocene deposits in faulted relations with the older terrane.

The Coastal Cordillera forms a long narrow tableland parallel to the
coast with an elevation from 1,200 to 1,500 feet, the flattish upper
surface of which is evidently a peneplain from which more recent beds

120) BULLETIN: MUSEUM OF COMPARATIVE ZOOLOGY.

have been denuded. These beds appear to have been the same ~
Cretaceous and Tertiary strata which occur in patches along the coast
in dislocated positions. The borders of the Coastal Cordillera are
determined by faults so that the Cordillera is itself a long relatively
uplifted fault-block or “horst.’”’ Relative to it, the western side of
the Longitudinal Valley on the east has sunk as has also the Pacific
Ocean border on the west. It is along the fault-zone on the west,
partly if not altogether, that the epicenters of great Chilean earth-
quakes mainly lie as was the case with that of Coquimbo in 1906,
which destroyed Valparaiso and neighboring towns.

It follows from this structure of the west coast of South America
throughout the greater part of Chile that the earthquakes which take
place along the faulted edges of the Coastal Cordillera arise from the
movements of this horst and are not directly connected with the
folded chain of the Andes on the east. Accepting as I think we must
the essential accuracy of Darwin’s and Fitzroy’s observations upon the
elevation of the coast about Concepcion at the time of the earthquake
of 1835, it does not follow, however, as Darwin thought, that the
Andes were simultaneously elevated.

In the Valparaiso earthquake of 1906 Professor Steffen has found
evidence of a local elevation amounting to half a meter on the western
border of the Cordillera at that place; but it is a question whether the
elevations which take place at the time of these earthquakes are
permanent. At Concepcion it is the opinion that a subsidence follows
the uplift of the coast. Nevertheless the horst stands as an indubita-
ble block of evidence of uplift in relation to the ocean bottom and the
Longitudinal Valley.

The date of beginning of the faults which bound the horst of
the Coastal Cordillera is Post-Eocene and apparently late Tertiary.
Whether movement has taken place in modern times along the fault
on the east I am not able to say. The Pleistocene gravels at San
Rosendo abut against the horst without signs of disturbance; and the
rivers, such as the Bio Bio, Calle Calle, and Maule, have cut channels
across it on their way to the sea from a superposed position. Neither
the Rio Bio Bio nor the Calle Calle, through whose valleys I passed in
traversing the width of the horst, display rapids such as might be due
to recent uplift at a greater rate than that of the cutting power of
these streams to maintain a free grade to the sea.

It seems most probable that the eastern fault bounding the Longi-
tudinal Valley is of early Pleistocene or late Pliocene date, and that
the present valley is due to erosion of the lavas and sediments part
passu with the cutting of the gorges.

WOODWORTH: GEOLOGICAL EXPEDITION TO BRAZIL AND CHILE. 12]

If one reads Chapter X on the Plains and valleys of Chile in Dar-
win’s Geological observations, he will find the author convincing
himself that the sea rather than the rivers now coming out of the
Andes is accountable for nearly every detail of the land form,— not
only the passes in the mountains and the valleys sculptured in the
mountain mass but also the plains of waterworn detritus particularly
where rounded pebbles are found.

This statement tallies with the general belief of the school of
geologists of the period in which Darwin formed his concepts. It was
naturally easy for him to admit terraces of any origin and marine
shells on the surface at lower levels along the coast of Chile as evidence
of elevation when he had found to his own satisfaction that the sea
had beaten against the high Andes where no recent marine shells are
found. ‘This much is necessary for the reader to understand in weigh-
ing the conclusions which may be presented concerning the elevation
of the coast of Chile.

It is only just to Darwin, however, to show from his own writings
how he abandoned the views he entertained concerning topographic
signs of elevation on the coast of South America when in later years
similar problems came to his attention in the British Islands.

Darwin (Francis Darwin, 1903, 2, p. 191) wrote Lyell in 1861,—
“T return Jameson’s capital letter. I have no comments, except to
say that he has removed all my difficulties, and that now and for
evermore I give up and abominate Glen Roy and all its belongings.
It certainly is a splendid case, and wonderful monument of the old
ice period. ... How many have blundered over those horrid shelves!”

Again in 1868, Darwin (Francis Darwin, 1903, 2, p. 211) wrote
Croll,— “I was formerly a great believer in the power of the sea in
denudation, and this was perhaps natural, as most of my geological
work was done near sea-coasts and on islands.

“But it is a consolation to me to reflect that as soon as I read Mr.
Whittaker’s paper on the escarpments of England, and Ramsay and
Juke’s papers, I gave up in my own mind the case; but I never realized
the truth until reading your papers just received. How often have
I speculated in vain on the origin of the valleys in the chalk platform
round this place, but now allis clear.” Still he clung to his interpreta-
tion of terraces and of shells on the surface in South America. As late
as 1872, Darwin wrote Lyell (Francis Darwin, 1903, 2, p. 164-165),—
“It seems to me very cool in Agassiz to doubt the recent upheaval
of Patagonia, without having visited any part of it; and he entirely
misrepresents me in saying that I infer upheaval from the form of the
land, as I trusted entirely to shells embedded and on the surface.

122 BULLETIN: MUSEUM OF COMPARATIVE ZOOLOGY.

“It is simply monstrous to suppose that the terraces on a dead level
for leagues along the coast, and miles in breadth, and covered with
beds of stratified gravels 10 to 30 feet in thickness, are due to sub-
aerial denudation.”

Darwin seems here to have forgotten what he wrote on the shells
embedded in a friable calcareous rock in the terraces of Coquimbo at
a height of 250 feet. “Although I examined so many hundreds of
miles of coast on the Pacific, as well as Atlantic side of the continent,
I found no regular strata containing sea-shells of recent species,
excepting at this place, and at a few points northward on the road to
Guasco. This fact appears to me highly remarkable; for the expla-
nation generally given by geologists, of the absence of stratified
fossiliferous deposits of a given period, namely, that the surface then
existed as dry land, is not here applicable; for we know from the
shells strewed on the surface and embedded in loose sand or mould,
that the land for thousands of miles along both coasts has been lately
submerged.” (1845, p. 344.)

The interpretation of the shells strewed on the surface and embedded
in the loose sand or mould constitutes the essential difference between
Darwin’s hypothesis of their marine deposition and the theory that
such surficial deposits pertain to the group of kitchen-midden.

My own observations of the coast of Chile were made about Corral
and Valdivia in latitude 40° S., at Taleahuano and Concepcion, in
latitude 35° 35’ S., and with less scope about Valparaiso in latitude
Say tek

The Coast at Corral and Valdivia. Of Valdivia, Darwin (1891, p.
236) states “I did not observe any distinct proofs of recent elevation;
but in a bed of very soft sandstone, forming a fringe-like plain, about
sixty feet in height, round the hills of mica-slate, there are shells of
Mytilus, Crepidula, Solen, Novaculina, and Cytheraea, too imperfect
to be specifically recognized.”

The plain or rather terrace which Darwin mentions extends from the
open coast outside of Corral entrance (Plate 29), about the sides of
the harbor of Corral and up the Valdivia River to the city of that name.
It is well shown about the sides of Manzera Island in the harbor of
Corral (Plate 30), and is a distinct feature along the banks of the
Valdivia River (Plate 31). It is a terrace of denudation sometimes
cut in the soft Tertiary sandstones mentioned by Darwin, sometimes
in conglomerates as at Corral, and sometimes in the schists. About
Valdivia and Teja Island this terrace I estimated to be fifty-five feet
above the sea-level and the same along the Rio Cruces Channel. The

cig
—

ae

WOODWORTH: GEOLOGICAL EXPEDITION TO BRAZIL AND CHILE. 123

same level holds for the seaward edge of the terrace on the south side
of Corral entrance facing the open sea, where, however, the surface
rises inwards towards the hills, being in part covered by deposits at
the mouths of gullies. The terrace thus has the appearance of being
a few feet higher on the coast than inland. At most, the terrace is a
few rods in width and carries here and there a few houses. On
the south side of Corral entrance near English Bay, there are drongs,
or stacks of the schists standing up on the surface of the terrace,
evidently sea-stacks left by marine erosion at about the sixty foot
level. Isaw none of these along the Valdivia River or inside of Corral
Harbor. I have no doubt that this bench marks a stand of the sea
when the coast stood relatively lower than now. Above this terrace
the hills are sharply ravined by long continued sculpturing of the
schists and no signs of beaches or wave action were found. ‘The next
higher plain is that of the summit of the Coastal Cordillera (Plate 29),
which is undoubtedly an old peneplain of Tertiary or older date.

The lack of beach deposits or signs of modern occupation of the
fifty to sixty foot terrace by the sea makes it necessary to place \its
origin in early Pleistocene or possible Pliocene times. It is long
anterior to such beaches as occur about the shores of the Champlain
or Hochelagan sea in northeastern North America.

South of Corral entrance near a group of houses I saw marine shells
on the terrace in the sod but they were in such a situation as to render
it altogether probable that they had no geological significance, being
nothing more than refuse of the aborigines.

At Valdivia and Corral it seems to me demonstrable that there was
in early Pleistocene of late Tertiary times a relatively lower stand
of the land than now by about sixty feet during a period long enough
for the river to widen out its valley on a rock-floor of schists and older
Tertiary sediments in a disturbed position. Whether these Tertiary
sediments occupy an old valley in the schists or owe their position to
down-faulting I can not state from the observations which I made.
Their disturbed attitude and the presence of cleavage in the con-
glomerates at Corral favors the view of infolding and down-faulting
of the Tertiary beds in the schistose terrane. After this period of
partial baselevelling the land rose about sixty feet relatively to the
sea.

The presumption that the channel of the Valdivia River and the
depression forming Corral Harbor have been excavated since the
uplift of the terrace began is also an argument for the ancient date
of the sixty foot terrace. The depths of water in the Valdivia River

124 BULLETIN: MUSEUM OF COMPARATIVE ZOOLOGY.

and connecting channels vary from forty-eight to ninety feet and
depths of seven and eight fathoms occur in the Rio Valdivia opposite
the town. While glacial gravels and boulders not far inland as at
Llanco Station show the presence of Pleistocene glaciers or glacio-
natant waters, I noted no distinct signs of glacial erosion of the
channels about Valdivia. It seems to me more probable, therefore,
that these depths indicate a depression of the coast since the formation

of the terraces rather than that they are due to glacial erosion inde--

pendently of the sea-level.

Below the fifty-five to sixty foot level, I noticed a small terrace on
Teja Island opposite Valdivia at a height about twenty-five feet
above sea-level, but this was not elsewhere observed except possibly
as noted below.

Along the coast south of Corral entrance near Palo Muerto Pt.,
there are several ancient sea-caves not now visited by the sea at high
tide. Near them is a trace of a bench at about 22 feet. The mouths
of these caves are not over ten feet above sea-level (Plate 33). The
bottoms of the caves are covered with a fine dusty cave-earth which
covers any pebble bed which may occur on their bottoms. One of
the caves (Plate 33) had been at the time of my visit recently occupied
by natives and a heap of shell refuse formed a considerable mound at
the entrance. In the same vicinity at Morro Gonzales there is a
natural bridge in the rock of the sixty foot terrace formed by a cave
penetrating to a small ravine in the rear. In front of the caves at
a few points where the coarse Tertiary conglomerate is developed
there is a flat or bench, cut back across these rocks, whose surface is
at least partly covered by high tide. It is evident that a slight
relative rise has taken place removing the caves from the action of
the sea, but while this may have been as much as ten feet it may not
have been over five feet as the depth of cave-earth would seem to
show. In embayments south of Corral entrance, there are old beach
ridges the highest of which comes sharply against the ravined surface
of the land behind it. The uppermost of these ridges agrees well with
the level of the mouths of the caves and does not indicate a rise of
more than ten feet and possibly not more than five feet above sea-
level; but, as Suess has pointed out, on the west coast of South
America beach ridges may be thrown up by seaquake waves or
tsunamis above the normal marine limit and thus are to be taken with
caution.

At Talcahuano and Concepcion. The literature concerning eleva-
tion about the Bay of Concepcion relates largely to the discussion

“WOODWORTH: GEOLOGICAL EXPEDITION TO BRAZIL AND CHILE. 125

-concerning the change of level accompanying the earthquake of 1835.
As Darwin noted, little or no evidence of the elevation claimed by him
and Fitzroy could be seen by a subsequent visitor. The Old Fort
at Penco (Old Concepcion) is shown in Plate 4. The alleged evidence
-of elevation prior to this famous incident pertains largely to marine
‘shells found on the hills about the Bay of Concepcion. Another
feature which undoubtedly is the result of relative uplift of the land
is the alluvial plain upon which the modern city of Concepcion stands.
* Ulloa (1772, 2, p. 252-254) who visited Concepcion Bay as early
as 1744 describes the occurrence of shell deposits from six to twelve
feet and more in thickness within four of five leagues from the shore
-and also on the tops of hills fifty toises (820 feet) high. He mentions
seeing one deposit at a height of twenty toises.

Darwin (1887, p. 310) speaks of “the vast number of sea-shells
scattered over the land, up to a height of certainly 600, and I believe,”’
he states, “of 1000 feet.”” He appears to have entertained no doubt
that these shells were evidence of an elevation of a former sea-bottom
to these heights as explaining the occurrence of the shells.

The shell deposits described by Ulloa in the plain about Concepcion,
though I did not see them, presumably are in their natural site as
they form a part of the tilted plain of the “Formation of sand and
shells” described by Pissis. This plain at Concepcion I estimated by
aneroid to be forty-five feet above sea-level. Westward towards
Talcahuano there are successive lower levels, that at the golf links
I made seventeen feet, and the next lower level adjacent to the bay of
Concepcion about ten feet. The surface has been more or less modified
by the blowing of sands by the winds. I saw several boulders in the
upper dark stratified sands of the plain between Concepcion and
-‘Taleahuano along the railway cuts. One of these blocks was eighteen
inches in diameter. These boulders were in an embedded position
-and would seem to demand floating ice for their transportation. It
is to be presumed that the water-laid portions of the deposits are
‘therefore of Pleistocene date.

Back of Concepcion the plain of sand abuts against the base of
‘the Coastal Cordillera at the hill called Cerro Caracol. I searched in
vain for any definite beach along the upper limit of the plain, but
-stream-action would long since have obscured such local deposits at
this juncture. Above this level to the top of Cerro Caracol the
granitic rock displayed only evidences of long continued atmospheric
weathering and the vertical lines of dissection worked out by running
water. Nowhere did I observe signs of the Pleistocene stand of the

126 BULLETIN: MUSEUM OF COMPARATIVE ZOOLOGY.

sea much above the level of the plain. A small cave in the granite
of the hillside in the public park by its angle of slope into the rock is
‘ shown to be artificial and having nothing in common with sea-caves.

A visit to Penco on the north of Concepcion gave little or no evi-
dence regarding a change of level. ‘The plain of dark sands is here
wanting, that deposit being largely an ancient delta of the Rio Bio Bio.
At the north end of Penco Beach, on a point of rock cut off by the
railway cut from the bluff of Tertiary sandstones which here form the
shore, at about twenty feet above sea-level I found shells thickly
strewn in the soil together with fragments of mammalian bone, a
typical mode of occurrence of Indian shell-heaps.

Talcahuano and Tumbres Peninsula I have given some account
of the coast near Tumbres Peninsula in my itinerary (p. 32). There
are good sea-caves on the inside of the Paps of Bio Bio at the present
sea-level but none above. The surface of the Tumbres Peninsula is
covered with a brownish weathered layer containing chips of quartz
and schist derived from the underlying rock. Waterworn pebbles
are absent except in situations about to be described as village sites.
There are sloping surfaces of the bed rock at discordant levels and
abundant traces of water action in cutting vales, but none of the
horizontal lines marking the prolonged action of waves in cutting
cliffs or heaping up beach materials.

On the southeast side of Tumbres at about 330 feet I found round
patches of marine shells in the sod of the tosca or surface deposit.
The shells in the first example seen occupied a space about twenty-one
feet in diameter in which the grass, in December, of a light color
because it was dead, showed clearly the shell covered area because all
about the shell tract a species of sorrel with a reddish stem and fruit
grew in abundance. This plant avoids the shell deposits probably
because of the lime. Observing this character of the vegetation, I
went to other tracts which from a distance showed the absence of the
sorrel and found them in each case the site of a thin shell deposit.
One of these shell deposits was upwards of forty feet in diameter. In
one of these areas I picked up a stone pestle, a cylindrical beach
pebble battered by wear at both ends. With the shells were a few
broken beach pebbles which seemingly could have had no use, but
such pebbles were altogether absent in the weathered soil and tosca
rock outside of the shell deposits. It is evident that these deposits
are of human origin. Darwin mentions shell deposits on Sentinella
Hill at about 400 feet elevation reported to him by Dr. Jenks, the
Assistant surgeon of the Beagle. At no point on the peninsula was I

Otis dar Dek,

WOODWORTH: GEOLOGICAL EXPEDITION TO BRAZIL AND CHILE. 127

able to detect anything like Pleistocene beaches or the work of waves
at elevations above that of the plain of Concepcion and this plain is
practically the only evidence of a former stand of the land lower than
the present sea-level since the time of the Tertiary and older deposits
which contain marine fossils.

Darwin was familiar with the fact that the natives carry large
quantities of shell-fish inland, but in commenting on the fragments of
shell-bearing forms such as sea-urchins found several miles inland from
the coast he doubted the human origin of many deposits on the ground
that it was improbable that the natives would transport such materials
to any distance from the coast. On the contrary, it is well known that
sea-urchins are an article of food in Chile. I saw in the market at
Tamuca baskets of a large sea-urchin with the spines attached. The
natives use for food many shell-fish which the unnaturalized European
does not think of utilizing and if one objects that in the Pre-Columbian
days such stores of food would decay in the course of a distant inland
journey due allowance must be made for savage tastes. The “hung”’
meats and festering cheeses of civilized countries show how much
latitude must be allowed for aberrant tastes. In the semi-arid and
arid climatic zones of the Coastal Cordillera many Mollusca would
become dessicated rather than putrid because of a belated consump-
tion. In fact the unusual distance to which shell accumulations are
found back from the coast of Chile appears to me quite compatible
with the physical environment of an aboriginal people whose hinter-
land in the Longitudinal Valley was in primitive times less fruitful
than the adjoining sea. Kitchen-middens occur here in every degree
of accumulation from the now sparsely scattered shells in the sod
marking a temporary abode to the thick shell-heap on the site of
permanent habitations. Shells on the surface or buried in the super-
ficial subaerial deposits along this coast are no more a criterion of
submergence beneath the sea than elsewhere and cannot be relied
upon as proving anything other than the agency of man.

The Red Soil.— Along the western slopes of the Coastal Cordillera
there are large tracts with a red soil, the characteristic effect of the
subaerial decomposition of rocks. The situation of these residual
deposits, the nature of the topography without beaches or marks of
wave action, and the abundant traces of the secular leaching by rains
and moisture, together with the gullying of the ancient surface on
which these deposits occur, alike bespeak ordinary atmospheric
processes as conditioning the formation of this red earth. The
occurrence of shells and other exuviae in this deposit is but an indica-

128 BULLETIN: MUSEUM OF COMPARATIVE ZOOLOGY.

tion of man’s occupation of the area for a few centuries perhaps in
the past.

Pissis (1875, p. 78-81) has described the distribution of the most
recent marine elevated deposits of Chile. Beginning on the north
in the plain which extends from the Bay of Antofogasta to that of

POS Joe

ose OS...
oe ree ee eae rid

MMAR EN ic 8 EW a
PRR DL a ceaaian: RE aede a e
/33 —_] —— qt

caldera |i Coquimbe— [st 2
ie a a A a
32 | cttta. Be =~ ~~», Concepcion [* —— 40
RRS HRT PERE Bee Te P
Miles. 0 199 200. 200 400 +500 600 700 800 809 1000 }100 1200

Fic. 35.— Rise to the north of the Recent uplifted marine deposits of Chile,
according to the data of Pissis (1875, p. 78-81). The dotted line shows
apparent warped surface from Caldera southward. The line a—b gives
the mean tilt-rate from Mejillones southward. The line c—d is drawn at
the same tilt-rate to touch the elevated rock bench at Corral.

Mejillones this plain, having there an altitude of forty meters, is
represented by correlated deposits at constantly lower levels on the
south as far as the 38th parallel. In the vicinity of Caldera the
deposits rise from twenty-five to thirty meters above sea-level.
About Valparaiso and San Antonio the level is maintained between
fifteen and twenty meters. The city. of Concepcion is built on the
sands at the mouth of the Rio Bio Bio and finally at Levau the de-
posits risé scarcely two to three meters above the sands of the beach.
Taking these data as given by Pissis and plotting the upper limit
between Mejillones and Levau, the tilt-rate is shown by the graphical
construction in Fig. 35. According to these observations made by
Pissis as early as 1875 the coast from a point north of Valdivia shows
a rise at the rate of about 0.13 ft. per mile for about 1,100 English
statute miles.

Valdivia and Corral lying to the south of this group of raised
deposits as noted above shows evidence of depression preceding the
last slight rise of the coast. Accepting Pissis’s altitudes for the height
of the uplifted plain and assuming the above rate of tilting, the
abandoned sea-caves south of Corral fall on a tilted plane about three
meters above that of the formation of sands and shells, and may
tentatively be regarded as at the sea-level under which the sands were
laid down.

WOODWORTH: GEOLOGICAL EXPEDITION TO BRAZIL AND CHILE. 129

_ The fifty-five to sixty foot bench at Corral is at that place about
fifty feet above the line of the abandoned sea-caves. What the
attitude of this terrace is south of Corral I am unable to state. That
the depression of the coast continues increasingly southwards is
possibly indicated by the extensive fiord zone however much these
channels may have been over-deepened by the Pleistocene glaciers.
But Darwin’s (1891, p. 233) account of sea-caves not now visited by
the waves on the island of Chiloe, though these small caves may be
more recent than those of Corral, shows the necessity of a resurvey of
this coast before assuming any uniform rate of tilting over great

distances.
= ieee z
2 EE TG NS We
7s SE OS
Bie ae canara HCE sé
panies ete ais 3
pcre tg reese Sut x%
. See ee a <7 | a
PE 28 AE Ee & by ys
CS De i Se See “as = - <s
oo eee nN: Oo» <4)
2S SS SaaS EE SEE <a
Se a7 D
22 ae “JE G
CS Se Ie + Re R
2 ———————————E a é ;
ey eagiene 3 v Gil
SS >. SR ONCEPCIN: re ngegM
———$_<—_—— = os *¥
eae = sa eee a ;
Bi Is = aa ee See eChiguayanta ee
SSeS = .Gualgui We ,
Senta Marve LC ORONELY iS
— = op OZ
cekenee esses feta = EQuilacoya.
= Ss SS ate (inTa (co mavidaZ
—~--— — aoe ee ef OH Bi fn, = :
- = 0 Sao Rio Laja
rae q <7 Cf
mee 2 - 5S. ROSENDO WS f4
= Nn ;
wate best ers = bas
Sees a ‘2
OE eit & =
ee oo cera re Az S,
= <S zd
ae eae eG =
= xt
= oe WwW!
Sa , aa et
—-LEBU- ae
<¢ Mitel Jide sar N
PP 1]

Fie 36.— Sketch map of country about Concepcion, Chile.

Terraces of the Rio Bio Bio within the Coastal Cordillera.— From the
delta plain of the Rio Bio Bio at Concepcion there is a series of river
terraces within the cafion of the river rising in altitude to San Rosendo
at the border of the Longitudinal Valley. The railway from Talca-
huano to San Rosendo follows the river, rising at a very uniform grade
now on a lower terrace and again passing through cuts in the margins
of a higher terrace. The following notes were taken on a railway
journey from Concepcion to San Rosendo. The elevations of the

130 BULLETIN: MUSEUM OF COMPARATIVE ZOOLOGY.

railway stations and distances from the terminus at Talcahuano are
from the official map of the railways of Chile. The estimates of the
elevations of the terraces are approximate only except at San
Rosendo where measurements were made with an aneroid barometer.

Talcahuano, O. kms. Altitude 3 M. (9.8 ft.). This is the elevation
of the border of the delta of the Rio Bio Bio.

Los Pexales, 7 kms. alt. 6 M. (19.6 ft.). This is a point on the
plain midway between Taleahuano and Concepcion.

Concepcion, 15 kms., alt. 9 M. (29.5 ft.). This is the elevation of
the railway station near the river. The plain rises north of this point
to an elevation of about forty-five feet.

Chiguayante, 25 kms., alt. 18 M. (59 ft.). This Station is within
the gorge where the river flows northwest. Above this point a stony
terrace appears.

Gualqui Station, 38 kms., alt. 21 M. (68.8 ft.). Terraces of allu-
vium judged to rise from thirty to thirty-five feet above railway, or
100 feet above sea-level. At about 120 feet above sea-level there is a
sloping dissected terrace in the decomposed bed rock of the gullied
sides of the cafon. The fine sand deposit continues as a terrace
bordering the river, apparently a continuation of the deposit forming
the Concepcion Plain. Above this terrace the side walls are deeply
gullied.

Quilacoya Station, 48 kms., alt. 25 M. (82 it.). The terrace here
rises to about 105 feet above sea-level, at its edge. The railroad runs
on a deposit of the dark sands. Old rock remnants of a sloping terrace
are to be seen on the ravined walls of the gorge, apparently indicating
an old baselevel of the river before noted as visible above Gualqui
Station. :

Talcamavida Station, 62 kms., alt. 34 M. (111.5 ft.). The fine dark
sands form a terrace rising twenty to twenty-five feet above the
Station or about 130 to 135 feet above sea-level. The general trend
of the gorge from this point to San Rosendo is southeast. Above
this Station in the terrace large angular stones, cobbles, and gravel
appear, indicative of flood conditions with ice-rafts,— typical glacial
river deposits.

Gomero Station, 69 kms., alt. 38 M. (124.6 ft.). The edge of the
terrace is here about twenty feet above the railway, or about 145 feet
above sea-level. Heavy gravel beds are exposed along the river above
this point, with waterworn cobbles.

Bunaraqui Station, 76 kms., alt. 41 M. (134.5 ft.). Top of terrace .
about twenty feet higher, or about 155 feet above sea-level.

WOODWORTH: GEOLOGICAL EXPEDITION TO BRAZIL AND CHILE. 131

Malvoa Station, 79 kms., alt. 43 M. (141.0 ft.). Good terrace on
opposite or southwest side of river. At railway station there are clays.

San Rosendo Station, 85 kms. alt. 46 M. (154.8 ft.). East of the
Station there was in 1908 an excellent exposure of gravels and sands
in a deep cut, rising about sixty-five feet above the railway, or to an
elevation of 220 feet above sea-level. The annexed
section is from a drawing made on the spot.
(Fig. 37).

This section reveals a succession of river deposits
varying from sands like those now transported in
this portion of its course by the Rio Bio Bio to
coarse gravels including large boulders such as
demand ice-rafts or the proximity upstream of
glacial conditions.

The lowest bed exposed (see Fig. 37) consists
of dark sand with pebbles of a voleanic rock.
Above this comes a coarse gravel including a
boulder of gneiss about seven feet long. Next
in the section is sand with thin bands of volcanic
pebbles. Above these layers comes about twenty-
five feet of dark sands with a thin band of volcanic
pebbles. Surmounting this and forming the sur-
face is a layer, about twenty-feet thick, of coarsely
bedded gravels and boulders, the cross-bedding of py6. 37. — section of
which suggests the structure of a delta front. Pleistocene terrace

There are in this section thus the records of two is Brae ance
episodes when the river transported to this point
coarse gravels and boulders which it is to be presumed indicate con-
temporary advances of the local glaciers or times of unusual melting
and discharge of coarse débris.

I saw no marine shells in any part of this terrace or on its surface,
and no evidence of the presence of the sea in the deposition of the
materials. The structure is quite like that of many glacial river
deposits in the inland portions of glaciated North America. The
lower bed of coarse gravel with boulders must have been deposited
at or above sea-level as is also the case with the uppermost bed of like
materials. The aggradation of the valley under glacial stream-action
affords a simple explanation of these deposits without recourse to the
hypothesis of a change of level in relation to the sea.

This section is the most reliable in its bearing on the altitude of the
terraces, since those lower down the river were estimated only from

132 BULLETIN: MUSEUM OF COMPARATIVE ZOOLOGY.

the elevation of their edges as seen from the train and the tops of such
terraces may rise higher than the figures above given. From Con-
cepcion at the seaward mouth of the gorge to San Rosendo along the
river is a distance of 70 kms. (42.7 miles). Taking sea-level in
Pleistocene time at Concepcion as forty-five feet, the San Rosendo
terrace gives a uniform rise up the river of 4.3 feet per mile for the
grade of the river, not in the least too steep to be within the range of
‘streamaction for the entire distance above Concepcion. It does not
appear to me therefore that the terraces of the Rio Bio Bio within
the Coastal Cordillera demand a lower stand of the land than is indi-
cated by the delta plain at Concepcion.

The traces of a higher floor of the valley shown in the faint rock-
terraces within the gorge of the Bio Bio are of very ancient date.
They appear to be more deeply dissected than the rock-bench at
Corral with which it would be hazardous to correlate them. The
tops of the low hills of Tertiary rocks piercing the plain at Concepcion
are possibly nearer the level of the Corral bench, but these and other
questions concerning ancient changes of level of baselevelled rock
surfaces on this part of the coast of Chile await further study. I am
only able to state that as the result of my own observations it appears
that an elevation since late Pleistocene times has taken place at
Concepcion amounting to something like forty-five feet; that at
Corral and Valdivia an older Pleistocene or late Pliocene uplift of
about sixty feet is indicated, and that a recent uplift of a few feet not
more than ten and perhaps not more than five feet is indicated, and
that, in conclusion, the coast of Chile affords a largely open field for
the|determination of the changes of level which it has undergone in
relation to the sea during Pleistocene and late Pliocene times, involvy-
ing a thorough revision of Darwin’s studies on this subject in the light
of criteria which were not in his time applied to the study of shore-
lines.

IX. STONE IMPLEMENTS AND POTTERY FROM
LAGUNA.

By R. B. Drxon.

A number of specimens from an aboriginal camp-site near Laguna,
Province of Santa Catharina, have been placed in my hands for
description, by Prof. J. B. Woodworth. The specimens comprise
stone implements and pottery sherds. The stone objects are of three
varieties, (1) hammer-stones, (2) smoothing-stones, (3) and celts.

WOODWORTH: GEOLOGICAL EXPEDITION TO BRAZIL AND CHILE. 133

The hammer-stones are of a simple, world-wide type, slightly pitted
on two sides where held between the thumb and fingers. The smooth-
ing-stones are from two to three inches in length, and show on one or
more sides evidences of rubbing and polishing. They were in all
probability used in the manufacture of pottery, for smoothing the
irregularities of the surface, both inside and out. The celt is of basalt,
having a length of a little over six inches, with a width at the cutting
edge of two and a quarter inches, and tapering somewhat toward the
butt. One side of the implement shows somewhat greater convexity
than the other and the entire surface was finished by grinding.

The pottery sherds collected represent four somewhat different
types. The first of these is a smooth, reddish ware, undecorated,
and with a thickness of three eighths of an inch. It was apparently
used for bowl-shaped vessels of rather small size. A second type is a
very coarse, red ware, from one half to five eighths of an inch in
thickness. The broken quartz and gravel used for tempering is
extremely coarse. The surface of the vessels was covered with thumb-
prints, often of large size, where the coils had been pressed together
in the process of manufacture. The vessels were probably of large
size. A third type of sherd shows a heavy and rather coarse gray
ware, the surface covered by a buff or yellowish slip. This ware is
nearly an inch in thickness, and the smoothed surface is decorated by
painted designs in red. The figures are rectilinear, and seem to be
aigzags and grecques. These vessels also must have been of large
‘size. A single sherd also with painted decoration but of much thinner
ware, and evidently a fragment of a small bowl, with decoration
‘consisting of cross-hatching around the margin only, may be regarded
as related to the thick, heavy ware in type. One other type is repre-
sented by a sherd of dark gray ware, very hard burned. In thickness
it is similar to the smooth red ware, but the surface is decorated by
parallel rows of thumb-nail impressions.

The character of the sherds would seem to indicate that the site
was one belonging to Indians of the Guarani group of the Tupi. These
Indians are known to have occupied the region in this immediate
vicinity at the period of the earliest European contact, and sherds
resembling the second and third types.here described, have been found
widely throughout this coastal region as well as over large areas of
Uruguay and along the lower Parana.

134 BULLETIN: MUSEUM OF COMPARATIVE ZOOLOGY.

X. BIBLIOGRAPHY.

Alcock, Frederick.
Trade and travel in South America. 2nd. ed., London, 1907.
Bodenbender, G.

Sobre la edad de algunas formaciones Carboniferas de la Republica Argen--

tina. Revista del Mus. de la Plata, 1895, 7, pt. 1, p. 131-148.
Branner, J. C.

Decomposition of rocks in Brazil. Bull. Geol. soc. of America, 1896, 7,.

p. 255-314.

Geologia elementar, preparada con referencia especial aos Estudantes
Brazileiros. Laemmert & C. Editores, Rio de Janeiro e Sado Paulo,
1908.

Bibliography of the geology, mineralogy, and palaeontology of Brazil.
Bull. Geol. soc. Amer., 1909, 20, p. 1-132.

Brobritzhoffer.
History of the Abipones. 2, p.6. (Cited by Darwin).
Bross, H.

Glaziale spuren in Parana, Brasilien. Centralblatt fiir mineral. geol. und

palaeont., 1909, p. 558-561.
Campos, Cicero de.

Report of a geological survey of a part of the state of Parand. Brazilian

engineering and mining review, 1908, 5, p. 3.
Darwin, Charles.

Observations of proofs of recent elevation of the coast of Chile, made
during the survey of His Majesty’s ship Beagle, ete. Proc. Geol. soc.
London, 1837, 2, p. 446-449.

Journal of researches in the natural history and geology of the countries
visited during the voyage of H. M.S. Beagle round the World. 2nd ed.,
London, 1845.

Geological observations. London, 1846.

Darwin, Francis.
More letters of Darwin. 2 vols., New York, 1908.
Derby, O. A.

Mittheilung eines briefes von Herrn O. A. Derby iiber spuren einer Carbonen
eiszeitin Sidamerika. Neues jahrbuch mineralogie, 1888, 2, p. 172-176.

Decomposition of rocks in Brazil. Journ. geology, 1896, 4, p. 529-540.

The serra do Espinhago, Brazil. Journ. geology, 1906, 14, p. 388.

The sedimentary belt of the coast of Brazil. Journ. geology, 1907, 15,
p. 63-75.

Estudios geologicos en el Brazil. Trabajos del Cuarto congreso cientifico
(1° Pan-Americano), 1911, 11, p. 498-507.

Dutton, C. P.
Earthquakes. New York, 1904.

WOODWORTH: GEOLOGICAL EXPEDITION TO BRAZIL AND CHILE. 135

‘Eckhardt, Wilh. R.
Palaeoklimatologie. Leipzig, 1907.
Ferreira, Francisco Ignacio.
Diccionario Geographico das Minas do Brazil. Rio de Janeiro, 1885.
Fricker, Karl.
The Antarctic regions. London, 1900.
Geikie, Sir Arch.
Text-book of geology. 4th ed. 2 vols., London, 1903.
Gorringe, H. H., Lieutenant-Commander, U.S. N.

The coast of Brazil. Vol. 1. From Cape Orange to Rio Janeiro. U.S.
hydrographic office. Bureau of navigation. No. 43. Washington,
1873.

Halle, Thore G.

On the geological structure and history of the Falkland Islands. Bull.
Geol. instit. Univ. Upsala, 1911, 11, p. 115-229. Colored geological
map and 10 pls.

Hartt, Chas. Fred.

Geology and physical geography of Brazil. Boston, 1870.
Hooker, Joseph D.

Himalayan journals. 2 vols., London, 1854.
Humboldt, Alex.

Personal narrative of travels to the equinoctial regions of America.

Bohn edition, 2 vols., London, 1852.
Hunter and Rosenbusch.
In Tschermak’s Mineral. und petrog., 1890, 11.
Kaemtz, L. F.
A complete course of meteorology. Translated by C. V. Walker. London,
1845.
Lesley, J. Peter.
Dictionary of fossils of Pennsylvania. 4 vols., 1889.
Leyell, Charles.
Principles of geology. 11th ed., 2 vols., New York, 1887.
Malte-Brun.

Geographie universelle, mise au courant par Th. Lavalee. 6 vols., Paris,

1857. ;

Oddone da Roma, Emilio.

- Quelques constantes sismiques trouvées par les macrosismes. Assoc.
sismol. internat. bureau central. Strassburg, 1907.

Oldham, R. D.

Report on the great earthquake of 12th June, 1897. Mems. Geol. surv.
of India, 1899, 29, p. 1-xxx, 1-379, with 44 plates and 3 maps.

Pissis, A.

Geografia fisica de la Republica de Chile. Paris, 1875. Atlas de 23

laminas.

136 BULLETIN: MUSEUM OF COMPARATIVE ZOOLOGY.

Pumpelly, R.
The relation of secular rock disintegration to loess, glacial drift, and rock-

basins. Amer. journ. sci., 3rd ser., 1879, 17, p. 183-144.
Rozas, Alfonso Rodriguez, y Cruzat, Carlos Gajardo.
La catastrophe del 16 de Agosto de 1906 en la Republica de Chile. Santiago
de Chile, 1906. )
Rosa, Jose Vieira de. °
Chorographia de Santa Catharina. Florianopolis, 1905.
Russell, Thomas.
Meteorology. New York, 1895.
Sayles, Robert W., and La Forge, L. A.

The glacial origin of the Roxbury conglomerates. Science, 1910, new ser.,

32, p. 723-724.
Siemiradski, J. v.
Geologische reisebeobachtungen in Siidbrazilien. Sitzungsber. k. Akad.
wiss. Wien., 1898, 107, abth. 1, p. 23-39, pl. 1.
Simroth, Heinrich.
Die pendulationstheorie. Leipzig, 1907.
Smith, H. H.
Brazil and the Amazons, and the coast. New York, 1879.
Steffen, H.

Contributiones para un estudio cientifico del terremoto del 16 de Agosto de
1906. Publicado en los anales de la Universidad en Mayo y Junio de
1907, Santiago de Chile, 1907.

Sundt, Lorenzo.
Breve resefia de la geologia Chilena. In Eduardo Poirier: Chile en 1908.
Appendice, p. 37-44. Santiago de Chile, 1909.
Ulloa, Don Juan and Don Antonio de. ’
A.voyage to South America. 3rd ed.,2 vols. London, 1772.
Wallace, Alfred R.

A narrative of travels on the Amazon and Rio Negro. 4th ed., London,

1892.

White, I. C.
Relatorio final apresentado a S. Ex. o Sr. Dr. Lauro Severiano Miller.

Comissaio de estudias das minas de Carvado de Pedra de Brazil. Rio de
Janeiro, 1908. (Printed in Portuguese and English).
Whymper, Edward.
Travels amongst the Great Andes of the Equator. New York, 1892.
Wieland, G. R.
The Williamsonias of the Mixteca Alta. Botanical gazette, 1909, 48, p
441-442.
Woodward, A. Smith.
Consideracdes sobre alguns peixes terciarios dos schistes de Taubeté,
Estado de Sado Paulo, Brazil. Revista do Museu Paulista. 1898, 3,
p. 63-75.

WOODWORTH: GEOLOGICAL FXPEDITION TO BRAZIL AND CHILE. 137

Woodworth, J. B.
Review of La face de la terre by Edouard Suess. Science, 1898, new ser.,
7, p. 803-806.

Papers and notes pertaining to the Expedition.

Papers and notes pertaining to the Expedition.
Greene, Jerome, and Warren, Joseph.

[Correspondence relating to appointment of Thomas Barbour, Archibald
Cary Coolidge, and Jay Backus Woodworth, delegates to Pan-American
Congress at Santiago de Chile]. Boletin 2, IV Congreso Cientifico (1°
Pan-Americano), 1908, p. 118-119.

Ward, R. DeC.

Government meteorological work in Brazil. Monthly weather review,
1908, 36, August, p. 254-256; Sept., p. 290-292.

Notes on weather and climate made during a summer trip to Brazil, 1908.
Monthly weather review, October, 1908, 36, p. 333-339.

The southern campos of Brazil. Bull. Amer. geog. soc., 1908, 40, p. 652-
662.

The national exposition at Rio de Janeiro. Pop. sci. mo., 1909, 74,
p. 105-123.

An outline of the economic climatology of Brazil. Bull. Geog. soc. of
Phila., 1909, 7, April, p. 53-66; June, p. 185-144.

Woodworth, J. B.

Shaler Memorial Expedition. Economic geologist, 1908, p. 359-361.

La expedicion memorial de Shaler en Brasil y Chile. (Read by title).
Boletin, 3, [V Congreso Cientifico (1° Pan-American), 1908, p. 151.

The Shaler Memorial Expedition to Brazil and Patagonia, 1908-09. Amer.
journ. sci., 1908, ser. 4, 26, p. 404.

Report of the delegates of the United States to the Pan-American scientific
congress held at Santiago, Chile, December 25, 1908, to January 5, 1909.
Washington, 1909. Appendix I. Report on Section 3, subsection 4,
Geology and related subjects, p. 34-35.

Permo-Carboniferous conglomerates in south Brazil. (Abstract). Bull.
Geol. soc. Amer., 1910, 21, p. 779.

i<

are

ina

ie wether se

4

AR TP AES ea ) ne

by wid. Cent ad wigetied
\ Peay ANE Dike eetirret |
oe Sewanee at wales ee wet ae sb |

Tiitahe Ri Me tees tt art hae IR LURE TVS

—. \, brevet. tative asia ~ he ae
a ad ys oe fas Leet 3) ee ;
‘thondy spon ei hay eee ee

Het sockapal) delivers ley a

Lo opaietél STR Pitt te hoetetery enna

. is eed

Ay ag 4g a

>
) oat ye ere

Pies ‘od tt ab sth ER i ae 9 a “aire

¥
i

Ah 15 EN <i ed Novd

nites (eos, eat ate BaP iitg | ie J aa PRA 7
oy beef tet yn AB Oy! use yavartils 'ehait
reer gan? FERS
yl an ae ih SR Nese 3 tary: eth Bs
Bo) ai - Sas dhet 2
it at ood gee = ite a any Tu td diss iat cia! if Bra
ay 4 eg he Aloat 1 Baie 8 Pd lly treat

reat 2 mebalsens? (i aoe ifs

Woopwortx.— Geological Expedition to Brazil and Chile.

PLATE 2.

GRANITE BOULDERS DUE TO WEATHERING.

View on the shore of Florianopolis Island, Santa Catharina; on shore of bay
south of the city of Florianopolis. This type of surficial boulder, due to
secular weathering of the granite, is common in the granitic regions along the
slopes and at the base of the Serra do Mar, particularly at sea-level. The
well-known boulders of the ‘‘ Furnas das Agassiz’ at Tijuca, which with other
phenomena gave rise, at the time of the Thayer Expedition to Brazil, to the
now abandoned hypothesis of the glacial origin of these detached blocks,
belong in the same class of effects of secular weathering since Tertiary times.
Page 113. ;

‘“SMOdONVINO1S ‘SYSOQINOS G3SYSHLVSM

NOLS8SO8 ''OO0 3dALOIISH "HLYOMIGOOM ‘8 “fT AB OLOHd

"t Wid ‘IUD W [izerg 0} -dxq

Woopworta.— Geological Expedition to Brazil and Chile.

PLATE 3.

The campos of Sado Paulo, between Faxina and Sao Pedro de Itararé a
few miles west of the former place on the Carboniferous sandstone terrane,
showing the dissected peneplain of southwestern Sao Paulo, and mode of
travelling by “‘trolley.” Page 13.

‘JLNOY N3 ALYVd HLIM ‘O1NWd OVS 40 SOdWVD 3HL

*"NOLSOB ‘'OO 3dALONNSH *"HLHOMGOOM “8 ‘*f AB OLOHd

PS

"€ avid ‘IUD W pzvag 0} ‘dxq

-_
; ry Pe a
¥ f
a
@
.

«4 :
- r ] f :
r 2 1
ae ms re ;
’ F
°
‘a - f

APR RIA Pi) 5): : en ee ee 4

ee a yh ete re
A re ; as 11 Wee oe) : ; : : Ys
Se Gk a “ .

Woopwortu.— Geological Expedition to Brazil and Chile.

PLATE 4.

Ruins of the old Spanish fort at Penco (ancient Concepcion) on the eastern
shore of Concepcion Bay, destroyed in the earthquake and tsunami of May
24, 1751. View looking N. 56° E, mag., December 13, 1908. These walls
are the sole standing remnant of the old town of Concepcion. Page 33.

‘JMIHO ‘OON3d LV LYO4S ATO 4O SNINY

*NOLSOB **O90 3dALOIT3H "“HLYOMGOOM °8 *f AB OLOHd

‘F ayrid ‘OTIYD ® [izvag 0} ‘dxq

- .
C at) Sets ees

Woopworts.— Geological Expedition to Brazil and Chile.

PLATE 5.

Pleistocene glacial gravels at San Rosendo, Chile, on the north bank of the
Rio Bio Bio just inside the Longitudinal Valley. December 24, 1908, looking
N. mag., from the railroad tracks just east of the Station. Page 131.

‘JTMIHOD ‘OON3SZ9Y NVS LV STSAVED TWIOVW1ID SN300LSI31d

*NO.LSO8 ‘'O0 3dALOINSH "“HLYOMGOOM ‘6 *f AB OLOHd

ae

"S$ a3UId "IUD W pizvag 0} ‘dxq

W oopwortH.— Geological Expedition to Brazil and Chile.

PLATE 6.

Reproduction of a model of a portion of the state of Sao Paulo made by the-
Geographical and Geological Commission of Sao Paulo and based on the
topographic surveys of Horatio E. Williams. The model brings out the
detail of the erosion of the metamorphic and igneous Pre-Devonian belt
forming the Atlantic slope of the highlands near Santos; the Serra do Japy
southwest of Jundiahy. The roughened surface northeast of Sio Paulo may
be regarded as the dying out of the Serra da Mantiqueira which range rises
rapidly to the northeast of the area shown in the relief map. Except for
the Tertiary basin in the smoothened area about Sao Paulo, the topography is.
developed on metamorphic and igneous rocks of the Pre-Devonian terrane.
Page 99.

Exp. to Brazil & Chile. Plate 6.

THE PLATEAU OF SOUTH BRAZIL IN EASTERN SAO PAULO.

HELIOTYPE CO., BOSTON

a ay #5 ; ?!
f - ‘
; pve ’ :
‘ Aj
’
La 7
4
teu | .
ary
~
>
ra)
as
.
eT %,
. -
.
.
*
-
7
‘
:
--*
Me ad \
\

W oopwortu.— Geological Expedition to Brazil and Chile.

PLATE 7.

The southern part of the city of Curityba, Parand, looking west over the
gently rolling surface of the peneplain developed on crystalline Pre-Devonian
schists. The surface lies at an altitude of about 900 meters. Page 102.

“WNVUVd NI SOIHdYONSLAW SHL NO NIVIdSN3d SHL ONY VSALINND

"NOLSOB ‘OO 3dALONNSH

"HLYOMGOOM °8 *f AB OLOHd

‘1 awwid ‘IYO ® [lzvag 0} -dxq

Tha ABs" A fi

aie 1a)

i
f J

Baath. cy i eee fi : si ag Paris
oe: a wer zig SME GRASS iit

Samay SAP Any teeta SRLS T

ah. as tek Py Ds rig ' 7 t ope iy . 7 |

WoopwortH.— Geological Expedition to Brazil and Chile.

PLATE 8.

The Devonian area in Parand looking northeastward, from the city of
Ponta Grossa standing on Devonian shales, to the rising, rounded back of the
westward dipping slope of the Devonian sandstone cuesta whose crest forms
the sky-line of the view. Page 103.

‘VWSSOYUD VINOd WOYS N33S VISAND SNOLSGNVS SHL 4O LS3YD SHL ONV VWSYV NVINOAZG 3HL

*NOLSOS ‘*O9 3dAL0I13H

“HLYOMGOOM *8 *f A

a OLOHd
Wry a

Wy

A Ney

"8 321d

“OTIYD WY [izvig 0} ‘dxq

ae } ’ r
i i 4 is
th i he oe oy
-& @7nL r
hy d it
4
i'd if, r,
ine Ae se ui oh of) 25
‘- b —s yan 4
7 a mu
yg hee
6 i e' fia
4
¥
, .
-
y
”

7

WoopwortHu.— Geological Expedition to Brazil and Chile.

PLATE 9.

The campos of Sao Paulo looking north from a point about eighteen kilo-
meters west of Bury (Porto Apiahy), 128 miles inland and westward from the
port of Santos, showing the smoothened surface of the peneplain of the Per-
mian area, which is usually more rolling than shown in the view. Freight
wagons and ant-hills in view. Treeless condition characteristic of large
tracts. Page 102.

/

*‘wOLSO8

6 aivIg

*OO

IALONISH

‘SNVYUYNSL SNOLSONVS NO fom1nvd OVS 4O SOdWVO 3H1L

*HLHYOMGOOM ‘8 ‘f AB *OLOHd

‘IYO W zag 07 -dxq

;

_) 4
ote
uy

olallaege a feok meeeNaly # acy Ve

7

W oopwortH.— Geological Expedition to Brazil and Chile.

PLATE 10.

The Morro do Monge (Monk’s Hill), near Lapa, Parand; looking 8. 46° E.
mag., from eastern limits of village, August 14, 1908, showing cliffs of heavy
bedded Permian sandstone overlying beds with waterworn gravels and out-
cropping layers carrying striated stones. Page 67.

bt o > i _ At pe ee aa

‘VNVUVd VWd¥1 YVSN SDNOW OO OYXOW SHI

*“NOLSO8 °'O9 3dALOINSH “HLHOMIOOM “8 *f AB OLOHd

‘OL MI ‘AYO W pizeag 0} ‘dxq

” Port a.

Woopwortu.— Geological Expedition to Brazil and Chile.

PLATE 11.

View from Restinga Secca on the railroad between Ponta Grossa and
Curityba looking S. W. over the Permian area in the valley of the Iguassi.
August 4, 1908. Page 104.

®
ee
7

‘

“VNVUVd VOOSS VONILSSY WOYS N3AS NSSVND!I SHL SO ASTIVA SHI NI VSY4V NVIWYSd

*NOLSO8 ‘°O9 3dALOINSH “HLYOMGOOM ‘8 *f AB OLOHd

‘IT QUI ‘OIUD W [zeag 07 -dxq

PLATE 12.

PLATE 12.

The gorge of the Iguassti at Serrinha, Parand, west of Curityba. The
Iguassti after leaving the baselevelled region about Curityba flows for several
miles across the sandstones and boulder-beds at the base of the Permian

system. The basal beds are regarded by some Brazilian geologists as possibly
Devonian. Page 65.

: " all

*NOLSOB ‘'O9 3dALOINSH
— e

‘=
%,

"ZI BIg

"YNVUWd ‘VHNIYYSS LY NSSVND! JHL 4O NONVO

"“HLYOMOGOOM °8 *f AB OLOHd

eee a? as ae

AUD ® [izeag 0} -dxq

se

F
a

* < +
7 — ik inv 8
es . os pad
, Syst iF 4
+ < eeeer aes Sa
+) eee

WoopwortH.— Geological Expedition to Brazil and Chile.

- PLATE 13.

Pleistocene surficial gravels at Tamandud, Paranda, in a railroad cut north
of the Station, on the westward dip-slope of the Permian sandstone cuesta.
These gravels occur on the interstream slopes here and there. Page 108.

"YNVUWd “VNONVWAWVL LY STISAVYD SNSOOLSI31d

*“NOLSOS °°O9 3dALOINSH "HLHYOMOGOOM °6 “Sf A@ OLOHd
o Bg a rr ~s a 2 me 19 ; : ~~ e im - —_ * . : 2 ae ORE “To
. « > J ,

‘C1 VI ‘IIYO W izveig 07 ‘dxq

PLATE 14.

rat TAME :
Section in the red residual cl: Mare fe ft avadg bord
Iguassi at Lago, south of Ponta Grossa, Parand. In
similar to the terra roca. Ant-hills show

Exp. to Brazil & Chile. Plate 14.

PHOTO BY J. B. WOODWORTH. HELIOTYPE CO., BOSTON

RED CLAYS AT LAGO, PARANA, VALLEY OF THE IGUASSU.,

Woopworrtu.— Geological Expedition to Brazil and Chile.

PLATE 15.

Sketch map of the geology of south Brazil, based upon surveys by Dr.
Orville A. Derby, Dr. I. C. White, the publications of Dr. J. C. Branner,
and upon field observations by Dr. Euzebio P. de Oliveira and the writer.
The boundary of the trap sheet and consequently of the sedimentary area
about Lages in Santa Catharina is drawn without any claim to approximation
of its true position on the south, and the eastern limits of the effusive trap
sheets in western Sao Paulo and northern Parand4 are likewise indefinite. No
attempt is made to delineate the numerous exposures of intrusive sheets
giving rise to the ferra roxa of Sio Paulo and northern Parand. Page 41.

NOLSO®S ''OO 3dALOINSH

Dey (TM LE OL &b) 2 tS : 2OL

prmovaietnmer Po

| Se

soypreP ew yur oO
yuanig BPP dF
fo cajou Wore

22 Vv O2IVM2'/

Aquay fo scout

wauf papsduoy

VAS:

| a) ¥ vNnbv If

¥ 5 rogudy
fo xard
-Woj Uv!
- UOnaT-F4/

|
SHI0L D7
“YF 404 Dau |
|
|

VZ

Y

‘Ud DlLUIL a

QW

dey unm
‘DISS VIG

Ar? i343)
aU LDU -U yy

Cola ae

Psy 4

VL, Wy

.
. tats
Pt ee ae
.

ONIO31

Mg

ys) 3

"Si aye “IUD ® [izvag 0} -dxq

Sa dies
“+7.
i a
i
%,
La iF
=
—
P..4
“a
» =
=
pet
2 ad Y
a

a ie

Woopwortu.— Geological Expedition to Brazil and Chile.

PLATE 16.
Ture Serra GERAL FROM MINAS IN THE TUBERAO VALLEY.

The dim wall in the background is the escarpment of erosion formed by the
Triassic trap sheets and the underlying sandstones, below which comes the
Permian shales. The foreground, looking up the Tuberdo just east of the
Lauro Miiller railway station is on the Permian plant-bearing shales. The
coal mines at Minas occur at a higher level to the left of the view. According
to Dr. I. C. White this wall from the top down presents the following section : —
1. Eruptive rocks, mostly diabase, amygdaloidal and otherwise, 600 M.
2. Massive grey and red sandstones and conglomerates with intruded sheets
and dikes of diabase. 200 M.

Red sandstones and shales, a massive red sandstone at the base resting with
slight unconformity upon limestone, approximately 100 M. The rocks
below are referred to in the accompanying text. Page 92.

—————— Cle eee

NOLSOB ‘OO 3dALOINISH

Quin

"QT 93d

‘ASTIVA OVESENL SHL NI SVNIW WOYS Wuy3s5D VWYYSS SHL

"“HLYOMGOOM °S 'f AB

OLOHd

‘IUD W [izeag 0} -dxq

“"

PLATE 17.

W oopwortu.— Geological Expedition to Brazil and Chile.

PLATE 17.
Rio pas Maromsas IN SANTA CaTHARINA.

View looking east across typical tributary of the Parana in the trap plateau
on the pack-train route between Corytibanos and Porto da Unido. Butia
palms and araucarian pines, with smoke of forest fire on left. Photo by
J. B. W., August 30, 1908. The road crosses the Marombas northwest of
Corytibanos. The view is looking southward. The river flows toward the
right, southwestward towards its junction with the Rio Canoas, a tributary
of the Rio Uruguay. The ferry is referred to on page 27.

"VNINVHLVO VWINVS SO NV3SLW1d dVY¥L SHL NI SVEWOUYVW Old SHL

*NOLSO8 ‘OOD 3dALONNISH "HLYOMGOOM “8 *f AB OLOHd

:

"LI aid ‘IYO ® [izeag 0} ‘dxq

——_— ——

Aba) hea Shoe 33 om) ‘eb ea) oa Oe wraawe

Ee rs 0S Vas a as
: | | re bn tn
oe ent Ds pe)

Woopwortu.— Geological Expedition to Brazil and Chile.

PLATE 18.

Augite porphyrite dike with numerous inclusions of sandstone, shale, and
fragments of vesicular lava, about six miles north of Lages on the road to
Corytibanos, Santa Catharina.

The cutting edge of the palaeontological hammer points out a small block
of red sandstone. The upper long hunting knife point rests on a fragment of
included trap; the German army knife lower down points to a light colored —
sandstone. View taken in a roadside wash-out, August 28, 1908. Page 95. —

‘| se ee

‘WNINVHLVD WLNWS ‘SS9V1 YVSN SNOISNIONI HLIM 3yI0 OISVE

*NOLSO8 ‘OO 3dALOINSH “HLHYOMGOOM °8 *f AB OLOHd

"gt aI ‘MYO W [izvrg 03 -dxq

ee

ce

Bf fron ohit IM pt coset Boag eS wig wis nied sina to Ge

| shioiny ab okin)

S40 CRReeeog al qen Jqreeinanr e: on howsit ei,gac nial’

toa wen balleda sew otuen MT: fie ob dahidloessiM o only

Raineasiy sinlog berunndint Sight pede pidge 1G
. eee sents. wal) ie gaqelt oor ation
3 eho: toute To nig) howe a Te NV eo .oovig’

ome agar He A282 teirgus

Woopwortu.— Geological Expedition tc Brazil and Chile.

PLATE 19.

Map of route from Rio Negro to Lages and return to Sao Jéao and Porto da
Unido da Victoria.

This map is based on a manuscript map in possession of the Servigo Geolé-
gico e Mineralogico do Brazil. The route was plotted from notes taken by
Dr. Euzebio Paulo de Oliveira. Light tricornered points represent camps
going southward, solid symbols those on the return journey. Dates are
given, e. g. viii. 27-28, should be read ‘‘night of August 27th to morning of
August 28th. See page 15.

é

ws
SS
a cA

pip Oty
Xx'0 ee

t
(ia 2,
ww G,

”

cena

mn)
4

s
=
2
z,
a
Corn!

4 Me,
famine,
i “we
~ “ey, 7) 17

“nga g"

Yo!

Lam

Exp

to Brazil & Chile.

wy
\AS

Plate 19.

ny er Se

Se)

re Esp

MN
nye,

“ay Wy,

iments

pe

fo Erno

19 da Vy, wen
Hin auapsne ayaa
on es

Wh

HO are anite

ve

82 (57? 10'27- WL Greenmib fo”

Zinta Greta, Saferdredp 170%
20°

MAP OF A PART OF SANTA CATHARINA, SHOWING ROUTE FROM RIO NEGRO

TO LAGES AND RETURN TO SAO JOAO.

HELIOTYPE CO. BOSTON

ie
» ,

Exp. to Brazil & Chile.

BOSTON

HELIOTYPE CO.

PHOTO BY J, B. WOODWORTH.

ILLA RAFFORD, SAO PAULO.

BOULDER OF CONGLOMERATE IN PERMIAN AT V

PLATE 21,

W oopwortH.— Geological Expedition to Brazil and Chile

PLATE 21.
TILLITE BED AND GLACIATED STONES NEAR SENGENS, PARANA.

This view was taken on the 16th of September, 1908, in the then relatively
fresh railway cuttings between Sengéns and Itararé, between kilometer posts
234 and 235 on the south side of the Rio Jaguaricatu. The wall of rock in the
background is a somewhat softened boulder-clay with scattered stones and
small boulders, one of which is seen protruding to the right of the white square
(block of field-labels). The loose stones in the foreground were picked up in
the immediate vicinity on the floor of the excavation. The largest boulder
is granite; the next but smaller on the left is a white sandstone, probably of
the basal Devonian; the stone on the extreme right is a dark greenish rock
not determined. To the left of the hammers is to be seen a well-striated
reddish brown sandstone, probably a fragment of the disrupted glacial floor.
Page 62.

‘YNVUVd ‘SNSONAS YVAN SSNOLS G3LvVIOWID ONY O39 SLITNIL

"“HLYOMGOOM °G ‘fT AB OLOHd

RS Ce A Cy i I os OP ARE DE RE MS ong nn 1 ee
. Be t. - 2 . ‘ ee _ o a % 4 a y

Re TA
PEEPS: =

‘NOLSOS oo aaonal
Sr oR, pies

oly

"LZ Id ‘IUD ® [izvag 0} ‘dxq

re

a

>> ee eo
[ a wre
° oe Pia Ua he

N
3
2
=
A

Woopwortu.— Geological Expedition to Brazil and Chile. :

PLATE 22.

BLockK OF FINE WHITE SANDSTONE NINE FEET LONG IN TILLITE BED NEAR
SENGENS, PARANA.

View taken September 16, 1908, in railway cut between kilometer posts
235 and 236 between Sengéns and Itararé, south side of track, showing a large
block 9 ft. (or three meters) long of a fine-grained white siliceous sandstone,
embedded in a typical glacial conglomerate or tillite. This bed lies a few
feet above strata of the same lithological appearance as the erratic. Page 62.

Plate 22.

Exp. to Brazil & Chile.

ney
CRB TON RT 0 Byes.

‘e.
PHOTO BY J. B. WOODWORTH.

ag
iss

any

BOSTON,

HELIOTYPE CO.,

BLOCK NINE FT. LONG IN TILLITE AT SENGENS, PARANA,

PLATE 23.

W ocopwortu.— Geological Expedition to Brazil and Chile.

PLATE 23.

SPHEROIDAL SEPARATION IN TILLITE NEAR Rio NEGRO, SANTA CATHARINA,
SOUTH OF THE RIVER OF THE SAME NAME.

View taken September 20, 1908, of a roadside cut, looking northwest,
about half an hour’s ride from Rio Negro on the road to Sao Bento, on the
south side of the Rio Negro. This bed overlies the shales shown in Plate 24.
See small map page 68. ;

"WNINWHLVO VLINVS ‘OYDSN O18 YVAN SLITML NI NOILVYVdSS TWOlWAHdS

*NOLSO8 ‘OO 3dALOIISH HLYOMGOOM ‘6 *f AB OLOHd

‘IID W [izvag 0} -dxq

Woopwortu.— Geological Expedition to Brazil and Chile.

PLATE 24.

Ice-rafted boulder and stones in shales near Rio Negro, Santa Catharina.
Same locality as Plate 28. Thisshale underlies the earthy bed shown in Plate
23, and contains near bottom of view a 1.5 inch band of the same material.
Surface shows the herbaceous growth of the open lands between forested
tracts. The boulder isa gneiss. Page 68.

*NOLSOS ° “O09 3dALON

‘hz aIVIg

~—- My em

‘VNINVHLVO VINVS ‘OYD3N O1Y YV3N SSTVHS NVINYSd NI SSNOLS GNV SY3QINO0d Galsvy-3d!

“HLYOMGOOM ‘Gf AD OLOHd

‘IYO W [izeag 03 -dxq

cy Pad

be
re ioe =

on to Brazil

0 > E ;

| , 7 A
= . >a

; &

A ; * “ i k

g ¢

E \

| peor

td 1

&

&

°

EB

a

° |

° a

5 |

‘VNINWHLVO WLINVS ‘SNV3TYO LV SLVYSWOTDNOD OIV1-YS LVM

*NOLSO® ''O9 3dALOINSH

“HLHYOMGOOM ‘8 ‘tf AB OLOHd

“Se aUId ‘IYO F [1zearg 03 -dxq

=
r
~
.
.
‘
»
’
.
*
o
,
4
.
] a
“

iwihe “Lee

at

-
- = 2
+ a . ae al
> ; : % Pe
* a *
as > , : ; a :
; ¢ : fpey
ah nt. ee > oa i ro
: s 7 | ae , mi
° > Y oe a)
2 : ie
= :
¥ te on LSS
* Riad =
ad 3
. oa
= -
‘ ~ < bat
Z : oe os ¢ a
A ay i
w os : > a ~
} uw : :
. ~
a a
Sg
iy 4
."
¥ a

a

g

AY

~

BR tials pte

.

‘ethan

* ~ "
. ‘’s

- “°

WoopwortTH.— Geological Expedition to Brazil :

7
»
re
'
»
~
=
°
aie
» e
4
”
-
.

| PLATE 96. °-

Permian shales resting on granite, with

Catharina. Page 74.

4

‘VNIUVHLVO VLINVS ‘SVNINW YVSN SLINVYD NO ONILSSY SATIVHS NVINYAd

*“NOLSO8 '*O9 3dALOINSH "“HLYOMGOOM ‘a ‘0

‘Qt aI ‘SHYD 2 jizeag 03 -dxq

a i. y
’ ~)
'
)
}
by
o~
,
7
‘
.
‘
t
ra
tA
J
Se fi

om
r? ahd a

Mag |
‘=
*

\
2 Led

Woopworrn.— Geological Expedition to Brazil and Chile.

’

°
PLATE 27.
“N ’ . ‘ y,*a1" < ‘. of. ;
Glaciated pebbles from tillite beds -of P
¢
¢
a =
s*
: it ae
a oe
; o * a hi iv o aee p! 4 ) 4
sh Ca La? ee WN ti fi) =

Exp. to grazil & Chile. Plate 27.

PEBBLES FROM TILLITE BEDS OF PARANA,

HELIOTYPE CO., BOSTON

3. a en y - a bat a . = - +) _“

‘

my

§

4

Geological Expedition to :

WoopwortTH.—

ale

7

ae ee

chart of Valdivia River and Corral entrance,

The coast

25.

Plate

Exp. to Brazil NX ¢

? ar eo

aN —"

pies it ay

* o
wel feiere + .

a oye,
i eel. 7, Syl Pee

i¢ ‘ are j

Pie me =a j

ree Ba 5 a

Woopworts.— Geological Expedition to Brazil and Chile. a
/ Cha?

74

PLATE 29.

’
: ‘

Corral and the Cordillera Costal. Page 122.
a , ’ ;

“JUMIHO ‘VYSTIIGHOOS WLSVOS SHL GNV IWeuYHOo

"HLYOMIOOM ‘8 ‘fT AB OLOHd

—

, tage” ero
ead oF co ey

Vs eheltre'h «

"62 aid ‘OTIUD 2 [Jzeag 0} “dxq

¥ —— <=, 0 oe |
- os ———

7
q uy My ’
al | Wy
fs a
{
’
7 . ’
ae ae
SNS ie PR Th ata atest AN Meee Fh aes pe gle ie , ie om
/ ‘
Ag
7 ;
A
¢
~ F h
. j ‘"
Jd -
; he
‘
r; 7
F
”
* a A .
‘
id
I ’ bd
Dae ae ae)
+s,

ied ek Re Seok a cee oe Ca ees os ere EN ee ee

WoopwortH.—Geological Expedition to Brazil and Chile.

PLATE 30.

Rock-terrace around Manzera Island, Corral Harbor, Chile. Page 122.

"
i “
7’
*
.
é
ried
*.
4
‘
n
:
>
ps
‘a ve
eo
4
ah » a
nt ee’,
~ 7” e+ ~"
Yuisd
Te im
wa
%

*“NOLSOS

‘OO 3dALONNSH

‘of a1uI1g

‘“YOSYVH WHYYOO ‘GNVIS! VYSZNVW GNNOUYV JOvYeYNsL YOOY

“HLYOMGOOM *8 *f AB OLOHd

—

“OTIYD W [lzvig 0} ‘dxq

nil

:
'
.

AWS MEM, WE ot etehlppiy.. J lastest rieu. : ‘

.
aan

“ ae ait, ry - . ,
: 7. ae, + ee ee yr ee” «Nee Pt Foe ; “ et. Fs : 2 /
: ae ea a PMI SOT ES on fees aitirahs 4” ae al eplyn per ivrits
, 7 ¢ a s. x y 7 ae

ny
. .
Y. F P Thy f t &
~~ 1 ae > > au » 13 ‘ y coos f a wi
, : ¢ r f pt y ¢ d : yey i ‘
cia a a gh ¥ t ue aj | wt r oe a ; ote o
oh Ger Ay tet ‘ \ i i ee Dah
r= , he Ps on = 7 - *
. i f VG 4 ras U

Wcopworrtu.— Geological Expedition to Brazil and Chile.

/

,

*

Plate aN.

to Brazil & Chile.

Exp.

i

Nh it
ie

yi

= 7 _—
oe ——

Nj "ih WAN

MONA \N :
| { Hl Hh ‘Nh
AR :

Nii Mi)

\\t

(

mi fi ml )'

yen (

1}

\ ii t) hy

STON,

BC

HELIOTYPE CO.,

PHOTO. BY J. Bs. WOODWORTH.

OPPOSITE CUTUPAI.,

ROCK TERRACE IN VALDIVIA RIVER

ay

ae § * ~'
mse * % .
: ied vid take iy Lal an A :

l
\

kn

f as * 2
f $ ,
re ii y* * J
J. aoe » £ a
t 7 au Bes: 4
. ee a
-
= i- ee ys
4 +." ~

W oopworru.— Geological Expedition to Brazil and Chile.

j

PLATE 32.

Drong of schist on fifty to sixty foot terrace near English Bay, south of
Corral entrance, Chile. Fame 123.

¢
‘
’
7
~
{
¥
&
aX
ra,
f
*
\
“re
i ¥
>
he 5
C- ¢n >
$ = »
r “a re 7 dil , ~" oh
4) iy , o if ‘
» i { *
R fe a ie e mat ¢
a ow ne ria a
i\% i’ ie A

“AVG HSIIONS YVAN SJOVYYSL ‘14 09-O0G NO LSIHDS 40 ONOYG
*NOLSOS8 ‘*00 3dAL0113H

“HLHYOMGOOM ‘8 ‘Pf Ag

HYD [izvag 0} -dxq

= ‘ 4 ad Z

Are

oe PLATE 33.

AM oa
2 Es i A eg +) it Sera Bes

a.

: f 2 ”
= 4

~

| te RPA tree
Sea-cave at upper marine limit near Palo Muerto
entrance, Chile. Page 124. ;

“we
" : ¥ %
;
"an ~
.
r
‘ — >
»
‘ *
‘ -
> ” y
P a ne. |
4 s
er fe
a 4 Py Ty “Pee 7
a s
‘ ‘ M4
= 3

"LNIOd OLYUSNW O1Wd YVAN LIANIT ANIYVW YAddN LV 3SAVO-V3S

. c ‘a
*NOLSO®8 °°O9 3dALOINSH HLYOMOOOM °S ‘fT A@ OLOHd

"CE aqvIig ‘IYO W® [izeaig 0} ‘dxq

PLATE 34.

Woopworts.— Geological Expedition to Brazil and Chile.

PLATE 34.
Coast chart of Concepcion Bay, Chile. Page 125.

‘

vrs .

Exp. to Brazil & Chile. Plate 34.

AAC Vou Wis Wa Lo)
> J “Pap

20

“my,

Yn
a

ay ,o* 'S $5 oI
= <> ¢ a
= S Oy
8 : ‘
ze =
=~
= ° 4
8 % _
a ?, Pid
z % %
= Sing x :
%

C7 ” ‘90 eo 110 as"
my iin anf a

OVEOINGS = FaATHOMS.

oe | 3 fetar

Boy the Mmnet Fort om thar Cas.
cs he proper <niranen tr smammer amet

thorough Qurepcina Avunmet
peal vepmtatlns 6 rats clear tomt om thew
cot este on be proce! tn 0 tard

PLATE 35.
)

Woopwortu.— Geological Expedition to Brazil and Chile.

PLATE 35.

Talcahuano and the country south of Concepcion Bay. Page 126. —

— - =e

“AVE NOIODdSONOOD SO HLNOS AYLNNOD SHL GNV ONVYNHVOTVIL

*NOLSO®8 ‘'O9 3dALOINSH "HLHOMGOOM ‘8 *f AB OLOHd

“SE aqRId ‘IUD W [izeig 0} ‘dxq

7\n i
& fl Pets. L.. ae
1, eles fon
sv ie 4
¥- ‘ Ps) oy a
Se

Woopvwortu.— Geological Expedition to Brazil and Chile.

_

PLATE 36.

-Tumbres Peninsula; looking west from deck of steamer at anchorage.
The flattish sky-line is that of a roughened peneplain. Page 126.

*NOLSO8 ‘*090 3dALOITSH

"Qf avid

“WINSNINSd SSYSWNL

*

ing
*
7

LBS fs ake ;

~

7

“HLYOMGOOM “6 “fT A@ OLOHd

“OIYD W [izeag 0} ‘dxq

,
ari
}

st

ar ws i. ‘

PLATE 37.

yn to Brazil and Chile.

— Geological Expedit

WoopwoRrTH.

ow

]

»
“ir,
*

te STAM
n on the Rio Bio

Modern Concepcio

NOLSO8

‘LE aid

is
tere, 4

‘OO 3dALOINSH

‘Old O18 O1Y AHL NO NOILd3DNOD NYSGOW

"HLYOMGOOM “dG ‘fT AS OLOHd

"IUD W [izesrg 04 -dxy

fam st a oi Es Yad

Bulletin of the Museum of Comparative Zodlogy
AT HARVARD COLLEGE,

Vou. LVI. No. 2:

GEOLOGICAL SERIES, Vol. X.

THE SQUANTUM TILLITE.

By Rospert W. SAYLES.

Wits TwetvE Puates.

CAMBRIDGE, MASS., U.S. A.
PRINTED FOR THE MUSEUM.

JANUARY, 1914.

Ee ke
ea:

'a*

ot) Tete aie To ae
(110% OP ae

‘S : i "wie

ree tA Ra tone

vA MU MAE SCL at AS

LSU A: CRE Oe aa ee

iQ) Saareal
3 Ke

. "
i ¥
bw
ee,
s id
Vv eee
a 2 ri
‘ eR eal
oP ee Ty
Ape ir
‘a ‘7 rey
¥ f ; WA wan aes
_ vant ae | ’
‘ yy ag) aan ar
; A a me
: a yay Py : .
la .

°°
»*
.
Se
—
<
ee
-
=
oe
-
~~

No. 2.— The Squantum Tullite.
By Rospert W. SAYLEs.
INTRODUCTION.

For many years the origin of the Roxbury conglomerate has been
a matter of debate. In 1869 Prof. N. S. Shaler stated his belief in
the marine origin of this formation (Shaler, 1869, p. 172-177). Later
on, in his lectures he considered the conglomerate to be of glacial
origin. In some of his published statements the same idea may be
found (Shaler et al., 1899, p. 37-38, 57-59, 64-67). Prof. W. O.
Crosby thought that the Roxbury conglomerate was of marine origin
(Crosby, 1890, p. 10-17). Mr. W. W. Dodge believed in the glacial
origin of the Roxbury conglomerate. (Dodge, 1875, p. 408-409).
Prof. J. B. Woodworth, although not publishing his views, has often
expressed his opinion that the conglomerate was of glacial origin.
Prof. George R. Mansfield did not indicate any decided view as to
the origin of the Roxbury conglomerate, but in his conclusions
(Mansfield, 1906, p. 256-257) favored strongly the glacial idea. In
the last paragraph of this section of his paper he says: “The great
quantity of large pebbles of relatively fresh granite and the abun-
dance of felspathic material in the sandstones and in the matrices
of the conglomerates suggest that much material was furnished to
the streams of that time by glaciers of which no direct evidence now
exists.”

In view of the fact that tillite of Permian age has been found in
every continent except North America, the discovery of what appears
to be tillite in the Roxbury conglomerate is important. Glacial beds
of Huronian age have already been found in Canada by Prof. A. P.
Coleman, (Coleman, 1907) and Mr. J. A. Taff (1909), has found ice-
borne boulders in mid-Carboniferous shales in Oklahoma.

On the 19th of December, 1909, I was gathering specimens of rock
with Dr. Laurence La Forge, just south of Boston, for the Harvard
Geological Museum. Dr. La Forge, who was at that time engaged in
making a geological survey of the Boston region for the United States
Geological Survey, had come upon a curious exposure of the Roxbury
conglomerate at a locality in Hyde Park, across the street from the
Becker Milling Machinery Company. The rock appeared to have all

142 BULLETIN: MUSEUM OF COMPARATIVE ZOOLOGY.

the characters of glacial till, and I suggested that it was tillite. We
visited another locality at the junction of Ware and Centre Streets in
Roslindale where Dr. La Forge had found the same characters in the
rock, and this time were accompanied by Profs. W. M. Davis and
J. B. Woodworth. All agreed that this rock looked very much like
tillite. We visited the Hyde Park locality, and Professors Davis and
Woodworth were strongly of the opinion that the rock was tillite,
though the proof was still wanting.

CRITERIA FOR THE DETERMINATION OF TILL OR TILLITE.

As all till varies a great deal in its character in different localities
and in the neighboring exposures of the same locality, it is difficult or
impossible to give a short definition of it. Strictly speaking, the only
characteristic common to all tills and tillites is, that they are made up
of a miscellaneous, unsorted, unstratified mass of rock-débris laid
down by a glacier. As arule till is composed of rock-flour or clay and
rock fragments, and in this mixture the matrix of rock-flour or clay
is a large part of the mass. ‘The matrix, though more often of clay,
is sometimes sand of varying texture.

Usually a majority of the stones in till are angular or bluntly angu-
lar (subangular), with several facets. Many have an oblong shape
with one or more flat sides or soles, smoothed or polished. Some of
the stones in till are rounded, water-worn pebbles, but these are usu-
ally in the minority. Where a glacier has pushed over gravel-beds,
many water-worn pebbles become included in the till without being
fractured or scratched. In southern Rhode Island the writer has seen
till made up almost entirely of rounded pebbles. In this case the
glacier retreated; the water from the melting ice deposited materials
at its front; the glacier then advanced over these deposits, which had
in the meantime become water-worn, making of them a mass which
can receive no other name than till. Here pebbles more than equalled
the sandy matrix in amount. |

Scratches or striae on the stones in the till are considered of prime
importance, in the proof of the glacial origin of a miscellaneous, un-
sorted, and unstratified mass like till. There is no doubt that striae
are most important, but they are not always conclusive proof of
glacial origin where present, and on the other hand the absence of
them is not conclusive proof of the non-glacial origin of any till-like
deposit. There are several mechanical agencies in nature capable
of producing striae besides glaciers; and over some areas in many parts

SAYLES: THE SQUANTUM TILLITE. 143

of the earth, the till contains no striated pebbles. It is not always
possible to say why striae are wanting in some regions. Where the
pebbles are mostly of some very hard rock like quartzite, the striae,
on account of the hardness of the substance, may be nearly absent.
I mention this absence of striae because so many geologists, on this
account, doubt the glacial origin of certain tillites.

Striated stones are frequently blunted at one or both ends, or turned
to a point at one end and blunted at the other end. They are also
in many cases bevelled on one or more sides. If a stone is too hard to
take the striations the blunting or bevelling effects may still be present.

In addition to the angular, subangular, oblong, facetted, bevelled,
striated pebbles, there is another mark of glaciation on stones in the
drift which has not received sufficient attention from geologists. This
characteristic, so often mentioned to me by Prof. J. B. Woodworth,
is of as much importance in determining the glacial origin of till or
tillite as the striated stones. I have reference to the concave fractures
which appear on many glaciated stones. These are as positive a
mark of glaciation as striae, and fractures of this nature can be found
as frequently as striae or even more so. Such fractures must be pro-
duced by crushing under a tremendous pressure of ice, in which case
the fractured pebble would probably lie between another rock frag-
ment and a rock-floor, or where there is no rock-floor, between two
other rock fragments (Woodworth, 1912, p. 76-77).

As a rock-floor or striated pavement has been found beneath several
tillites, some geologists have not been convinced of the glacial origin
of tillites which have no grooved rock-floor or striated pavements.
The principal objection to Professor Coleman’s glacial deposit in the
Huronian was just this point; but no rock-floor in the case of some
accepted tillites has been found (Coleman, 1908, p. 354).

Glacial deposits are made up of a great variety of rocks. In till or
tillite may be found rock fragments which have been transported many
miles from the parent ledge. This great variety is not found in gravels
formed under cliffs where there has been no glacial action, nor in local
river-gravels where the stream has had a short course and acted on
the local rocks.

The crushed fragments of the till or tillite show fresh rock and not
weathered materials.

In many deposits of till and tillite pieces of the local, underlying,
at the time unconsolidated beds over which the glacier has passed
may be found, together with those derived from a distance.

According to Stone (1899, p. 29-30), the lower part of a bed of till

144 BULLETIN: MUSEUM OF COMPARATIVE ZOOLOGY.

has usually a finer matrix than the upper part, and in the lower part
the striated pebbles are much more abundant.

In beds of till or tillite intercalated layers or beds of gravel, sand,
or clay may be found. Strictly speaking till is unstratified, but Sir
Archibald Geikie (1903, p. 1310) writes as follows: — “In general,
boulder clay is unstratified, its materials being irregularly and tumultu-
ously heaped together. But rude traces of bedding may not infre-
quently be detected, while in some cases, especially in the higher clays,
distinct stratification or intercalated seams of sand or gravel may be
observed.” Prof. James Geikie (1895, p. 14) describes these intercala-
tions as follows: —“ Another characteristic of till is the not infrequent
occurrence of included nests and lenticular layers, and occasionally
thick beds of gravel, grit, sand, and brick-clay. Sometimes these lie
in approximately horizontal or gently inclined positions, but usually
they are more or less disturbed, and often curiously curled up and
contorted, so as to present the appearance of having been rolled over
upon themselves along with the clay in which they are inclosed.”’

Considering all the different characteristics of till and tillite, it is
necessary to conclude that the proof of till or tillite lies, not in the
finding of any one or two glacial characters which a given deposit may
possess, but in the sum total of the characters found. In the different
exposures of a bed which has the appearance of till or tillite, there
should be found a majority of the characteristics of till. The list of
criteria given above contains the most important characters, and the
following resumé is given as an aid in forming an opinion of the glacial
characters found in the tillite described in this paper.

A. An unstratified mass of miscellaneous, and unsorted rock mate-

rials. 7
The matrix usually of rock-flour or clay, but sometimes of sand.
The included rock fragments usually angular or subangular,

with several facets.

Some rounded, water-worn pebbles or boulders.

Striated pebbles, rock fragments, and boulders.

Stones blunted at one or both ends, or rather pointed at one end
and blunted at the other end.

Fragments bevelled on one or more sides, the sides usually not
parallel but making an angle with each other.

Concave fractures.

Till or tillite resting on a grooved or striated rock-floor or striated
pavement.

Great variety in kinds of materials.

Hh Oo eo OD

Da

SAYLES: THE SQUANTUM TILLITE. 145

The lower part of the till usually has a finer matrix and more
striated pebbles than the upper part.

The materials in till showing that when deposited there was no
evidence of weathering or decomposition.

Materials in till usually composed of more local than foreign rocks.

Near the top of till, intercalated stratified beds may be found.

Included “nests”’ and layers often contorted.

a

DESCRIPTION OF LOCALITIES.

In descriptions of conglomerates the words “pebbles” and “boul-
ders”’ are used, as a rule, very indefinitely. In this paper I propose
to use the word pebble for sediment from % of an inch up to 5 inches;
the word boulderet for rock fragments from about 6 inches up to | foot
in diameter; the word boulder for rock fragments over 1 foot in diame-
ter.

Locality 1. Hyde Park. At Hyde Park the sandstone and conglom-
erate beds just under the tillite, strike N 30° E, and dip 70°S. About
100 feet of the tillite is exposed. The matrix of the tillite is very fine,
originally of clay. So far as I have observed them the pebbles,
boulderets, and boulders are of granite, felsite, and quartzite, but
other varieties may be found later. The shapes of the rock fragments
indicate glaciation; angular and subangular for the most part, but
here and there rounded. Dr. La Forge found a well-striated pebble
embedded in the matrix, in the presence of Dr. Ellsworth Huntington
and the writer on July 11th, 1910. This pebble is shown in Plate 2.
I have since found two others from this outcrop. There is a large
boulder of granite in the matrix within a few feet of the place where
Dr. La Forge found his striated pebble, beyond the fence back of the
Becker Milling Machinery Company’s storehouse. This boulder
measures somewhat over three feet in diameter, and is angular. This
spot is very near the bottom of the tillite, where it would be most
natural to find striated and well-glaciated pebbles. There are no
intercalated beds visible in the tillite of this outcrop, and absolutely
no stratification of any kind. No fragments of slate have been found,
and this is an important fact to be dwelt upon later. I have found
no melaphyre fragments or pebbles. Shearing is intense. See
Plate 1.

Below the tillite there are some transition-beds of conglomerate and
sandstone about 100 feet thick.

Criteria found (see page 144): — A, B, C, D, E, F, G, H, J, K, L, M.

146 BULLETIN: MUSEUM OF COMPARATIVE ZOOLOGY.

Locality 2. Milton Upper Mills. At Milton Upper Mills, about
one tenth of a mile east of Blue Hill Avenue on Eliot Street, 200 feet
from the road on the left, in woods, there may be seen an outcrop
which has many of the characteristics of tillite. The strike and dip
were not obtainable. There are about fifty feet exposed. The
matrix is gritty argillite. Pebbles are angular, subangular, and
rounded. ‘There is no stratification seen.

Criteria found.— A, B, C, D, F, J, L.

Locality 3. Roslindale. In Roslindale, south of the Arnold Arbore-
tum, fifty feet east of the junction of Weld and Centre Streets, there is
an old quarry in the tillite. No strike or dip can be found near the
tillite. The thickness is problematical on account of the heavy cover-
ing of Pleistocene drift. About 200 feet are exposed in a north-south
direction. The matrix is argillaceous as at Hyde Park. Boulders
up to two feet are common. The pebbles are angular, subangular,
and rounded. Blunted pebbles with concave fractures are common,
and also bevelled pebbles. I believe this outcrop to be near the
bottom of the tillite as at Hyde Park. I have found one pebble with
very faint striae, well bevelled and blunted. There are no signs of
stratification in this bed. Shearing has been intense.

Criteria found: — A, B, C, D, F, G, H, J, K, L, M.

Locality 4. Roslindale. Another exposure in Roslindale may be
seen south of the above. Strike and dip not observed. About
twenty feet only are exposed, in a north-south direction. The matrix
here is arenaceous. Pebbles and boulders up to a foot in diameter
may be found, of angular, subangular, and rounded shapes. No new
kinds of rocks have appeared. I have found no shale or slate frag-
ments. A striated quartzite pebble was obtained haying three striae.
Two of these are parallel, and a third makes an angle of 11° with the
others. This pebble is angular with blunted ends and a concave
fracture. Cleavage well developed at this locality:

Criteria found: — A, B, C, D, E, F, J, L.

Locality 5. South of Forest Hills Cemetery. About 600 yards
southwest of the Forest Hills Cemetery locality, and 100 yards east
of the trolley line, there is a fair exposure of the tillite. The strike
and dip were not obtainable. The thickness is also unknown. On
the strike of the Cemetery exposure given below, there are 150 feet
visible. The matrix is gritty argillite. The pebbles, boulderets, and
boulders are angular, subangular, and rounded, consisting of granite,
felsite, quartzite, melaphyre. I have not seen any shale or slate
fragments. No striated pebbles have yet been found. There are no

SAYLES: THE SQUANTUM TILLITE. 147

signs of stratification, and no visible contacts with underlying or
overlying beds. Cleavage is well developed.

Criteria found: — A, B, C, D, J, L.

Locality 6. Forest Hills Cemetery. In the Forest Hills Cemetery
there is a small outcrop of tillite. This rock may be found about 100
feet from the fence north of Walkhill St., opposite the crematorium.
Here the strike at the contact with slate is N 72° E, and the dip 70° S.
It is not possible to obtain the thickness. About twenty-five feet are
exposed. Measuring toward the west there is a distance of 335 feet
to an outcrop of sandstone, and the latter is exposed twenty-five feet.
The sandstone is in contact with the main body of the Roxbury con-
glomerate. The characters of the tillite are much like those at Squan-
tum Head, to be described later. Thin layers of shale or slate may be
seen in the tillite a few feet from the slate contact. The matrix is
arenaceous; the pebbles of all shapes, angular, subangular, and
rounded. The materials of the- pebbles are the same as at all the
exposures with no additional varieties yet found. The largest frag-
ment found was about one foot in diameter. Several shale or slate
fragments were found mixed with the pebbles. There are no transi-
tion-beds to the Cambridge slate, and in this the deposit resembles
that at Squantum Head. Cleavage is well developed. No striated
pebbles were found.

Criteria found: — A, B, C, D, J, L, M, O.

Locality 7. New Calgary Cemetery. At New Calgary Cemetery,
one fifth of a mile south of Walkhill St., close to the fence on the east,
a small outcrop of what appears to be tillite may be seen. The strike
_ of the beds just south is E 6° S, and the dip 76° W. The thickness
here is not obtainable. The matrix is a gritty argillite. No stratifi-
cation is visible. The pebbles have angular, subangular, and rounded
shapes. The position of this bed is right for tillite, the transition-
beds coming to the south, and south of these the main body of the
Roxbury conglomerate, while to the north comes thick slate. Very
little more can be said of this outcrop except that the pebbles so far as
visible are of granite, and felsite.

Criteria found: — A, B, C, D, J.

Locality 8. Morton Street. About fifty feet south of Canterbury
Street on Morton Street there is an outcrop which resembles very
closely the tillite at the junction of Blue Hill Avenue and Harvard
Street. No strike or dip have been obtainable. The matrix is arena-
ceous. Pebbles are angular, subangular, and rounded. Cleavage
noted.

148 BULLETIN: MUSEUM OF COMPARATIVE ZOOLOGY.

Criteria found: — A, B, C, D, F, H, J, L, M.

Locality 9. Franklin Field. At the junction of Blue Hill Avenue
and Harvard Street there is an outcrop of tillite. No strike or dip are
available. The breadth of the outcrop is about 100 feet. The
matrix is a sandy slate. The included rock fragments are angular,
subangular, and rounded. One striated pebble has been found. ‘This
pebble is shown in Plate 10. Blunted and bevelled stones are common.
Several concave fractures have been found. ‘The rock fragments are
composed of granite, felsite, and quartzite. I have not found mela-
phyre, and there are no slate masses. There are no intercalated
layers. The transition-bed or beds to the Roxbury formation are
conglomerate, and not slate or sandstone as in localities farther south
and southeast. Shearing was as intense at this locality as at any other.

Criteria found: — A, B, C, D, E, F, G, H, J, K, L.

Locality 10. Atlantic. About one half of a mile southwest of the
aviation field at Atlantic, on a wooded knoll, a very important expo-’
sure of the tillite may be seen. The series of beds here commences
well down in the Roxbury conglomerate proper, and in almost con-
tinuous outcrops ends near the middle of the tillite. This is the best
exposure of the beds underlying the tillite, with the tillite bed itself
well exposed. Commencing at the most northern extremity of the
knoll the Roxbury has a strike of N 42° E, with a dip of 45° S. The
transition-beds below the tillite have a strike of N 48° E, and a dip
of 70° S. Thicknesses of beds in this section are as follows: — Rox-
bury conglomerate 520 feet; a sandstone bed twenty-five feet; con-
glomerate, sandstone, and ‘slate 120 feet; contorted slate and sand-
stone forty-seven feet; conglomerate and sandstone 123 feet; tillite
298 feet. This gives a total for the section of 1,133 feet. In correct-
ing the thickness of the Roxbury conglomerate an average dip of 57°
was used, the dip at the bottom being 45° and at the contact with the
first bed of sandstone 70°.

At this point it is necessary to describe some of the beds underlying
the tillite, for the reason that two of these beds resemble a bed in
Brighton. About fifty feet above the main body of the Roxbury
formation, and just above a sandstone layer about 3 feet thick, there
appears a bed of conglomerate with thin layers of slate and fragments
of slate. ‘These fragments are very irregular in shape and vary in
size, although most of them are not over six inches in diameter. It is
evident that this deposit is water laid, and that the fragments have
been washed along with the pebbles. In view of the fact that much
larger fragments of slate are found in the tillite at Squantum .and

SAYLES: THE SQUANTUM TILLITE. 149

Atlantic, and that the same kind of argillaceous masses are found in
the tillite of Australia (Wilkinson, p. 194) and other places, and in
Pleistocene till, it is necessary to suppose — unless some better agency
can be advanced — that moving ice tore up all these shale or slate
fragments. A moving ice-sheet would tear up a clay-bed; part of
the torn up mass would be over-ridden and dragged up into the till,
and some of it would be seized by the glacial torrents and carried
forward and deposited somewhere in front of the ice. It seems per-
fectly possible that this latter method explains these lumps deposited
with the conglomerate. A large fragment would be rolled over along
the bottom more easily than a rock of corresponding size, on account
of the lower specific gravity of clay and lessened liability to wedging
on account of ready marcellation when moved against an obstacle.
In this way a large fragment might be moved for some distance and
when finally brought to rest would be much reduced in size.’ As for
the fragments at Atlantic, the ice-front could not have been very
far away, and a retreat must have followed, for above this horizon a
slate bed is found. Another advance is indicated by another con-
glomerate bed with slate fragments and tillite. Above these beds
come about 25 feet of sandstone with fine conglomeratic layers, and
some shale or slate fragments, and in contact with this the main
body of the tillite. Knowing that moving ice does disrupt clay-beds,
and the proximity of the tillite to these shale or slate fragment hori-
zons, this association makes it very probable that ice was the agency
responsible for the fragments. I am well aware that such fragments
may also be due to the undermining of clay-beds by streams, and
the falling in of clay fragments so undermined, but the evidence in
these cases under consideration points to ice-action.

The tillite has a very ragged contact with this underlying sandstone,
as if the surface of that deposit had been disrupted by violent move-
ment such as that of ice. In places the tillite appears to have been
pushed down into the sandstone. The lower part has a fine argilla-
ceous matrix with here and there coarse patches. About ten feet
from the bottom the matrix is uniformly fine. Farther up in this
main tillite bed the matrix becomes coarser and at the point highest
up resembles very closely the tillite at Squantum Head. —

1 By experiment I have found that frozen clay disintegrates as soon as it comes
in contact with water. The temperature of the water used was about 32° F. Plastic
clay in an unfrozen condition does not disintegrate with anything like the same
rapidity. Itis inferred from this that the clay fragments under discussion were not
frozen.

150 BULLETIN: MUSEUM OF COMPARATIVE ZOOLOGY.

The included fragments at the bottom are composed of granite, mela-
phyre, felsite, quartzite, and very large masses of slate. The largest
of the last measures at least six feet long by four feet wide. These
slate fragments become scarcer upward, and almost disappear. As this
exposure has been found very recently, no striated pebbles have as.
yet been discovered. The boulders, boulderets, and pebbles observed
are angular and subangular for the most part, with rounded water-
worn individuals here and there. Pink granite predominates over all
other varieties of rock. This is true both for the underlying Roxbury
conglomerate and the tillite. Melaphyre is well represented and in
large fragments, as at the Squantum exposures. An intercalated bed
of conglomerate about two feet thick in the tillite, about 150 feet
above the sandstone is the equivalent of a similar but thicker bed at
Squantum Head. Such beds in tillite are very variable. Shearing
has given a well-developed cleavage with sharp dips in a northerly
direction.

Criteria found: — A, B, C, D, F, H, J, K, L, M.

Locality 11. Atlantic-Squantum Knoll. To the north of Atlantic
on the road to Squantum and about three fourths of a mile southeast
of the aviation field there is a little wooded knoll where tillite is ex-
posed. It is not possible to be sure of strike or dip. Some inter-
calated beds occur on the shore, but they appear to be in blocks which
have been moved by Pleistocene ice-action. About 100 feet of the
tillite bed is exposed in separate outcrops. The matrix at this ex-
posure is very fine, suggesting the lower part of the tillite. The
included pebbles, boulderets, and boulders are of granite, melaphyre,
felsite, and quartzite. The shapes of the fragments are as usual,
angular and subangular, with a very few water-worn pebbles. As this
exposure has been discovered very recently no search for striated
pebbles or other marks of glaciation has been made. A block of
melaphyre was found in the southeastern extremity of the outcrop
showing a very angular outline. This block measures four feet long
and one foot wide. Cleavage is well marked.

Criteria found: — A, B, C, D, J, K, L, M.

Dr. F. H. Lahee, of the Massachusetts Institute of Technology,
has examined a specimen of tillite from this locality and writes me
(14 February, 1913) as follows: —

“The specimen of tillite, from which I had the two thin sections
made, was obtained on the eastern coast of a small hill at the head of
Quincy Bay, three quarters of a mile southeast of the aviation field
(Locality 11).

SAYLES: THE SQUANTUM TILLITE. 151

“The exposures at this place are large blocks which show occasional
well marked, laminated strata of a fraction of an inch to two feet in
thickness. Since these beds all dip vertically and strike in the same
direction, I believe that they are practically in situ. Alternating
with these strata are layers that absolutely lack any evidence of water
sorting and water deposition.

“The sorted layers, sometimes of uniform width for many feet,
consist of mudstone and sandstone. In the thickness of the entire
cliff section they may represent 15 or 20 percent. The unsorted
layers contain angular and irregularly shaped fragments of pinkish
granite (the species so common in the Roxbury conglomerate), gray
quartzite, greenish felsite, and dark green, chloritized melaphyre.
There are some rounded pebbles. These fragments and pebbles are
very variable in size. They range from grains of quartz and feldspar
ze to zs of an inch in diameter to large boulders, the largest seen being
four feet long. They are contained in a compact, greenish gray paste
or matrix which comprises 50 to 75 percent of the bulk of the rock.
Having no parallel structure of any sort — bedding or schistosity —
the paste breaks with an uneven fracture. Although the term
‘tillite’ is applicable to the whole section, I use it here with reference
only to the unstratified, unsorted portions. My thin sections were
cut from a hand specimen of this tillite.

“When examined with the microscope, the finer part of the rock is
found to be composed of minute grains of quartz and feldspar, very
small laths of sericite, and a highly refracting, granular substance,
uniformly distributed, which is probably epidote. The quartz and
feldspar are so fine that little can be distinguished. The sericite laths
show a tendency to parallel orientation, thus indicating some shearing
in the rock, but not enough to produce a visible schistosity in the hand
specimen. The laths are small and of nearly uniform dimensions.
The paste may be said to consist of particles having the same size as
these laths, or smaller. This mica constitutes between 20 % and
25 % of the matrix.

“Tn the paste are scattered grains of quartz and feldspar and small
fragments of granite, quartzite, and melaphyre, as seen in the hand
specimen. ‘These grains and fragments are usually angular. The
larger quartz particles exhibit slight wavy extinction and some crack-
ing, and also fine peripheral granulation accompanied by the marginal
insertion of sericite laths, characteristic of the early stages of dynamic
metamorphism. Of the feldspar grains, examples of orthoclase, micro-
cline, microperthite, and plagioclase (albite to oligoclase) were de-

152 BULLETIN: MUSEUM OF COMPARATIVE ZOOLOGY.

termined. While most of these are considerably altered, either to
sericite or to calcite, as the case may be, a few are remarkably clear
and fresh.

““ Now, as regards the origin of this tillite, there are two agents which
have been ascribed to its formation,— vulcanism and ice. If this
were a product of extrusive action, it would be called a tuff or an
agglomerate, and in either case it should reveal signs of a former glassy
nature of all or part of its components. To my mind, there is no sug-
gestion of such an original structure. The rock does not at all re-
semble the tuffs and agglomerates found elsewhere in the Boston
basin. On the other hand, if this were a typical till in a consolidated
state, its finer parts should be composed largely of rock-flour; kaolin
should not be an abundant constituent. Unfortunately the finer
portion of the feldspar elements has gone to form sericite and calcite,
and the original source of these secondary minerals is therefore not
determinable; but the considerable amount of larger feldspar grains,
many of them very little altered, suggests that kaolin was not abundant
originally.

“Tn conclusion, then, I may say that the megascopic and microscopic
study of this rock lead me to believe that ice was the most important
factor in its deposition; but that water, too — standing or gently
moving — was concerned in its origin. I could find no evidence of
contemporaneous erosion throughout this section. Both upward
and downward, stratigraphically, in the section, the tillite beds grade
into the water-laid strata. Apparently the ice was either floating or
had its weight much reduced by partial flotation.”

In the above letter Dr. Lahee suggests possible flotation for the
ice. I have not found any evidence that such was the case. In view
of the large fragments of slate in the tillite, and the disrupted beds
found at Squantum Southeast, not to mention the immense block of
sandstone, fifteen feet in diameter, which is evidently part of a dis-
rupted bed, I cannot agree with Dr. Lahee on the flotation idea.
Again the first beds encountered at the top of the tillite are conglomer-
ate of coarse texture, sandstone layers, then sandstone and slate, and
last slate without sandstone. These beds would indicate that the
water in front of the ice was at first shallow, and the slate would
indicate that subsidence was in progress as the ice retreated.
This transition may be seen to best advantage at the most southerly
part of the Squantum Southeast exposure where the slate appears on
the shore.

Locality 12. Squantum Southeast. At the end of the road running

SAYLES: THE SQUANTUM TILLITE. 153

farthest east on the peninsula of Squantum there are found many ex-
posures of the tillite, and these different exposures range from near
the bottom to the very top of the tillite, with transition-beds to the
overlying slate. The strike of the slate here is N 63° E, and the dip
68° S. The thickness of the tillite is probably about 600 feet. The
matrix near the bottom is argillaceous, while higher up and at the top
it is a gritty argillite, and in one place is arenaceous. Boulders of
many kinds may be found, but the largest and most numerous are of
granite. The pebbles are angular, subangular, and rounded, and many
have blunt ends and concave fractures. Prof. J. B. Woodworth

Fig. 1.— Disrupted sandstone in tillite: Squantum Southeast.

found a striated pebble at this locality. The striae are very faint,
but with the aid of a pocket-lens appear perfectly characteristic.

As stated above, the tillite here has a probable thickness of about
600 feet. There is no good evidence that the bed is doubled by fold-
ing and from the most northerly exposure to the upper transition-beds
the thickness‘is about as given. The lower contact with transition-
beds is probably hidden beneath drift and beach-shingle. I believe
that these lower transition-beds and a part of the Roxbury conglomer-
ate underlie the tillite here and have not been faulted out. At the
most northerly exposure and hence the lowest part of the tillite there

154 BULLETIN: MUSEUM OF COMPARATIVE ZOOLOGY.

is found a large mass of sandstone about 15 feet in diameter, the upper
part of which is stratified with layers of slate alternating with sand-
stone. It appears to have been a part of an underlying bed which had
been disrupted and dragged laterally upward into the till. The entire
mass is surrounded with tillite. This is by far the largest block
yet discovered in the tillite. In view of the fact that a bed of sand-
stone underlies the tillite at Atlantic, it is inferred that this sand-
stone mass is a part of a similar bed disrupted and included in the
tillite. That the lower transition-beds and the Roxbury conglom-
erate are present, but hidden from view in this locality is also inferred
from the large amount of fragments of the conglomerate found in the
drift on Squantum just where they should be found. The stone-walls
are made up chiefly of this variety of rock. The sizes of the pebbles
in the Roxbury drift fragments average the same as in the Roxbury at
Atlantic; lithological characters are the same. The texture of the
Roxbury conglomerate in different localities is extremely variable,
and hence to find the same texture in these fragments as at Atlantic,
on nearly the same strike and only a short distance away, would
indicate that they were not transported far but came from this vicin-
ity. The great preponderance of these Roxbury fragments over other
kinds of rock would also indicate a local origin. It is from this evi-
dence that I have felt justified in believing that the tillite here has
not been doubled, but is a bed shown in its actual thickness.

Near the very top of the tillite there is a multitude of slate frag-
ments, which had their cleavage developed after inclusion in the tillite.
They have very irregular shapes, none of them being water worn,
and many of them bent and twisted as if they had been in a plastic
condition when the ice moved over the surface. It is evident that
here the ice plowed over a clay-bed, breaking it up and dragging the
clay fragments, thus disrupted, upward into the till. Some of these
fragments are large, one measuring four feet long by two feet wide in
the other dimension. Most of them are about a foot long. Shearing
has changed the original shapes in some cases. It is clear, however,
that the bending, twisting, and tearing observed in these cases has
not been brought about by shearing. See Plate 8.

Dr. Ellsworth Huntington, who was with Dr. La Forge and myself
on our first visit to this locality, found a very remarkable case of
deformation in a slate bed. The bed in question is about three feet
thick. A part of this bed has been turned up so that it makes an
angle of about 90° with the original bedding for several feet, and about
ten feet farther on the bedding again assumes its original position.
There are two cases of this sort here not far apart.

Locality Finder

s

>

brid viilsoo

RR eR ne eee me = me ? _— *g " 7 ~—
eT te, ; > 4 he AT eye Bs 22a ky:

wily the Same 90 Hen omy @ ‘shag d

Al he Looe
Can

M OF Gonkt ana nes reduce:

ho? 1G 9 {
be! i with lewers of slate Alternaling
2 apart af an pag. be: hadw
() lateratiy -cpward into the Hib
nile! Withee ‘ Lika ia a far the bk
a the tilige, Tp. view af the fact Sat ai beta
_

tilamtic, it is inferred that: bs

ort of. a shader hed. disrupted and inchadeaam
the lower. traneition-bacts and the oxtail <
hit hidden frou? view in this locality’ aime
cc amount of fragments of-the congiomes te ni
Sciadtum just where they should be found,
any oF £2 FeLIety of Took, ihe ARAN
ry drift. (ragaientm average the santé aaj te
hh 4 Ceaerme fr Are. the ¢ samme Vie "
iC1G rate in chferest localiti i is extreme :
si id the sare ¢igtere in Gutse fragmentses

i
«
v4
an

stance AW
thet tne f ita (eddy sent cw came frowrt!

Teal Hpepos i) ok > miry Tragenen ti Of
i. whid haa ihicnte-& A <2 aaeusrn, v i$” fy ny

at have fett fietiiet ae Sethetitie thet the tillite hi
loubled, inet 14:5 Fost t MeRMCAY Lh ee smal th inkmesianys

~~

SAYLES: THE SQUANTUM TILLITE. 155

In the tillite of New South Wales, Australia, there is a case very
similar to those just described. The exposure was found by Mr. C.
S. Wilkinson in 1879 in Permian, or Permo-Carboniferous tillite.
It bears so closely on our problem here that it will be well to quote
Wilkinson in part. He says:— “In the section exposed in the
quarries at Fort Macquaire, Woolloomooloo, Flagstaff Hill, and other
places, may be seen angular boulders of the shale of all sizes up to 20
feet in diameter, embedded in the sandstone in a most confused man-
ner, some of them standing on end as regards stratification and others
inclined at all angles. ‘These angular boulders occur nearly always
immediately above the shale beds, and are mixed with very rounded
pebbles of quartz: they are sometimes slightly curved as if they had
been bent whilst in a semi-plastic condition, and the shale beds occa-
sionally terminate abruptly, as though broken off. Had the boulders
of soft shale been deposited in their present position by running water
alone, their form would have been rounded instead of angular. It
would appear that the shale beds must have been partly disturbed by
some such agency as moving ice, the displaced fragments of shale
becoming commingled with the sand and rolled pebbles carried along
by the currents.” (C. S. Wilkinson, p. 194).

Where the slate fragments appear near the transition-beds the pro-
portion of pebbles to matrix is large, suggesting thin ice acting for a
short time. There are some sandstone beds intercalated in the tillite
which have a strike differing by 8°-10° from the strike of the main
body of the slate just above. According to James Geikie (1895, p. 24),
this is characteristic of beds intercalated in till. These beds must
have dipped to the west when deposited.

Cleavage is well developed throughout these outcrops. See Plate 9.

Criteria found: — A, B, C, D, E, F, H, J, K, L, M, N, O.

Locality 13. Squantum Head. At Squantum Head about three
fourths of a mile north of the exposure just described, there is a massive
outcrop of tillite. Strike on the north at contact with the slate N 48°
E, dip 25° S. These strikes and dips were taken west of a north and
south fault line to be described later. The thickness is probably 600
feet. The matrix is arenaceous and argillaceous. Boulders, boul-
derets, and pebbles are of all shapes and sizes up to three and one half
feet in diameter. The proportion of rounded pebbles is larger than at
the other Squantum exposure, although angular ones are very common,
and the latter show the usual shapes due to glaciation. Dr. Arthur
Keith, in the presence of Dr. La Forge and the writer, found one -
pebble which he considered at the time to be ice worn, and I found
a pebble bearing several glacial striae (Plate 10). The pebbles are of

156 BULLETIN: MUSEUM OF COMPARATIVE ZOOLOGY.

the usual kind found in the Roxbury formation: — granite, felsite,
melaphyre, sandstone, quartzite, with some shale or slate fragments.
The slate fragments found so abundantly in the exposure at Squantum
farther south, and at Atlantic, are fewer here. A few may be seen
near the contact with the slate. Near the middle of the tillite on the
top of the hill there is an intercalated bed of water-laid gravel averag-
ing about twenty feet thick. ‘This bed may be seen again at the point
of the Head on the north side. ‘There is a north-south fault between
the two exposures. On the glacial hypothesis it is apparent that the
ice retreated and advanced again. On the shore to the north of a
quarry which is on the top of the hill, may be seen a bed of sandstone
about twenty feet thick and perhaps fifty feet from the bottom of the
tillite at its contact with the lower slate. As this sandstone comes
between two beds of tillite, it indicates another retreat and advance of
the ice.

The fault mentioned above, cuts the tillite on a line near the front
of the barn at the end of the road, and may be seen at the place on the
shore where slate is encountered south of the dwelling house. Pro-
ceeding in a straight line from this point past the barn, the fault may
be located on the north shore.

Dr. F. H. Lahee has observed plications in the slate south of the
tillite bed which deserve notice here. The plications occur in layers
of slate, and above and below such plicated layers the slate bedding
has not been disturbed. The upper parts of the folds have been cut
off, showing that the folding went on during the deposition. Dr.
Lahee suggests that floating ice became grounded and compressed
the layers, and later on when the same ice or other ice floated over
these layers, the tops were cut off. Prof. James Geikie (1895, p. 271-
274) has noted like plications in clay beds overlying the till at Porto-
bello, Scotland, and he suggested grounding ice-rafts, as Lahee did, for
the plicating agency. I have noted the same kind of folding in slate
at Crow Point, Hingham, and at the Chestnut Hill fault locality, on
Beacon Street west of Hammond Street, Chestnut Hill, but have not
seen evidence of the cutting off of the folds at these localities. At
the Atlantic exposure also folds in slate may be seen, with arches
cut off as at Squantum Head. It is not impossible that the tops of
these folds were removed by a swifter flow of water, as evinced at
Atlantic by ripple-mark of fine sandstone above the folds. The
same kind of folds, but not cut off, may be found in Pleistocene clays
in many places in this country. Near Hanover, N. H., I have found
many folds of this description. In view of the fact that the layers

— al

SAYLES: THE SQUANTUM TILLITE. 157

above and below the folds have not been deformed, it is difficult to
see how the folds could have been formed by simple gravity, especially
when it is noted that the folding and deposition were nearly contem-
poraneous.

There is some difference of opinion among geologists who have
visited Squantum Head as to whether the tillite bed is doubled by
folding and part of the exposure inverted. Dr. La Forge thinks that
the strata on the north of the Head are inverted, and that the slate

found both north and south of the tillite is the same bed. From a

nile
Ht

®

sy

Fie. 2.— Hypothetical cross section of Squantum taken at right angles to
the strike at Squantum Southeast. 1, Cambridge slate; 2, Tillite beds: 3,
Lower slate; 4, Roxbury conglomerate.

study of the Atlantic locality and a comparison of the beds there
with those at Squantum Head I have come to a different conclusion.
It is necessary to recall the order of the beds at Atlantic. In the midst
of the tillite at Atlantic there is an intercalated bed of gravel of
small thickness. Near the middle of the main tillite at Squantum
Head there is an intercalated bed of gravel from 15 to 30 feet thick.
Under the main tillite formation at Atlantic there is a bed of sandstone
about twenty-five feet thick. There is a bed of sandstone of about
the same thickness on the north side of Squantum Head in contact
with the main body of the tillite. Under the tillite at Atlantic there

158 BULLETIN: MUSEUM OF COMPARATIVE ZOOLOGY.

is, in all probability, as previously shown, another bed of tillite. At
Squantum Head a bed of tillite lies to the north of the sandstone.
Under the tillite at Atlantic there occur contorted slate and sandstone
layers with a predominance of slate. Just north of the tillite at Squan-
tum Head contorted slate with a few sandstone layers appear. Now
these beds are not duplicated at Squantum Head as they should be if
doubled by folding. The order is what it should be if the beds were
not doubled. It is true that such intercalated beds as these are very
variable, and a bed in one outcrop might not correspond to a similar
bed similarly placed in another outcrop, but the close correspondence
of these beds at Atlantic and Squantum Head is more likely to mean
a similar order of deposition, and the chances of coincidence are rather
small.

Slate layers, or “nests” are found, and also a few small fragments of
slate. These layers or “nests” of slate may be found in the first few
feet of tillite on both the north and the south sides of the Head. This
deposition would be possible either in an advance, or retreat, or sta-
tionary condition of the ice, so it might mean either top or bottom, and
could not be limited to one or the other. If there is no duplication of
beds by folding the thickness of the tillite is 600 feet, otherwise 300
feet.

There is some evidence of floating ice at Squantum Head, in
boulders found in the slate. Plate 7 shows such a boulder. This
one is of amygdaloidal melaphyre twenty-seven inches long and four-
teen inches wide, and was found at the western extremity of Squantum
Head near the contact of the tillite with the slate.

Shearing has been very intense at Squantum Head, producing a
cleavage with sharp dip to the northeast.

Criteria found:— A, B, C, D, E, F, G, H, J, K, L, M, N, O.

Locality 14._ Brighton. In a vacant lot west of 55 North Beacon
Street, there is an outcrop which has been a puzzle to local geologists.
The strike at this locality is E 8° S, and the dip 28° N. The matrix,
which is less abundant than the included pebbles, varies from arena-
ceous to argillaceous. There is stratification, and some assorting.
The thickness exposed is about seventy-five feet. The pebbles are
mostly rounded with a few angular and subangular examples. No
striated pebbles have been found. Slate fragments abound. Mans-
field (1906, p. 75) writes as follows in regard to this outcrop: —“ This
ledge has given rise to some controversy because of the appearance of
slate masses that resemble clastic material but are two feet or more
in length and nearly a foot in width. It has been maintained on the

SAYLES: THE SQUANTUM TILLITE. 159

one hand that the slate masses are pebbles and on the other that they
are pockets of slaty material laid down during the deposition of the
conglomerate.”

The slate masses referred to above are very similar to the slate
lumps in the beds under the tillite at Atlantic. I consider their
origin to be similar; disruption by moving ice and transportation by a
glacial stream would explain it. No undoubted tillite has appeared
north of Squantum Head and Roslindale, but clay fragments could
have been transported by glacial streams beyond the ice-front for
some distance. It is difficult to understand how clay particles could
have been deposited in isolated pockets in so swift a stream as is
indicated in this exposure by the size of the pebbles and boulders.
Some of the boulders measure over a foot in diameter.

Cleavage is found at this exposure. It does not appear that this
rock is tillite.

Locality 15. Waban. About half way between Eliot and Wa-
ban railroad stations there is an outcrop on the south side of the
track. The strike is N 38° E, and the dip 53°N., as determined from
overlying slate. This rock is not tillite but has every appearance of
-being on the tillite horizon. It is a very coarse conglomerate. The
largest boulders are at least two feet in diameter, and are of angular,
subangular, and rounded shapes. Melaphyre tuff appears to under-
lie this conglomerate. Above the conglomerate are beds of sandstone
transitional to a thick body of slate, which appears to be the Cam-
bridge slate. This very coarse conglomerate may well be outwash
material from the glacier. It is several miles west of the most westerly
outcrop of tillite.

Locality 16. Moon Island. At the most eastern extremity of
Moon Island, which is, as a matter of fact, artificially connected with
Squantum by a viaduct, there is an outcrop of the tillite. The
strike is N 70° E, and the dip about vertical. The matrix is very fine
suggesting the lower part of a tillite bed. There is no stratification
and the included rock fragments are mostly angular, and subangular.
No striated pebbles have been searched for. Some intercalated beds
of sandstone and conglomerate may be seen. As this place was found
very recently it does not appear on the locality map, and the criteria
for tillite are not yet as complete as possible. Moon Island is a drum-
lin, and at the tillite outcrop well-exposed till lies on the tillite.

Criteria found:— A, B, C, D, J, K, L.

Locality 17. Huit’s Cove. At Huit’s Cove, Hingham, on the east
shore of Weymouth Back River, there is an exposure of the very top

160 BULLETIN: MUSEUM OF COMPARATIVE ZOOLOGY.

of the tillite and transition-beds and the uppermost part of the slate
formation. The strike here is N 12° E, and the dip 70° N. The
tillite exposure proper is very small and little can be said of it. The
matrix of the tillite is sandy. The pebbles and boulders are angular,
subangular, and rounded. The transition-beds are very much like
those at Squantum Southeast:—large boulders and slate fragments
mingled in an unstratified mass, with here and there thin layers of
sandstone. An ice-rafted boulder was extricated from the slate, and
many more may be seen.

Melaphyre appears about fifteen feet below the tillite, but whether
as a flow, dike, or sill has not been satisfactorily determined.

Crosby studied this locality some years ago and wrote (1894, p. 249)
as follows concerning the tillite: — “The pebbles are of all sizes up
to a yard or more in diameter, the largest observed being a boulder
of coarse granite over 5 feet in length. Furthermore the various
sizes are jumbled together promiscuously without evident assorting
or stratification, looking not unlike an indurated till or boulder clay.”
Crosby discovered some exotic limestone here, which he thinks came
from the north. He did not prove his conclusions, however, on
this point (Crosby, 1894, p. 265-266).

Cleavage is well marked.

Criteria found: — A, B, C, D, F, J, L, M, N, O.

Locality — Arnold Arboretum. While the proof of this paper was
in press, I discovered a large tillite locality in the Arnold Arboretum
north of Peters Hill. It is the ridge covered by evergreens on the
eastern margin of the Arboretum grounds. This is along the same
strike as Locality 3, but farther northeast. No contacts with other
beds have been seen, so it is impossible to obtain strike and tip.
Criteria observed during one visit: — A, B, C, D, J, M.

REVIEW OF LOCALITIES WITH CRITERIA.

The list of localities on the following page with the criteria of tillite
found at each, are arranged as nearly as possible according to the
extent of outcrop, and favorable conditions for search.

The list below shows that where there is ample opportunity, abun-
dant criteria are usually found. It must be noted that a thorough
examination of some outcrops has been impracticable as yet, owing
to location in private grounds or cemeteries. Other outcrops are so
limited in extent that they show only the presence of the formation.

SAYLES: THE SQUANTUM TILLITE. 161

This will explain why some of the descriptions give so few glacial
characters, and does not mean that such an outcrop would not reveal
many more glacial characters if opportunity were granted to hammer
and blast.

In spite of the small outcrops and those which I have not been able
to investigate properly, the average percentage of criteria for all
localities so far as the present investigation has gone, is slightly over
61%. Taking the first five of the localities given in the list, and the
percentage of criteria is 80%. The best locality of all, Squantum
Head, gives a percentage of 93%. A further search for striated
pebbles by the blasting method would probably raise these percentages
materially. So far nothing but geological hammers and chisels have
been used. The striated rock-floor or pavement is entirely wanting.
There is no prospect of finding this on account of the nature of the
beds underlying the tillite.

Locality 12. Squantum Southeast:— A,B,C,D,E,F,H,J,K,L,M,N,O.

os 13. Squantum Head :— A,B,C,D,E,F,G,H,J,K,L,M,N,O.
* 3. Roslindale:— A,B,C,D,F,G,H,J,K,L,M.

4 1. Hyde Park:— A,B,C,D,E,F,G,H,J,K,L,M.

is 10. Atlantic:— A,B,C,D,F,H,J,K,L,M.

* 5. So. Forest Hills:— ALB: C.D, JA.

~ 9. Franklin Field:— A,B,C,D,E,F,G,H,J,K,L.

m 11. Atlantic-Squantum Kn:—A,B,C,D,J,K,L,M.

' 17. Huit’s Cove:— A,B,C,D,F,J,L,M,N,O.

“16. Moon Island :— A,B,C,D,J,K,L.

* 8. Morton St:— A,B,C,D,F,H,J,L,M.

6. Forest Hills Cemetery :— A,B,C,D,J,L,M,O.
2. Milton Upper Mills— A,B,C,D,F,J,L.

4. Roslindale:— MBG) dvb
7. New Calgary Cemetery :—A,B,C,D,J.

In view of the sum total of the evidence found in these different
localities, I am forced to the opinion that there is true tillite in the
Boston Basin. No other theory explains the evidence so far brought
to light.

DOUBTFUL LOCALITIES.

About 500 feet northwest of Canterbury Street on Morton Street
southeast of Forest Hills Cemetery, there is a conglomerate outcrop.

In an open field 300 yards south of Walkhill Street and one fourth
of a mile southeast of Harvard Street there is an outcrop of slate,
sandstone, and perhaps tillite.

162 BULLETIN: MUSEUM OF COMPARATIVE ZOOLOGY.

On Blue Hill Avenue opposite Hazelton Street, just north of Walk-
hill Street, there is an outcrop of slate and an unstratified conglomeratic
mass resembling tillite. Other doubtful deposits are as follows: —

On railroad between Wollaston and Quincy.

Black’s Creek, one fourth mile west of railroad.

North Quincy, one half mile northeast of Atlantic station.

Between Florence St. and Hyde Park Avenue, near Mt. Hope
station.

CLEAVAGE.

As noted in the description given of the tillite, evidence of intense
shearing is found in every locality. The cleavage dips, as a rule,
in a northeasterly direction. The shearing is beautifully shown in
some of the pebbles from the tillite, which have been split in two and
the parts turned as if on a pivot. In a boulderet found at Squan-
tum Head one half has been sheared from the other about one third
of an inch at one end, while at the other end only slight displace-
ment has been effected. Some of the pebbles have been indented,
and others flattened and stretched. A great many have a puck-
ered or wrinkled appearance suggesting flow-effects. Striations due
to diastrophic movements may be found frequently and are en-
tirely different from the glacial striae. Almost all the surfaces of
the rock fragments in the tillite have been thus affected in some
manner. With all the shearing, and other diastrophic movements
which the pebbles in the tillite have been through it is not to be won-
dered at that glacially striated pebbles and boulders are rarely found.
Occasionally one of the surfaces of a pebble has been so placed in the
matrix of the tillite as to escape the violent diastrophic movements.

Some of the tillite exposures have been weathered and it is nearly
useless to look for striae in these. At Hyde Park where Dr. La Forge
found the best striated pebble yet brought to light, the rock has been
freshly blasted and there is more hope of a successful search. ‘There
is also an advantage here in a search for striations, in that the bottom
of the tillite is exposed. As mentioned above, till contains finer
materials and more striated pebbles at the bottom than at the top
(Stone, 1899, p. 29-30). Wherever the bottom of the tillite has been
found the matrix is much finer than in the places where the top is
exposed. The difficulties experienced in extricating pebbles from
the fresh matrix of the bottom of the tillite has been very great.
Most of those taken out have been broken in many fragments. All
of the striated pebbles but one were found near the bottom of the

SAYLES: THE SQUANTUM TILLITE. 163

tillite. Until some outcrop is discovered where cleavage is absent, or
much less than at any exposure yet found, it is not likely that many
pebbles with glacial striae will be found.

THE MUD-FLOW IDEA.

Stanislaus Meunier tried to prove that the till of Switzerland was of
mud-flow origin. He did not prove his theory for Switzerland and
if he had been familiar with the immense areas of till in North America
he might have come to a different conclusion. That till does flow under
the ice when full of water, no glacial geologists will deny. There is
no reason why it should not flow under such tremendous pressures,
and flow-structure in till is often found. (Meunier, 1899).

The mud-flows most commonly known are composed of mud and
disaggregated rock. The rocks from which flows are most likely to
come are shales or slates or argillaceous schists. Granite and sand-
stone or conglomerate would not be so apt to flow even in a disin-
tegrated condition. Mud-flows of this kind require a rather steep
gradient, and are limited to comparatively small areas. The materials
in the tillite under discussion are fresh and angular, showing no weath-
ering, and are not of the kinds found ordinarily in mud-flows. There
is no evidence of a steep gradient in the Roxbury series. The area
of the tillite is more than 100 square miles, so far as known by
outcrops and allowance for folding. The total original area was
probably several hundred square miles or even more.

The mud-flows of volcanic origin are usually composed of a large
amount of volcanic materials, such as pumice, ash, scoria, bombs,
lapilli, etc. They are also, ordinarily, of comparatively small dimen-
sions. In the tillite no such evidence of a voleanic mud-flow has been
found.

The voleanic action, however, near the tillite horizon shown in the
lava flow at Brighton and certain other places, may have no other effect
than to cloud, for many, the whole issue of the glacial origin of the Rox-
bury series. Torrential waters as well as mud-flows are common in
volcanic eruptions and some volcanic materials are found in the tillite,
although the quantity is negligible when compared with the non-
volcanic materials. In some regions glacial deposits are made up
chiefly of volcanic ejections. In Glaciers of North America Prof.
I. C. Russell quotes Dr. C. Willard Hayes as follows: —“ The moraine
in front of the Klutlan is the largest accumulated by any of the inte-
rior glaciers. It is composed very largely of the white volcanic tufa

164 BULLETIN: MUSEUM OF COMPARATIVE ZOOLOGY.

already described, but with this are mingled many angular fragments
of amygdaloidal lava and a few of granite and gneiss. Much of the
moraine has been removed by streams flowing from the glacier, but
remnants 2000 feet or more in thickness extend nearly across to the
highland north of the valley.” (Russell, 1901, p. 106).

Although striated rock fragments might be found in a mud-flow,
I have yet to find a reference in the literature to such, from an actual
mud-flow. Even if striated stones were found, it is not likely that all
the other evidence of glaciation would be found. Those who would
have the tillite under discussion a mud-flow, have also the onus pro-
bandi on their side.

THE AGE OF THE ROXBURY SERIES.

The exact age of the tillite is still uncertain. The lithological
characters of the Roxbury series resemble closely those of the Carbon-
iferous and Permian of the Narragansett and Norfolk Basins. ‘The
Roxbury series, which consist of the Roxbury conglomerate, the Squan-
tum tillite, and the Cambridge slate, is newer than the Cambrian as
proved by pebbles in it of the granite which cuts the Cambrian. The
Roxbury series lie, without much doubt on the same granitic surface
of erosion which underlies the Carboniferous of the Narragansett
and Norfolk Basins.

All that can be said at present is, that the tillite is of Permo-
Carboniferous age. The fact that the Permian glaciation was so
widespread, and that new evidence of it is coming in so rapidly, makes
it very probable that the tillite is of Permian age. No fossils of de-
terminative value have been found, although Burr and Burke did
find a fossil tree-trunk in the Roxbury conglomerate proper. (Burr,
H. T., and Burke, R. E., 1900, p. 179-184).

HISTORY OF THE TILLITE.

A study of the sediments of the Boston Basin gives some idea of the
physiography of the region, during late Carboniferous or Permian
times. The area in which the sediments were deposited extended
far and wide beyond the present limits of the deposits. That the
area of deposition was low relatively to the surrounding country is
certain, but that it was at sea-level is not so easily determined.
Towards the close of deposition the land must have been subsiding as
shown by the thick bed of slate over the tillite. In order for till to be

SAYLES: THE SQUANTUM TILLITE. 165

preserved as tillite, it must ordinarily be on a surface which is sub-
siding at or soon after the time of the retreat of the ice-sheet. Any
till deposit above sea-level on a stationary or rising surface would
almost invariably be eroded long before later subsidence could remove
it beyond the wear and tear of the elements. Whether the slate above
the tillite is of marine or fresh water origin it is not possible at present
to say. No clearly marine fossils have been found in it, and so far as
this negative evidence goes it is more probable that this slate is of
lacustrine origin. The absence of fossils, however, does not settle the
question. Marine life in the Permian seas was scarce or wanting
altogether in many places, and furthermore fossils are not found
everywhere in the marine clays of Massachusetts and Maine and other
places where marine clay of Pleistocene age outcrops. If volcanoes
were situated then as now near the continental margins, the sea might
not have been many miles away, for volcanic action’ was associated
with the deposition of these beds as shown by melaphyre flows in
several places in the Basin. According to Bailey Willis (1909, p.
403-405) land extended at least 100 miles in a southeasterly direction
from Boston and probably much farther than this. That there was
high land to the southeast appears probable also from a study of the
tillite. The evidence so far points to a southeasterly origin for the
ice which formed the tillite. A discussion of this question of direction
comes naturally in the history of the appearance of the tillite as shown
best in the Atlantic exposure, and in a study of some features of the
tillite found at the southeastern Squantum exposure.

The Roxbury conglomerate proper at Atlantic exposes a thickness
of about 520 feet. The lowest part shows rather small pebbles averag-
ing about one inch in diameter. Farther up the pebbles increase in
size gradually, while in the transition-beds below the tillite the pebbles
are larger, averaging about four inches. It would seem very probable
that this gradual increase in the size of the pebbles heralded the com-
ing ice-sheet by wetter conditions or by a shorter distance from the
source, as the ice drew nearer. If the larger size of the pebbles was
due to more water and greater velocity, the pebbles should be as
rounded as formerly, but if the approach of the ice was the cause of the
size, the pebbles should be more angular as well as larger. The latter
appears to be the case.

Above the Roxbury a sandstone bed was formed, indicating slower
stream-action. A bed of conglomerate was then laid down, indicating
swifter stream-action. Another sandstone bed was then deposited.
At this point a new phenomenon is met with. Above this last men-

166 BULLETIN: MUSEUM OF COMPARATIVE ZOOLOGY.

tioned sandstone comes a conglomeratic mass which differs from the
Roxbury in having the fragments and lenticular layers of slate, men-
tioned above in the description of this locality. From a study of this
bed I infer that the ice had come near when these fragments of clay
were deposited. Just above this bed come about forty-seven feet
of slate and sandstone layers with ripple-mark and some boulderets
from eight to ten inches in diameter. At this time the ice must
have made a temporary halt or retreat. At least deeper or slower
water conditions prevailed. Certain layers in this slate bed are con-
torted, and immediately above and below these layers there are no
signs of contortion. The tops of the arches are eroded, thus proving
that the contortions were made while the deposition of the slate
beds was going on. The ripple-mark suggests a stream of slow speed
which might very well have eroded the tops of these folds. Above
these slate and sandstone layers occurs another conglomeratic mass
with more slate fragments and probably a bed of tillite. At this place
there are no good exposures for fifty or sixty feet; but tillite, which I
believe to be in satu outcrops, in one place. There is no tillite im-
mediately to the north or northwest of this spot, so it does not seem
probable that this outcrop is a boulder. Above this horizon comes a
bed of sandstone about twenty feet thick. In the midst of this sand-
stone are some very thin layers of conglomerate and a few slate frag-
ments, one of which measures eight by ten inches. These last men-
tioned beds indicate another advance and retreat of the ice-sheet.
The relatively thick sandstone bed shows that the ice retreated for
some distance and might or might not indicate an interglacial epoch.

Above this sandstone comes the main body of the tillite. The differ-
ence between the tillite and the water-laid conglomerate which contains
the slate lumps is obvious. The contact between the sandstone and the
tillite is very ragged, showing disruption of the sandstone. ‘The tillite
pierces the sandstone as if pushed into it. With the exception of very
thin layers of slaty material no intercalated beds are met with in this
exposure for about 150 feet, when a bed of conglomerate and sandstone
is found not over two feet thick. It is probable that this bed is the
equivalent of the bed of coarse gravel found at Squantum Head and
indicates a retreat of the ice. Above this bed the tillite is continuous
as far as the outcrops extend, but it is evident that not much more
than one half of the tillite is exposed at this locality.

To obtain an idea of the sequence of events near the top of the
tillite a study of the exposure at the southeast Squantum locality is
best, as the section is almost all exposed to view for several hundred

SAYLES: THE SQUANTUM TILLITE. 167

feet along the shore. Of this exposure only the uppermost part will
be considered.

Commencing on the little high-tide island opposite the end of the
road, thin intercalated beds of sandstone and slate are found. One
of these beds has a plication in an east-west direction which may have
been made by ice-push. It does not seem probable that this plication
was caused by diastrophic movement, not only because the movement
was at right angles to the main direction of folding, but also because
there are no signs of plication above or below this bed. It is of course
possible that there was diastrophic movement transverse to the main
direction of folding, but if this had been the case here it would seem
that there should be some evidence of it above and below the plication.

Above this first intercalated bed, near the top of the tillite, there are
two more similar beds, and between each, undoubted tillite. In places
there are very fine layers of slaty material not more than one sixteenth
of an inch thick. Pebbles are pressed into these thus cutting them off
and deforming them. These tiny clay-threads suggest melting of the
ice and trickling of water laden with clay, between the ice and the
till.

A large block of pink granite, in the tillite on this island, six feet
long and one foot wide, is important in showing transportation with-
out wear. (Plate 9). The block is angular. It is not easy to see
how this block could have been transported in its present fresh con-
dition by any other agency than an iceberg or a glacier.

Returning to the main land and proceeding in a southerly direction
along the shore, the transition-beds from the tillite to the main slate-
body can be studied with ease. The beds intercalated in the tillite
grow in thickness towards the top, suggesting longer retreats of the
ice each time. The proportion of pebbles to matrix increases, and
slate fragments of all shapes and sizes make their appearance. The
tillite now suggests very thin ice acting for short periods, for the peb-
bles are very abundant. Retreats and advances were of shorter
duration. The reappearance of the slate fragments at the top of the
tillite is to be explained, I believe, in these advances and retreats of
the ice. The ice retreated, and deposits of gravel, sand, and clay
were made on the ground left vacant by the retreat. Again the ice
advanced, ploughing up the beds formed at its front and making a new
till composed of parts of gravel, sand, and clay-beds.

Disrupted sandstone and slate beds come above this slate lump
horizon, and then appears the main body of the slate, the highest
member of the series in the Boston Basin. The ice had then retreated

168 BULLETIN: MUSEUM OF COMPARATIVE ZOOLOGY.

permanently trom the Basin, and the land had subsided, and continued
to subside until several hundreds of feet of clay had been deposited.
The direction of movement of the glacier which produced the tillite
is most important. There are a number of considerations which in-
dicate a direction from the southeast to the northwest. Though not
certainly due to ice thrust, the plication of the intercalated bed men-
tioned above, points to such a direction of movement. Again in
the description of this locality (page 155) it should be noticed that the
beds intercalated in the tillite strike at an angle of from eight to ten
degrees more east than the main body of the slate higher up. This
must mean either a diastrophic change in the attitude of the beds, or
that the intercalated beds sloped downwards towards the level of the
water in which the slate was deposited. There is no evidence of an
eroded zone between the transition-beds and the slate, so it does not
appear that there is any unconformity. The beds in question slope
from the east towards the west. According to Prof. James Geikie
(1895, p. 24), beds intercalated in till are diagonal and not as a rule
horizontal, and slope towards the ice-front. It would appear that the
beds in the tillite at Squantum Southeast dipped westward, and if this
was the case, and the difference in strike is not due to diastrophic move-
ment, there would seem to be good reason for believing that the ice
came from an easterly direction. Then again, a consideration of the
slate fragments might also indicate an east-west direction of ice move-
ment. In the tillite at Hyde Park, Milton Upper Mills, Roslindale,
I have not observed slate fragments. At Squantum, and Atlantic
the rock fragments in the tillite show a majority of pink granite, with
melaphyre and quartzite coming next in abundance. If the ice had
come from the north, the granite fragments could be explained, but
not the melaphyre. If it had come from the west, the melaphyre
fragments could be explained, but no pink granite of the variety found
in the tillite is known in that direction. If the ice came from the south
the pink granite could be accounted for, but not the melaphyre. If
the ice came from the southeast, however, both the pink granite and
the melaphyre are explained, for at Nantasket, Cohasset, and Hing-
ham these rocks are found in situ. The fact that the largest
boulders found in the tillite are of pink granite and melaphyre, and
that these are found together, suggests a place of origin for both near
the same locality. I have not forgotten that Pleistocene drift may
hide some outcrops, and that the above suggestion cannot be proved,
but so far as known outcrops go, it is a legitimate speculation, and when
joined to the other evidence of the direction of ice movement appears

SAYLES: THE SQUANTUM TILLITE. 169

logical and what would be expected if the ice came from the southeast.
Another indication of the direction of ice movement is found in the
limited westward extension of the tillite. To the west beyond Roslin-
dale no true tillite has yet been found. Southwest of Mt. Benedict
in the woods there is a layer of very large boulders in the conglomerate.
One of them measures over three feet in diameter and most of them
are over two feet. ‘There is a suggestion here of outwash materials
and swift water. At Waban in what appears to be the tillite horizon
there are more large boulders, but no tillite. To the north of Squan-
tum Head and Roslindale no undoubted tillite has been found. All
of these considerations point to an easterly or southeasterly place of
origin. The fact that no terminal moraine has been found is no proof
that there was none. The width of a frontal moraine belt varies from
a few feet to twenty miles for a continental glacier. A wide belt
would probably appear somewhere in these highly folded strata of the
Boston Basin, but a narrow belt might easily have been folded under
or already eroded and thus lost to sight. Outwash materials, however,
would extend for miles beyond the terminal moraine, and that some
of the coarse gravels west of Roslindale are of such origin appears
possible.

It is impossible to say whether the glacier which formed the tillite
was of the continental or piedmont type. ‘The large thickness of the
tillite might indicate either, for thick till is not limited to continental
glaciers, but is found in the low lands of the Alps at the present day.
The thickest till is almost always found in the valleys (J. Geikie, 1895,
p. 24). The boulders and fragments of limestone found by Crosby
at Huit’s Cove, Hingham, in the tillite, seem to be real exotics, and
this might indicate that the ice came from some distance. It is
necessary to suspend judgment on this question of type of the glacier.
The Malaspina glacier is about seventy miles long and twenty-five
miles wide. A glacier of this size would answer all the requirements
of the discoveries in the Boston Basin. ‘The extent of the tillite
precludes anything smaller than a piedmont glacier.

In the vicinity of Squantum and Atlantic the tillite is seen to be
made up of three separate beds divided by the two intercalated beds
mentioned above. If the intercalated beds near the top are con-
sidered, the tillite is divided still farther. Whether the two main
intercalated beds indicate interglacial epochs is a question of impor-
tance. That such beds indicate milder conditions there can be no
doubt, but that such milder conditions would mean an interglacial
epoch of long duration is more difficult to prove. All that can be said,

170 BULLETIN: MUSEUM OF COMPARATIVE ZOOLOGY.

therefore, in regard to these two beds in the tillite is, that they prove
milder conditions and temporary retreats of the ice-sheet, and that the
cause of glacial periods fluctuated in the distant geological past much
as it did during the Pleistocene period. It would be a difficult matter
even with the aid of fossil plants, to prove an extended interglacial
epoch in such a limited deposit as that found in the Boston Basin,
unless other evidences of interglacial conditions were present.

ACKNOWLEDGEMENTS.

It was through Dr. Laurence La Forge that I first saw the rocks
discussed in this paper: He pointed out to me localities 1, 2, 5, 7, 12,
14, 15, 17. During trips with him I obtained a clear insight into
the structure and stratigraphy of the rocks in the Boston Basin as
determined by him. Subsequent research independently corroborates
the ideas he held in 1910, with the exception of his interpretation of
the structure of Squantum. The light which my recent work on this
locality has thrown, obliges me to differ from him in regard to the
structure of this section of the field, and the reasons for my difference
of opinion have been given in the description of Locality 13. Prof. J.
B. Woodworth has given me invaluable criticism and advice. Ac-
knowledgements are also due to Profs. W. M. Davis, A. C. Lane,
J. E. Wolff, Ellsworth Huntington, and George R. Mansfield, to Drs.
F. H. Lahee, and Arthur Keith, to Mr. Burton M. Varney, and Mr.
George M. Flint.

SAYLES: THE SQUANTUM TILLITE. 171

BIBLIOGRAPHY.
Barton, G. H.
See Crosby, W. O., and Barton, G. H.
Bascom, Florence.
Voleanics of Neponset valley, Massachusetts. Bull. Geol. soc. Amer.,
1900, 11, p. 115-126.
Blackwelder, Eliot.
See Willis, Bailey, Blackwelder, Eliot, and Sargent, R. H.
Blanford, W. T.
On the nature and probable origin of the superficial deposits in the valleys
and deserts of central Persia. Quart. journ. Geol. soc. London, 1873, 29,
p. 493-501.
Branner, J. C.
The supposed glaciation of Brazil. Journ. geol., 1893, 1, p. 753-772.
Brock, R. W.
See McConnell, R. S., and Brock, R. W.
Briickner, E.
Die vergletscherung des salzachgebietes. Geogr. abhandl., 1886, 1,
p. 1-183, 3 pls., 3 maps.
See Penck, A., Briickner, Edward, and Du Pasquier, Léon.
Burke, R. E.
See Burr, H. T., and Burke, R. E.
Burr, H. T.
A new Lower Cambrian fauna from eastern Massachusetts. Amer. geol.,
1900, 25, p. 41-50.
The structural relations of the amygdaloidal melaphyr in Brookline, Newton,
and Brighton, Mass. Bull. M. C. Z., 1901, 38, p. 51-70, pl. 1-2.
Burr, H. T. and Burke, R. E.
The occurrence of fossils in the Roxbury conglomerate. Proc. Bost. soc.
nat. hist., 1900, 29, p. 179-184, pl. 1. |
Chamberlin, T. C., and Salisbury, R. D.
Geology. 3 vols., New York, 1904-1906.
Coleman, A. P.
A Lower Huronian ice age. Amer. journ. sci., 1907, ser. 4, 23, p. 187-192.
Glacial periods and their bearing on geological theories. Bull. Geol. soc.
Amer., 1908, 19, p. 347-366.
Crosby, W. O.
Contributions to the geology of eastern Massachusetts. Occ. papers Bost.
soc. nat. hist., 1880, 3.
Geological history of the Boston Basin. Proc. Bost. soc. nat. hist., 1890,
25, p. 10-17.
Geology of the Boston Basin. Occ. papers Bost. soc. nat. hist., 1894-1900,
4, pt. 2, p. 249, 265-266; pt. 3, p. 289-293, 499-505.

172 BULLETIN: MUSEUM OF COMPARATIVE ZOOLOGY.

Crosby, W. O. and Barton, G. H.
Extension of the Carboniferous formation in Massachusetts. Amer. journ.
sci., 1880, ser. 3, 20, p. 416—420.
Dale, T. N.
A contribution to the geology of Rhode Island. Proc. Bost. soc. nat. hist.,
1883, 22, p. 179-201, pl. 1-3.
David, T. W. E.
Evidences of glacial action in Australia in Permo-Carboniferous time.
Quart. jour. Geol. soc. London, 1896, 52, p. 289-301, pl. 12.
Davis, W. M.
Causes of Permo-Carboniferous glaciation. Journ. geol., 1908, 16, p. 79-82.
A journey across Turkestan. Carnegie inst. Publ., 1905, 26, p. 21-120.
Dodge, W. W.
Notes on the geology of eastern Massachusetts. Proc. Bost. soc. nat. hist.,
1875, 17, p. 388-419.
Du Pasquier, Léon.
See Penck, A., Briickner, E., and Du Pasquier, L.
Emerson, B. K.
On the Triassic of Massachusetts. Bull. Geol. soc. Amer., 1891, 2, p. 451-
456, pl. 17.
Emmons, 8. F.
Review of the geological literature of the South African Republic. Journ.
geol., 1896, 4, p. 1-22.
Fairchild, H. L.
Beach structure in Medina sandstone. Amer. geol., 1901, 28, p. 9-14,
pl. 2-6.
Foerste, A. F.
The paleontological horizon of the lmestone at Nahant, Mass. Proc.
Bost. soc. nat. hist., 1889, 24, p. 261-263.
See Shaler, N.S., Woodworth, J. B., and Foerste, A. F.
Geikie, Archibald.
Text book of geology. Ed. 4, 2 vols., London, 1903.
Geikie, James.
The great ice age. Ed.3, New York, 1895.
Gilbert, G. K.
Lake Bonneville. U.S. G.S. Monograph, 1890, 1.
Green, A. H.
A contribution to the geology and physical geography of the Cape Colony.
Quart. journ. Geol. soc., London, 1888, 44, p. 239-269.
Halle, T. G.
On the geological structure and history of the Falkland Islands. Bull. Geol.
inst. Univ. Upsala, 1912, 11, p. 115-229.
Hayes, C. W.
An expedition through the Yukon District. Nat. geog. mag., 1892, 4,
p: 117-162, pl. 18-20.

SAYLES: THE SQUANTUM TILLITE. 173

Haynes, W. P.
Discovery of bivalve Crustacea in the Coal measures near Pawtucket, R. I.
Science, 1913, new ser., 37, p. 191-192.
Howchin, Walter.
Australian glaciations. Journ. geol., 1912, 20, p. 193-227.
Howe, Ernest.

Landslides in the San Juan Mountains, Colorado, including a considera-
tion of their causes and their classification. U.S. G. S. Prof. paper,
1909, 67.

Huntington, Ellsworth.

The basin of eastern Persia and Sistan. Carnegie inst. Publ., 1905, 26,
p. 217-318.

Some characteristics of the glacial period in non-glaciated regions. Bull.
Geol. soc. Amer., 1907, 18, p. 351-388, pl. 31-39.

Johnson, W. D.
The high plains and their utilization. U.S. G.S. Ann. rept., 1901, 21,
pt. 4, p. 601—742, pl. 113-156.
La Forge, Laurence.
See Sayles, R. W., and La Forge, Laurence.
Lahee, F. H.

Relations of the degree of metamorphism to structure and to acid igneous
intrusion in the Narragansett Basin, Rhode Island. Amer. journ. sci.,
1912, ser. 4, 38, p. 249-262, 354-372, 447-469.

McConnell, R. G. and Brock, R. W.

Report on the great landslide at Frank, Alta. 1903. Ann. rept. Dept.

interior Canada, 1908, part 8.
Mansfield, G. R.
The origin and structure of the Roxbury conglomerate. Bull. M. C. Z.,
1906, 49, p. 89-272, 7 pls.

Meunier, Stanislas.

La géologie expérimentale. Paris, 1899.
Penck, Albrecht.

Die vergletscherung der Deutschen Alpen. Leipzig, 1882.
Penck, Albrecht, and Briickner, E.

Die Alpen in eiszeitalter. Leipzig, 1903.
Penck, Albrecht, Briickner, E., and Du Pasquier, L.

Le systeme glaciaire des Alpes. Guide Congres geol. intern. Zurich, 1894.
Preller, C. 8. Du R.

On fluvio-glacial and interglacial deposits in Switzerland. Quart. journ.
Geol. soc. London, 1895, 51, p. 369-386.

On Pliocene glacio-fluviatile conglomerates in subalpine France and Switzer-
land. Quart. journ. Geol. soc. London, 1902, 58, p. 450-467.

Pumpelly, Raphael W.

Physiographic observations between the Syr Darya and Lake Kara Kul,

on the Pamir, in 1903. Carnegie inst. Publ., 1905, 26, p. 121-216.

174 BULLETIN: MUSEUM OF COMPARATIVE ZOOLOGY.

Ramsey, A. C.

On the occurrence of angular, subangular, polished, and striated fragments
and boulders in the Permian breccia of Shropshire, Worcestershire, &c.;
and on the probable existence of glaciers and icebergs in the Permian
epoch. Quart. journ. Geol. soc. London, 1855, 11, p. 185-205.

Russell, I. C.
Glaciers of North America. Boston. 1901.
Salisbury, R. D.

Agencies which transport materials on the earth’s surface. Journ. geol.,
1895, 3, p. 70-97.

See Chamberlin, T. C. and Salisbury, R. D.

Sargent, R. H.
See Willis, Bailey, Blackwelder, Eliot, and Sargent, R. H.
Sayles, R. W., and LaForge, Laurence.

The glacial origin of the Roxbury conglomerate. Science, 1910, new ser.,

32, p. 723-724.
Shaler, N.S.

[Note on the geological section at Chestnut Hill Reservoir, Mass.]. Proc.
Bost. soc. nat. hist., 1869, 13, p. 172-177.

Note on the geological relations of Boston and Narragansett Bays. Proc.
Bost. soc. nat. hist., 1875, 17, p. 488-490.

On the geology of the Cambrian district of Bristol County, Massachusetts.
Bull. M. C. Z., 1888, 16, p. 13-26, map.

Shaler, N. S8., Woodworth, J. B., and Foerste, A. F.
Geology of ‘the Narragansett Basi. U. S. G. S. Monograph, 1899, 33.
Stone, G. H.

On the scratched and facetted stones of the Salt Range, India. Geol. mag.,
1889, Dec. 3, 6, p. 415-425.

The glacial gravels of Maine and their associated deposits. U.S. G. 8.
Monograph, 1899, 34.

Taff, Joseph A.

Ice-borne boulder deposits in mid-Carboniferous marine shells. Bull.

Geol. soc. Amer., 1910, 20, p. 701-702.
Waagen, Wilhelm.

Mittheilung eines Briefes von Herrn A. Derby iiber Spuren einer carbonen

Eiszeit in Siidamerika. Neues jahrb., 1888, 2, p. 172-176.
‘White, C.D.

Carboniferous glaciation in the southern and eastern hemispheres, with

some notes on the Glossopteris-flora. Amer. geol., 1889, 3, p. 299-330.
‘Wilkinson, C. S.

Note on the occurrence of remarkable boulders in the Hawkesbury rocks.

Journ. and proc. Royal soc. N. 8. W., 1880, 13, p. 105-108.
‘Williams, G. H.

The distribution of ancient volcanic rocks along the eastern border of North

America. Journ. geol., 1894, 2, p. 1-31.

SAYLES: THE SQUANTUM TILLITE. 175

Willis, Bailey.
Paleogeographic maps of North America. 8. Latest Paleozoic North
America. Journ. geol., 1909, 17, p. 403-405.
Willis, Bailey, Blackwelder, Eliot and Sargent, R. H.
Research in China. Carnegie inst. Publ., 1907, 54, 1, part, 1.
Wolff, J. E.
The great dike at Hough’s Neck, Quincy, Mass. Bull. M. C. Z., 1882, 7,
p. 231-242.
Woodworth, J. B.
Geological expedition to Brazil and Chile, 1908-1909. Bull. M.C. Z., 1912,
56, p. 1-138, 37 pls.
Boulder beds of the Caney shales at Talihina, Oklahoma. Bull. Geol.
soc. Amer., 1912, 28, p. 457-462, pl. 23-24.
See Shaler, N.S., Woodworth, J. B., and Foerste, A. F.

SarLes.— The Squantum Tillite.

PLATE 1.
TILLITE AT HypE Park.

Behind the store house of the Becker Milling Machinery Co. Note the un-
stratified, unassorted nature of the rock, and the cleavage. The rock is com-
paratively fresh.

Mid

SUT TEL

4 i
Ma
7)
:
“"
b

A
sit
Aa,
a

sf
;
iti
pal
ul
T

on
ro
it
i
'
. t 4 , |
Op

Say.txEs— The Squantum Tillite.

PLATE 2.
STRIATED PEBBLE FROM Hyper Park.

Shows striae in numerous directions. On the right is the typical concave
fracture due to pressure of ice. A crack due to shearing is shown above, dip-
ping at an angle of about 60° towards the right. Another crack crosses this one
and dips in the opposite direction. The latter was made in extracting the
pebble from the matrix. There is also a fresh fracture on the upper left.
Note the wide type of striae. The pebble is quartzite. Found in tillite matrix
by Dr. Laurence La Forge, in the presence of Dr. Ellsworth Huntington and
the writer, 11 July, 1910. A trifle under natural size.

31894d G3LlvViels

Zz aNd SUILL

‘
..

Vit te =ocP

Sayues.— The Squantum Tillite.

PLATE 8.
PLEISTOCENE PEBBLE SHOWING WIDE STRIAE.
Found by Prof. J. B. Woodworth at Pondville, Mass., in washed gravel

one hundred and fifty feet from an ice-front; shows the faintness of the
grooves and the wide type of striation; compare Plate 2. Natural size.

B3vIYLS SOIM DONIMOHS 319883d ANSOOLSIFId '
NIdUNL AD “OLOHd

FULLELL

7 : ; 2

art - a! ta oD
% oh Pern
Sayies.— The Squantum Tillite.
°
-

SEA CLIFFS ON THE SOUTH OF

‘«
why i
i
Ww
~
.
he
:
av
4 -
. . ,
’
- -
»*
. ww
}
hb)
> -
}
'
.
a
“wn id
ae
™ 5
4 o
~ ‘ ats s
.
Sal a sa
f ae >
¥ rs * *

Plate 4

PHOTO. BY LINCOLN

TILLITE AT SQUANTUM HEAD

‘

r
i ee q
> a pf Laer 4

pk
Petey P00,

SayLEes.— The Squantum Tillite.

PLATE 65.
TILLITE AT SQUANTUM HEapD.

A small detail section of a part of the exposure shown in Plate 4. The exact
position is indicated by the hammer. Note the angular, subangular, and
rounded rock fragments in an argillo-arenaceous matrix.

QV3SH WALNVNOS LV SLITNIL

NIOONIT AG “OLOHd

FUTTEL

we

SayLes.— The Squantum Tillite.

This exposure is at the base of the c

cleavage on the rie appears like slate on ace sar

of the matrix,

OQV3SH

WOALNVYNOS LY

SLIVWL

SAIAVS “M “YAP OLOHd

SUT[ELL

or Pot TON a, we ‘ dg Pe
y = 74 y »
J oe. i
; 4 d ft Oe ea’, J
£ 7 ¢ i ro sa
on aw ts {i uy HB
Pid
i
Sartes.— The Squantum Tillite.
-

PLATE 7.

Boutper IN ls AT Squanrom Hi

Amygdaloidal melaphyre boulder i in slate show
Several other cases of this kind may be found near Bai
boulder was found at the most western part of the Squan
Size of boulder 20 x 9 inches.

<
‘
. pul
ee
a mw
ye
TA /
ue

/

~

Nid

QV3H WNLNVNOS LIV 3LVIS NI YsqqTNOg

SJTAVS M ‘YAS *OLOHd

A a a SE

ay

4 + Sa eles

Sayutes.— The Squantum Tillite. 7

PLATE 8.
LARGE SLATE FRAGMENT IN TILLITE. —
This large slate fragment was dragged upward into the tillite from a bed of

slate intercalated in the tillite. The bed shows disruption as if by moving ice.
Squantum Southeast.

SITAVS M ‘YAS ‘OLOHd

8 Id

’

ONT mnmnpe iT --- 42 0148

8 APAIF

|

pe ee fs af -
Rae Seeger: ¥i ROTA wre ARE

vets B50 Miyooret sedT — Atinwry aniq in doold ralegnk
WEY 3) 8 anoienantG “aobaw

: 2

Sayuxres.— The Squantum Tillite.

PLATE 9.

GRANITE BLOCK IN TILLITE AT SQUANTUM SOUTHEAST.

Angular block of pink granite. | There isno evidence of movement by flowing
water. Dimensions 6 ft. x 1 ft.

6 avId

ASVAHLNOS WNLNVNOS iv 3LITMNL NI XO0718 SLINVED

SJIAVS “M ‘HY AG “OLOHd

SayLtes.— The Squantum Tillite.

PLATE 10.
PEBBLES FROM THE TILLITE.

Beginning at the upper left hand corner the pebbles from left to right are:—
quartzite pebble from Roslindale showing subangular shape and fracturing;
granite pebble from Roslindale showing subangular shape and concave frac-
tures; melaphyre pebble from Squantum Head showing bevelling and striae;
quartzite pebble from Franklin Field shown to better advantage in Plate 11;
quartzite pebble from Roslindale showing bevelling, faint striae, and concave
fractures; here an attempt was made to show ice bevelling on both sides of a
central ridge, the ridge is indistinct in this picture. The last specimen was
found by Dr. Arthur Keith of the United States Geological Survey and con-
sidered by him at the time as of glacial origin.

Plate 10

Tillite

BY TURPIN

PHOTO.

TYPICAL PEBBLES FROM THE TILLITE

axa a
We hi :
aye ult ei rarey ; ;
-
ia 4 . ’ ‘
m

SaYLEes.— The Squantum Tillite

PLATE 11.
EBBLE FROM TILLITE AND PEBBLE FROM TILL.

On the left is a quartzite pebble from the tillite at Franklin Park, showing one
blunted end, concave fractures, striae, and ice shaping. On the lower surface
there are a number of striae running in different directions. This pebble has
been slightly squeezed at the pointed end; the contact of the matrix with the
pebble may be seen on the lower surface. A small piece of matrix adheres to
the pebble on the upper left hand side.

The pebble on the right is the type found very ccvhenetrs in till. This one
was found at Mystic in the Pleistocene till of a drumlin. In general char-
acters it is very much like the pebble from Franklin Park.

Plate 11

Tillite

PHOTO. BY TURPIN

PERMIAN AND PLEISTOCENE PEBBLES

INA a ee kt ee

La ee
ey ca

Ms

af
%

Sayutes.— The Squantum Tillite.

Fig. 1.
Fig. 2.
Fig. 3.
Fig. 4.
Fig. 6.
Bags) 6:
Pig: 7;
Fig. 8.
Fig. 9.
Fig. 10.
Fig. CP,
Fig. 12.
Fig. 13.
Fig. 14.
Fig. 15.

NOTE.

PLATE 12.
LOCALITY MAP.

Hyde Park; across the street from the Becker Milling Machinery
Co.
Milton Upper Mills; about 345 of a mile east of Blue Hill Avenue,
on Eliot Street, 200 feet from the road, in the woods on the left.
Roslindale, about 50 feet east of the junction of Centre and Weld
Streets.

Roslindale; about + of a mile southwest of the first Roslindale ex-
posure.

Forest Hills; about 300 yards south of the Cemetery, and 100 yards
east of the trolley line.

Forest Hills Cemetery; the most southeasterly exposure of rock in
the Cemetery.

New Calgary Cemetery; 4 of a mile south of Walkhill Street on
Harvard Street, close to the fence on the left.

Morton and Canterbury Streets; Morton Street east of Forest
Hills Cemetery, about 50 feet south of Canterbury Street.

Franklin Field; at the junction of Blue Hill Avenue and Harvard

Street.
Atlantic; 14 miles southwest of the aviation field, in a wooded
cow pasture.
Squantum Knoll; ona little wooded knoll to the right of the highway
which connects Atlantic and Squantum.
Squantum Southeast; at the end of the road which extends farthest
east on Squantum. Also on alittle high-tide island near the shore.
Squantum Head; at the Head, along the shores, and on the hill.
Brighton; in a vacant lot west of 55 North Beacon Street.
Waban; about half way between the Eliot and Waban railroad sta-
tions, to the south of the track.

The map was printed before Locality 16 was found.

Plate 12

Tillite

SBOSTON:

‘Governare
ele al >» .
”
Le

Tho
N

b, -
2
= ae

0 x vk
: td: |
S : 8

: Ge

LOCALITY MAP

Bulletin of the Museum of Comparative Zodélogy
AT HARVARD COLLEGE.
Var. DVL. No. 3.

GEOLOGICAL SERIES, Vol. X. SHALER MEMORIAL SERIES, No. 2.

EXPEDITION TO THE BALTIC PROVINCES OF RUSSIA
AND SCANDINAVIA.

PART 1.— THE CORRELATION OF THE ORDOVICIAN STRATA
OF THE BALTIC BASIN WITH THOSE OF EASTERN
NORTH AMERICA.

‘

By Percy E. RayMonp.

Wits Eicut Puates. .

CAMBRIDGE, MASS., U.S. A.
PRINTED FOR THE MUSEUM.
‘ Juuy, 1916.

oa, ald ante ST

‘ ‘ #1 ~*

ye
se be 4 Boe tel it
; ’ Meg } >
oes Nahe TAR Ce
| », " a ie 4 ? ea
| ATEN eas ee
siege
werd ch te swe aur ont ts onde)
t OAD A as £4 att cx
‘Saee es ae oe Ge
< § toer ee tat RS.fA% Rh Sot “Sart ane
t+ AO DOR ORE ASE eee
AIVAVT TS ADS TAS
3
A ' AA} 7 a ‘3 ae a Tse 1 z¥4 7) ti) HOT PAAR) -
AHATAAM t+ Deed PTT SRA OPE ae
v4 fu TIES HES
’
\
. \
; 4
07 OTA VEL HE
ii

~

+
as

“SAM IDCUEEMAD
gO9 RCE
arg

*
.
~ ‘
I
| a tS
bs, +4 ie i
y. i
y Oe K, ee
\. wn Pe

" ae yh oe - {oe a

No. 3.— Expedition to the Baltic Provinces of Russia and Scandinavia,
1914.

Part 1.— THE CORRELATION OF THE ORDOVICIAN STRATA OF THE BALTIC
BASIN WITH THOSE OF EASTERN NorRTH AMERICA.

By Percy E. Raymonp.

TABLE OF CONTENTS.

’ _ Pace.
Preface : 179
The Ordovician seetas in lie Gdveausents of Petroavad snd Bathinda
Russia . i ‘ ; : ‘ « eh Sh
The Lower and Middle Gadovieis of Saunier , { , , . 206
Correlation of the American with European formations . : «eee
Correlation of the Trenton in America J . . é eee
General discussion of Russian early Ordovician isan . é ie
Detailed sections in Russia . , : , . : : : Cee
Bibliography ‘ : : : ; ; , : : : . ~2e0

Explanation of Plates

PREFACE.

THE writer has been engaged since the summer of 1900, in the study
of the stratigraphy and faunas of the Middle Ordovician formations
of the northeastern United States and Canada. During all that time
the need of a personal knowledge of the Ordovician strata of northern
Europe had become more pressing, so that he was exceedingly glad of
a grant from the Shaler Memorial fund which enabled him to spend
several months of the summer of 1914 in Russia and Scandinavia.
With the codperation of Dr. W. H. Twenhofel, who studied especially

-the Upper Ordovician and Silurian strata, a rather complete, though

necessarily hasty survey of the Lower Palaeozoic strata of the districts
mentioned was completed between May 1 and September 30 of 1914,
and large collections secured. The outbreak of the war interfered
with the completion of the work as planned, and quite seriously
affects the present report, since all the collections of three men during
a month spent in Esthonia were in Russia at the time of the outbreak
of hostilities and it is somewhat doubtful if they ever reach this
country. Our conclusions on some points are therefore of a tentative
nature, and if we should be so fortunate as to receive the collections,
a supplementary report may be necessary.

180 BULLETIN: MUSEUM OF COMPARATIVE ZOOLOGY.

The results of a trip of this sort are very largely dependent upon
the assistance of others, and I have an unusually large number of
courtesies to acknowledge. To my colleagues in the Division of
Geology at Harvard University I am deeply indebted for making the
expedition possible. To Prof. Charles Schuchert I owe much for
advice, and for letters of introduction over a route which he had
himself travelled. Mr. I. P. Tolmacev, Curator-in-chief of the Mu-
seum of the Imperial Academy of Science in Petrograd, made arrange-
ments and secured letters which greatly facilitated our work in Russia.
It was he, and the excellent assistant, Mr. Carl Lackschewitz, whom
he secured to travel with us, who enabled us to work continuously and
comfortably during our stay in Russia. Mr. Lackschewitz, a grand-
son of the celebrated Middendorf, and grand-nephew of Fr. Schmidt,
in whose footsteps we were following, proved our invaluable interpreter,
advisor, and business agent. To the knowledge and skill of Mr. O.
Knyrko preparator at the Museum in Petrograd, I am indebted for
much beautifully preserved material collected during the ten days in
which he acted as guide in the region south of Lake Ladoga. To the
many other gentlemen who assisted us in Russia, some of whom are
named in the introduction to Dr. Twenhofel’s report, I also wish to
express my thanks.

Just as I write these acknowledgements comes the sad news of the
death of Professor Dr. Johan Christian Moberg, but for whose kindly
assistance my visit to Sweden would have been of little value. Cutting
short his own field season, Professor Moberg devoted himself for
almost three weeks to guiding me over the Cambrian and Silurian
deposits of Scania. Without his intimate knowledge of the extremely
restricted outcrops in this region, my work would have been fruitless,
indeed, impossible in the restless moments of the first weeks of the war.

In Norway we were greatly indebted for guidance and hospitality
to Professor Kiaer and Dr. Holtedahl who made possible a great deal
of work and collecting jn a very short time. Nor must I omit an ex-
pression of my obligation to M. Pierre Pruvost of Lille for escorting
me during three pleasant days spent among the Palaeozoic outcrops
of the Ardennes. It is with more than usual feeling that I make
these expressions of sincere gratitude to those who assisted me,
especially as there is only too much cause to fear that they may never
see these lines.

Finally, I wish to express my appreciation of Dr. Twenhofel’s
kindness in undertaking his part of the work, and of his hearty
codperation in the field.

:
<4
{
“7
'

*¥

RAYMOND: CORRELATION OF THE ORDOVICIAN STRATA. 181

THE ORDOVICIAN STRATA IN THE GOVERNMENTS OF
PETROGRAD AND ESTHONIA, RUSSIA.

% LOCATION.

The Cambrian and Ordovician strata of western Russia outcrop in
the northern parts of the Governments of Esthonia and Petrograd,
| forming a narrow strip about 380 miles long, extending from the island
[- of Dago along the southern shore of the Gulf of Finland to the eastern
boundary of Esthonia at the Narowa River, thence eastward inland
to the Sjass River, south of Lake Ladoga, and eighty-five miles east
of Petrograd. This strip is roughly triangular, having at its widest
portion in Esthonia, a breadth of thirty miles, and narrowing to a
‘ point a short distance east of the Sjass. In Esthonia, west from Lake
Peipus, the Ordovician is followed by the Silurian; while in the
Government of Petrograd, the Devonian conceals the Silurian and
overlaps successively lower and lower beds of the Ordovician, until
east of the Sjass, it conceals all but a narrow band of Cambrian.
This overlap of the Devonian on the Ordovician in the Government
of Petrograd does not, however, indicate at all the absence of Silurian
. in the southern part of that district, as many geologists have believed.
Throughout the whole band, the older strata dip gently to the south,
a dip which they apparently received in pre-Devonian times. Thus
each successively higher stratum has its outcrop in an east-west band
lying southward from its neighbor, and the Devonian, lying uncon-
formably on these beds, conceals older and older ones according to
the amount of its northward extent. (Plate 1).

PREVIOUS WORK.’

As early as 1821 an Englishman, William Strangways, (52), pub-
lished a detailed description and map of the strata in the vicinity of
Petrograd, and since then various writers have described in great
detail the Cambrian, Ordovician, and Silurian of this region. The
principal writers on the Geology, as distinguished from the Palaeon-
tology, have been Murchison, (35), Eichwald, (8), Schmidt, (42, 44, 45,
47), Mickwitz, (33), and Lamansky, (29). Recently Bassler, (1),
has published a short resumé of the results of the work of Schmidt
and Lamansky.

182 BULLETIN: MUSEUM OF COMPARATIVE ZOOLOGY.

The palaeontologists have been particularly active in describing
the Ordovician fossils of this region, as may be seen by the very long
list of species given by Bassler, (1). Part of this work, on the trilo-
bites by Schmidt, (48), bryozoans by Bassler, (1), and cystids by
Jaeckel, (24), is modern, as are also descriptions of certain groups of
Ostracoda, Cephalopoda, Brachiopoda, and Gastropoda by various
writers. The bulk.of the Brachiopoda, Pelecypoda, and corals, still
await monographic treatment, though some of these groups are now
in the hands of specialists.

OBJECT OF THE PRESENT PAPER.

Although so well known and fully described, there still exist in
text-books many inaccurate statements about the region under dis-
cussion and there is no modern general treatment of the whole area.
For these reasons, and because previous papers are mostly in German
and Russian, and without illustration, there exists in the minds of
most American geologists only a very vague idea of the character of
the Russian deposits, and the writer therefore feels justified in re-
traversing this old and well-known ground, and hoping to add some-
thing to what has previously been observed.

During the seven weeks spent in this area I was able to cross the
outcrop of the Cambrian and Ordovician on the Sjass, the Walchow,
and the Lawa at Wassilkowa, all south of Lake Ladoga, at Papowka,
south of Petrograd, at Narwa, from Ontika south to Jewe, from Port
Kunda south through Wesenberg and Taps to Borkholm, from
Reval and Baltishport southwest through Kegel, Wassalem and Lyck-
holm to Hapsal, and also visited the principal localities on the northern
half of Dago. I was thus able to see all the principal sections and
type-localities and crossed the Cambrian-Ordovician belt at right
angles to the strike at frequent intervals throughout the whole length
of the outcrop.

Although the Borkholm and Lyckholm are considered by the writer
to belong to the Ordovician they are treated only incidentally in this
paper, but are fully discussed by Dr. Twenhofel (Bull. M. C. Z. 56,
no. 4).

\ NATURE OF THE EXPOSURES OF THE STRATA.

Throughout the whole area underlain by the Ordovician strata the
country is comparatively flat, and the majority of the hills which do

~

i

RAYMOND: CORRELATION OF THE ORDOVICIAN STRATA. 183

occur are composed of glacial debris. Along the whole northern
boundary of the area there is an abrupt escarpment, facing northward,
extending from the island of Odensholm through Baltishport and
along the southern shore of the Gulf of Finland to the mouth of the
Narowa, and thence across country south of Petrograd to the Sjass.
Where this borders the sea it is usually very steep, often perpendicular
or overhanging. In the Government of Petrograd it is a steep slope,
but usually without exposures of rock except where cut by streams or
by the opening of quarries. This cliff, or “Glint,”! is of variable
height; only fifteen feet on Odensholm, it reaches its maximum height
of 206 feet at Ontika in the eastern part of Esthonia, and probably
averages 75 to 100 feet. The strata composing the top of this cliff
are, remarkably enough, practically always the same, being the rather
hard magnesian limestone of the lower part of zone C or the “ Echi-
nosphaerites” layers. Such being the case there are many excellent
exposures of the part of the Ordovician below this horizon, and,
where there is not too much talus at the foot of the cliff, the upper
part of the Lower Cambrian is usually shown.

From Petrograd eastward, no strata of the Ordovician are exposed
above the Echinosphaerites beds. The higher strata are to be found
in the part of the Government of Petrograd west of the metropolis,
and especially in the Government of Esthonia. These beds are very
seldom seen in natural sections, being practically always uncovered
only by the opening of quarries. All the quarries are of comparatively

‘small extent and very shallow, so that there is never more than one

formation exposed in any one quarry, and contacts between forma-
tions above the Kuckers have never been seen.

Over large areas in Esthonia the strata lie very close to the surface,
and even very shallow ditches often penetrate the rock. Among such
ditches the “Graben” on the estate of Baron Toll at Kuckers, near
Jewe, is famous as the principal locality of the Kuckers formation.
Many other ditches, often very small ones, were examined during this
trip and often afforded the only outcrops over considerable areas.

FoORMATIONAL NAMES.

The names applied to the formations in this district of Russia bear
a direct relation to the above described occurrence of outcrops. In
the classification proposed by Schmidt the important divisions of the

1 From the Danish Klint, a reminder of the settlement of this country by the Danes in
the 12th century.

184 BULLETIN: MUSEUM OF COMPARATIVE ZOOLOGY.

Cambrian and Ordovician were lettered in ascending order A to F,
and many of the subdivisions designated by numbers as, Cj, Cs, C3 and
finally some of these subdivisions were further divided by Lamansky,
as for instance B,,;., By;g and B,,,. The strata also received names
suggested by their lithological characteristics or faunal contents, as
for instance, B, was also known as the “Glauconite sand” and B,,, as
the “Orthoceras limestone” or “Vaginatenkalk.” The divisions
from A to Cig are to be found in the escarpment, and these together
received the collective name of the “Glint” but no formation in this
part of the series has received a separate geographic name.

The strata above Cig are exposed as has already been explained,
principally in quarries, and therefore each formation has received
a name from the locality which has furnished either the best expo-
sures or the best fossils. Thus C2 is known as the Kuckers, E as the
Wesenberg, etc.

In this particular case, both the system of lettering and the system
of descriptive names is objectionable, and for the sake of uniformity
the writer suggests a set of geographic names for the older formations
of the series. The system of lettering fails, because A, and Az; of
Schmidt prove not to be Cambrian but Ordovician, thus splitting A
between two great systems. ‘There is likewise a difficulty about D»
which will appear later. A mixed table of descriptive names, part
derived from lithological and part from faunal characteristics is never
satisfactory, and in this case the names seem particularly inapplicable.
Thus the “ Orthoceras”’ limestone is not by any means the only forma-
tion in Russia in which Orthoceras is abundant, and the term has not
the same meaning here as when applied to Ordovician strata in Sweden
or Norway. Likewise C; is called the “ Echinosphaerites”’ limestone,
though Echinosphaerites is equally common at some localities in Co,
C3 and the lower part of Dy.

DESCRIPTION OF FORMATIONS.
Cambrian.

Esthonia formation. A, and part of Ag (Blauer Thon and lower part
of Ungulitensand) of Schmidt. Lower Cambrian.

Since it has been so well described by Schmidt (42), Mickwitz (33),
and Holm (20), the writer paid comparatively little attention to the
study of the Cambrian, but examined the exposures at Reval, Port

“pan dn pte 58 eh DELTA .

i ie

9h aoe Ba

of Nee

RAYMOND: CORRELATION OF THE ORDOVICIAN STRATA. 185

Kunda, Ontika, Peuthof, and Papowka, and studied especially the
contact of the Cambrian and Ordovician near Baltishport and at
Narwa. The best section seen was at the cliff Peuthof, which is
north of the station Waiwara, a few miles west of Narwa. Alternating
strata of light colored sandstone and blue clay-shale were there well
exposed, but the strata can be studied more in detail at Port Kunda,
where Mickwitz found specimens of Schmidtiellus mickwitzi, the
mesonacid which first afforded definite pupa of the Lower Cambrian
age of these strata.

It has been repeatedly stated that the “Blue Clay” underlies the
sandstone of the Cambrian, but I did not find this to be the case.
Everywhere the highest layer of the Cambrian appeared to be a hard,
usually almost white, sandstone. The upper bed, where its thickness
could be seen, was usually not over fifteen to twenty-five feet thick,
and beneath it was a bed of blue clay-shale of variable thickness.
Below this again one finds sandstone and alternations of shale and
sandstone continue to the base of the cliff, and, according to borings
in Reval and Petrograd, such alternations continue downward about
600 feet to the gneiss. The fossils have been found in the upper zones,
within fifty feet of the top of the formation, and there is no reason to
believe that strata of any age other than Lower Cambrian are present
in this formation. The “Blue Clay” of the Lower Cambrian has
received considerable notoriety, as it has often been reported as a soft,
unconsolidated blue clay which could not be distinguished from clay
of glacial age. Masses of this sort were seen at two places, at Papowka
south of Petrograd, and on the shore at Ontika. In neither case was
the clay actually in position either under or between layers of sand-
stone, but it lay in such a position that it could be readily conceived
that it was Cambrian clay which had worked out from a layer nearby.
In both cases it was very full of water, and it is probable that it
represented a portion of a stratum of shale which had been worked
up by the action of frost, water, and a creeping movement, until all
traces of the original stratification had been destroyed. Where mined
from the layers for the cement plant at Port Kunda, the clay is well
stratified, and hard. It is, however, very fine grained, soapy to the
touch, and a very fine plastic clay. The quickness with which it loses
its stratification on weathering is probably due to its fine grain and
the readiness with which it takes up water, rather than to the fact
that it has never been consolidated.

In discussing the finds of “Olenellus” at Kunda and near Reval by
Mickwitz, Marcou (31) proposed for the Lower Cambrian strata as

186 BULLETIN: MUSEUM OF COMPARATIVE ZOOLOGY.

developed at these two localities the geographic name Esthonian,
which may now with propriety be applied to all the Lower Cambrian
strata in the whole district. Of the two localities first mentioned,
Kunda, and Strietsberg near Reval, the former presents the better
outcrop, and it may be taken as the type-section of the Esthonian.

Ordovician.

Packerort formation. Upper part of As, and A3 (upper part of
Ungulitensand and the Dictyonemaschiefer or Alunschiefer), of
Schmidt. (Plate 7).

The most instructive section of the formation to which this new
name is given is to be found at the base of the perpendicular 80-foot-
high-cliff upon which is built the light-house Packerort at the end of
the peninsula north of Baltishport. At the base of the cliff one sees,
partly in the water, eight feet of hard, almost white, coarse-grained
sandstone, representing the top of the Esthonia formation. Resting
upon this is a bed of conglomerate, the matrix of which is an iron-
stained sandstone, and which contains well-rounded boulders ranging
from a few inches up to four feet in the greatest diameter. The
boulders are all of sandstone, with some small pebbles of quartz, and
are very numerous, making up the whole of that part of the formation
which rests upon the Cambrian. This conglomeratic layer is very
irregular, and only two or three feet thick. It is succeeded by alter-
nations of thin layers of sandstone and a very dark gray, friable, soft
shale. Above this comes a very irregular layer, five to ten feet thick,
of cross-bedded, coarse-grained sandstone with great numbers of
Obolus apollinis in the upper part. Then follows a band of thin-
bedded dark gray shale like that below, but in certain layers, contain-
ing great numbers of graptolites, principally Dictyonema flabelliforme.
As these strata rest upon the undulating surface of the sandstone
below they have a variable thickness, from thirteen to eighteen feet.

The conglomerate at the base of the formation was seen also along
the river north of the railroad bridge at Narwa. The pebbles at that
locality were all rather small, the largest seen being ten inches in

diameter. The interbedding of the shale and sandstone was seen also _

at Asserien, and although the sandstone and shale of this formation
are usually in distinct bands, these sections show that the two are
intimately associated and belong to the same time. The conglomerate
and sandstone indicate the shore phases of the earliest Ordovician
transgression; the shale with the graptolites a later shallow water

RAYMOND: CORRELATION OF THE ORDOVICIAN STRATA. 187

phase. The still later deeper water limestone phase, with the Cera-
topyge fauna, seems never to have reached this region.

Mickwitz fully recognized the physical evidences of an erosion
period between the Lower Cambrian and the time of deposition of
the Obolus sandstone, and in the preface to his paper on Obolus (33),
he gives some excellent detailed sections of the Obolus sandstone.
He recognized the basal conglomerate at Packerort, which he illus-
trated by a diagram, and he gave also a diagrammatic representation
of the strongly eroded top of the Lower Cambrian sandstone at Jam-
burg on the bank of the Luga, east of Narwa. At this latter locality,
the Obolus sandstone, with conglomerate, fills hollows and cracks in
the Lower Cambrian sandstone.

The dark shale at the top of the formation is thicker at Packerort
than in any other section, and, as has been noted already by Schmidt,
it thins to the eastward, until it is entirely absent at Narwa. Still
further east it comes in again and is seen far eastward, being four and
one half feet thick at Papowka and one foot on the Lawa at Wassil-
kowa. Bassler cites the variable thickness of the Dictyonema shale
as evidence of erosion before the time of deposition of the overlying
“Glauconite sandstone.”’ The evidence on this point does not seem
entirely clear, and the presence at Narwa of only four inches of “ Glau-
conite sand”’ at the point where all the shale is missing does not favor
that interpretation, as one would expect the greatest amount of sand
in the deepest erosion hollows. . Moreover, the glauconite sand is
thickest where the shale is thickest, and suggests the alternative
explanation, partly borne out by its fauna, that the “Glauconite
sandstone” may really belong to the Packerort formation, representing
the deposits of the third and emergent phase of the cycle. It should
be noted that the Dictyonema shale is usually unfossiliferous, fossils
being common only along the western portion of its outcrop, the most
western locality being on the island Odensholm where they are found
in loose pieces cast up on the shore and the most eastern so far re-
ported being on the Isenhof stream between Asserien and Ontika.!

1 Eichwald (17), however, reports Dictyonema from as far east as Zarskoe Selo, south of
Petrograd, and also at Narwa, where there is no Dictyonema shale. According to Schmidt
(47), Dictyonemas have been found in lenses of limestone at the latter locality, and this was
probably the source of the ones reported by Eichwald. This of course suggests that the Glau-
conite sand at Narwa may not be a representative of the real Glauconite sand as developed at
other localities, but a residium from the Dictyonema shale. Lamansky (29, p. 197) states
that graptolites have also been found in the Lower Linsenschicht at Narwa, and that they were
sent to Dr. Holm for study, but I have found nothing more in the literature about ‘them.
Lamansky thought that they would prove to be the same as those found by Holm in Oeland
(Tetragraptus fauna).

188 BULLETIN: MUSEUM OF COMPARATIVE ZOOLOGY.

At Packerort the most fossiliferous layers are in the upper part of
the shale, and it might be inferred from this that in the eastern region
the upper layers had been removed by erosion, but it is equally possible
that they were never deposited there.

The characteristic fossils of the Packerort formation are the various
species of Obolus, chiefly 0. apollinis, in the sandstone, and Dictyonema
flabelliforme in the shale. Species of other genera of inarticulate
brachiopods are found in the sandstone, and the shale has furnished
several species of graptolites which have not yet been satisfactorily
identified. If Schmidt’s (44, p. 16) figures are to be trusted there may
be a Didymograptus in this fauna.

The Obolus or Unguliten sandstone, has, like the Lower Cambrian
clay of the same region, often been cited as an example of a formation
which has never been consolidated. At nearly all exposures it is a
friable sandstone which crumbles readily under the hammer, but in
certain places it has considerable hardness, and one receives the
impression that the present condition is due to the removal of the
cement through leaching. The surface water enters the sandstone
through joints in the overlying limestone, and being checked in its
further downward passage by the Cambrian clays, naturally moves
through the sandy beds.

Walchow formation. B,and B,, (Glauconitsand and the Glauconit-
kalk) of Schmidt; B,, Byras Brg, Byry and Bis of Lamansky.

While the Patient formation is best developed at the extreme
western end of the Ordovician outcrop, the succeeding formation finds
its best expression in the east. This, however, is not due to the fact
that the Walchow formation was deposited in a sea invading from the
east, for the opposite seems to be the case, but because the upper
layers have been eroded away at the west, as has already been shown
by Lamansky (29). The lower members of the formation, the “Glau-
conitsand”’ and the “ Glauconitkalk”’ are better developed in the west
than in the east, and the deposits of the same age as the Glauconitkalk
are still thicker in Sweden.

On the Walchow and on the Lawa at Wassilkowa this formation
has five bands easily distinguished on lithological grounds, each with
its own faunal characteristics. The measurements given here are
those of the section on the Lawa which presents a more satisfactory
natural section than any seen on the Walchow. (Plate 4).

The lowest bed is a soft, easily disintegrated green sandstone, six
feet in thickness. Upon it rests two or three layers of limestone,
making a total thickness of six feet, which usually form a bold pro-
jection from the cliff, being preceded and followed by softer strata.

i? re

=
oo i a

|

"

]

,
alt

.

4

=

RAYMOND: CORRELATION OF THE ORDOVICIAN STRATA. 189

These layers are often vividly colored, being generally red or purple
with patches and spots of green and yellow, and usually contain
quantities of rather large green grains of glauconite. The characteristic
fossil is Megalaspis planilimbata. Above these layers comes a band,
thirteen feet in thickness, of thin-bedded, shaly limestone and shale
in which Asaphus bréggert and Onchometopus volborthi are found.
These strata weather to a soft gray mass, and above them are harder
layers of limestone with less shale, making the fourth division, eleven
feet in thickness. This also is a blue-gray limestone, and contains
Asaphus lepidurus and Megalasyis gibba in numbers. At the top
of the formation is another thinner-bedded, softer, gray and green
limestone, characterized by a great abundance of Asaphus expansus,
and containing also A. lamanskii, and Nileus armadillo, this limestone
being about ten feet thick. This makes the total thickness of the
formation on the Walchow and Lawa about forty-six feet.

When followed westward this formation becomes thinner and
usually at the expense of the upper members, though the green sand
may thin to practical disappearance. Thus, on the Papowka, the
green sand is only one foot thick, the Megalaspis planilimbata or
lowest limestone bed is seven feet thick, and is followed by twelve feet
of shaly limestone, the greater portion of which contains the Onchome-
topus volbortbi fauna, while at the top, Asaphus lepidurus and Megalas-
pis gibba are found. The layers with Asaphus expansus are gone
entirely. Further west, the shale almost entirely disappears from this
part of the section, though there is usually a thin shaly layer or a
shaly parting. The limestone of the section becomes very thin, but
the three faunas, M. planilimbata, Onchometopus volborthi, and Asaphus
lepidurus, persist as far west as Reval, though further west the Asaphus
lepidurus fauna is lost, and of the zone with Asaphus bréggeri and
Onchometopus volbortht only a thin remnant remains in the section at
Packerort. At this latter locality the green sand has the greatest
thickness known, eleven feet, followed by two and a half feet of hard
green limestone with large grains of glauconite and many trilobites,
among them Megalaspis planilimbata, then one foot three inches of
thin-bedded limestone and shale, this containing Asaphus bréggeri.
The limestone of the formation is therefore only three feet and nine
inches in thickness, the two younger faunas are absent entirely, and
the strata containing the others are very thin. Besides the absence of
the younger faunas there is other evidence which indicates that erosion
has taken place since the deposition of the upper strata of this forma-
tion. In the section at Packerort, the thin-bedded limestone is fol-
lowed by a conglomerate in which there are large numbers of pebbles

190 BULLETIN: MUSEUM OF COMPARATIVE ZOOLOGY.

of green and gray limestone and pieces of shale; and at Catherine
Park, Reval, the green “Glauconitkalk”’ is likewise followed by a six
inch layer of conglomerate in which there are pebbles of limestone
full of glauconite. This unconformity is not newly discovered, but
was distinctly foreshadowed in Schmidt’s papers, and was definitely
worked out by Lamansky, who, however, placed the layers containing
Asaphus expansus in the overlying formation instead of with the older
strata, as the evidence seems to require. (Plate 2).

The doubtful member of this formation is the green sand. It is
placed here, because there is an undoubted break in the sedimentary
record between the Dictyonema shale and the M. planilimbata lime-
stone. In Norway and Sweden one finds between these two forma-
tions the Ceratopyge limestone, with a fauna which, though it occupies
no great thickness of strata in Scandinavia, really endured for a very
~ long period of time. During this interval no deposition was taking
place in the region in Russia here discussed, and, apparently, neither
was there any great erosion, the district standing nearly at sea-level.
Durmg some part of this time the green sand seems to have accumu-
lated, perhaps as a beach sand, at least at first, but probably reworked
as a whole or in part by the invasion of the sea in which were deposited
the Walchow sediments. It differs from an ordinary beach sand not
only in its green color, but in the presence of much fine clay. It
usually shows neither stratification nor cross-bedding. The fauna isa
scanty one. In the west, on the Baltishport peninsula and near
Reval, a few specimens of Obolus lingulaeformis Mickwitz, a Lingula,
a Siphonotreta and conodonts have been found. At Papowka,
Lamansky has referred to the “Glaukonitsand” a sandy part of the
limestone, and has obtained from it a considerable fauna which he
considers to be distinct from the regular M. planilimbata fauna and
allied to the Ceratopyge fauna of Scandinavia. This fauna is how-
ever, too closely allied to the M. planilimbata fauna to indicate the
presence of either a Ceratopyge or Lower Didymograptus fauna, and
the strata containing it would seem to go naturally with the limestone
rather than with the sandstone of the section.

So far as I have seen it, the fauna of the green sand seems to be
allied with that of the Ungulite sandstone below, rather than with the
limestone above. The sand and clay content of the bed may easily
have been derived from the denudation of the underlying Packerort
formation, which was undoubtedly uplifted and subjected to erosion
at some localities, even though we can not now point definitely to the
particular places, and, such being the case, it seems more probable
that the sand belongs to the later and not the earlier sedimentation.

RAYMOND: CORRELATION OF THE ORDOVICIAN STRATA. 191

In the east, on the Lawa, Walchow, and other rivers of that district,
the limestone of the Walchow formation is rather soft, and disinte-
grates very readily on exposure, so that great numbers of very beauti-
fully preserved fossils may be obtained, especially in the extensive
quarries on both sides of the Walchow above Old Ladoga.

The lowest limestone has a decidedly green color, due to the presence
of a considerable quantity of glauconite. The glauconite is in small
grains about .5 mm. in diameter, and makes a considerable part of the
limestone. In Esthonia the corresponding bed is much harder, is a
deeper green in color and contains more and larger grains of glauconite,
up to 2 mm. in diameter.

All of the limestone in the formation on examination in thin section
proves to be made up almost entirely of fragments of fossils, largely
trilobites, but also ostracods, bryozoans, and brachiopods. All these
are in small pieces, sometimes 2 mm. long, but generally much less.
' The cement consists of exceedingly small grains of crystalline calcite.
Stray grains of glauconite are seen in most of the slides from all horizons
in the formation. The red and brown colors of some of the limestone
prove to be due to a stain and grains of limonite surrounding the
crystals of the matrix and filling the zoecia of the bryozoans. The
limonite halo around the grains of glauconite suggest that the source
of the iron compound may possibly be found in the decomposition of
that mineral. (Plate 5).

In a slide from Putilowa especially, very few of the glauconite
grains are unaltered, but almost all show a border of limonite, and
the grains contain much of a dark alteration product along cracks.

The glauconite from the glauconite sand and limestone has been
analyzed by Kupffer (27), some of whose analyses are quoted below.

i 2 3 4

SiO, 51.93 51.24 50.91 52.38
Al,O3 9.20 $231'22 ud, St 10.53
FeO; Lavo 13.44 16.54 Eee ys
FeO 4.73 3.06 4.80 4.36
MgO 3.79 3.93 Sx62 4.96
CaO 30 10 30 OS
KO 8.02 7750 8.09 8.00
NaO .20 ap | 14 04
HO Owe 8.20 6.48 5.88
Quartz 40

99.40 100.00 100.69 100.00

192 BULLETIN: MUSEUM OF COMPARATIVE ZOOLOGY.

The analysis given in column 1 was from the glauconite sand at
Karya-Oro near Ontika, Esthonia, and 2 from the glauconite lime-
stone at the same place; 3 was from the glauconite limestone, and 4
from the glauconite sand, at Baltishport.

For comparison, one may quote the following, the first two from
Clarke’s Data of geochemistry, p. 494, and the other three from an
abstract of a paper by Glinka. (Zeitschr. kryst. u. min., 1898, 30,
p- 390).

1 2 3 4 5

SiO, 5156 53.61 48 .95 49.53 52.96
Al.O3 6.62 9.56 1.66 vst Abi S4 12.76
Fe:O3 1565 21.46 23 .43 20.06 13.56
FeO 8.33 1.58 4552 5.95 2.34
MgO .95 2 87 2.97 2.92 4.11
CaO .62 1.39 mit | .56 A ga
Na,O 1.84 .42 .98 a6 AT
H:O 10.32 5.96 4.93 4.91 4.91
K,O 4.15 3.49 9.54 9.31 8.69
MnGooumantr, '% trace b the ae

99.55 100.34 100.35 99 .54 99.80

The glauconite in column 1 is from a greensand marl, Hanover
Co., Virginia; 2 is the mean of four analyses of deep-sea deposits
from the Challenger Report; 3 is a glauconite from the Cretaceous
sandstone at Padi, Government of Saratow, Russia; 4 is from an
Eocene sandstone in the Urals; 5 from the Glauconite limestone at
Udriass, Esthonia.

It will be noted that the Russian Ordovician glauconite contains
_less iron, more alumina, much more magnesia, and more potash than
the other glauconites listed.

Kunda formation. By, (Vaginatenkalk) of Schmidt; By, and
By of Lamansky.

This well-known formation may be seen throughout the whole
extent of the Ordovician from the Sjass to the Island Rogo, off Baltish-
port, but is best exposed in Esthonia. I have selected Kunda as the
type-section because it is there well exposed and richly fossiliferous.
A ‘drain recently dug by the Cement Company at the extensive
quarries about three miles south of Port Kunda on their private rail-
road exhibits a complete section of the formation, which is here fifteen
feet thick. A large quarry, opened during the summer of 1914 will,

RAYMOND: CORRELATION OF THE ORDOVICIAN STRATA. 193

if completed as planned, also furnish a complete section through the
limestone. This formation is quite thick in the eastern exposure on
the Walchow but is very poorly exposed, only the basal portion being
cut by the extensive quarries, and the greater part being seen only at
places along the river bank just below Dubowiki.

At the base of the formation one finds the so-called “ Lower Linsen-
schicht,” a rather soft clayey limestone six to twelve inches in thick-
ness, full of small flattened grains of about the size, shape, and color
of small Leperditias, with which Schmidt first confused them. These
small “lentils” have a concretionary form and have been shown on
chemical analysis by Kupffer (27), to consist of clay containing iron
oxides and calcium phosphate.

The layer containing these linsen does not seem to have been formed
under abnormal chemical conditions, for it is fossiliferous, often highly
so, being in fact noted as the best stratum for Pliomera fischeri and
Lycophoria nucella, and the fossils are of full size and show no abnor-
malities. It is probable that the “Linsenschicht” really represents a
basal conglomerate for the Vaginatenkalk, for in places, as at Reval
and Packerort, there is a real conglomerate which replaces the “ Lower
Linsenschicht.”’

East from Reval this conglomerate was not seen, and west from
that city there is no “ Lower Linsenschicht,”’ but both the conglomerate
and the “ Lower Linsenschicht” represent the basal bed of the forma-
tion.

The “linsen”’ of the Linsenschicht are almost opaque in the thinnest
sections which can be obtained of these soft rocks, but show a definite
concentric structure.

An analysis of some of them from Ontika by Kupffer is as follows:

MMM etre re renN yt, sive ttns, Mitt MSF ies te ot) ey Pa 5.93
enratrre rete” Se tiiOr VEST ee LOO ON eS a 3.95
FeO; SE EE aE OF ms SRN E CAC ean. Uc ighally Guat fb eure we eh eles 6 69 92
Mn,03 Pe an eT eee Ge alc Wiiuticet aon seavtee tile te leh e tal aren Swe ca NS ow ets eck 26
GL eae eee ey wae et tipo EI 1.21
Sree eee rr eee Ne SS 46
PA LAL Lom ee ee fe ee oS. 1.99
I a es ee a ee Obes SE a, 12.42
Camera cette Fo ORS yk .68
Merge et ey oo ee eerie. PS IS e, 3.09

194 BULLETIN: MUSEUM OF COMPARATIVE ZOOLOGY.

This analysis may be compared with those of the odlitic hematite
ore of the Clinton of New York, listed by Newland and Hartnagel
(Bull. 123, N. Y. state mus., 1908, p. 62).

1 2 3 . 4 5

FeO; 69.17 42.97 79.98 63 .00 71.82
SiOz 11.457. DO dae 9.98 12 6e. 11.34
Al,O3 3.92 4.13 - 2 4: 5.45 3.91
MnO 19 Pt tr. 45 1.63
OD cor rchcc eae BOT, 154 6:2 3.97
MgO iy | 1.96 ‘3 aw ed 2.21
S .28 837 2s

P.O; 1.726 1.534 1.239 1 2.096

Comparing these analyses, it will be noted that the “linsen” have
about the same iron content as some of the Clinton odlites, and about
the same amount of clay, manganese, and phosphates, but less silica,
lime, and magnesia. The high silica content of the Clinton ore is due
to the presence of nuclei of sand in the spherules, whereas the nuclei
of the linsen, when such can be observed at all, seem to be calcitic
fragments of fossils. |

The linsen have much the same size and shape as the spherules in
the odlitic Clinton ore, most of them being from .5 to 1 mm. in diameter,
and somewhat flattened or lentil shaped. This flattening, in the case
of the Clinton odlites, has been ascribed to pressure, but in the case
of the linsen it seems to be the original form, for, while these dises
often lie parallel to the bedding, very large numbers of them do not,
but are imbedded at all angles.

The mode of occurrence of these linsen has some bearing upon the
rival theories of the origin of the odlitic sedimentary hematite ores.
The view put forward by Shaler was that they were replacements of

original limestone effected by the circulation of ground water, while -

C. H. Smyth, Jr., considers these ores to be original sedimentary
deposits. The Russian occurrences are explainable only by Professor
Smyth’s views, since: —

First; the linsen occur in a limestone which is not otherwise odlitie.

Second; the linsen occupy definite layers which can be traced
laterally some 300 miles through a series of gentle undulations, always
maintaining a definite horizon, as shown by evidence of fossils, and
without any relation to the present water table.

Third; the lower linsenschicht passes laterally into a true conglomer-

«
ae 7 “

i

RAYMOND: CORRELATION OF THE ORDOVICIAN STRATA. 195

ate showing that it was formed at or near the shore, and at a time
immediately subsequent to a period of erosion.

The Kunda formation has a somewhat irregular thickness, being
thickest in the east and very thin at the west. On the Walchow it is
thirty-two feet according to Lamansky, at Papowka it is thirty-four
feet with the top not seen, at Ontika eighteen and one half feet, at
Asserien fourteen and two thirds feet, at Kunda fifteen feet, at Reval
four feet, and three and one half feet at Packerort. The abundant
fauna at Reval is practically the same as that at Kunda, but as the
fauna seemed to be the same all through the section at Kunda, this in
itself would not indicate whether the thinness at Reval was due to
erosion at the top of the formation or to a smaller original deposition.

Schmidt states that west of Reval the Orthoceras limestone passes
into a sandstone, but I myself saw no evidence of this, either at Baltish-
port or on the Island Rogo. At these localities the formation consists
of a rather thick-bedded hard limestone without many fossils, and at
the base is a conglomerate made up of pebbles of green glauconitic
limestone and irregular pieces of dark shale, these latter proving on
analysis by Kupffer to contain, in some cases, a large percentage of
phosphoric acid.

The fauna of the Kunda formation is dominated by Mollusca,
mostly cephalopods and gastropods. Pelecypods are rare, making
here their first appearance in the Russian section. Typical fossils
are Vaginoceras vaginatum, V. commune, Maclurites helix, Estonio-
ceras lamellosum, Asaphus raniceps, Pliomera fischeri, Lycophoria
nucella, and Pterygometopus sclerops.

WIERLAND GRouP. Cl, 2 and 3, and part of D, (Echinosphaerites
limestone, Kuckers schicht, Itfer schicht, and basal portion of the
Jewe schicht), of Schmidt.

As already noted, Schmidt gave geographic names to all the strata
above the “Glint,” but these names are of very unequal value, some
of them designating true formations, and others indicating merely
the quarry at which a certain fauna or type of strata was seen. The
faunas of all three of the formations named above are very closely knit
together by the presence of Echinosphaerites aurantium and species of
Chasmops. The name Wierland which I have applied to the group
is that of the district in which most of the localities for Kuckers and
Itfer are located, and in which the lower divisions are well developed.
The lower members are also given geographic names to correspond to
the two upper members named by Schmidt. There is really a greater
faunal change between the Reval and Dubowiki members than be-
tween any other two in the group.

196 BULLETIN: MUSEUM OF COMPARATIVE ZOOLOGY.

Dubowiki formation. C, (Upper Linsenschicht and lower part of
the Echinosphaerites limestone), of Schmidt.

This formation, like the Kunda limestone is best exposed at the
east, where it reaches its best development on the Walchow River at
St. Michael Archangel, opposite Dubowiki, just above the steamer
landing and below the railroad bridge. At this locality the base of
the formation is not seen, but fourteen feet of soft caleareous mudstone
are exposed, the base twenty-three and one half feet above water level
in the river. This outcrop is capped by twelve feet of the harder
dolomitic limestone of the Reval formation. Schmidt and Lamansky
state that the upper part of B,,, (the Kunda formation) is to be seen
in the basal parts of the quarries at St. Michael Archangel, but I was
not able to find it, and so did not see the contact between the two
formations here. The upper part of the Kunda is, however, exposed
along the riyer bank about a mile below the steamboat landing and
with the prevailing low dip should still be above the water-level at St.
Michael Archangel. The thickness of the Dubowiki at this locality
is therefore uncertain. It can not be more than thirty-seven feet or
less than fourteen feet in thickness, and is probably twenty-five to
thirty feet, as Dubowiki fossils, which seemed to be in place, were
found within ten feet of the water’s edge.

Schmidt and Lamansky agree that there is no “Linsenschicht”’ at
the boundary between B,,, and C, at this locality. Following the
Dubowiki westward it is present at various sections, but always
thinner than at the typical locality. It is well exposed in the cement
quarry and on the railroad south of Asserien, where it is fifteen and
one half feet in thickness. It is here a hard compact limestone,
unlike the soft marly beds at Dubowiki; and at this locality, as well
as at Ontika and all the other localities in Esthonia the “Upper
Linsenschicht”’ is present at the base of the formation. This linsen-
schicht is not a definite, rather thin band, like the Lower Linsenschicht,
but the linsen are smaller, less abundant, and scattered through a
thickness of six or seven feet. Continuing westward, the Dubowiki
formation thins out entirely, so that at Reval the upper Linsenschicht
is reduced to a thickness of one foot and at Baltishport to ten inches,
and it is at the base of the Reval formation instead of the Dubowiki.
The Upper Linsenschicht is therefore a tangential formation and
represents the invading base of the Wierland group.

The fauna of the Dubowiki retains some survivors of previous
faunas, though very few species are common to this formation and
those below. This formation is particularly marked by the intro-

RAYMOND: CORRELATION OF THE ORDOVICIAN STRATA. 197

duction of Echinosphaerites and Chasmops, and the acme of the varia-
tion of the genus Asaphus. Echinosphaerites is not found in the
Linsenschicht at the base of the formation, though it does oecur with
linsen a few feet above the base at some localities (Ontika and As-
serien).

The soft limestone from St. Michael Archangel, on the Walchow,
weathers to a nearly white flour which, when wet, forms a very sticky
mud. A thin section shows that this rock is made up almost entirely
of very small fragments of fossils, few of which reach 1 mm. in length
and none have more than one fourth that thickness. The most
abundant fragments are of some organism with minute tubules,
possibly a Solenopora. Bryozoa, Ostracoda, and trilobites seem to
furnish a large part of the material. The fragments are much more
finely comminuted than in the limestone of the Walchow.

Characteristic fossils are: — Echinosphaerites aurantium, Clitam-
bonites adscendens, Porambonites aequirostris, Chasmops nasuta, Cerau-
rus exsul, Illaenus tauricornis, Asaphus cornutus, and A. kowalewski.

Reval formation. Cig (upper part of Echinosphaerites limestone)
of Schmidt.

Resting upon the Dubowiki at Dubowiki on the Walchow, and
through the greater part of Esthonia, and upon the Upper Linsen-
schicht from Reval westward, is a hard, compact, sparingly fossili-
ferous limestone, frequently magnesian in character, to which the
name Reval may be applied, as it is very extensively quarried at that
locality. The thickness and lithological character of this formation
are remarkably uniform all the way from Baltishport to Dubowiki
and it is a favorite quarry rock wherever accessible. The beds vary
in thickness from an inch to about a foot and afford both building and
flagging stone. It is extensively used for both purposes in Reval,
Narwa, and Petrograd. Certain of the layers are traversed by vertical
tubes suggesting worm-burrows. The thickness varies from twenty-
five to thirty-five feet. Fossils are not very common, and in many
eases dolomitization has gone on to such a degree that the rock has a
porous appearance and the fossils are represented by hollow molds.

The rock has about the same color and appearance as the Galena
of Minnesota, and there does not seem to be much question but that
the dolomitization has here taken place in beds originally composed
mostly of limestone. In thin section the rock from Dubowiki shows
irregularly intergrown areas of very small crystals of calcite with
irregular boundaries, and areas in which the crystals are of dolomite,
about twice the size of those in the areas of calcite, and with definite

198 BULLETIN: MUSEUM OF COMPARATIVE ZOOLOGY.

crystal form. This rock is not made up of fragments of fossils, like
that of the calcareous formations below.

The leading fossil of this formation is Christiania oblonga (Pander).
Some large cephalopods are found, and in the Government of Petro-
grad, Cryptocrinites laevis (Pander) is a characteristic fossil, and
Echinosphaerites aurantium (Gyllenhahl) is occasionally found.

The Reval is the youngest formation which appears in the “Glint.”

Kuckers formation. Cz, (Kuckerssche schicht or Brandschiefer)
of Schmidt.

The Kuckers formation takes its name from the estate of Baron
Toll, about five miles northeast of the railroad station Jewe; but the
strata are very little exposed at that locality and collecting is now very
poor. The same strata are reported by Schmidt to be exposed in a
few natural sections along streams but the places were difficult of
access and we saw the Kuckers only at the typical locality and at
Reval. The formation occupies the low, level land at the top of the
escarpment which faces the Gulf of Finland, and the numerous
beautiful fossils which it has produced have come from ditches dug
to drain this sort of land. (

The base of this formation can be seen in the extensive quarries
at Reval where the upper three or four feet are a-bluish gray calcareous
shale and thin-bedded shaly limestone containing numerous cystids,
including Echinosphaerites aurantium, Caryocystites balticus, and C.
aranea.

In the trench at Kuckers the strata consist of gray and reddish
earthy limestone and soft reddish shale known from its combustibility
as the “ Brandshiefer.”’ Schmidt has listed localities in this formation
all the way from the village of Djatlizy south of Gostilizy and west
of Petrograd to the point where it goes beneath the waters of the Gulf
of Finland southwest of Baltishport. Its large fauna is rather easily
recognized. Chasmops odini, Ceraurus spinulosus, Porambonites
teretior, Plectambonites sericeus, Oxoplecia dorsata, Echinosphaerites
aurantium, and Platystrophia lynx were the more common species seen
by the writer. Of these, the Oxoplecia dorsata is most valuable as a
cosmopolitan form, but, in Russia, confined apparently, to this horizon.

Schmidt estimated the thickness of the Kuckers at from thirty to
fifty feet.

Itfer formation. Cs; (Itfersche schicht) of Schmidt.

This member is named from an exposure on the estate of Baron
Wrangel at Itfer, northeast of Wesenberg. This exposure, a small
quarry, is now completely overgrown and nothing is to be seen. |

RAYMOND: CORRELATION OF THE ORDOVICIAN STRATA. 199

was able to find the formation in another small, shallow, old quarry
at Wannamois, and in a small ditch at Tolks. The strata here consist
of thin-bedded (layers two to four inches thick) gray and almost white
limestone, rough to the touch, full of silica, and with silicified fossils.
The fossils included Echinosphaerites aurantium, which species we also
found in the lower layers of strata assigned to the next formation, the
Jewe, at the quarry south of the Guthof at Kuckers, and at Aluver,
north of Wesenberg. As this species does not occur in the typical
Jewe, I propose to extend the Itfer to include all the strata at the above
localities which contain Echinosphaerites. These strata lack the
shale of the Kuckers member and are lithologically unlike the Jewe,
as they weather to a grayish white instead of a rusty yellow.

The geographical distribution of the Itfer is unknown. It is
difficult to trace, as it has few fossils peculiar to it, and no very dis-
tinctive lithological characteristics. It has not been identified outside
the vicinity of the typical locality, but Baron Toll called our attention
to an outcrop of strata on his estate which were stratigraphically a
few feet above the typical Kuckers in the “Graben,” and which may
prove to be Itfer. They consisted of a thin-bedded soft, earthy, gray
limestone, and contained too few fossils to permit of positive identifi-
cation of age.

Schmidt estimated the thickness of the Itfer at twenty to thirty feet.

Jewe formation. Dj, (Jewesche schicht, except for the basal portion),
of Schmidt, but not including the Kegel and Wassalem.

The Jewe is a formation with distinct lithological characteristics,
contains a well-marked and easily recognized fauna, and is well ex-
posed along a line extending from Gatschina in the Government of
Petrograd to the coast near Spitham in the northwestern corner of
Ksthonia. |

At the type-locality, Jewe, in an abandoned quarry south of the
railroad there is an exposure of about twelve feet of light gray to
yellow magnesian limestone of earthy texture. Some layers are more
shaly than others and weathering brings this out strongly. Still
higher strata of the same formation are to be seen a mile to the south-
west on the Gut Eichenheim, where, in similar strata, fossils are some-
what more plentiful.

A much better exposure of the Jewe is that at Aluver on the rail-
road to Kunda, three ntiles northeast of Wesenberg. Here about
twenty-five feet of the Jewe are shown in a quarry, with the upper part
of the Itfer, full of Echinosphaerites, exposed at the lower part. The
rock is a fairly compact bluish limestone with earthy texture; on

200 BULLETIN: MUSEUM OF COMPARATIVE ZOOLOGY.

weathering it becomes a mass of rusty yellow fragments. Fossils are
very plentiful in the upper part of the quarry. The Jewe covers a
large area north of Wesenberg and small quarries and ditches furnish
many exposures. I saw the Jewe further west beyond Nemme, about
seven miles southwest of Reval and at St. Mathias, five miles south of
Baltishport. At both these localities the fossils and lithology were
the same as at Jewe itself and the formation is throughout its extent
a very distinctive one. The most common and characteristic fossils
are: — Platystrophia lynx (very robust variety), Clitambonites schmidti,
Hemicosmites extraneus, and Poramborites ventricosus. Equally charac-
teristic are the peculiar conical bodies figured by Schmidt (44, p. 331).
These appear to be of organic origin, but their exact nature is not
known.

Kegel formation. Dz and Ds; (the Kegel, Wassalem and, west of
Reval, the “ Wesenberg’’) of Schmidt.

At the typical locality, at Kegel, southwest of Reval, about eight
feet of strata are exposed in two quarries about one and one half miles
west of the station. The strata here are limestone without shale, in
layers two to six inches thick. When fresh the limestone is blue and
fine grained, but weathers to a yellow shaly mass. The fossils weather
more rapidly than the matrix and the rock is left full of holes. The
most abundant fossil is Cyclocrinites spasski, which occurs in immense
numbers. Clitambonites anomalus and Asaphus kegelensis are also
quite common. The country south of Kegel is very flat and the rock
everywhere near the surface. Following the railroad or highway
southwest from Kegel station, the Kegel beds with their characteristic
fossils are seen in ditches and shallow quarries till one comes to a
broad low ridge which is made up of a very different rock, to which
the name Wassalem has been given. At the large quarries in Wassa-
lem, the strata are light gray to white, very coarse-grained massive
limestone, the lower ten feet with feebly developed partings, the upper
three feet very irregularly bedded and containing some shaly lenses.
The lower part is quarried in large blocks, up to three feet in thickness,
for use as a marble. In this portion there are few fossils, other than
joints of the columns and plates of Hemicosmites. Weathered pieces
show that the rock is practically made up of these. The upper three
feet contain lenticular and cross-bedded strata and lenses of fine-
grained buff limestone with numerous specimens of Ilaenus. Fossils
may be found in this upper portion, especially in pockets where the
limestone has decomposed, leaving a mass of yellow, calcareous earth.
The most common fossils are bryozoans and Hemicosmites. The

——————

|
|
j
;
:
;
,

RAYMOND: CORRELATION OF THE ORDOVICIAN STRATA. 201

Wassalem thus shows at its type locality many of the characteristics
ofareef. The outcrop of the Wassalem has a width of about two miles
and south of it one finds blue and buff, very fine-grained dense lime-
stone, somewhat purer than that at Kegel, but with the same fossils,
Cyclocrinites spasskii (or C. roemeri, as Stolley calls it) being very
abundant. This limestone appears to belong to the Kegel, though
it has previously been called Wesenberg. The reasons for this belief
are givaén on page 202. The Bryozoa described by Bassler as coming
from the Wassalem were very probably derived from a lense of the
fine-grained buff limestone associated with the reef, for their appear-
ance and the lithology of their matrix is entirely unlike that of the
typical Wassalem. (Plate 6).

Wesenberg formation. E (Wesenberger schicht, partim), of Schmidt.

The strata of this formation are well shown in three or four shallow
quarries about one and one half miles southeast of the town from which
it derives its name. The limestone is a very fine-grained, dense, blue
to yellowish buff rock, so fine grained as to have received the name of
“lithographic stone.” It is usually in layers three to five inches in
thickness, the layers separated by thin shaly partings. The good
fossils adhere to the limestone and stand out in relief when the shale
is washed away. The deepest quarry shows a face of sixteen feet,
the lower eight feet being compact light blue limestone and the upper
eight feet somewhat less compact and more magnesian limestone
which becomes yellowish on weathering. Lithologically these strata
differ from the rocks of the Kegel] at Kegel in being less earthy, more
compact, and in containing definite partings of shale. Fossils are
exceedingly abundant in these quarries, the most conspicuously com-
mon being Amphilichas holmi, Homolichas eichwaldi, Chasmops
wesenbergensis, and Encrinurus seebachi.

The Baltic railroad runs in a southwesterly direction from Wesenberg
to Taps and thus traverses the outcrop of the Wesenberg diagonally.
Between the two stations there are several small cuttings, one of them,
about two miles east of Taps, being in strata very near the top of
the formation. A tiny quarry, a few rods south of the railroad and
near the stream just east of Taps shows strata even higher in the
formation. In both localities the rock is a hard fine-grained yellow-
ish to buff limestone, without shale, and fossils were exceedingly
scarce. In the small quarry I obtained Clhtambonites wesenbergensis,
a Discoceras, and Chasmops wesenbergensis, fossils which are char-
acteristic of the Wesenberg at the type-locality. A half mile south
of these outcrops one finds the Lyckholm, with typical fauna.

202 BULLETIN: MUSEUM OF COMPARATIVE ZOOLOGY.

The Wesenberg was believed by Schmidt to be a thin formation,
the thickness being estimated by him at not more than thirty feet.
From the width of the outcrop in the vicinity of Wesenberg, one would
expect a somewhat greater thickness.

DISTRIBUTION OF THE KEGEL AND WESENBERG FORMATIONS.
,

According to Schmidt the outcrops of the Kegel and Wesenberg
strata form parallel bands extending from the western part of the
Government of Petrograd to the western border of Esthonia. In the

eastern part of this belt the Wesenberg is said to overlie the Kegel °

proper, while in the western area the Wassalem intervenes between
the Kegel and Wesenberg. I regret to say that I have not been able
to trace these formations in the field as I should like to, but from what
I have seen in the course of traverses in the neighborhood of Wesen-
berg and Taps, and between Baltishport and Hapsal, and the débris
on the northern end of the Island of Dago, I very much doubt whether
these formations do outcrop as parallel belts. The distribution of
the Kegel is given by Schmidt in detail, as follows: — The most
easterly outcrop is at Poll (a short distance east of Wesenberg) where
the Kegel is said to outcrop in the ravine and the Wesenberg on the
bank above; then north of Wesenberg, at New Sommerhusen, west
of Taps on the railroad between Kedder and Rasick, at Penningby,
Nappel, Jelgimaggi (south of Reval) Friedrichshof, Kegel, Habbinem,
and Kreuz.

I did not see the locality at Poll, but visited the old quarry at New
Sommerhusen, where the lithology and fossils are both typical of the
Jewe, and not at all Kegel. The locality “north of Wesenberg”’ is
probably an outcrop on the road to Haljal, and about three miles
north of Wesenberg. Here, where the road mounts a slight terrace,
is an exposure of nine feet of bluish and yellowish compact limestone
containing many fossils, among which were Amphilichas holmi and
the large Porambonites so common in the quarries at Wesenberg, and
which are believed to be characteristic of that formation. If the
strike here is approximately east and west, as it is supposed to be,
and as it actually is in most places, then the strata at this locality
must be but a short distance above the top of the Jewe, which out-
crops at Aluver and New Sommerhusen at approximately the same
level. Ina ditch at Welch, about five miles northwest of this out-
crop, I obtained, through Herr von Dane, some specimens whose

———— |

RAYMOND: CORRELATION OF THE ORDOVICIAN STRATA. 203

matrix reminded me of the Kegel, but which were not diagnostic
species. North of Welch there are numerous outcrops of the Jewe.
If the Kegel be indeed present in the neighborhood of Wesenberg, it is
either very thin or else does not carry the fauna of the Kegel at Kegel.

The next locality mentioned by Schmidt, that between Kedder and
Rasick, is forty miles west of Wesenberg, and less than thirty miles east
of Kegel. I did not visit this Jocality myself and am unable to find any
adequate faunal lists for it or for any of the other localities mentioned
between it and Kegel. Owing to its proximity to the latter place,
however, it is very probable that one finds here a real Kegel fauna.

Schmidt gives the following localities from east to west, for the
Wesenberg: — from Polja, on the River Pljussa in the western part
of the Government of Petrograd, then at Paggar, Piilse on the stream
Isenhof, at Poll, Raggafer, Wesenberg and other outcrops on the
railroad in that vicinity, at K6rwekiill north of Taps, at Wait south and
a little east of Reval, then southwest of Reval at Forby, Munnalas
and Paekiil, and as boulders on the Islands Oesel and Dago. Of these
localities I have seen only Wesenberg and the localities on the rail-
road as far as Taps, and the loose boulders on Dago, but have also
seen material from Munnalas. The fossils listed by Schmidt from
Potja, Paggar, Piillsse, Poll, Raggafer, and Korwekiill leave no doubt
that these eastern localities belong to the Wesenberg. When one
inspects the lists, usually very meager, from the more western localities,
beginning with Wait, one finds however, a marked change. In these
localities the common, and usually the only fossil, is Cyclocrinites
spasskii, a typical Kegel fossil, but one so rare at Wesenberg that I was
unable to find it, though Schmidt has listed it from that locality.

On the shore at Kertel, on the northern side of the Island Dago,
numerous angular blocks of limestone are found which are not seen
in place, but which are evidently derived from a ledge not far below
water-level. The blocks contain great numbers of Cyclocrinites
spasskii and lesser numbers of other typical Kegel fossils. At the new
factory at Hohenholm, west of Kerte], this same limestone was seen
in a trench immediately in contact with the Lyckholm.

The fauna of the Kegel has never. been carefully listed, the best
enumeration being that given by Schmidt (44, p. 34). This one is,
however, subject to considerable revision, and contains fossils found in
both the Jewe and the Wesenberg. In both the Kegel and Wesenberg
the trilobites are most important, because best known, and a study
of their distribution throws considerable light on the present subject.

Of nine trilobites which are supposed to be restricted to the Wesen-

204 BULLETIN: MUSEUM OF COMPARATIVE ZOOLOGY.

berg, four are found at Wesenberg only, one only at Wesenberg and
Raggafer, and four are found in the western as well as the eastern
localities. These are Homolichas eichwaldi and Isotelus remigerum,
which occur at Forby as well as at the eastern localities, Chasmops
wesenbergensis which is found at Wait and Forby, and loose on Dago,
and Pterygometopus nieszkowskii, found at Wait and Munnalas.

There are eleven species of trilobites reported from the Kegel; six
of which are common to the Jewe and Kegel and thus of no importance
in this discussion; four, Pterogometopus kegelensis, Chasmops brevispina,
Basilicus kegelensis, and Illaenus linnarssoni, are found only in the
western localities; and a single one Asaphus lepidus var. kegelensis, is
reported from both east and west. This last species has no particular
value, for it is very like Asaphus lepidus jewensis; and it is reported
by Schmidt not only from New Sommerhusen, which we know to be
Jewe, but also from localities in the Government of Petrograd east of
the limits which Schmidt himself set on the distribution of the Kegel.
Of the six trilobites found in both the Kegel and Jewe, five are found
in the Kegel in the typical region, while one, Chasmops mutica, is
listed as a Kegel species only from its occurrence at New Sommer-
husen. That there should be five species common to the lithologically
unlike Jewe and Kegel, and no species common to the lithologically
alike Kegel and Wesenberg strikes one as strange.

The results of the study are rather suggestive. Subtracting the
one species reported from the “Kegel” at New Sommerhusen, there
are nineteen species of trilobites reported from the Kegel and Wesen-
berg. Of these no one is reported as common to the two, while six of
the species in the Kegel occur in the Jewe below. Of the four which
may be considered strictly typical of the Kegel, not one is found at
any locality of the Kegel east of the locality on the railroad near
Kedder, forty miles west of Wesenberg. Of the nine trilobites in
the Wesenberg, five are restricted to the typical region about Wesen-
berg and do not occur in the western region, and four are reported in
both eastern and western localities, three of them at Forby, one at
Wait, one at Munnalas, and one on Dago.

The most abundant fossil in the Kegel at Kegel itself is Cyclo-
crinites spasskii, using that term in its old, broad sense.! Following

1 This usage is, I believe, fully justifiable. All of the five species described by Stolley (51),
from Esthonia were found by him associated in the same blocks, so that, so far as their strati-
graphic value is concerned, one specific name is as good as five. Most of Stolley’s specimens
seem to have come from loose boulders at localities south of the actual outcrop of strata con-
taining these species.

a =a

OE a

RAYMOND: CORRELATION OF THE ORDOVICIAN STRATA. 205

the railroad or highway southwestward from Kegel, one continues to
find Cyclocrinites as the common fossil until the coarse-grained, white
limestone of the Wassalem is reached. After crossing the outcrop
of this formation, the beds above are similar to those below, though
with less shale, and still full of the Cyclocrinites.

- Cyclocrinites seems to be confined very largely to the district west
of the longitude of Reval. It is reported by Schmidt from the Jewe
at Jewe and from the Wesenberg. At Jewe I succeeded in finding a
few small specimens of Coelosphaeridium cyclocrinophilum, and this
is probably the fossil which Schmidt had seen. At the quarries at
Wesenberg I saw no Cyclocrinites, though I looked for it particularly,
especially on my second visit, after I had collected many specimens
at Kegel and in the loose blocks on Dago. It is therefore, I think,
safe to assert that Cyclocrinites is a very rare fossil, if present at all,
at Wesenberg.

Stolley reports no species from the quarries at Wesenberg, though
he visited that locality, and also had access to the material collected
by Schmidt (in Dorpat). Stolley (51) described or reported five
species, Cyclocrinites balticus, C. schmidti, C. mickwitzi, C. roemeri, and
C. spasskii, from Esthonia, all from the region southwest of Reval,
and in the strata above the Wassalem.

At the United States National Museum I have seen specimens
collected by Professor Schuchert at Wesenberg while in company with
Akademiker Schmidt, and which are labeled Cyclocrinites spasskiv.
These specimens are none of them spherical, though some of them
might be interpreted as fragments of spheres. Moreover, they do not
show the surface structures of Cyclocrinites, and they do show that if
they were originally spherical, they were not hollow spheres, but had a
structure extending nearly to the center, as in Coelosphaeridium. I,
myself, collected many similar specimens, as they are very common at
Wesenberg. They are certainly not Cyclocrinites, and probably not
Coelosphaeridium, but this identification, which was probably made
by Professor Schmidt, explains the listing of Cyclocrinites from Wesen-
berg.

Summarizing what has been said on the preceding pages, it appears
that: — ,

lst, the fauna of the strata above the Wassalem is more like that of
the Kegel than that of the Wesenberg.

2nd, that the typical Wesenberg fauna is not found in the same
region as the typical Kegel fauna, but that both the Wesenberg and
the Kegel rest upon the Jewe and are followed by the Lyckholm.

206 BULLETIN: MUSEUM OF COMPARATIVE ZOOLOGY.

The question suggests itself as to whether the Kegel may not be of
the same age as the Wesenberg, instead of being older as has been
supposed. On the basis of the faunas this must be at once answered
in the negative, for there are only one or two of the long ranging
species which are common to the Jewe and Wesenberg, while there

are quite a number of species, particularly trilobites and brachiopods,

common to the Jewe and Kegel.

The presence of a few of the Wesenberg trilobites at localities south
of the outcrop of the Wassalem suggests that there may be a thin edge
of the Wesenberg in that region, probably overlying the strata with
the Cyclocrinites, but I did not have time to search for outcrops which
might have shown such relations. It seems more probable, however,
that these trilobites are not restricted to the Wesenberg horizon,
but are found in the Kegel as well.

It seems very possible therefore, that the Lyckholm rests at the
west on the Kegel and further east upon the Wesenberg, and there is
undoubtedly an unconformity at the base of the Lyckholm, for there
is at most, only a very small fraction of the normal thickness of the
Wesenberg present south of Wassalem. The relations of the forma-
tions may be as represented on Plate 8.

THE LOWER AND MIDDLE ORDOVICIAN OF SWEDEN.

To make a direct correlation between the various subdivisions of
the Ordovician in Russia and North America is impossible, the testi-
mony of the few species common to the two areas being entirely out-
weighed by the general unlikeness of the faunas. It was a realization
of this fact which caused a visit to Sweden and Norway after studying
the Russian sections. I visited the rather complete and easily ac-
cessible section at Kinnekulle, and other sections in Vasterg6tland at
Hunneberg, Ekedalen, and Alleberg near Falképing. In Ostergét-
land I collected at the large quarries at Borghamm and at the old
quarry at Vistani, the Husbyfjol of the literature on trilobites, and
visited a number of very poor localities in the vicinity of Motala. In
Scania I had the very kind guidance and assistance of Professor Dr.
J. C. Moberg, without whose help it would have been impossible for
me to have understood the very imperfectly exposed sections on the
Fagelsing and at Jerrestad. I was not able to visit Oeland on account

{
‘-
>
|=

RAYMOND: CORRELATION OF THE ORDOVICIAN STRATA. 207

of the war, neither did I see Dalecarlia, but Dr. Twenhofel made some
collections for me from the district about Rittvik, on Lake Siljan.
In Norway, I visited, under the guidance of Dr. Holtedahl, the
Ordovician sections in the vicinity of Christiania and, partly with Dr.
Holtedahl and partly with Professor Kiaer, some of the exposures of
Stage 4 in the Ringrike district.

The description of the Palaeozoic strata of Sweden has been very
ably summarized in English by Professor Moberg in his Historical-
Stratigraphical Review of the Silurian of Sweden and a briefer sum-
mary of certain facts in regard to the lower beds of the Ordovician
has been published by Fearnsides (59). I am indebted to these two
papers, and to the original sources from which their facts were derived,
for the greater part of what is here set forth in regard to the geology
of Sweden. The greater part of the Ordovician of Norway is still
inadequately known. The account here is derived chiefly from the
works of Brégger (93-95), and Holtedahl (97). No account in English
of this section, except that in Geikie’s Geology, has, so far as I know,

~ appeared.

The strata of Ordovician age in Sweden are found in isolated patches,
usually of small area. These patches may be grouped in bands,
having a roughly parallel NE-SW alignment.

The northern band, that in Jemtland and Lapland, has the greatest
extent, running far north parallel to the mountains on the boundary
between Norway and Sweden. These rocks are, however, except in
the southern part, largely metamorphosed. The next band to the
south of this has its best exposures in Dalecarlia, and there is a very
small area in Gistrikland, especially on the little island of Lim6én near
Gifle. From the evidence of boulders and fossiliferous sand in cracks
on the Aland islands, it would appear that this band may once have
connected with the Esthonian strata. To the westward the strata
of the Christiania district of Norway are in line with these patches.

The next band is in south-central Sweden, and includes the deposits
in Nerike, Ostergétland, and Vistergétland, while the fourth band is
at the extreme south and includes Scania and Oeland, while the Silu-
rian of Gotland is in the-same line. The strata are best exposed and
least disturbed in Vastergétland and Oeland. A brief description of
the strata at some of the principal sections in each region follows. It
is rather interesting to note that most of the principal lakes of Sweden
are connected with these limestone patches. Thus we find Lake
Storsjén in Jemtland, Siljan in Dalecarlia, Hjiilmaren in Nerike, and
the largest lakes, Vinern and Vattern in connection with the Palaeo-
zoic deposits of Viastergétland and Ostergétland.

208 BULLETIN: MUSEUM OF COMPARATIVE ZOOLOGY.

JEMTLAND.

The most recent general account of the Ordovician of Jemtland is
that given by Wiman (87), to which account must be added certain
facts obtained later by Moberg (75), Wiman, and Hadding (60).

The section of Palaeozoic rocks there is given by Wiman as follows:

Pentamerus kalk.

?
Chasmopskalk with Graptolithenschiefer.
Orthocerenkalk.
Underer Graptolithenschiefer.
Ceratopyge kalk?

Olenidenschiefer.
Alunschiefer
| Paradoxidenschiefer.

Quarzit.

The so-called Ceratopyge limestone is conglomeratic at the base,
containing fragments of the Olenus shales. Above it is a limestone
with much glauconite, and at the top a somewhat pure limestone.
This limestone contains some fossils, referred by Moberg definitely
to the Ceratopyge fauna but which seem to indicate fully as much
affinity with the Planilimbata limestone. Moberg lists, from Tos-
sasen: — Orthis christianiae, Niobe laeviceps, and a Cyrtometopus. At
Kl6fsjé he found Niobe insignis and a Megalaspis like M. stenorha-
chis Ang. The thickness of this limestone is not stated but one would
infer that it was about one meter.

This is succeeded by a green and gray graptolite-bearing shale,
which with the limestone below, make a total thickness of fifteen
meters. In the shale are lenses and one continuous bed of limestone.
The shales have afforded Wiman: —

Pliomera sp. Didymograptus filiformis Thg.
Megalaspis sp. D. hirundo Salter.

Leptaena sp. Phyllograptus sp. ind.
Tetragraptus serra. Tetragraptus quadribrachiatus.

In the limestone he found: —

Megalaspis sp. Ampyx sp.
Niobe laeviceps Dalm. Orthis sp.

RAYMOND: CORRELATION OF THE ORDOVICIAN STRATA. 209

To the writer it would appear that this entire series, including the
limestone at the base, belonged to the Phyllograptuschiefer and that
we have here Planilimbata limestone and Phyllograptus shales inter-
bedded. If this is the case, then the Dictyonema shales, Obolus
sandstone, and the Ceratopyge zones are absent.

The limestone of the section, the Orthocerenkalk, about thirty-
seven meters in thickness, is thus subdivided: —

Platyuruskalk.
Gigaskalk.
Asaphuskalk.
Limbatakalk.

Of the Limbatakalk a thickness of only 1.35 meters of dark red
limestone was seen, the guide fossil, Megalaspis limbata being present.

In the gray Expansuskalk the following typical fossils were obtained,
among others: —

Orthocerenkalk

Megalaspis heros. Lycophoria nucella.
Asaphus expansus. Orthis callactis.
Ampyx nasutus. Orthis calligramma.

The Gigaskalk is a rather thick-bedded red limestone with Mega-
lass gigas and grades into the red, coarse-grained Platyuruskalk
with the guide fossil, Asaphus platyurus, and cephalopods. Above
this is a gray limestone, occupying the position of the Chironkalk of
other sections, but without fossils.

The Graptolithenschiefer have been investigated fully by Hadding,
(60) who found, immediately overlying the unfossiliferous gray lime-
stone, about eight meters of black shale with layers and lenses of dark
limestone, the whole characterized by graptolites and trilobites. He
has designated this band as the zone of Climacograptus putillus, and
finds there, among other fossils: —

Didymograptus superstes Lapw. Nileus armadillo Dalman.
Diplograptus perexcavatus Lapw. Ogygiocaris dilatata Briinn.
Climacograptus putillus Hall. Trinucleus coscinorrhinus Ang.
C. scharenberg. Lapw. Telephus bicuspis Ang.

Triarthrus beckt humilis Hadding. Robergia micropthalma Linrs.

In slightly higher beds, with less limestone and more shale, he
obtained, with others: —

Dicellograptus sextans exilis E.& W. Ogygiocaris dilatata Briinn.
Nemagraptus gracilis remotus E. & W. Nileus armadillo Dalm.

210 BULLETIN: MUSEUM OF COMPARATIVE ZOOLOGY.

This he designates as the zone of Nemagraptus gracilis, and it in
turn is followed by similar shales containing Dicranograptus clingant.
Of the fauna associated with this latter graptolite Hadding gives no
list, but the shales containing it are presumably those which Wiman
refers to the Chasmopslager, with Chasmops sp., Asaphus lundibundus
Taqt., Illaenus fallax Holm, I. gigas Holm, and Caryocystis granatum
Wbg. According to Moberg the Chasmops limestone is in places
found resting on the Orthoceras limestone.

Above the Chasmops zone, but never seen in contact with it, is the
Brachiopod shale, from which only a few determinable fossils have
been obtained, among them Encrinurus multisegmentatus Portl.,
Atrypa crassicosta Dalm., Leptaena rhomboidalis Wilckens, and Plasmo-
pora conferta Milne Edwards & Haime.

GASTRICKLAND.

Nearly all the fossils from this region have been obtained from
boulders found in the drift (Wiman, 89), but the record is of consider-
able interest, from its rather close similarity to the Esthonian develop-
ment. It appears to be the only part of Sweden where the Lower
Cambrian was developed in a region where the Middle Cambrian was
absent. The Chironkalk seems also to have been developed, in part,
as a Linsenschicht (corresponding to the Upper Linsenschicht of the.
East Baltic) and the Chasmopskalk appears to have a development
comparable to a part of C; of the Russian section.

Unfortunately only the strata from the Ceratopyge limestone to the
Limbata limestone are found in place.

The Lower Cambrian is indicated by boulders with fragments of
Olenellus, Agraulos, Ellipsocephalus, and Mickwitzia.

The Obolus sandstone is indicated by boulders.

Fragments of shale with Ceratopyge forficula and two species of
Shumardia have been found.

The Ceratopyge limestone is in place on the Island Limon and is
.83 meters thick. It contains the usual fossils.

Above this hmestone is a clay with glauconite and noduiel of lime-
stone. It is 1.17 meters thick and contains few fossils, Lingula ? sp.,
Acrotreta sp., and Torellella sp. being the only ones reported. This
corresponds to the “Glauconite sand”’ of Esthonia. By Wiman it is
united with the Ceratopyge limestone rather than with the Plani-
limbata limestone above.

RAYMOND: CORRELATION OF THE ORDOVICIAN STRATA. 211

The Planilimbata limestone is described as being brownish red in
color, with green, violet, and yellow spots and streaks, thus reminding
one strongly of the same stratum on the Walchow. It is 3.5 meters
in thickness. A rather large fauna is reported, including, Pliomera
actinura, Megalaspis planilimbata, Niobe laeviceps, Harpina excavata,
and Orthis christianiae, reminding one of the fauna which Lamansky
found at the top of the “Glauconite sand” at Papowka.

The Limbatakalk is a lighter colored rock than the limestone below,
and may be gray. It has a thickness of 5.45 meters. Megalaspis
limbata and other fossils are present.

The Expansuskalk is known from boulders which contain many
fossils, including the typical Asaphus expansus, A. raniceps, Megalasyis
acuticauda, M. heros, Lycophoria nucella, ete.

The Gigaskalk is represented by a single boulder.

The Platyuruskalk is usually found as boulders of red limestone,
and along with Asaphus platyurus contains many cephalopods, such
as Orthoceras conicum, Vaginoceras wahlenbergi, Lituites lituus, ete.

The Chiron kalk is found in boulders, sometimes containing “lin-
sen.” The fauna contains Asaphus kowalewski, A. cornutus, Illaenus
chiron, I. schmidti, and Christiania oblonga, and distinctly suggests
the C, of Russia.

The older Chasmopskalk is lithologically like the Chironkalk, and
in the boulders are found Porambonites schmidti, Platystrophia lynz,
Christiania oblonga, Echinosphaerites, Ptilograptus suecicus, Climaco-
graptus, and Diplograptus.

The boulders assigned to the younger Chasmopskalk or Raerousns.
kalk contain among others, Chasmops maxima, Illaenus fallax, I.
oblongatus, Porambonites ventricosus, and Platystrophia lynz.

Boulders of the so-called “Ostseekalk,’’ also occur which, in the
North Baltic area, is partly fine-grained “lithographic stone’? com-
parable to the Wesenberg limestone of Eethania, while other boulders
are of a different sort.

Wiman lists the fossils found in a large number of these boulders,
which would seem to have been derived from formations very similar
to the Kegel, Wesenberg, Lyckholm, and Borkholm of Esthonia.

Interesting species are Chasmops wesenbergensis, Encrinurus see-
bachi, Lichas eichwaldi, and Clitambonites wesenbergensis, all of which
occur in the Wesenberg at Wesenberg.

Cyclocrinites schmidti and C. balticus of course suggest the Kegel,
while Platystrophia lynx and Oxoplecia dorsata occurring together,
remind one of the Kuckers.

‘

“ake BULLETIN: MUSEUM OF COMPARATIVE ZOOLOGY.

Encrinurus multisegmentatus, Lichas laxatus, Tetradium wrangeli,
Atrypa imbricata, and Halysites parallelus are all typical species of the
Lyckholm.

DALECARLIA.

Miss Elsa Warburg (84), gives the following section as typical of
the region.

Leptaena limestone

| Red Trinucleus shale 15 m.
; Gray limestone 5-9 m.
Trinucleus shales Black Trinucleus shale 6 m.
| Masur limestone 9-15 m.
Ch ee ee Macrourus limestone 9m.
ce aiden Cystidean ls. 15 m.+
Ancistroceras |s. Upper gray Orthoceras
Chiron ls. limestone.
Orthoceras limestone aril a Upper red_ limestone.
a a ee Asaphus ls. Lower gray limestone.
Limbata ls. Lower red limestone.
Planilimbata ls. 3.08 m.
Ceratopyge ls. .14-.16 m.
Ceratopyge limestone } Glauconite sand 10 m. |
Obolus conglomerate .15-.80 m.

In Dalecarlia there is no Cambrian, the Obolus conglomerate resting
on the granite. It contains Obolus apollinis and is followed by a thin
bed of greenish gray glauconitic clay-shale which contains some
fragments of Obolus. Above comes a thin bed of glauconitic limestone
which contains Obolus fragments and Lycophoria laevis Stolley.
This Wiman correlates with the Ceratopyge limestone on the basis
of the latter fossil.

At a single locality (Skattungbyn) Térnquist found in shales with
interbedded slabs of limestone, Tetragraptus serra, T. quadribrachiatus,
Dichograptus octobrachiatus, Phyllograptus densus, and three species
of Didymograptus. The limestone contained Pliomera térnquisti
Holm, Megalasmdes dalecarlica, Ampyx pater, and Agnostus térnquisti
Holm. In general however, the Planilimbatakalk is present, followed
by the Limbata limestone. Megalaspis planilimbata, M. limbata,

:
,
t

RAYMOND: CORRELATION OF THE ORDOVICIAN STRATA. 213

Niobe laeviceps, and Nileus armadillo are found in these strata. The
Asaphus limestone contains the typical species, Asaphus expansus,
Lycophoria nucella, etc., and it may be noted that the Lower Linsen-
schicht is developed in this district. The Gigas and Platyurus lime-
stones have many cephalopods, but few trilobites. Both the Chiron
and Ancistroceras limestones contain their typical fossils and are
followed by fifteen meters of limestone containing Chasmops odini,
Echinosphaerites aurantium, and Oxoplecia dorsata.

The Macrourus limestone contains Chasmops maxima Schmidt.

The black Trinucleus shale contains Trinucleus seticornis, Calymene
trinucleina Linrs., Remopleurides radians Barr., Dalmanella argentea
His., and the graptolites Dicellograptus anceps, Diplograptus pristis,
D. truncatus, and Lasiograptus margaritatus. The gray limestone is
reported as containing numerous fossils, which however, occur only
as fragments, and I have seen no list.

The red Trinucleus shale contains few fossils. Remopleurides dor-
sospinifer, Proetus brevifrons, Agnostus trinodus, Trinucleus, and Pseu-
dosphaerexochus laticeps have been reported. The Leptaena limestone
contains a very large fauna, comparable to that of the Lyckholm
and Borkholm. There has been a great deal of discussion about
the relative positions of the Leptaena limestone and Trinucleus shale
in Dalecarlia. Although their faunas are quite different, yet both are
characterized by an influx of Bohemian species, and both show the
beginnings of a fauna like the Silurian. Furthermore, the presence
of Dicellograptus anceps and D. complanatus are indicative of the
youngest Ordovician age of the Trinucleus shales. The Leptaena
limestone seems to show the physical characteristics of a “reef,”
though not perhaps of a coral reef, as Nathorst has suggested.

VASTERGOTLAND.

The strata of the third belt are best exposed in Vastergétland where
the Cambrian, Ordovician, and Silurian rocks are practically hori-
zontal and well shown on the sides of small “mountains” in which
they have been protected from erosion by a capping sheet of diabase.
On account of its quarries, Kinnekulle presents unusual opportunities
for studying the Orthoceras limestone and the section there is one of
the classic ones in Swedish geology. Throughout this region the
Ordovician rests upon Upper Cambrian formations, and the Dictyo-
nema shales are usually either very feebly developed or entirely

214 BULLETIN: MUSEUM OF COMPARATIVE ZOOLOGY.

absent, and the Obolus sandstone has not been reported at all. At
most localities the Ceratopyge limestone has a much better develop-
ment than in the more northern belts, though in a few places it is
entirely absent. In most sections, the Planilimbata limestone is
absent and the graptolite-bearing shales replace it. The Asaphus
limestone of this region is similar to that of Oeland and unlike that of
other regions in that Asaphus expansus itself is absent from the fauna
and the Asaphuskalk is divided into two members by a stratum which
is almost entirely made up of the cystid Sphaeronis pomum Gyllen-
hahl.

Kinnekulle.
Brachiopodenschiefer 5 meters.
Trinucleusschiefer 32 meters.
Or She pte Chasmopskalk 10 meters.
Orthocerenkalk 52 meters.
Underer Didymograptusschiefer 9 meters.
Ceratopygekalk 1 meter.

The Ceratopyge limestone is a light gray limestone with light green
glauconite and considerable pyrite in certain layers. Beside the
typical fossils, which are somewhat abundant, Lycophoria laevis Stolley
has been reported from this locality by Wiman (90).

The Orthoceras limestone here has been subdivided into four
divisions, on the basis of color.

The lower twenty meters are of a deep red color and are known as
the “Lower Red.” Above is found a band three meters in thickness
of light gray limestone, the “Lower Gray.” At the top of this there
is a sudden change again to deep red limestone and shale, not well
exposed except for the two meters at the base, but perhaps thirteen
meters thick. This is the “Upper Red,” and it is followed by the
“Upper Gray,” sixteen meters of which could be measured. Fossils
are common in the lower part of the Lower Red” and in the “ Lower

Gray” but rather rare elsewhere, and it does not seem to have been ©

possible so far to make exact subdivisions on the basis of fossils. The
color divisions do not, however, seem to correspond to the subdivisions
which would be made on the basis of the fauna. The “Lower Gray”’
includes the “Sphaeronis bank” and with the upper part of the “ Lower
Red” and the lower part of the “ Upper Red,” represents the Asaphus-
kalk. On the basis of fossils it would appear that all the usual zones
of the Orthoceras limestone, except the Planilimbatakalk and possibly
the Gigaskalk, are present.

RAYMOND: CORRELATION OF THE ORDOVICIAN STRATA. 215

Limbatakalk — The greater part of the “Lower Red” probably
belongs to this zone. Megalaspis limbata, Nileus armadillo, and
Symphysurus palpebrosus are common here, and a number of other
species were collected.

Asaphuskalk — I did not find fossils other than Sphaeronis pomum
and Megalaspis heros very common in this zone. Several other species
have been reported, among them Phacops sclerops, Cyrtometopus
clavifrons, Asaphus raniceps maximus, etc. The numerous cephalo-
pods assigned to this zone in lists seemed to be derived from the lower
part of the “Upper Red” and possibly to indicate the Gigas rather
than the Asaphuskalk. Among these are Vaginoceras wahlenbergi
(Foord), Bathmoceras linnarssoni (Ang.), and Estonioceras proteus
Holm. Megalaspis gigas itself has not been found, but the cephalopods
mentioned indicate, I believe, its zone.

Platyuruskalk — Asaphus platyurus and Orthoceras tortum are
reported from the upper part of the “Upper Red.”

Chironkalk — The “Upper Gray”’ evidently represents the zones
of both Illaenus chiron and Ancistroceras. JIllaenus chiron Holm,
Ogygrocaris dilatata sarst Angelin, Ancistroceras undulatum Boll, and
Discoceras teres Kichwald have been reported.

The next formation, the Chasmopskalk, is not exposed on = side
of Kinnekulle which I visited, but it is reported as being a dark green
graptolite-bearing shale with lenses and layers of impure limestone.
The graptolites are not listed, but the limestone is said to contain
Chasmops sp., Remopleurides sealineatus, Ptychopyge? glabrata, Am-
pyx rostratus, Echinosphaerites aurantium, ete. The thickness is ten
meters.

The Trinucleus shales, which are not well exposed, are said to be
thirty-two meters thick and consist of two shales separated by a thin
limestone. The lower twelve meters consist of black and greenish
shales separated by two meters of limestone from eighteen meters of
the red shale above. Among the forms listed from the upper shales
are Remopleurides radians, co verrucosa, Trinucleus wahlenbergi,
and Dionide euglypha.

The Brachiopodenschiefer are said to be represented by twenty-six
meters of calcareous shale below, followed by two and four tenths
meters of impure sandy limestone. Fossils do not appear to be com-
mon, Dalmanites mucronatus, D. pulchellus, and Homalonotus platy-
notus being the chief ones reported.

I visited Alleberg, but found it impossible to get good fossils without
spending more time than was at my disposal.

216 BULLETIN: MUSEUM OF COMPARATIVE ZOOLOGY.

Here the upper Trinucleus shales, also called the Staurocephalus
shales, are reported to be very fossiliferous and add Stauwrocephalus
clavifrons, Acidaspis centrina, and Maneinge: parabola to the list of
species found at Kinnekulle.

OELAND.

The Ordovician on Oeland rests on Upper or Middle Cambrian
strata and the basal member may be either the Dictyonema shale or
the Ceratopyge shale. This has been especially well brought out in
an instructive diagram by Fearnsides (59). The most complete
section of the “Ceratopyge Region”’ and Cambrian is shown in south-
ern Oeland, where, ignoring subzones, the following strata are found,
in descending order.

( Glauconite shale.
Ceratopyge limestone.
Glauconite shale.
Shumardia shale.
Dictyonema shale.

(

Ordovician

Peltura limestone.

Upper Cambrian Olenus shale.

Paradoxides forchammeri zone.
P. tessini zone.
| P. oelandicus zone.

Middle Cambrian

——

It has been shown that in passing northward both the Dictyonema
and Shumardia shales pinch out, but at the northern end of the section
both come in again, the Dictyonema shale being in certain localities
replaced by the Obolus conglomerate. At the southern end of the
island the Ordovician rests upon the youngest known beds (zone of
Peltura scaraboides) of the Upper Cambrian. About midway between
the northern and southern ends the Peltura beds disappear and the
Ordovician rests for a short distance on the lower part of the Upper
Cambrian. From Borgholm for several miles north the Ordovician
rests on the next lower zone of the Cambrian, the Paradoxides forcham-
mert zone of the Middle Cambrian. Still further north this gives
place to the next lower zone, that of P. tessint. The strata of the
Paradowxides tessint zone are here sandy, and it is where the Ordovi-
clan rests on them that the Obolus conglomerate is developed, thus
indicating the local origin of the material in the Lower Ordovician.

:

ec nt ee a eee a ye eG]

hi . hae
i a ial i

eho Ee

e - de Teepe

ones:

e

4)
s
+

RAYMOND: CORRELATION OF THE ORDOVICIAN STRATA. 217

Finally at the northern end of the island, the Ordovician rests on strata
of the lowest of the Middle Cambrian zones, that with Paradoxides
oelandicus. This section shows very clearly that there was an uplift,
tilting, and erosion after the deposition of the Peltura beds of the
Upper Cambrian and before the deposition of the Dictyonema shales
of the Ordovician, and thus emphasizing once more that the natural
place to draw the boundary between the Ordovician and Cambrian is
at the base of the Dictyonema shales (or the equivalent Obolus sand-
stone).

In spite of the considerable amount of work which has been done on
Oeland, I am unable to find that any section has been published in
which the thicknesses of all the strata have been given.

The youngest strata found in place in Oeland belong to the Lower
Chasmops or Echinosphaerites limestone, but the Macrourus lime-
stone, Trinucleus shale, and Leptaena limestone (Lyckholm) are all
represented by numerous boulders.

The Echinosphaerites limestone is seen only in northern Oeland,
and only Echinosphaerites aurantiwm and Illaenus chiron seem to be
reported from it.

The Orthoceras limestone is very well developed on the island and
it was here that its subdivision on the basis of fossils was first accom-
plished by Moberg (71). The zones, in descending order are: —

Ancistroceras limestone. The fauna is reported to contain An-
cistroceras undulatum Boll, Remopleurides, Ptychopyge, Nileus arma-
dillo, Illaenus chiron, and Orthoceras.

Chiron limestone. This is a limestone containing Illaenus chiron
Holm, Ptychopyge aciculata Ang., P. testicaudata Steinh., Megalaspis
pagiata Tqst., Ogygiocaris dilatata sarsi Ang., Telephus bicuspis Ang.,
Iituwites lituus. Monf., and Didymograptus geminus His.

Platyurus limestone. This zone has only a few fossils reported,
these being Asaphus platyurus maximus, Ptychopyge brachyrachis
Remele, Rhynchorthoceras cf. angelini Boll:, Echinosphaerites auran-
tuum Gyllenhahl, and Hyolithes inaequistriatus Remele.

From the Gigas limestone, only Megalaspis gigas Ang. has been
reported.

Upper Asaphus limestone. Moberg states that this is a reddish,
rarely white, crystalline limestone with a large fauna of small, mostly
undescribed trilobites, and that there is nothing elsewhere which
exactly corresponds to this zone. He cites Nieszkowskia tumidus,
Asaphus sp., Illaenus esmarki, I. centrotus, Nileus armadillo, and
Niobe frontalis as among the species present.

218 BULLETIN: MUSEUM OF COMPARATIVE ZOOLOGY.

Below this is the stratum filled with Sphaeronis pomum Gyllenh.
as at Kinnekulle.

Lower Asaphus limestone. This j is a gray limestone from which
Pterygometopus sclerops, Megalaspis heros, Ptychopyge applanata,
Niobe frontalis, Illaenus esmarki, Ampyx nasutus, Orthis obtusa, and
Gly ptocystites cf. leuchtenbergi have been reported. Asaphus expansus
is not found in Oeland.

Holm (65), has described, from a glauconitic gray limestone at
Halludden near Béda in northern Oeland, Jsograptus gibberulus
(Nicholson), Didymograptus minutus Tqst., Tetragraptus bigsbyi Hall,
and Phyllograptus angustifolius Hall. The limestone containing these
fossils is said to belong to the Lower Asaphus zone, but may possibly
be in the Limbata zone.

Limbata limestone. From this zone Moberg reports Megalaspis
limbata, Niobe laeviceps, two pelecypods, two gastropods, and “ Rhyn-
chonella’’ digitata Leuchtenberg.

Planilimbata limestone. From this limestone, which is often quite
glauconitic, Megalaspis planilimbata Ang. and Holometopus limbatus
Ang. have been obtained.

The Ceratopyge (76) zone is well developed in Oeland, but, as has
previously been mentioned, it is variable in its constitution. In the
southern half of the island a limestone is present in the upper part,
included between two glauconitic shales, and beneath the lower shale
is another shale characterized by Shumardia. In the northern half
of the island the Shumardia shale is absent, also the limestone, and
there remains only a shale bearing Ceratopyge. The total thickness
seems to be small, with a maximum of about two meters.

The Dictyonema shale is, as stated above, present in both the
northern and southern portions of the island, but absent for a consider-
able space through the middle. The thickness in the southern part
of the island is about two meters.

ScANIA.

In Seania the strata of the Ordovician consist very largely of
graptolite-bearing shales, these shales resting upon the Olenus shales
of the Upper Cambrian. At various horizons, however, beds of
limestone are intercalated in the shales. No one region presents a
complete section, and the following composite section contains some
beds which are found only in East Seania, some found only in West

at ne

A
{
a
p

:

RAYMOND: CORRELATION OF THE ORDOVICIAN STRATA. 219

Scania, while others are developed in both regions. The table is an
amplification of one given by Moberg (75).

Zone of Phacops eucentra and Stauro-
cephalus clavifrons.

Trinucleus or |
Zone of Amypax portlocki.

Upper Dicellograptus shales

Zone of Pleurograptus linearis.
Zone of Calymene dilatata and Dicra-
nograptus clingant.

Chasmops or
Middle Dicellograptus
shales (with limestone)
Zone of Nemagraptus gracilis.
Zone of Climacograptus putillus.
Zone of Glossograptus hinckst.

Lower Dicellograptus shales

Upper Didymograptus shales Zone of Phyllograptus cf. typus.

Orthoceras limestone
Zone of Isograptus gibberulus.

Zone of Phyllograptus angustifolvus.
Zone of Didymograptus balticus.

Zone of T etragraptus phyllograptordes.

Zone of Didymograptus geminus.
Lower Didymograptus shales |

Ceratopyge limestone
Shumardia shale
Dictyonema shale

The fauna of the Trinucleus beds has been described by Olin (79),
who enumerates many species. Of trilobites forty-three species were
listed, and it is of particular interest to note that twelve of these are
species common to Sweden and Bohemia. Prominent among such
species are Cheirurus pectinifer Barrande, Remopleurides radians Barr.,
Calymene incerta Barr., C. pulchra Beyr., Phillipsia parabola Barr.,
“ Asaphus”’ ingens Barr., Trinucleus bucklandi Barr., Ampyx gratus

_ Barr., A. portlockt Barr., A. tenellus Barr.; and Aeglina rediviva Barr.

In addition to these common species, the genera Aeria, Staurocephalus,
and Dionide in themselves suggest the Bohemian fauna. Other
important species in the fauna are Phacops (Dalmanites) eucentra Ang.,
Agnostus trinodus Salter, and Stygina latifrons Portlock. The im-
portant graptolites are Dicellograptus complanatus Lapw., and Diplo-
graptus truncatus Isapw.

The Chasmops beds contain a rather small fauna,-but although
there are many limestone bands in the shales, there are fewer trilo-
bites and more graptolites. Olin lists among others Calymene dilatata

220 BULLETIN: MUSEUM OF COMPARATIVE ZOOLOGY.

Tullb., Remopleurides sexlineatus Ang., Ampyzx rostratus Sars, Dal-
manella argentea (His.), Corynoides calicularis Nich., Dicranograptus
clingant Carr., Diplograptus quadrimucronatus Hall, Climacograptus
scharenbergi Lapw., and Ptychopyge ? glabrata Ang. The zone of
Dicranograptus clingani is at the base of the Chasmops beds, and the
zone of Pleurograptus linearis at the top. |

The fauna of the Lower Dicellograptus shales has been monographed
by Hadding (60), who lists the following as characteristic fossils of
the thin-bedded gray-black shales which make up the strata of the
three subzones in Scania.

Subzone of Nemagraptus gracilis— Nemagraptus gracilis remotus
Elles & Wood, Lasiograptus mucronatus Hall, Obolus elatus Hadding,
Plectambonites sericeus restrictus Hadding.

Climacograptus putillus zone.— Climacograptus putillus Hall, C.
caudatus Lapw., Dicellograptus vagus Hadding, and Dicranograptus
erregularis Hadding.

Glossograptus hincksi zone.— Glossograptus hincksi Hopk., Crypto-
graptus lanceolatus Hadding, and Diplograptus perexcavatus Lapw.

The Upper Didymograptus beds consist of green and gray shales and
the fauna seems to be incompletely known.

In the Didymograptus geminus beds.— D. geminus Hisinger,
Lonchograptus ovatus Tullb., Climacograptus confertus Lapw., Ptero-
graptus scanicus Moberg, P. elegans Holm, and species of Diplo-
graptus and Cryptograptus have been reported.

From the lower zone.— Phyllograptus cf. typus, Didymograptus
cf. bifidus, Climacograptus and Cryptograptus are reported. It seems
to be a sort of transition zone, in which the Diprionian graptolites
which are so abundant in the beds above make their first appearance,
but accompanied by some of the survivors of the more ancient
fauna.

The Orthoceras limestone is a hard, rather pure, dark blue limestone
and is at present inadequately exposed. Angelin (58) described
twenty species of trilobites from the quarries at Fagelsang, some of the
more important of which are *Ampya nasutus Dalm, Asaphus acumi-
natus (Boeck), *Cyrtometopus clavifrons, *Megalaspis limbata (Sars
and Boeck), *Nileus armadillo Dalm, Symphysurus palpebrosus
(Dalm), *Niobe frontalis Ang., *Pterygometopus sclerops (Dalm),
Trinucleus coscinorrhinus Ang. The species marked with an asterisk
occur in the Limbata or Expansus limestones of Sweden and in the
zones By,g and Biya of Russia.

According to Strandmark (81), beds of shale containing graptolites

<

:
‘ ig
:
2
&
i

a es

a a

RAYMOND: CORRELATION OF THE ORDOVICIAN STRATA. 221

occur between the limestone strata of the Orthoceras limestone, the
species found being Tetragraptus bigsbyi Hall, Didymograptus extensus
Hall, and Phyllograptus cor Strandmark.

The fossils of the Lower Didymograptus shales have been described
by Térnquist (83). These shales are much better developed in south-
eastern Scania than about Lund.

The zone of Isograptus gibberulus contains Isograptus gibberulus
Nich., Didymograptus extensus Hall, D. patulus Hall, and D. mobergi
Taqst. .

The zone of Phyllograptus angustifolius has Didymograptus prae-
nuntius Tqst., Phyllograptus angustifolius Hall, and Tetragraptus
pendens praesagus Tqst.

The zone with Didymograptus balticus has a larger fauna, among the
species being Didymograptus constrictus Hall, D. balticus Tullb.,
Tetragraptus quadribrachiatus Hall, Dichograptus octobrachiatus Hall,
and Clonograptus subtilis Tqst.

The lowest zone contains Tetragraptus phyllograptoides Linrs., T.
approximatus Nich., T. serra (Brong.), Didymograptus undulatus Tqst.,
and D. holmi Tqst.

The Ceratopyge zone is poorly developed in Scania, but both the
Ceratopyge limestone and Shumardia shale have been shown to be
present, with the usual fossils.

The Dictyonema shales are subdivided into three zones.

C, zone with Bryograptus kjerulfi and Dictyonema norvegica.

B, zone with Clonograptus tenellus in four varieties.

A, zone of Dictyonema flabelliforme in which, interstratified with
shales, are beds of limestone with Hysterolenus térnquisti and H.
lericaudus.

Very little seems to be known of the thicknesses of strata in Scania.
On seeing the outcrops in the field, one receives the impression that

*they are very small. Ina boring made at Stabbarp, northeast of Lund,
the Chasmops beds were encountered between eighty-six and ninety-
two meters below the surface and shales of the Phyllograptus ct. typus
zone at a depth of 102 meters, thus indicating a thickness of about ten
meters for the Lower Dicellograptus shales.

At Jerrestad in eastern Scania Olin measured a section 425 em. in
thickness, all in the Trinucleus zone, and most sections which it is
possible to measure are of this order or of less thickness.

222 BULLETIN: MUSEUM OF COMPARATIVE ZOOLOGY.

SUMMARY.

From the above summary of the sections in the principal Ordovician
belts of Sweden, certain facts should appear.

It will be noted that the Dictyonema zone is present in almost all
sections, and that the character of the deposit is controlled by the
underlying beds, a fact which in itself is evidence of a cessation of
sedimentation, uplift, and erosion. In North Oeland, where the
basement beds are the Paradowides tessini sandstone, an Obolus sand-
stone is developed, and in the northern belt, Dalecarlia-Gastrikland,
where the older strata are pre-Cambrian crystallines and Lower
Cambrian sandstones, one finds Obolus conglomerate and Obolus
sandstone. That the Obolus sandstone is of the same age as the
Dictyonema shale is of course abundantly proved by their interstratifi-
cation with each other at several places in Esthonia. In southern
Oeland, in Scania, and at Christiania, where the underlying strata
are the black shales and limestones of the Upper Cambrian, the Dictyo-
nema zone is developed as a shale without sand and has even, in places,
irregular layers of limestone.

Over a certain area, in Nerike and part of Vastergétland, the
Dictyonema zone is absent. Thus, at Kinnekulle the Ceratopyge
limestone rests upon the limestone and shale of the Upper Cambrian,
at Ekedalen the Planilimbata limestone rests directly on the Upper
Cambrian, and at Hunneberg the Ceratopyge limestone rests on the
Upper Cambrian limestone in most places, though in some spots about
three inches of shale belonging to the Dictyonema zone have been
reported. The arrangement of these localities free from Dictyonema
deposits suggest a large, low lying island composed of uplifted Upper
Cambrian strata, which was progressively submerged during Lower
Ordovician time, but not completely covered till toward the end of
Planilimbata time. Correlated with this may be the distribution of
the Lower Didymograptus shale.

As has already been pointed out repeatedly by the Swedish geolo-
gists, when the Lower Didymograptus shales are present in any sec-
tion, the Planilimbata limestone is absent, and when the latter is
present, the former is absent. This has naturally led to the deduction
that the two formations were deposited at the same time and owe their
dissimilar faunas to the very different conditions of sedimentation.
On this point, the evidence, though strong, does not seem to be abso-
lutely conclusive. It would seem that, should it happen that lime-

RAYMOND: CORRELATION OF THE ORDOVICIAN STRATA. 223

stone were found interstratified with the graptolite shale, it should
contain some at least of the species found in the strata where limestone
only was deposited. In Dalecarlia, Holm (62), found a trilobite
fauna in limestone interstratified with the Lower Didymograptus
shales. Six species of trilobites were identified, and five of the species,
Pliomera tirnquisti, Megalaspides dalecarlica, Ampyx pater, Agnostus
térnquisti, and Trilobites brevifrons were new and are all restricted
to this one locality. The sixth species, Niobe laeviceps Dalman
is not a guide fossil, ranging from the Ceratopyge limestone to the
Asaphus limestone. The pygidium of Pliomera térnquisti is not known
but the cephalon and thorax show that it is not a true Pliomera
but a Cyrtometopus allied to the forms found in the Ceratopyge
limestone. Similar species are, however, found in higher strata.
Megalaspides is not yet definitely placed outside this occurrence in
Dalecarlia. Wiman (88), has described Megalaspides nericiensis
from the Shumardia shale in Nerike, but there is some doubt as to
whether this Shumardia shale belongs to the Ceratopyge zone or to a
horizon in the Planilimbata limestone.

Wiman (89), also reports pygidia of Megalaspides from boulders of
Planilimbata limestone in Gistriklind, and Lamansky (29) described
Megalaspides schmidti from a pygidium found in B, at Papowka. Iden-
tifications based on pygidia alone seem rather unsafe in this genus, the
pygidium being so Asaphus-like. The genus has not hitherto been
suspected in the Ceratopyge limestone, but at Hunneburg I found a
large hypostoma of the “forked” type in the same strata with Euloma
and Symphysurus, and as no other member of the Asaphinae is known
at so low a horizon, it of course suggests Megalaspides. Agnostus
tornquisti and Trilobites brevifrons are of no value in the discussion.
Ampyzx pater is similar to Ampyx nasutus of the Limbata and Asaphus
limestones rather than to the species so far described from the Cera-
topyge limestone.

The fauna found by Holm in Dalecarlia is then not very useful in
the correlation of the Lower Didymograptus beds with the Planilim-
bata limestone. It contains no species restricted to the Planilimbata
limestone, and the general composition of the fauna is such that,
lacking a guide fossil of either, it could be referred to the strata either
above or below it. Over most of Dalecarlia the Planilimbata lime-
stone is present, but very poor in fossils. In Jemtland, as indicated
on page 208 there does seem to be some mingling of the species of the
Lower Didymograptus shales with those of the Planilimbata limestone.

The Didymograptus shale is absent from Oeland, is best develeped

224 BULLETIN: MUSEUM OF COMPARATIVE ZOOLOGY.

in eastern Scania, present in western Scania, well developed in Viaster-
gétland at Hunneberg and Kinnekulle, and present though less
developed in other places. As previously noted, the shale is absent
at Ekedalen, and Fearnsides has pointed out that the shale thickens
in going away from Skéofde and Ekedalen. Munthe gives the follow-
ing thicknesses at places on the Skofde sheet of the Geologic Map:
Ulunds, .6 meters, Baickagarden .23 meters, Kapplunda .2 meters,
Persberg, absent, Skultorps Norra .9 meters. At Klefda in the west-
ern Falbygden region, not far from Falk6ping, the thickness is three
meters, at Hunneberg, eleven meters (capped with diabase), and at
Kinnekulle, ten meters. As to fauna, it is only in eastern Scania that
the four subfaunules are developed. Here one sees, in descending
order:— (4), zone of Isograptus gibberulus, (3), zone of Phyllograptus
angustifolius. (2), zone of Didymograptus balticus. (1), zone of
Tetragraptus phyllograptoides.

In connection with the thinning of the shales toward Ekedalen it is
important to note that at Hunneberg one finds the zones 1 and 2, at
Kinnekulle 1, 2, and 3, the Limbata limestone succeeding the shale,
but in the thin sections in the Sk6fde area Phyllograptus angustifolius
is reported as the most common graptolite, indicating the presence
there of zone 3. In other words, the shale has thinned by the loss of
the lower members, thus showing overlap in that direction and sus-
taining the idea that there was an island of Cambrian strata here in
early Ordovician times. The Ceratopyge formation also thins in this
direction, as shown by Munthe’s measurements in the Skéfde sheet:
Ulunda .3 meters, Backagarden .12, Kapplunda .4, Persberg .3,
Skultorps Norra .9 and Skultorps Sodra .15. It has been argued
(Wiman, 90) that the foot of conglomeratic and glauconitic limestone
resting on the Cambrian at Ekedalen represented the Ceratopyge
limestone, but this seems very improbable both on faunal and strati-
graphic grounds.

It has been suggested that the lower part of the Lower Didymo-
graptus beds replace the Ceratopyge limestone in certain regions, but
the very general occurrence of the Tetragraptus phyllograptoides
faunule in shales resting on the youngest of the Ceratopyge zones
negatives this idea. That the Limbata limestone does replace the
upper part of the shale is, however, readily shown by the occurrence of
the fauna of the Isograptus gibberulus subzone in or above the Limbata
limestone on Oeland.

In connection with the development of the Orthoceras limestone,
certain things should be noted. For instance, in those sections where

:

RAYMOND: CORRELATION OF THE ORDOVICIAN STRATA. 225

the Orthoceras limestone is thickest, as on Oeland or at Kinnekulle,
Asaphus expansus, the guide fossil of the “ Expansusschicht”’ is absent;
and there is found, above the strata containing a fauna which includes
a part of the species usually found associated with Asaphus expansus,
a layer with such quantities of the cystid Sphaeronis pomum as to make
a veritable cystid reef. No such reef is seen where Asaphus expansus
is present, as in Ostergétland or the Christiania district. Both
Sphaeronis and Asaphus expansus are reported from Dalecarlia, but,
according to Térnquist the former species is exceedingly rare, and not
quite typical, and that district will probably not prove to be an excep-
tion to the general rule.

Very little has been done toward working out the details of the
various sections in the Orthoceras limestone, so that the sections
at Kinnekulle and Oeland are really the only ones which can be com-
pared. On Oeland, one finds above the bed with Sphaeronis pomum
the Upper Asaphus limestone, with numerous undescribed trilobites.
Moberg states that this zone occurs nowhere else and it certainly is not
present at Kinnekulle, where a cephalopod fauna is found in the red
limestone above the Sphaeronis bed. This cephalopod fauna is that
normally found in the Gigas limestone, and it seems that there is in
the Oeland section a zone which is lacking at Kinnekulle. Whether
the absence of Asaphus expansus from the thick sections is explainable
by the predominance of red sediments, or whether the Expansus beds
are actually absent is not at present apparent. It should be observed
that the extra thickness in these great sections is largely accounted
for by the unusual development of the Planilimbata limestone, though
of course the added zones above the Gigas limestone have something
to do with it.

In Oeland, at Kinnekulle, and in Delarne, one finds considerable
limestone above the Gigas limestone, which by the Swedish geologists
is included in the Orthoceras limestone. This limestone contains
three faunal zones, according to the Swedish geologists, but the faunas
of the three seem very much alike. The lower zone contains Asaphus
platyurus and Echinosphaerites aurantium, fossils found in Ci, of
Russia, and I/laenus chiron is found in Cig of that country, but the
other guide fossils are mostly species not found in Russia. The pres-
ence of Didymograptus geminus in the middle zone of the three in
Oeland is noteworthy, for it serves to complete the parallelization of
the Scanian and Oeland sections.

The presence of Ogygiocaris, which seems to have an exceedingly
narrow vertical range, in the Nemagraptus zone in Jemtland and in the

226 BULLETIN: MUSEUM OF COMPARATIVE ZOOLOGY.

Chiron limestone of Kinnekulle and Oeland, serves to indicate that
the upper part of the Orthoceras limestone in these localities is the
equivalent of the Nemagraptus gracilis shale as well as the Geminus
shale. And the correlation of the Chasmops limestone with the grapto-
lite shales of the Dicranograptus clingani and Pleurograptus linearis
zones is assured by the presence of several species common to the two
types of deposits.

CORRELATION OF THE AMERICAN WITH THE
EUROPEAN FORMATIONS.

In attempting a correlation over such a distance one is of course
obliged to depend very largely on fossils; and it is necessary to assume
an hypothesis which all the evidence seems to support, namely, that
cosmopolitan faunas reach their wide distribution within a very short
- space of time. In the present case, it is necessary to depend very
largely upon the graptolites which seem to be more widespread than
any of the other organisms. Graptolites are, of course, almost absent
from Russia, so that it is necessary to correlate the Russian with the
Swedish sections by means of other fossils. The trilobites, being best
known, have been used most, but in certain cases species of brachio-
pods, cystids, or cephalopods have proved of prime value.

After making many groupings of the formations and testing many
tentative correlations, it has seemed that the most logical arrange-
ment is secured if the principal weight is given to the graptolites. It
appears that these organisms had spread very much more rapidly than
any of the other animals, except for a few thin-shelled brachiopods and
‘trilobites which may have been dispersed by the same agency as the
graptolites. When relatively short distances are in question, it seems
that the bottom dwelling animals were able to keep pace in their
migrations with the graptolites, at least sufficiently closely so that we
detect no difference in the geological record, but when long distances
are traversed, the bottom animals lag very considerably behind.
Striking cases are those of Shumardia, which preceded the first Tetra-
graptus fauna in Scandinavia, and reached America only with the last
Tetragraptus and the first Diplograptus, and of Echinosphaerites,
which preceded the Nemagraptus gracilis fauna in Scandinavia and
followed it in America.

RAYMOND: CORRELATION OF THE ORDOVICIAN STRATA. 227

LOWER ORDOVICIAN.

'The drawing of the lower boundary of the Ordovician at the base
of the shales with Dictyonema flabelliforme instead of at the top of that
zone is due to Moberg (73), and seems to have been suggested to him
in the first instance by the discovery low in the zone of a trilobite
(Hysterolenus térnquisti), which bore a marked resemblance to Cera-
topyge. The suggestion was a remarkably good one and is borne out
both by faunal and stratigraphic relations.

On the faunal side, one may cite the occurrence in this zone of the
oldest graptolites, thus marking the introduction of an entirely new
faunal element. Also, a few species are common to the shales of the
upper part of the zone and the Shumardia or Ceratopyge shales at
the base of the Ceratopyge zone.

The best argument is, however, based upon the evidence which the
geographical distribution of the deposits show of a great transgression
of the sea at this time, and the evidence of preceding erosion. This
evidence has already been detailed above, and it is necessary here only
to call to mind the conditions in the several regions.

In Esthonia the Dictyonema shale is interstratified with the Obolus
sandstone, leaving no doubt that the two are of the same age. The
Obolus sandstone, with conglomerate at the base in places, rests on
Lower Cambrian sandstone.

In Oeland, in passing from north to south, the base of the Ordovi-
cian (Obolus sandstone, Ceratopyge shale, and Dictyonema shale)
rests on successively higher and higher strata, varying in age from
lower Middle Cambrian to upper Upper Cambrian, the Obolus sand-
stone being developed over the Middle Cambrian Tessini sandstone.

In Dalecarlia, the Obolus sandstone overlaps onto the crystalline
rocks, but eastward in Gastrikland boulders of both Lower Cambrian
sandstone and of Obolus sandstone indicate that conditions there were
formerly as in Esthonia.

These facts show very clearly that at the end of the Cambrian there
was uplift accompanied by some, though probably not great tilting,
and a considerable amount of erosion, before the deposition of the
strata of the Dictyonema zone.

In America, the Dictyonema flabelliforme fauna is known from a
number of localities in the northeastern part of the United States and
Canada, but nowhere are these strata found in such position that their
relation to other strata can be definitely determined.

228 BULLETIN: MUSEUM OF COMPARATIVE ZOOLOGY.

Ruedemann (115) has summarized all that is known about its
occurrence in this country, and he, Matthew (104), and Hahn (103),
have pointed out the similarity of development of subzones at St.
John, New Brunswick, and Scandinavia. Present evidence indicates
that the Dictyonema flabelliforme fauna is the oldest of the Ordovi-
cian faunas in both Europe and America, though there is a possibility
that the strata with the Acerocare fauna in Scania and the Niobe
fauna in Wales may have to be added to the Ordovician series.

Brogger (99), attempted a correlation of the Ceratopyge zone with
certain strata in America but though his paper was extremely sug-
gestive, no very definite correlations were at the time possible. And
even now, we know no typically developed Ceratopyge fauna in
America. Walcott (120) has described a trilobite from the lower part
of the Goodsir formation of British Columbia as Ceratopyge canadensis,
and though it seems very doubtful if this is a Ceratopyge, it probably
belongs to the Tremadoc fauna. The writer has described from the
same region Hemigyraspis mcconnelli (110), a form indicative of the
Asaphellus beds of the Tremadoc of Wales. I have also described,
from the Tribes Hill of New York and the equivalent Stonehenge of
Pennsylvania, species of Hemigyraspis and Symphysurus also indica-
tive of the Ceratopygefauna. ‘This latter correlation is of considerable
importance, for the Stonehenge is the lower member of the Beekman-
town in western Central Pennsylvania. The Hemigyraspis fauna
occurs in the upper twenty-seven feet of the Stonehenge, which has a
thickness of 662 feet, and above the Stonehenge are 2570 feet of strata
before the top of the Beekmantown is reached. In the lower Stone-
henge, a Dictyonema has been found (Hahn, 103).

In Russia the Ceratopyge zone is probably not represented. There
is of course a temptation to make the Glauconite sandstone the equiva-
lent of the whole Ceratopyge zone, especially as one sees in Sweden
glauconitic sands associated with the Ceratopyge limestone. In
Dalecarlia there is a glauconite sand beneath the Ceratopyge limestone,
but in Gastrikland, where the section is very like the Russian section,
there is a glauconitic sand above the Ceratopyge limestone, and it
seems very probable that this sand is an extension of that so well
developed in Esthonia. It seems to belong with the Limbata lime-
stone above. This sandstone is thickest in Esthonia, and probably
indicates an emergence in that district so that there was some erosion,
perhaps accounting for the absence of the Dictyonema shale at Narwa,
and shore or shallow water conditions in Esthonia during the deposi-
tion of the Lower Didymograptus shales in Scandinavia.

|
|

RAYMOND: CORRELATION OF THE ORDOVICIAN STRATA. 229

The absence of the Lower Didymograptus fauna in Esthonia seems
explainable on the basis of lack of suitable physical conditions. It is
well known that the abundance of graptolites increases in proportion
to the degree of fineness of grain and amount of carbonaceous material
in shale. The Middle and Upper Cambrian strata of Sweden are
vast storehouses of very fine grained, highly carbonaceous shale.
Possibly these deposits extended at one time nearly or quite to
Esthonia. As has been shown, the end of the Cambrian was a
time of considerable denudation, and the Cambrian sediments could
furnish a vast supply of black mud, which, on account of its fineness,
could be transported long distances. Hence the widespread deposit
of Dictyonema shales. The shales however, rapidly covered the
sinking land, and were in turn covered, over large areas, by the Cera-
topyge limestone, so that, when the Lower Didymograptus fauna
occupied this region, only limited areas of Cambrian strata, such as
the island already mentioned in Vastergétland at Ekedalen and Skéfde,
were subject to erosion. There may have been a small rather general
uplift at this time, indicated in Sweden by the change from limestone
to shale sedimentation, and in Esthonia by the glauconite sand. To
consider the Dictyonema and Lower Didymograptus black shales as
reworked Cambrian shales seems more plausible than to think of them
as due to certain peculiar conditions under which black shales seem
usually to be formed. In any event, Esthonia was at this time out-
side the territory which could be supplied with reworked upper
Cambrian muds, while sands were immediately available and the
graptolite fauna did not reach the region.

To correlate the Russian “Orthocera’s limestone” (B,,, B,,,;) with
any formation in America on the basis of graptolites is rather com-
plicated but it can, I think, be done fairly satisfactorily. In the first
place there is general agreement, on the evidence of numerous species
of trilobites, cephalopods, and brachiopods common to both, that the
zones from the Planilimbata limestone to the top of the Gigas lime-
stone in Sweden and Norway are the equivalent of the zones B,, and

~ By, in Russia. As to the exact correlation of the subdivisions there

is not so great unanimity of opinion, but as to the bounding formations,
the Planilimbata limestone and the Gigas limestone on one side; and
the Glauconite limestone (B,,.) and the Orthoceras limestone (B,,,,),
on the other, there can be no question.

In Sweden the position of this limestone in respect to the graptolite
succession is definitely fixed. We know that the Planilimbata lime-
stone succeeds the Ceratopyge zone, and that the Tetragraptus phyllo-

230 BULLETIN: MUSEUM OF COMPARATIVE ZOOLOGY.

graptoides zone succeeds the Ceratopyge zone, and precedes the Lim-
bata limestone. The Planilimbata limestone can not be older than
the oldest of the Tetragraptus zones. On Oeland, Isograptus gibber-
ulus, and other graptolites of the Lower Didymograptus zone occur
in the Lower Asaphus limestone, which Lamansky correlates with the
Russian zone B,,;g. In Norway the Gigas limestone is followed by
black shale with Didymograptus geminus. On Oeland, the Gigas
limestone is followed by the Platyurus limestone, and that in turn by
the Chiron limestone, which contains Didymograptus geminus. The
top of the Gigas limestone, and likewise of the Orthoceras limestone
(B,:1,) of Russia, is therefore somewhat older than the Didymograptus
geminus fauna. In Scania, the Didymograptus geminus fauna is found
in shales resting on other shales containing Phyllograptus cf. typus and
Didymograptus cf. bifidus. This in turn rests upon the Orthoceras
limestone of that region. This limestone, in turn follows shale with
the Isograptus gibberulus fauna. The limestone contains a large
trilobite fauna which, however, in our incomplete knowledge of the
faunas of the Swedish “Orthoceras limestone,” it is not safe to corre-
late directly with the faunas of other regions. Since it follows the
zone with Isograptus gibberulus, one would naturally correlate the
Orthoceras limestone of Scania with the Lower Asaphus limestone of
Oeland. If this correlation is correct, then the Phyllograptus cf.
typus beds of Scania would appear to represent some or all the Upper
Asaphus limestone, Gigas limestone, and Platyurus limestone of
Oeland. But for reasons to be given later, the Platyurus limestone is
probably to be eliminated from this list. This does not of course,
prove definitely that the Gigas limestone is of the age of Phyllograptus
cf. typus, though there is a strong presumption, since both are older
than the zone with Didymograptus geminus. Fortunately however,
Schmidt found Phyllograptus in B,,, in Esthonia, and Phyllograptus
sp. has been reported from above the Gigas limestone of Norway.

The limestone of the Gigas and lower zones, down to the bottom of
the Planilimbata zone of Sweden, and the zones B,, and B,,, of Russia
must therefore be placed within the range of Phyllograptus in the
graptolite succession. It remains now to see what that means in the
American sequence.

The occurrence of the Tetragraptus-Phyllograptus-Didymograptus
fauna in the shales of the Lévis formation of Canada has long been
known, and recent work by Ruedemann (115) and the writer (111)
has shown the order of succession of the faunules within the forma-
tion. The relation of the Lévis to the Beekmantown of America

RAYMOND: CORRELATION OF THE ORDOVICIAN STRATA. 23]
is, however, not yet fully settled, though evidence accumulated by
Dr. Ulrich (119) shows that the Lévis and Beekmantown are probably
of the same age. This evidence is based largely on the occurrence of
graptolites in dolomite with a Beekmantown fauna near Smithville,
Lawrence Co., Arkansas. At this locality, Phyllograptus ‘licifolius,
P. angustifolius, Didymograptus bifidus, and D. amplus have been found
with brachiopods and fragments of trilobites. Ina bed just above the
one containing graptolites, Plethospira cassina, Subulites obesus, and
Eurystomites kellogt were found. These latter fossils are characteristic
of the Cassin limestone and occur about midway in the section of the
Beekmantown on the eastern side of Lake Champlain. In Vermont
the strata of the Beekmantown are principally limestone and dolomite,
the total thickness being about 1200 feet, the top being unknown, as
the upper beds were eroded before the deposition of the Chazy. The
top of the Cassin limestone is 470 feet below the top of the Beekman-
town and the formation is 100 feet thick. No graptolites have been
found in the Beekmantown in Vermont. There seems no doubt
however, that the mollusks cited above are characteristic of the Cassin
and, therefore, the graptolites found in Arkansas belong in the Cassin
or in an older formation, and are surely Beekmantown in age.

In England all the species of Phyllograptus and Tetragraptus are
in the Arenig, and in America the various zones of the Lévis contain
species of these genera, and all the zones are so knit together by com-
mon species that it seems quite evident that all belong to a continuous
series. Phyllograptus ilicifolius is a long ranging species at Lévis, but
Didymograptus bifidus is found in shale in the middle of the section.
The fossils in Arkansas thus suggest that the middle of the Lévis
corresponds to the middle of the Beekmantown, and that the two are
approximately equivalent.

From these considerations one feels justified in concluding that all
the strata characterized by Phyllograptus, Didymograptus bifidus,
and Tetragraptus, both in Europe and America, are equivalent, and
represent the deposits of Beekmantown (Arenig) time. It has been
pointed out in the previous detailed discussions that the Dictyonema
and Ceratopyge zones of Scandinavia are related to the strata over-
lying them, so that the final correlation would be that the formations
from the Packerort to Kunda (inclusive of both) (A:-B,,,) of Russia
are equivalent to the strata from the Dictyonema zone to the Gigas
limestone of Scandinavia and to the Canadian (Beekmantown) of
America.

Zoe BULLETIN: MUSEUM OF COMPARATIVE ZOOLOGY.

LowER MIDDLE ORDOVICIAN.

As has already been stated, there is throughout northern Europe,
a distinct change in fauna after the deposition of the last of the strata
correlated above with the Beekmantown of America.

In Russia, in Vastergétland, and on Oeland, where there is a con-
tinuous section of calcareous strata this change is less marked than in
Norway, where the Gigas limestone is succeeded by black shales, but
it is shown as well in Scania where the strata of the section are mostly
shales.

In Russia, C; marks the introduction of the Echinosphaerites-
Christiania fauna, in which nearly all the species are different from
those found in the strata below. In Norway, the Gigas limestone is
followed by black shales with the Didymograptus geminus fauna, and
in Scania the shale with Phyllograptus cf. typus is followed by shale
with Didymograptus geminus. On Oeland, the Gigas limestone is
followed by limestone with Asaphus platyurus, and that in turn by
limestone with Illaenus chiron and Didymograptus geminus. 'Through-
out Scandinavia then, the zone with Didymograptus geminus is the
earliest zone of the series succeeding the Arenig. In Great Britain
likewise there is a zone of Didymograptus murchisoni, a species practi-
cally identical with D. geminus, which is considered as the lowest zone
of the Llandeilo. In Great Britain there are two species of Dicello-
graptus in the fauna of this zone, and no species of either Phyllo-
graptus or Tetragraptus are present. The latter statement is true
also of Scandinavia, and the zone is known at all localities as being
the oldest in which there is a profusion of diplograptids. This zone
is not yet positively identified in America. Ruedemann has provi-
sionally correlated the Bed 7 of the Deep Kill section of New York
with the European strata containing the Didymograptus murchisoni
fauna, but although Phyllograptus and Tetragraptus were not found
by Ruedemann, the remainder of the fauna is so nearly identical with
the fauna of the Diplograptus dentatus zone of the Lévis section, where
both Tetragraptus and Phyllograptus do occur, that I am inclined to
believe that the zone 7 belongs with the Lévis.

In Scandinavia the Didymograptus geminus beds are followed by the
Lower Dicellograptus shales, containing the zones, in ascending order,
of Glossograptus hincksi, Climacograptus putillus, and Nemagraptus
gracilis. Unfortunately, a great deal remains to be done before a very
satisfactory correlation of the Echinosphaerites limestone of Norway

RAYMOND: CORRELATION OF THE ORDOVICIAN STRATA. 233

with the Lower Dicellograptus shale of Sweden can be made. Echino-
sphaerites has not yet been reported directly associated with a grap-
tolite fauna containing diagnostic species. ‘The shales with the Lower
Dicellograptus faunas are found at the north in Jemtland and at the
south in Scania, in both of which places Echinosphaerites is absent.

On Oeland, Echinosphaerites occurs at two horizons. The first
occurrence is in the Platyurus limestone, in strata below the Chiron
limestone, which contains Didymograptus geminus. The second
appearance is in Chasmops limestone, which is the horizon in which
it is found at Kinnekulle and in Vastergétland generally.

At Kinnekulle, Echinosphaerites is found in the Chasmops lime-
stone, which is a formation consisting of limestone interstratified with’
shale, the shale holding undetermined graptolites. Very little seems
to be known of the detailed distribution of the faunas in the Chasmops
formation in Vastergétland but the general consensus of opinion seems
to be that the Echinosphaerites is confined to the lower portion. Of
the trilobites listed from the Chasmops limestone in Vastergétland
and Dalecarlia, Remopleurides sexlineatus, Ptychopyge glabrata, Ampyx
rostratus, and Agnostus trinodus all occur in the zone of Dicrano-
graptus clingant in Scania. The zone of Dicranograptus clingani is
the one next above the zone of Nemagraptus gracilis, both in Scania
and in Jemtland.

In Gistrikland and Dalecarlia conditions seem to be somewhat
similar to those in Russia, for in Delarne the Platyurus limestone
contains a layer practically made up of “linsen,” and boulders from
Gistrikland referred to the Chiron limestone contain “linsen”’ and
such typical Russian species as Asaphus kowalewski, A. cornutus, and
Christiania oblonga. In other boulders, said to be lithologically lke
the Chiron limestone, but referred by Wiman to the Chasmops lime-
stone, Echinosphaerites aurantium, and Christiania oblonga are found.

In the Christiania district of Norway, as has already been stated,
the Gigas limestone is succeeded by black shale and limestone, forty
to forty-five meters thick, with Didymograptus geminus. Other
significant fossils found here are Asaphus platywrus, Ogygiocaris dila-
tata, and Lituites lituus, fossils found in the Platyurus, Chiron, and
Ancistoceras zones at Kinnekulle and on Oeland, leaving no doubt as
to the correlation of these strata. Following this zone which is known
as 4 aa, is the zone 4 a8, the zone of Echinosphaerites aurantium. Here
the strata are dark blue to black limestone and dark shale inter-
stratified and the thickness is about fifty meters. The Echinosphae-
rites are confined to certain layers and various trilobites are present,

234 BULLETIN: MUSEUM OF COMPARATIVE ZOOLOGY.

among them, Ogygiocaris dilatata, Nileus armadillo, Trinucleus cos-
einorrhinus, all of which are found in the zones of Climacograptus
putillus and Nemagraptus gracilis in Jemtland.

It would seem from the comparison of sections that Echinosphaer-
ites appeared in northern Europe first in Oeland and migrated thence
into Norway and Russia. Its occurrence in Oeland in strata older
than those containing Didymograptus geminus shows definitely that
it antedated there the first appearance in Norway, for in the latter
country it is found first in strata resting upon those containing D.
geminus. In Russia the sequence of faunas, Echinosphaerites first
and then Echinosphaerites and Christiania, is exactly the same as in
‘Norway, strongly suggesting that the Echinosphaerites did not reach
that country sooner than it did Norway. There is physical evidence
in Russia of an interruption of sedimentation after the Kunda forma-
tion was deposited, while there is but slight evidence of a break be-
tween the Reval and the Kuckers. This, coupled with the faunal
evidence, particularly the total absence of the Ogygiocaris fauna in
Russia, indicates such a correlation as I have shown in the table.

The Ogygiocaris fauna is Norway is found best developed in 4 aa
but many of the species pass over into 4 a8, among them the Ogygio-
carts dilatata itself. In Jemtland Ogygiocaris is found in the zones of
both Climacograptus putillus and Nemagraptus gracilis and serves to
connect the Nemagraptus zone with the first Echinosphaerites zone
in Norway (4 af).

NoORMANSKILL.

The Didymograptus geminus fauna has not yet been identified in
America, but the Nemagraptus gracilis fauna is well known from the
Normanskill shale of New York. At the type-locality, however, the
Normanskill shale is very much faulted, folded, and twisted, and its
correlation with the formations of the standard section is not yet
established. Ruedemann at first considered it to be of early to Middle
Trenton age but later inclined to correlate it with the Black River.
Ulrich (119) has considered it still older, placing it below the Low-
ville, but above the typical Chazy, making it a member of his Blount
group, which he places between the Stones River and the Lowyville.

In Virginia typical Lower Dicellograptus faunas (Nemagraptus
gracilis zone) have been seen in sections where the sequence is normal
at two localities, but in neither is the evidence fully established. In
looking over material collected by Drs. E. O. Ulrich and George W.

e

RAYMOND: CORRELATION OF THE ORDOVICIAN STRATA. 235

Stose from thin-bedded dark limestone at the Mattheson limestone
quarry near Abingdon, Virginia, I found numerous specimens of a
Robergia very like R. micropthalma and a Telephus like 7. bicuspis.
These trilobites were associated with diplograptid graptolites which
Dr. Ulrich assures me are of typical Normanskill species. Such an
association is found in the Climacograptus putillus zone of the Lower
Dicellograptus shale at Anderson, Jemtland, Sweden, and the asso-
ciation is too remarkable a coincidence to indicate anything other
than approximate contemporaneity of the two formations. Stose
gives the following section at Abingdon.

Sevier shale — Eden fossils in upper and Trenton fossils in Feet.

lower part. S00
Moccasin limestone. Red limestone with few fossils. . 400
Ottosee limestone, with Echinosphaerites. 200

Athens shale. Calcareous shale and dark blue shaly lime-

stone above; dark gray fissile shale below.

The fauna mentioned was found in lower

part. 500-600
Stones River limestone. 400
Knox dolomite (of Beekmantown age).

From this section it may be seen that the Normanskill is younger
than some part of the Stones River, and considerably older than the
Sevier shale, the lower part of which seems to be of Trenton age,
though to what part of the Trenton it appertains is not yet evident.

The other section containing the Normanskill fauna was described
by Powell (105) and is near Salem, Virginia. He records the following,
in descending order: —

Medina sandstone (Silurian). Feet.
8. Shale, very fossiliferous. No list of fossils. 1200
7. Blue to black shale and limestone. | 300
6. Red and gray sandstone without fossils. 400
5. Blue and black limestone without fossils. 500
4. Black carbonaceous shale with numerous graptolites. 32

species are listed, including Nemagraptus gracilis, Clima-

cograptus putillus and Dicellograptus sextans. Triarthrus

and Trinucleus are also recorded. 560
3. Coarse grained dark limestone with bands of marble. 250
2. Pure “dove” colored limestone with a Tetradium and

large gastropods. 50

1. Cherty magnesian limestone with Maclurites. 500

236 BULLETIN: MUSEUM OF COMPARATIVE ZOOLOGY.

Professor Powell was good enough to spend three days in showing
the writer this section, and my interpretation of it differs somewhat
from that in his published account. The cherty magnesian limestone
at the base (zone 1) appears to belong to the Beekmantown, not the
Chazy, and Ophiletas were found in the upper beds. Zone 2 is a fine-
grained buff limestone with numerous gastropods and some trilobites.
At the base is a conglomerate with pebbles of magnesian limestone
and chert in a caleareous matrix. This formation is to be correlated
with the Mosheim of southwestern Virginia and eastern Tennessee,
and that in turn is correlated with the lower part of the Stones River
of central Tennessee.

The coarse-grained dark limestone of zone 3 is very fossiliferous,
some of the genera present being Hormoceras, Amphilichas, Illaenus,
Isotelus, Orthis, Dinorthis, Plaesiomys, Oxoplecia, Leptaena, Plectam-
bonites, and Solenopora, besides numerous bryozoans. This fauna
is much more like that of the Black River of New York than it is like
any fauna of the Chazy in the typical region, Oxoplecia and Plectam-
bonites in particular being unknown in the Chazy. On the other
hand, the fauna is more or less like that of the Holston and Lenoir of
eastern Tennessee, and these latter formations seem to be of Middle
Chazy age. The relation of this formation to the one below is exactly
like the relation of the Leray to the Lowville in New York. The line
of separation between the dark, impure limestone above and the pure
light-colored limestone below is a sharp one, and yet the top of the one
formation and the bottom of the other are combined to form a single
layer; a so-called “welded contact.”

The Athens shale (zones 4 and 5) is a dark fossiliferous shale in the
lower portion, and passes rather gradually into an almost entirely
unfossiliferous blue limestone above. Nemagraptus gracilis and Didy-
mograptus occur in the lower part of the shale, while Dicellograptus,
Climacograptus, and the beautiful synrhabdosomes of Diplograptus
are most abundant at about the middle. Ampyx americanus appears.
to begin its range with these latter fossils, being here accompanied by a
Triarthrus, and extends up into the limestone of zone 5. In the upper
part of its range, I found it accompanied by Cryptolithus, Robergia,
and Acrothele. The Athens is plainly equivalent to the Normanskill
of New York, and the Lower Dicellograptus shales of Sweden.

Zone 6, the Tellico sandstone, is practically unfossiliferous here as.
elsewhere. Upon it rests a thick mass of shale with some thin-bedded.
limestone. This formation is generally called the Sevier in south-
western Virginia, and has not yet been studied in sufficient detail to

ae = SS ~~

7
*

RAYMOND: CORRELATION OF THE ORDOVICIAN STRATA. 237

permit of exact subdivision or correlation. In this section, the lower
500 feet or thereabouts appear to be of Trenton age, while the re-
mainder of the strata are probably Eden and Maysville, but where to
draw the line between the two is not yet fully determined. Professor °
Powell is engaged upon a further study of the section, and will soon be
able to give more detailed information.

The lower 100 feet of the part of the Sevier referred to the Trenton
consists chiefly of shale, and its fossils are Calymene, Dalmanella,
and Rafinesquina. The next 125 to 150 feet consist of alternations

- of thin-bedded limestone and shale, the latter predominant. In this

zone Cryptolithus tessellatus is common, associated with Ceraurus
pleurexanthemus, Calymene, Sinuites, Plectambonites, and Dalmanella.
About 100 feet above this zone, hemispheric bryozoans are common,
and with them are Parastrophia hemiplicata, Dinorthis, and Sinuites.
The Parastrophia is not the pauciplicate form found in the Lower
Trenton of New York and Quebec, but like the form in the Middle and
Upper Trenton of Ontario. In this same zone Professor Powell found
some graptolites which appear to be Lasiograptus eucharis, a Middle
Trenton and Utica graptolite in New York. The upper 100 feet of
the strata here referred to the Trenton consist almost entirely of thin-
bedded blue limestone, and at the very top are great numbers of
Zygospira, Plectambonites, and Pholidops, while some layers are full
of gastropods.

The rocks above the Trenton consist principally of shale, becoming
more and more sandy toward the top. In the middle are some
massive calcareous strata, and an occasional layer of limestone is met

with at various horizons. The whole reminds one very much of the

Lorraine of New York, and the fossils emphasize that impression.
The upper 100 feet, more or less, belongs to the Bays sandstone, and
has the typical fauna, Platystrophia (or Orthorhyncula) stevensoni
Grabau, Byssonychia walkerensis Grabau, B. radiata (Hall), and other
pelecypods and brachiopods. This fauna is generally considered to
be of Upper Maysville age.

The section, as interpreted above, seems to place the Normanskill
definitely as post-Middle Chazy and pre-Trenton, the Athens and the

Tellico together occupying the position usually filled by the Upper

Chazy and Black River. Since the Normanskill graptolites occur in
the lower part of the formation, they would probably be of Upper
Chazy age.

238 BULLETIN: MUSEUM OF COMPARATIVE ZOOLOGY.

ECHINOSPHAERITES AND CHRISTIANIA FAUNAS.

The position of the Echinosphaerites beds in the American Ordovi-
cian section can not be said to be definitely established. The knowl-
edge which we have of these beds is due largely to Dr. Ulrich and to
Dr. Bassler (98), and by the former of these investigators the fossil is
reported as occurring at three horizons, one below and two above the
Lowville. \ | |

The evidence concerning the younger of these occurrences, in so far
as it has been published, may be found in the description by Stose, of
the Chambersburg — Mercersburg map-area, and in the Revision of
the Palaeozoic Systems by Ulrich. The following section (here re-
arranged) is given by Stose (118).

Feet.
10. Soft yellowish green sandstone with few fossils, said to
be of Eden species 1200+
9. Black carbonaceous shale, with Climacograptus spinifer,
Corynoides calicularis, Leptobolus insignis, Triarthrus
becki, ete., in lower 100 feet. 800+
8. Shale and thin-bedded limestone with many small
fossils, including Triarthrus becki, Cryptolithus tessella-
tus, Ampyx, Caryocaris, etc. 150+

All the above strata are referred to the Martinsburg shale.
7. Caleareous shale and shaly limestone, with Christiania

trentonensis, Plectambonites asper, P. pisum, Oxoplecia,
Parastrophia hemiplicata, ete. Echinosphaerites in the

upper ten feet. 150+
6. Dark gray, largely thin-bedded limestone with Nidu-
lites favus, Ampyx, Plectambonites asper, ete. 237+

5. Dark gray limestone in which Echinosphaerites is very
common, Ampyx, Receptaculites, Oxoplecia, and

brachiopods also common. 60+
4. Grayish dense thin-bedded limestone with Tetradiwm
cellulosum, Zygospira recurvirostris, ete. 150+

These limestones above are all grouped as the Chambersburg
limestone.

¢
;
:
‘

RAYMOND: CORRELATION OF THE ORDOVICIAN STRATA. 239

3. Thin-bedded, pure, fine-grained limestone with Leper-
ditia fabulites. 275+
2. Massive pure limestone and layers of black chert.
Maclurites magnus, Tetradium syringoporoides, Ampyx
halli, and brachiopods. 200+
1. Massive and thin-bedded pure and magnesian limestone. 600+
Beekmantown limestone.
Zones 1 to 3 are correlated with the Stones River.

According to Dr. Ulrich, the section near Chambersburg, Pa., has
a “lower Echinosphaerites bed,” forty to fifty feet in thickness, resting
upon about 150 feet of limestone referred by him to the Lowville
(4 of section on preceding page). This Echinosphaerites bed, which
lithologically is a very earthy limestone is said by Ulrich (119, p. 322
to be overlain by about 300 feet of hard dark limestone with Nidulites,
and that in turn by 270! feet of strata in which thin beds of lime-
stone are interstratified with thick beds of shale. ‘These strata are
characterized by Christiania. A few feet below the top is the upper
Echinosphaerites zone, and here Christiania is most abundant. The
Christiania beds are capped by the Martinsburg shale, which con-
tains Triarthrus, Cryptolithus, and Corynoides.

At Strasburg, Virginia, still according to Ulrich, the lower Echino-
sphaerites bed rests upon a cherty limestone 100 feet thick, and the
Echinosphaerites is accompanied by brachiopods and bryozoans which
suggest to him a correlation with the Decorah shale of Minnesota
(called Black River by Ulrich). Above this zone are the massive beds
with Nidulites, 207 feet thick, followed by a forty foot bed of argilla-
ceous gray limestone and calcareous shale, containing Echinosphaer-
ites (upper zone) and brachiopods, with other fossils characteristic
of the Christiania fauna. This bed is followed above by 300 feet of
thin-bedded argillaceous light gray limestone and calcareous shale,
passing at the top into true shales. This limestone is referred to the
Martinsburg, since it has, in shaly beds ten to thirty feet above the
base, Corynoides cf. C. gracilis, C. calicularis, Climacograptus spinifer,
Leptobolus insignis, ete.

This manner of occurrence is in striking accord with that in Norway,
where Echinosphaerites occurs first in stage 4a8 without Christiania,
and then at a higher horizon 4ba with that fossil.

At Bellefonte, Penn., according to observations made by Mr. R. M.

1 These figures are given on the authority of Ulrich and do not agree with the sections pub-
lished by Stose.

240 BULLETIN: MUSEUM OF COMPARATIVE ZOOLOGY.

Field and the writer, a zone of dark limestone, containing such typical
Leray (Black River) species as Columnaria hallt and Maclurites logani,
is followed by more argillaceous limestone containing Echinosphaerites
and a large number of other fossils. Christiania has not yet been
found in the Bellefonte section, but this section does definitely show
that the Echinosphaerites zone is there younger than the Leray —
Black River of New York. As shown by Mr. Field, there is essential
agreement between the section at Bellefonte and that at Chambersburg
and Strasburg, so that all three of these occurrences of Echinosphae-
rites may be dated definitely as post-Leray.

According to Ulrich (119), the Kimmswick limestone at Thebes,
Illinois, and Cape Girardeau, Missouri, has at the top a bed of erystal-
line limestone, from five to thirty feet in thickness, which contains
Echinosphaerites and Comarocystites, among other fossils. The
Kimmswick at this locality can not be definitely placed in the section,
except that it is post-Lowville. In the Nashville dome in Tennessee
a formation correlated by Ulrich with the Kimmswick and containing
Echinosphaerites has been found at Aspen Hill, where it is forty feet
thick, and followed by the Hermitage, the Bigby, and the Catheys
formations. The contact with the underlying formation is not shown
but Ulrich states that there is no doubt that it rests upon the
Carters, which is the equivalent of the Leray or Lowville of New York
so that it may safely be stated that here again the Echinosphaerites
bed is post-Leray. At this locality we have the Echinosphaerites
without Christiania, and the zone apparently corresponds to the lower
zone at Chambersburg, Strasburg, and Bellefonte. In this case the
formation containing the Echinosphaerites is limited above by the
Hermitage formation, a formation which can not be correlated with
any New York formation, but which corresponds to the Logana of
Kentucky and is also found at Bellefonte above the zone of Echino-
sphaerites. The Hermitage is followed above by the Bigby limestone,
which contains a fauna corresponding to that of the Prasopora zone,
or Middle Trenton of New York and Ontario. The Kimmswick lime-
stone, and the corresponding Echinosphaerites zone in Pennsylvania
and north-central Virginia, may therefore be correlated with some
confidence with the lower part of the Trenton of New York.

The other occurrence of Echinosphaerites in the Appalachian region
is in the Ottosee formation of southwestern Virginia and eastern
Tennessee. Dr. Ulrich believes that the Ottosee is older than the
Lowville, and, if this can be shown to be correct, then this zone is
older than the two already discussed. A good section showing the

RAYMOND: CORRELATION OF THE ORDOVICIAN STRATA. 241

position of the Ottosee in relation to the other formations is that given
by Bassler (98) from outcrops on Walker Mountain north of Marion,
Smyth Co., Virginia. The strata there named “Holston marble and
associated strata’’ were later named Ottosee by Ulrich.

Feet.
Sevier shale.
Brown to olive and gray shales. 1500
Moccasin limestone.
Impure, argillaceous limestone. 300
Ottosee formation.
(e) Unfossiliferous drab shales. 40
(d) Nodular limestone and yellowish to gray Bak |
holding many Bryozoa. 30
(c) Massive gray and pink marble with numerous
Bryozoa, Solenopora and Stylaraea parva. 30
(b) Clayey nodular limestone and shale. Some of the
layers are crowded with Receptaculites. 50
(a) Massive crystalline limestone. 40
Athens shale.
Dark to black shale with black slaty limestone at the
base. Linguloids and trilobites are abundant at the
base. / 500+
Stones River formation. 170

Knox dolomite (Beekmantown in age).

The Stones River of this section is stated by Bassler to contain in its
upper part a typical Chazy fauna, though the only species mentioned
are Maclurites magnus and Stylaraea parva.

‘ The trilobites mentioned as occurring in the Athens are not listed,
but it is known that this formation in at least two places in Virginia
carries the Nemagraptus gracilis (Normanskill) fauna. (See p. 234).

From the Ottosee Bassler lists Echinosphaerites, Batostoma seviert,
Scenellopora radiata, Diabolocrinus .vesperalis, Solenopora compacta,
and Receptaculites biconstrictus. It is quite true that this fauna is
somewhat unlike that of the Echinosphaerites zone in the other
localities, but there is nothing in the composition of the fauna
itself to indicate that it is older than Black River. Receptaculites
biconstrictus is similar to the Black River R. occidentalis, which occurs
with Echinosphaerites at Chambersburg and Bellefonte, Pa., Batos-
toma is principally a Black River genus, and the other fossils of the
Ottosee have a rather indefinite value in correlation.

244 BULLETIN: MUSEUM OF COMPARATIVE ZOOLOGY.

coryphe, Calymene, Conularia, and Echinosphaerites. The zone is
forty to forty-five meters thick about Christiania.

The strata of 4b@ consist of thin-bedded dark blue limestone with
shaly partings and the thickness is about twelve meters. Here one
finds much the same fauna as in 4ba, but Christiania has disappeared.
A Platystrophia of large size, like the P. lynz of the Itfer and Jewe was
collected from this zone.

4by is another zone of much shale and some thin-bedded limestone,
with a variable thickness, usually from thirteen to sixteen meters.
Chasmops extensa is the guide fossil and Brégger has not listed any
others. I myself found no fossils worth saving at this horizon.

4bé, the last of the zones of 4b, consists of interbedded dark blue
limestone and almost black shale, the thickness being about twelve
to seventeen meters. In this zone are found the last and the largest
of the Echinosphaerites, and a very Trenton-like fauna, in which I
was interested to note two common American forms, a Parastrophia
somewhat like P. hemiplicata, and a Triplecia very like T. nuclea.
Cyclocrinites spasskit makes its first appearance in Norway at this
horizon, and Illaenus, Ampyx, Trinucleus, Remopleurides, Cybele,
Chasmops, etc., are present. Brodgger (94) correlates this zone with
the Jewe of Russia with which I entirely agree, only adding that the
presence of Cyclocrinites spasskw suggests also the Kegel. It seems
quite possible that there is a break in the sedimentary record in the
Christiania district at this point.

In the Christiania area the zone 4bé is followed by the zone 4ca,
the beginning of the Trinucleus seticornis fauna, correlated with the
Trinucleus shales of Sweden. In the district Mjésen, north of Chris-
tiania, however, Holtedahl (97) has found a different succession,
and strata which, in my opinion, are to be intercalated between
4bé and 4ca, and not to be correlated with 4c as Holtedahl has done.
As in Christiania, stage 4 in Mjésen is introduced with an Ogygio-
caris zone, containing Ogygiocaris dilatata, Didymograptus geminus and
many other species, this zone having a probable thickness of twenty
to thirty meters. This is followed by a thin zone of calcareous shales,
three to four meters thick, and it in turn by black shales with limestone
nodules, the thickness unknown. The fauna consists very largely of
gastropods and cephalopods, of which many species are listed. The
fauna is connected with the preceding zones by’the presence of Ogy-
giocaris dilatata, but no species pass on into the overlying strata.
We have here probably a very unusual development of the strata of
the age of the zone of Nemagraptus gracilis.

RAYMOND: CORRELATION OF THE ORDOVICIAN STRATA. 245

The next formation, B, of Holtedahl’s section, consists of practically
unfossiliferous shales and thin-bedded sandy limestone which passes
above into By, sandy limestone and shale containing Coelosphaeridium
cyclocrinophilum. In the upper part of the same formation is a large
fauna which includes many species found in the Jewe of Esthonia.
Cybele grewingki Schmidt, Chasmops marginata Schmidt, Porambonites
schmidti Noetling, Coelosphaeridium cyclocrinophilum Roemer and
Mastopora concava Eichwald are among the striking species which
Holtedahl lists as common to the two regions. The thickness of this
formation is unknown but it must be over forty meters.

Then follow two very fossiliferous zones both characterized by an
abundance of Cyclocrinites, and making a total thickness of about
twenty-three meters. These formations have a fauna entirely com-
parable to that in the Kegel of Esthonia. Holtedahl lists the follow-
ing species found in both B; and the Kegel :—

Basilicus kegelensis Schmidt, Chasmops maxima Schmidt, Chas-
mops bucculenta Sjogren, Pterygometopus kegelensis Schmidt, Bucai-
niella lineata Koken, Leptaena aff. schmidti Térnquist, Platystrophia
biforata Schlotheim, Triplecia insularis Eichwald, Cyclocrinites oelandi-
cus Stolley, C. vanhoeffeni Stolley, and C. balticus Stolley.

Holtedahl correlated these latter zones, Bsa and Bsgb, with the
Trinucleus shales of Sweden, the Trinucleus shales and Isotelus
limestone of Norway, and the Kegel and Wesenberg of Esthonia.
In regard to the correlation with the Trinucleus shales, Holtedah!]
himself says that there are only two species, Illaenus linnarssoni and
Remopleurides dorsospinifer common to the two. Neither of these
species is a guide fossil, I//aenus linnarssoni in particular having a
very long geological range. It is worthy of note that none of the
Bohemian species which} make so important a part of the fauna of
the Trinucleus zone of Sweden and the Christiania district is found
in B; of the section in the Mjésen district, but the fauna is strictly of
the Russian type and belongs to another basin. In this case, the
difference can hardly be due to difference in facies, for the Trinucleus
beds of the Christiania district contain a large amount of limestone;
not a sandy limestone, it is true, but neither is the Kegel a sandy
limestone. Furthermore, we are not here dealing with graptolites
or other fossils highly sensitive to environmental conditions, but with
general faunas.

In Jemtland, the Trinucleus shales appear to be absent, and among
the boulders of “Oestseekalk” are found many of the characteristic
fossils of the Kegel and Wesenberg limestones, indicating that these

246 BULLETIN: MUSEUM OF COMPARATIVE, ZOOLOGY.

formations may in former times have extended across this north-
ern region. In Dalecarlia, however, which is well to the north,
the Trinucleus shales are well developed, and contain the typical
southern fauna. The Kegel and Wesenberg are certainly older than
the Trinucleus shales.

The Chasmops limestone of Sweden has been little studied except
for the work of Olin on the trilobites in Scania, and the time is not yet
ripe for accurate correlations. The Upper Chasmops or Macrourus
limestone of Dalecarlia contains such species as Chasmops maxima
Schmidt, found in Russia in the Jewe and Kegel, while the lower
Chasmops limestone of the same region contains such typical Kuckers
and Itfer fossils as Echinosphaerites aurantium, Chasmops odini, and
Oxoplecia dorsata. 'The Chasmops limestone is therefore correlated
as indicated by these fossils. As mentioned above, there are certain
trilobites which serve to connect the Chasmops limestone with the
shales in Scania containing Dicranograptus clingant and Pleurograptus
linearis. Between the zone of Nemagraptus gracilis and that of
Dicranograptus clingant, there is, it seems, a considerable gap, so that
the section in Scania is far from complete.

ZONE OF DICRANOGRAPTUS CLINGANI.

In southern Sweden the zone of Nemagraptus gracilis is succeeded
by the zone of Dicranograptus clingani, or as it is sometimes called,
the Middle Dicellograptus beds.

These beds, have recently been studied in great detail on Born-
holm, politically an apanage of Denmark, but geologically in the
Scanian province. Hadding (61) has recognized four subzones on
this island. These are in descending order:

4 Zone with Climacograptus styloideus Lapworth.

3. Zone with Dicranograptus clingani Carruthers.

2. Zone with Amplexograptus vasae (Tullberg).

1. Zone with Climacograptus rugosus Tullberg. 7

Among the fossils of zone 4 may be mentioned, beside the name
fossil, Glossograptus quadrimucronatus, Diplograptus truncatus, Dicello-
graptus pumilis, Leptograptus flaccidus macer, and inarticulate brachio-
pods.

Zone 3 has a large fauna, including among others, Amplexograptus
vasae, Climacograptus brevis, Dicellograptus forschammeri, D. pumilis,
and Corynoides curtus. There are also several inarticulate brachio-
pods and Dalmanella argentea.

Ce

—

RAYMOND: CORRELATION OF THE ORDOVICIAN STRATA. 247

In ‘zone 2 the only graptolites are Amplexograptus vasae and Cory-
noides curtus.

Zone 1 contains Climacograptus scharenbergi and C. rugosus.

In America the zone of Dicranograptus clingani, though that species
itself is not present, is found in the shales of eastern New York and
Canada. These shales were long called Utica, but Ruedemann
(113 and 117) has shown that they are older and equivalent to the
caleareous Trenton of Central New York. The shales are also
involved in the mass of the so-called Hudson River shales in the Hudson
Valley. They are the Cumberland Head shales of the Champlain
Valley, the Snake Hill shales of the Hudson Valley, and the Cana-
joharie shales of the Mohawk Valley.

At Canajoharie, the type-locality for the formation of that name,
Ruedemann (117) found the following sequence of faunas :—

At the base, are seventeen feet of dark blue fossiliferous limestone
with interstratified shale beds. This limestone is basal Trenton and
contains the Cryptolithus fauna in the limestone, while the interstrati-
fied shale afforded Corynoides calicularis, Diplograptus amplexicaulis,
and Mesograptus mohawkensis, so that we here have an indication of
the fauna of the shaly equivalent of the Glens Falls limestone.

Above this comes the Canajoharie shale, which is strongly cal-

careous in its lower portion, becoming truly argillaceous above.

In the lower sixty-five feet of the Canajoharie, Diplograptus ampleai-

~ caulis is the only graptolite, while in a zone between 65 and 120 feet

above the base of the formation, Corynoides calicularis, Diplograptus
putillus, and Lasiograptus eucharis are added. Brachiopods, pelecy-
pods, and crustaceans are fairly common in both zones. In the zone
from 115 to 150 feet above the base Glossograptus quadrimucronatus
cornutus appears, associated with some of the previously mentioned
graptolites. These pass up through the next 120 feet, to 270 feet
above the base of the formation. Triarthrus becki is not noted until
a height of 190 feet above the base of the formation, when it appears
suddenly in great numbers.

Thus, combining the information obtained from the Canajoharie
and Sprakers sections — these localities are only about two and one
half miles apart — Ruedemann made out four faunal zones.

4. Zone of Climacograptus spiniferus, Diplograptus vespertinus,

and Lasiograptus eucharis.

3. Zone of Lastograptus eucharis, Trocholites ammonius, etc.

2. Zone of Glossograptus quadrimucronatus cornutus, Corynoides

calicularis, ete.

248 BULLETIN: MUSEUM OF COMPARATIVE ZOOLOGY.

1. Zone of Diplograptus amplexicaulis, Corynoides calicularts, ete.

As one goes further west the limestone at the base of the section
becomes thicker and it is found that the graptolites characteristic of
the lower part of the shale at Canajoharie are not present, but the
lowest of the shale beds, resting on the limestone, contain species
which first appear high in the section at Canajoharie.

At Dolgeville, half way between Canajoharie and Utica, where the
lower 200 feet of the Trenton is limestone, the first shale bed contains
Glossograptus quadrimucronatus, Climacograptus typicalis, and Lasio-
graptus eucharis, which is interpreted by Ruedemann as being a fauna
younger than that of zone 2 above, since Climacograptus typicalis is
best developed in the still higher Utica shales.

In the typical Trenton at Trenton Falls, Diplograptus amplexicaulis
is arather common graptolite in the lower part of the formation, where
it is found in limestone. (See Fig. 1, p. 257).

THE TYPICAL UTICA, OR ZONE OF PLEUROGRAPTUS LINEARIS.

At Holland Patent and South Trenton, New York, where the black
shale of the typical Utica rests upon the top of the Trenton limestone
not far from the type-locality, the following fauna has been found: —

Callograptus compactus (Walcott), Dicranograptus nicholsoni Hop-

Climacograptus typicalis Hall, kinson,
C. putillus (Hall), Glossograptus quadrimucronatus
Mastigograptus simplex (Wal- Hall. var.

cott), Lasiograptus eucharis also occurs
M. tenuiramosus (Walcott), nearby and in beds a little
Pleurograptus linearis (Carruth- higher in the section.

ers), L. bimucronatus timidus Ruede-
Leptograptus annectans (Walcott), mann.

This may be taken as the typical Utica fauna, but it may be re-
marked that it has not been found in its entirety, at any other locality.
Associated with these graptolites one finds also Triarthrus becki and
Cryptolithus tessellatus. It will be noted that in the above fauna there
are certain graptolites which are found at a lower horizon, such as
Climacograptus typicalis, Glossograptus quadrimucronatus, and Lasio-
graptus eucharis. Mastigograptus simplex has until recently been
found only at Holland Patent and the immediate vicinity. Lasio-
graptus bimucronatus timidus, Leptograptus annectans, and Mastigo-

aS ee ee eer ee

«a

RAYMOND: CORRELATION OF THE ORDOVICIAN STRATA. 249

graptus tenuiramosus, are restricted to that locality and the Fulton of
the Cincinnati district. Pleurograptus linearis, which is so important
a fossil in northern Europe, is in this country known only from a single
specimen found at Holland Patent. Callograptus compactus is also
known only from Holland Patent. There are varieties of Dicrano-
graptus nicholson in older shales, but the typical form is restricted to
the horizon of the Utica of Holland Patent. It is found also in the
Fulton shale of the Cincinnati district and near Saratoga. In at-
tempting a correlation of any other fauna with that of the typical
Utica it must be borne in mind that some of the typical species of that
fauna are so rare that but few specimens are known. Practically all
the species of graptolites which are not restricted to the locality at
Holland Patent are species -of considerable vertical range. The
occurrence, however, of Pleuwrograptus linearis stamps that fauna as
younger than that with Dicranograptus clingani and probably equiva-
lent to the upper Chasmops shale of Sweden.

The Fulton shale at the base of the Eden at Cincinnati contains a
graptolite fauna very like that at Holland Patent. Some of the
species are: —

Mastigograptus tenuiramosus Climacograptus typicalis Hall,
(Walcott), C. putillus (Hall),
Leptograptus annectans Walcott, Lasiograptus bimucronatus timi-
Dicranograptus nicholsont Hop- dus Ruedemann.
kinson,

Associated with these are Cryptolithus tessellatus and Triarthrus

beckt. In higher beds of the Eden of the same region one finds Dictyo-
nema arbuscula Ulrich, Mastigograptus gracillimus (Lesquereux),
Chaunograptus gemmatus Ruedemann, and Climacograptus typicalis
Hall.
_ From the above we may derive that the more widespread species
of the typical Utica (Pleurograptus linearis fauna) are Mastigograptus
tenuiramosus, Leptograptus annectans, Lasiograptus bimucronatus
tumidus, and Cryptolithus tessellatus. Climacograptus typicalis, Dicran-
ograptus nicholsoni, and Triarthrus becki are also species which though
not strictly restricted to the Utica, are to be found in every outcrop
of that formation.

From the above it is evident that while some of the graptolites have
a long range in the black shales of New York, there are species which
seem to be characteristic of certain horizons. Thus we have in the

250 BULLETIN: MUSEUM OF COMPARATIVE ZOOLOGY.

shale equivalent of the Lower Trenton, Corynoides calicularis and
Diplograptus amplexicaulis, in the shale equivalent of the remainder
of the Trenton (as exhibited in the Trenton Falls section) Lasio-
graptus eucharis, Glossograptus quadrimucronatus var., Climacograptus
putillus, ete., and in the shale above the Trenton of Trenton Falls,
Leptograptus annectans and the other fossils listed above.

If we now return to the Lower Mohawk Valley, we find above the
Canajoharie shale the Schenectady formation, composed largely of
sandy shale and sandstone, but containing some graptolites and other
fossils. Our knowledge of this fauna is due almost entirely to Dr.
Ruedemann. That author has correlated the Schenectady with the
Upper Trenton, but, to my mind, on very inadequate evidence.
Some of the more important fossils of the formation are :—

Azygograptus sp. C. typicalis Hall.
Mastigograptus cf. M. simplex Wal- Lasiograptus eucharis Hall.
cott. Rafinesquina ulrichi James.

Diplograptus vespertinus Ruedemann. T'rocholites ammonius Conrad.
Climacograptus bicornis ulttmus Ruede- Triarthrus becki Green.
mann. Cryptolithus tessellatus Green.
and numerous eurypterids.

Ruedemann states that this fauna bears a Utica aspect, but that
the graptolites point as much toward the Canajoharie fauna as toward
the typical Utica. In this connection we must note the absence of
Diplograptus amplexicaulis, Corynoides calicularis, and Glossograptus
quadrimucronatus. It is true that none of the graptolites listed are
confined to the Utica of the Holland Patent type, but both Climaco-
graptus typicalis and Lasiograptus eucharis are very abundant in the
typical Utica. The presence of a Mastigograptus comparable to M.
simplex also suggests Utica, as does the presence of Eurypterida.
Finally, and most important, is the presence of Cryptolithus tessellatus,
a fossil which to Dr. Ruedemann suggested the Trenton age.

The geological range of Cryptolithus tessellatus seems to be quite
generally misapprehended. It is frequently thought of as a fossil
which occurs almost anywhere in the Trenton, whereas, as a matter
of fact, it is restricted to certain definite horizons and is not every-
where present.

The earliest appearance of Cryptolithus in New York is very near
the base of the Trenton, where it is exceedingly abundant in the
Glens Falls formation. At this horizon it is very abundant near
Quebec, at Montreal, in the Champlain Valley, and the Mohawk

i
4
2
?
q

RAYMOND: CORRELATION OF THE ORDOVICIAN STRATA. 251

Valley, in all of which places it occupies a thin formation, its vertical
range never exceeding forty feet. A very few specimens have been
found in the lowest layers at Trenton Falls, but it is absent from more
northern outcrops of the Trenton. It is not found in the Trenton
anywhere in the region of New York west of the Adirondacks, it is
absent from the Trenton of Ontario and Quebec west of Montreal,
and it is absent from Minnesota. The second occurrence in New
York is not in the Trenton, but in the typical Utica, at Rome and the

,
; vicinity. It occurs also in the Frankfort, and still higher, in the
Pulaski.

‘ In the vicinity of Quebec the second appearance of Cryptolithus is
. in the light-colored sandy shale about 400 feet above the top of the

limestone of the Trenton, and above the dark “ Utica”’ shale.

At Bellefonte, Penn., the earliest appearance of Cryptolithus tessel-
latus is, as in New York, just above the limestone containing the
Leray fauna, and it reappear’ in the upper fifty feet of the 600 foot
Trenton section, at the point where the limestone begins to pass over
into shale, and just before the first appearance of Triarthrus beck.
In Kentucky, Cryptolithus appears first in the Logana (Hermitage),
only a few feet above the base of the Trenton, and does not reappear
till the Cynthiana, just at the top of the Trenton or base of the Eden.
On the Ohio River at Cincinnati it is in the Cynthiana, and the Lower
Eden, and appears again in the Maysville.

The occurrences are so exceedingly alike, and there is so great an
indifference displayed as to the character of the sediments, that I am
inclined to look upon Cryptolithus as an exceedingly good horizon
marker. If this be the case, then the Schenectady formation is to be
correlated with the Utica, and, probably, the Frankfort.

Se ee ee

CORRELATION OF THE TRENTON IN AMERICA.

One great obstacle to any correlation of the kind attempted in
this study is the fact that we have as yet reached no satisfactory
solution to the problems presented by our American Ordovician strata.
By far the best correlation tables for the Ordovician are those recently
presented by Drs. Ulrich and Bassler. My own differs radically from
theirs, and I am therefore compelled to traverse the principal outlines
of the subject in justification of the departures which I have made
from former schemes.

252 BULLETIN: MUSEUM OF COMPARATIVE ZOOLOGY.

It is only some twenty years since it became known that there is in
the United States more than one kind of “Trenton.” Naturally,
the history of the attempts to correlate the various kinds of “Trenton ”’
has been made in that time. The most troublesome of the still
unsolved problems is the exact relation of the “’Trenton” (Rysedorph,
Chambersburg, Quebec City, Chickamauga, Sevier, etc.) of the Appa-
lachians to the “'Trenton”’ (typical Trenton of New York, Trenton of
Ontario, Minnesota, etc.) of the interior.

The Appalachian Trenton, if I may so call it, extends from the
destroyed end of the range at Gaspé in intermittent aligned exposures
as far as Georgia. A beginning on the description of its fauna was
made by Ruedemann (114) in his paper on the fossils in the pebbles of
the Rysedorph conglomerate, but practically nothing more has been
done along that line. Until the fauna is described the problem will
remain unsolved. We have, however, some inkling of what the fauna
is like, and notice that while in general similar to the Trenton faunas
of the interior, it differs in containing Echinosphaerites, Christiania,
Nidulites, Tretaspis, and Lonchodomas in abundance, these genera
being unknown in the interior Trenton. Many undescribed forms are
also peculiar to this Appalachian area, but the above familiar genera
are sufficiently striking. An entering wedge in the solution of the
problem has been driven home by the demonstration that the principal
zone of Echinosphaerites is, over wide areas, resting upon the Leray-
Black River. Here there is then a point of contact between the
Appalachian and interior provinces. Dr. Ulrich will agree to this,
but will include practically all of the limestone at Chambersburg,
for instance, with the Nidulites, Christiania, and Upper Echinosphae-
rites zones, in the Black River. To show that they represent the
Trenton is a difficult, perhaps at present, impossible task, but I shall
endeavor to present my reasons for so regarding them. ‘To do this,
I must start with the section in New York and proceed by a round-
about western route to reach eastern Pennyslvania.

TRENTON IN NEw YORK.

The type-locality is in New York State, at Trenton Falls. The
section at Trenton Falls is unsatisfactory, in that the formation is
not there exposed to a low enough level to show the formation upon
which the Trenton rests. But a few miles east of Trenton Falls, at
Rathbone Brook, is another section which supplements the one at

RAYMOND: CORRELATION OF THE ORDOVICIAN STRATA. 253

Trenton Falls, and zones 2 and 1 of the section below are to be under-
stood to be exposed only in small part at Trenton Falls, but com-
pletely at Rathbone Brook. The Utica is likewise not exposed directly
at Trenton Chasm, but at several places in the immediate vicinity.

Composite section (Raymond, 112). Zones 7 to 3 and upper part
of 2 exposed at Trenton Falls.

8. Thin-bedded black and brown carbonaceous shale with
Triarthrus becki, Cryptolithus tessellatus, Pleurograptus
linearis and many other graptolites. The contact with
the limestone below is sharply defined and there are no
transition beds. Utica shale (typical). Thickness
about 300 feet.

7. Light gray, coarse-grained lithified coquina in thick beds.

Rafinesquina deltoidea, Hormotoma trentonensis and other Feet.

fossils. 26
6. Thin-bedded blue limestone with shaly partings. Ra- ‘

finesquina deltordea the common characteristic fossil. 92

5. Thin-bedded blue limestone with thick shaly partings.
Prasopora simulatrix and other common Trenton fossils
abundant. 100

4. Thin- and thick-bedded limestone, dark in color and
fine-grained. Dziplograptus amplexicaulis a common

fossil. 35
3. Thin-bedded dark limestone with Triplecia extans and
other fossils. | 20

2. Thin-bedded dark limestone with some inter-bedded
coarse-grained layers. Cryptolithus tessellatus the char-

acteristic fossil. Trematis terminalis, Platystrophia -

trentonensis, Calymene senaria and many other fossils

present. 41
1. Thin-bedded gray limestone with an abundance of Dal-

manella rogata, and some other fossils. 32

The Leray-Black River is beneath 1. I wish to call especial atten-
tion to the fact that there are here two zones of Cryptolithus tessella-
tus; one in the forty-one feet of limestone near the base of the section
(this is the Glens Falls limestone) and one in the Utica shale.
Cryptolithus tessellatus is not a facies fossil, as its occurrence in this
section shows for it is in both dense fine-grained blue limestone and
coarsely crystalline gray limestone (coquina rock) in the lower zone,
and in the Utica it is in a very fine-grained carbonaceous shale. In
northern New York at the eastern end of Lake Ontario, it is found in

254. BULLETIN: MUSEUM OF COMPARATIVE ZOOLOGY.

green, yellow, and brown sandy and calcareous shale, and in sand-
stone in the Lorraine (Pulaski).

When the Trenton is traced northward from Trenton Falls it is
found well exposed along the tributaries entering the Black River
from the west. In this region the Cryptolithus bed is no longer seen
at the base of the formation, but the middle beds with Prasopora
simulatriz are the same as at Trenton Falls. At the top, however, a
considerable thickness of strata are added, of a kind not seen at
Trenton Falls, and containing a fauna not found there. Above thin-
bedded limestone containing Rafinisquina deltoidea, these beds being
the equivalent of the upper strata at Trenton Falls, one finds thick-
bedded impure dark gray limestone which on weathering breaks down
into a rubbly mass. This limestone contains many gastropods and
some other fossils, the most characteristic being Hormotoma trentonen-
sis, Fusispira subfusiformis, and Cyclospira bisulcata.

It is interesting to note that the species which are most characteristic
of the upper beds and most of which are not found at Trenton Falls,
were described originally from this region. Thus the type-localities
of Fusispira subfusiformis are Adams, Jefferson County, where only
these upper beds are exposed, and Turin, Lewis County. Trocho-
nema ambigua, Fusispira vittata, and Cyclospira bisulcata were all
described from specimens obtained at Adams, and Holopea paludini-
formis and Subulites elongatus were both found originally in the higher
Trenton strata at Watertown. Of all these species, only the last has
been found at Trenton Falls, and in New York they are characteristic
of strata above the strata exposed at Trenton Falls, and the highest
Trenton limestone exposed in the State.

ONTARIO.

Crossing into Ontario, the section is practically identical with that
in northern New York, except for certain loca] developments. The
strata in the middle of the Trenton remain the same as those at Tren-
ton Falls, but the lower beds do not carry Cryptolithus. The lower
beds do, however, show a decided recurrence of Black River conditions.
and faunas, as high as 100 feet above the base of the formation, so that
there is here practically continuous sedimentation after the Leray, the
Trenton fauna gradually replacing the Leray fauna. In this respect
the section reminds one greatly of Kentucky, as will be seen later.

The section, in descending order, is as follows, the section at Ottawa
being taken as typical of the region.

RAYMOND: CORRELATION OF THE ORDOVICIAN STRATA. 255

6. Dark brown to black carbonaceous shale with Triarthrus
spinosus, T. becki, T. glaber, Leptograptus annectans,
Glossograptus quadrimucronatus, and other graptolites.
Gloucester formation. (New Name). 50-75

5. Interstratified limestone and dark shale. Characteristic
fossils are Ogygites canadensis, Oxoplecia calhouni, Zygo-
spira uphami, Plectambonites rugosus, etc. Collingwood
formation. 25-30

4. Thick-bedded dark gray limestone with very little shale. ;
Characteristic fossils are Fusispira subfusiformis, F.
nobilis, and many other gastropods, Cyclospira bisul-
cata, Strophomena trilobata, and, in the lower part, Ra-
finesquina deltoidea. Picton formation. 100

3. Gray limestone, thin-bedded and with much interstrati-
fied shale in the lower twenty-five feet, less shaly but not
very thick-bedded above. “Prasopora beds” or true
Trenton. In the shale at the base Clitambonites ameri-
canus is the guide fossil, though many others are present.

In the strata a short distance above these a profusion of
echinoderms are found, among them being Pleurocystites
squamosus, P. filitextus, Agelacrinites inconditus, and
Comarocystites punctatus. Prasopora simulatrix is so
very abundant throughout these strata that I have
usually spoken of them as the Prasopora beds. 100

2. Coarse-grained light gray thick-bedded limestone,
thirty-three feet in thickness, resting upon sixty-six
feet of blue to gray fine- to coarse-grained limestone con-
taining great quantities of black chert in layers and
flattened cakes. The upper beds contain an abundance
of: Stromatocerium and Solenopora and in a nearby
locality, Tetradiwm racemosum. The.lower beds have
shaly partings in which great numbers of fine echinoderms
have been found, particularly crinoids. Among the
characteristic fossils are Edrioaster bigsbyi and Cleiocrinus
regius. Hull formation. 100

1. Thick-bedded dark gray limestone with partings of shaly
limestone containing numerous fossils, among them
Triplecia extans, Phragmolites compressus, Orthis dis-
paralis, Strophomena filitexta, and Receptaculites occi-
dentalis. This formation is not well exposed at Ottawa.
Rockland formation. 35

256 BULLETIN: MUSEUM OF COMPARATIVE ZOOLOGY.

It will be noted that in this section there is a formation added at
the top which is even younger than any found in New York and that
it is characterized by a fauna differing in a considerable number of
species from the fauna found below. The strata containing this
fauna are at the base, limestone, but above pass over into dark shale
which becomes finally typically “Utica” in facies. The transition
from the limestone of the Picton into the limestone of the Collingwood
is gradual and some of the species of the Picton continue into the
Collingwood; there is probably no break in sedimentation.

MINNESOTA.

These formations, or at least most of them, can be traced across
Ontario to the westward and in Minnesota the following section may
be seen (Winchell and Ulrich, 121).

Feet.
8. Massive dolomitic limestone with Maclurites and Mac-
lurina. Stewartville dolomite. 50
7. Fine grained to subcrystalline limestone with some
argillaceous layers in the upper portion. Among the
fossils are Rafinesquina deltoidea, Zygospira uphami,
Fusispira nobilis, Fusispira inflata, Cyclospira bisulcata,
etc. 56
6. Cherty limestone with Orthis tricenaria, Clitambonites
_ americanus, Parastrophia hemiplicata, ete. Clitambon-

ites bed. 9
5. Mostly thin-bedded limestone, argillaceous in the lower

portion and becoming more pure toward the top. 36

Zones 7 to 5 belong to the Prosser limestone. |
4. Blueshales with branching sponges. Fucoid bed. 6
3. Blue shales with Bryozoa and Ostracoda. Phylloporina

beds. 14
2. Shales with limonite, sometimes odlitic in structure.

Many Pelecypoda. Ctenodonta bed. 9
1. Dark green soft shale with numerous Bryozoa. Rhini-

dictya bed. 23

Zones 4 to 1 make up the Decorah shale.
There are two subdivisions of the Trenton strata in Minnesota
which may be correlated directly with formations in Ontario. These
are the upper strata of the Prosser limestone, (zone 7 of the section

a
> E
«
- ri
4
a

RAYMOND: CORRELATION OF THE ORDOVICIAN STRATA. 257

above), and the Clitambonites bed (zone 6). The zone 7, with its
large gastropods, particularly the Fusispiras and Trochonemas, Cyclo-
spira bisulcata, Rafinesquina deltoidea, and Strophomena trilobata, is
an exact counterpart of the Picton of Ontario.

The Clitambonites beds of both Ontario and Minnesota are char-
acterized by the same species of Clitambonites and Parastrophia, and
there are many more fossils common to the two.

The Decorah shales of Minnesota have been correlated by Dr.
Ulrich with the strata above the Leray-Black River and below the
cystid beds at Kirkfield, Ontario, and on this point we are in agree-
ment.

The Stewartville dolomite is not present in Ontario, nor have any

NORTHERN DOLGEVILLE =PATTERSONVILLE ALBANY
AY. MINN. ONTARIO if TRENTON CANAJOHARIE SCHENECTADY
| FALLS

STEW. ¥.
fetoey =)
CO UPPER PICTON Sse UTICA:
F EaiEEEEEREEEEEEEEEEEREREEEE EEE
te te es =f ——_—
Te Se == LOWER PICTON reo

Ht Ob rosser eagles ioe el 2S os es oe a Bd =a =
wit MORES Seer PRASOPORA BED Snsatoss CANAJOHARIE==

F 2 ak [=]

t= DECORAHL I 1

{el Catt a ae a imi ; C1
CURDSVILLE Foo ROCKLAND

Fig. 1.— Correlation of the sandy and shaly strata of eastern New York with the
calcareous strata of the more western localities. The strata represented by the
dots are predominantly sandy shales and sandstones, with subordinate amounts of
black shale. The next beds below are very fine-grained shales, usually very dark
in color, and the remaining strata, represented by the “‘brick”’ design, are prin-
cipally limestone. For Plat’ville, read Platteville.

of its characteristic fossils been found there. In Minnesota a part
of the Prosser fauna passes over into the Stewartville, and there does
not appear to have been any break in the sedimentary record, so that
there was apparently here a persistence of limestone deposition after
it had ceased in Ontario.

Reviewing what has been said of the preceding sections, it will be
seen how the shales in their westward progress transgressed higher and

258 BULLETIN: MUSEUM OF COMPARATIVE ZOOLOGY.

higher beds of the Trenton series and how, as time went on, their own
faunal content became changed, showing that it was not a sudden
migration of the sea over a tilted and partially eroded series of earlier
deposits, but that the near shore black shales were able constantly to

encroach upon the portion of the sea where limestone was forming,

till it finally progressed over the entire northern and eastern portion
of the sea.

If we now turr to the south, we find that in Kentucky and Tennessee,
clear water conditions prevailed throughout the Trenton and conse-
quently there were very different faunas, a fact best expressed in the
presence there throughout the Trenton of the corals Columnaria and
Tetradium.

KENTUCKY.

The following is a section in central Kentucky, after Foerste (102).

7. Granular limestone above, argillaceous limestone and Feet.

clay below. This formation is referred to the Eden by

the Kentucky geologists. Along the Ohio River oppo-

site Cincinnati it contains Cryptolithus in abundance

and in northern central Kentucky it contains a fauna

very closely allied to that of the Eden and Maysville.

Cynthiana formation. 40
6. Granular limestone in the upper five feet, nine feet of

dense white limestone below, and twenty to twenty-five

feet of fine-grained grayish limestone at the base. The

fauna is large, containing many gastropods and pelecy-

pods, several brachiopods, and two species of Tetradium,

T. columnare and T. fibratum. Perrysville formation. 30
5... Granular limestone above, with seven to ten feet of fine-

grained argillaceous limestone below. Columnaria alveo-

lata is present in the upper part, and the lower bed is the

one from which the Brachiospongia have been obtained.

It contains the oldest Platystrophia and Clitambonites

found in the Kentucky section. Flanagan formation. 60-70
4. Granular limestone with Stromatocerium. Benson for-
mation. 35

3. Argillaceous limestone with interbedded thin clayey
layers. Prasopora simulatrix, Rhynchotrema increbes-
cens, and Hebertella frankfortensis appear first at this

RAYMOND: CORRELATION OF THE ORDOVICIAN STRATA. 259

horizon, most of them passing up through the Perrys-

ville. Three species of Fusispira, and other gastropods

are present. Wilmore formation. 70
2. Fine-grained limestone alternating with clayey layers of

similar thickness. Heterorthis clytie, Dalmanella fertilis,

Leptaena tenucstriata, and Cryptolithus tessellatus are char-

acteristic species. Hermitage formation. 22
1. Granular limestone with cystids, Edrioaster bigsbyz, Orthis
tricenaria, Dinorthis pectinella, ete. Curdsville formation. 23

On a first survey of the lithological characteristics of the section
in central Kentucky, one is impressed by the large amount of light-
colored, fine-grained and coarse-grained rather pure limestone and
the lack of dark-colored shale.

Columnaria occurs in the Curdsville, Flanagan, and Cynthiana
formations, and Tetradium in the Hermitage, Perrysville, and Cyn-
thiana, so that the Wilmore and Benson are the only formations
without corals.

Because of the presence of species of Amygdalocystites, Pleuro-
cystites, and Edrioaster in the Curdsville in Kentucky it has become
the custom to correlate this zone with the cystid zone of Ontario.
Ulrich, and following him, Bassler, have correlated the Prosser of
Minnesota with the Curdsville of Kentucky, a correlation not borne
out, I think, by the evidence.

The Curdsville fauna of Kentucky is a pure derivative of the Black
River, only the echinoderm fauna being added to a rather typical
Leray-Black River assemblage of fossils.

In Minnesota the Pleurocystites occur in a very. different associa-
tion. They are found in the lower part of the Fusispira beds (zone 7
of the section above) where they are associated with Strophomena
trilobata, Cyclospira bisulcata, and Rafinesquina deltoidea, all Upper
Trenton species in New York and species which are never found so
low as the Black River. This zone is also above the Clitambonites
bed, which can not be correlated with anything older than the Wilmore
of Kentucky.

In Ontario there are three “Curdsville”’ zones, two above and one
below the bed which is correlated with the Clitambonites bed of
Minnesota. All are seen in the section at Ottawa, where the zones
are separated by seventy-five feet of strata containing two distinct
faunules.

The lower zone, to which I have given the formation name Hull, is

262 BULLETIN: MUSEUM OF COMPARATIVE ZOOLOGY.

County. Also in the neighborhood of Amsterdam. I have better
collections from near Lake Ontario, on Sandy Creek, near Ellisburgh,
Jefferson County.”

Since receiving this letter I have visited all these localities, the first
two of which were already familiar to me from previous work. There
can be no doubt that the last locality contains the typical Fusispira
fauna, for the strata there are of the very highest Trenton, just be-
neath the Utica shale, and it is in the immediate vicinity of the places
from which the principal species of the Fusispira fauna were originally
described. The other two localities are located where only basal
Trenton strata are exposed. “The vicinity of Poland’ probably
means the exposures between Poland and Newport, the localities of
the well-known Rathbone Brook section and the “Moshier quarry”
in the Leray-Black River. The lower part of the Trenton (Crypto-
lithus beds) in this section contains some layers with gastropods, but
I have found Fusispira here only in the Leray-Black River where it is
associated with a number of other gastropods in a large fauna (107).

At Amsterdam the section is practically the same as at Poland.
Only the lower part of the Trenton is exposed, and it rests upon a very
small thickness of the Leray-Black River, which formation is not very
fossiliferous. If any members of the Fusispira fauna are found here,
they must be in a very different association from that in Minnesota,
and are certainly at a much lower horizon.

The last place mentioned by Ulrich, is in New Jersey, at Jackson-
burg, where the formations present are equivalent to those at Amster-
dam and Poland in New York and the same remarks will apply to it.
Personally I am unable to see in the faunas of any of these localities
anything to suggest the Fusispira fauna of Minnesota and even if the
Fusispira fauna be there, the section in New York affords ample proof
that the Lower Trenton beds with Cryptolithus are below the typical
Trenton of Trenton Falls and the Fusispira fauna occurs in the Upper
Trenton beds, above the typical Trenton of the Trenton Falls section.

The above discussion is necessary to justify my position in placing
the Prosser very much higher in the section than it is placed by other
writers.

CENTRAL PENNSYLVANIA.

The section in central Pennsylvania at Bellefonte has been described
in outline by Professor Collie (100), and has lately been reinvestigated
at my suggestion by Mr. R. M. Field, in whose company I was able

RAYMOND: CORRELATION OF THE ORDOVICIAN STRATA. 263

to spend a few days on the work. The study of the collections is still
incomplete, and considerable field-work remains to be done before any
satisfactory correlation can be made, but the preliminary results seem
to indicate a correlation with the formations in Kentucky rather than
with those in New York. The important fact, mentioned before, of
the identification of the fauna of the Leray-Black River in the strata
just above the pure quarry rock gives a datum plane for correlation.
Above this zone, but still in the strata called “ Black River” by Collie,
is found the fauna with Echinosphaerites, and in the lowest “Trenton”
(A 8 of Collie), one finds fossils such as Cryptolithus tessellatus, Orthis
tricenaria, Dinorthis, and others which suggest the Hermitage of
Kentucky. The “Trenton” has a thickness of 600 feet, and at the
top, Cryptolithus comes in a second time, as in the Cynthiana of
Kertucky. In our present state of knowledge, it must be confessed
_ that the thickness of the section, the general likeness of the faunas,
and the two occurrences of Cryptolithus, are the principal bases on
which the correlation is made. So far the upper Echinosphaerites
zone with Christiania has not been found at Bellefonte, but the general
similarities of the sections at Chambersburg and Bellefonte are such
that we fully expect work in the intervening areas to establish a fairly
secure correlation between them.

GENERAL DISCUSSION OF RUSSIAN EARLY ORDOVICIAN
FAUNAS.

The correlations attempted above have been based largely upon
stratigraphic position and “guide fossils.”” It remains to compare
in a little more detail the faunas of the strata which have been corre-
lated.

WaALCHOW AND KUNDA FORMATIONS.

The part of the correlation which I must most justify to American
students is, probably, that in which I agree with most Continental
authors, in assigning the Walchow and Kunda formations, B,, and
B,,,, to the horizon of the American Beekmantown. Therefore, I
wish, in addition to what has already been brought out, to discuss the
faunas of these two formations in some detail. Bassler has correlated
the two formations, in a general way, with the lower part of the Black

'

264 BULLETIN: MUSEUM OF COMPARATIVE ZOOLOGY.

River, and it has been by no means unusual, even among European
text-book writers, to consider the upper formation (Orthoceras lime-
stone) as of Black River age, this being primarily due to the abundance
of large cephalopods and of trilobites in both the Orthoceras limestone
and the Black River limestone. Bassler’s argument seems to be based
largely upon the state of development rather than on identity of
species of the bryozoan fauna, for of eleven species in the Glauconite
limestone and fourteen in the Orthoceras limestone, only two in each
are identified by him with American species.

Lamansky has listed 142 species and varieties of fossils from these
two formations, and Schmidt and Bassler have since added enough
more to bring the number to about 186. Of these, seventy-seven are
trilobites, forty-five brachiopods, nineteen cystids, twenty-six bryo-
zoans, eleven cephalopods, four pteropods, and four gastropods.

Trilobites.

Eighteen genera of trilobites are listed, and of these, most of the
prominent ones, namely, Asaphus, Onchometopus,! Niobe, Pseuda-
saphus, Ptychopyge, Cyrtometopus, Cybele, Pliomera, and Plato-
polichas, do not occur in America. Of the remaining nine genera,
Nileus and Eoharpes are found in the typical Beekmantown, Megalas-
pis may occur in the Beekmantown, provided the few American species
which have been referred to that genus really belong,” and Illaenus is
common in the Beekmantown, while the other genera, Ceraurus,
Remopleurides, Lichas, Pterygometopus and Ampyx make their
first appearance in American faunas in the Chazy or later formations.
To these five genera belong eighteen species, mostly rare trilobites,
and of the eighteen species, eight are confined to the upper of the two
formations.

These figures indicate very clearly the total unlikeness of the trilo-
bite faunas of the Beekmantown of America and the Walchow and
Kunda of Russia. Of eighteen genera only four are common to the
two, nine are entirely unknown in America, and four make their
first American appearance in the Chazy. In America, Lichas appears
first in the Silurian.

1The American species referred to this genus by the writer must probably be included
with Brachyaspis. ‘
2 Megalaspis beckeri Slocum is almost certainly not a Megalaspis.

RAYMOND: CORRELATION OF THE ORDOVICIAN STRATA. 265

Brachiopoda.

The Brachiopoda are listed by Lamansky under twelve generic
names, but here comparisons are less satisfactory as the species have
not been studied critically. Orthis as used in this list includes Orthis
s. s.,and Dalmanella, and should include Platystrophia, two species of
which occur, but are not listed by Lamansky. Acritis should also be
added. This increases the list to fifteen genera, two of which we may
at once drop, Leptaena as being meaningless in the present state of
our knowledge of the three species referred to it, and Lingula as being
_ cosmopolitan. Of the thirteen genera then remaining, eight, Poram-
bonites, Lycophoria, Plectella, Pseudocrania, Acritis, Pseudometop-
toma, Philhedra, and Siphonotreta, are unknown in the Ordovician
of America. Orthis is known from the American Beekmantown,
Dalmanella is probably there, though doubts have been cast on some
of the species so referred, and Strophomena may be there, but the
reported cases are questioned. Clitambonites appears first in the
Chazy, and Platystrophia in the Trenton. In passing, it may be said
that Orthis obtusa Pander, which is very abundant, belongs to an un-
described genus, unknown in America, and that Orthis parva Pander,
which Wysogorski (57) states can not be a Dalmanella because im-
punctate, is in reality exceedingly punctate. Orthis is very common
and exceedingly variable in these deposits, but all the species agree
in having a much lower cardinal area and a much wider delthyrium
than the species which we in America know as a typical Orthis (p. ex.
Orthis tricenaria). Orthis panderiana Hall and Clarke, of our Beek-
mantown, is much more like the typical Orthis of the Walchow.

Bryozoans.

As previously stated, Bassler describes twenty-six species and
varieties from these two formations, four species and one variety being
identified as common to Russian and American deposits. <Arthro-
clema armatum is said to be common to the Walchow and to the
Nematopora and Fusispira beds of Minnesota (Upper Trenton).
Dianulites petropolitanus, which in Russia ranges from the Walchow to
the Wesenberg, is also identified in the same upper Trenton strata in
Minnesota. Batostoma fertile and its variety circulare are said to be
common to the Kunda formation of Russia and the Stictoporella bed
(Black River) of Minnesota. Hemiphragma irrasum, found in the

266 BULLETIN: MUSEUM OF COMPARATIVE ZOOLOGY.

Stictoporella, Rhinidictya, Phylloporina, and Clitambonites beds
(Black River to Middle Trenton) of Minnesota, was identified from
the Kunda of Russia. These forms are evidently of little value in
direct detailed correlation, since the two species found in B,, would
correlate that formation with the Upper Trenton, while the two found
in B,,, correlate that formation with the Black River and Lower
Trenton. One of the species has a range equal to almost the whole
Ordovician of Russia, and another has a very long range in America,
so that no very valuable conclusions can be derived from them.

In considering the Bryozoa, it must be remembered that as yet only
a few specimens belonging to the genus Nicholsonella, have been
found in the American Beekmantown, and that the fairly large bryozoan
fauna of the Chazy is as yet undescribed. The comparison of the range
of genera in America and Russia does not, therefore, mean much, for
many of the American genera now supposed to start in the Black River
will probably be found to have their beginnings in the Chazy. There
are, however, one or two interesting points to be noted in this connec-
tion. Bassler distributes the twenty-six species which he describes
under eighteen genera (compare with seventy-seven species of trilobites
in eighteen genera, and forty-five species of brachiopods in fifteen
genera). Of these eighteen genera, only three are not found in Amer-
ica (compare with eight out of eighteen in the trilobites and eight of
fifteen in the brachiopods). Three more are very peculiar, in that
their American range begins much later than in Russia. One of them
is known in this, country from the Richmond to the Mississippian,
another from the Niagara to the Coal Measures, and the third is
wholly Devonian and Mississippian. A single one is found in the
American Beekmantown, and the remaining eleven are known from
the Stones River or Black River to the Richmond, excepting for one
or two which do not start till the Middle or Upper Trenton. Looking
over the table showing the range of the various species in Russia, we
find that four species are confined to B,,, one is confined to B,, and
B,,, nine are found only in B,,,, eight pass from B into C, but do not
extend further, and that three begin in B and continue into D, E, or F.
Thus, among the Bryozoa, there are only five species common to B,,
and B,,,, and of these only one does not continue into C, while there
are eleven common to B and C, which is very unlike the condition
which obtains among the other classes. For instance, in the seventy-
seven species of trilobites, nineteen are common to B,, and B,,,, and
only thirteen pass from B into C, and some of these cases must be
considered as doubtful, since they come from that district on the

RAYMOND: CORRELATION OF THE ORDOVICIAN STRATA. 267

Walchow where the boundary between B,,, and C, is indefinite.
Still it is evident that thirteen out of seventy-seven is a much smaller
proportion than eleven out of twenty-six. The only conclusion which
it is possible to draw from the above rather remarkable array of evi-
dence derived from the bryozoans is that both the species and genera
have too great a vertical range to allow their use in direct correlations.

Cystidea.

Lamansky lists fifteen species of cystids, belonging to six genera (if
Bolboporites can be called a genus), but he places no particular specific
names after Cheirocrinus. Asa matter of fact there are a considerable
number of species of cystids not enumerated by Lamansky, but that
does not affect the present purpose.

Of the six genera listed, three are unknown in America. Of the
remaining three, Glyptocystites and Bolboporites appear in the Chazy,
and Cheirocrinus in the Trenton. As with the Bryozoa, it must be
remembered that the cystidean fauna of the Beekmantown is unknown.
There is plenty of proof that cystideans were present, but most of our
Beekmantown rocks are lithologically ill adapted either for the preser-
vation or recovery of fossils.

Cephalopods.

Vaginoceras is of course the common genus in the Kunda, and that
genus is represented in America by a single species found in the Chazy.
Other genera, like Estonioceras and Planctoceras are unknown here.

Gastropods.

Gastropods make their first appearance in Russia in the Kunda,
and the variety there issmall. Of the four genera, two, Maclurites and
Sinuites, are found in the Beekmantown, while Raphistoma appears
first in the Chazy and Salpingostoma in the Stones River.

The fauna as a whole.

Bringing together what has been said above, it appears that out of
the 186 species considered, five, all Bryozoa, have been considered as
identical] with species found in America, and these appear in the two
countries in reverse order, and thus have no significance. Of the

268 BULLETIN: MUSEUM OF COMPARATIVE ZOOLOGY.

genera, including organisms of all kinds except the Bryozoa, about
half are not found in America. Of the genera common to the two
countries, a few make their first appearance in America in the Beekman-
town, a much larger proportion first appear in the Chazy or Stones
River, and a few do not appear in America till Middle or late Trenton
or even later. It is pointed out, however, that even in the case of
certain of the genera reputed to be in both countries, the Russian
representatives differ in marked ways from American species, and
closer research is bound to show greater differences instead of greater
likeness between the faunas of the two countries.

The fauna of the American Beekmantown is very imperfectly known,
but the classes of fossils so far as relative abundance is concerned, are
ranked in the order of: — first, cephalopods; second, gastropods; third,
trilobites; fourth, ostracods; and finally brachiopods, cystids, and
bryozoans, all in small numbers. On the other hand, the Walchow
and Kunda formations have great numbers of trilobites and brachio-
pods, many cystids, cephalopods, and bryozoans, very few gastropods,
a few ostracods, and crinoids. The two groups agree in the absence
of pelecypods and corals.

If there were no other evidence than that afforded by the time of
the first appearance of certain genera in Russia and America, it might
well happen that the Walchow and Kunda formations might be corre-
lated with the Chazy, but I do not see that they could be correlated
with any younger strata. The comparison of the Russian with the
Scandinavian sections, however, places such a correlation out of the
question, and when one compares the faunas of these zones with the
fauna of the Ceratopyge formation of Scandinavia, he realizes the
antiquity of many of the groups, especially of the trilobites. Mega-
laspis, which is almost entirely confined to the Walchow and Kunda
in Russia, is well represented in the Ceratopyge limestone of Scan-
dinavia, one species being apparently common to the two formations
and regions. Symphysurus, Nileus, Niobe, Eoharpes, and Ampyx
are other genera connecting these formations with the Ceratopyge
limestone, and even species of Nileus and Niobe are said to be com-
mon to the two.

On a priori grounds, we would not expect the Walchow and Kunda
faunas to have much in common with the Beekmantown, if they
are of the same age. The American fauna is an autochthonous one
developed in the interior continental sea, out of the late Cambrian
fauna. The fauna in Esthonia was, on the other hand, an invading
fauna which was derived primarily from the Ceratopyge fauna of

:

RAYMOND: CORRELATION OF THE ORDOVICIAN STRATA. 269

Scandinavia, but that in turn though it received certain contributions
from the late Cambrian faunas of the region in which it developed,
owed much of its richness to types developed further south in the
Tremadoc of Bohemia, France, and England. Having once gotten a
foothold in northern Europe, the fauna developed very rapidly there,
but apparently in an enclosed basin, for this fauna, as a whole, is
unknown outside Scandinavia and Russia. Here, however, the factor
of bottom control must be taken into account. As we have shown
(p. 222), it is generally recognized that the black shales with the
Didymograptus-Tetragraptus-Phyllograptus fauna were deposited
at the same time as the “Orthoceras limestone,” and in the same
sea, but under different physical conditions. As is well known,
the graptolite faunas did migrate, and very widely, but they did not
carry the bottom fauna with them. If we adopt the rather generally
accepted opinion that the graptolites were pelagic animals, supported
either by floats or by their attachment to floating bodies such as sea
weeds, we may conceive that the graptolites may have been distrib-
uted within a very short time, by the power of ocean currents, over
very wide areas, while the influence of a strong current or of a cold
current, impinging upon headlands, or the presence of vast expanses
of sandy or muddy bottoms, may have long delayed the migration of
bottom-living animals. We seem to have an excellent example of
this in the case of Shumardia and certain associated species of the
Ceratopyge limestone. In Sweden and Norway, Shumardia is rather
abundant in the shale and limestone making up the Ceratopyge zone,
and this zone is above the shale with Dictyonema flabelliforme, but
below that of Tetragraptus and Phyllograptus. The Shumardia
limestone of America, however, (at Point Lévis) is very high in the
Tetragraptus-Phyllograptus series, so high even as the beginning of
the range of Diplograptus. This case is the more striking since there
are several species of the Scandinavian Shumardia-Ceratopyge fauna
(Shumardia pusilla (Sars), Agnostus sidenbladhi Linnarsson, and Sym-
physurus elongatus Moberg and Segerberg) in this limestone high in
the Lévis.

Under these conditions, if these species of the Ceratopyge fauna
could not arrive in America until late Beekmantown, it is not surprising
that many genera which originated in Europe during Beekmantown
time, should not have arrived in America till the Chazy. I do not
wish to be understood to advocate a general principle that homotaxial
formations of separated continents are really one stage apart in age,
but each particular case must be decided on its own merits.

270 BULLETIN: MUSEUM OF COMPARATIVE ZOOLOGY.

WIERLAND GROUP.

This series, held together by the presence throughout of Echino-
sphaerites aurantium, shows a decided change of faunas from bottom to
top. The fauna of the oldest formation, the Dubowiki, is dominated
by its asaphids, but these trilobites immediately lose their importance,
and though they continue through the remainder of the Ordovician,
they are present in limited variety and numbers. As previously
stated, there are thirteen species which pass over from the Walchow
and Kunda into the Dubowiki, but very few of them survive beyond
this formation.

The Wierland fauna is, however, in general, an outgrowth of that
which preceded it in the same area. Among the trilobites, notable
new arrivals in this group are: — Chasmops, Sphaerocoryphe, Pseudo-
sphaerexochus, Sphaerexochus, Acidaspis, Hoplolichas, Homolichas,
Cyphaspis, Lonchodomas, and Ogygites (Basilicus of Schmidt). All
of these genera, with the exception of Acidaspis and Cyphaspis could
have arisen as modifications of types already in this region, so that we
have, as true invaders only these two genera.

Among the brachiopods the important new genera are Christiania
and Oxoplecia. The place of origin of these genera is unknown, but
from their short ranges and limited variety in Russia, it seems probable
that they are present there as migrants. Other brachipods intro-
duced here are Plectambonites and Rafinesquina, but the dominant
forms are the Clitambonites and Porambonites which continue from
the formations below.

Among the gastropods, Bucania, Cymbularia, Eccyliopterus,
Salpingostoma, and Subulites make their appearance first in this
formation; while of the echinoderms, we find here the oldest species
of Caryocystites, Echinosphaerites, Cryptocrinites, Cystoblastus,
Cyathocystis, Hemicosmites, Hybocrinus, and Protocrinites. Among
the sponges, Receptaculites is introduced at this time.

In running over this list of genera newly introduced into the Russian
Ordovician during Wierland time, we are struck by the fact that we
are here dealing with more familiar genera. Barring some of the
cystids and one or two other genera, all these genera are known in
America, and most of them from the Middle Ordovician. The greater
part of these genera seem to have developed in the North European
basin and to have migrated thence to America.

Certain of the genera are, in America, restricted to a belt along the

ss
—————— elhlh..LlUc

RAYMOND: CORRELATION OF THE ORDOVICIAN STRATA. 271

eastern Appalachians and to a small area in western Tennessee and
Missouri (Kimmswick of Missouri), and do not occur in the usual
‘ Ordovician fauna of the great interior basin. The eastern Appalachian
belt extends from Gaspé to Alabama, and includes the “Trenton” at
Percé, the Quebec City at Quebec, the “Trenton” of southeastern
Quebec, the Chazy and Rysedorph of New York, the Chambersburg
of Pennsylvania, the Liberty Hall and Murat of central Virginia, the
Holston, Ottosee, Lenoir, and Athens of southwestern Virginia and
eastern Tennessee, and the “Trenton” of Alabama. This series
includes formations of various ages from Chazy to Upper Trenton.

Russian genera which in America are restricted to the strata men-
tioned above are Sphaerexochus, Lonchodomas, Christiania, Oxo-
plecia, and Echinosphaerites. Genera prominent in this group, but
only sparingly represented in the interior Ordovician are, Pseudo-
sphaerexochus, Remopleurides, and Sphaerocoryphe.

Of the other genera introduced during Wierland time, Ogygites does
not appear in America till late Trenton (Collingwood), Cymbularia is
unknown, as are also Cryptocrinites, Cystoblastus, Cyathocystis,
Hemicosmites, and Protocrinites. The other genera are more or less
common throughout the Middle Ordovician faunas of America.

Some of the more striking of the Wierland guide fossils are absent
from the Chazy, thus preventing what seems otherwise a very satis-
factory direct correlation. These are Echinosphaerites, Christiania,
and Oxoplecia. The fauna of the Chazy is, in fact, a curious mixture
of native and immigrant ‘types. All its large molluscan fauna is
probably derived directly from the Beekmantown fauna, and most of
its brachiopods are also American in origin. There is certainly nothing
in northern Europe like its great profusion of rhynchonelloids, and its
Orthidae and Strophomenidae may as well be native types as invaders.
Even Camarella, the Chazy representative of the Porambonitidae, is
probably of American stock, and does not really represent the Russian
Porambonites. In Clitambonites, however, we have a true immigrant,
which did very well for a time.

When we come to the trilobites, however, we begin to see the in-
vaders. In the list we notice not only certain native genera, Bathyu-
rellus, Isotelus, Isoteloides, Thaleops, and Glaphurus, but also many
others, which are actually Russian or derived from Russian stocks.
These are, Russian: — Eoharpes, Lonchodomas, Remopleurides,
Nileus, *Ceraurus, Pseudosphaerexochus, Nieszkowskia, Sphaero-
coryphe, Sphaerexochus, and Pterygometopus; derived from Russian
stock: — Cybeloides, Pliomerops, Vogdesia, and Heliomera.

PIM: | ae

Qe BULLETIN: MUSEUM OF COMPARATIVE ZOOLOGY.

These genera may be divided into two groups, first those appearing
first in the Walchow and Kunda in Russia, and second, those making
their first appearance in the Wierland in Russia. In the first group we
find Eoharpes (also in Beekmantown), Remopleurides, Nileus,
Ceraurus, Nieszkowskia, and Pterygometopus. In the second group
are only Lonchodomas, Pseudosphaerexochus, Sphaerexochus, and
Sphaerocoryphe. Cybeloides, Pliomerops, Vogdesia, and Heliomera
are all derived from forms appearing first in the lower group in Russia.

From this analysis of the Chazy fauna it appears that, while the
fauna is in very large proportion of American origin, it has present in
it a considerable Russian element which is largely derived from genera
present in the Walchow and Kunda formations. Why certain genera
such as Echinosphaerites, Oxoplecia, Plectambonites, and Christiania
from Russia, and Tretaspis, Nidulites, and Agnostus from Scandi-
navia, which reached North America somewhere about this time, did
not get into the typical Chazy is a perplexing question. So far as
bottom control is concerned, there seems to have been no barrier, and
only two possible explanations occur, namely, either their natural
habitat was fully occupied, or they did not reach America till later
than the time of the typical Chazy; the latter explanation seems, in
view of all the facts, the more probable.

DETAILED SECTIONS IN RUSSIA.
Sections arranged in order from east to west.

SECTION ON THE WatcHow. (Given by Lamansky (29)).

Bi, Thick-bedded compact limestone at Sapolek and

Bylstchina, and at the base of the quarries at St. Mich-

ael Archangel. 6 meters = 18.35 m.
Biss. Rusty and spotted limestone at the base, followed

by one or more layers of “Linsenschicht,”’ these being

succeeded by red and yellow spotted layers. The Linsen-

schicht is one meter above the base. 3.5m. = 12.35 m.
Bia. Bluish green limestone with fine-grained glauconite at

the base, but none above. 3m. = 8.350%
B,,,- Compact limestone with some glauconite. Loses

compactness quickly on weathering. 2.4 to 2.7 m. = 5.85 m.

RAYMOND: CORRELATION OF THE ORDOVICIAN STRATA. 273

Big. Thin-bedded limestone with red and yellow spots.

Very little glauconite. 1.8 m. = 3.45 m.
Biz. Thick-bedded, red, yellow, violet, and gray-green
limestone. 1.65 to 1.80 m. = 1.65 m.

SECTION ON THE WatcHow. (Measured by the writer).

The highest beds are exposed in the quarries at Dubowiki. Zones
11 to 9 were measured near the steamboat landing at St. Michael
Archangel, and the remainder of the section on the eastern bank of
the Walchow opposite Iswos.

11. Light gray to yellowish magnesian limestone in beds
: one inch to one foot in thickness. Some layers are fine
grained and dense, while others are porous, the fossils
occurring as hollow moulds. Christiania oblonga com-
mon and characteristic. Reval formation. 25 ft. = 112 ft.+
10. Soft gray limestone which breaks down readily to a
sticky calcareous mud. Entire trilobites and other
perfect fossils common. Echinosphaerites aurantium

abundant. :
Dubowiki formation. Base not exposed. 14 ft. = 87 ft.+
9. Concealed to river level. 23 ft. = 73 ft.

(According to the thickness given by Lamansky the
greater part of the concealed interval here would be filled
by the “Orthoceras” limestone (Kunda formation)).
8. Red and green limestone enclosing the “Lower Linsen-
schicht.” 3 ft. = 50 ft.
7. Thin-bedded limestone and shale, with 3 feet of rather
solid limestone at the top. Byys (Bie of Lamansky).

9 ft. = 47 ft.
6. Thick-bedded gray limestone with shaly partings.
(Bix) 10 ft. 6 in. = 38 ft.
5. Thin-bedded, shaly limestone. (By,,). 5 ft. 6in. = 27 ft. 6 in.
4. Heavy-bedded red limestone with abundant glauconite
and echinoderm fragments. (By,.). 6. ft. i=/22 ft.
3. Sandy shale. 1 ft. = 16 ft.
) Zones 3-7, and the lower part of zone 8, below the
Linsenschicht, belong to the Walchow formation.
2.. Concealed. 3 ft. = 15 ft.
1. White and gray sandstone to the river’s edge. Zones 1
and 2 belong to the Packerort formation. 12 ft.= 12 ft.

274

BULLETIN: MUSEUM OF COMPARATIVE ZOOLOGY.

River Lawa at WassitKowa. (Measured by the writer).

Thin, irregularly bedded, gray limestone and shale.

20 ft. = 72 ft.
Bright red and green variegated limestone in thin beds.
Byis- (Birra (partim) of Lamansky).: Aft; 1.52 Ft.
Thick-bedded light, blue-gray limestone. (B,,,). 11 ft. = 48 ft.
Thin-bedded, irregular, blue limestone with thick shaly

partings. (By). 13 ff, 2 a ae
Heavy-bedded red and green vetane with few fossils.

(Be) 6 Tt; = 22 O.
Green sand and clay. (Glauconite sand). 6 ft..= 18 ft.

Zones 4 to 8 and the lower part of 9 make up the
Walchow formation. The “Linsenschicht”’ is present in
9, but unfortunately I did not note its exact position.
Black shale. (Dictyonema shale). 1 Ft 12 Ft.
White, gray, and yellow sand, very irregularly bedded.
It crumbles readily at the touch and contains numerous
specimens of Obolus apollinis along the planes of cross-
bedding. Zones 2 and 3 represent the Packerort forma-
tion. 10 ft.
Very hard red sandstone, base not seen. 1 ft.

PaPowKA, SOUTH OF PrTroGRAD. (Measured by the writer).
A

=

ole

Thick-bedded, gray, fine-grained limestone with numer-
ous Orthoceratites. It contains two red layers, each one
foot in thickness. 18 ft. = 57 ft.
Red limestone full of Orthoceratites. 1 ft. = 39 ft.
Heavy-bedded dolomitic limestone which contains cavi-
ties from which fossils have been dissolved. Ortho-
ceratites in top foot. 7 ft. = 38 ft.
Shaly limestone with “linsen.”’ 5in. = 31 ft.
Zones 6-9 represent the Kunda formation.
Thin-bedded limestone with much clay and shale.

12 ft. = 30 ft. 6 in.
Thick-bedded green and red limestone. 7 ft. = 18 ft. 6 in.
Green sandy shale. 1 ft. = 11-4. 6 m.
Zones 3-5 represent the Walchow formation.
Black shale with only microscopic fossils. 4 ft. 6 in. = 10 ft. 6 in.

RAYMOND: CORRELATION OF THE ORDOVICIAN STRATA. 275

1. Green, white, and red sandstone with numerous frag-
ments of Obolus. Base not seen. 6 ft. = 6 ft.
Zones 1 and 2 represent the Packerort formation.

SEcTION AT Narwa. (Given by Schmidt (47) in the Guide for the
excursions of the St. Petersburg meeting of the International Geo-
logical Congress).

Cig. Echinosphaeritenkalk. Dolomite. 3.0m. = 16.9 m.
C;,. Upper Linsenschicht. 3m. = 13.9 m.
Big. Vaginoceras limestone. Dolomite. 3.0m. = 13.6 m.
Birra. Lower Linsenschicht. 3m. = 10.6 m.
B,,. Glauconite limestone. 3.3:m. = 10:3.m.
B,. Glauconite sandstone with concretions of bituminous
limestone with Dictyonema. 2m.= 7. m.

Ay. Red Obolus sandstone showing cross-bedding. 2.6 m. = 6.8 m.
A,,. White fucoidal sandstone with sand concretions. ;
4.2m.= 4.2m.

Narwa. (Measured by the writer). Plate 6.

8. Gray limestone, dolomitic, heavy-bedded, and with few
fossils at the top, thin-bedded and with numerous

Orthoceratites in the lower portion. 10 ft. = 36 ft. 10 in.
7. Linsenschicht. 15 in. = 26 ft. 10 in.
Zones 7 and 8 represent the Kunda formation.
6. Hard thick-bedded gray limestone. 3 ft! =;25 ft 7.in.

5. Impure red limestone with numerous M. planilimbata.
Top layer a much weathered rusty one, full of holes.
Sf. Ze hi 7 in.
4. Red clay-shale. 3in. = 14 ft. 7 in.
3. Green clayey sand. 4in. = 14ft.4in.
Zones 3-6 represent the Walchow formation.
2. Yellow, white and reddish massive and cross-bedded
sandstone with very numerous Obolus. At the base is
a layer of rounded, hard, sandstone pebbles, three, four,
and ten inches in diameter. Much pyrites in lower two
or three feet, sometimes in hard layers. 11 ft. = 14 ft.
Zone 2 represents the Packerort formation.
1. Almost white, thin-bedded sandstone. Base not seen.
3 ft. = 3 ft.
Zone 1 represents the Esthonia formation.

276 BULLETIN: MUSEUM OF COMPARATIVE ZOOLOGY.

ASSERIEN. (Zones 16 to 13 are well shown in the quarries and along
the railroad about two miles south of the cement plant at Asserien.
Zones 13 to 1 are well exposed at and below the large quarry at the
edge of the cliff about a mile west of the plant. Measured by Dr.
Twenhofel and the writer).

16. Fine-grained magnesian limestone in layers one to threé

inches thick. Very few fossils. 6 ft..=| 764. Lim
15. Thick-bedded, fine-grained compact magnesian lime-

stone with an abundance of Orthoceratites in the lower

three feet. Some layers are porous and show open moulds

of fossils. 8 ft. = 70 ft. 1 in.

Zones 15 and 16 belong to the Reval formation.
14. Thick- and thin-bedded gray limestone without much

shale. Echinosphaerites especially abundant in the

lower part. 7 ft. 6 in, = 62%t. 1 in,
13. Light gray, rather thick-bedded limestone without

shale, but becoming soft and shaly on weathering. This

zone contains more or Jess “linsen” throughout, but

they are especially abundant in the lower three feet.

8 ft. = 54 ft. 7 in.

Zones 13 and 14 make up the Dubowiki formation.
12. Thick- and thin-bedded gray limestone which becomes

soft and shaly on weathering. Many Vaginoceras

present. 14ft. 8in. = 46 ft. 7 in.
11. Lower “Linsenschicht.” 9 in. = 3P4te/11 ip.
Zones 11 and 12 represent the Kunda formation.
10. Thick-bedded gray limestone. 3 ft. 6 in. = 32 ft. 2 in.

9. Thin-bedded limestone and shale. 6in. = 27 ft. 8 in.
8. Thick-bedded, hard, reddish limestone with glauconite.
3 ft. 6in. = 27 ft. 2 in.

7. Clay and limestone. 8 in. = 23 ft. 8 in.
6. Green clay and sandstone. 6 ft. 3in. = 23 ft. Oin.
Zones 6 to 10 form the Walchow formation.
5. Thin-bedded, soft, yellowish clay-shale. 4 ft. = 16 ft. 9 in.
4. Red and yellow limonite. 3 in. = 12 ft. 9 in.
3. Thin-bedded black shale with thin layers of sandstone
in the lower half. 7 ft: =[12 £626 in.

2. Light yellow and red cross-bedded sandstone, with
lenses of thin, black shale in layers one fourth to one inch
in thickness and with black deposits on the bedding
planes. 3ft.3in. = 5 ft. 6 in.

RAYMOND: CORRELATION OF THE ORDOVICIAN STRATA. 277

1. Yellow and white sandstone with numerous Obolus,
one layer. 2 ft.3 in. = 2 ft. 3 in.
Zones 1-5 represent the Packerort formation.
Concealed to water-level.

OntiKA. (Top 206 ft. above sea-level. Measured by Dr. Twen-
hofel and the writer).

11. Thin-bedded soft limestone with Echinosphaerites aurantium.
| 5 ft. = 78 ft. 5 in.

10. Thin-bedded light gray limestone with numerous
Orthoceratites and many small “linsen.” Upper
Linsenschicht. 7 ft.6in. = 73 ft. 5 in.
Zones 10 and 11 represent the Dubowiki formation.

9. Thin-bedded light gray shaly limestone with numerous
Orthoceratites. A nine inch shaly parting is present

four ft. above the base. 18 ft. 6 in. = 65 ft. 11 in.
8. Lower Linsenschicht. 8 in. = 47 ft. 5 in.
Zones 8 and 9 represent the Kunda formation.
7. Heavy-bedded gray limestone with nine inches of shale
at top. 5 ft. = 46 ft. 9 in.
6. Thin-bedded nodular and shaly limestone 6 in. = 41 ft. 9 in.
5. Heavy-bedded reddish limestone with green patches,
weathering yellowish. 4ft.9in. = 41 ft. 3 in.
4. Green sandy shale. , 4 ft. = 36 ft. 6 in.
Zones 4 to 7 represent the Walchow formation.
3. Hard red limonite layer. 3 in. = 32 ft. 6 in.
2. Black shale, no fossils seen. 44iti Gin. = /32i8 3 mm:
1. Cross-bedded reddish and white sandstone which dis-
integrates to loose sand. Base not seen. 25 ft. = 25 ft.

Zones 1 to 3 represent the Packerort formation.
Concealed to water, 127 ft.

CATHERINE Park, Revat. (Measured by the writer).

7 Thin-bedded often shaly limestone. The highest layers
in the quarries. Echinosphaerites and other cystids
. common. Kuckers formation. 4 ft. = 47 ft.
6. Light gray compact fine-grained limestone in beds 8 to
15 inches in thickness. Fossils are very rare. Certain
layers contain vertical, tube-like fillings suggesting
borings. Largely quarried. 28 ft. = 43 ft,

278 BULLETIN: MUSEUM OF COMPARATIVE ZOOLOGY.

5. Upper Linsenschicht. 1 ft. = 15 ft.
Zones 5 and 6 represent the Reval formation.

4. Two and one half feet of gray limestone with one foot
of yellowish dolomite below it. 3 ft. 6 in. = 14 ft.

3. Conglomerate with pebbles of green glauconitic lime-

stone. 6 in. = 10 ft. 6 in.
Zones 3 and 4 represent the Kunda formation.

Green glauconitic limestone with most glauconite in the

top layer. 7 ft. = 10 ft.
1. Red calcareous sandstone with glauconite grains.
Base not seen. 3 ft. = 3 ft.

Zones 1 and 2 represent the Walchow formation.

Cuurcu Jecetecut. (Village Joa on the Jaggoul road, west of

Reval. Section by Schmidt (47)).
Upper Linsenschicht. 38m. = 8.0m.
Vaginoceras limestone. 3.2m. = 7.7 m.
Lower Linsenschicht. 2m. = 4.5m.
Glauconite limestone. 3.1m. = 4.3m.
Glauconite sandstone. 8m.= 1.2m.
Dictyonema shale. Base not seen. 4m. =

.4 m.

BELOW PACKERORT LIGHT-HOUSE. (Two miles north of Baltish-
port. Measured by the writer). Plate 7.

10. Heavy-bedded dolomitic limestone with few fossils.
Weathers dull yellow. Best exposed in the neighbor-
hood of Baltishport. Top not seen. 25 ft. = 80 ft. 1 in.

9. Linsenschicht. 10 in. = 55 ft. 1 in.
Zones 9 and 10 represent the Reval formation.
8. Thick-bedded limestone, the bottom layer full of pebbles

of limestone and irregular fragments of phosphatic shale.

Kunda formation.
7. 'Thin-bedded limestone and shale.

3 ft.6in. = 54 ft. 3 in.
1 ft.3 im. = 50 ft. 9 in.

Green limestone, very full of glauconite, especially at

2 ft. 6 in. = 49 ft. 6 in.
11 ft. = 47 ft.

the base.
- 5. Soft, green sandstone.
Zones 5 to 7 represent the Walchow formation.
4. Dark clay-shale, weathering to a light gray.

layers contain numerous graptolites.

Some
The shale rests

RAYMOND: CORRELATION OF THE ORDOVICIAN STRATA. 279

in hollows in the sandstone below and the strata bend up
and down with the irregularities of the subjacent sur-
face. At the base is a bed of limonite two to three inches -
in thickness. 18 to 13 ft. = 36 ft.
Thick, irregularly cross-bedded sandstone with some
specimens of Obolus in the upper layer. Thickness
variable. 5 to 10 ft. = 23 ft.
Thin strata of black shale alternating with layers of
yellow or red sandstone. The shale is in beds one to ten
inches thi¢k and is similar in character to that above.
The basal portion consists of conglomerate with boulders
of sandstone from one to five feet in diameter, and also
small pebbles cemented together by limonite and pyrite.

“ohne 5 ft. = 13 ft.
Zones 2 to 4 represent the Packerort formation.
Hard, white, coarse-grained sandstone. Base not seen.
Top of Esthonia formation. 8 ft. = 8 ft.

280 BULLETIN: MUSEUM OF COMPARATIVE ZOOLOGY.

BIBLIOGRAPHY.

In the list which follows I have enumerated all the works consulted during
the preparation of this report, either as to the geology or the identification
of fossils. I have not attempted a full bibliography, but, in the case of Russia,
have added a few titles which I have not seen, but which are referred to as
authorities for certain statements obtained from papers I have seen.

RUSSIA.

1. Bassuter, R. 8. The early Paleozoic Bryozoa of the Baltic provinces.
Bull. 77, U.S. N. M., 1911.

2. Bonnema, J. H. Beitrag zur kenntnis der ostrakoden der kuckersschen
-schicht (C2). Mitteil. Min. geol. inst. reisch. Univ. Groningen,
1909, 2, pt. 1.

3. Boruc, N. Die kleinen organizmen des untersilurs des Ostsee-
Ladoga glintes. Bull. Berg-ingen. gessellsch., 1904, no. 6 (in Rus-
slan).

4. Born, AxeL. Ueber neuere gliederungsversuche im esthlindeschen
hoheren untersilur. Centralb. min. geol. u. pal., 1913, p. 712.

5. Bucy, L. von. Beitrage zur bestimmung der gebirgsformation Russ-

lands. Archiv min., geogn. bergbau. hiittenkunde, 1840, 15.

6. Dysowski, W. N. Monographie der Zoantharia Sclerodermata rugosa
aus der silurformation Estlands, Nord-Livlands, und der Insel
Gotland. Archiv naturk. Liv-, Ehst-, und Kurlands, 1874, ser. 1, 5.

a: Die chaetetiden der ostbaltischen Silurformation. 1877.

8. EzcHwaup, E. Geognostico-zoologicae per Ingriam marisque Baltici
provincias nec non de trilobitis observationes. Casani, 1825.

9. Zoologia specialis. St. Petersburg, 1829, 1.

10. Silurische schichtensystem von Esthland. St. Petersburg, 1840.

1; Die urwelt Russlands. St. Petersburg, 1842, pt. 2.

12: Neuer beitrag zur geognosie Esthland’s und Finland’s. Beitrag
kenntn. d. russ. reiches, 1843, 8. .

13. Die urwelt Russlands. St. Petersburg und Moscau, 1840-48.
4 vols.

14. Die infusorienkunde Russlands, mit drei nachtrigen. Bull.
Soc. nat. Moscow, 1844-52.

15. Die grauwackenschichten Liev- und Esthlands. Moscow, 1854.

16. Beitrag zur geograph. verbreitung der fossilien. Thiere Russ-
lands, Moscow, 1857.

17. Lethaea Rossica. Stuttgart, 1860.

18.
19.

20.
21.
22.

23.
24.
25.
26.

27.

28.
29.

30.

| ol.
32.
33.
34.
30.
36.

37.

RAYMOND: CORRELATION OF THE ORDOVICIAN STRATA. 281

Heutmerson, G. von. Der artesische brunnen zu Petersburg. Bull.
Acad. imp. sci. St. Petersburg, 1864, 8, p. 185.

Horrman, E. Simmtliche bis jetzt bekannte trilobiten Russlands.
Verhandl. Russ. kais. min. gesellsch., 1857.

Hoi, G. Bericht ueber geologische reisen in Ehstland, Nord-Livland
und im St. Petersburgen Gouvernement in den Jahren 1883 und
1884. Verhandl. Russ. kais. min. gesellsch., 1886, 22.

Hoynincen-Hevunn, F. von. Beschreibung der silurischen craniaden
des Ostseeliinder. Verhandl. Russ. kais. min. gesellsch., 1899,
ser. 2, 36.

Supplement zu der beschreibung der silurischen craniaden der
Ostseelinder. Verhandl. Russ. kais. min. gesellsch., 1900, ser.
2, 38.

Hyatr, A. Phylogeny of an acquired characteristic. Proc. Amer.
philos. soc., 1894, 32. (Cephalopoda).

JAEKEL, O. Stammesgeschichte palmatozoen. 1. Thecoidea und Cys-
toidea. Berlin, 1899.

Kiarr, JoHAN. The Lower Silurian at Khabarova. Norwegian North
Polar exp. 1893-96, 1902, 4, no. 12. |

Koxen, E. Die gastropoden des baltischen untersilurs. Bull. Acad.
imp. sci. St. Petersburg, 1897, ser. 5, 7, no. 2.

Kuprrer, A. Ueber die chemische constitution der baltisch-silurischen
schichten. Archiv naturk. Liv-, Ehst-, und Kurlands, 1874, ser.
1, 5.

Kutoraa, S. Ueber die brachiopoden-familie. der Siphonotretacae.
Verhandl. Russ. kais. min. gesellsch., 1848, no. 12.

Lamansky, W. Die aeltesten silurischen schichten Russlands. Mem.
Comité geol. St. Petersburg, 1905, 20.

LEUCHTENBERG, Max, Herzog von. Beschreibung einiger neuen thier-
reste der urwelt aus den silureschen kalk-schichten von Zarskoje-
Selo. St. Petersburg, 1843.

Marcou, J. The Lower and Middle Taconic of Europe and North
America. Amer. geol., 1890, 5.

Micxkwitz, A. Vorlaiifige mittheilung ueber das genus Obolus Eichwald.
Bull. Acad. imp. sci. St. Petersburg, 1890, 1.

Ueber die brachiopoden-gattung Obolus Eichwald. Mem.
Acad. imp. sci. St. Petersburg, 1896, ser. 8, 4, no. 2.

Miuuer, F. Beitrige zur orographie und hydrographie Estlands.
St. Petersburg, 1872.

Mourcuison, R. I., VerRNEvIL, E. pg, and Kryseruine, A, le Comte de.
Russia and the Ural Mountains. London and Paris, 1845, 2 vols.

NreszkowskI, I. Monographie der trilobiten d. Ostsee-provinz. Archiv.
naturk. Liv-, Ehst-, und Kurlands, 1857, ser. 1, 1.

Nogtuine, F. Beitraige. zur kenntniss der cephalopoden aus silur-
geschieben der provinz Ost-Preussen. Jahrb. Koénig-Preuss. geol.
landesan., 1884,

282

38.
39.
40.
41.

42.

43.

44.
45.

46.
47.
48.
49.

50.

51.

52.

53.
54.

55.

56.

57.

BULLETIN: MUSEUM OF COMPARATIVE ZOOLOGY.

PanpDER, Cur. H. Beitrage zur geognosie der russichen reiches. St.
Petersburg, 1830.

PHALEN, A. v. D. Die gattung Orthisina. Mem. Acad. imp. sci. St.
Petersburg, 1878, ser. 7, 24, no. 8.

ScHAMARIN, A. Chemische untersuchung des brandschiefers von
Kuckers. Archiv naturk. Liv-, Ehst-, und Kurlands, 1874, ser. 1, 5.

Scumipt, C. Die grauen untersilurischen Thone der nordkiiste Ehst-
lands. Archiv. naturk. Liv-, Ehst-, und Kurlands, 1858, ser. 1, 2.

Scumipt, F. Untersuchen ueber die silurische formation von Ehst-
land, Nord-Livland und Oesel. Archiv. naturk. Liv-, Ehst-, und
Kurlands, 1858, ser. 1, 2.

Einige neue und wenig bekannte baltisch-silurische petrefacten
Mem. Acad. imp. sci. St. Petersburg, 1874, ser. 7, 21, no. 11.

Revision der ostbaltischen silurischen trilobiten, nebst geog-
nostischer uebersicht des ostbaltischen silurgebiets. Mem. Acad.
imp. sci. St. Petersburg, 1881, ser. 7, 30, no. 1.

On the Silurian (and Cambrian) strata of the Baltic provinces
of Russia, as compared with those of Scandinavia and the British
Isles. Quart. journ. Geol. soc. London, 1882, 38.

Ueber eine neuentdeckte untercambrische fauna in Estland.
Mem. Acad. imp. sci. St. Petersburg, 1888, ser. 7, 36, no. 2.

Excursion durch Estland. Guide excurs. VII Cong. geol.
international, 1897, no. 12. .

Revision der ostbaltischen silurischen trilobiten. Mem. Acad.
imp. sci. St. Petersburg, 1881, ser. 7, 30 — 1907, ser. 8, 20, no. 8.

ScHRoEDER, H. Untersuchen ueber silurische cephalopoden. Pal.
abh., 1891, n. f. 1, heft. 4.

Stottey, E. Untersuchen ueber Coelosphaeridium, Cyclocrinus,
Mastopora und verwandte genera des Silur. Archiv anthrop. u.
geol. Schleswig-Holsteins, 1896, 1, heft 2.

Neue siphoneen aus baltischen Silur. Archiv. anthrop. u. geol.
Schleswig-Holstein, 1898, 3, heft 1.

Srraneways, W.T.H.F. Description of strata in the brook Pulcovea,
near the village of Great Pulcovca, in the neighbourhood of St.
Petersburg. Trans. Geol. soc. London, 1821, 5, no. 2.

Geological sketch of the environs of St. Petersburg. Trans.
Geol. soc. London, 1821, ser. 1, 5, no. 2.

An outline of the geology of Russia. Trans. Geol. soc. London,
1822, ser. 2, 1.

Struve, H: Die artesischen wasser und untersilurischen thone zu
St. Petersburg. Mem. Acad. imp. sci. St. Petersburg, 1865, ser. 7,
S,:no. hl,

Vo.portu, A. von. Ueber Schmidtia und Acritis, zwei neue brachio-
poden gattungen. Verhandl. Russ. kais. min. gesellsch., 1869.
Wysogorsk!, J. Zur entwicklungsgeschichte der orthiden im ost-

baltischen Silur. Zeitsch. Deutsch. geol. gesellsch., 1900, 52, heft. 2.

RAYMOND: CORRELATION OF THE ORDOVICIAN STRATA. 283

SWEDEN.

Ancetin, N. P. Palaeontologia Scandinavica. Pars 1, Crustacea
formationis transitionis. 1851-1854. (Lindstrém’s edition, 1878).

Frearnsipes, W.G. The Lower Ordovician rocks of Scandinavia, with
a comparison of British and Scandinavian Tremadoec and Arenig
rocks. Geol. mag., 1907, dec. 5, 4.

Happina, AssaAr. Undre Dicellograptusskiffern i Skane. Lunds univ.
arsskrift, 1913, n. f., afd. 2, 9, no. 15.

Der mittlere Dicellograptus-schiefer auf Bornholm. Lunds
univ. arsskrift, 1915, n. f. afd. 2, 11, no. 4.

Hoitm, G. Ueber einige trilobiten aus dem _ Phyllograptusschiefer
Dalekarliens. Bihang Kgl. Svenska vet. akad. Handl., 1882, 6, no. 9.

De Svenska arterna af trilobitslagtet Illaenus (Dalman).
Bihang Kgl. Svenska vet. akad. Handl., 1882, 7, no. 3.

Kinnekulles berggrund. Sver. geol. unders., 1901, ser. C, no.
172.

Om Didymograptus, Tetragraptus, och Phyllograptus. Sver.
geol. unders., 1895, ser. C, no. 150. English translation, by Ellis
and Wood. Geol. mag., 1895, dec. 4, 2.

Linpstr6m, AxEL. Beskrifning till kartbladet Venersborg. Sver. geol.
unders., 1887, ser. Ab, no. 11.

Linpstr6m, G. F. List of the fossil faunas of Sweden. Stockholm,
1888.

LINNARSSON, G. On the vertical range of the graptolite types in Sweden.
Geol. mag., 1876, dec. 2, 3.

Linnarsson,. G. and Tutupere, 8. A. Beskrifning till kartbladet

Vreta kloster. Sver. geol. unders., 1882, ser. Aa, no. 83.

Marr, J. W. On the Cambrian and Silurian rocks of Scandinavia.
Quart. journ. Geol. soc. London, 1882, 38, p. 313.

Mosere, J. C. Anteckningar om Oelinds orthocerkalk. Sver. geol.
unders., 1890, ser. 6, no. 109.

Om den af Trinucleus coscinorrhinus Angelin karakteriserade
kalkens alder. Sver. geol. unders., 1892, ser. C, no. 125.

Nya bidrag till utredning af fragen om griinsen mellan under-
silur och kambrium. Geol. f6ren. férhandl. Stockholm, 1900, 22,
p. 5238. .

Guide for the principal Silurian districts of Scania. Geol. féren.
forhandl. Stockholm, 1910, 32, haft 1.

Historical-stratigraphical review of the Silurian of Sweden.
Sver. geol. unders., 1911, ser. C, no. 229.

Mosera, J. C., and SecerserG, C. O. Bidrag till kiinnedomen om
Ceratopygeregionen med sirskild hinsyn till dess utveckling i
Fogelsingstrakten. Lunds univ. irsskrift, 1906, n.f. afd. 2, 2, no. 7.

284

cif é
78.

vo.

80.
81.

82.

83.

84.

85.

86.
87.
88.
89.
90.

St.

92.
93.

94.

BULLETIN: MUSEUM OF COMPARATIVE ZOOLOGY.

Mounrue,H. De geologiska hufvuddragen af Vastgétabergen och deras
omgifning. Sver. geol. unders., 1906, ser. C, no. 198.

Beskrifning till kartbladet Kalmar. [and] Beskrifning till
kartbladet Ottenby. Sver. geol. unders., 1902, ser. Ac., no. 6, 7.

Ouin, E. Om de Chasmopskalken och Trinucleusskiffern motsvar-
ande bildningarne i Skane. Lunds univ. arsskrift, 1906, n.f., afd. 2,
no. d.

Post, L. von. Bidrag till kinnedomen on ceratopygeregionens ut-
bildning inom Falbygden. Sver. geol. unders., 1906, ser. C, no. 203.

STRANDMARK, J. EK. Undre graptolitskiffer vid Fagelsang. Geol.
foren. forhandl. Stockholm, 1902, 23, no. 7, p. 548.

Tornauist, 8. L. Undersékningar 6fver Siljansomradets trilobitfauna.
Lunds univ. arsskrift, 1884, 20. Also Sver. geol. unders., 1884,
ser. C, no. 66.

Researches into the graptolites of the lower zones of the Scanian
and Vestrogothian Phyllo-Tetragraptus beds. Lunds univ. arsskrift,
1901, 37, afd. 2, no 5; 1904, 40, afd. 2, no. 2.

WarsureG, E. Geological description of Nittsjo and its environs in

Delarne. Geol. féren. forhandl. Stockholm, 1910, 32, haft 2.

WESTERGARD, A. H. Studier 6fver Dictyograptusskiffern och dess
grinslager med sarskild hansyn till i Skane férekommande bild-
ningar. Lunds. univ. arsskrift, 1909, n.f., afd. 2, 5, no. 3.

Wiman, C. Ueber das silurgebiet des bottnischen meeres. Bull.
Geol. inst. Upsala, 1893, 1, no. 1.

Ueber die silurformation in Jemtland. Bull. Geol. inst. Upsala,
1893, 1, no. 2.

Ein shumardiaschiefer bei Lanna in Nerike. Ark. zool. K.
Svenska. vet. akad. Stockholm, 1905, 2, no. 11.

Studien iiber das nordbaltische Silurgebiet. Bull. Geol. inst.
Upsala, 1905, 6; 1908, 8.

Die Silurbildungen in Vastergétland. Guide excurs. 11th
Internat. geol. congress, 1910, no. 22.

Wav, C., and Hepsrrém, H. Beskrifning till blad 5 [Oeland]. Sver.
geol. unders., 1906, ser. Ala.

Norway.

Byor_Kke, K. O. Geologisk kart med beskrivelse-over Kristiania By.
Norges geol. unders., 1898, no. 25.
BroéacEr, W. C. Die silurischen etagen 2 und 3 in Kristianiagebiet
und auf Eker. Kristiania, 1882.
Spaltenverwerfungen in der gegend Langesund-Skien. Nyt
mag. naturv., 1884, 28.

95.

96.

97.

100.
101.

102.

103.
104.
105.

106.

107.
108.
109.
110.
111.
112.

113.

RAYMOND: CORRELATION OF THE ORDOVICIAN STRATA. 285

Bréacer, W.C. Geologisk kart over Oerne ved Kristiania. Nyt mag.
naturv., 1890, 31.

Getz, A. Graptolitfoerende skiferzoner i det Trondhjemske. Nyt
mag. naturv., 1890, 31.

Ho.TEepAHL, O. Studien ueber die etage 4 des norwegischen silur-
systems beim Mjdésen. Videnskabs-selskabets skrifter. 1. Math.-
naturv. klasse, 1909, no. 7.

AMERICA.

Basser, R. S. The cement resources of Virginia west of the Blue
Ridge. Bull. 2-A, Virginia geol. survey, 1909.

Bréccrr, W. C. Ueber die verbreitung der Euloma-Niobe fauna in
Europa. Nyt mag. naturv., 1896, 35, p. 164.

Cotu, G. L. Ordovician section near Bellefonte, Pennsylvania.
Bull. Geol. soc. Amer., 1903, 14, p. 407.

Cumines, E. R. The Lower Silurian system of eastern Montgomery
county, New York. Bull. 34, N. Y. state mus., 1900.

Forrste, A. F. The phosphate deposits in the Upper Trenton lime-
stone of central Kentucky. Kentucky geol. survey, 1913, ser. 4, 1,
pt. 1.

Haun, F. E. On the Dictyonema fauna of Navy Island, New Bruns-
wick. Ann. N. Y. acad. sci., 1912, 22, p. 135.

MarruEw, G. F. Ona new horizon in the St. John group. Canadian
rec. sci., 1891, 4, no. 7.

Powe tL, 8. L. Discovery of the Normanskill graptolite fauna in the
Athens shale of southwestern Virginia. Journ. geol., 1915, 23.
Prosser, C. S. and Cuminas, E. R. Sections and thickness of the

Lower Silurian formations of West Canada creek and in the Mo-
hawk valley. 15th ann. rept. N. Y. state geol. survey, 1895.
Raymonp, P. E. The faunas of the Trenton at the type section and
at Newport, N. Y. Bull. Amer. paleontology, 1903, 4, no. 17.

The Chazy formation and its fauna. Ann. Carnegie mus.,
1906, 3.
The Ordovician of Montreal and Ottawa. Guide book no. 3,
excurs. 12th Internat. geol. congress, 1913.
Description of some new Asaphidae. Bull. 1, Victoria mem.
mus., 1913.
The succession of faunas at Lévis, P. Q. Amer. journ. sci.,
1914, ser. 4, 38, p. 523.
Thej Trenton group in Ontario and Quebec. Summary rept.
Director Geol. survey dept. mines Canada for 1912, 1914.
Ruepemann, R. The Hudson River beds near Albany, and their
taxonomic equivalents. Bull. 42, N. Y. state mus., 1901.

286

114.

115.
116.
147.
118.
119.

120.

121.

BULLETIN: MUSEUM OF COMPARATIVE ZOOLOGY.

RUEDEMANN, R. The Trenton conglomerate of Rysedorph hill, Rens-
selaer county, N. Y., and its fauna. Bull. 49, N. Y. state mus.,
1901.

The graptolite facies of the Beekmantown formation in Rens-
selaer county, N. Y. Bull. 52, N. Y. state mus., 1902.

The graptolites of New York. Memoirs 7, 11, N. Y. state
mus., 1904, pt. 1; 1908, pt. 2.

The Lower Siluric shales of the Mohawk valley. Bull. 162,
N. Y. state mus., 1912.

Stosre, G.W. Geological atlas U.S. G.S. Mercersburg-Chambers-
burg folio, 1910. ’

Uxricu, E. O. Revision of the Paleozoic systems. Bull. Geol. soc.
Amer., 1911, 22.

Watcorr, C. D. Cambro-Ordovician boundary in British Columbia,
with description of fossils. Smithsonian misc. coll., 1912, 57,
no. 7.

WIincHELL, N. H. and Utricu, E. O. The Lower Silurian deposits
of the Upper Mississippi province. Paleontology Minnesota,
1897, 3, pt. 2.

=e.

Sa t
Raymonp.— Correlation of the Ordovician Strata. o

; |
\ x
7 2
.
. r
we TATE Aytt 4 b
= in #
Sa be
hn. « ct: é ;
¥
4 .
< 4 5
\
-
>
~~ ¢
\ =
«
|
t
' e 4
7 ,
Ls
= \ : <~
“ ‘«

t PLATE 1.
b RELI

<i a

4 of.
2,
P|
ac
/
|
J
}
7
‘
x ~ _
.
« ‘ \
te rs lel oo) oe €
. ; el ee

BULL. MUS. COMP. ZOOL.

EXPED. TO BALTIC PROVINCES. PLATE 1

LAKE LADOGA

GULF OF FINLAND

ASH

—_—— IN THE -—-—

BALTIC PROVINCES OF RUSSIA

V
‘e)
EN
a?)
s8
Q
34
A,
Ho
gf
eS
SS
pa]

'
|

yl
‘Si
if

MODIFIED FROM BASSLER (AFTER SCHM/OT)

CAMBRIAN

S/LURIAN

ORDOVICIAN

HELIOTYPE ©CO., BOSTON

EE

out B77)

_ PLATE 2.

te |
<<
~

Raymonp.— Correlation of the Ordovician Strata.

PLATE 2.

Comparative table of sections of the Lower Ordovician strata in the Govern-
ments of Petrograd and Esthonia, Russia. This diagram is redrawn with
some alterations from one published by Lamansky. A: is the Lower Cambrian
Esthonia formation, A: the Obolus sandstone, A; the Dictyonema shale,
B,; the Glauconite sandstone, By; the Glauconite limestone with its subzones,
and Biz the Orthoceras limestone (Kunda formation). The Bis is the
Birra of Lamansky. A» and A; make up the Packerort formation and B,
and By; are the Walchow formation.

LYOYTWI Vd

EXPED. TO BALTIC PROVINCES. PLATE 2

BULL. MUS. COMP. ZOOL.

JLVYTWOTINOD OBBS

LOSE SEP EASESER:
IESE EMIRSAR

LSESEESTRERSOSEITSN A:

ANOLSTW/7

IILINOINYTO
BWZ ZA Ue AA
FEZAVZZ/A 77

ISSEG EE SOROS IES
WESSEBESSROCFORR IRE
SSOESEIPSEESIRERG:

ISI ISSINGEE,
SESE EESMAAESASAE

LIF TIIII IT OILS
SPT STOTT
|

%

f}
“Ht
Wy
“4, “eg
Mh A

ANOLSONES
JILINODTYTI

INOLS TWIT

arerareat

DO EE ee =

VYMO0d td

ALL

FNOLSONES
$7090
SIS IIITIISIERGA
AO OSEISAERSESA ST:
DU eine oes

SLIT)
LL IEO
- Pr

MOHITUM

TELE LITETITLL
LLL TILE LL ILA
ULL LAA AL ASL hhh
SLISIDIDISDSSDSADIS I A
SISIISISIIGISSI AAG

144 1%

HELIOTYPE CO., BOSTON

vi

be

Eo) ;

in

oO) AM be whiter) ~~ serenade
Soh ae
LAO ea ors

, Bra
é % ,
\ — 4 ;
> ; = =

-

wel x

et nt re

podulaiy wild gh nk
mrot-ettd hedbaila:

yy) :
fe

*

484 ws gy

¢

*

|
et

' id ¥. ; 7 iB ahs fet at
4 Evy, What Ket RT
{ wy eg, ae ae Ms
ra 2 See ee
ae 7 a “i % a2 4
Pope Rae Te Tate
: ‘ At ae
Mian: a
* pt
? os
Raymonp.— Correlation of the Ordovician Strata. ;
7

PLATE 3.

Sketch showing the probable relationship of the various Ordovician forma-
tions in a west-east section from the shore of the Baltic to the Walchow River.
With the Kuckers and Itfer is included the Jewe. ©

Not drawn to scale.
_
?
:
ed “i
ee
1 €
7
aos JON. i
- "Al a aS ae 2
Se, el it 1 _ 4
; (es LS ye! te -
er ce bee Tak ais a i ‘% ;

PLATE 3

EXPED. TO BALTIC PROVINCES.

BULL. MUS. COMP. ZOOL.

TB fae) eee wes ll i eee
ae 3S) 0 a Oe es ee ee
= eee ieee cseneneereeree renee
a A SE a Te ee ee ee
a sO OD Te eS ee
0 ee ee
Ee 2 ee ee Erg

i re ? te ee
6 a ae Ma eT BY Bree

rr arr
oe

wt eterre
eolcle

si | | ET
cme mea eran et
rs cae Gi iy BE ey coe ae ge
: ee ee ee ee a Pe es oe es
of e*, Hes 2)

CO al FS EY Ie a

St

OHI7TYM TVYDOULTd QYTENTSTM = TUN7Y = LYOCHSLLIVE-WTIVSSEM ~— LS)

BOSTON

HELIOTYPE CO.,

Raymonp.— Correlation of the Ordovician Strata.

PLATE 4.

Fig. 1. An exposure of the lower beds of the Ordovician (Walchow and
Packerort formations) on the Lawa at Wassilkowa. The top of
the Lower Cambrian is at water-level; the Obolus sandstone, the
Dictyonema shale, and the Glauconitic sandstone all occur in the
fifteen feet of strata beneath the projecting beds seen in the middle
of the figure. The conspicuous beds are the lowest limestone
and belong to the zone of Megalaspis planilimbata (Bira). Above
this are seen the softer strata of Brig, then the harder limestone
of Biz,, and at the top, the zone of Asaphus expansus, or Bris
(Birra Of Lamansky).

Fig. 2. Quarry in the upper part of the Walchow and the lower part of the
Kunda at Putilowa. Both these localities are south of Lake
Ladoga and east of Petrograd, Russia.

. BULL. MUS. COMP. ZOOL EXPED. TO BALTIC PROVINCES. PLATE 4

HELIOTYPE CO,, BOSTON

vr

an

A,

Raymonp.— Correlation of the Ordovician Strata.

Fig. 1.

Fig. 2.

Fig. 3.

PLATE 6.

Thin section of glauconitic limestone from Putilowa, Russia. The
darker grains are glauconite; note the extent to which they are
altered to opaque limonite, especially in the large grain near the
edge to the left.

Another section from Putilowa in which glauconite is somewhat more
abundant and less altered.

A thin section of limestone from the zone of Asaphus expansus (top
of the Walchow) in a quarry opposite Iswos, Russia.

Note that all three of these slides show the limestone to be made
up almost entirely of fragments of fossils, largely trilobites, brachio-
pods, and cystids. All magnified about 10 diameters.

PLATE 5

EXPED. TO BALTIC PROVINCES.

BULL. MUS. COMP. ZOOL

:

|
)

HELIOTYPE CO., BOSTON

wid We Grd beds ps

peajesittii

2 Spat: pis js
: tay Bes: sri h fie)
eg me aiwdl le

eA, bore ay 7 e$ “Ae

ree 2425 f , 4 * BMS ews

Raymonp.— Correlation of the Ordovician Strata.

PLATE 6.

Fig. 1. Cross-bedded limestone at the top of the Wassalem at the type-
locality.

Fig. 2. Cliff on the western bank of the Narowa at Narwa, Russia. The
Obolus sandstone is seen at the base, and is 11 feet thick. The
reéntrant at the top of the sandstone indicates the position of the
Glauconite sandstone, which is only 4 inches thick. The lower
half of the remaining height of the cliff is occupied by the Walchow
formation, and the upper half by the Kunda.

BULL. MUS. COMP. ZOOL. eulale: a tai vic Prognetce Paaree

HELIOTYPE CO., BOSTON

et ; ours

mpireserer? rclosig

srr “ort. ovine?

Raymonp.— Correlation of the Ordovician Strata.

PLATE 7.

Fig. 1. Basal conglomerate of the Obolus sandstone at Packerort, Russia.

Fig. 2. The ‘“Glint” beneath Packerort light-house. Height, 80 feet. At
the bottom are seen the upper 8 feet of the Lower Cambrian, and
the contact with the Ordovician is indicated by the lower under-
cutting, which is at the base of the Obolus sandstone. Above
this sandstone may be seen the Dictyonema shales and then the
Glauconite sandstone, making the second undercutting. Above
this are the thin representatives of the Walchow and Kunda forma-
tions, and at the top, the overhanging Reval dolomite, the Dubo-
wiki being absent.

BULL. MUS. COMP. ZOOL.

BOSTON

HELIOTYPE CO.,

Raymonp.— Correlation of the Ordovician Strata.

PLATE 8,

Table showing the writer’s interpretation of the sections and correlation
of the subdivisions of the Ordovician of Esthonia and Scandinavia with those
of certain localities in North America.

The first section is from the coast at Kunda south through Wesenberg, the
second from the coast at Packerort near Baltishport south through Wassalem
to the vicinity of Hapsal: all of these places being in Esthonia, Russia. The
next five are all in Sweden, and the eighth is a generalized section at Chris-
tiania, Norway. Of the remainder, all, except the one at Ottawa, Canada, are
in the United States. The one in the Champlain Valley includes strata in both
New York and Vermont, Chambersburg and Bellefonte are both in Pennsyl-
vania, the section from Kentucky is that near Lexington, and the one in
central New York is a combination of sections near Trenton Falls and Utica.

BULL. MUS. COMP. ZOOL.
‘ Kunda Baltishport
. Wesenberg Hapsal

Wesenberg

mm
limestone
MS
Chasmops

limestone
a a

ey

Reval

Se

Reval

AM

| Kunda

Walchow

Ancistroceras
limestone

Chiron

limestone

Platyurus
limestone

Gigas
limestone

}Expansus Is. _| Is.

Ordovician

‘yaaoe oy Eze

Tetragraptus
zone

W taupe

| Packerort

Lower

Ceratopyge ls.

Glauconite ss.

Obolus
conglomerate

| Packerort |

Roa

EXPED. TO BALTIC PROVINCES. PLATE 8

|
|
|

, Ottawa Central

al

Frankfort
Stewartville
eae | Upper Picton |
f Prosse
Prasopora Prasopora
zone zone
Hull Glens Falls
Decorah

ve | eee Feb

iil
HY

Little Falls

| Phere |
ye
jaime |

S| I

cal a hoe ees

—* i E

4,
aes

<

7
4

a | -* rt Fe

Ee ee ie Ve ee eee ee i as

Bulletin of the Museum of Comparative Zodlogy
AT HARVARD COLLEGE.

Vout. LVI. No. 4.

GEOLOGICAL SERIES, Vol. X. SHALER MEMORIAL SERIES, No. 3.

EXPEDITION TO THE BALTIC PROVINCES OF RUSSIA
AND SCANDINAVIA, 1914.

PART 2.— THE SILURIAN AND HIGH ORDOVICIAN STRATA OF
ESTHONIA, RUSSIA AND THEIR FAUNAS.

PART 3—AN INTERPRETATION OF THE SILURIAN SECTION
OF GOTLAND.

By W. H. TWENHOFEL.

Wits Five Puartes.

CAMBRIDGE, MASS., U.S. A.:
PRINTED FOR THE MUSEUM.
Juty, 1916.

No. 4.— Expedition to the Baltic Provinces of Russia and Scandinavia,
1914.

Part 2.— THE SILURIAN AND HiaH OrpovicrAN STRATA OF ESTHONTA,
RuSSIA, AND THEIR FAUNAS.

By W. H. TweEnHoFELt.

TABLE OF CONTENTS.

. Part 2.
PAGE.
Introduction : : ; i ; : ; : . . 289
The Russian Section . : : ; ; ; . ;’ ae
The Ordovician System : : : . ; ; ‘ . 293
The Silurian System . : : ; - 814
PaRT 3.
Introduction , c . : , : jf : é . 341
Conclusions . : 3 ‘ ; . : ; P “ .. om

Explanation of Plates

INTROD UCTION.

THE present paper is based on work done under the auspices of the
Shaler Memorial fund. In company with Prof. Perey E. Raymond,
an expedition was undertaken in the summer of 1914 for the purpose
of examining the Cambrian, Ordovician, and Silurian strata exposed
around the Baltic in Russia, Sweden, and Norway, with the object
in view of attempting a correlation with strata of the same systems
of eastern North America. The writer was primarily concerned with
the highest Ordovician and Silurian, while the Cambrian and the
remainder of the Ordovician were studied by Professor Raymond,
although mutual assistance was rendered. The outbreak of the war
to some extent necessitated a curtailment of the study, but in my case
this amounted to the loss of only a few days’ work in northern Got-
land and a considerable shortening of the time allotted to Norway.
The failure to see certain sections of northern Gotland will be ulti-
mately overcome by detailed collections which have been made (1915)

290 BULLETIN: MUSEUM OF COMPARATIVE ZOOLOGY.

from these sections under the direction of Dr. Henry Munthe of the
Swedish Geological Survey. Nearly all of southern Gotland and
parts of northern Gotland were examined, much of the study being-
under the guidance of Dr. Munthe. About ten days were spent in
Dalarne studying the Leptaena kalk, while Professor Raymond made
collections from the Brachiopod shales and Silurian horizons of
southern Sweden.

In the Kristiana region of Norway, the Malmo and Ringerike
Silurian sections were studied, the latter under the guidance of Drs.
Johan Kiaer and Olaf Holtedahl of the University of Kristiana, and,
as the facies of the latter section is quite similar to that of the Silurian
of eastern America, while the other Norwegian Silurian sections are
of a quite different lithology, the failure to see the latter is not of
great importance, especially since they have been exhaustively studied
by Professor Kiaer.

The Russian Silurian and higher Ordovician were carefully exam-
ined, each of the type-sections of Schmidt being studied, in addition
to many other outcrops.

It was also desired to see some of the English sections; but this *
proved to be impossible. This misfortune was largely made good,
however, through the courtesy of Dr. Audrey Strahan, Director of
the Geological Survey of England and Wales, and different members
of his staff, particularly Mr. John Pringle, who placed, at my dis-
posal for study, complete detailed collections from Silurian and high
Ordovician strata, together with the general collections of the Museum
of Practical Geology. The kindness of Dr. F. Cowper Reed of
Cambridge University gave the opportunity to examine such collec-
tions at the Sedgwick Museum of Cambridge as had been made from
the Keisley and Chair of Kildare limestones, while Dr. F. A. Bather
of the British Museum very courteously permitted an examination
of desired parts of that Museum’s magnificent collections.

In this paper merely a preliminary discussion of the Russian sec-
tions is given, as the fossils have not yet been studied.

The opportunity is taken at this time to acknowledge the many
courtesies and the unselfish kindness received from all from whom
assistance was desired. Thanks are particularly due to Baron Frey-
tag-Loringhoven of the island of Oesel, Baron Toll of Kuckers, Baron
Maidel of Eichenheim, Baron Rosen of Lyckholm, Herr E. von Wahl
of Addifer, the Directors of the cement plants at Port Kunda and
Asserien, Dr. 1. P. Tolmacev of the Imperial Academy of Petrograd,
Drs. Henry Munthe, Johan Kiaer, Olaf Holtedahl, Audrey Strahan,

TWENHOFEL: EXPEDITION TO THE BALTIC PROVINCES. 291

F. A. Bather, F. R. C. Reed, Mr. John Pringle, as well as many others
who rendered assistance. Thanks are also due Professor Raymond for
‘the cordial comradeship and great assistance afforded during the
study of the Baltic sequence. Dr. R. S. Bassler has kindly read the
paper for which I am grateful.

THE RUSSIAN SECTION.

Work on the Russian section of Esthonia was begun on June 15,
and closed on July 17. Fossils were collected wherever it was pos-
sible. These have not yet reached the United States and pending
their arrival no detailed discussion of the correlation will be given.

Previous work. Many foreign geologists have studied the Estho-
nian sections of the Russian Ordovician and Silurian and their
faunas. To attempt a review of the work of these students in detail

is out of question, and one must be content with merely a brief refer-
ence to a few of those whose contributions are of greatest importance.
Those whose studies are perhaps of most value are:—

Pander. Beitrige zur geognosie der russichen reiches. Niesz-
kowski. Versuch einer monographie der in den silurischen schichten
der Ostseeprovincen vorkommenden trilobiten. Archiv. fur naturk.
Liv-, Est-, und Kurlands, 1857, ser. 1, 2. Dybowski. Monographie
der Zoantharia Sclerodermata rugosa aus der silurformation Est-
lands, Nord-Livlands, und der Insel Gotland. Archiv. fur naturk.,
Liv-, Est-, und Kurlands, 1873, ser. 1, 5. Koken. Bull. Acad. imp.
sci. St. Petersburg, 1897, ser. 5, 7, no. 2. Eichwald. Bull. Soc. nat.
Moscow, 1854, 1855. The most extensive contributor to the knowl-
edge of the Baltic section in Russia was F. Schmidt, who in numerous
papers, beginning in 1858, has described the section and its faunas.’
For more than fifty years he wandered over the Russian Baltic sec-
tions, collecting fossils, studying the stratigraphy, and making observa-
tions in many fields.

A more recent paper by Bassler ? discusses the Ordovician section,
comprehensively describes the Bryozoa, and attempts a correlation ~
with American equivalents. The latest paper is by Axel Born,’

1Schmidt. Untersuchungen iiber die silurische formation von Ehstland, Nord-Livland und

Oesel. Archiv. fur naturk. Liv-, Est-, und Kurlands, 1858, ser. 1, 2, p. 1-248, 465-475.
2 Bassler. The early Paleozoic Bryozoa of the Baltic Provinces. Bull. 77, U.S. N. M., 1911.
3 Born. Centralbl. min. geol. pal., 1913, no. 22, p. 712-720.

292 BULLETIN: MUSEUM OF COMPARATIVE ZOOLOGY.

who takes exception to some of Bassler’s conclusions relating to the
stratigraphy. A work which discusses the history, archaeology,
stratigraphy, and natural history of the Russian Baltic provinces
appeared in Riga in 1911, and bore the title, Baltische Landeskunde.
It was the work of several writers, that which pertained to the Palaeo-
zoic geology having been written by A. von Mickwitz. It is a good
work for general reference.

Introductory discussion of the Esthonian section. 'The entire se-
quence of the Russian Baltic section consists of evenly bedded, almost
horizontal limestones and subordinate shales with an occasional thin
division of sandy material. There is a very gentle dip southwest-
ward, generally imperceptible. The entire Ordovician and Silurian
may have a thickness of 725 feet, of which about 350 feet are
Ordovician and 375 feet Silurian.

Natural outcrops are not common, rarely existing save at the
seashore, where the Cambrian and basal divisions of the Ordovician
are exposed in the cliffs, or, as they are called in Esthonia, glints. On
the island of Oesel the Silurian is exposed in the sea cliffs, which on
this island are known as panks. Had one to depend on natural
exposures little could be learned of the stratigraphy of the land a few
miles distant from the sea. Everything is grassed over and, if a
surface be made bare, a few years suffice to completely cover it again.
Fortunately, however, the need of limestone for burning or con-
struction purposes in the past, led to the opening of many small
quarries, and in later times artificial exposures have been further
increased by the digging of ditches for the drainage of swamps or
roads. Through these, a partial understanding of the stratigraphy
of the interior has been attained. Of late years the demand for lime
and stone appears to have decreased, or to have been supplied from
elsewhere, since many of the quarries studied by Schmidt are now
wholly or partially grassed over, so that their examination is difficult,
and, in some cases, impossible. Large cement plants have been built
at Port Kunda and Asserian, and these have developed extensive
exposures of the lower divisions of the Ordovician; but nothing of the
higher beds.

Up to the present it has been impossible to learn the exact sequence
of strata above the Echinosphaerites limestone and it is rare that one
is able to discover the contact of any formation with those adjacent.
Hence the determination whether certain strata are continuous with
others, from which they differ through horizontal variation of sedi-
ments, or whether they lie at a different horizon, has not been possible.
It will probably be long before the sequence is completely known.

TWENHOFEL: EXPEDITION TO THE BALTIC PROVINCES. 293

To some of the divisions, geographical formational names were
applied by Schmidt, while others were named after some character-
istic fossil or lithic characteristic. Each division coming within the
writer’s problem has been redefined, Schmidt’s geographical terms
have been retained, with, so far as determinable, the limitations
imposed by him. To divisions named after characteristic fossils,
geographical formational names have been applied.

The names of the fossils listed have been taken from Schmidt,
Koken, Bassler, and others, and some are field identifications.

THE ORDOVICIAN SYSTEM.

Introduction. The Ordovician was subdivided by Schmidt into the
following members which are named with their thicknesses (approxi-
mate only) from the summit downward.

F2. Borkholm limestone. 20 ft.
Fl. Lyckholm limestone. 50-60 ft.
E. Wesenberg limestone. 30 ft.
D3. Wassalem limestone. 10 ft.
D2. Kegel limestone. 10 ft. |
D1. Jewe limestone. 100 ft.
C3. Itfer beds. 10 ft.
C2. Kuckers shale. 10 ft.
C1. Echinosphaerites limestone. 30 ft.
B3. Vaginaten limestone. — 10 ft.
B2. Glauconite limestone. 30 ft.
Bl. Glauconite sandstone. 30 ft.

Of these only the Lyckholm and Borkholm limestones come within
the writer’s problem.

Lyckholm formation. The most easterly known exposures of this
formation are along the lower course of the Pungern Brook just before
it empties into the northern end of Lake Peipus. Bending northward
the outcrops follow just north of the 59th parallel, and reach the sea
on the Nuck6 peninsula, north of the city of Hapsal. The Island of
Worms probably belongs wholly to the Lyckholm and there are exten-
sive outcrops on the northern side of Dago. The thickness of the
formation is not known exactly, but it is estimated to be about fifty
or sixty feet.

294 BULLETIN: MUSEUM OF COMPARATIVE ZOOLOGY.

The division received its name from Lyckholm, the home of Baron
Rosen, which is situated on the long Nucké peninsula, opposite the
city of Hapsal. The type-exposure, and apparently the only one in
the vicinity, lies about one eighth of a mile north of the dwelling.
Formerly it was quite extensive, but at present little is shown and the
old dump heaps have been picked over so frequently, that, without
further quarrying, it is now impossible to obtain a representative col-
lection. The rock consists of gray and white to yellowish white,
rather thick-bedded, partially crystalline limestone, the individual
beds being separated by thin shale partings. Not more than five or
six feet are, or were exposed. It appears very probable that these
beds belong to the lower division of the formation.

Several small exposures have recently been developed at Hohen-
holm on the Island of Dago and from five to six feet of strata have
been exposed, consisting of blue and gray limestone with thin shale
partings. The bedding is not well defined. The Kegel appears to
lie about five or six feet below the surface, and it seems to be visible
at the water’s edge a short distance north, and just back of the factory
at Hohenholm, a ditch reaches a limestone of Kegel aspect, and so
Dr. Raymond considers it. Schmidt reports the occurrence of the
Wesenberg limestone on the shore by the village of Rootsi, only a short
distance east ! but what he saw was probably the Kegel. The older
Hohenholm exposures could not be found. These beds contain
essentially the same fauna as those at Lyckholm; Halysites catenu-
laria, the typical Porambonites gigas, and Triplecia insularis, in addi-
tion to other fossils, having been collected.

At Paope, about four miles southwest of Hohenholm, is a quarry of
considerable size which during the summer of 1914 was being worked.
Not over six or seven feet of rock are exposed, consisting of soft impure
bluish gray limestone of semicrystalline texture with beds of blue shale
separating the limestones. ~“ The beds of this quarry probably lie a
little above those of the Hohenholm exposures, as the locality is south
of that place and the quarry is situated on a little higher ground. The
fauna is the same as that of Lyckholm and Hohenholm and the lith-
ology is also quite similar. Fossils are very abundant.

Kertel is the name of a village on the northern side of Dago and
the quarries are located about a mile south of the village, near Pallo-
kill Krug (Plate 2, fig. 1).2~ About six feet of somewhat heavy-

1Schmidt. Loc. cit., 1858, p. 136.
2 A krug is a place where lodging and food may be obtained.

TWENHOFEL: EXPEDITION TO THE BALTIC PROVINCES. 295

bedded, semicrystalline, grayish blue limestones and thin shales are
exposed. Throughout most of the thickness the bedding is illy defined.
Some of the beds are quarried for construction purposes and these are
from four to six inches thick. The basal beds are more shaly than
those above and locally consist of 50% shale. The limestone is
similar to that of the other localities mentioned above.

At Pallokiill Chapel, three to four miles south of Kertel, on the road
to Helterma, is an exposure of what appears to be the Kegel. The
outcrop is in the woods a short distance back of, and southwest of the
chapel, and the beds dip from ten to fifteen degrees northward. Un-
less the tilting of the Kegel is purely local and involves the higher beds,
it follows that the Lower Lyckholm rests unconformably on the former.
That the relations are disconformable appears fairly certain.

Another exposure of what appears to be the Lower Lyckholm was
seen at Kappa-Koil, south of Reval, on the railroad to Pernau; but
the old quarry was almost wholly grassed over and no fossils were
collected.

Near Muddis Krug, about two miles east of the railroad station,
Taps, is an outcrop of the Lyckholm which is of considerable im-
portance as the exposed beds are not far above the contact with the

_Wesenberg and the locality is nearly at the eastern end of the Silurian

territory. A long low cutting on the railroad about a mile and a
half east of Taps exposes a dense fine-grained, almost unfossiliferous
limestone. In a small quarry nearer Taps, beds of a similar character
are exposed in which Chasmops wesenbergensis and a few other fossils
were collected which show the strata to belong to the Wesenberg.

About a half mile south of the railroad cutting and a half mile west
of the Meinkerb residence is the old quarry referred to by Schmidt as
“near Muddis Krug.” Only two feet of irregularly bedded, whitish
earthy limestone. are now visible. The fauna from these beds con-
sists of such typical Lyckholm forms as Porambonites gigas, Pseudo-
lingula quadrata, Triplecia insularis, Halysites, Heleolites, and large
gastropods of the genera Hormotoma and Subulites, forms essentially
identical with those found in the Lyckholm beds on the Island of Dago
at the locality where it overlies the Kegel, and they show that the
lower beds of the formation are the same in the two widely separated
regions.!

The strata described in the preceding paragraphs do not exceed

1 For the information relating to the outcrops near Taps I am indebted to Professor
Raymond.

296 BULLETIN: MUSEUM OF COMPARATIVE ZOOLOGY.

twenty-five feet in thickness and all apparently belong to the Lower
Lyckholm. The lowest beds seen appear to be those-of Hohenholm
and Taps.

The best present collecting localities for the lower divisions of this
formation are Paope and Pallokiill, both on the island of Dago. At
these two places the quarries are in operation and new exposures are
constantly being made. Other localities in addition to those men-
tioned where the lower beds appear to outcrop are Neuenhof, Od-
dalem, Saximois, Sutlep, Forel, and Kirna.

Characteristic fossils of the Lower Lyckholm are Pseudolingula
quadrata, Porambonites gigas, Triplecia insularis, Subulites gigas, and
Salpingostoma dilatatum. The species which are known to occur in
the lower portion of the Lyckholm are :—

Calapoecia canadensis Billings.
Clathrodictyon cf. vesiculosum Nicholson and Murie.
Columnaria fascicula Kutorga.
Halysites catenularia (Linné).

escharoides (Lamarck).
Paleofavosites asper (d’Orbigny).
Proheliolites dubius Schmidt.
Streptelasma cf. corniculum Hall.
Syringophyllum organum Linné.
10. Anaphragma mirabile Ulrich and Bassler.
11. Ceramopora intercellata Bassler.
12. Chilotrypa immatura Bassler.
13. Corynotrypa abrupta Bassler.

eg et te oo be

14, barberi Bassler.

15: ~ dissimilis (Vine).
16. Dianulites colliferus Bassler.

Li. grandis Bassler.

18. suleatus Dybowski.
19. Dittopora colliculata (Eichwald).

20. Diplotrypa petropolitana Nicholson.
21. Glauconema plumula Wiman.
22. Hallopora elegantula (Hall).

- 23. Monticulipora dagoensis Bassler.
24. Nematopora fragilis Ulrich.
25. Orbignyella expansa baltica Bassler.
26. Orbipora fungiformis Eichwald.
27. Pachydictya bifureata (Hall).

_ ln te

TWENHOFEL: EXPEDITION TO THE BALTIC PROVINCES.

Ptilodictya flabellata Eichwald.
gladiola Billings.
Stellipora constellata Dybowski.
Stomotopora arachnoidea (Hall).
Atrypa imbricata Sowerby.
Clitambonites sinuatus (Pahlin).
verneuili (Eichwald).
Dalmanella elegantula estona (Wysogorski).
wimani Mickwitz.
Dinorthis solaris (von Buch).
Orthis actoniae Sowerby.
callactis Dalman.
concinna Lamansky.
flabellum Sowerby.
lyckholmensis Wysogorski.
oswaldi von Buch.
vespertilio Sowerby.
Platystrophia biforata lynx (Eichwald).
fissicostata (McCoy).
Plectambonites schmidti (Térnquist).
Porambonites gigas Eichwald.
Pseudolingula quadrata (Eichwald).
Rafinesquina deltoidea (Conrad).
Strophomena tenuistriata Sowerby.
Triplecia insularis (Eichwald).
Byssonychia cf. radiata (Hall).
Bucania contorta Eichwald.
cornu Koken.
crassa Koken.
eycloides Koken.
radiata (EKichwald).
Ectomaria kirnaensis Koken.
Eunema rupertre Eichwald.
sulcifera Eichwald.
schmidti, Koken.
Euomphalus carinifer Koken.
gradatus Koken.
laminosus Koken.
Holopea ampulacea Eichwald.
coronata Koken.
Isospira bucanoides Koken.

BULLETIN: MUSEUM OF COMPARATIVE ZOOLOGY.

Lytospira valida Koken.
Murchisonia exilis Eichwald.
insignis Eichwald.
Pleurotomaria chamaeconus Koken.
nodulosa Schmidt.
notabilis Eichwald.
notlingi Koken.
numismalis Koken.
plicifera Eichwald.
rotelloidea Koken.
Pycnomphalus borkholmensis Koken.
Salpingostoma dilatatum (Eichwald).
Sinuites bilobatus (Sowerby).
Subulites bullatus Koken.
gigas Eichwald.
inflatus Eichwald.
subulus Koken.

. .Worthenia aista Koken.

esthona Koken.
silurica (Eichwald).
vermetus Koken.
Cyrtoceras sphynx Schmidt.
Discoceras antiquissimum (Eichwald).
Orthoceras arcuolyratum Hall.
Arges wesenbergensis Schmidt.
Calymene stacyi Schmidt.
Chasmops eichwaldi Schmidt.
Cybele brevicauda Angelin.
Encrinurus multisegmentatus Portlock.
seebachi Schmidt.
Goldius laticaudus (Wahlenberg).
Harpes wegelinus Angelin.
Homolichas angustus (Beyrich).
Illaenus angustifrons Holm.
caecus Holm.
linnarssoni Holm
mascki Holm.
roemeri Volborth.
Platylichas hamatus Schmidt.
laxatus McCoy.
Amphilichas laevis Eichwald.
lineatus Schmidt,

.

TWENHOFEL: EXPEDITION TO THE BALTIC PROVINCES. 299

111. Prionocheilus pedolobum (F. Roemer).
112. Proetus kertelensis Schmidt.

113. ramisulcatus Nieszkowski.
114. Remopleurides emarginatus Tornquist.
115. Sphaerexochus angustifrons Angelin.
116. Sphaerocoryphe cf. granulata Angelin.
117. ‘Trinucleus seticornis Hisinger.

From some portion of the Lyckholm, and perhaps the lower part,
were derived the species whose names follow: —

Aulocoporella cepa (Roemer).

Aulocopodium aurantium Oswald.

Cyclocrinites spasski Eichwald (very doubtful).

Solenopora spongioides Dybowski.

Acantholithus astericus Roemer.

Alveolites? hexagona Schmidt.

Coccoseris micraster Lamansky.
ungerni Eichwald.

Halysites parallellus Schmidt.

10. undulatus Kiaer.

11. MHeleolites hirsutus Lamansky.

$2. inordinatus Sowerby.

13. interstinctus Linnaeus.

14. parvistellus Roemer.

15. Labechia conferta Edwards and Haime.

16. Lyellia bacillifera Lamansky.

ee SP ae eae

Li. conferta Edwards and Haime.
18. tubulata Edwards and Haime.
19. Petraia darcoceras Dybowski.

20. silurica Dybowski.

21. Protaraea cf. vetusta Hall.

22. Streptelasma europaeum Roemer.
23. Tetradium wrangeli Schmidt.

24. Glaphyrocystis compressa Jaekel.

20: wohrmanni Jaekel.

26. Hemicosmites grandis Jaekel.

Dk verrucosus Jaekel.

28. Aulacomerella angusta Huene.

29. macroderma (Eichwald).

30. Craniella? papillifera Huene.

300 BULLETIN: MUSEUM OF COMPARATIVE ZOOLOGY.

31. Dinobolus schmidti Davidson.

32. Eleutherocrania gibberosa Huene.

33. Orthis sadawitzensis Wysorgorski.

34. Pseudocrania cranoides Huene.

35. Pseudometoptoma concentricum Huene.

36. curvatum Huene.

37. monopleurum Huene.
38. Strophomena assmussi Verneuil.

39. luna Térnquist.

AO. semipartita Roemer.

41. Modiola devexa Eichwald.

42. incrassata Eichwald.

43. Cymbularia aequalis Koken.

44. Murchisonia scrobicula Koken.

45. spectabilis (Schmidt).
46. Conularia cf. trentonensis Hall.

47. Tentaculites anglicus Salter.

48. Cyrtoceras angulosum Schmidt.
49. Endoceras hasta Eichwald.

50. Orthoceras cuneolus Eichwald.

51. exaltatum Eichwald.
en fenistratum Eichwald.
5B5 ibex Eichwald.

54. Ceraurus cf. glaber Angelin.
55. Homolichas eichwaldi Nieszkowski.
56. Platylichas docens Schmidt.

The upper beds of the Lyckholm formation are best shown at Pirk
(Plate 2, fig. 2) and Saremois, the former about three miles northwest
of Herkiill and the latter hardly more than a mile in the same direction.
Pirk consists merely of a watermill and a few houses and the exposure
is in a small cliff on the right bank of the creek, just below the mill-
dam. About nine feet of yellowish white and white limestones of a
somewhat chalky or marly consistency are shown, the sequence con-
sisting of alternations of slightly different varieties of the same kind
of rock and the whole having a somewhat massive appearance with
illy defined bedding. Through sun and frost action it spalls off with
conchoidal fracture. Fossils are quite common, corals comparatively
rare, large gastropods are the most abundant and the Maclureas were
collected near the base.

Between Pirk and Kappa-Koil, at various localities on the higher

TWENHOFEL: EXPEDITION TO THE BALTIC PROVINCES. 301

lands and from one to five miles from the latter place, are outcrops of
thick-bedded dolomitic limestone with very poorly preserved fossils
and these chiefly corals. These strata may lie between the Lower
Lyckholm of Kappa-Koil and the strata of Pirk, or may be some
division of the Borkholm.

The exposures at Saremois, a part of the Herkiill estate, are cites
extensive, occurring in ditches which have been crisscrossed over a tract
of about forty acres. Most of the ditches are excavated through
alluvium, but in many places what appear to be buried knolls of the
country rock have been cut through. The lithology and fauna are
the same as at Pirk; but the strata are thought to be from ten to fifteen
feet higher in the section and they must lie almost immediately below
the lowest beds of the Borkholm formation, as the latter outcrop in
the woods between Herkiill (about a mile southeast) and Saremois,
and the vertical distance between the two outcrops amounts to but a
few feet. Corals are quite common at Saremois, while rather rare at
Pirk, due probably to the small size of the exposure at the latter place.

Strata with a lithology and fauna similar to that above described
also occur near the village Rannakiill, about six miles northeast of
Hapsal. Thin bands of gray shale separate the four to six inch beds
of yellowish white, somewhat marly-like limestone.

From these different localities have been collected the species whose
names follow. Those which are not followed by the authority also

occur in the Lower Lyckholm: —

Clathrodictyon cf. vesiculosum.
Halysites catenularia.
Paleofavosites asper.
Syringophyllum organum.
Dianulites grandis.
Diplotrypa petropolitana.
Graptodictya obliqua Bassler.
Atrypa imbricata.

Orthis actoniae.

1 aaa flabellum.

11. Platystrophia biforata lynx.
12. Rafinesquina deltoidea.

13. Byssonychia cf. radiata.

14. Bucania cornu.

15. crassa.

16. radiata.

ea Sigs a es Th ae

302 BULLETIN: MUSEUM OF COMPARATIVE ZOOLOGY.

17. Eunema (?) piersalense Koken.
18. Holopea ampulacea.

19. Maclurea neritoides Eichwald.
20. Murchisonia insignis.

21. Subulites subulus.

22. Worthenia aista.

23. Cyrtoceras sphynx.
24. Discoceras antiquissimum.
24a. Arges wesenbergensis.
25. Chasmops eichwaldi. ;
26. Cybele brevicauda. |
27. Encrinurus multisegmentatus. :
28. Goldius laticaudus. .
29. Harpes wegelinus.

30. Homolichas angustus.
31. Illaenus angustifrons.

ae linnarssoni.

oO; roemeri.

34. Isotelus platyrachis Steinhardt.
35. robustus Roemer.

36. Platylichas laxatus.

37. Proetus ramisulcatus.

38. Pseudosphaerexochus conformis.

39. roemeri Schmidt.
40. Sphaerocoryphe cf. granulata.

Out of the total of forty species there are but seven which do not
occur in the Lower Lyckholm and, so far as has been determined from
the field-study, the fauna is the same as that of the lower beds as dis-
played at Lyckholm, Hohenholm, Pallokiill and Paope.

From glacial boulders on Gotland Dr. Carl Wiman has collected ‘
fossils of Lyckholm age.' These boulders were no doubt derived 3
from strata outcropping beneath the Baltic and with the boulders
of Lyckholm limestone were others from the Borkholm. He has
collected the following Esthonian Lyckholm fossils, of which some, ;
however, may have come from Borkholm limestones :—

1. Acantholithus astericus.

2. Halysites catenularia.

a escharoides.

4, parallelus. ‘

1Wiman. Bull. Geol. inst. Univ. Upsala, 1902, 5, pt. 2, no. 10.

TWENHOFEL: EXPEDITION TO THE BALTIC PROVINCES. 303

5. Lyellia bacillifera.
6. Proheliolites dubius.
7. Atrypa imbricata.
8. Clitambonites sinuatus.
9 verneuili.
10. Dalmanella wimani.
11. Orthis oswaldi.
12. Platystrophia biforata lynx.
13. Strophomena assmussi.
14, semipartita.
15. Chasmops eichwaldi.
16. Isotelus robustus.
17. Proetus ramisulcatus.

Dr. Bassler from his faunal studies of Baltic Russian strata stated
that it appeared very probable that a great time break exists within
the Lyckholm formation, and “that the Wesenburg and early Lyck-
holm show affinities with the early Trenton, and that the upper
Lyckholm and Borkholm closely resemble certain divisions of the Rich-
mond group.”! He describes the Lower Lyckholm as a “ Magnesian
limestone holding Maclurea, Subulites and other gastropods related
to American Trenton species” ? and “The bryozoans of the lower
Lyckholm are, like those of the American Galena, distinctly Trenton
in character.” In his table of correlation he places the Lower Lyck-
holm as the equivalent of the Stewartsville dolomite of the Mississippi
Valley.2 In drawing his conclusions Bassler labored under the
extreme difficulty of never having seen the strata in the field and he
also had to accept the statements on the labels attached to the speci-
mens which he studied. .

Born‘ dissents from the conclusion of Bassler relating to a time
break in the Lyckholm and denies its existence. He also disagrees
with Bassler on the correlation of the Lyckholm; but as he fails to
definitely state its stratigraphic equivalent in the American section,
referring it to two possible horizons, this objection need not be further
considered. Schmidt® described the Lower Lyckholm as a white,
thick-bedded limestone, rich in silica and carrying few corals; while

1 Bassler. Bull. 77, U.S. N. M., 1911.

2 Bassler. Loc. cit., p. 9.

3 Bassler. Loc. cit., p. 17, 38.

4Born. 1913, Centralbl. min., geol., pal., 1913, no. 22, p. 714-19.

* Schmidt. Mem. Acad. sci. St. Petersb., 1881, ser. 7, 30, no. 1, p. 37.

304 BULLETIN: MUSEUM OF COMPARATIVE ZOOLOGY.

the upper part is a gray marly limestone which in places is full of
corals. Either division is readily distinguishable from the Wesen-
berg (Schmidt states that the Lower Lyckholm resembles the Wesen-
berg), for the latter is far more hard, more compact, and finer grained
and partakes more of the aspect of a lithographic stone. I dissent
from the general statement of Schmidt that few corals occur in the
Lower Lyckholm. Locally they are not uncommon, but there are
not nearly so many as in the Upper Lyckholm and the Borkholm.

Bassler apparently was misled by the labels accompanying his
specimens in considering that the Lower Lyckholm is characterized
by Maclurea and Subulites and in respect to the gastropods, he
appears to have reversed the sequence. Maclurea may occur in the
Lower Lyckholm; but I have not seen it there, while it is common
in the Upper Lyckholm. Subulites and large Hormotomas range
throughout, but the latter are more abundant in the upper beds. In
seeking for equivalents in American deposits, it is my Judgment that
no correlative value can be placed on these large gastropods of general
Middle Ordovician aspect. Bassler now agrees in this view, for since
his correlation of the supposed Lower Lyckholm he has learned of
numerous examples of the reappearance of faunas, and of the close
resemblance of the North American Black River and early Trenton
faunas to those of the Richmond.

Although Bassler placed considerable emphasis on the Bryozoa
in drawing his deductions relating to a time break, he appears to have
relied more on the general aspect of the fauna. In respect to the
Bryozoa it would be well to review those forms which he had in mind
as indicative of Middle Ordovician time; these are Corynotrypa bar-
bert, Diplotrypa petropolitana, Dittopora colliculata, Stellipora constel-
lata, and Stomotopora arachnoidea. Each of these species is said to
occur at Hohenholm, of which the present exposures are quite cer-
tainly the Lower Lyckholm and such was probably true of the older
ones as the topography does not lend itself to the view that strata
much higher than those now shown were formerly exposed. Ditto-
pora colliculata and Diplotrypa petropolitana also occur at Pallokiill
and the latter at Paope; but as they occur at these latter localities
with at least thirteen other species of the Lyckholm Bryozoa, consid-
ered as evidence of “Silurian affinity,” it follows that they must
almost wholly be neglected as evidence of age. Diplotrypa petro-
politana, moreover, is so long ranging that it has hardly any strati-
graphic value. Corynotrypa barberi and Stomotopora arachnoidea
occur at Hohenholm growing on Richmondian and Silurian corals

i
.
;
4
|

—E—— SS SS CULL Cl

TWENHOFEL: EXPEDITION TO THE BALTIC PROVINCES. 305

and Ptilodictya flabellata from the same locality is referred by Bassler
to the Borkholm.

Since it is really the lower beds which contain nearly all the Bryozoa
described by Bassler and few are from the upper beds, it must be
inferred that Bassler’s conclusion in so far as it relates to a time break
arose from a mislabeling of the specimens as to stratigraphic position,
and he “now agrees that not only the Upper Lyckholm is of Richmond
age, as stated in his work, but that the Lower Lyckholm also should
be so correlated.” ! ,

It appears very probable that the Lyckholm rests on the Wesenberg
or the Kegel in apparent conformability, although the Pallokiill
Krug and Pallokiill Chapel occurrences throw some doubt on this
conclusion. There is, however, a sharp lithic and a considerable
faunal change in passing from the Wesenberg to the Lyckholm. Of
the great array of Lyckholm corals there is nothing seen in the Wesen-
berg.

The species whose names follow are common to the Lyckholm and
underlying strata; but many of the gastropods and some of the other
Lyckholm forms are varietally different. The letter following the
name of a species refers to the name of the underlying formation in
which the species occurs. (See p. 293).

1. Cyclocrinites spasski K.
2. Dittopora colliculata W.
3. Diplotrypa petropolitana W.
4. Stomotopora arachnoidea E. Ku.
5. Plectambonites sericeus W.
6. Porambonites gigas W.
7. Rafinesquina deltoidea W.
8. Strophomena assmussi J. K. W.
9. Bucania contorta . W. |
10. cycloides | W.
zi, radiata W.
12. Isospira bucanoides ~ W.
13. Murchisonia exilis W.
14. insignis ee
15. Pleurotomaria notabilis W.
16. Sinuites bilobatus W.
17. Subulites inflatus W.

1 Personal communication.

306 BULLETIN: MUSEUM OF COMPARATIVE ZOOLOGY.

18. Subulites subulus W.

19. Worthenia esthona Ku. L.

20. silurica V.E. Ku. I. J. K. W.
21. Arges wesenbergensis W.

22. Cybele brevicauda 8

23. Encrinurus multisegmentatus W.

24. seebachi W.

25. Homolichas eichwaldi (not certain as to its occurrence below).
26. Illaenus linnarssoni W.

27. roemeri W.

28. Prionocheilus pedolobum W.

29. Sphaerexochus angustifrons W.

This is about fifteen per cent of the Lyckholm fauna, but several
of the species are long ranging and, hence, have no stratigraphic value;
most of the gastropods and some of the other forms are said to be
varietally different, and in the case of a few (Cyclocrinites spasski)
the identifications are uncertain. The number of common species of
chronologic value, therefore, becomes small, and these must probably
be looked upon as survivals from the older stage, as similarly occurs
in the Richmond of America, and the difference between the Lyckholm
and Wesenberg faunas argues for a time break of considerable impor-
tance, but one that does not appear to be of systemic value.

Borkholm formation. The eastern limits of this formation are near
Lake Peipus, the strata making their first appearance at the surface
near Pastfer and Miintenhof. Thence the outcrops of the formation
extend westward over a narrow band just south of the 59th parallel.
A few exposures, as the Nyby outlier, occur in Lyckholm territory.
Except for the little island of Wohhi to the east of Dago,! the forma-
tion does not appear to have been noted on the islands. The divi-
sion is probably not over twenty feet thick. The type-locality is in
the park at Borkholm, the estate of Herr von Rennenkampf. The
type-exposure is in a quarry, now no longer worked. According to
Schmidt, there were formerly about fifteen feet of strata shown;
but the present exposures are probably not so good as in earlier days,
since little more than twelve feet are now uncovered and few fossils
other than corals are obtainable.

The lowest beds are said to consist of a crinoidal limestone with a
thickness of about a foot. Above this lies a coarse-grained, yellowish
gray and white limestone which appears to contain few fossils other

1Schmidt. Bull. Acad. sci. St. Petersb., 1881, ser. 7, 30, no. 1, p. 39.

_——_——

|
:

TWENHOFEL: EXPEDITION TO THE BALTIC PROVINCES. 307

than corals. Brownish limestones with thin calcareous shale partings
succeed and at the top is a coral limestone, made up almost wholly of
these organisms. This coralline limestone, according to Schmidt
and Mickwitz, affords an excellent datum plane, which can readily
be followed throughout the entire district. Eastward it is said to
undergo dolomitization. About a mile west is a new quarry which
during the summer of 1914 was in operation, and where fossils are
more readily obtained. The lithology is about the same as the middle
portion of the Borkholm section and about six feet are exposed.

Near Herkiill, northwest of the dwelling and on the east side of the
post road, is a small quarry in the Borkholm. Here the basal ex-
posures consist of a fossiliferous, crystalline gray limestone with
mammillary upper surface. Then follow fifteen inches of black shale,
containing fucoidal impressions, and above this three feet of crystal-

line, coralline, and crinoidal limestone with many poorly preserved

coral heads. The beds of this limestone are from two to ten inches
thick.

The Herkiill locality made known by Schmidt lies southwest of the
dwelling, and the former exposures were made by a drainage ditch.
At the north end of this ditch Schmidt noted the occurrence of the
white Borkholm limestone in place. This is overlain by what appear
to be thin-bedded limestones and shales which carry an abundance of
fossils. No rock is visible in place, but the debris from the ditch has
been stirred by plowing and the fossils are readily collected. Both
fauna and lithology are similar to that of the locality described in the
next paragraph. Schmidt further stated that the contact between
the Silurian (Jérden Schicht) and the Borkholm could be seen here,
the Jorden strata appearing at the top of the ditch near its midlength.

_ Nothing of this is now visible.

The Nyby exposures, north of Hapsal, are the best that were seen
of the Borkholm formation. The quarries are situated on a low
terrace north of the residence of the Nyby estate, and are readily
found by their nearness to an old stone windmill.

The section exposed in these quarries, given from the summit
downwards, is as follows: —

1. Impure, dark gray, semicrystalline limestone in four to six inch
beds, separated by thin beds of gray shale. In places 50% of the
limestone consists of the stems of large crinoids. Four feet six inches.

2. Impure, dark gray, poorly crystalline limestone (beds four to
six inches thick) and thinner beds of gray calcareous shale. ‘Two feet.

Fossils are extremely common throughout, particularly corals

308 BULLETIN: MUSEUM OF COMPARATIVE ZOOLOGY.

belonging to the Heleolitidae, Calapoecia, Favosites, and Halysites.
Bryozoa are very abundant and there are also many brachiopods.
The fauna of the formation is as follows: —

1. (?) Aulocopodium aurantium.1

2. Acanthodus tubulus Dybowski.

3. Acantholithus astericus.

4. Calapoecia canadensis.

5. Clathrodictyon cf. vesiculosum.

6. Coccoseris microporus Eichwald,

7. Coelophyllum amalloides Dybowski.

8. Columnaria fascicula.

9. Cyathophyllum middendorfi Dybowski. ;
10. Endophyllum contortiseptum praecurser Weissermiill.
11. Favosites forbesi Milne Edwards.
12. Halysites catenularia.

13. escharoides.

14. parallelus.

15. undulatus.

16. Heleolites inordinatus.

17. Lyellia bacillifera.

18. cancellata Lamansky.
19. conferta.

20. Paleofavosites asper.

21. Pholidophyllum tubulatum Schlotheim.
22. Streptelasma elongatum Phillips.
23. europaeum.

24. Syringophyllum organum.

25. Hemicosmites tricornis Jaekel.
26. Chasmopora tenella (Eichwald).
27. Corynotrypa dissimilis.

28. Fenestella striolata Eichwald.
29. Glauconema plumula.

30. strigosa (Billings).
31. Hallopora elegantula.

32. Lichenalia concentrica Hall.

33. Nematopora lineata (Billings).
34. Pachydictya bifurcata.

35. Phaenopora ensiformis (Hall).

1 Species without authority also oceur in the Lyckholm.

TWENHOFEL: EXPEDITION TO THE BALTIC PROVINCES. 309

Protocrisina exigua Ulrich.
Pseudohornera orosa (Wiman).
Ptilodictya gladiola.
Sceptropora facula Ulrich.
Atrypa imbricata.

marginalis Dalman.

cf. nodostriata Hall.

undifera Schmidt.

Clitambonites sinuatus.

verneuili.

Discina gibbi Lamarck.

Eleutherocrania gibberosa.

Leptaena rhomboidalis (Wilckens).

Orthis actoniae.

oswaldi.

sadawitzensis.

Platystrophia biforata lynx.

Plectambonites schmidti.

sericeus (Sowerby).

Pseudocrania cranoides.

Pseudometoptoma concentricum.

monopleurum.

Streptis sp.

Strophomena expansa Sowerby.

luna.
tenuistriata.

Triplecia insularis.

Byssonychia cf. radiata.

Ectomaria nieszkowski Koken.

Eunema rupestre.

schmidti.

Euomphalus dimidiatus Koken.
gradatus Koken.
helicoides Koken.

Helicotoma superba Koken.

Murchisonia meyendorfi Koken.

Pycnomphalus borkholmemsis.

Trochonema minor Koken.
panderi Koken.
peraltum Koken.

Tryblidium esthonum Koken.

lindstrémi Koken.

310 BULLETIN: MUSEUM OF COMPARATIVE ZOOLOGY.

78. Worthenia borkholmensis Koken.

79. tolli Koken.

80. Tentaculites annulatus Schlotheim.
81. Discoceras antiquissimum.

82. Orthoceras calamiteum Portlock.

83. Calymene stacyl.

84. Encrinurus multisegmentatus.

85. Homolichas angustus.

86. Illaenus angustifrons depressus Holm.
87. roemeri.

88. Isotelus robustus.

89. Platylichas cicatricosus Loven.

90. margaratiferus Nieszkowski.
91. Proetus ramisulcatus.

92. Pseudosphaerexochus conformis.

93. roemeri.

94. Sphaerocoryphe cf. granulata.

95. Leperditia brachynoti Schmidt.

Other places in Esthonia where the Borkholm may be seen are Ruil
and Pastfer in the east; and Nomkiill, Kurro, Kerrofer, Affel, and
Noistfer in the west.

On Gotland Wiman obtained from glacial boulders scraped by ice
from Borkholm limestones, which probably lie beneath some parts of
the Baltic, the following Esthonian species :—

1. Acantholithus astericus.
2. Favosites forbesi.
3. Halysites catenularia.
4. escharoides,
5. parallelus.
6. Lyellia bacillifera.
7. Glauconema plumula.
8. Rhinidictya borkho!mensis.
9. Atrypa imbricata.
10. undifera.
11. Cltambonites sinuatus.
12. Dalmanella wimani.
13. Leptaena rhomboidalis.
14. Orthis oswaldi.
15. Platystrophia biforata.
16. Chasmops eichwaldi.

TWENHOFEL: EXPEDITION TO THE BALTIC PROVINCES. 31]

17. Isotelus robustus.
18. Platylichas cicatricosus.
19. Proetus ramisulcatus.

STRATIGRAPHIC AND FAUNAL RELATIONS OF THE BoRKHOLM FoRMA-
TION TO THE LYCKHOLM.

There are no reasons for believing that the Borkholm is other than
conformable on the Lyckholm and directly continuous thereto.
Furthermore, the faunas are essentially identical and at least fifty-
three of the ninety-five Borkholm species occur in the Lyckholm.
Several of these common species belong to the genus Halysites. I
believe it will ultimately be found necessary to unite several of these;
but this will not change the strong faunal similarity. Each formation
contains many species of Silurian aspect; this is shown to a greater
degree in the Borkholm, where Silurian forms are conspicuous and
abundant, but the Ordovician expression dominates.

TimME EQUIVALENTS OF THE LYCKHOLM AND BoORKHOLM.

At present these can be given in only a general way. When the
collections have been secured and submitted to a careful study itis
hoped to give a detailed correlation.

Considered as a whole the faunas bear an aspect not generally fami-
liar to American stratigraphers. The large Porambonites; the asso-
ciation of Halysites and members of the Heliolitidae with large
Hormotomas, Maclureas, and Subulites; and the abundance of large
Triplecias are faunal groupings not occurring in America, yet certain
associations are similar and the faunas have many components which
are present in American deposits, the writer being constantly im-
pressed with certain striking similarities.

The points of contact with American sections are greatest in num-
ber in the Anticosti formations,! where the facies and faunal associa-

1The Anticosti section is, figuratively speaking, on the frontier of the American Upper
Ordovician and Silurian, and hence would be most likely to show the greatest faunal similari-
ties to the strata of like age of northwestern Europe. The Anticosti section consists of eight
formations of which the lower four are Ordovician and the upper four Silurian. Named from
the summit downward, the formations are as follows.

Silurian Chicotte. Ordovician Ellis Bay
1223 feet Jupiter River. 939 feet. Charleton
Gun River. English Head

Becsie River. Macasty.

ae BULLETIN: MUSEUM OF COMPARATIVE ZOOLOGY.

tions are quite similar to those of Esthonia. The faunal elements
common to the two regions, however, are not so many as one could
wish.

A correlation is most readily made if the Borkholm formation be
taken as a point of departure. It has its closest faunal relations with
the Ellis Bay formation of the Anticosti section, where are found the
following identical or closely related forms:—

Calapoecia canadensis.
Clathrodictyan vesiculosum.
Halysites catenularia. .
Paleofavosites asper.
Protaraea vetusta (this form in Esthonia occurs in the Lyck-
holm only).
6. Zaphrentis affinis (a similar form in the Borkholm is called
Streptelasma elongatum).
7. Corynotrypa dissimilis.
8. Hallopora elegantula, var. nov.
9. Glauconema strigosa.
10. Nematopora lineata.
11. Phaenopora ensiformis.
12. Protocrisina exigua (Charleton, not Ellis Bay).
13. Ptilodictya gladiola.
14. Sceptropora facula (Charleton, not Ellis Bay).
15. Stomotopora arachnoidea (not in Borkholm, but Lyckholm).
16. Atrypa marginalis.
17. Clitambonites verneuili diversus (Shaler).
18. Platystrophia regularis Shaler (European type with two plica-
tions in the sinus and three on the fold).
19. Leptaena rhomboidalis.
20. Pseudolingula elegantula (Shaler) (P. quadrata in the Lyck-
holm).
21. Plectambonites sericeus.
22. Byssonychia sp. nov.
23. Calymene meeki Foerste.
24. Encrinurus multisegmentatus.

ST ete ea ae

This constitutes a total of twenty-four out of ninety-five species,
over twenty-five per cent of the Borkholm fauna, and coupled with
this, is the further fact that each witnesses the occurrence in great
numbers of a multitude of tabulate corals of Silurian aspect, which in

— s+. +:

| ‘

a ON ee Ss a eee

TWENHOFEL: EXPEDITION TO THE BALTIC PROVINCES. 313

each region make their first appearance in significant numbers in the
underlying formation. Considering that the Borkholm fauna has
not been carefully studied for correlation with North America, the
common occurrence of such a great number of species is rather remark-
able. If only the Bryozoa be considered it is found that eight of the
fourteen Borkholm species are also found in the Anticosti section, six
in the Ellis Bay and two in the Charleton. It is therefore probable
that when other phyla have been subjected to as careful comparative
study as have the Bryozoa, that a greater number of species will be
found common to the two regions. Since not so close faunal contact
or general expression is shown with any other Anticosti formation, it is
held as extremely probable that the Borkholm represents the whole or
a part of the Ellis Bay formation. The Lyckholm, hence, represents
some part of the Charleton formation. Large Subulites and Hormo-
tomas are found in the upper part of this formation in association with
Halysites and members of the Heliolitidae. There are, however, no
Maclureas. Common or similar species are as follows:—

1. Calapoecia canadensis.

2. Halysites catenularia.

3. Paleofavosites asper.

4. Streptelasma rusticum Billings (S. corniculum is said to occur in

Esthonia).

5. Zaphrentis affinis.

6. Corynotrypa dissimilis.

7. Nematopora lineata.

8. Protocrisina exigua.

9. Sceptropora facula (Borkholm).
10. Clitambonites verneuili diversus.
11. Plectambonites sericeus.

12. Pseudolingula elegantula.
13. Byssonychia sp. nov.

14. Sinuites cf. bilobatus.
15. Calymene meeki.

16. Proetus alaricus Billings.

Formerly I was inclined to believe that the Lyckholm and Borkholm

found their equivalents in the lower parts of the Ellis Bay and the

English Head and Charleton formations.!. This view is now modified
as stated above and it is believed that the English Head and perhaps

1Twenhofel. Bull. 3, Victoria memorial museum, 1914, p. 19.

S

314 BULLETIN: MUSEUM OF COMPARATIVE ZOOLOGY.

the Lower Charleton do not find representation in the Esthonian sec-
tion, since the English Head formation appears younger than the
Wesenberg and older than any part of the Lyckholm.

If the views previously expressed ! as to the chronologic position of
the Ellis Bay formation be correct, it follows that the correlation of
the Borkholm with the Ellis Bay formation leads to the conclusion
that the former is not represented in the known deposits of the North
American Interior. As the Charleton formation is certainly the
equivalent of the Richmond beds of the Interior, the conclusion follows
from the above correlation that the Lyckholm beds are of Richmond
age, and such the writer considers them, referring them with little
doubt to the upper portion of that formation. It may be noted that
Bassler correlated the Upper Lyckholm and the Borkholm with the
Richmond which he considers of early Silurian age. In the latter
view I can not follow him.

These two divisions have repeatedly been correlated with the Lep-
taena kalk and Brachiopod shales of Sweden and the Gastropoden
kalk and Meristella crassa zone of Norway and I now see no reasons
for dissenting from this view. Correlation has also been made with
the Caradoc or Bala of the British Isles, but detailed work is neces-
sary to determine the exact position. Its closest relations appear to
be with the Keisley limestone of northwestern England and the Chair
of Kildare limestone of Ireland.

THE SILURIAN SYSTEM.

Introduction. The most easterly observed occurrence of the
Silurian strata of Esthonia is near the village Pastfer to the north-
west of Lake Piepus. It thence extends westward, south of Herkiill
and Hapsal, forming the surface rock of the south end of Dago and
the whole of Moon and Oesel. On the mainland its outcrop forms a
strip about 40 miles wide from north to south.

Except that there is no appreciable departure from parallelism of
bedding, the stratigraphic relations of the lowermost division of the
Silurian to the Borkholm formation are not known. At no place was
the contact with the Ordovician seen. According to Schmidt, a
contact was formerly visible at Herkiill, but at present there is no

1 Twenhofel. Loc. cit., 1914.

|
:

Wt

TWENHOFEL: EXPEDITION TO THE BALTIC PROVINCES. 31

exposure of the lower beds of the Silurian at that place. As will be
shown later, however, there is an extensive time break at the base of
the Silurian.

The lithology of the Silurian is somewhat different from that of the
Ordovician, in that there is a greater degree of dolomitization of
the limestones, coralline limestone constitutes a larger proportion of the
rock, and, except in the higher beds, shales are far less conspicuous.

The faunas are quite different. The corals in many species are the
same as those of the Ordovician and some of the Borkholm Bryozoa also
appear to cross the system boundary; but the gastropods, pelecypods,
cephalopods, brachiopods, and trilobites are almost altogether different.

One can hardly fail to escape the conclusion that the Silurian de-
posits of Baltic Russia are those of quite shallow water and near a
shore. The horizontal variation of sediments, the extensive reef
deposits made by animals which must certainly have lived near the
surface, and the masses of shell breccia, as in the Borealis banks, can
hardly lead to another view. If this view be correct it follows that
there should be considerable horizontal lithic variation and this ap-
pears to obtain.

There are certain features of the faunas which are of interest. These
features arise from the fact that common species of other parts of the
Baltic are not found in Russia. The lowest beds of Gotiand, just a
short distance away, are characterized by an abundance of Strick-
landinia lirata (Sowerby), while Bilobites bilobus (Linné) of slightly
higher beds is an equally abundant Gotland fossil. Coelospira hemi-
spherica (Sowerby), is a common Middle Llandovery species of the
Kristiana region of Norway and of England and it also is not found in
the Silurian of Esthonia.

It is not clear how the absence of these species is to be explained.
One explanation would be that the equivalents of the Gotland beds
containing Stricklandinia lirata and Bilobites bilobus were never de-
posited in Baltic Russia, and that beds are also absent which are the
equivalents of those of the Kristiana region of Norway carrying
Coelospira hemispherica. Since the first two mentioned fossils belong
to different horizons in the Gotland section, the above explanation
leads to the assumption of two time-breaks in the Esthonian Silurian
at the places where these fossils should be present; that is, somewhere
within the Addifer and St. Johannis formations.

The preceding explanation is by no means satisfactory. Marine
animals have their habitats determined by a combination of an ex-
tremely great number of varying conditions, and the absence of the

316 BULLETIN: MUSEUM OF COMPARATIVE ZOOLOGY.

species mentioned may as well have been due to the absence of some
of the conditions that were necessary for their presence. This expla-
nation is in keeping with the fact that Bilobites bilobus is not universally
distributed in the Gotland division in which it occurs, while in Norway
Kiaer assigns it to different horizons in different facies.

The divisions of the Esthonian Silurian as determined by Schmidt !
are as follows, the highest division being named first.

Upper Oesel zone, stage K or 8, thickness, 50-75 ft.

Lower Oesel zone, stage I or 7, thickness, 100 ft.

Pentamerus estonus zone, stage H or 6, thickness, 50 ft.

Raikull beds | substage G3 or 5, thickness, 20 ft.
Borealis bank ; stage G ‘i G2 or 4, thickness, 10 ft.
Jérden beds J © G1 thickness, 20-30 ft.

For these names the following are proposed : —

Oesel formation for stage K.

St. Johannis formation for stage I.
Addifer formation for stage H.
Tamsal formation for stage G.

There are several reasons why stage K. should be subdivided; but,
as the outcrops are hardly sufficient to determine the exact sequence
and the relation of the upper subdivision to the lower, it is considered
wiser to leave the formation undivided. ‘The matter will be further
considered under the proper heading.

Tamsal formation. The three lowermost divisions of the Silurian
were grouped together by Schmidt, and there appear to be no good
reasons for changing the grouping. Where well defined, the Borealis
Banks are worthy of distinction; but as these shells appear to enter
into both the strata above and below, it would appear unwise to make
a formational separation from either. At no place can the three
divisions be seen together.

(a) Jorden zone. The Jérden zone makes its most eastward ap-
pearance just west of Lake Peipus and thence can be traced to the
sea-coast a short distance south of Hapsal. Of the islands, Dago
appears to be the only one on which it occurs. Since no section is
known showing the contact with both the Borkholm and the succeed-
ing Borealis banks, its thickness is quite uncertain. Schmidt?
estimated that it may lie between twenty and thirty feet.

1Schmidt. Quar. journ. Geol. soc. London, 1882, p. 525.
2Schmidt. Loc. cit., 1882, p. 526.

ee ee here |

ee ee a eee

ee ee ee

TWENHOFEL: EXPEDITION TO THE BALTIC PROVINCES, 317

The type-section of this zone is exposed in an old quarry near and
across the road from Jérden church (Plate 3, fig. 1). Only about
seven feet are shown. At the base are four feet of mottled gray and
red dolomitic limestone and this is overlain by three feet of the Borealis
bank. No fossils were seen in the Jérden strata at this point.

The zone was again seen about three fourths of a mile northwest of
Tamsal and also southeast of that village. West of Tamsal about two
feet are exposed and there is perhaps a little more in the woods to the
southeast of the village. At Podrang, south of Tamsal, it is probable
that the basal portion of the thirteen feet which are shown there may
‘belong to the Jérden. The rock at these various places consists of a
somewhat reddish, coarse-grained limestone which is not very fossili-
ferous.

The division was also seen near Hapsal on the Weisenfeld estate,
about a mile west of the dwelling. The quarry is quite large and the
exposures are particularly good. About six feet are exposed of which
the lower three feet belong to the Jérden zone and consist of heavy-
bedded (beds four to six inches thick) gray, crystalline limestone con-
taining Favosites, Clathrodictyon, and rarely Pentamerus borealis
Eichwald. Above lie three feet of thin-bedded, yellowish, semi-
crystalline limestone which contains a great abundance of Favosites,
species of the Heliolitidae, Halysites catenularia, Clathrodictyon, and
entire and broken Pentamerus borealis. On the island of Dago the
outcrop of Puhhalep probably belongs to this division. The expo-
sure consists of from four to five feet of coarsely crystalline, thin-
bedded (beds 1-3 inches), dolomitic limestone which contains little
other than corals. Other outcrops on Dago are those of Grossenhof
and Kallasto.

The fossils of this division are few in number and generally not well
preserved. They have not yet been studied by the writer and the
species listed with few exceptions are taken from others. Near the
base is a small fauna which promises much of interest, but until care-
fully studied nothing will be said concerning it. Schmidt gives as the
fossils most characteristic of the Jorden beds, the following species: —

Cyathophylloides kassariensis Dybowski.
Densiphyllum thomsoni Dybowski.
Donacophyllum losseni Dybowski.
schreneki Dybowski.
Favosites gothlandicus Lamarck.
Halysites agglomeratus Eichwald (= H. catenularia?).

Soe Pa

318 BULLETIN: MUSEUM OF COMPARATIVE ZOOLOGY.

7. Heliolites interstinctus.

8. Paleofavosites aspera.

9. Atrypa imbricata var.
10. Coelospira duboysi (Verneuil).
11. Dinobolus davidsoni Salter.
12. Orthis bouchardi Verneuil.
13: davidsoni Verneuil.
14. Rhynchonella aprinis Verneuil.
15. Schuchertella pecten (Linné).
16. Encrinurus punctatus Wahlenberg.
17. Proetus cf. distans. Lindstrém.
18. planedorsatus Schmidt.
19. Leperditia schmidti Kolm.

There have also been recognized Clathrodictyon vesiculosum, Favosites
forbesi, Halysites catenularia, and Pentamerus borealis.

(b) Borealis banks. This is one of the most remarkable deposits
of fossil shells I have ever seen. They are almost wholly Pentamerus
borealis, generally single valves, although in the western outcrops it is
possible to obtain entire shells.

Respecting the continuity of this zone it is not possible to speak
with certainty; but the general facts of the distribution of marine
organisms would lead to the conclusion that it is not of wide extent
and it seems very probable that it thins westward. It also appears
quite probable that in places the character of deposition which gave
rise to the Jérden beds may have persisted upward while the shell-
banks grew around, or within such places, that is, the banks are of a
more or less local distribution.

The zone makes its first appearance west of Lake Peipus near St.
Simon’s church, and, narrowing toward the east, the belt of its out-
crop reaches the coast a few miles south of the city of Hapsal, and finds
its further continuation on the island of Dago. Schmidt, (Loc. cit.,
1882, p. 526) gives the thickness as forty feet.

Where the division is typically shown little else is to be seen other
than myriads of Pentamerus borealis and a few corals. The number
of the former is simply inconceivable. The rock is wholly a mass of
the shells, which are chiefly preserved as casts. Where the shells are
not the dominant component, the rock is either a limestone or a dolo-
mite.

About the best locality to see this division, and the one which has
been selected as the type-locality of the Tamsal formation, is north-

TWENHOFEL: EXPEDITION TO THE BALTIC PROVINCES. 319

west of the village and railroad station, Tamsal. The rock is there
quarried for lime, an extensive plant being in operation. This is
taken as the type-locality of the zone, since Schmidt gives none, as
well as the type of the formation. About twelve feet of strata are
visible here, the basal two feet belonging to the Jérden beds and the
upper ten feet to the Borealis banks.

The rock is a gray and yellowish gray Pentamerus borealis breccia.
No two valves were seen together and little else can be seen. Clath-
rodictyon is the only other common fossil.

The rock at the Weissenfeld locality previously described (p. 317)
is not so exclusively composed of Pentamerus borealis, and entire shells
are not uncommon. At Helterma, on the island of Dago, the beds
outcrop on the beach and, as at Weissenfeld, the rock is not so com-
pletely a mass of Pentamerus, although parts are fully as much so.
Syringophyllum organum, Clathrodictyon vesiculosum and other corals

are not uncommon here. This outcrop may represent the upper

portion of the division. The Borealis banks are also shown by eight
or nine feet in the Podrang quarry near Tamsal, and there the shells
do not make such a great proportion of the beds.

From the localities cited it appears probable that only the lower
portions of this zone have been seen, unless the exposures at Helterma
lie above the base. Schmidt was inclined to believe that the Borealis
banks lose their character as a unit upon the island of Dago, reaching
this opinion from finding P. borealis in the Jérden beds at Kallasto.
This is by no means certain, as this shell also occurs in the Jérden beds
at Tamsal, and in some places it should be more abundant than else-
where in this basal division, and perhaps this is the case at Kallasto.
As a shell breccia, the division probably has only a local development.

The fauna of this division has never been carefully studied. It is
probably not large, since the conditions of fossilization were not such
as to lend themselves to the preservation of delicate forms. The
omnipresent fossil is Pentamerus borealis. Syringophyllum organum,
Favosites gothlandicus, Clathrodictyon vesiculosum, and small bryozoans
occur more or less rarely.

(c) Ratkiill beds. This division is the thickest of the three belong-
ing to this formation, Schmidt stating that 100 feet may be present.
It appears to be continuous with the Borealis banks, and receives its
name from Raikiill, the estate of Count Keyserling. In general, the
rock is a coralline limestone, in some places completely dolomitized,
in others a mass of corals, and in still others well-bedded crystalline
limestone. The division makes its appearance about twenty miles

320 BULLETIN: MUSEUM OF COMPARATIVE ZOOLOGY.

west of Lake Peipus, reaches the coast south of Hapsal and forms a
part of the south end of Dago. In the east it outcrops over an area
about thirty miles wide from north to south, but south of Hapsal the
belt of its outcrops is only a few miles wide.

The quarry of the type-locality is in a forest about three miles north-
west of the Raikiill residence, and is reached by a road branching off
from the main road a little to the west of the entrance to the estate
(Plate 3, fig. 2). About ten feet are exposed. The lower eight feet
consist of well-bedded, probably dolomitic limestone with the beds
varying in thickness from three to eight inches and these are exten-
sively quarried for construction purposes. Very few fossils are
present. The upper two feet consist of thin-bedded rough limestone
in which are many corals, chiefly Favosites gothlandicus, Halysites
catenularia, and Clathrodictyon vesiculosum.

Certain beds of this zone are much more extensively exposed near
the village Lippa, situated about three miles south of Raikiill. The
quarry covers an acre or two and appears to have been continuously
operated for many years, but only about four feet of strata are ex-
posed. The rock is a hard, white, crystalline limestone with which is
interstratified a little softer, almost microcrystalline limestone of the
same color. The former is dolomitic, and both are in beds from three
to six inches thick. At the top are myriads of corals, many of which
are silicified. They are not uniformly distributed through the rock,
but are aggregated in patches and the species are the same as men-
tioned for the Raikiill locality and, in addition, many Heliolitidae are
also present. A layer of the softer crystalline limestone is particularly
characterized by numerous, fine, large specimens of Leperditia keyser-
lingt Schmidt.

A more extensive exposure of this division is near Weissenstein on
the Miintenhof estate, where the beds exposed reach a thickness of
twenty feet. The lowest fifteen feet consist of well-bedded bluish
and yellowish crystalline dolomitic limestone in which only obscure
fossils were seen and these were near the top. The beds are from
four to six inches thick, and one about two feet below the top contains
flattened mud pebbles with the horizontal diameters reaching an inch
and the vertical diameters about a fourth as great. Overlying these
strata are from five to six feet of cavernous yellow dolomite with
gnarled structure and a general absence of bedding. This is probably
an old coral reef and it still contains numerous poorly preserved
specimens of Favosites and Clathrodictyon. Many of these are
merely skeletons and after the exterior has been broken, crumble on
being touched.

%
%

J
:
:
B
f
|

TWENHOFEL: EXPEDITION TO THE BALTIC PROVINCES. 52]

On the coast south of Hapsal, near the village Pasko, are extensive
exposures of a white crystalline limestone. The outcrops are on the
beach, and the rock appears to have been formed from an old coral
reef. Well-preserved coral masses strew the beach and have con-
tributed a great deal of material to the many stone walls of the ad-
joining lands. These are the most fossiliferous exposures of this
division that were seen. The corals consist of Favosites, Heliolitidae,
Halysites, Clathrodictyon, eo aa Acervularia and several
species of horn corals.

Other localities where the Raikiill beds are exposed are Mexhoff,
Piometz, Endama, Allenkiill, Teknal, Wieso, Wédja, Tenja, and the
island Kassar to the south of Dago.

Besides the corals mentioned as being present in the Raikiill beds,
there are the trilobites Encrinurus punctatus, Illaenus livonicus Holm,
Phacops elegans Sars and Boeck, Proetus planedorsatus; the bryozoans
Vineularia megastoma Eichwald and V. nodulosa Eichwald. Brachio-
pods appear to be quite rare. At Raikiill and Wahlokiill Schmidt
states the occurrence of the graptolite Diplograptus estonus Schmidt !
and at the latter place Deiphon forbesi Barrande.

Addifer formation. This formation was designated by Schmidt
the Pentamerus estonus beds by reason of the great numbers of that
brachiopod which are commonly present. The section exposed in the
quarry at Addifer, the estate of Herr E. von Wahl, is taken as the
type.

The outcrops of the Addifer formation begin about twenty miles
west of Lake Peipus near Pedja Brook and reach the sea south of
Hapsal. It has not been identified on the islands, but may underlie
the extreme southern end of Dago, and probably does underlie Soela
Sound between the islands of Dago and Oesel. The thickness of the
formation has never been accurately determined, but it appears that
not more than fifty feet are present.

The formation is separated from that underlying, chiefly on the
basis of the introduction of Pentamerus estonus Eichwald, Atrypa
reticularis (Linné) and LEospirifer radiatus (Sowerby). So far as
known, its stratigraphic relations to the Raikiill beds and St. Johannis
formation are those of conformability.

The designation, Pentamerus estonus beds, does not carry the idea
that the rocks are largely composed of the shells of one species, as is
the case in the P. borealis beds. In this formation the pentamerids

1 Mem. Acad. imp. sci. St. Petersb., ser. 7, 30, no. 1, p. 44.

-

oe BULLETIN: MUSEUM OF COMPARATIVE ZOOLOGY.

lie in colonies at various levels, in some places composing the whole
of the rock and in others only a few or none being present.

The greater portion of the rock of the formation appears to be com-
posed of a yellowish white, fine-grained limestone which is only partly
crystalline; or a coarse-grained, quite crystalline, dolomitic limestone.
At some levels or localities there is also a little shale. Schmidt
(Loc. cit., 1882, p. 526) states that in the east dolomites prevail,
while in the west the lithology is one of coralline limestone. In a
general way this conclusion appears justified. Also, in the eastern
limits, either the upper beds of this formation, or the lower beds of the
succeeding, consist of unfossiliferous sandstone which Schmidt at first
considered Devonian; but he later learned that the sandstone is
overlain by the fossiliferous beds of the St. Johannis formation which
necessitated his referring the former to the Silurian.?

Addifer, the type-locality, is in the eastern area. The exposures are
in an old quarry which has recently been reopened. About ten feet
are shown of which the lower six feet are typical Pentamerus estonus
limestone. The brachiopods are not uniformly distributed, but occur
in patches and in these they are extremely abundant. Many are
silicified, but others show no microscopic trace of silica. Flint nodules
are very abundant and many are several inches long. The floor of
the quarry is composed of a coarse-grained, quite crystalline, dolomitic
limestone. How thick this is, was not learned. The upper four feet
of the quarry section consist of a whitish yellow, fine-grained, partly
crystalline limestone, which, except for a few poorly preserved corals,
appears to be without fossils. It contains a few white flint nodules.
Near one of the barns about the residence of this estate a ditch has
recently been constructed, and in it the same sequence as in the quarry
is shown.

East of the railroad station, Wochma, about fifteen miles from
Addifer, are the old quarries of Koksfer, and Arrosaar. They are
largely grassed over, but a few of the old dumps permit collecting.
The rock appears to be a coarsely crystalline dolomitic limestone.
Fossils seem to be rare, consisting chiefly of Halysites, Cyathophyllum,
Favosites, and Zaphrentis. Pentamerus estonus was not seen. Many
yellow or brownish yellow flint concretions are present. These strata
are higher than those of Addifer.

At Pajus, a few miles east of Addifer is a small exposure of about
five feet of soft yellow limestone containing no observed fossils, but a

1Schmidt. Loc. cit., 1881, p. 46.

TWENHOFEL: EXPEDITION TO THE BALTIC PROVINCES. 323

considerable number of flint concretions occur. Another quarry at
Pajus has an exposure of about eight feet of finely crystalline hard
limestone in beds up to eight inches thickness. A few poorly preserved
corals were the only fossils seen. These two exposures are near the
base of the Addifer formation, and they may belong to the Raikiill
beds.

At Nudi, about five miles south of Raikiill, and on lower ground, is
an old abandoned quarry, once extensively worked. About four feet
of horizontal strata are exposed of which the basal two feet consist of
thin-bedded, gray crystalline limestone with thin shale partings, and
the upper two feet of thin-bedded, poorly crystalline white limestone.
Fossils are quite common throughout and in some layers they are very
abundant. Altrypa reticularis, Pentamerus, like P. oblongus from the
Clinton of New York, and Eospirifer radiatus are those most common.
These strata probably are situated quite near the base of the formation.

The formation is excellently exposed at Kattentack near the village
Turpel. About four feet of thin-bedded limestones are exposed in
which corals are the chief fossils, these consisting of Favosites, Lyellia,
Halysites, and Zaphrentis. Large Pentamerus estonus and Orthis ef.
callagramma Dalman are also not uncommon.

Other localities where the strata of the formation can be seen are
Kerro, Fennern, and Keskfer.

The fauna of the Addifer formation is evidently quite large, but has
not been well studied or collected. Important species are: —

Alveolites labechi Milne Edwards.
Darwinia speciosa Dybowski.
Favosites gothlandicus.

hisingeri Milne Edwards and Haime.
Halysites catenularia.

distans Eichwald.

exilis Eichwald.
Syringopora bifurecata Lonsdale.
Vincularia megastoma.
10. nodulosa.
11. Atrypa reticularis.
12. Eospirifer radiatus.
13. Pentamerus estonus.
14. Schuchertella pecten (Linné).
15. Strophomena euglypha Dalman.
16. Orthoceras canaliculatum Sowerby.

he ge eee ee ge eh ae

324 - BULLETIN: MUSEUM OF COMPARATIVE ZOOLOGY.

17. Leperditia abbreviata Schmidt.
18. Calymene frontosa Lindstrom.
19. Encrinurus punctatus.

20. Goldius estonicus Schmidt.

21, cf. marklini Angelin.
22. Illaenus livonicus Holm. ‘
23. proles.

24. Phacops elegans.
25. Proetus concinnus Dalman.

St. Johannis formation. By Schmidt this formation was desig-
nated the Lower Oesel zone, or stage I. It makes its first appearance
on the mainland just west of the village of Fennern, where it overlies
the Addifer formation and appears from beneath the Devonian. It
forms the surface rock of the north end, and probably most of the
central and southern parts of Moon, and all of the northeastern half of
Oesel. The exposures of the mainland are negligible; but on the
islands there are many and no difficulty need be experienced in ob-
taining a representative collection. Several.of the best localities are
the island of Schildau between Moon and the mainland, St. Johannis
on the northeast corner of Oesel and Mustel or Panga Pank (cliff) on
the northwest corner. It is the only Silurian formation of this region
of which a nearly completely continuous section is shown.

The formation has a thickness of about 100 feet, consisting of a
variation of dolomites and limestones interbedded with highly cal-
careous shales. As has been previously noted, perhaps there are beds
of sandstone in the lower part of the formation in its eastern exposures.
In the western areas many of the beds are markedly dolomitic, and
some of the dolomitic beds carry hardly any fossils.

Nearly the entire formation is shown at Mustel Pank on the north-
west coast of Oesel. At sea-level (Schmidt, 1858, p. 60) are beds of
limestones which are chiefly composed of corals. These are succeeded
by gray or blue calcareous shales rich in fossils. There are probably
not more than ten feet of these. Above follow compact gray lime-
stones with shale partings, and this zone traced horizontally appears
to pass over into gray or yellow, often porous, crystalline dolomites.
Somewhere in this Mustel Pank section is said to be the horizon of
the occurrence of Schidosteus mustelensis Pander.

One of the most westerly of the good fossil localities, and one of the
few of the mainland was Kerkau which Schmidt ! considered to lie at

1Schmidt. Loc. cit., 1858, p. 61.

—
bo
ee |

TWENHOFEL: EXPEDITION TO THE BALTIC PROVINCES.

about the same horizon of the marl beds at St. Johannis. At that
place the rock is said to be quite dolomitic. The exposure is no
longer in existence. Fossils which have been collected there are: —

Eospirifer radiatus.

Leptaena rhomboidalis.

Meristina tumida (Dalman).

Platystrophia biforata lynx.

Spirifer crispus Dalman.

Orthoceras canaliculatum:

Calymene tuberculata Brunn.

Encrinurus punctatus.

Oncholichas ef. gothlandicus Angelin.
ornatus Angelin.

Proetus concinnus osiliensis Schmidt. .

See Ore SS Ibo oboe

pod ped

At Leal, on the western end of the mainland, the once extensive
quarries are now to a considerable extent overgrown. The rock is a
fine-grained dolomite which becomes yellowish white on exposure.
Originally there may have been twenty feet exposed, but at present
not more than one third of this is shown. Few fossils appear to be
present.

At Kuiwast on the eastern side of the Island of Moon, a brownish
gray to yellowish white, fine-grained crystalline dolomite outcrops on
the beach. At most levels it has a gnarled structure and appears to
be largely composed of coralline material, although the fossils have
been completely destroyed. In a few places the bedding is regular.

A similar rock, but with bedding in some places well defined, out-
crops at many localities on Moon. At the base of Igo Pank and
Pussininna Pank occur soft, bluish gray and yellow, not well or finely
crystallized dolomites with well-defined bedding. The overlying
rocks are cavernous dolomites which, being more resistant to wave-
erosion than the softer underlying strata, overhang the latter in many
places, producing a wild and picturesque coast. Fossils are quite
rare in the rocks at the top of the cliffs, but in the softer beds beneath
they are more common. The former beds are probably the same as
those of Kuiwast, and they appear to compose most of the surface
strata of the northern interior of Moon.

Fossils from the basal beds of the two Panks mentioned are: —

1. Hindia fibrosa Roemer.

2. Halysites catenularia.

326 BULLETIN: MUSEUM OF COMPARATIVE ZOOLOGY.

3. Ptychophyllum patellatum Schlotheim.
4. Atrypa reticularis.
5. Camarotoechia bidentata (Hisinger).
6. Dalmanella elegantula.
7. Dinorthis rustica osilensis (Schrenck).
8. Leptaena rhomboidalis.
9. Spirifer crispus.
10. Strophomena euglypha.
11. Wilsonia wilsoni (Sowerby).
12." Omphalotrochus discors (Sowerby).
13. Orthoceras canaliculatum.
14. Encrinurus punctatus.

At Kaggowasar on Moon have been collected Acidaspis marklini
Angelin, Bumastus barriensis Murchison, and Proetus cf. P. verrucosus
Lindstrém.

At Orrasaar on the east coast of Oesel, about 150 yards north of the
post station, occur many blocks of a marly semicrystalline, yellowish
gray dolomitic limestone which contains numerous small fossils. The
blocks were probably pushed up by the ice from beneath the sound
between Moon and Oesel, and appear to have been derived from about
the same horizon as the rock at the base of Igo Pank. From these
blocks have been obtained: —

Lyellia tubulata Lonsdale.

Favosites gothlandicus.

Ptychophyllum patellatum.

Wilsonia wilsoni.

Omphalotrochus funatus.
sculptis.

Pleurotomaria alata Wahlenberg.

Phragmoceras pyriforme (Sowerby).

Orthoceras canaliculatum.

10. Beyrichia klodeni McCoy.

11. Calymene tuberculata.

12. Encrinurus punctatus.

13. Proetus concinnus osiliensis.

Be Th ea ee ee ee

On the island of Schildau between Moon and the mainland the same
sequence appears to obtain as on Moon, but somewhere in the lower
beds there is a level containing many fine fossils similar to those of
the locality at St. Johannis described in the next paragraph. I did

ite i i et ee i |

=

TWENHOFEL: EXPEDITION TO THE BALTIC PROVINCES. 327

not visit this island, but saw a collection from these rocks in the
cabinet of Herr E. von Wahl.

The best locality for excellent fossils of this formation is without
doubt the one which is taken as the type. This is near St. Johannis
church on the northeast corner of Oesel. About a mile southeast of
the church the rock is a compact, crystalline dolomitic limestone with
hardly any fossils. Beneath this lies a somewhat softer dolomitic
limestone in which are many large corals, belonging to Clathrodictyon,
Favosites, Halysites, and Syringopora. These beds probably corre-
spond to some part of the upper beds of Mustel Pank and Moon.
Underneath are the strata of St. Johannis church, consisting of white,
porous, marl-like limestones. Directly at St. Johannis the rock is not
seen in place, but blocks are not uncommon on the shore, and about a
mile northwest it forms a cliff. It yields readily to the action of water
and, since the fossils are more resistant than the inclosing rock, they
have not been destroyed; but concentrated on the beach. Where ex-
posed northwestward there is a cliff about ten feet high with the marly
limestones quite thick-bedded and bedding not well defined. Fossils
are not so common as at St. Johannis, as here there has been no con-
centration.

There is no place in all the Russian Baltic where fossils are so
abundant and so beautifully preserved. Species which have been
collected at St. Johannis are: —

1. Hindia fibrosa.
2. Campophyllum irregulare Dybowski.
3. Cystiphyllum cylindricum Lonsdale.
4. Favosites gothlandicus.

5. Halysites escharoides.

6. distans.

rs exilis.

8. Ptychophyllum patellatum.
9. Vincularia nodulosa.

10. Atrypa reticularis.

11. Camarotoechia bidentata.

12. Cyrtia exporrecta (Wahlenberg).

13. Dalmanella elegantula.

14. Dinorthis rustica osilensis.

15. Eospirifer radiatus.

16. Meristina tumida.

17. Leptaena rhomboidalis.

328 BULLETIN: MUSEUM OF COMPARATIVE ZOOLOGY.

18. Rhipidomella hybrida (Sowerby).
19. Plectambonites transversalis (Dalman).
20. Spirifer crispus.

21. Schuchertella pecten.

22. Strophomena euglypha.

23. Wilsonia wilsoni.

24. Omphalotrochus funatus.

25. sculptis.

26. Cornulites vagans Schrenck.

27. Tentaculites ornatus Sowerby.
28. Orthoceras annulatum Sowerby.
29. canaliculatum.
30. Beyrichia klodeni.

31. Acidaspis marklini Angelin.

32. Bumastus barriensis.

33. Calymene tuberculata.

34. Cyphaspis elegantula Angelin.
35. Encrinurus punctatus.

36. Proetus concinnus osiliensis.

37. Aulacodus obliquus Eichwald.

Numerous crinoid stems and at least four additional species of
Bryozoa are also present.

Strata of this same horizon are said to make their appearance on
the small island of Keinast between Oesel and Moon, and Schmidt
reports Palaeocyclus porpita (Linné) and Caryocrinus ornatus Hall to
have been collected there. It is extremely probable, however, that
the latter identification is incorrect. Farther westward on Oesel at
Surika Pank and Hundwa Pank, the lithology of the formation be-
comes less dolomitic. As a consequence, fossils are better preserved
and, among others, Schmidt notes the occurrence of Thecia swinder-
nana Goldfuss, Eridophyllum rugosum Milne Edwards and Haime,
Camarotoechia borealis (Schlotheim), Rhynchotreta cuneata (Dalman),
and Leperditia baltica Hisinger.

Oesel formation. This represents the Upper Oesel group of Schmidt
and with it the Silurian of Esthonia ends. The surface exposures of
the formation appear to be confined to the southwestern side of Oesel,
but there is a possibility that it finds representation among the higher
rocks of Moon. It appears to be a direct continuation of the St.
Johannis formation and, since there is no appreciable variation of dip,
the stratigraphic relations are considered conformable. Probably

a

TWENHOFEL: EXPEDITION TO THE BALTIC PROVINCES. 329

between fifty and seventy-five feet are represented, consisting of a
complex of dolomites, limestones, and shales. In the interior of Ogesel
to the northwest of Arensburg, the rocks are chiefly yellow erystalline
limestones, and dolomites and a similar lithology obtains on the
northern end of its outcrops on the western shores, to as far south as
Sarepa.

On the south coast to the east of Arensburg, the strata consist of
heavy-bedded limestones and shales, and the lithic facies on the
Sworbe to the southwest of the city is very like that of the southern
coast, although there is not so much of the massive limestone, and
calcareous shales appear to haye a larger representation. These
strata are younger than those described in the preceding paragraph.

Since there are two lithic units in this division it may ultimately
prove necessary to give formational rank to each. I have hesitated
to do this because of the apparent impossibility of determining a basis,
or locating a plane of division. The latter, however, is a common and
almost universal difficulty throughout the Esthonian region. None
has been determined between the St. Johannis and this formation,
but the latter is assumed to begin with the appearance of Whitfieldella
didyma (Dalman), W. prunum (Hisinger) and the eurypterids. For
convenience, however, the two divisions of this formation will be
described as the Sagaristi and Kaugatoma zones, after the localities
which are considered typical of the lower and upper divisions.

(a) Sagaristi zone. The most northern and lowest exposures of
the interior are those of Sagaristi, Uddafer, and Ladjal. At Sagaristi
not more than four feet of strata are exposed, consisting of interstrati-
fications of coarsely crystalline gray and yellow limestones (beds one
to three inches thick), white to yellowish white poorly or finely
crystalline dolomitic limestones and thin calcareous shales. The
dolomitic limestones appear to be much used for flags and building
stones. Fossils are extremely common in patches, but the fauna is

dominated by the abundance of Whitfieldella didyma and Rhipi-

domella hybrida. The former in some places occurs by hundreds.
The same strata with a similar lithology are well exposed in quarries
extending from Uddafer to Ladjal, but the thickness shown is con-
siderably greater. The following fossils have been collected from
these localities: — .

1. Clathrodictyon vesiculosum.
2. Favosites gothlandicus.
3. Laceripora cribrosa Eichwald.

330 BULLETIN: MUSEUM OF COMPARATIVE ZOOLOGY.

4. Syringopora reticulata Hisinger.
5. Whitfieldella didyma.
6 prunum.
7. Rhipidomella hybrida.
8. Murchisonia cingulata Hisinger.
9. Platyschisma helicites (Sowerby).
10. Pleurotomaria undata Sowerby.
11. Holopella obsoleta Sowerby.
12. Jlionia prisca (Hisinger).
13. Pterinea retroflexa Wahlenberg.
14. Orthoceras imbricatum Wahlenberg.
15. Leperditia baltica.
16. Encrinurus? obtusus Angelin.
17. Eurypterus fischeri Eichwaldi.

The exposures northwest of Arensburg and on the west coast are
rarely extensive, consisting of small quarries in the interior and a few
low cliffs on the coast.

Padel, about nine miles from Arensburg and on the road to Rotzikiill,
has an exposure of fully four feet of very finely crystalline limestone
which in fresh exposures is probably thick-bedded, but under the
action of sun and frost the rock separates into thin slabs. The fauna
is the same as that at Sagaristi, and Whitfieldella didyma is the most
common fossil. About a mile west of Padel is the Koggul quarry.
The elevation is in the neighborhood of ten feet higher and the beds
may be a little higher stratigraphically, but there is no certainty re-
garding this point. About ten feet of thick-bedded, chocolate colored
limestone are exposed. Fossils are not abundant; but large omphalo-
trochoid gastropods and small W hitfieldella didyma are not uncommon.
Other fossils from here are Crotallocrinus rugosus Miller, Aviculopecten
danbyt (McCoy), Goniophora cymbaeformis Sowerby, Megalomus
gothlandicus Lindstrém, and Encrinurus punctatus. A form closely
related to Eospirifer radiatus also occurs at this locality. Another
small quarry at Limadau exposes about two feet of very hard compact
limestone containing hardly any fossils.

Rotzikiill is the noted eurypterid locality, made famous through
the labors of Schmidt, Holm, and others, and which has long been the
European Mecca for students of this group of organisms. The eury-
pterid layer is well exposed in a small quarry on the shore to the south
of the village and the base of the quarry is from six to eight feet above
sea-level. At the base are eighteen inches of white fine-grained

>

~ ’ .
a oe ee ee 2

.
\

ee ———— ee

TWENHOFEL: EXPEDITION TO THE BALTIC PROVINCES. 331

dolomite in which the Eurypterus fauna is magnificently preserved.
Eurypterids from here are Bunodes lunula Eichwald, B. rugosus Niesz-
kowski, B. schrencki Nieszkowski, Eurypterus fischeri Eichwald, Pseu-
doniscus aculeatus Nieszkowski, and Pterygotus sp. An Orthoceras,
O. tenuis Wahlenberg, is quite common in the eurypterid layer and
there are also two cephalaspidian fishes, Thyestes verrucosus Eichwald
and Tremataspis schrencki Schmidt. This layer is overlain by about
eighteen inches of mottled yellow and gray, nearly unfossiliferous
limestone from which have been collected a few fish scales, Leper-
ditia angelina Schmidt and Platyschisma helicites.

What is probably the same eurypterid horizon is exposed on the
seashore at Attel, about six or seven miles southwest. The rock
nearest the water-level consists of two feet of gray to white dolomitic
limestone in one to four inch beds. Some of the beds are crowded
with Leperditia angelina while others have many eurypterids. Masses
of Clathrodictyon up to two feet in diameter are associated in the
same beds with the eurypterids and toward the top is a thin layer
which has the appearance of being a conglomerate, but it may be that
the “pebbles” are of organic origin. The upper portion of the beach
is covered with debris, and it is probable that from five to six feet of
strata are concealed. At the top of the beach is a low cliff in which
about four feet of thick-bedded gray limestone are shown. Below the
dolomitic beds at the water’s edge there appear to be beds of coralline
limestone with many stromatoporoids.

Between Rotzikiill and Attel are the exposures of Lello Brook.
These are a little above the eurypterid beds and show about six feet
of thin-bedded limestone, containing many Platyschisma helicites, and
in some layers numerous Leperditia angelina. Many fish remains,
chiefly scales, have also been collected. Species of fish from this

place are: —

Coccopeltus asmussi Pander.
Coelolepis carinatus Pander.

goebeli Pander

laevis Pander.

schmidti Pander.
Coscinodus agassizi Pander.
Ctenognathus murchisoni Pander.
Cyphomalepis egertoni Pander.
Dasylepis keyserlingi Pander.
Dictyolepis bronni Pander.

SOTO SEO be

—_

G02 BULLETIN: MUSEUM OF COMPARATIVE ZOOLOGY.

11. Loptholepis schmidti Pander.
12. Melittomelepis elegans Pander.
13. Otontodus rootsikiillensis Pander.
14. Prionacanthus brandti Pander.
its! dubius Pander.
16. Rytidolepis quenstedti Pander.
17. Stigmolepis oweni Pander.

18. Strosipherus indentatus Pander.
19. laevis Pander.

20. serratus Pander.
21. Trachylepis formosus Pander.

South of Attel are small exposures at Karral and Sarepa, but not
more than four feet of thick-bedded gray limestone are shown at either
place. The horizon is perhaps a little higher than the highest beds of
Attel, but may be the same. At Sarepa Cyphaspis elegantula,
Illaenus sulcatus Lindstrom and other fossils have been collected. At
Hoheneichen, a few miles inland from Sarepa, a few feet of thick-
bedded limestone are exposed in which Labechia conferta Lonsdale
occurs in great abundance, and from which Phlebolepis elegans Pander
and Schidiosteus mustelensis Pander have also been collected.

(b) Kaugatoma zone. This zone is excellently exposed along the
southern coast to the east of Arensberg and on the Sworbe, the south-
western peninsular extension of Oesel. It is more fossiliferous than
the Sagaristi zone.

On the southern shore there are quite a number of important
exposures of which the more extensive are those of Moritz, Kasti,
Nessomay, and Lode, the last being west of Arensberg.

The Moritz locality is on the road to Nessoma and is about four
miles east of Arensberg. No rocks in place are now visible, the quarry
having been abandoned, and it is at present largely overgrown. The
strata appear to have been alternations of thin shales and coarsely
crystalline, black and gray limestones. Corals and crinoid stems ex-
cepted, good fossils do not appear to have been common. Corals
are abundant, chiefly Clathrodictyon and Acervularia luxurians
Eichwald. The crinoid stems belong to Crotallocrinus and the most
common brachiopod is Whitfieldella prunum. Some of the limestones
have the Caudigalli effect on their upper surfaces.

Near Kasti two small quarries were examined, each about a mile
westward from the dwelling of the Kasti estate. One is on the shore
and exposes about three feet of gray crystalline limestone (beds four

;
x
>

ee

:
:

Ww

TWENHOFEL: EXPEDITION TO THE BALTIC PROVINCES. oot

to six inches thick) with thin shale partings. Few fossils appear to
be present. The second quarry is situated on the side of a hill at an
elevation of about fifteen feet above the shore quarry. The rocks
consist of gray crystalline limestones interstratified with gray cal-
careous shales. Only about three feet are shown. A single limestone
bed near the base has the Caudigalli effect on its upper surface, and
this same surface has many beautifully preserved Schuchertella pecten
and Rhipidomella hybrida, of which a great many are single valves
showing excellent interiors. Numerous fragments of Calymene
spectabilis Angelin are also present. The other limestone beds contain
many corals and crinoid stems like those of Moritz.

About a mile farther eastward, at about the same elevation and’
presumably about the same horizon, is a large quarry which has
recently been reopened. About five feet of strata are exposed, con-
sisting of very fossiliferous gray crystalline limestones, and thin cal-
careous gray shales. Some of the shale bands are filled with small
Spirifers and Rhynchonellas, while the limestone beds are thickly
crowded with large Crotallocrinus stems. The old dump heaps afford
splendid collecting. Common fossils of the beds of this quarry are: —

1. Clathrodictyon sp.
2. Favosites gothlandicus.
3. Spongophyllum contortiseptum Dybowski.
4. Hallopora elegantula.
5. Monticulipora fletcheri Milne Edwards and Haime.
6. Ptilodictya lanceolata Hisinger.
7 flexa Hisinger.
8. Chonetes striatellus Dalman.
9. Orthis canaliculata Lindstrém.
10. Homeospira salteri (Davidson).
11. Spirifer elevatus Hisinger.
12. Tentaculites curvatus Boll.
13. inaequalis Eichwald.
14. Calymene spectabilis Angelin.

The Nessoma quarries are just above sea-level, and cover several
acres. The strata consist of thick-bedded black to dark gray mottled
limestone with thin shale partings. Not more than five feet are ex-
posed. Some of the beds of limestone are fully a foot thick, and
afford excellent dimensional stone. The quarries are being actively
developed, and at the time of my visit about a score of men were em-
ployed, the stone being shipped to Riga, Pernau, Hapsal, Arensberg,

334 BULLETIN: MUSEUM OF COMPARATIVE ZOOLOGY.

and perhaps other towns. Fossils are quite common, but generally
not well preserved. Those most common are Acervularia luxurians,
Clathrodictyon, large stems of Crotallocrinus, Whitfieldella prunum,
Ctenognathus murchisont Pander, and Gomphodus sandelensis Pander.
The horizon is practically identical with that of Moritz.

The Lode locality is about two miles west of Arensberg, and the
present exposures are in an old well of which the top is not more than
ten feet above sea-level. About six feet of interstratified gray and
blue limestones and shales are exposed, which appear to belong to
nearly the same horizon as that of Kasti or Moritz. The fossils were
collected on the dumps, those most common being Crotallocrinus
rugosus Miller, Monticulipora fletchert, Rhipidomella hybrida, Homeo-
spira saltert, Spirifer elevatus, Whitfieldella prunum, Calymene specta-
bilis, and Proetus conspersus Angelin.

‘About the most fossiliferous locality of the Kasti zone is that of
Kaugatoma Pank on the Sworbe, where strata are exposed which
excellently typify the lthology of the upper division of the Oesel
formation. The height of the cliff at Kaugatoma Pank hardly ex-
ceeds twelve feet, but the exposures are increased by several feet
which are shown on the beach. At the water’s edge are three to four
feet of thin limestones and calcareous shales, the latter being filled
with extremely fine fossils. The upper portion of the beach is covered
and about four feet of strata are concealed. The lower five feet of the
cliff consist of thin limestones and shales like those below, and contain
essentially the same fossils. At the top of the cliff are six or seven
feet of heavy-bedded limestone which is quite similar to that of Nes-
soma and elsewhere on the southern coast. The species whose names
follow have been collected.

1. Acervularia luxurians.
2. Alveolites repens Fought.
3. Clathrodictyon vesiculosum.
4. Cyathophyllum articulatum Hisinger.
5. Favosites cristatus Blum.
6 forbesi Milne Edwards and Haime.
7 hisingeri Milne Edwards and Haime.
8. Omphyma subturbinata d’Orbigny.
9. Syringopora reticulata.
10. Crotallocrinus rugosus.
11. Fenestella subantiqua d’Orbigny.
12. Monticulipora fletcheri.

TWENHOFEL: EXPEDITION TO THE BALTIC PROVINCES. 335

13. Ptilodictya lanceolata.

14. Camarotoechia nucula (Sowerby).

15. Chonetes striatellus.

16. Leptaena rhomboidalis.

17. Retzia salteri.

18. Rhynchonella diodonta Dalman.

19. Spirifer elevatus.

20. schmidti Lindstrom,

21. Strophomena filosa Sowerby.

22. Spirorbis imbricatus.

23. Autodetes calyptratus (Schrenck).

24. Euomphalus? catenulatus Wahlenberg.
25. cornuarietis Wahlenberg.
26. Holopella obsoleta Sowerby.

27. Pterinea reticulata Hisinger.

28. retroflexa Wahlenberg.

29. Orthoceras bullatum Sowerby.

30. Calymene intermedia Lindstrém.

- 3i. spectabilis.

32. Proetus conspersus.

Farther southwest on the same coast is the Ohhesaare Pank in
which strata of about ten feet in thickness are exposed, consisting of
interstratifications of gray and red sandy limestones and calcareous
shales of various shades of brown, the latter dominating in the middle
of the section. The sandy limestones contain many small remains of
fishes, while the shales carry numerous pelecypods of which Grammysia
cingulata Hisinger is probably the most abundant. The invertebrate
fauna is quite similar to that of Kaugatoma Pank, but the presence
of fish remains and the absence of crinoids are marked features of
difference. These differences are probably to be referred to differ-
ences in the ecologic facies during the times of sedimentation. The
complete fauna from Ohhesaare Pank is as follows: —

1. Favosites cristatus.
forbesi.
Trachypora porosa Dybowski.
Dianulites elegantulus Dybowski. :
Hallopora elegantula.
maculata (Dybowski).
-Ptilodictya lanceolata.
Chonetes striatellus.

eee ee et

BULLETIN: MUSEUM OF COMPARATIVE ZOOLOGY.

Dalmanella elegantula.
Retzia salteri.
Spirifer elevatus.
Cardiola interrupta Broderip.
Modiolopsis complanata Sowerby.
Pterinea reticulata.
retroflexa.
Autodetes calyptratus.
Holopella obsoleta.
Tentaculites annulatus Schlotheim.
inaequalis.
Orthoceras bullatum.
tracheale Sowerby.
Beyrichia wilkensiana Jones.
Leperditia phaseolus Hisinger.
Calymene intermedia.
spectabilis.
Phacopidella downingiae (Murchison).
Gomphodus sandelensis.
Lophosteus superbus Pander.
Monopleurodus ohhesaarensis Pander.
Nostolepis striatus Pander.
Onchus curvatus Pander.
murchisoni Agassiz.
tricarinatus Pander.
Oniscolepis crenulatus Pander.
dentatus Pander.
magnus Pander.
serratus Pander.
Pachylepis costatus Pander.
glaber Pander.
Pterichthys harderi Pander.
Rabdacanthus truncatus Pander.
Tolypelepis undulatus Pander.

It is possible that strata of the Oesel formation occur on the island
of Moon. Schmidt states that material which was brought to him
by a peasant, which was said to have been obtained on Moon from an
excavation for a spring, was of the same lithic aspect as the rock of
the Kaugatoma zone of the Oesel formation, and contained Whit-
fieldella prunum, Spirifer elevatus, and Homeospira salteri. The gen-

oe os

TWENHOFEL: EXPEDITION TO THE BALTIC PROVINCES. 337

eral unreliability of unscientific statements relating to locality, how-
ever, gives to the above data no other value than that it should be
kept in mind in future explorations on Moon.!

TIME EQUIVALENTS OF THE SILURIAN FORMATIONS.

Until the fossils have been carefully studied all correlations must be
merely tentative and based on published lists and field-identifications.
While the conclusions reached may contain something of error, it does
not appear likely that the probability of error is very great.

The Silurian faunas of Esthonia do not appear to have nearly so
much in common with the North American faunas of the same time,
as existed between those of the highest Ordovician beds of the two
countries. The common Russian Silurian corals are largely the same
as those of North America, but that has little meaning as all of them
are long ranging. The gastropods and pelecypods are decidedly dif-
ferent, the Bryozoa have been subjected to little comparative study,
so that reliance must chiefly be placed on the trilobites and brachio-
pods. As with the Ordovician, comparison will be made with the
Anticosti section.

The Addifer formation affords the best points of contact for therein
Atrypa reticularis, Eospirifer radiatus, Pentamerus estonus, and Syringo-
pora bifurcata make their first appearance. In the Anticosti section,
these species or their equivalents, appear for the first time in the very
topmost portion of the Gun River formation and are typical of the
Jupiter River. Using first appearances as a point of departure, leads
to the correlation of the Addifer formation with possibly the upper-
most zone of the Gun River and certainly the lower and middle por-
tions of the Jupiter River, that is, Clinton. One, however, notes the
conspicuous absence of such common Clinton forms as Coelospira
hemispherica, the Stricklandinias, and Bilobites bilobus, as well as
others not so commonly occurring.

The Tamsal formation would then be represented by the middle and
perhaps the lower portions of the Gun River formation; but, corals
excepted, there is little specific faunal evidence supporting the correla-
tion. The general aspect of the fauna, however, is not so old as that
of the Becsie River faunas. A conclusion can be reached if the faunas
of a region intermediate in position be selected for common comparison.

1 Schmidt. Mem. Acad. imp. sci. St. Petersb., 1881, ser. 7, 30, no. 1, p. 49.

338 BULLETIN: MUSEUM OF COMPARATIVE ZOOLOGY.

The most diagnostic fossil in the Tamsal formation is Pentamerus
borealis. In the Kristiana region, this is confined to the upper portion
of Kiaer’s zone 6c and to 7a, that is, to the uppermost zone of the
Norwegian Lower Llandovery and the lowest zone of the Upper
Llandovery, where it is associated in the latter zone with Bilobites
bilobus, and in the former with Plectambonites quinquecostata McCoy
(probably a different species since McCoy’s form came from the
Ordovician), Coelospira hemispherica, Stricklandinia lens Sowerby,
and Camarotoechia decemplicata (Sowerby); all of which in the same
or closely identical species make their first appearance in the upper-
most zone of the Gun River formation, and this fact points to the
probability that this zone, and probably some parts of the preceding
zones, represent the Tamsal formation in the Anticosti section. If
this view be correct, it follows that the equivalent of the Becsie River
formation is not present in the Russian Silurian of the Baltic, and
hence there is a considerable time break between the Borkholm and
Jorden beds. It is worthy of note that the faunal break at this point
is far greater than that between the Lyckholm and Wesenberg.

From one locality a small collection was made which in the field
reminded me of the Cataract fauna, but until this collection has been
studied and compared it is thought best to place no emphasis upon it.
At any rate, it would not depreciate the correlation just made.

The St. Johannis formation is almost certainly the equivalent of
the Wenlock of Scandinavia, England, and west Europe generally,
and the Rochester and Niagara of North America. With the Wen-
lock of Norway, Kiaer’s zones 8a to 8d, the formation has at least
twenty-eight species in common, of which the more diagnostic are
Paleocyclus porpita, Thecia swindernana, Camarotoechia borealis,
Cyrtia exporrecta, Dalmanella elegantula, Dinorthis rustica, Eospirifer
radiatus, Meristina tumida, Rhipidomella hybrida, Rhynchotreta cuneata,
Spirifer crispus, Wilsonia wilsont, Omphalotrochus discors, Calymene
tuberculata, Encrinurus punctatus, and Bumastus barriensis. Corre-
lation of the St. Johannis formation is therefore made with stage 8
(Wenlock) of the Norwegian sections. Since many of the fossils to
which reference has been made in this paragraph also occur in the
lower zones of Kiaer’s stage 9 (Lower Ludlow) it is possible that the
upper portion of the St. Johannis formation may find an equivalence
in the lower portion of stage 9.

With the English Wenlock or Salopian series, the St. Johannis
formation has nearly the same species in common as it has with the
Norwegian section as well as some additional species and the correla-

TWENHOFEL: EXPEDITION TO THE BALTIC PROVINCES. 339

tion is hence made with greater confidence. It is also possible that
the upper beds of the St. Johannis formation have an equivalence
with the Lower Ludlow of England.

As is to be expected, the case is not so sharply clear for the American
sections; but no less than twenty species occur in the American
Niagaran ! or are represented by forms closely related. These are: —

Hindia fibrosa.

Cystiphyllum cylindricum.

Favosites gothlandicus.
hisingeri.

Halysites catenularia.

Paleocyclus porpita.

Atrypa reticularis.

Cyrtia exporrecta.

Dalmanella elegantula.

10. LEospirifer radiatus.

11. Leptaena rhomboidalis.

12. Meristina tumida.

13. Plectambonites transversalis. :

14. Rhipidomella hybrida.

15. Rhynchotreta cuneata.

16. Schuchertella pecten.

17. Spirifer crispus.

18. Wilsonia wilsoni.

19. Calymene tuberculata.

20. Encrinurus punctatus.

oe a ee SE

There is little doubt that the above list will be lengthened when
the St. Johannis fauna has been comparatively studied. Heliolitidae
are present which are quite similar to the American species and such
is the case with other forms of other groups. Since the fauna has by
no means been carefully studied, only a few over fifty species having
been recognized, this makes over thirty per cent of the fauna which
has been recognized in the American Niagaran. It is also possible that
the St. Johannis formation extends a little higher than the highest
beds of the Niagaran.

In the Anticosti section the formation may find representation in
the upper portion of the Jupiter River formation and almost certainly
in the Chicotte, but the Anticosti series does not extend high enough
to completely represent the Russian formation.

1 Niagaran as here used excludes the Clinton.

340 BULLETIN: MUSEUM OF COMPARATIVE ZOOLOGY.

The Kaugatoma formation is of Ludlow age, and has many faunal
elements in common with rocks of that division in both Scandinavia
and Great Britain, but with the Norwegian section the common spe-
cies are not so many as desirable. Species of high diagnostic value
which also occur in the Kristiana region of Norway in strata of Ludlow
age (Kiaer’s stage 9) are Camarotoechia nucula, C. borealis, Chonetes
striatellus, Retzia saltert, Spirifer elevatus, Grammysia cingulata, Megalo-
mus gothlandicus, and Cardiola interrupta (occurs in the strata of the
preceding stage 8).

In the English Ludlow the same (Megalomus gothlandicus excepted)
and other additional species occur, among the latter being Aviculo-
pecten danbyi, Cardiola interrupta, Orthonota cymbaeformis, Orthoceras
bullatum, Crotallocrinus rugosus, and also numerous eurypterids, some
.of which are closely related to those of the Oesel formation. While
the identifications of some of the species named above are perhaps
erroneous, enough are certain to assure synchroneity.

With the North American section the Oesel formation has few spe-
cies in common and these are of almost universal distribution and of
long range. The fauna of the Eurypterus beds, however, affords a
most important point of faunistic contact. Schmidt first identified
the common eurypterid (E. fisheri)’as E. remepes De Kay and later
both he and Holm agreed that it is little more than a geographic
variety of the latter.1 E. remipes is the common form of the Bertie
Waterlime of the Salina group of New York and in addition to these
two forms, there are others which are closely related. If then, the
Sagaristi zone and the Bertie Waterlime be considered synchronous,
it follows that the Kaugatoma zone finds an equivalence in the Coble-
skill, Rondout Waterlime, and Manlius limestones.’

1Schmidt. Bull. Geol. soc. Amer., 1892, 3, p. 59.
2 No correlation has been made with the Gotland section for the reason that the data are not

well enough in hand to warrant the attempt.

' TWENHOFEL: EXPEDITION TO THE BALTIC PROVINCES. 341

Part 3.— AN INTERPRETATION OF THE SILURIAN SECTION or GOTLAND.

INTRODUCTION.

THE writer’s hope of discovering new facts which might illuminate
some of the intricate problems of Gotland’s stratigraphy, and of being
able to place before American students a detailed section of the strata,
is his excuse for delving into this much studied field. The present
discussion, which is intended to be preliminary to a more extensive
treatment of the subject, attempts an interpretation of the lithic and
faunal peculiarities of the stratigraphy. In the first part is given a de-
scription of some of the more important features which bear on the
problem. This is followed by an interpretation of these features.
Lastly, the various conclusions which have been reached are sum-
marized.

In this study of the section, I was greatly assisted by Dr. Henry
Munthe of the Swedish Geological Survey, a courtesy which is deeply
appreciated.

LITHIC AND FAUNAL CHARACTERISTICS OF THE GOTLAND STRATA.

The Silurian rocks of Gotland have long been famous because of
the abundance and excellent preservation of their fossils, and many of
the great leaders of European geology have gone to the island as pil-
grims to a shrine. The sequence of strata and the faunas have fur-
nished the theme of numerous papers, and few of the masters of geology
of northern Europe have refrained from in some way referring to the
geology of this “Cross Roads of the Baltic.” In spite, however, of the
fact that numerous students have studied the various sections, there
still is much that is uncertain and obscure. This arises from the
absence of extensive exposures in the interior of the island, the decided
horizontal variation of the sediments and their enclosed faunas, the
numerous undulations of the strata, and the extensive discontinuities
in the shore sections.

Without entering into a detailed discussion of the different views
which have been held, attention will merely be called to the two
divergent lines about which they may be grouped. Murchison (1847)
considered that the strata dipped from northwest to southeast, that
the beds of North Gotland are the oldest and underlie those of southern
and central Gotland. In this view, he was later stoutly supported

342 BULLETIN: MUSEUM OF COMPARATIVE ZOOLOGY.

by Dr. Friedrich Schmidt. The other view appears to have been
first broached by Hisinger. According to him, the strata of Gotland
are essentially horizontal with only a slight dip to the east, so that
the same beds are found over many parts of the island. This view
received strong support from Helmerson, F. Roemer, Angelin, Bather,
Stolley, Wiman, Dames, and Lindstr6m.

Among the latest published studies of the stratigraphy of Gotland
are those from the pens of Drs. Henry Munthe and Herman Hedstrém,
the former having studied the southern and the latter the northern
half of the island. With some modification, Munthe appears to be in
general agreement with the second view, while Hedstrém appears to
adopt the views of Murchison and Schmidt with, however, considerable
changes.

In general, the strata of Gotland have been said to fall into two
lithic divisions; a lower, consisting of calcareous shades, thin lime-
stones, odlite, and sandstones, and an upper, consisting of coralline
and crinoidal limestones with subordinate shales. This grouping has
been previously noted by Holm, Munthe, and others. It has generally
been considered that the boundary between the two divisions is fairly
sharp, but it is by no means certain that such is the case, since much
which has been considered evidence of sharp division is capable of
a different interpretation.

The detailed work of Munthe over the southern part of Gotland
has given the classification which follows. The divisions are named
from the summit downward.

la Ascoceras limestone.

1b Youngest crystalline limestone.

Ilionia or Spongiostroma limestone.

Upper Sphaerocodium bed and odlite.
Sandstone with clay.

Lower Sphaerocodium bed. |

Dayia flags.

Marl shale with lenses and bands of limestone.

“IO OP W bO

For northern Gotland Hedstrém has offered the following sub-
divisions, named in order from the highest to the lowest.
VII. Nonstratified reef limestones or stratified crinoidal and coralline
limestones.
VI. Leperditia shales.
V. O6litic limestones.

1 Munthe. Guide book 11th. internat. geol. congr., 1910, no. 19, p. 49-50.

TWENHOFEL: EXPEDITION TO THE BALTIC PROVINCES. 343

IVb. Ostracodan limestone, or marl shales and crinoidal limestone.
IVa. Bottom stratum with Stromatoporan limestone and Spongio-
stroma.
III. Upper cliff level of varying composition in different localities.
II. Lower cliff level.

I. Stricklandinia marl.!

Below the lowest visible strata of the northern end of the island are
others which have been revealed by boulders on the shore near Visby.
These consist of red shales containing the coral Arachnophyllum, on
account of which they have been called the Arachnophyllum shales.

Each of the above classifications is subject to the same criticism,
in that divisions have been named after fossils which by no means
always occur, or after some lithic characteristic which is only locally
present.

Attitude of the strata. 'This is the bone of contention over which
opinion has clashed.’ In many places there appears to be a definite
dip southeastward, but in an extremely great number of other places
it can as definitely be determined that a pronounced dip in the oppo-
site direction obtains. Since, however, the lowest visible division, or
zone, is exposed only at the northern end of the western shore, near
Visby, it appears reasonable to assume that there is a general gentle
southeastward dip, from which, however, there are many reversals
of inclination. The island of Storo Karlso excellently illustrates the
directional variations of inclination. On the western side of the
island the Marl shales are beautifully exposed to the south and north
of the point on which the light-house stands and the beds dip into the
sea. On the eastern side there are no exposures of the Marl shales,
but overlying beds are extensively exposed and dip eastward into the
sea, while on the coast of the main island directly opposite (Gotland),
the Marl shales again appear with southeastward inclination (Plate 4).
At the Lau Canal the beds dip northwest at an angle of about fifteen
degrees. These facts bear but one interpretation, 7%. e., very little
reliance can be placed on local inclinations in reaching conclusions
relating to the general attitude, and from inclinations alone one can
reach any conclusion desired, merely by selecting the proper place to
reach the conclusion.

Variation of sediments and faunas. There is nothing more striking
in the Gotland section than the horizontal variation of the sediments
and their contained faunas. These variations are to be seen in many

1 Hedstrém. Guide book 11th. internat. geol. congr., 1910, no. 20, p. 8.

344 BULLETIN: MUSEUM OF COMPARATIVE ZOOLOGY.

localities of which only a few will be considered. One of the best
exposures of the beds immediately above and below the Dayia flags
is that of Lau Canal. This section has been described in detail by
Munthe ! and I merely wish to call attention to a few of the lithic and
faunal variations. At the northwestern end of the canal the lithology
is that of grayish blue laminated shales in which fossils — Atrypa
reticularis excepted — are not common. Proceeding southeastward,
a rising dip leads to a reef-like mass of limestone. This is composed
of tabulate and stromatoporoid corals and is succeeded by jointed
shales filled with an abundance of rhynchonelloids. Another small
coral reef follows and, after that, another mass of shales similar to
those first described and with an equal paucity of fossils. Rhyncho-
nelloids are very uncommon in the first and third shales; but extremely
abundant in the shales between the coral masses, while large specimens
of Atrypa reticularis are relatively common in the first and third
sections of shales, but are hardly present in the second. All of this
variation may be seen in less than fifty feet, and the strata in question
are believed to be of the horizon of the Ilionia bed.

Another excellent place to observe horizontal lithic and faunal
variations is at Hoburgen klint in the higher beds of reef and other
- limestones. In the klint the great unstratified masses of coral rise
many feet, and are bordered on each side by stratified crystalline and
other limestones of a quite different lithic and faunal aspect, the latter
deposits filling up the irregularities in the sides of the former. The
interfingering of coral and sediment and the irregular contacts are
extremely striking and instructive. These latter strata generally dip
away from the reef. Still another excellent place to study the varia-
tion in sediments is Burgen Ridge, on the northern side of which a
mass of hard reef limestone is underlain by shaly reef and crinoidal
limestone, this being limited below by the Dayia flags. On the
southern side, on the contrary, the rock occupying exactly the same
stratigraphic position consists of odlite and limestone conglomerate
(Plate 5, fig. 2) with pebbles up to an inch in diameter, sandy lime-
stone, and calcareous shales and limestones underlain by the lower
Sphaercodium bed and the Dayia flags. On the northern side Sphae-
rocodium has not been noted. All of these variations occur in about
one fourth mile. Another excellent locality is on the island of Storo
Karlso where numerous fine exposures may be seen in the sea cliffs
in which faunal and sedimentary variation are shown. Dozens of

1 Munthe. Sveriges geologiska unders., 1902, ser. C. no. 192, p. 9.

/

TWENHOFEL: EXPEDITION TO THE BALTIC PROVINCES. 345

other exposures present similar features, and the fact of such variation
is altogether too evident to be opposed. It appears very doubtful if
any single division is continuous without much variation over the
entire island. This, and other features to be noted later, have made
the stratigraphic determinations and correlations extremely difficult.
Since changes in sediments are invariably accompanied by faunal
changes the problem is one of great complexity and much difficulty.
Unconformities in the Gotland section. Discordances of strata are
not uncommon in the Gotland section; but where seen by the writer
they have no great significance and bear no other interpretation than
that of contemporaneous erosion and refilling, that is, they were made
by erosion of the sea-bottom during the times of deposition of the
sediments, or they have been produced by the overgrowth of corals on
sediments or the covering of reef growths by sediments. Some of the
conglomerates which have been mentioned by various writers have
been considered as evidence of a transgressing sea. Such conglomer-
ates are present in great number, but they always occur as lenses and
are always local. Many of the rocks which have been called con-
glomerates are certainly not such, since the “pebbles”? have a con-
centric structure and are either of odlitic or organic origin. At any
rate, these conglomerates bear no relation whatsoever to a transgress-
ing sea. Holm! considers the probable existence of an unconformity
at the top of the lower division whose summit is placed above the
odlitic zone of southern Gotland, while Hedstrém mentions a dis-
cordance which is situated at about the same position, that is, between
the Lower and Upper Gotlandian beds? and at about this same level
the Silurian scorpion, Palaephonus nuncius Thorell and Lindstrém
was discovered in association with marine forms. Holm thought that
the presence of this land form bore on the question of discordance and
the probability of a Middle Silurian land interval on Gotland,’ but to
the writer it fails to bear that, as the only, or even the probable inter-
pretation. It might have attained the bottom of a shallow sea in.
many ways. Itis by no means rare today to see land insects and other
land organisms floating in the sea, and A. Agassiz states that “It was
not an uncommon thing to find at a depth of over one thousand
fathoms, ten or fifteen miles from land, masses of leaves, pieces of
bamboo and of sugarcane, dead land shells, and other land débris,
undoubtedly blown out to sea by the prevailing tradewinds. We

1 Guide book 11th. internat. geol. congr., 1910, no. 19, p. 8, quoted by Munthe.
2Hedstrém. Guide book 11th. internat. geol. congr., 1910, no. 20, p. 9.
3 Guide book 11th. internat. geol. congr., 1910, no. 19, p. 8, quoted by Munthe.

346 BULLETIN: MUSEUM OF COMPARATIVE ZOOLOGY.

frequently found floating on the surface masses of vegetation, more or
less water-logged, and ready to sink. The contents of some of our
trawls would certainly have puzzled a paleontologist; between the
deep-water forms of crustacea, annelids, fishes, echinoderms, sponges,
etc., and the mango and orange leaves mingled with the branches of
bamboo, nutmegs, and land shells, both animal and vegetable forms.
being in great profusion, he would have found it difficult to decide
whether he had to deal with a marine or a land fauna.” !

It is not improbable that similar conditions obtained in Silurian
days. Eurypterids occur at about the same level, and their presence
is susceptible of the same explanation, even if they were not them-
selves marine.

In the south of Gotland I do not consider that any evidence of an
unconformity indicating a land interval is present nor was any seen in
the waterfall section at Wisby, but the observations made at that
section were not extensive, nor detailed. Hedstrém described a dis-
conformity between the Lower and Upper Gotlandian in the precipice:
south of Gustafsvik, indicated by a “thin marly and gravelly water-
worn layer, containing, amongst other things, worn Gastropods of the
genera Trochus, Pleurotomaria, Horistomia etc.,” and he has little
doubt of the actual occurrence of erosion, but he does not state whether
he considers this as indicative of a land interval.? I have not seen the
locality and so can not give further data, but the presence of pebbles
and worn shells may be equally well explained by contemporaneous.
erosion in a coral reef channel. It appears that the question of land
in Middle Silurian time, within the present area of Gotland, remains:
to be proven, and the idea is favored that marine conditions prevailed
over the present limits of Gotland throughout the time of deposition.
embraced between the youngest and oldest deposits.

Evidence of shallow water. That the Gotland rocks are the deposits.
of very shallow water is evidenced by the extraordinary development
of coral reefs, the strongly developed edaphic modification shown by
the fossil faunas, and the extensive lateral gradation of sediments.
These show that the sea bottom was sufficiently shallow to closely
respond to land conditions, and that there were numerous local bar-
riers so near the surface as to produce a great variety of differing con-
ditions. Among these local barriers the most important were the
coral reefs which protected colonies within them and hindered migra~
tion over, or around them.

1 Agassiz. Three cruises of the Blake, 1888, 1, p. 291.
2 Hedstr6ém. Guide book 11th. internat. geol. congr., 1910, no. 20, p. 23, fig. 4.

TWENHOFEL: EXPEDITION TO THE BALTIC PROVINCES. 347

The sandstone layers in the south of Gotland are extremely sug-
gestive and interesting. They appear to grade northward into shales,
suggesting that the source of the materials was southward. The
sands are the deposits of quite shallow water, and are probably not
wind and wave deposits of a shore, the absence of decided crossbedding
rather strongly arguing against the latter view. They contain fossils
only near the base and near the top and are overlain by odlites. The
cleanness of the sands and the. absence of crossbedding and fossils
suggests the drift sand zone of Godwin-Austin, a deposit of fairly
quiet water; while the overlying odlites with fossils could well have
accumulated in the shallow water after either some extra-Gotland
territorial physical changes or changes in the development of the
coral reefs, modifying the direction of the currents, had caused the
sand to cease to drift. A most important fact connected with the
sandstone beds is that in the sea over southern Gotland they elimi-
nated the corals during the duration of the deposition of the sands,
and hence no coral reefs appear to pass through them:

The coral reefs. Many students have commented on the extensive
development of the reefs of Gotland and they are well worthy of com-
ment. They are present in practically every parish of the island,
and the surface and shore topography of Gotland have been largely
controlled by the fact of their occurrence, and these in turn to a large
extent have influenced the past and the present activities of the
Gotlanders. The organisms which played the greatest part in the
development of the reefs belong to the Stromatoporoidea, and there
are probably a half score of these for every colony of all the others
put together. Crinoids also made great contributions to the material
of the reefs. Some of the reefs are of large size. At Hoburgen klint
one of them is fully one hundred and fifty yards wide and reaches from
the odlite to the top of the cliff, a thickness of about seventy-five feet.
The corals of the reefs are frequently little changed from their original
condition, and in pockets between the coral growths are small masses
of clay and sand which are remarkable for the excellent fossils which
they contain. In general, the coral masses offer greater resistance
to erosion than do the surrounding or underlying rocks, so that in the
cliffs they project as salients or overhanging masses and in the fields
as knolls, on the northward or stoss side of which, the side from which
the ice came, the surrounding rock has all, or nearly all, been eroded
away, while on the side opposite the direction of ice movement the
protection of the coral masses has preserved to some extent the sur-
rounding rock, which is generally a crystalline limestone. These

348 BULLETIN: MUSEUM OF COMPARATIVE ZOOLOGY.

knolls are called by the people “Klintar,” and they constitute the
hearts of the elevations of the island. Probably the original summit
of every one of these reef masses has been eroded away, as well as later
rock which was deposited about their margins. In the cliffs the pecul-
iar forms developed in the coral masses have led to various legends, as
the “Old Man of Hoburgen,” ete. Some of the coral colonies are
quite large. At Hoburgen a single stromatoporoid colony which shows
continuous growth, but has two included marl pockets, is ten and a
half feet wide and eight feet high.

At Staffs klint, the reefs extend to and into the Marl shales, tend-
ing to show that at that place growth apparently was continuous from
the times of Marl shale deposition up to, and probably beyond, the
time of the deposition of the Spongiostroma bed, thus bridging the
gap between the upper and lower divisions.

INTERPRETATION OF THE GOTLAND SECTION.

Any interpretation of this famous section must recognize the
importance of the coral reefs. They control the present shore and
surface topography of the island and they must have exerted a large
influence upon, even if they did not control, sedimentation, during
the times of their development. What were the conditions under
which they were built? It has been shown that at Hoburgen and
Staffs klint, and could be shown in many other places, that they
extend through many beds, and it is assumed that they grew more or
less continuously while the sediments were depositing around them.
A question of great importance in this connection is whether the rock
of the coral reefs is of the same age as that of the other beds which
at any particular level lie on their flanks. The conditions around
modern coral reefs help to solve this problem. These, at least in
many, and probably most instances, rise with steep slopes from
deeper waters “and if elevated above the sea, they would stand as
broad ramparts separated by passages mostly 20 to 200 feet deep.” *
Vaughan describes a like bottom topography about the Florida reefs,
but the passages separating the reefs are “usually 9 to 12 feet”’ and
“always less than 10 fathoms.’”’? The surface of a modern coral reef
is very irregular and “there are deep cavities among the congregated
corals, in which a lead will sometimes sink to a depth of many feet

1 Dana. Coral and coral islands, 1890, p. 389.
2 Vaughan. Papers Tortugas laboratory Carnegie inst. Washington, 1910, 4, p. 109.

a,

—_ eee

TWENHOFEL: EXPEDITION TO THE BALTIC PROVINCES. 349

and even fathoms,” ! and these depressions of the surface afford
lodging places to the multitude and variety of organisms, which live
in, on, and about a growing coral reef, a variety of life to which atten-
tion has been called by nearly every student of coral reefs.

Many feet, in the vertical sense, may and usually do separate the
upper surface or summit of a coral reef from the bottom of the sea
about the margin, while the horizontal distance between the summit
of the reef and the bottom is usually not great. Conditions of this
nature make it possible for individuals of the same species to live on
top of the reef among the coral colonies and also twenty-five to a
hundred feet lower down only a few feet distant horizontally, and it is
certain that some of the shells of animals living on the top will be
washed into lower waters by the waves which dash over the reef during
high tide or during storms. Since variation of depth up to one
hundred feet is not sufficient to control the distribution of a great
number of species of marine organisms, it follows that many species
will thrive in multitudes on the top of the reef and over the adjacent
bottom.

Ultimately the reef may and probably will become surrounded with
sediments of an age somewhat younger than that of the reef on the
same level with itself. Stating the matter differently, the rocks com-
posing the reefs would have their time equivalents in strata holding a
lower vertical position in the section.

Exactly similar conditions must have obtained among and about
many, if not all, of the ancient coral reefs and the writer is quite
positive that they existed in connection with the reefs of Gotland,
which stood above the bottom as is proven by the fact that cases are
visible where coral colonies have fallen from the sides of the reefs to a
lower level. That the tops of these elevations were irregular and filled
with deep and shallow holes is shown by the included masses of clay,
many of which are filled with excellently preserved fossils.

If the organisms of this sea were distributed over this sea bottom
both upon the tops of the reefs and the lower levels about their margins
there should be some evidence of this distribution in the existing fossils.
In this consideration the distribution of the corals themselves has no
validity, because in the cases of most of them they range more or less
throughout the entire section. The chief reliance has been placed on
the brachiopods. The natural stratigraphic position of Bvlobites
bilobus (Linné) is in the lower half of the Gotland section, but near

1Dana. Loc. cit., p. 145.

300 BULLETIN: MUSEUM OF COMPARATIVE ZOOLOGY.

Etelhem Dr. Henry Munthe collected this species ! on the summit of
one of the knolls of the Ascoceras limestone which are so abundant in
that region, while on the summit of another knoll of the same region
I collected Conchidium conchidium (Dalman), C. biloculare (Linné)
in the midst of a coralline limestone of the Ascoceras type, and on
Storo Karlso Pentamerus estonus Lindstrém and Conchidiwm conchi-
dium occur in association in the midst of coralline limestones belonging
to the rocks of the highest part of the island. Orthis rustica Sowerby
is another species which can be found far above its natural strati-
graphic position in the midst of the highest rocks of the island, the
writer having collected it in clay pockets in the coral reef at Hoburgen.

It has been quite generally considered that the highest rocks of
Gotland are the youngest, yet we find therein species whose natural
stratigraphic positions are lower in the section. | I believe that the
high coralline rocks of Gotland are not the youngest of that island,
that they are older than any rock (there are a few exceptions) which
lies at the same level on their flanks, and that the coralline rocks
should be correlated with strata holding a much lower level. A recent
letter from Dr. Munthe shows that he is coming to the same conclusion
for he says, writing under date of October 2, 1915, “Last summer I
have ascertained that the ‘Marl Shales’ (Lindstrém’s bed e) are
synchronous with a part of the reef-limestone (“Ascoceras limestone
in part’) in Garde etc.” I believe the conditions are something like
those represented in Figure 1.

Conditions like these make local correlation extremely difficult,
and when one considers that collections have been made without a
recognition of the possibility of their occurrence, it is readily seen to
what extent extra-Gotland correlation would be complicated. It is
believed that in this way are explained a great many of the various
difficulties which have arisen in connection with the Gotland section.

Assuming, then, that the corals stood above the bottom and rose
to or nearly to the surface of the water, it follows as a consequence that
during the period of the development of the reef that it grew upward
much more rapidly than the sediments accumulating around it.
Later, however, when the reef reached the surface of the water its
rapid growth ceased,? and sediments accumulated around it far more
rapidly than it grew upward, the coral reef providing through its

1 Personal communication.

*This statement assumes that there was no sinking of the sea-bottom. If the sea-bottom
were sinking, growth upward would continue until the downward movement ceased, after
which the conditions described would obtain.

TWENHOFEL: EXPEDITION TO THE BALTIC PROVINCES. 551

attempted upward growth a part of the materials composing the sedi-
ments. It became possible, then, for very small thicknesses or even
small patches of material in the upper portion of a reef to represent
considerable thicknesses of the sediments which were accumulating on
the flanks of the reef. Stating the matter differently, in the early
stages of reef development, great thicknesses of unstratified reef
limestone should be represented by much thinner masses of stratified

A’

Figure 1.— Diagram showing possible conditions in and about a coral reef. The
line A—A’ represents sea-level during the development of the reef and the deposits
about it. The line B—B’ represents the present surface. Sediments other than
those of the reef are stratified. ‘The unmarked portions of the reef above the
present surface are assumed to have been eroded away after uplift had taken place.
Unmarked portions between the present surface and the former sea-bottom, M and
M”, are assumed to be filled up with sediments which were developed after the
reef was formed and hence are younger than the reef. ’

M-M” are deposits of the same age. On the margins of the reefs the stratifica-
tion is inclined.

C-C” are colonies of the same species of organism on the reef and adjacent bot-
tom. Sediments containing the shells of these species will be of the same age,
but the reef rock on each side and above the colonies of the reef is older than the
sediments which fill the cavities.

b” represent the base or beginning of a coral colony.

a-a’ are stratified sediments within the reef. The time equivalents of these
sediments are below M and M”’.

P-P are outcrops of inclined stratified rock with everything else, other than
some exposures of the reef rock, hidden. The exposures might readily lead to the
conclusion that the reef on the left is older than that on the right, and that the strata
are progressively younger toward the right or, if the fossils indicated otherwise,
that a fault lay between the two reefs.

limestone lying at a lower level, and, after the reef reaches the water-
level, almost negligible thicknesses of reef limestone should find their
equivalence in much thicker beds of stratified limestone, also at a
lower level. Further, in all the stages of reef development the later
animals would live above the shells of the earlier animals, and also in
the hollows of the reef below them, and when the whole became turned
into stone the shells of different times would become almost hope-
lessly mixed.

352 BULLETIN: MUSEUM OF COMPARATIVE ZOOLOGY.

If the above arguments have validity, and I am firmly convinced
that they do, it follows that little weight can be attached to coral
reef faunas in attempting correlation.

In the channels between modern reefs varied conditions obtain,
and Dana describes the deposits as variable to a high degree, at one
place coral sand, at another coral mud, at others clay mud with little
coralline material and he states that “The facts show that the rocks ~
formed in such channels may be of all the kinds that occur in reef
regions — coral and shell conglomerate, compact impalpable limestone,
limestone full of Orbitolites, or containing, as well, remains of other
species of the seas, and also rocks made of clay, mud, sand or pebbles
of: the mountains or high lands adjoining.” !

As illustrative of the data given in the above quotation Dana might
have cited the reefs of Gotland, as every word is strictly applicable.
There, on the same level, are fine-grained limestones made from lime
muds, limestone conglomerates, lime sandstones, clays, shell breccias,
etc. In the passages between modern reefs the “tidal currents often
have great strength, and are much modified and increased in certain
places, or diminished in others, by the position of the reef with refer-
ence to the lands” ? and there is little doubt that local erosion replaces
deposition to a considerable extent as the growth and erosion of the
reefs change the direction of the currents, thus developing local uncon-
formities. In this way the many discordances of the Gotland section
are readily explained.

These varied conditions upon, within, and about a coral reef will
be and are reflected in the faunas of the bottom, each species selecting
that bottom and those conditions on and under which it best thrives,
and animals of the same species might be found on the top of the reef
and at its foot, as, for example, might have occurred (and where it is
believed it did occur) at Hoburgen where about seventy-five feet sepa-
rate the existing summit from the base.

The sediments which are deposited on the flanks of a coral reef,
or any similarly elevated mass, are inclined away therefrom. The
compacting of the strata would increase the inclination while local
slumping would intensify the undulatory structure. As a consequence
anticlinal and synclinal structures would be developed. Some of the
undulations of the Gotland section may, and appear to have been
developed by subsequent movement, but it is certain that many are
contemporaneous and are directly referable to the influence of the

1Dana. Loc. cit., p. 152.
2Dana. Loc. cit., p. 151.

TWENHOFEL: EXPEDITION TO THE BALTIC PROVINCES. 390

coral masses. Some of them may also have been subsequently formed
by the settling of the sediment over the coral ridges, which would
remain rigid while that filling the basins on each side would settle.
At any rate, few or none of these inclinations and undulations have
any significance in connection with the stratigraphic relations.

Among coral reefs the strata in the different channels are more or
less isolated, and they would dip away from the reefs in a way that
would appear to carry them far below other strata in the direction of
inclination, and if this basin or synclinal-like structure could not be
discerned the conditions would lead to divergent interpretations.
I do not wish to argue that younger strata are not present in the south-
eastern part of the island, nor that strata of the same age occur over
all parts of the island. I believe the truth lies between the two
extremes, but wish to emphasize the fact that in each of the channels
between the reefs there will be apt to be a different lithology, a differ-
ent structure, and a different fauna from that in any other channel.
Since the rate of deposition in each channel will be different from that
in another, it follows that strata of a certain time interval will lie in
one channel at a higher or lower level than that of another channel,
and each will be below the rock of the reef which represents the same
time interval.

That to a considerable degree, the conditions described represent
those obtaining in the Gotland section I feel assured. These views
open a most interesting field for investigation, which should be begun
by a careful mapping of every coral reef which can be discovered on
Gotland, and this should be followed by a careful study of the inclina-
tions of the strata, and these should also be mapped in order to see
the relation of these inclinations to the coral reefs. Then the deposits
and fauna of each basin between the reefs should be studied in an
endeavor to correlate the strata of the different basins. When this is
done the controversy as to whether certain strata on one side of the
island lie underneath or above certain strata on the other side will have
disappeared, and data will have been collected upon which something
besides hypothesis can be erected. The detailed work that Drs.
Munthe and Hedstrém are doing is laying the foundation to the ulti-
mate untieing of this Gordian knot of Baltic stratigraphy.

354 BULLETIN: MUSEUM OF COMPARATIVE ZOOLOGY.

CONCLUSIONS.

1. The coral reefs exercised a controlling influence in the develop-
ment of Gotland lithology and stratigraphy.

2. Little reHance can be placed on coral reef faunas in correlation,
since species of different horizons are apt to occur in association.

3. The fauna of any basin about a coral reef is apt to exhibit many
differences when compared with the fauna of another basin.

4. Discordances of strata are readily developed in sedimentation
about coral reefs and have little or no significance.

5. Conglomerates and worn shells should be present in deposits

about coral reefs, and neither indicates a transgressing sea nor a land

interval.

6. Directional and quantitative variations of inclination of strata
have wide limits about coral reefs, and should always be considered
in coral reef stratigraphy.

7. In general, the rock of any reef is older than that lying at the
same level on its flanks.

8. Strata of the same time interval will lie at different elevations in
the different channels between the reefs.

9. The ultimate solution of the problems of Gotland stratigraphy
will probably be reached by a study of the relations of the sediments
and the faunas to the coral reefs.

1" , ra
{> a § a f iw 4 adel Po ha j s
aia eager 4 h poten «tt
i} ry vis z ¥ ;
7 rye

?

j ar f
ms / vf "% +) ,
an w* 4

cs

“a
4

Ar eer es!

n

, : it

Be edt Yo isan RAP ols

erie % 2
| # ;

Sah Ae Oh ERO RES EL Ct Ne ee a

: ' Fe, ; ie % 1a F 33 ,
> ‘ t di 5 { #
. ; 5 4 “". M ma : “,
‘ Ma Bake SBA
? pir way a }
ae ¥ y is
i J ne ‘ Aaa)
’ '8 4 7h
« J r,t ve 7
i ome
ye el Ys :
; Ce A
, say Hf, Ay r
e ul >)
er ann) oS “ink

Twennore..— Expedition to the Baltic Provinces. this 43 ‘s a ko IS eae
er | ae ee ucts. ae

F ~ ‘
"Y Pit © Sea,
.
\ \
-
f Cel aNd Bi
} ee
G {
? £ iz ed. fd wor > : i
Ce
i .) ,: Lied Whey \ id “
‘ to3 TES
. i : . an [
° /} ‘ rat,
J |
1 7 ‘
| 7 ’
{ ihe of bead i
7 , y
\
; t
/ ° ,
3 ad ae Cs
: STOR ES ie i Cy Oe
. . bed 7) . -
ae j LETRA en
r 5 . . t 5 er:
i r Palaeozoic in the Baltic Provinces.
| M ; ‘ we’) OP ea
. if i 5“ 4
+ ‘ ta
a { oe ; My
« ; * fi
f ‘Fa
Pl Sg “toad
’ 4 : 4 i
. " rs A r Ae ‘ 4 7 #7 ‘yi
Pie
‘ ‘eet Toles A Se ‘ ’
S
/ P ; ‘
3
a
= ~
‘
} \
‘
.
’ “ ‘s ’
K - -
‘ -
j
;
.

“*~
' ‘
5 »
' g A . 7 J
~* es | - iy rv
y ‘ ¥ ry; ‘OS ae
A iS 2 : ?
. 7 Ls |
- «| be ; 2 *
( i. uit r Ppa! Ad *
| ve | ' «
. j ~ p é
'
a f + "

EXPED. TO BALTIC PROVINCES. PLATE 1

BULL. MUS. COMP. ZOOL.

LAKE LADOGA

GULF OF FINLAND

ae,
ES
1k ©
jee
} Ay
St
am
ae
=3
o)
|

IN THE ——
BALTIC PROVINCES OF RUSSIA

—_—_——

iu!
HP
tte
teeny tite

MODIFIED FROM BASSLER (AFTER SCHM/0T)

GULF OF R/GA

S/LURIAN

ORDOVICIAN

CAMBRIAN

: OX ties ve
TwENHOFEL.— Expedition to the Baltic Provinces.
ag ele oo eo
. Fig. Ah ey ioe. in the Lewes De ho }
Tig, oe ig of the Upper Lyckholm at Pirk, Bsthonia. oy
| 7 z ;

< iV

ee P

. ik x vs

2 ? ; :
¥ ’ + ~
o. ¥ 7 ts id
Fi "y ‘i ™
; \ rs
wt ‘
ie ‘ iat
”
4 + 2 -
s
=
;
!
’
~ é
.
‘

BULL. MUS. COMP. ZOOL.

TO

EXPED.

HELIOTYPE CO., BOSTON

: if f ¥ si
< . } ™~ 4 rl
_ | a a .
yi fy :
’ fr
»> o*.
ih ‘, ’
d H 4
; =
= - 2
3 “A a
“ . ; 7
fe iN ’ e
ae Yy
is é 4

’ .
Whee |. 4

ra i

TweENnHOFEL.— Expedition to the Baltic Provinces.

PLATE 3.

Fig. 1— Contact between the Jérden Schicht and the Pentamerus borealis
banks at Jorden, Esthonia. The type-section. | .

The hammer is on the line of division. The spots in the upper

strata are the shells and molds of Pentamerus borealis.

Fig. 2.— Quarry face in the Raikiill formation at the type-locality in Esthonia.

ein.

BULL. MUS. COMP. ZOOL

HMELIOTYPE CO...

EXPED. TO BALTIC PROVINCES.

BOSTON

PLATE 3

TweENHOFEL.— Expedition to the Baltic Provinces.

PLATE 4.

Fig 1.— East side of the Island of Storo Karlso off Gotland, Sweden, showing
the strata dipping into the sea.
Fig. 2.— West side of Storo Karlso showing the strata dipping west.

BULL. MUS. COMP. ZOOL.

EXPED. TO BALTIC PROVINCES.

PLATE 4

ule

7

Te eae , at
a

x se Ais = ae alte Sieh dh wie y
Mente 5 ae vA ape et A dar)

“O41 1) seg TES ia a
F I ha 4 PF bi if seh ty Gu!
% » Nahe jelg we

ke 4. ai

‘ ary é
( a 4 * Te a we
oye pt
TWwENHOFEL.— Expedition to the Baltic Provinces. a
7
‘

_ Fig. 1— South side of Storo Karlso showing great variation of i

“i in the vicinity of small reef-like masses of coral. ane
Fig. 2.— Limestone conglomerate on the south side of Burgen Ridge,

Nearly all the pebbles are rounded and show great wear. — . few

years appear to suffice for the rock to disintegrate. The coral |

reef is in the heart of the ridge. The camera was held about

three feet above the ground.

’
’
. ae
» v,
gis
i a ey
A Tiel ten
payed
. /
=
ae 445
eo often oy
‘ a af pee '
ee a Ped
it nde ie ‘ ¥
7 ’ ‘ae ae) é
. ape «peel aa J % Bere PP se,

anmee tawsunae Ae

EXPED.

ener rom

TO BALTIC PROVINCES.

PLATE 5

i hie

fd
j
>
.

—_

ae at

ct Sal

‘

it

ny

.¢
ver heat t
ve
Pret 7
ima \
rey , re |
rr fer .
‘
i.
wet Sly
f
a
~
~~
ms
¥
*
om
ase
ae
ee
“ ? i
2
n
aris)
vores
ue
y
sx >
a 7 J
% sl e
eet ae
va
fn
i 6
Ris
Wx
=~
+
f ‘
'
/ =
.
]
t
‘
-
-
4
‘
® ¥

%
‘
”
f

“Se
ae

Pay

—

Pa he el So
Se $

QL Harvard University. Museum
ui of Comparative Zoology

H3 Bulletin

v.56.

Biological

& Medica!

Serials

PLEASE DO NOT REMOVE
CARDS OR SLIPS FROM THIS POCKET

UNIVERSITY OF TORONTO LIBRARY

‘
voy
‘ :
Py - #
'
i “¢

we PPS

rw

«eel