DIFFICULTIES AND SOURCES OF ERROR IN THE
DETERMINATION OF FOSSIL PLANTS.
“Robinson Crusoe did not feel bound to conclude, from the single
human footprint which he saw in the sand, that the maker of the
impression had only one leg.”
HUXLEY’S _Hume_, p. 105, 1879.
The student of palaeobotany has perhaps to face more than his due
share of difficulties and fruitful sources of error; but on the other
hand there is the compensating advantage that trustworthy conclusions
arrived at possess a special value. While always on the alert for
rational explanations of obscure phenomena by means of the analogy
supplied by existing causes, and ready to draw from a wide knowledge of
recent botany, in the interpretation of problems furnished by fossil
plants, the palaeobotanist must be constantly alive to the necessity
for cautious statement. That there is the greatest need of moderation
and safe reasoning in dealing with the botanical problems of past ages,
will be apparent to anyone possessing but a superficial acquaintance
with fossil plant literature. The necessity for a botanical and
geological training has already been referred to in a previous chapter.
It would serve no useful purpose, and would occupy no inconsiderable
space, to refer at length to the numerous mistakes which have been
committed by experienced writers on the subject of fossil plants.
Laymen might find in such a list of blunders a mere comedy of errors,
but the palaeobotanist must see in them serious warnings against
dogmatic conclusions or expressions of opinion on imperfect data and
insufficient evidence. The description of a fragment of a handle of
a Wedgwood teapot as a curious form of Calamite[137] and similar
instances of unusual determinations need not detain us as examples of
instructive errors. The late Prof. Williamson has on more than one
occasion expressed himself in no undecided manner as to the futility of
attempting to determine specific forms among fossil plants, without the
aid of internal structure[138]; and even in the case of well-preserved
petrifactions he always refused to commit himself to definite specific
diagnoses. In his remarks in this connection, Williamson no doubt
allowed himself to express a much needed warning in too sweeping
language. It is one of the most serious drawbacks in palaeobotanical
researches that in the majority of cases the specimens of plants are
both fragmentary and without any trace of internal structure. Specimens
in which the anatomical characters have been preserved necessarily
possess far greater value from the botanist’s point of view than
those in which no such petrifaction has occurred. On the other hand,
however, it is perfectly possible with due care to obtain trustworthy
and valuable results from the examination of structureless casts and
impressions. In dealing with the less promising forms of plant fossils,
there is in the first place the danger of trusting to superficial
resemblance. Hundreds of fossil plants have been described under the
names of existing genera on the strength of a supposed agreement
in external form; but such determinations are very frequently not
only valueless but dangerously misleading. Unless the evidence is
of the best, it is a serious mistake to make use of recent generic
designations. If we consider the difficulties which would attend an
attempt to determine the leaves, fragments of stems and other detached
portions of various recent genera, we can better appreciate the greater
probability of error in the case of imperfectly preserved fossil
fragments.
[Sidenote: EXTERNAL RESEMBLANCE.]
The portions of stems represented in figures 20 and 21, exhibit
a fairly close resemblance to one another; in the absence of
microscopical sections or of the reproductive organs it would be
practically impossible to discriminate with any certainty between
fossil specimens of the plants shown in the drawings. Examples such
as these, and many others which might be cited, serve to illustrate
the possibility of confusion not merely between different genera of
the same family, but even between members of different classes or
groups. The long slender branches of the _Polygonum_ represented in
(fig. 21) would naturally be referred to _Equisetum_ in the absence
of the flowers (fig. 20 _B_), or without a careful examination of the
insignificant scaly leaves borne at the nodes. The resemblance between
_Casuarina_ and _Ephedra_ and the British species of _Equisetum_, or
such a tropical form as _E. debile_, speaks for itself.
[Illustration: FIG. 20.
_A._ _Restio tetraphylla_ Labill. (Monocotyledon).
_B._ _Equisetum variegatum_ Schleich. ╮ (Vascular Cryptogam).
_C._ _Equisetum debile_ Roxb. ╯
_D._ _Casuarina stricta_ Dryand. (Dicotyledon).
_E._ _Ephedra distachya_ Linn. (Gymnosperm). (_A_–_E_ ½ nat. size).]
[Illustration: FIG. 21. _Polygonum Equisetiforme_ Sibth. and Sm. _A._
Showing habit of plant. ½ nat. size. The two flowers towards the
apex of one branch, drawn to a larger scale in _B_. _C._ Node with
small leaf and ochrea characteristic of _Polygonaceæ_. From a plant
in the Cambridge Botanic Garden.]
[Illustration: FIG. 22. _Kaulfussia æsculifolia_ Blume. From a specimen
from Java in the British Museum herbarium. ⅓ nat. size.]
Endless examples might be quoted illustrating the absolute futility,
in many cases, of relying on external features even for the purpose
of class distinction. An acquaintance with the general habit and
appearance of only the better known members of a family, frequently
leads to serious mistakes. The specimen shown in fig. 22 is a leaf of a
tropical fern _Kaulfussia_, a genus now living in South-eastern Asia,
and a member of one of the most important and interesting families
of the Filicinæ, the Marrattiaceæ; its form is widely different from
that which one is accustomed to associate with fern fronds. It is
unlikely that the impression of a sterile leaf of _Kaulfussia_ would be
recognised as a portion of a fern plant.
Similarly in another exceedingly important group of plants, the
Cycadaceæ[139], the examples usually met with in botanical gardens
are quite insufficient as standards of comparison when we are dealing
with fossil forms. Familiarity with a few commoner types leads us
to regard them as typical for the whole family. In Mesozoic times
cycadean plants were far more numerous and widely distributed than
at the present time, and to adequately study the numerous fossil
examples we need as thorough an acquaintance as possible with the
comparatively small number of surviving genera and species. The less
common and more isolated species of an existing family may often be
of far greater importance to the palaeobotanist than the common and
more typical forms. This importance of rare and little known types
will be more fully illustrated in the chapters dealing with the
Cycadaceæ and other plant groups. Among Dicotyledons, the Natural Order
Proteaceæ, at present characteristic of South Africa and Australia,
and also represented in South America and the Pacific Islands, is
of considerable interest to the student of fossil Angiosperms. In a
valuable address delivered before the Linnean Society[140] in 1870
Bentham drew attention to the marked ‘protean’ character of the members
of this family. He laid special stress on this particular division of
the Dicotyledons in view of certain far-reaching conclusions, which had
been based on the occurrence in different parts of Europe of fossil
leaves supposed to be those of Proteaceous genera[141]. Speaking of
detached leaves, Bentham says:—“I do not know of a single one which, in
outline or venation, is exclusively characteristic of the order, or of
any one of its genera.” Species of _Grevillea_, _Hakea_ and a few other
genera are more or less familiar in plant houses, but the leaf-forms
illustrated by the commoner members of the family convey no idea of the
enormous variation which is met with not only in the family as a whole,
but in the different species of the same genus. The striking diversity
of leaf within the limits of a single genus will be dealt with more
fully in volume II. under the head of Fossil Dicotyledons.
[Sidenote: VENATION CHARACTERS.]
There is a common source of danger in attempting to carry too far the
venation characters as tests of affinity. The parallel venation of
Monocotyledons is by no means a safe guide to follow in all cases as
a distinguishing feature of this class of plants. In addition to such
leaves as those of the Gymnosperm _Cordaites_ and detached pinnæ of
Cycads, there are certain species of Dicotyledons which correspond in
the character of their venation to Monocotyledonous leaves. _Eryngium
montanum_ Coult., _E. Lassauxi_ Dcne., and other species of this
genus of _Umbelliferæ_ agree closely with such a plant as _Pandanus_
or other Monocotyledons; similarly the long linear leaves of _Richea
dracophylla_, R. Br., one of the Ericaceæ, are identical in form
with many monocotyledonous leaves. Instances might also be quoted of
monocotyledonous leaves, such as species of _Smilax_ and others which
Lindley included in his family of Dictyogens which correspond closely
with some types of Dicotyledons[142]. Venation characters must be used
with care even in determining classes or groups, and with still greater
reserve if relied on as family or generic tests.
It is too frequently the case that while we are conversant with the
most detailed histological structure of a fossil plant stem, its
external form is a matter of conjecture. The conditions which have
favoured the petrifaction of plant tissues have as a rule not been
favourable for the preservation of good casts or impressions of the
external features; and, on the other hand, in the best impressions of
fern fronds or other plants, in which the finest veins are clearly
marked, there is no trace of internal structure. It is, however,
frequently the case that a knowledge of the internal structure of
a particular plant enables us to interpret certain features in a
structureless cast which could not be understood without the help of
histological facts. A particularly interesting example of anatomical
knowledge affording a key to apparently abnormal peculiarities in a
specimen preserved by incrustation, is afforded by the fructification
of the genus _Sphenophyllum_. Some few years ago Williamson described
in detail the structure of a fossil strobilus (_i.e._ cone) from the
Coal-Measures, but owing to the isolated occurrence of the specimens
he was unable to determine the plant to which the strobilus belonged.
On re-examining some strobili of _Sphenophyllum_, preserved by
incrustation, in the light of Williamson’s descriptions, Zeiller was
able to explain certain features in his specimens which had hitherto
been a puzzle, and he demonstrated that Williamson’s cone was that of a
_Sphenophyllum_. Similar examples might be quoted, but enough has been
said to emphasize the importance of dealing as far as possible with
both petrifactions and incrustations. The facts derived from a study of
a plant in one form of preservation may enable us to interpret or to
amplify the data afforded by specimens preserved in another form.
[Sidenote: DECORTICATED STEMS.]
The fact that plants usually occur in detached fragments, and that
they have often been sorted by water, and that portions of the same
plant have been embedded in sediment considerable distances apart,
is a constant source of difficulty. Deciduous leaves, cones, or
angiospermous flowers, and other portions of a plant which become
naturally separated from the parent tree, are met with as detached
specimens, and it is comparatively seldom that we have the necessary
data for reuniting the isolated members. As the result of the partial
decay and separation of portions of the same stem or branch, the wood
and bark may be separately preserved. Darwin[143] describes how the
bark often falls from _Eucalyptus_ trees, and hangs in long shreds,
which swing about in the wind, and give to the woods a desolate and
untidy appearance. In the passage already quoted from the narrative
of the voyage of the Challenger, illustrations are afforded of
the manner in which detached portions of plants are likely to be
preserved in a fossil state. The epidermal layer of a leaf or the
surface tissues of a twig may be detached from the underlying tissues
and separately preserved[144]. It is exceedingly common for a stem
to be partially decorticated before preservation, and the appearance
presented by a cast or impression of the surface of a woody cylinder,
and by the same stem with a part or the whole of its cortex intact
is strikingly different. The late Prof. Balfour[145] draws attention
to this source of error in his text-book of palaeobotany, and gives
figures illustrating the different appearance presented by a branch
of _Araucaria imbricata_ Pav. when seen with its bark intact and more
or less decorticated. Specimens that are now recognised as casts of
stems from which the cortex had been more or less completely removed
before preservation, were originally described under distinct generic
names, such as _Bergeria_, _Knorria_ and others. These are now known
to be imperfect examples of Sigillarian or Lepidodendroid plants.
Grand’Eury[146] quotes the bark of _Lepidodendron Veltheimianum_ Presl.
as a fossil which has been described under twenty-eight specific names,
and placed in several genera.
Since the microscopical examination of fossil plant-anatomy was
rendered possible, a more correct interpretation of decorticated and
incomplete specimens has been considerably facilitated. The examination
of tangential sections taken at different levels in the cortex of such
a plant as _Lepidodendron_ brings out the distribution of thin and
thick-walled tissue. Regularly placed prominences on such a stem as the
_Knorria_ shown in fig. 23 are due to the existence in the original
stem of spirally disposed areas of thin-walled and less resistant
tissue; as decay proceeded, the thinner cells would be the first to
disappear, and depressions would thus be formed in the surrounding
thicker walled and stronger tissue. If the stem became embedded in mud
or sand before the more resistant tissue had time to decay, but after
the removal of the thin-walled cells, the surrounding sediment would
fill up the depressions and finally, after the complete decay of the
stem, the impression on the mould or on the cast, formed by the filling
up of the space left by the stem, would have the form of regularly
disposed projections marking the position of the more delicate tissues.
The specimen represented in the figure is an exceedingly interesting
and well preserved example of a Coal-Measure stem combining in itself
representatives of what were formerly spoken of as distinct genera.
[Illustration: FIG. 23. A dichotomously branched Lepidodendroid stem
(_Knorria mirabilis_ Ren. and Zeill.). After Renault and
Zeiller[147]. (¼ nat. size.) The original specimen is in the
Natural History Museum, Paris.]
The surface of the fossil as seen at _e_ affords a typical example of
the Knorria type of stem; the spirally disposed peg-like projections
are the casts of cavities formed by the decay of the delicate cells
surrounding each leaf-trace bundle on its way through the cortex of
the stem. The surface _g_ exhibits a somewhat different appearance,
owing to the fact that we have the cast of the stem taken at a slightly
different level. The surface of the thick layer of coal at a shows
very clearly the outlines of the leaf-cushions; on the somewhat deeper
surfaces _b_, _c_ and _d_ the leaf-cushions are but faintly indicated,
and the long narrow lines on the coal at _c_ represent the leaf-traces
in the immediate neighbourhood of the leaf-cushions.
[Sidenote: IMPERFECT CASTS.]
It is not uncommon among the older plant-bearing rocks to find a
piece of sandstone or shale of which the surface exhibits a somewhat
irregular reticulate pattern, the long and oval meshes having the
form of slightly raised bosses. The size of such a reticulum may
vary from one in which the pattern is barely visible to the unaided
eye to one with meshes more than an inch in length. The generic
name _Lyginodendron_[148] was proposed several years ago (1843) for
a specimen having such a pattern on its surface, but without any
clue having been found as to the meaning of the elongated raised
areas separated from one another by a narrow groove. At a later date
Williamson investigated the anatomy of some petrified fragments
of a Carboniferous plant which suggested a possible explanation
of the surface features in the structureless specimens. The name
_Lyginodendron_ was applied to this newly discovered plant, of which
one characteristic was found to be the occurrence of a hypodermal band
of strong thick-walled tissue arranged in the form of a network with
the meshes occupied by thin-walled parenchyma. If such a stem were
undergoing gradual decay, the more delicate tissue of the meshes would
be destroyed first and the harder framework left. A cast of such a
partially decayed stem would take the form, therefore, of projecting
areas, corresponding to the hollowed out areas of decayed tissue, and
intervening depressions corresponding to the projecting framework of
the more resistant fibrous tissue. A precisely similar arrangement of
hypodermal strengthening tissue occurs in various Palaeozoic and other
plants, and casts presenting a corresponding appearance cannot be
referred with certainty to one special genus; such casts are of no real
scientific value[149].
The old generic terms _Artisia_ and _Sternbergia_ illustrate another
source of error which can be avoided only by means of a knowledge of
internal structure. The former name was proposed by Sternberg and the
latter by Artis for precisely similar Carboniferous fossils, having the
form of cylindrical bodies marked by numerous transverse annular ridges
and grooves. These fossils are now known to be casts of the large
discoid pith of the genus _Cordaites_, an extinct type of Palaeozoic
Gymnosperms. _Calamites_ and _Tylodendron_ afford other instances of
plants in which the supposed surface characters have been shown to be
those of the pith-cast. The former genus is described at length in
a later chapter, but the latter may be briefly referred to. A cast,
apparently of a stem, from the Permian rocks of Russia was figured in
1870 under the name _Tylodendron_; the surface being characterised by
spirally arranged lozenge-shaped projections, described as leaf-scars.
Specimens were eventually discovered in which the supposed stem was
shown to be a cast of the large pith of a plant possessing secondary
wood very like that of the recent genus _Araucaria_. The projecting
portions, instead of being leaf-cushions, were found to be the casts
of depressions in the inner face of the wood where strands of vascular
tissue bent outwards on their way to the leaves. If a cast is made of
the comparatively large pith of _Araucaria imbricata_ the features of
_Tylodendron_ are fairly closely reproduced[150].
A dried Bracken frond lying on the ground in the Autumn presents a very
different appearance as regards the form of the ultimate segments of
the frond to that of a freshly cut leaf. In the former the edges of
the pinnules are strongly recurved, and their shape is considerably
altered. Immersed in water for some time fern fronds or other leaves
undergo maceration, and the more delicate lamina of the leaf rots away
much more rapidly than the scaffolding of veins. Among fossil fern
fronds differences in the form of the pinnules and in the shape and
extent of the lamina, to which a specific value is assigned, are no
doubt in many cases merely the expression either of differences in the
state of the leaves at the time of fossilisation or of the different
conditions under which they became embedded. Differential decay and
disorganisation of plant tissues are factors of considerable importance
with regard to the fossilisation of plants. As Lindley[151] and later
writers have suggested, the absence or comparative scarcity of certain
forms of plants from a particular fossil flora may in some cases be
due to their rapid decay and non-preservation as fossils; it does not
necessarily mean that such plants were unrepresented in the vegetation
of that period. The decayed rhizomes of the Bracken fern often seen
hanging from the roadside banks on a heath or moorland, and consisting
of flat dark coloured bands of resistant sclerenchyma in a loose sheath
of the hard shrivelled tissue, are in striking contrast to the perfect
stem. A rotting Palm stem is gradually reduced to a loose stringy mass
consisting of vascular strands of which the connecting parenchymatous
tissue has been entirely removed. It must frequently have happened that
detached vascular bundles or strands and plates of hard strengthening
tissue have been preserved as fossils and mistaken for complete
portions of plants.
[Sidenote: MINERAL DEPOSITS SIMULATING PLANTS.]
Apart from the necessity of keeping in view the possible differences
in form due to the state of the plant fragments at the time of
preservation, and the marked contrast between the same species
preserved in different kinds of rock, there are numerous sources of
error which belong to an entirely different category. The so-called
moss-agates and the well-known dendritic markings of black oxide of
manganese, are among the better known instances of purely inorganic
structures simulating plant forms.
An interesting example of this striking similarity between a purely
mineral deposit and the external form of a plant is afforded by some
specimens originally described as impressions of the oldest known fern.
The frontispiece to a well-known work on fossil plants, _Le monde des
plantes avant l’apparition de l’homme_[152], represents a fern-like
fossil on the surface of a piece of Silurian slate. The supposed plant
was named _Eopteris Morierei_ Sap., and it is occasionally referred
to as the oldest land plant in books of comparatively recent date. In
the Museum of the School of Mines, Berlin, there are some specimens
of Angers slate on some of which the cleavage face shows a shallow
longitudinal groove bearing on either side somewhat irregularly
oblong and oval appendages of which the surface is traversed by fine
vein-like markings. A careful examination of the slate reveals the fact
that these apparent fern pinnules are merely films of iron pyrites
deposited from a solution which was introduced along the rachis-like
channel. Many of the extraordinary structures described as plants by
Reinsch[153] in his Memoir on the minute structure of coal have been
shown to be of purely mineral origin.
The innumerable casts of animal-burrows and trails as well as the casts
of egg-cases and various other bodies, which have been described as
fossil algae, must be included among the most fruitful sources of error.
It requires but a short experience of microscopical investigation
of fossil plant structures to discover numerous pitfalls in the
appearance presented by sections of calcareous and siliceous nodules.
The juxtaposition of tissues apparently parts of the same plant, and
the penetration by growing roots of partially decayed plant débris,
serve to mislead an unpractised observer. In sections of the English
‘calcareous nodules’ one very frequently finds the tissue of Stigmarian
appendages occupying every conceivable position, and preserved in
places admirably calculated to lead to false interpretations. The
more minute investigation of tissues is often rendered difficult by
deceptive appearances simulating original structures, but which are
in reality the result of mineralisation. It is no easy matter in
some cases to discover whether a particular cell in a fossil tissue
was originally thick-walled, or whether its sclerous appearance is
due to the deposition of mineral matter on the inside of the thin
cell-membrane. Examples of such sources of error as have been briefly
referred to, and others, will be found in various parts of the
descriptive portions of this book.
[Illustration: FIG. 24. _A._ Section of partially disorganised tissue
attacked by some boring animal. _c_, _c_, coprolites; _d_, a tunnel
made by the borer through the plant tissue.
_B._ Transverse section of a Lepidodendroid leaf, of which the
inner tissues have been destroyed and the cavity filled with
coprolites; simulating a sporangium containing spores. (A and B
from specimens in the Botanical Laboratory collection, Cambridge.)]
[Sidenote: TRACES OF WOOD-BORERS IN PETRIFIED TISSUE.]
There is one other form of pitfall which should be briefly noticed.
In sections of petrified plants one occasionally finds clean cut
canals penetrating a mass of tissue, and differing in their manner of
occurrence and in their somewhat larger size from ordinary secretory
ducts. Such tunnels or canals are probably the work of a wood-boring
animal. An example is illustrated in fig. 24 _A_. Similarly it is not
unusual to meet with groups or nests of spherical or elliptical bodies
lying among plant tissues, and having the appearance of spores. Such
spore-like bodies appear on close examination to be made up of finely
comminuted particles of tissue, and in all probability they are the
coprolites of some xylophagous animal. Examples of such coprolites are
shown in fig. 24 _A_[154], and in fig. 24 _B_ an interesting manner of
occurrence of these misleading bodies is represented. The framework
of cells enclosing the nest of coprolites in fig. 24 _B_, represents
the outer tissues of a Lepidodendroid or a Sigillarian leaf; the inner
tissues have been destroyed and the cavity is now occupied by what may
possibly be the excreta of the wood-eating animal.
Some of the oval spore-like structures met with in plant tissues
may, as Renault has suggested, be the eggs of an Arthropod[155].
In a section of a calcareous Coal-Measure nodule in the Williamson
collection (British Museum)[156] there occur several fungal spores
or possibly oogonia lying among imperfectly preserved Stigmarian
appendages. Associated with these are numerous dark coloured and larger
bodies consisting of a cavity bounded by a simple membrane; the larger
bodies may well be the eggs of some Arthropod or other animal.
[Sidenote: PHOTOGRAPHY AND ILLUSTRATION.]
In looking through the collections of Coal-Measure plants in the
Museums of Berlin, Vienna and other continental towns, one cannot fail
to be struck with the larger size of many of the specimens as compared
with those usually seen in English Museums. The facilities afforded
in the State Collieries of Germany to the scientific investigator may
account in part at least for the better specimens which he is able
to obtain. It would no doubt be a great gain to our collections of
Coal-Measure plants if arrangements could be made in some collieries
for the preservation of the finer specimens met with in the working of
the seams, instead of breaking up the slabs of shale and consigning
everything to the waste heaps. There is one more point which should be
alluded to in connection with possible sources of error, and that is
the essential importance of accuracy in the illustration of specimens,
especially as regard type-specimens. It is often impossible to inspect
the original fossils which have served as types, and it is of the
utmost importance that the published figures should be as faithful
as possible. M. Crépin[157] of Brussels, in an article on the use of
photography in illustrating, has given some examples of the confusion
and mistakes caused by imperfect drawings. It does not require a long
experience of palaeobotanical work to demonstrate the need of care in
the execution of drawings for reproduction.