GEOLOGICAL SURVEY OF CANADA,

ALFRED R. C. SELWYN, F.G.S., Dtrector.

PvEPO

).N" THK i,

FOSSIL PLANTS

LOWER CARBONIFEROUS AND MILLSTONE GRIT

FORMATIONS OF CANADA-

BV

J. W. DAWSON, LL.D., F.R.S., F.G.S.

MONTREAL :

PRINTED BY JOHN LOVELL, ST. NICHOLAS STREET.

1873.

THE LIBRARY

BRIGHAM YOUNG UNIVERSITY

PROVO. UTAH

Digitized by the Internet Archive
in 2010 with funding from
Brigham Young University

http://www.archive.org/details/reportonfossilplOOdaws

GEOLOGICAL SURVEY OF CANADA,

ALFRED R. C. SELWYN, F.G.S., Director.

REPORT

OS THE

FOSSIL PLANTS

OF THE

LOWEE CAEBONIFEEOUS AND MILLSTONE GEIT
rOEMATIONS OF CANADA-

BY

J. W. DAWSON, LL.D., F.R.S., F.G.S.

MONTREAL :

PRINTED BY JOHN LOVELL, ST. NICHOLAS STREET.

1873.

\

THE LIBRARY

BRIGHAM YOUNG UNIVERSITY

PROVO. UTAH

REPORT

ox THE

FOSSIL LAND PLANTS

OF THE

LOWER CARBONIFEROUS AND MILLSTONE GRIT
FORMATIONS OF CANADA.

BY J. W. DAVv^SON, LL.D., F.R.S., F.G.S.

To Alfred R. C. Selwyn, Esq., F.G.S.,

Director of the Geological Survey of Canada.'

Sir,

The present Report may be regarded as a sequel to that of 1871, on
the Flora of the Devonian and Upper Silurian.

It mil serve to characterise by vegetable fossils the beds which form
the lower members of the Carboniferous System, and to distinguish these
lower or " False" Coal-measures from the Devonian on the one hand and
the Middle or Productive Coal-measures on the other.

It will also serve to illustrate the Flora of a somewhat obscure and
neglected part of Palaeozoic time, that which constitutes the dawn of the
Carboniferous Period.

To facilitate comparison, until a detailed Report on the more extensive
Flora of the Middle and Upper Coal-formations can be issued, I have
added a catalogue of the plants of those formations.

I have the honour to be.

Your obedient servant,

J.^W. DAWSON.
McGiLL College,
Montreal, June 10, 1873.

CONTENTS.

I. Introductory sketch of the Geology of the Lower Carboniferous Coal Measures and
Millstone Grit, with the equivalent formations abroad.

II. Descriptions of the Plants of the Lower Carboniferous Coal Measures.

III. Descriptions of the Plants of the Millstone Grit.

IV. Catalogue of Plants of the Middb and Upper Coal Formations.

V. Note on the external characters of Lepidodendroid and Sigillaroid trees.
YI. Appendix — New Sigillariae and Lepidodendron.

I. INTRODUCTION.

The formations to which this Report relates extend from the Lowest
Carboniferous Beds upward to the equivalent of the Millstone Grit inclu-
sive, and underlie the Middle or Productive Coal-formation.

In the region under consideration they overlie unconformably the
Devonian beds of which the Flora was described in my Report of 1871,
as well as the Upper Silurian and other older Formations.

The term Lower Carboniferous as used in this Report is, therefore,
equivalent to Sub-Carboniferous of some American geologists, and desig-
nates a group of beds distinguished both stratigraphically and by fossils
from the Devonian or Erian below, passing into the Coal-formation above,
and properly included within the Carboniferous system, of which its basal
portion forms the lowest member, while its upper portion immediately
underlies the Millstone Grit, which may be regarded as forming the transi-
tion from the Lower Carboniferous proper to the Middle Coal-formation.

Where most fully developed in Nova Scotia and New Brunswick, these
formations may be thus sub-divided in ascending order : —

1st. The Horton Bluff Series or Lower Carboniferous Coal-measures,
consisting of hard sandstones and shales often calcareous,
associated with conglomerate and grit, and in some places with
highly bituminous shales. They contain underclays and thin
coaly seams, remains of Plants, Fishes and Entomostracans,
and foot-prints of Batrachians, but no strictly marine remains.
This group was first estabhshed as a distinct sub-division of
the Carboniferous in Nova Scotia, by Sir C. Lyell and the
writer in 1843 and 1817.
2nd. The Windsor Series, ov Lower Carboniferous Limestone and
Gypsiferous Beds. This is a marine formation holding char-
acteristic shells and corals of the Lower Carboniferous period,
and containing in addition to the limestones thick beds of
sandstone, Karl and clay, usually red, and of gypsum. First
defined by Sir C. Lyell in 1843.

6 CANADIAN FOSSILS.

3rd. The 31illstone Crrit Series, consisting of sandstones and shales,
often red, and conglomerate, associated -lyitli dark coloured
beds holding fossil plants and Naiadites, and -with a few
underclays and thin seams of coal. The name Millstone Grit
■was first applied to these as a distinct group by Mr. R.
Brown in 1844. The group was distinctly indicated in Sir
W. E. Logan's section of the South Joggins in 1843, and in my
paper of the same year on the Lower Carboniferous Rocks of
Eastern Nova Scotia.
4th. Above these are the Middle Coal-fonnation and the Ui^per
or Newer Coal-formation.
The three lower members are thus described in "Acadian Geology,'
where will also be found an account of the fossils discovered in them up to
1868, and many local details : —

" The Loiver Carloniferous Coal measures, or Lower Coal measures. —
In some localities these resemble in mineral character the true coal mea-
sures. In others they present a great thickness of peculiar bituminous and
calcareous shales. They usually contain in their lower part thick beds of
conglomerate and coarse sandstone, which in some places prevail to the
exclusion of the finer beds. The characteristic plants of these beds are
Lepidodendron corrugatum and Cyclopteris Acadica, with Dadoxylon antiquius.
They also contain locally great quantities of remains of fishes, and many
Entomostracans, among which are Leaia Leidyi and an JEstheria, also
Leper ditia suhrecta, Portlock, Beyrichia colUmdus, Eichw., and a Cythere.

This formation is not everywhere distinguishable at the base of the Car-
boniferous, and is variable in its characters. It is seen in southern Cape
Breton, in the county of Sydney, and in Hants ; but its most remarkable
and interesting exposures are at Horton Bluff, and at Hillsborough and
other places in southern New Brunswick. In the last mentioned locality,
it affords the remarkable bituminous mineral known as Albertite."

" The Loiver Carboniferous Marine Formation. — The essential features
of this formation are thick beds of marine limestone, characterized princi-
pally by numerous brachiopods, especially Productus Cora, P. semireti-
culatus, Athyris subtilita, and Terebratula sufflata, with other marine
invertebrates. Associated with these limestones are beds of gypsum, and
they are enclosed in thick deposits of sandstone, clay and marl, of pre-
vailing red colours.

The thickness of this formation seems to be very variable, and in some
districts it is represented almost entirely by conglomerates, while in others
it abounds in limestone and gypsum. It is very largely developed in
Hants and Colchester counties, and rises from beneath the Millstone Grit

LOWER CARBONIFEROUS PLANTS. 7

in Cumberland, Pictou, and Cape Breton. Smaller areas occur in several
other-parts of the province of Nova Scotia, and it is extensively developed
in New Brunswick. It affords all the gypsum exported from Nova Scotia
and New Brunswick."

" The '^Millstone Grit" Formation, — This name, though not in all cases
lithologically appropriate, has been borrowed from English geology to
designate the group of sandstones, shales, and conglomerates, destitute of
coal, or nearly so, and with few fossil plants, which underlies the coal
measures. In its upper and middle part it includes thick beds of coarse
gray sandstone holding prostrate trunks of coniferous trees (Badoxylon
Acadianuni). In its lower part, red and comparatively soft beds pre-
vail. This formation is exposed in the same localities with the Middle
Coal formation, and especially in the south Joggins section, where it
attains to the enormous thickness of between 5000 and 6000 feet."

In some localities the lower member is absent, the marine limestones
resting on the older rocks. In other locahties the marine member is
absent or very slenderly developed and the Lower Carboniferous
Coal-measures and Millstone Grit are united together. In this case, how-
ever, the lower series is usually represented by coarse conglomerates with
few fossils.

Equivalent Formations Abroad.

This is a subject of some importance, more especially with respect to the
Horton series or Lower Carboniferous Coal-measures, as errors have been
committed both in the way of confounding these with the Coal-measures
above and with the Devonian below ; and in works of general geology very
little attention is usually given to them as a distinct group. With regard
to the Marine Limestones, their equivalency to the Lower Carboniferous
Limestones of other countries is undoubted. The Millstone Grit also admits
of very little difference of opinion as to its equivalents. In the following
lists I have given the equivalents of the Horton series and Millstone Grit
series as they appear to me to be settled by stratigraphy and fossils.

1. Equivalents of the Lower Carboniferous Coal Measures or Horton Series.

(1). The Vespertine Group of Rogers in Pennsylvania.

(2). The Kinderhook Group of Worthen in Illinois.

(3). The Marshall Group of Winchell in Michigan.

(4). The Waverley Sandstone (in part) of Ohio.

(5). The Lower or False Coal Measures of Virginia.

(6). The Calciferous Sandstones of McLaren, or Tweedian Group of Tate in Scotland.

(7). The Carboniferous Slate and Coombala Grits of Jukes in Ireland.

(8). The Culm and Culm Graywacke of Germany.

(9). The Graywacke or Lower Coal-measures of the V^osges, as described by Schimper.

8 CANADIAN FOSSILS.

(10). The Older Coal-formation of the Ural, as described by Eichwald.

(11). The so-called "Ursa Stage" of Heer includes this, but he has united it with Dero-

nian beds, so that the name cannot be used except for the local derelopment of

these beds at Bear Island, Spitzbergen.

All of the above groups of rocks are characterised by the prevalence of
Zepidodendra of the type of Zi. corrugatum, L. Veltheimianuni and L. Glin-
canum ; Pines of the sub-genus Pitus of Witham , Fdlaeoxylon of Brongniart ;
and peculiar Ferns of the genera Cyclopteris, Cardiopteris and Sphen-
opteris. In all the regions above referred to they form the natural base of
the great Carboniferous System.

2. Equivalents of the Millstone Grit Series.

1. The Serai Conglomerate of Rogers in Pennsylvania, &c.

2. The Lower Coal formation Conglomerate and Chester Groups of Illinois (Worthen.)

3. The Lower Carboniferous Sandstone of Kentucky, Alabama and Virginia.

4. The Millstone Grit and Toredale Rocks of Northern England, and the Culmiferous of

Devonshire.

5. The Moor rock and Lower Coal Measures of Scotland.

6. Flagstones and Lower Shales of the South of Ireland and Millstone Grit of the North

of Ireland.

7. The Jungste Graywacke of the Hartz, Saxony and Silesia.

The vegetable fossils of this Group differ from those of the beds below
the Marine Limestones, and contain forms resembling or identical with
those of the Middle Coal-formation, into which indeed both lithologically
and as to fossils the Millstone Grit passes by imperceptible gradation.

Distribution in New Brunswick, Nova Scotia and Lower Canada.

Under this head I shall very shortly sketch the geographical arrange-
ment of the formations in question, referring to previously published
memoirs and reports for details.

In Gaspe and the Bay de Chaleur and along the northern margin of the
New Brunswick Carboniferous District, the Lower Carboniferous Formation
presents the characters of the Bonaventure Formation of Sir William
Logan : the marine limestones being absent or little developed, and the
prevailing rocks being conglomerates and sandstones with few fossils.
(Logan, Report of 1863. Robb, Report of 1869. Acadian Geology, P. 227.)

In Southern New Brunswick the Lower Carboniferous Coal-Measures
are remarkable for the great thickness of bituminous and bitumino-cal-
careous shales which they contain. These rocks hold the remarkable vein
of Albertite worked in this district. They contain numerous remains of
fishes and also of the characteristic Lower Carboniferous plants. (Bailey
and Matthew, Report of 1871. Acadian Geology, P. 231.)

LOWER CARBONIFEROUS PLANTS. 9

In Southern New Brunswick and North Western Nova Scotia the
Millstone Grit is also largely developed. At the South Joggins, where
this formation and the Middle Coal-formation probably attain their maxi-
mum thickness, the equivalent of the Millstone Grit occupies in Sir Wil-
liam Logan's section a vertical thickness of no less than 5972 feet, and
consists of red and gray sandstones, red and chocolate shales and conglo-
merates with some dark shales, underclays, bituminous limestones and
thin unproductive coals. It contains species of Sigillaria, Zepidodendron^
Calamites, Badoxylon and Cordaites. (Logan, Geol. Survey of Canada,
1845. Acadian Geology, P. 176.)

On the south side of the Cumberland coal-field, the Lower Carbonife-
rous beds appear to return to the type of the Bonaventure formation, and
to consist principally of conglomerate and sandstone not rich iu fossil
plants, and these principally of the Millstone Grit horizon.

Crossing the ancient metamorphic ridge of the Cobequids, we find on
their southern flanks conglomerates representing the lowest Carboniferous
rocks. Above these there is a slender developement of the marine lime-
stones and a great thickness of hard sandstones and shales representing
the Millstone Grit and perhaps the lower part of the Middle Coal-forma-
tion. These rocks form a long belt extending from Cape Chiegnecto till
it unites with the Pictou Coal-field on the eastward. Their general
arrangement appears to be that of a narrow trough much broken by
faults. They aflford a good representation of the Flora of the Millstone
Grit. (Acadian Geology, P. 263 et seqr.)

On the south side of Minas Basin and Cobequid Bay a very wide area
is occupied by Lower Carboniferous Rocks ; and at the cUfl" of Horton
Bluff and other places in its vicinity, these beds which, from their large
development in this locality may be named the Horton Series, are very
well exposed and contain abundance of their characteristic fossils. For
their detailed description I may refer to my paper of 1858, Journal of
Geol. Society, Vol. XV., P. 63. (See also Acadian Geology, P. 252.)

Similar rocks are seen and have been described by the author near
Windsor, at Walton and Noel, and at Five Mile River on the Shubena-
cadie, in all these places rising up from under the Lower Carboniferous
hmestones. (Journal of Geol. Society, Vol. IV., P. 59 : Vol. VII., P.
335 ; Acadian Geology.)

Further East, on the Salmon River, and on the West, Middle and East
Rivers of Pictou, there is a great development of rocks of the Millstone
Grit Series, consisting largely of chocolate sandstones and shales, often
very hard and with bands of gray and dark-coloured beds holding plants.
In this region the marine limestones extend upward into the Millstone

10 CANADIAN FOSSILS.

Grit, SO that it is difficult to establish any distinct line of separation, and
the Lower Carboniferous coal measures seem to be absent. (Journal of
Geol. Society, Vol. I., P. 26, 1843. Logan and Hartley, Reports on Pic-
tou Coal-Field, 1869. Acadian Geology, P. 316 et seqr.)

In the Pictou coal-field tjiere are certain hard sandstones holding
obscure fossil plants which come up from beneath the Millstone Grit
on the Middle River, and which I have regarded as Devonian. It is, how-
ever, barely possible that they may represent the Lower Carboniferous
Coal-measures, otherwise wanting in this district.

The great and exceptional conglomerate of the Pictou coal district,
known as the New Glasgow Conglomerate, appears to be a shingle bed of the
UpperMillstone Grit or Middle Coal-formation epoch. It stretches with some
interruptions from Merigomish to Roger's Hill and jNIt. Dalhousie near
the Eastern end of the Cobequid Ridge, or about twenty miles, and is un-
doubtedly connected with the different development of the beds of the
Coal-formation on the South and North of this line ; and it implies very
great and violent denudation of the Lower Carboniferous sandstones during
the Coal-formation period, as the fragments contained in it are largely
composed of these sandstones, and are often of great size. (Acadian
Geology, P. 321 et seq. Logan, Report on Pictou, 1869.)

At the extreme eastern end of the Pictou coal-field, where it is in con-
tact with the Upper Silurian at McCara's Brook, the Lowest Carboniferous
beds are conglomerates with interstratified trap, above* which is marine
limestone overlaid by the Millstone Grit Series. (Journal of Geol. Society,
Vol. L, P. 329. Acadian Geology, section opposite Page 125).

In the carboniferous area of Antigonish County we again meet with the
dark shales and sandstones of the Horton Group, holding their character-
istic plants and underlying the marine limestones and gypsums. I noticed
these beds as occurring at Right's River in 1843, * and Dr. Honeyman,
who subsequently traced them farther to the Eastward, has kindly placed
in my hands a small but interesting collection of their fossil plants.

The long belt of Carboniferous Rocks extending along the West Branch
of the St. Mary's River, has the mineral character and fossils of the
Millstone Grit Series in those places where I have examined it, except
near Guysboro, where there are Lower Carboniferous hmestones, and in
the Strait of Canso, near C. Porcupine, where the basal conglomerates
appear. (Acadian Geology, P. 350).

In Cape Breton a well-characterised representation of the Lower Car-
boniferous Coal-measures or Ilorton Series is seen in the sandstones, gray

• Journal Geol. Society, Vol. I., Page 329.

LOWER CARBONIFEROUS PLANTS. * 11

and black shales and conglomerates which underhe the limestone and
gypsum of Plaister Cove, while the Millstone Grit seems to be represented
by the thick sandstones underlying the Coal-field of Richmond County.
(Journal of Geological Society, Vol. V.; Acadian Geology, P. 390 et seqr.)

In Northern Cape Breton, from the Cape Dauphin section, as described by
Mr. 11. Brown, it would appear that the Lower Carboniferous Coal-
measures are slenderly represented or concealed by faulting. Mr. Brown
has, however, recognised the Millstone Grit as underlying the Sydney
and Glace Bay Coal-fields, and attaining to a thickness of 1800 feet. It
consists largely of gray sandstone, and holds SUjillariae, Calamites and
Zepidodendra, (Bro^Yn, Journal Geol. Society, Vol. Ill, P. 258. Ibid.
Vol. VI., P. 116.)

From a collection of fossils made by Mr. R. Bell in Western Newfound-
land and presented to the Museum of the McGill University by Donald
Ross, Esq., it appears that the Lower Carboniferous limestone of that
Island holds the same fossils with that of Nova Scotia, and that it is over-
laid by a series of beds corresponding to the Millstone Grit. This forma-
tion, however, contains beds of coal of workable size, abounding in remains
0^ Lepidodendra, so that it would seem that in Newfoundland, as in Scot-
land, the workable coals extend farther down in the series than is the case
to the southward.

The following table will clearly show the relations of the formations in
question in the more important localities referred to above :

12

c3

CG

o

id
o

a

O

1-:!

i-i

J
in

En

<»-. .^ ^

-T !n *

.^^

c = o

O -> 5 •

o

5 ^

§-":h|

.1^ - «

0-' o.'ii

6

a

III

fe -i^

c^

a, - O

8 of Plaisti

I'assago

kc, C. D

associate

o

kO

_3

<
a

o

f=4

" O 1. o

S>;^3

c 2 S

ll

CO

-^ X _o_a

tr S *^ -.

M

1
1

"Si? = ^

2 C^ o

^5

'-O

t-3

M

D

6

■§.^«b

o 2

^■■^ S cS

^ '^ '■

oO

s^ga

-sS^ 2

^•3

>- 'rH

o 2-g .

cS dj

cup

CC' "

Si

si

o

5 = a -
m 3 5j S

.2 o 5 o >

.?; s "S -^

C M o

3¥.

3

'S
_3

CO
u
o

Oh

i=«a^s

E^SsS

g^S

;§

K<J

X

1-?

CO

§ [

o'S 0 .

oS i

J-" 3

»-.

<1 j

^-^

.2

5®

a.

of Wir
adie. Stewi
Jrookfield,
atud beds.

Sandston
lalf-way 1
Noel, Wi
liivcr, &c

t>

S ^= X

m g-'o

'S~' . O

s

rt,^— cJ

Q) 5 , O

r^ - t- ^^

a
.2

5

fill

o sS'H
Exol

-rf

O

uj

M

i&

O t£ o

C 3

1=:

o .

III

3 5?

o s

n

o

CI

S cJ'Si

a; ^^ y
o re's -2

11 :i

alcareous and Bitumino
Shales of Albert Mil
llillsborougli.

1

■« S

3.2

.2 2

o
Q

o

c»

J

O

«

2

"O S

o

>

3

a

o
3

P/2

CO o

a

In

o

fcH

15 -;

>

3

_^ 3

a.3

1^
■ U

ci

o

-«*.>

OS'S

■5

3
o

o S

3 a

ceo
ci

3i3
"" a

2

S

"S 3
7- .2

ll

is
o

P 3

1^

Co 3

o -. o

cog

fi

r»

■^3
5'^

trsS

a

3

«

3 o 2

Q

z

CC

y .D W

' ^ •

2

5 —

.5 o

is

> ^

6

2

s

wer Carhonifer
Limestone.

Windsor Group.

ti i

11
a i

c =

?3

^1 i
1^ a

^g

^ -

^

LOWER CARBONIFEROUS PLANTS. 13

The above introductory details are rendered necessary by the doubts
and obscurity in which the Lower Carboniferous plant-bearing beds have
been involved. I would only farther make the following general remarks.

1. Lines of separation between the different members of the Carboni-
ferous must necessarily to be some extent arbitrary, and may differ in
different localities.

2. Wherever the marine members of the system are developed, a mark-
ed difference is observable in the flora under and above those members,
depending on the time during -which the area was submerged ; but where
■the marine members are not developed, a gradual passage in the flora may
be observed.

3. Marine limestones occur locally even in the Upper Coal-formation,
-and these must of course not be confounded with the true Lower Carboni-
ferous limestones. I have also suggested in Acadian Geology that in
those parts of Nova Scotia where the marine limestones are largely deve-
loped in thickness, as in Hants county, they may possibly encroach on the
time elsewhere occupied by the Millstone Grit and Middle Coal formation,
just as in other locahties, as at Salmon River, where the Millstone Grit is
largely developed, isolated beds of limestone sometimes occur in its
upper part.

4. In the area to which this report relates, the close of the Devonian
was accompanied by great physical changes which removed the Devonian
flora. In the Lower Carboniferous period a meagre flora different from
that of the Devonian took possession of the land. This was again partially
removed by the subsidence leading to the deposition of the Lower Carbo-
niferous limestones, and the Millstone Grit lying on these, forms, as to its
£ora, the dawn of the great Middle Coal-formation.

5. While the local elevation, subsidences and denudations within the
Carboniferous period were sufiicient to cause some limited cases of uncon-
formability*, these are not comparable with those between the Devonian
and the Carboniferous ; and the Devonian fauna and flora are as a whole
quite distinct from those of the Carboniferous,! though there are some
species of plants common.

6. In Eastern America, as in Great Britam, the conditions of coal accu-
mulation seem to have set in earlier to the Northward. The coal beds
of Newfoundland belong to the Millstone Grit series. Those of Pictou are
exclusively in the Middle Coal series, and apparently in its lower part.
Those of the Joggins seem to be rather higher in the series than those of

* Journal of Geol. Society, Vol. 1, P. 32, Bailey's Report on New Brunswick; Acadian
Geology, P. 150, section.

fFor illustrations of the Devonian and Carboniferous fauna in the area in question, 1
may refer to theReports of Mr. Billings and to my Acadian Geolo;fy.

14 CANADIAN FOSSILS.

Pictou, and in the United States there are workable beds of coal in the
Upper-Coal measures which are barren in Nova Scotia. This connects
itself with the fact illustrated in my previous Report on the Devonian flora,
that this flora in North America seems to have extended itself
from the north east ; a view which Heer and Prof. Asa Gray seem also to
entertain with respect to the Tertiary floras.

7. The following remarks on the Lower Carboniferous Plant-beds, pub-
lished in 1858,* should perhaps be reproduced here : —

" In Nova Scotia these older coal-measures, as compared with the true
coal-measures, are more calcareous, abound more in remains of fishes, and
have fewer vegetable remains and indications of terrestrial surfaces. They
occur generally along the margin of the coal-areas, near their old shores ;
and as we might expect in such circumstances, they are associated with,
or replaced by beds of conglomerate derived from the neighbouring high-
lands of Silurian and Devonian rocks. Where these conglomerates are
absent, we usually find very frequent alternations of sandstones with sandy
and calcareous shales, giving a homogeneity of appearance, together
with, at the same time, very frequent changes and diSerences in mineral
character. The general aspect is that of muddy estuarine deposits, very
slowly accumulating, and discoloured and darkened by decaying organic
substances, partly of aquatic, and partly of terrestrial origin.

" Both the supply of sediment and the growth and preservation of
vegetable matter were on a smaller scale than in the coal-period,
the only exception being the bituminous limestone and associated dark
shales of the latter, which in composition and aspect often much resemble
the beds now under consideration.

" These characters cause the Lower Carboniferous coal-measures to
present a very striking contrast with the coarse and often reddish sedi-
ments which prevail in the marine parts of the Lower Carboniferous
series in the area in question.

" Before leaving this comparative view, it is necessary to remark that it is
precisely in those districts where the true coal-measures are least developed
that the lower series is most important. This is not likely to be the result
of accident. It shows that the physical and vital conditions of the coal-
measures originated as early as those of the mountain-limestone, that
locally these conditions may have been contemporaneous throughout the
whole period, but that in some localities the estuary and swamp-deposits
first formed were so completely submerged as to be covered by marine
deposits, while in others early marine beds were elevated and subjected
to the conditions of gradual subsidence and fillmg up indicated in the
great coal-measures of the South Joggins, Fictou.and Sydney."

• Journal of Geol. Society, Vol. XV.

LOWER CARBONIFEROUS PLANTS. 15-

II. PLANTS OF THE HORTON SERIES OR LOWER CARBO-
NIFEROUS COxiL MEASURES.

Coniferce.

Dadoxylon, Endlicher.

Trunks of coniferous trees are found in all the geological formations
from the Lo^yer Devonian to the Modern, and in almost every case sepa-
rated from the foliage and fruit. Hence they have to be classified sepa-
rately ; and botanists, attaching httle value to the characters of the merely
vegetative organs of plants, have in general given them little attention.
Goeppert has, however, by an elaborate investigation established a syste-
matic classification of some coniferous woods, which Kraus and others have
endeavoured to modify and improve.

The difficulties arise mainly from the great similarity in the wood of
somewhat distinct types of conifers, and the differences which are due to
the different states of preservation of the specimens of the same species
of wood. These difficulties can be overcome only by careful and multi-
plied microscopic examinations of numerous examples of wood recent and
fossil, and of the latter in different states of preservation. In this place
we may confine ourselves to the consideration of the Lower Carboniferous
species, and those more nearly allied to them.

Witham in 1833 described several Carboniferous species of pine wood,
under the generic name Pinites, separating under the name Pitus species
which appeared to have the disks on the cell-walls separate and in trans-
verse lines. Witham's name was changed by Goeppert to Araucarites, to
indicate the similarity of these woods to Araucaria, Pinites being reserved
for trees more closely allied to the ordinary Pines. Endlicher, restrictino-
Araucarites to foliage, &c., of Araucaria-like trees, gave the name Dadoxy-
lon to the wood ; and this, through Unger's " Genera and Species," has
gained somewhat general acceptance. Endlicher also gave the name
Pissadendron to the species which Witham had called Pitus ; but Rron-
gniart proposed the name Palaeoxylon to include all the species with
thick and complex medullary rays, whatever the arrangement of the discs.
In Schimper's new work Kraus substitutes Araucaroxylon for Endlicher's
Badoxylon, and includes under Pissadendron all the species placed by
Brongniart in Palaeoxylon.

To understand all this confusion, it may be observed that the characters-

16

CANADIAN FOSSILS.

available in the determination of Palaeozoic coniferous wood are chiefly
the form and arrangement of the wood-cells, the character of the bordered
pores or discs of their walls, and the form and composition of the medul-
lary rays.

The character on which Witham separated his genus Pitus from
Pinites is, as I have ascertained by examination of slices of one of his
original specimens kindly presented to me by Mr. Sanderson of Edinburgh,
dependent on state of preservation, the imperfectly preserved discs or
areolations of the walls of the fibre presenting the appearance of separate
and distinct circles, while in other parts of the same specimens these discs
are seen to be contiguous and to assume hexagonal forms, so that in this
respect they do not really differ from the ordinary species of Dadoxylon.
The true character for subdividing those species which are especially
characteristic of the Lower Carboniferous, is the composite structure of
the medullary rays, which are thick and composed of several radial piles
of cells placed side by side. This was the character employed by Bron-
gniart in separating the genus Palaeoxylon, though he might with con-
venience have retained Witham's name, merely transferring to the genus
the species of Witham's Pinites which have complex medullary rays.

The following table will clearly show the distinctive characters and
relations of the genera in question, as held by the several authors above
referred to : —

Wood of Palceozoic Conifers.

"Woody Fibres.

Medullary Rays.

Generic Names.

Geological Age.

No Discs.

One or two Series of
Cells.

Aporoxylon, Unger.

Devonian.

Spiral ligneous lining,
Discs few or none.

Uncertain, probably
composite.

Prototaxites,* Dawson.

Devonian.

Discs in one Series con-
tiguous, or in several

Complex or of two or
more Series of Cells.

(Pitus, Witham.

J Palceoxylon, Brongniart.

( Pissadendron, Endlicher.

Middle & Lower Car-
boniferous and De-
vonian.

Series spirally ar-
ranged.

Simple or of one Kow
Cells.

f Araucarites, Goeppert.
J Dadoxylon, Endlicher.
( Araucaroxylon, Schimper.

Upper Carboniferous
and Permian.

*The criticisms of Mr. Carruthers on this genus, 1 purpose to meet in a proper manner elsewhere.

LOWER CARBONIFEROUS PLAXTS. 17

The two first genera above named are not known to ascend above the
Devonian. The third begins as far as known in the Middle Devonian
and extends to the Middle Carboniferous. The fourth begins in the Middle
Carboniferous ^and extends to the modern world. The true Dadoxjlons
have large transversely floored pith-cylinders of the type of Sternbergia,
and this is also known to have been the case with some species of the
Palaeoxylon group. As I have elsewhere shown, there is reason to believe
that the foliage of the true Dadoxylons was that known to Botanists as
Walchia (Report on Prince Edvvard Island, 1871), and especially cha-
racteristic of the Upper Carboniferous and Permian.

In Nova Scotia D. (Palaeoxylon) antiquius is characteristic of the
Lower Carboniferous Coul measures ; B. (Palaeoxylon) Acadianum of the
Millstone Grit and Middle Coal formation, and D. materiarium of the
Upper Coal formation. D. antiquius, described below, is of precisely the
same type with Witham's Pitus antiqi,a and P. primaeva from the same
horizon (Tweedian) in Scotland, but is specifically distinct.

Dadoxylon (Palaeoxylox) antiquius, Dawson. — Plate I. Figs. 1 to 5.
(Cand. Nat. Vol. VIII., 1863. Journal of Geol. Society of London,
May 1866. Acadian Geology, P. 473 & 425.)

Description. — Wood-cells narrow, thick-walled, with two or three rows

of discs, which are contiguous, alternate and hexagonal, and

contain oval oblique pores. Medullary rays of three or four

series of cells with twenty or more superimposed, nearly as wide

as the wood-cells. Rings of growth indistinct but visible.

This species differs from Pitus (^Palaeoxylon) antiqua of Witham in

its longer and narrower medullary rays, and its narrower wood-cells with

smaller pores and thicker walls. It is a distinct though closely allied

species.

It is unfortunate that I have given a specific name approaching so near-
ly to that of Witham ; but as the species are closely allied and discoveries
of their other organs may necessitate different names, I think it better to
leave it for the present.

The wood-cells of this species present in the cross-section that thick-
walled appearance and rotundity of the cavity of the cells which causes
them to appear to have intercellular matter and to resemble Prototaxites or
very young tixine twigs. Detailed comparisons, however, lead me to
believe that this results neither from any distinct structure interposed
between the fibres nor from any proper lining of the cell-walls, but that it is an
effect of long maceration in water, causing a swelling of the cell-walls,so that
specimens of the same wood differently preserved may either appear with

B

18 CANADIAN FOSSILS.

thickened cell-walls, cylindrical or nearly so, or with square and coitt-
paratively thin-walled cells or fibres.

The only specimen of this species at present in my collection is a frag-
ment of a trunk collected by the late Dr. Harding at Horton, and showing
neither the pith nor the outer bark. No other parts of the tree have in
so far as I am aware been found.

The figures show magnified camera tracings without any reference to
pictorial effect.

Sigillarice.

I have as yet found no well characterised stems or leaves of Sigillarise
in the Lower Carboniferous Coal-measures. The old stems of Lepidoden-
aron corrugatum in certain states assume a vertically ribbed appearance,
but this has no connection with the true ribs of Sigillariae, being only a
vertical cracking of the bark, owing to the expansion of the stems. It
occurs also in the Cyclostigma of the Upper Devonian of Ireland.

One small and flattened specimen of fossil wood found at Horton Bluff,
has the characteristic rariporous tissues of Sigillaria, but it shows none of
the external markings. I have figured its structure in Fig 6.

Specimens of the ordinary Stigmaria ficoides are found at Horton
Bluff, and also at Five-mile River, and also the variety stellata with radiat-
ing marks around the areoles. A fine specimen of this last was in the col-
lection of the late'Dr. Harding of Windsor, but I have not now access to
it.

The Stigmariae figured in Plate IV, I believe from their associations to
belong to Lepidodendron, but it is not at present possible to distinguish by
external characters alone the Stigmarise of this genus from those of Sigilla-
rise.

Eichwald mentions as fossils of the Lower Carboniferous of Russia, S.
reniformis and the stellate and ordinary varieties of Stigmariae.

Equisetacece.

Calamites.

Small stems of a Calamite, apparently 0. cannaeformis, or C. trami-
tionis (radiatus} are the Lower Carboniferous of Horton.

LOWER CARBONIFEROUS ILANTS. 19

Lycopodiacece.
Lepidodendeon.
Lepidodendron corrugatum, Dawson. — Plates II, III, IV. Plate V,
Figs. 33 to 3G & 38.

(Quarterly Journal of Geological Society, Vol. XV. Acadian Geology,
P. 451.)
Description : —

Ealit of groiith. — Somewhat slender, with long branches and
long slender leaves having a tendency to become horizontal
or drooping.
Markings of Stevi. — Leaf-bases disposed in quincunx or spirally,
elongate, ovate, acute at both ends, but more acute and
slightly oblique at the lower end ; most prominent in the upper
third, and with a slight vertical ridge. Leaf-scars small,
rounded, and showing only a single punctiform vascular
scar. The leaf-scar on the outer surface is in the upper third
of the base ; but the obliquity of the vascular bundle causes
it to be nearly central on the inside of the epidermis. In young
succulent shoots the leaf-scars are contiguous and round as
in Cyclostigma, without distinct leaf-bases (Fig. 83). In
this state it closely resembles L. OUvieri, Eichwald. *

In the ordinary young branches the leaf-scars are con-
tiguous, and closely resemble those of L. elegans Erongt.
(Fig. 13). As the branches increase in diameter the leaf-
scars slightly enlarge and sometimes assume a verticillate
appearance (Figs. 14, 15). As they still farther enlarge they
become separated by gradually increasing spaces of bark,
marked with many waving strine or wrinkles, (Figs. 16 to 20.)
At the base of old stems the bark assumes a generally wrinkled
appearance without distinct scars (Figs. 21 & 32.)

The forms represented in Figs 34, 35, 36, I was at first
disposed to regard as specifically distinct ; but I now think
they may be merely varietal. Fig. 36 shows a ribbed appea-
rance, and the scars arranged in vertical rows. It may be
compared with the Knorria form, Fig. 29. It is deserving
of remark that the inner surface of the epidermis in the old
stems is more delicately corrugated than the outer surface.
Knorria or decorticated states. — Of these there is a great variety,
depending on the state of preservation, and the particular
• Lethaea Rossica Plate Y, Figs. 12, 13.

20 CANADIAN FOSSILS,

lajer of bark forming the actual surface. Fig. 24 shows the
common Knorria form with the leaf-bases represented by
longitudinal ridges. Fig. 26 shows a form in which the vascu-
lar bundles appear as cylindrical truncate projections. Fig.
27 shows the leaf-bases prominent, and Fig. 27 and 28 show
the appearance of longitudinal ribbing produced bj the ex-
pansion of the bark. Fig. 82 shows the decorticated base of
an erect stem. Fig. 88 shows one of the deeper layers of the
bark with leaf-scars of a transverse form, the ordinary form
being seen in the lower part of the same specimen.
'Structure of Stem. — This is not perfectly preserved in any of my
specimens, but one flattened specimen shows a central
medulla with a narrow ring of scalariform vessels surround-
ing it, and constituting the woody axis. The structure is thus
similar to that of i. Sarcourtii, which I regard as probably
the same with the closely allied European species L.
Veltheimianum.
Leaves. — These are narrow, one-nerved, curving somewhat rapidly
outward, Figs. 10, 11, 12, 23, 25. They vary from one to two
inches in length.
Roots. — I have not seen these actually attached, but they occur
very abundantly in the underclays of some erect forests of
these plants at Horton Bluff, and are of the character of
Stigmaripe, Figs. 30, 31. In some of the underclays the long
flattened rootlets are excessively abundant, and show the
mark of a central vascular bundle.
Fructification. — Cones terminal, short, with many small acute
imbricate scales. Spore-cases globular, smooth (Figs. 10, 22.)
On the surface of some shales and sandstones at Horton
there are innumerable round spore-cases of this tree about
the size of mustard seed. Fig. 22. Large slabs are some-
times covered with these, and thin layers of shale are filled
with flattened specimens.
This is the characteristic species of the Lower Carboniferous Coal-measu-
res, occurring in great profusion at Horton Bluff and its vicinity, also at
Sneid's Mills near Windsor, Noel and Five-Mile River, at Norton Creek
and elsewhere in New Brunswick (^Matthew's Col.), and at Antigonish
(Honeyman's Col.)

I have received from the lowest carboniferous beds of Ohio specimens
of this species.* According to Rogers and Lesquereux similar species

* Journal of Geo. Society, Nov. 1862, P. 313.

LOWER CARBONIFEROUS PLANTS.

21

exist in the Vespertine of Pennsylvania, associated with Sticrmaria.
L. ohsmrum, Les, from the Lower Carboniferous of Illinois resembles
decorticated specimens of this species, and the fossil from the same re<non
referred with doubt by that author to L. Velthelm> anum may be the same.
The Eiropaan equivalents of this species are undoubtedly L. Crlincanum,
Eichwald, from the Lower Carboniferous of Russia, and i. Veltheimiamim,
Sternberg, from the Lower Carboniferous of Western Europe. So closely
indeed does the last species resemble L. co^rugatmn, that Schimper'and
other European Palaeo-botanists conversant with the protean forms of
these species ami knowing ours only by imperfect figures, may well be
excused for regarding them as identical. They are undoubtedly related
species of the sama group, and in habit of growth L. comigatum is in

some respects intermediate between the others,
seen in the following: table : —

Leave?

Branches.

Stems.

L. Glincanum.

Long-linear, erect

( Stout. Leaf-bases elon-
-I gate, elliptic, contigu-
(_ ou3,with a strong keel.

( Leaf-bases rerr long,
{ larger than in next
■j species, narrowed at
] extremities, separated
[ by narrovr lines.

L. comigatum.

Linear, horizontal
curved downward.

Their diiferences will be

L. Veltheimianum.

Linear, short, curved up-
ward.

Old Trunks.

Leaf-scars.

Strobiles.

( Leaf-bases widely s
\ parated by corrugat(
( bark.

36-

gated

f Round or oval with
( three vascular marks

lu

(

Unknown.

Spore-cases j Cnknnwa.
] Unknown.

Slender. Leaf-bases with
a short keel, contigu-
ous, elongate, elliptic.

Leaf-bases long, acumi-
nate below, separated
by narrow lines.

Leaf-bases widely sepa-
rated by corrugated
bark.

Rhombic or rounded,
very small, only one
distinct mark.

Small, with
scales.

triangular

Slender. Leaf-bases elon-
gate, rhombic, keeled.

Leaf-bases oblong, rhom-
bic, acuminate, sepa-
rated by narrow lines.

Leaf-bases slightly sepa-
rated. Bark splitting
in long gashes.

Transversely
with three
marks.

rhombic,
vascular

Internal
structure

Long and cylindrical,
scales elongate.

Probably globular.

L. Harcourtii, probably
belongs to this species.

Globular, smooth.
Type of L. Harcourtii

In the Knorria or decorticated states I do not think that these three
species can be clearly distinguished. L. Glincanum^ however, presents
coarser markings than the others. In the above diasrnoses I have relied
for the European species mainly on the figures and descriptions of Eich-
wald and Schimper.

The relations of this plant to the Devonian Cyclostigma are of some
interest. There can be no doubt that L. corrugatum, and its allies approach
more nearly to Cyclostigma than do m ost of the other Carboniferous Lepi-

22 CANADIAN FOSSILS.

dodendra. This resemblance consists in two facts. (1). The round shape
of the young leaf-bases. (2). The expansion of the bark in such a man-
ner in the course of growth as to separate the leaf-scars by finely corru-
gated spaces of epidermis, whereas, as I have elsewhere pointed out, * in
the ordinary Carboniferous Lepidodendra the leaf-bases themselves expand
with the growth of the stem or they remain unchanged and become sepa-
rated by deep gashes of the bark.

The differences, however, are more important. They are as follows : —
(1). In Cyclostigma the leaf-scars remain always round and unipunc-
tate, and they never acquire leaf-bases properly so-called. The characters
which exist only in young succulent stems of Lepidodendron corrugatum are
persistent in Cyclostigma, which is thus an inferior form. (2). The stro-
biles of the Devonian Cyclostigma of Ireland are very different in the form
of the scales from those of Lepidodendron, though the spore-cases are
similar to those of the species now under consideration. (3). Cyclostigma
is in so far as known confined to the Devonian period.

The above are the matured conclusions from careful examination of
specimens from Ireland kindly communicated to me by Mr. Bailey. Heer
and Carruthers also have argued this question in the Journal of the Geo-
logical Society, and are both in my judgment somewhat in error, though
in opposite senses. The former attaches too much importance to detached
fragments of Cyclostigma, and is in error as to its geological horizon ; the
latter confounds Cyclostigma with Lepidodendra of the type of that now
under discussion.

For some of my specimens of L. corrugatum I am indebted to Mr.
Matthew of St. John, Prof. Hartt of Cornell University, Mr. Elder late of
Acadia College, Horton, and Dr. Honeyman.

Lepidodendron Sternbergii, Brongt. L. obovatum and L. dicho-
tomum, Sternberg. Plate VI. Figs. 42 to 45, and 42 a b and 44 a.

This species, as found in the Lower Carboniferous of Horton, was
figured by me under the above name in the Journal of the Geological
Society, vol. XV, and such specimens as I have since obtained have con-
firmed the determination. Of the various forms referred to this species,
my specimens resemble those figured by Schimper in plate LVIII, Fig. 4,
and in plate LX, Figs. 3, 4, 5. They are identical with L. (^Sagenaria)
dichotomum as figured by Geinitz, in the Plants of the Coal-formation of
Saxony, Plate III, Figs. 2 to 12,

It is, however, to be observed that the Horton specimens resemble i.
Cottaii of Sauveur, in the only important point in which that species seems

• Journ. of Geol. Society, 1866. P. 713.

LOWER CARBONIFEROUS PLANTS. 23

to differ from L. Stembergii, namely in the leaf-bases being truncated at
the lower end. This maj, however, be an accident of growth.

Though L. Sternhergii is found in the Lower Carboniferous, it is more
abundant in the Middle Coal-formation. The specimens figured are from
Horton. Fig. 43 represents a young branch, Fig. 42 a somewhat older
branch, and Figs 44, 45 a still older stem in which the leaf-bases alone
are preserved without the scars. The cone and scales represented in Fig.
41 may possibly belong to this species.

Lepidodendron tetragonum, Sternborg. — L. quadrangulare, linger. —
Plate V. Fig. 39, 39 a.

The species named L. tetragonum by Sternberg was founded on a spe-
cimen too badly preserved to show its distinctive characters, and linger
identifying this with another and possibly distinct species also described
by Sternberg, has given the name quadrangulare to both. Geinitz has
used the name for another species, possibly a Lepidophloios, and Goeppert
has identified it with his L. sexangidare, which seems to include both a Lepi-
dophloios QXid a Lepidodendron closely allied to, if not identical with
num, Schimper has mixed up my Lepidofliloios tetragonus with Geinitz'
Lepidodendron tetragonum, a totally distinct form, and with a species named
by Presl quadrata and referred by him to his sub-genus Bergeri i.

In the midst of this confusion I think it best to fall back on Stern-
berg's name, which certainly applies to something very near to the present
species, and which I used to designate it in Acadian Geology and in my
Synopsis of Carboniferous fossils in 1863. The characters of the genera
Bergeria and Lepidophloios certainly do not apply to tliis plant. It is a
true Lepidodendron, though with remarkably broad and regularly rhombic
areoles. Unfortunately its leaves and fruit are unknown to me, though
specimens of the bark of the stem have been obtained both from the Middle
Coal-formation and Lower Carboniferous Coal-measures.

Its characters will be seen by a glance at Fig. 39. The leaf-bases
are flat, of a rhombic form, approaching to square, and separated by deep
and wide grooves. The leaf-scar is small and of the same form with the
leaf-base, and situated quite at its upper end. In my specimens it shows
only the central vascular mark. Lepidodrendra of this type are charac-
teristic of the Lower Carboniferous in Europe as well as in America.
Mr. Carruthers has figured a similar Lepidodendron collected by Mr.
Daintree in Queensland, Australia, in rocks supposed by him to be Devon-
ian. Mr. Selwyn has obtained similar specimens in the Carboniferous of
Victoria ; so that about the beginning of the Carboniferous period species
of this group seem to have been very widely distributed.

2-k CANADIAN FOSSILS.

Lepidodendron fenestratum of Eichwald may be a decorticated form of a
similar species, but if it shows the outer bark, it should perhaps be referred
to Lepidophloios.

In comparison with this plant I have represented in Fig. 40 a specimen of
a strobile deprived of its outer scales, and showing rhombic scars with a
sub-central vascular mark. This strobile is from Horton and may be part
of a mature fruit of L. corrugatum.

LepidodendPwOX aculeatum, Sternberg. Plate V. Fig. 37, & 37 a.

To this species I would refer the imperfectly preserved specimen from
Horton figured in Plate V, Fig. 37. It is from the collection of Prof.
Elder, and is clearly distinct from any of the above, and not distinguishable
from specimens of i. aculeatum in a similar state from the Millstone Grit and
Middle Coal-formation.

It will be observed that of the above four species, X. corrugatum is spe-
cially distinctive of the Lower Carboniferous, in which it is extremely
abundant. The others are common to the Lower Carboniferous and true
Coal-measures, and attain their maximum of abundance in the latter.
Indeed, the specimens of them found in the Lower Carboniferous are so
few that only extensive collection could have discovered them ; and in
this matter I am much indebted to the co-operation of Prof. Hartt and
Prof. Elder, both of whom made large collections when resident in Hor-
ton, and have liberally given me access to them.

DiPLOTEGIUM.

■ Plate VI, Fig. 46, 46 a.

To this genus I may with doubt refer the specimens one of which is
represented in the figure above referred to. Similar plants in a better
state of preservation and probably specifically distinct occur in the Middle
Coal-formation.

It is difficult to form any opinion of the precise nature of these plants.
They appear to have borne flat leaves or leaf-bases, and were probably
allied to Lepidophloios. In Nova Scotia they are rare, and always, so far as
I have seen, in small fragments.

Lycopodites.

Lycopodites plumula. S. N. Plate I. Figs. 7, 8, 9.

Description. — Stems slender, branching, slightly corrugated or tuber-
culated, and bearing flat linear leaves either pinnate or

LOWER CARB0XIFER0U3 PLANTS. 25

apparently so o-wing to compression. The leaves are par-
allel-sided, obtuse and nerveless, and are slightly constricted
and decurrent at the base. The tissues of the stem seem to
have been dense and are carbonaceous, but show under the
miscroscope only what appear to be remains of cortical
cells. The specimens were collected by Rev. Dr. Honey-
man in Lower Carboniferous shale near Springville on the
East River of Pictou.
This plant belongs apparently to the same group of so-called Lycopodites
with L. Vanuxemii of the Devonian of New York* and L. pennceformis,
Goeppert, from the Jungste Grauwacke of Silesia. Filicites gracilis of
Shumard from the Lithographic limestone of the Upper Devonian in
Missouri is very similar. Whatever the nature of these curious fossils,
they seem to be characteristic of the Lower Carboniferous and Upper
Devonian.

The botanical relations of these plants must remain subject to doubt,
until either their internal structure or their fructification can be discovered.
In the mean time I follow Goeppert in placing them in what we must
regard as the provisional genus Lycopodites. On the one hand, they are
not unlike the slender twigs of Taxodium and similar Conifers, and the
highly carbonaceous character of the stems gives some colour to the sup-
position that they may have been woody plants. On the other hand, they
might in so far as form is concerned be placed with algae of the type of
Brongniart's Chondrites obtusus, or the modern Caulerpa plumaria. Again,
in a plant of this type from the Devonian of Caithness to which I have
referred in a former memoir, the vernation seems to have been circi-
nate, and Schlmper has conjectured that these plants may be ferns, which
seems also to have been the view of Shumard. The following remarks
on this group of plants are copied from a paper by the author, on new
Devonian plants, in the proceedings of the Geological Society, March
1871.

" In his recently published ' Paleontologie,' Schimper (evidently from
inattention to the descriptions and want of access to specimens) doubts the
Lycopodiaceous character of the species of this genus described in my
papers in the Journal of this Society from the Devonian of America. Of
these L. Richardsoni and L. Jlatthewi are undoubtedly very near to the
modern genus Lycopodium. L. Yanxixemii is, I admit, more problemati-
cal ; but Schimper could scarcely have supposed it to be a fern or a fucoid
allied to Caulerpa had he noticed that both in my species and the allied

• Report on Fossil plants of Devonian, <fec. Journal of Geol. Society, XVIII, 314, PI
XYII.

26 CANADIAN FOSSILS.

L. pemi'xformis of Goeppert, which he does not appear to notice, the pin-
nules are articulated upon the stem, and leave scars where they have
fallen off. When in Belfast last summer I was much interested at finding
in Prof. Thomson's collection a specimen from Caithness, which shows a
plant apparently of this kind, with the same long narrow pinnae or leaflets,
attached, however, to thicker stems, and rolled up in a circinate manner.
It seems to be a plant in vernation, and the parts are too much crowded
and pressed together to admit of being figured or accurately described :
but I think I can scarcely be deceived as to its true nature. The circi-
nate arrangement in this case would favour a relationship to ferns ; but
some Lycopodiaceous plants also roll themselves in this way, and so do
the branches of the plants of the genus PsihpJiytony

Goeppert has described from the Carboniferous Limestone of Sile sia*
a plant which he names Cycadites taxodinvs, and which has some resem-
blance to the above species ; this plant has, however, a distinct midrib in
each leaflet.

A resemblance also suggests itself to Graptolites of the genus Rastrites
but the structure and mode of preservation of these fossils show that they
were not of the corneous nature of GraptoKtes. The somewhat problema-
tical Buthograptus from the Trenton limestone of Wisconsin, which Hall
seems to regard as very doubtfully allied to Graptolites, more closely
resembles in general form the plants now under consideration, and may
have been aMied to them.

CORDAITES.

C. BORASSIFOLIA, Corda, must be reckoned as one of the plants of this
formation, though more abundant in the Millstone Grit and Middle Coal-
formation. At Horton there also occurs a narrow leaf, probably of this
genus, and which seems to have been very long and parallel-sided. I
have represented a fragment of it in Plate VI. Fig. 49.

FiKces.

Cyclopteris (Aneimites) Acadica, Dawson. Plate VII. Figs 53 to
63.— (Journal of Geological Society, Vol. XVII, P. 5.

Description. — Frond very large, the main stipe being sometimes three
inches in diameter. Stipe finely and regularly striate,
dichotomous several times, and finally loosely pinnate, each
pinna having a long petiolule, and bearing a group of three

Jahrback 1866.

LOWER CARBONIPEROUS PLANTS. 27

or more pinnse which are flabellate and simple or divided into
round lobes, with radiating and bifurcating veins. Fertile
pmnse with recurved petiolules, and borne on the divisions of
the petiole near their origin. Form and arrangement of
thecge unknown.

This magnificent fern is eminently characteristic of the Lower Car-
boniferous Coal-measures. It is found in great abundance on the surface s
of shales at Horton, and has also been found at Norton Creek, New
Brunswick, by Mr. Matthew, and in fragments at many other places. Its
representative species in Eastern Europe is C. nana of Eichwald, and in
the culm of Germany C. tenuifolia of Goeppert. It is probably an ally of
C. obtusa of the Devonian.

There can be no doubt as to the manner in which its fructification was
borne, in which respect it resembles the modern ferns of the genus Anemia
and also the Devonian ferns of the genus Archaeopteris ; but unfortunately
the precise form of its spore-cases is unknown.

In the mean time I adhere to my sub-genus Aneimites established in
I860*, and place with this plant A. obtusa, A. BocBUi and A. valida of
the Devonian. For the reasons of this, and for a discussion of the
affinities of these ferns, I refer to my Report on the Devonian and Silurian
Plants, P. 48. In the plate, Figs 52, 53, 54, 55, show the stipe and its
divisions. Figs 61,62 and 54 a, show the recurved petioles, on which the
fructification was borne. Fig. 03 is a portion of a frond in vernation.
Fgs. 56 to 60 show the pinnules and venation.

Hymenophyllites ? — (Plate VII Figs. 64, 64 a.)

Fragments of a fern which may belong to this genus have been found
by Mr. Matthew at Sneid's Mill, Windsor, and are represented in the
figure above referred to. They may, however, be fragments of petioles
of a delicate frond of which the pinnules have perished through decay.

Rhachiopteris.— (Plate VI, Figs. 47, 48.)

This is a peculiarly pitted fern-stipe which must belong to some fern
different from the above and perhaps dotted with hairs or ramenta. It
resembles the Devonian stipe which I have named B. 'punctata, and also
the plant described by Eichwald under the name Schizopteris foveolata.

* Journ. Geol. S. Vol. XVII.

28 CANADIAN FOSSILS.

Fruits, (&c.
Cardiocarpum.
Cardiocarpum tenellum S. N. — (Plate VI, Fig. 50, 50 a.)

Description. — Elongated oval, margin very narrow, a slight notch at
the apex leading to a median furrow. Nucleus small, striated
longitudinally, with traces of transverse striae.
This little fruit, which must for the present be placed in the genus

Cardiocarpum, is found scattered over the surfaces of shales at Horton,

but not abundantly.

Pinnularia.

Pinnularia crassa, Dawson. Acadian Geology, P. 480. — (Plate VI,
Fig. 51.)

The figure represents a portion of a slab covered with fragment of this
plant from Horton. It was described in Acadian Geology as probably a root
of Asterophyllites, but no species of that genus has yet been found in our
Lower Carboniferous coal measures, and though probably a root, it remains
uncertain to what plant it belonged. It differs from P. Oapillacea in its
coarser texture.

LOWER CARBONIFEROUS PLANTS. 29

III. PLANTS OF THE MILLSTOXE GRIT.

Coniferce.

Dadoxtlon (Paljeoxylon) Acadianum, Dawson. — (Cand. Nat. 1863.
Journal of Geol. Society, 1866, PI. Y. Acadian Geology, P. 473.

I may refer to the works above cited for the description of this species,
which is the characteristic fossil pine of the Millstone Grit and Middle
Coal-formation, as D. materiariwn is of the Upper Coal-formation. I
have not yet found a Sternbergia pith in D. Acadianum ; but loose Stern-
bergiae, which may have belonged to it, sometimes occur in the sandstones
of the Millstone Grit.

Sigillarice.

Both drifted and erect -Sigillariae occur in the middle division of the
Millstone Grit at the South Joggins (Division 6 of Sir W. E. Logan's
Section), but I have not any specimens sufficiently perfect for deter-
mination. Stigmariae are not infrequent, and I have specimens of this
genus from this horizon in several other parts of Nova Scotia and also in
Newfoundland.

Equisetacece.

Plants of the genus Calamites are numerous in the Millstone Grit, and
in this respect that formation agrees with the L^pper Coal-formation, in
which Calamites again become prevalent. The most abundant specimens
in my collection belong to three forms, one of which I regard as identical
with C. Cistii of Brongniart. Another is a broad-ribbed species which
may belong to C. Steinhaueri or C. canncefornv's, but is too imperfect for
description. A third is represented by numerous and very perfect
8pecim3n5, and is apparently C. undulatus of Brongniart. The two first
require merely a passing notice. The last deserves more detailed
consideration.

Calamites Cistii, Brongt. — (Plate VII, Fig. 65.)

This species is found in the Millstone Grit ; at Apple River ; Riversdale ;
St. Mary's River; St. George's Bay, Newfoundland; &c. It ranges
from the Millstone Grit to the Upper Coal-formation.

30 CANADIAN FOSSILS.

Calamitbs cann^formis, Brongt.

Specimens possibly referable to this species are found at Calvary River
near Riversdale. It occurs in this horizon elsewhere, and also in the
Devonian.

Calamites undulatus, Brongt. (Plate VII, Figs. 66 to 73.)

This species is stated by Brongniart to be distinguished from the
C. Suckoviif the characteristic Calamite of the Middle Coal-formation, by
its undulated ribs marked with peculiar cellular reticulation. He
suggests that it may be merely a variety of C. SucJwvii, an opinion in
which Schimper coincides ; but since I have received large additional
collections from Mr. Elder, containing not only the stems and branches
but also the leaves and rhizomes, I am constrained to regard it as a
distinct though closely allied species. I shall endeavour to illustrate
these points with the aid of the Figures in Plate VIII.

The rhizomata, Figs 73, 73 a, 73 b, are slender, being from one to two
inches in diameter and perfectly flattened. They are beautifully covered
with a cellular reticulation on the thin bark, and show occasional round
areoles marking the points^ of exit of the rootlets. I have long been
familiar with irregular flattened stems thus reticulate, but have only
recently been able to connect them with this species of Calamite.

The main stems, Figs. 66 to 68, present a very thin carbonaceous bark
reticulated like the rhizomes. They have flat broad ribs separated by
deep and narrow furrows, and undulated in a remarkable manner even
when the stems are flattened. This undulation is, however, perhaps an
indication of vertical pressure while the plant was living, as it seems
to have had an unusually thin and feeble cortical layer, and the undula-
tions are apparently best developed in the lower part of the stem. At
the nodes the ribs are oftsn narrowed and gathered together, especially
in the vicinity of the rounded radiating marks which appear to indicate
the points of insertion of the branches (Fig. 67). At the top of each
rib we have the usual rounded areole probably marking the insertion of a
primary branchlet (Figs. 68, 68 a.)

The Branches (Figs. 70, 71) have slender ribs and distant nodes, from
which spring secondary branchlets in whorls, these bearing in turn small
whorls of acicular leaflets much curved upward (Figs. 72, 72 a, 72 b),
and which are apparently round in cross section and dehcately striate
They are much shorter than the leaves of Calamites Suckovii, and are
less dense and less curved than those of 0. nodosus, which I believe to
be the two most closely allied species.

LOWER CARBONIFEROUS PLANTS. 31

Lesquereux notices this species as characteristic of the Lower part of
the Carboniferous in Arkansas.

It will be observed that I regard the striated and ribbed stems not as
internal axes, but as representing the outer surface of the plants. This
was certainly the case with the present species and with (7. Suchovii and
C. nodosus. Other species, and especially those which belonged to
Calamodendron, no doubt had a smooth or irregularly wrinkled external
bark ; but this gives no good ground for the manner in which some
writers on this subject confound Calamites with Calamodendra, and both
with Asterophyilites and Sphenophyllum. With this no one who haa
studied these plants, rooted in their native soils, and with their appendao^es
still attached, can for a moment sympathize. One of the earliest
geological studies of the writer was a bed of these erect Calamites, which
he showed to Sir C. Lyell in 1844 and described in the Proceedings of
the Geological Society in 1851, illustrating the habit of growth as
actually seen well exposed in a sandstone cliff. Abundant opportunities
of verifying the conclusions formed at that time have since occurred, the
results of which have been summed up in the figures in Acadian Geology,
which though they have been treated by some botanists as merely
restorations, are in reality representations of facts actually observed.

On these subjects, without entering into details, and referring for these
to the elaborate discussions of Schimper, Williamson and M cNab, and to
my paper on the subject. Journal of Geol. Society, vol. 2 7, p. 54, I may
remark : —

1. That the aerial stems of ordinary Calamites had a thin cortical layer,
with lacunss and fibrous bundles and multiporous vessels — the whole not
differing much from the structure of modern Equiseta.

2. Certain arborescent forms, perhaps allied to the true Calamites, as
well as possibly the old underground stems of ordinary species *, assumed
a thick-walled character in which the tissues resembled the wedges of an
exogen, and abundance of pseudo-scalariform fibres were developed, while
the ribbing of the external surface became obsolete or was replaced by a
mere irregular wrinkling.

3. Sufficient discrimination has not been exercised in separating casts
of the internal cavities of Calamites and Calamodendron from those repre-
senting other surfaces and the proper external surface.

4. There is no excuse for attributing to Calamites the foliage of Annu-
laria, Asterophyilites and Sphenophyllum, since these leaves have not
been found attached to true Calamite stems, and since the structure of

• Williamson, Trans. Roy Socy. McNab in Proceedings of Edin. Botanical Society.

8-2 LOWER CARBONIFEROUS PLANTS.

the stems of Asterophyllites as described bj Williamson, aad that of
Sphenophyllum as described bj the ■Tvriter,=»^ are essentially different from
those of Calamites.

5. As the species above described indicates, good external characters
can be found for establishing species of this genus, and these species are
of value as marks of geological age.

LycopodiacecB.
Lepidodendron aculeatum, Sternberg. — (Plate IX, Figs. 75, 75 a, b, c.)

To this species I refer without much doubt the specimens from
McKay's Head collected by Prof. Elder, one of which is represented in
Fig. 75. This species, especially if, as Schimper supposes, identical with
L. undulatam^ ranges both in Europe and America through the whole
extent of the Carboniferous system.

With the flattened stems of this Lepidodendron were found great quan-
tities of the branches, leaves and cones represented in Figs. 77, 78, 80,
and which in all probabiUty belong to this or to the next species — proba-
bly to this; but as the leaves of i. aculeatum are not, I believe, known,
I have'no further evidence of this except their juxtaposition.

Lepedodendrox radiato-plicatum, S. N. — (Plate IX, Figs. 76, 76 a,
b, c.)

Description, — Large stems with very elongate and acuminate rhombic
leafbases having a strong central furrow and the leaf-
scar a little above the centre, and trigonal in form with a
single central vascular scar. The spaces between the leaf-
bases marked with strong wrinkles radiating from the leaf-
bases. Leaf-bases nearly 2 centimetres in extreme length
and 5 millimetres in breadth.
This plant seems quite distinct from anything else I have met with, and

is in very fine preservation, so that the characters can be made out dis-
tinctly. Being quite flattened, the denser tissues of the leaf-bases of the
opposite sides have mutually impressed each other, so as to give a trans-
versely tuberculated or ridged appearance as indicated in Fig. 76. The
leaves more immediately associated with this species are the very elongate
ones represented in Fig. 79.

Lepidodendron selaginoides, Sternberg. — (PI. IX, Figs. 82, 83.)

This characteristic Lower Coal-formation species is found abundantly
in connection with the Coal-beds of St. George's Bay, Newfoundland,

• Journal of Geol. Society, 1866.

n

LOWER CARBONIFEROUS PLANTS. 33

■which I regard as in or near this horizon, and also in the lower part of
the Middle Coal -formation or Upper Millstone Grit of Cape Breton.

Lbpidophloios Acadianus, Dawson. — (Plate IX, Fig. 85.)

A branch of this species occurs in the roof of a small coal at Salmon
River, which must he near the upper limit of the Millstone Grit, and is
the lowest geological position to which I have traced the species.

CoRDAiTES BORASSIFOLIA, Unger. — (Plate VIII, Fig 74.)

This species is very characteristic of the Millstone Grit and Middle
Coal-formation in all parts of Nova Scotia and New Brunswick. I have
represented in Fig. 74 a slab covered with its leaves, from Apple River.

Filices.

The ferns which I have obtained from the Millstone Grit are not very
numerous, and with one exception of species already fully described.

Odontopteris antiqua, Dawson.— (Plate X, Figs. 86, 87.)|Cyclopteris
antiqua, J. G. S. 1866, P. 154, PL XIII, F. 95. Acadian
Geology, P. 481.

Description. — Tripinnate ; petioles slender ; pinnules oblong, obtuse, decur-
ent on the petiole, not contiguous. Terminal pinnules much
elongated ; venation simple, divergent. This plant approaches
more nearly to the peculiar species of Cyclopteris found in the
Devonian, than any of the others I have seen in the Carboni-
ferous.
The original specimens on which this species was founded were found
on loose slabs at R. Hebert and Maccan River ; but I have since found
fragments in situ at the Calvary River, near Riversdale.''' It seems to
come within the technical characters of Odontopteris, but ''^certainly
approaches very nearly to some of the Devonian ferns of the genus Arch(X-
opteris. Its fructification is needed to settle its true place.

Odontopteris ?— (Plate X, Fig. 88.)

A small fragment in Mr. Elder's collection from McKay's Head.

Cardiopteris ?— (Plate X, Fig. 89.)

A few detached pinnules in Shale at Calvary River may indicate a
species of this characteristic Lower Carboniferous genus.

0

34 CANADIAN FOSSILS.

Alethopteris Lonchitica, Brongt (var, heterophylla.) — (Plate X,
Fig. 90.)

The specimen figured was collected by the late Dr. Gesner at Moose
River. The same species occurs in Mr. Elder's collections from McKay's
Head ; and as I have elsewhere shown it extends under different varietal
forms through the whole of the Middle Coal-formation.

Pecopteris abbreviata, Brongt, — (Plate X, Fig. 91.)

The specimen figured is from the Coal above referred to at Salmon
River, and appears to be a part of a frond of this common Coal-formation
species.
Hymenophyllites furcatus, Brongt.— (Plate X, Fig. 92.)

Collected at Calvary River. ^

Sphenopteris 0BTU3IL0BA, Brongt. — (Plate X, Fig, 93.)

Small fragments in Prof Elder's collection from McKay's Head.
Spii;:xopter[S Hoe.vingiiausi, Brongt. — (Plate X, Fig. 94.)

Specimens having the characters of this species, though not well pre-
served, are in Prof. Bell's collections from St. George's Bay, Newfound-
land. This species ranges from the Middle Devonian to the Millstone'Grit,
r.nd is one of the links connecting together the Devonian and Carboniferous
Periods.

Pal^opteris. — (Plate X., Fig. 95.)

A tree fern collected by Mr. Matthew at Gardiner's Creek, and which
I was at first disposed to regard as a stem of Cordaites, but which seems
to be a fern trunk nearly allied to my P. Acadica from the Upper Coal-
formation.

LOWER CARBONIFEROUS PLANTS. 35

IV. PLANTS OF THE MIDDLE AND UPPER COAL-FORMATIOxV.

These have already been illustrated in some detail in my memoir of 1866
already referred to, and in Acadian Geology ; and though much new mate-
rial has accumulated and many species have been inadequately figured, it
may be some time before any full illustration of these plants can be
attempted. For this reason, I give here a list of the species recognised in
these groups of strata.

To avoid repetition of references, I may state that the new species
established by Bunbury will be found in his papers in the Journal of the
Geological Society, Vols. Ill and VIII. Those established by myself will
be found in my paper in the same Journal, Vol. XXII, and my Acadian
Geology, 2nd edition, 1868.

1. Middle Coal-foruation, Nova Scotia and New Brunswick.

Dadoxylon Acadianum, Dawson.

annulatum, Dawson.

Sigillaria (Fav-tilaria) elegans, Brongaiart.
(Fav.) tessellata, Brongn.

(Rh.) Schlotheimiana, Brongn.

(Rh.) Saullu, Brongn.

Brownii, Dawson.

reniformis, Brongn.

laevigata, Brongn.

catenoides, Dawson.

striata, Dawson.

(Clathraria) Menardi, Brongn.

(Asolanus) Sydnensis, Dawsou.

organum, Lindley & Hutton.

elongata, Brongn.

flexuosa, L. & H.

pachyderma, L. & H.

(Fav.) Bretonensis, Dawsou.

eminens, Dawson.

Dournaisii, Brongn.

Knorrii, Bronga.

Stigmaria ficoides, Brongn.
Antholites Rhabdocarpi, Dawson.

■^— • pygmaea, Dawson.
Trigonocarpum Hookeri, Dawson.

Sigillariae, Dawson.

intermedium, Dawson.

avellanum, Dawson.

— — minus, Dawson.

36 CANADIAN FOSSIL&.
rotundum, Dawson.

RhabdocarpuB ?

Calamites cannseformis, Brongn.

ramosus, Artis.

Voltzii, Brongn.

Nova-Scotica, Dawson.

nodosuB, Bchloth.

Equisetites curta, Dawson.
Asterophyllites foliosa, L. & H.

grandis? Sternberg,

— trinervis, Dawson.

Sphenophyllum emarginatum, Brongn.

saxifragifolium, Sternberg.

Schlotheimii, Brongn.

erosum, L. & H.

Pinnularia capillacea, L. & H.

ramosissima, Dawson.

Noeggerathia dispar, Dawson.

flabellata, L. <fe H.

Cyclopteris (Neuropteris) obliqua, Brongn,

(? Neuropteris) ingens, L. & H.

oblata, L. & H.

fimbriata, Lesquereui.

hispida, Dawson.

Neuropteris rarinervis, Bunbury.

perelegans, Dawson.

Voltzii, Brongn.

flexuosa, Sternberg.

Loshii, Brongn.

acutifolia, Brongn.

conjugata, Goeppert.

attenuata, L. & H.

dentata, Lesq.

Soretii (Brongn,)

. cyclopteroides, Dawson.

Odontopteris subcuneata, Bunbury.
Dictyopteris obliqua, Bunbury.
Lonchopteris tenuis, Dawson.
Spbenopteris munda, Dawson.

decipiens, Lesquereui.

— — — gracilis, Brongn.

artemisisefolia, Brongn.

Sphenopteris Canadensis, Dawson.

Lesquereuxii, Newberry.

microloba, Guttbier.

obtusiloba (?), Brongn.

Phyllopteris antiqua, Dawson.
Alethopteris Grandini, Brongn.

. pteroides, Brongn.

■ Serlii, Brongn.

grandis, Dawson.

Pecopteris abbreviata, Brongn,

plumosa, Brongn.

polymorpha, Brongn.

LOWER CARBONIFEROUS PLANTS. 37

acuta, Brongn.

■ iongifolia, Brongn.

taeniopteroides, Bunbury.

cyathea, Brongn.

sequali3, Brongn.

Sillimani (?), Brongn.

villosa, Brongn.

Pecopteries Bucklandi, Brongn.
— — ■ — oreopteroides, Brongn.

decurreng, Lesq.

Pluckenetii, Sternb.

Hymenophyllites pentadactyla, Dawson.

Palaeopteris Hartii, Dawson,

Caulopteris, L. & H.

Psaronius, Cotta.

Megaphyton magnificum, Dawson.

humile, Dawson.

Lepidodendron Pictoense, Dawson.

rimosum, Sternberg.

Sternbergii, Brongn.

decnrtatum, Dawson.

dilatatum, Lindley & Hutton.

binerve, Bunbury.

tumidum, Bunbury.

gracile, Brongn.

elegans, Brongn.

plumarium, L. & H.

selaginoides, Sternb.

Harcourtii ? Witham.

clypeatum (?), Legqx.

aculeatum, Sternberg.

plicatum, Dawson.

—— personatum, Dawson.
Lepidostrobus variabilis, L. & H.

squamosus, Dawson

longifolius, Dawson

sp.

■ trigonolepis, Bunbury.

Lepidophyllum majus (V), Brongn.

intermedium, L. & H.

Lepidophloios Acadianus, Dawson.

prominulus, Dawson.

tetragonus, Dawson.

Diplotegium retusum, Dawson.
Knorria Sellonii, Sternberg.
Cordaites borassifolia, Corda.
Cardiocarpum fluitans, Dawson.

bisectum, Dawson.

•— ^— sp. like C. marginatum.

sp, allied to C. latum, Newberry.

Sporangites papillata, Dawson.
glabra. Dawson.

38 CANADIAN FOSSILS.

2 Upper Coal Formation, Nova Scotia, New Brunswick and Prince Edward IsiANOi
Dadoxylon materiarium, Dawson.
Walchia (Araucarites) gracilis, Dawson.
Walchia (Araucarites) robusta, Dawson.

Sigillaria (Rhytidolepis) scutellata, Brongn. /

Calamodendron approximatum, Brongn.
Antholithes squamosa, Dawson.
Rhabdocarpus insignis, Dawson.
Trigonocarpum Noeggerathii, Brongn.
Calamites Suckovii, Brongn.

Cistii, Brongn.

dubius, Artis.

gigas, Brongt.

Asterophyllites equisetiformis, L. & H.
Annularia sphenophylloides, Zenker.

longifolia, Brongn.

Sphenophyilum longifolium, Germar.
Cyclopteris heterophylla, Goeppert.

oblongifolia, Goeppert.

Neuropteris cordata, Brongn. (and var. angustifolia) .

gigantea, Sternb.

— — heterophylla, Brongn.

auriculata, Brongn.

rarinervis, Bunbury.

Odontopteris Schlotheimii, Brongn.
Sphenopteris hymenophylloides, Brongn.

latior, Dawson.

Alethopteris lonchitica, Sternberg.

nervosa, Brongn.

muricata, Brongn.

Pecopteris arborescens, Schloth.

rigida, Dawson.

unita, Brongn.

oreopteroides (7), Brongn.

Beinertia Goepperti, Dawson.
Palfeopteris Acadica, Dawson.
Lepidodendron undulatum, Sternberg.
Lepidophyllum lanceolatum, L. & H.

trinerve (?), L. & H.

— sp.

Lepidophloios parvus, Dawson.
Cordaites simplex, Dawson.

To this last list it is possible that a few species may be added from my
more recent collections in Pictou County.

Of the above species those which may be regarded as the most charac-
teristic of the Newer Coal-formation, as distinguished from other members
of the Carboniferous system, are perhaps Dadoxylon materiarium, the
species of Walchia, Annularia sphenophylloides, Neuropteris auriculata,
Pecopteris arborescens, and Cordaites simplex. Calamites gigas is found only
in the newest beds of the Upper Coal-formation.

LOWER CARBONIFEROUS PLANTS. 39

v.— NOTE ON THE CHARACTERS OF SIGILLAROID AND LEPI-
DODENDROID TRUNKS.

From the loose manner in which plants of these groups are named and
described by botanists, it seems useful to give some definitions of the parts
employed in their determination.

It may be premised that the modes of determination in Fossil
Botany are necessarily different from those employed in recent Botany.
The Palceobotanist must have recourse to characters derived from the
leaves, the scars left by their fall, and the internal structures of the stem.
These parts, held in Httle esteem by botanists in describing modern
plants, and much neglected by them, must hold the first place in
the regard of the fossil botanist, whereas the fructification, seldom pre-
served, and generally obscure, is of comparatively little service- It is to
be remarked also that in such generalized plants as those of the Palaeozoic,
remarkable rather for the development of the vegetative than of the re-
productive organs, the former rise in importance as compared with their
value in the study of modern plants.

In Sigillarioe, Lepidodendra, kc, the following surfaces of the stem
may be presented to our inspection : —

1. The outer surface of the epidermis without its leaves, but with the
leaf-bases and leaf-scars more or less perfectly preserved. On this sur-
face we may recognize — (1) Cellular swellings or projections of the bark to
which the leaves are attached. These may be called Leaf-bases, and they
are sometimes very prominent. (2) The actual mark of the attachment
of the leaf situated in the most prominent part of the leaf-base. This
is the Leaf-scar. (3) In the leaf-scar when well preserved we can see one
or more minute punctures or prominences which are the points where the
vascular bundles passing to the leaf found exit. These are the Vascular
scars.

When the leaves are attached, the leaf-scars and vascular scars cannot
be seen, but the leaf-bases can be made out. Hence it is important, if
possible, to secure specimens with and without the leaves. In flattened
specimens the leaf-bases are often distorted by pressure and marked with
furrows which must not be mistaken for true structural characters. Th«
leaf-bases, which are in relief on the outer surface of the stem, of course
appear as depressions on the mould in the containing rock, in which the
markings often appear much more distinctly than on the plant itself.

40 CANADIAN FOSSILS.

2. The outer surface of the epidermis may have been removed or may
be destroyed by the coarseness of the containing rock. In this case the
leaf-bases are usually preserved on the surface of the outer or corky bark,
but the leaf-scars and vascular-scars have disappeared. This gives that
condition of Lepidodendroid trees to which the name Knorria has been
applied. When plants are in this state careful inspection may sometimes
discover traces of the leaf-scars on portions of the stem, and thus enable the
Knorria to be connected with the species to which it belongs.

3. The outer or corky bark may be removed, exposing the surface of
the inner or fibrous and cellular bark, which in the plants in question is
usually of great thickness. In this case neither the leaf-bases nor the
scars are seen, but punctures or little furrows or ridges appear where the
vascular bundles entered the inner bark. Specimens in this state are
usually said to be decorticated, though only the outer bark is removed .
It is often difficult to determine plants in this condition, unless some por-
tion of the stem can be found still retaining the bark ; but when care is
taken in collecting, it will not infrequently be found that the true outer
surface can be recovered from the containing rock, especially if a coaly
layer representing the outer bark intervenes between this and the inner
impression. Specimens of this kind taken alone, have been referred to
the genera Knorria, Bothrodendron and Halonia.

4. In some cases, though not frequently, the outer surface of the
ligneous cylinder is preserved. It almost invariably presents a regularly
striated or irregularly wrinkled appearance, depending upon the vertical
woody wedges, or the positions of the medullary rays or vascular bundles.
Specimens of this kind constituted some of the Endogenites of the older
Botanists, and the genus ScMzodendron of Eichwald appears to include
some of them. Many of them have also been incorrectly referred to
Calamites.

5. In some cases the cast of the medullary cylinder or pith may
alone be preserved. This may be nearly smooth or slightly marked by
vertical striae, but more usually presents a transverse striation, and not
infrequently the transverse constrictions and septa characteristic of the
genus Sternbergia. Loose Sternbergiae aflford little means of connecting
them with the species to which they belong, except by the microscopic
examination of the shreds of the ligneous cylinder which often cling to
them. *

These facts being premised, the following general statements may be
made respecting some of the more common Palaeozoic genera, referring,
however, principally to the perfect markings as seen on the epidermis.

• See my paper, Journal of Gaol. Society, Vol. XXVII -

LOWER CARBONIFEROUS PLANTS. 41

iSigillaria. — Leaf-bases hexagonal or elongated, or confluent on a
vertical ridge. Leaf-scars hexagonal or shield-shaped. Vas-
cular scars three, the two lateral larger than the central. This
last character is constant, depending on the fact that the
leaves of Sigillaria have two or more vascular bundles. All
so-called Sigillariae having the central vascular scar largest, or
only one vascular bundle, should be rejected from this genus.
In young branches of branching SigillarijB the leaf-scars
sometimes appear to be spiral, but in the older stems they
form vertical rows ; interrupted, however, by transverse rows
or bands of Fruit-scars, each with a single large central
vascular scar, and which have borne the organs of fructifi-
cation. Artlirocaulis of McCoy is founded on this pecuharity.

Syringodendron. — Difi'ers from Sigillaria in the leaf-scars, which
are circular and with a single vascular bundle. It is a matter
of doubt whether these plants were of higher rank than Sigil-
laria tending toward the pines, or of lower rank tending toward
Cyclostigma. Their leaf-bases form vertical ridges.

Lepidodendron. — Leaf-bases rhombic, oval or lanceolate, moder-
ately prominent. Leaf-scars rhombic or sometimes shield-
shaped or heart-shaped, in the middle or upper part of the
leaf-base. Vascular scars three — the middle one always
largest and corresponding to the single nerve of the leaf; the
lateral ones sometimes obsolete.

In older stems three modes of growth are observed. In
some species the expansion of the bark obliterates the leaf-bases
and causes the leaf-scars to appear separated by wide spaces
of more or less wrinkled bark, which at length becomes longi-
tudinally furrowed and simulates the ribbed character of
Sigillaria. In others the leaf-bases grow in size as the trunk
expands, so that even in large trunks they are contiguous
though much larger than those on the branches. In others
the outer bark, hardening at an early age, is incapable of
either of the above changes, and merely becomes cleft into
deep furrows in the old trunks.

Lepidophloios. — Leaf-bases transverse and prominent— often very
much so. Leaf-scars transversely rhombic or oval with three
vascular scars, the central largest. Leaves very long and
one-nerved. Large strobiles or branchlets borne in two
ranks or spirally on the sides of the stem, and leaving

42 CANADIAN FOSSILS.

large round scars ( Cone-scars) often with radiatingimpression»
of the basal row of scales.

Species with long or drooping leaf-bases have been included
in Lepidophloios and LomatopUoios. Species with short leaf
bases and cone-scars in two rows have been called Ulodendron
and some of them have been included in Sigillaria (sub-genus
Clathraria). Decorticated stems are Bothrodendron and
Halonia. Some of the species approach near to the last genus,
especially to the Lepidodendra with rhombic leaf-bases like
L. tetragonum.

Cycloatigma. — Leaf-bases undeveloped. Leaf-scars circular or
horse-shoe shaped, small, with a central vascular scar. In old
trunks of Cyclostigma the leaf-scars become widely separated,
and sometimes appear in vertical rows. Young branches of
Lepidodendron sometimes have the leaf-scars similar to those
of Cyclostigma.
Leptoplileum. — Leaf-bases flat, rhombic; leaf-scars obsolete; vas-
cular scar single, central. The two last genera are character-
istically Devonian.
In contradistinction fi'om the trees above mentioned, the following
general statements may be made respecting other groups.

In Conifers the leaf-bases are usually elongated vertically, often scaly
in appearance, and with the leaf-scar terminal and round, oval or rhombic,
and with a single well-marked vascular scar.

In Calamites, Calamodendron and Asterophyllites the scars of the
branchlets or leaves are circular or oval with only a single vascular
scar, and situated in verticils at the top of well-marked nodes of the
stem.

In Tree-ferns the leaf-bases are large and usually without a distinct
articulating surface. The vascular- bundles are numerous. Protopteris
has rounded leaf-scars with a large horse-shoe shaped bundle of vessels
above and small bundles below. Caulopteris has large elliptic or oval |leaf
scars with vascular scars disposed concentrically. Palseopteris* of Geinitz
has the leaf-scars transversely oval and the vascular bundles confluent
in a transverse band with an appendage or outlying bundle below. Stem-
matopteris has leaf-scars similar to those of Caulopteris, but the vascular
bundles united into a horse-shoe shaped band.

*This name, preoccupied by GeinitZj has been inadvertently misapplied to the Devonian
ferns of the genus ArchEeopterie.

SIGILLARIA LORWAYANA.— Dn.

Zones of Fruit-scars at (a, a, a). (6) Leaf-scar enlarged.

(c) Fruit-scar enlarged.

LOWER CARBONIFEROL'S PLANTS. 43

VI.— APPENDIX.— ON SOME NEW FORMS OF SIGILLARIA
AND LEPIDODENDRON.

While the above Report vras in progress, some interesting specimens
have been received from Mr. Hill, of the Emery Mine, Cape Breton, and
which throw hght on some obscure forms of the Middle Coal formation pre-
viously known, but not well understood. As these are of some
interest, I give a preliminary notice of them here, though they deserve to
be fully figured and described.

1. — Sigillaria Lorwayana, S.N. *

This has long been known to me by specimens from Sydney in the
collection of Mr. R. Brown, but I did not feel justified in describing it as
new. It is a very beautiful form of that type in which the leaf-bases are
very large and contiguous to each other, and its nearest alHes are my
S. Bretonensis and Schimper's ^S*. Lalayana. Its description is as
follows : —

Leaf-bases about 8 m.m. broad and 5 m.m. high, in trunks of moderate
size, hexagonal with rounded angles, or approaching to oblong, sometimes
a shght indentation below causes them to appear reniform. They are
contiguous or nearly so in vertical rows, being separated from each other
only by a slight ridge. The rows are separated by spaces of wrinkled
bark nearly half as wide as the leaf-bases. Vascular scars near the top of
the leaf-base, each having two minute and often confluent points and two
larger and lunate lateral punctures.

Fruit-scars arranged in transverse rows forming a girdle, each member
of the girdle consisting of 2 to 7 contiguous .vertical scars placed in the
spaces between the leaf-scars in the vicinity of an articulation, where the
rows of leaf-scars are not continuous, as if there had been an interruption
of growth. These articulations are from two inches to a foot apart
vertically. The scars are depressed or sunk into the stem, rounded or
angular by pressure, and having in the centre a small sunken ring and dot.

The bark appears to have been thin. Flattened specimens are sometimes
a foot in diameter.

When the epidermis is removed, the inner surface appears rugose longi-
tudinally,and there are transverse leaf-scars, each with two vascular points,
the whole presenting the appearance of the type Leioderma.
* Xamed from the "Lorway " Coal seam, near which it is found.

44 CANADIAN FOSSILS.

The fruit-scars are evidently modified leaf-scars passing into these.
They have thus no affinity either in form or relation with the large round
cone-bearing-scars o? Lepidojloios, and they must either have borne single
ovules or modified leaves with marginal fruit. In the former case the
shape of the base would indicate that the fruits were of the nature of
Trigonoearpa. In the latter case they were probably of the nature of
Antholites, (which are not, as Carruthers supposes, to be confounded with
Cardiocarpa.) Or they may have been racemes of Trigonoearpa, as I
suggested many years ago.* Still it is possible that some of the plants
included in Sigillaria may have borne Cardiocarpa, and the mode of
association of these in racemes, as shown by Carruthers in one of the
species, is a further evidence of this possibiUty.

In any case it is perfectly plain both from considerations of homology and
from the form and close approximation of the fruit-scars, that they could
not have borne strobiles of the nature of those figured by Goldenburg
and Schimper as fruits of Sigillaria.

Some of the trunks are imbedded in what seem to be masses of flattened
fern stipes, but which might be readily mistaken for leaves, and the stems
of this and the next species are not unfrequently filled or interlaced on the
surface with rootlets of Stigmaria belonging to succeeding generations of
trees.

2. — Clathraria Menardi, Rrongt.

I have elsewhere discussed the probable affinities of those Sigillariae
which have rhombic or elliptical leaf-bases arranged spirally, and which in
this respect resemble Lepidofloios, to which, I have no« doubt, soma of the
species belong ; but which, as I have elsewhere shown, also resemble the
young branches of Sigillariae of the type of S. elegans.-f

In Mr. Hill's collection there are some very instructive specimens of
this type from the Reserve Coal Mine. They have transversely elliptic
and acuminate leaf-bases about 5 m.m. broad, and similar, but for the large
size, to those of S. Menardi of which they may represent a luxuriant
variety. The leaf-scars are small protuberances above the centre of the
base and each with two vascular punctures. The surface under the epi-
dermis is longitudinally striate, and the scars appear as single punctures
arranged in quincunx, and giving the plant at first sight the appearance of
a Stigmaria with very small and distant areoles.

• Acadian Geology p. 437, 438 and 459.
t Acadian Geology p. 435.

LOWER CAKBONIFEROUS PLANTS. 45

The plant evidently branched dichotomously. Its appearance was
that of very young branches of *S'. elegans magnified in dimensions. Old
decorticated trunks might readily be referred to the sub-genus Leioderma,
as presenting merely a vrrinkled surface without ribs or leaf-bases.

In one specimen the exogenous manner of the growth is well seen in the
overlapping of the bark in a ridge at the junction of the branches.

3. — Lepidodendron Sternhergii audits allies.

The descriptions of this species in the books are of such a nature as to
give rise to the suspicion that in addition to the L. longifolium which
Schimperhas separated, other species may be included under the name.

In Mr. Hill's collection, and also in Mr. R. Brown's, and among plants
received some years ago from Mr. H. Poole, all from the Middle Coal-for-
mation of Cape Breton, there are specimens which differ so markedly from
the ordinary L. Sternhergii that I must regard them as a distinct species
or at least varietal form. They may be described thus : —

Leaf-bases in old trunks broadly oval, acuminate at both ends, about 6
m.m. long and 4 m.m. broad, separated from each other by a border of
wrinkled epidermis. Leaf-scars transversely rhombic with three small sub-
equal vascular points below the middle, and a tubercle at the apex.

In young trunks the leaf-bases become shorter and contiguous, and in
this state resemble those of young branches of L. Sternhergii and L.
deeurtatum, though in good specimens distinguishable. The leaves are
narrow and as much as 5 inches long, one-nerved, though appearing two
nerved' near the base.

Certain remarkable leafy strobiles with bracts three or four inches lono-,
Lepidostrobus longifolius, are probably referable to this species or to L.
decurtatum.

In the present species or variety the branches are very spreading, or
given off at very obtuse angles. Hence, if a distinct species, I would name
it L. dispansum.

EXPLANATION OF THE PLATES.

{Lower Carboniferous).
Plate i, Dadoxyloii, &c.

Fig. 1. Dadoxylon antiquius, — Longitudinal section, tangential.

2. Longitudinal section, radial.

3. Transverse section.

4 and 5. Discs and pores.

6. Fragment of wood of Sigillaria, 6 a, 6 b, portions of fibres.
7 and 8. — Lycopodites plumula.

9. portion enlarged.

Plate ii. Lepidodendron corrugatum.
Fig. 10. End of branch with young strobile.
11, 12. Branches with leaves.
13, 14, 15. Branches with leaf-bases, (a) enlarged.

16 to 20. Portions of old stems, showing leaf bases separated by wrinkled bark, (a)
enlarged.

21. Very old stem decorticated.

22. Spore cases, natural size, and (a) enlarged.

Plate hi. Lepidodendron corrugatum.
Fig. 23. Leafy branch, 23 (a), (b), (c). Leaves natural size and enlarged.

24. Decorticated branch — (Knorria).

25. Branch with leaves.

26 to 28. Various decorticated Knorriae, (a) markings enlarged.
29. Bark showing vertical ribs, (a) rib enlarged.

Plate iv. Roots of Lepidodendron, &c.
Fig. 30. Extremity of a flattened branch of Stigmaria, showing areoles and rootlets, 30 a, b,
areole enlarged and rootlet showing mark of vascular bundle.

31. Stigmaria, (a) areoles.

32. Portion of an erect stem of Lepidodendron corrugatum, from near the base.

Plate v. Lepidodendra.
Fig. 33 to 35. Lepidodendron corrugatum, — Variety Cyclostigmoides — young branch and
portions of old stems, (a) scars enlarged.

36. Variety with scars in vertical rows on ribs, 36 a, portion enlarged.

37. Lepidodendron aculeatum, (a) single leaf-base.

38. Lepidodendron, probably corrugatum, showing marking of inner surface of bark.

39. Lepidodendra tetragonum, (a) areole enlarged.

40. Portion of strobile with scales removed, (a) scar enlarged

Plate vi. Lepidodendra, &c.
Fig. 41. Fragment of Strobile, (a) scales detached.

42, 43. Lepidodendron Sternbergii, branches, (a) scar and leaf-base enlarged.
44, 45. Branches showing leaf-bases, (a) leaf-base enlarged

46. Diplotegium, (a) leaf-base enlarged.

47, 48. Stipe of Fern (Rhackiopteris).

49. Cordaites, (a) venation enlarged.

50. Cardiocarpum tenellum.

51. Pinnularia crassa.

CANADIAN FOSSILS. 47

Plate tii. Cycloj^teris Acadica, &c.

52, 53, 54. Stipes of Cyclopteris Acadica; (a) attachment of fructification. 54 a, striation

enlarged.
55 to 60. Pinnules of the same.
61 to 63. Remains of fructification of the same.
64. Hymenophyllites ? (a) fragment enlarged

{Millstone Grit).
Plate viii. Calamites, <|"c.
Fig. 65. Calamites Cistii.

66 to 68. Calamites widulatus, — Portions of stem, showing ribs, nodes and attach-
ments of branches and branchlets ; 68 (a) markings enlarged.
69 to 71. Branches of the same, showing attachments of branchlets,", 70 a, peril en en
larged

72. Branchlet and leaves ; 72 a and b, portions enlarged.

73. Fragment of Rhizoma of the same; 73 a and b, markings enlarged.

74. Slab with Cordaites borassifolia.

Plate ix. Lepidodendra, ^e.
Fig. 75. Lepidodendron aculeatum; 75 a, inner surface of bark, 75 b, c, leaf-bases and leaf-
scars

76. Lepidodendron radiato-plicatum ; 76 a, b, c, leaf-bases, leaf-scar and vascular-scar.

77. Branch (fertile) of Lepidodendron.

78. Fragments of Strobiles; (a) scale, (b) group of scales, (c) axis.
79 to 81. Leaves and cone-scales of the same.

82. Lepidodendron Selaginoides ; (a) leaf-base enlarged.

83. Branches and leaves of the same.

84. Old decorticated stem of the same.

85. Lepidofloios Acadianus, fragment of stem ; (a) scar of cone.

Plate x. Filices.

Fig. 86, 87. Odontopieris antiqua,

88. Odontopteris, sp.

89. Cardiopteris.

90. Alethopteris lonchitica.

91. Pecopteris abbreviata.

92. Hymenophyllites furcatus ; (a) portion enlarged.

93. Sphenopteris obtusiloba.

94. Sphenopteris Hoeninghausi ; (a) portion enlarged.

95. Palxovteris.

lieoiogicai' fciurvey ot Uaiaada.

Flate I.

i-iCUOA C^LiUi Mo'itT^Al

DADOXYLON 5cC- LOWER CARBONIFEROUS.

Geological Survey of Canrxda

Plate II.

Mjw;?.?

!m7

^}f

so

l,p

* /lull.'/* ''"

20 a

4 iV^^'/.f

iHffil'i^ lift,* wm 'imi

LEPIDODENDRON CORRUGATUM._ LOWER CARBONIFEROUS.

Geological Survey of Canada.

Plate Iir

A.L.D. de

I.EGCO (ic^ Litii Montreal.

LEPIDODENDRON COKRUGATUM,_ LOWER CARBONIFEROUS.

Geoloqical Survey of Canada.

Plate IV.

A.r. D del'

'FOCO A c° Lit], Mon'real

GecIcQieal Survey of Canada..

Plate V.

ft hi

U^U

^^-^

^f v?/ ii ^^1 ^ i I

^

^^

^ , ^ r*' -^'^■^

ofe^

LEPIDQIENDRA,- LOWER

-loGC & C9 Steam

CARBOl^IFEROUS.

i^:ntrea:.

Geolcqieal Survey of Canada.

Plate VI.

EGGO &; C? Steam Liu-.. Mciitreal.

LEPIK) DmjAJd-. LQWP,. ^ CARB.Q.RIEE.K O.U.a .

GeoIogiGal Siji-vev of Ganada.

W:.^~

ate VI.

iij5: i. Z'. Sieam Lim. Montreal.

"RJ. CARBQl^IFEEOUS.

Geological Survey of Canada

p>-- Yn

_, INL U Oj

Ci > 1 ' ^'

Geological Survey of Canada

Plate K.

s " 'W I ■' / *^

■ ' '^^ ttMto™ ii^ -ft 1^' -^^ ^

ALU. deU

LEGGO &C°bth Montreal

r[LICES_ l^Px^ho^T 0 NE ' ^u^lF.

a p T T

Geological Survey of Canada

Rate X.

LEGC 5 Jt C° L.th. Vfor.treil.

7ILI CES_ MILLS T 0 lil GRIT

1^1,',=,;*" VOUNG UNIVERSrrv

3 1197 00663 6333

DATE DUE

DEMCO 38-297