LIBRARY

UNIVERSITY OF CALIFORNIA

DAVIS

(L

sCr^'-^-'^-^-^.^^

Digitized by the Internet Archive

in 2007 with funding from

IVIicrosoft Corporation

http://www.archive.org/details/biologicallectur07maririch

BIOLOGICAL LECTURES

FROM

THE MARINE BIOLOGICAL LABORATORY
OF WOODS HOLL

1899

BOSTON, U.S.A.

GINN & COMPANY, PUBLISHERS

Cbe ^ttl^enseum preset

1900

LIBRARY
UNIVERSITY OF CAHIFORNM

Copyright, 1900
By GINN & COMPANY

ALL RIGHTS RBSBRVED

CONTENTS.

LECTURE PAGE

I. The Evoltitio7i of the Sporophyte in the Higher

Plajits. Douglas Houghton Campbell . i

II. The Nature of the Evidence exhibited by Fossil
Plants^ and its Bearing upon our Knowledge of
the History of Plant Life. D. P. Penhallow 19

III. Influence of Inversions of Temperature, Ascending

and Descending Curreitts of Air, upon Distri-
bution. Prof. D. T. Macdougal ... 37

IV. Significajice of Mycorrhizas. Prof. D. T. Mac-

dougal 49

V. Institut. Edward Thorndike 57

VI. The Associative Processes in Animals. Edward

Thorndike 69

VII. The Behavior of Unicellular Orgartisms. Her-
bert S. Jennings 93

VIII. The Blind-FisJies. Carl H. Eigenmann . . . 113
IX. Some Governing Factors usually neglected in Bio-
logical Investigatiofts. Alpheus Hyatt . 127
-- X. O71 the Development of Color in Moths and But-
terflies. Alfred Goldsborough Mayer . 157
XL The Physiology of Secretion. A. Mathews . . 165
XII. Regeneration: Old ajid New Interpretations.

T. H. Morgan 185

XIII. Nuclear Divisio7i in Protozoa. Gary N. Calkins 209

XIV. TJie Significance of the Spiral Type of Cleavage

and its Relation to the Process of Differentia-
tion. C. M. Child 231

XV. The Aims of the Quantitative Study of Variation.

C. B. Davenport 267

XVI. On the Nature of the Process of Fertilization.

Jacques Loeb 273

iii

FIRST LECTURE.

THE EVOLUTION OF THE SPOROPHYTE IN THE
HIGHER PLANTS.

DOUGLAS HOUGHTON CAMPBELL.

The questions relating to the origin of organic structures
must always possess great interest for the biologist, and when I
was asked to speak before the students at Woods Holl, it
seemed to me that a discussion of recent views bearing on the
development of the spore-producing structures of the higher
plants, i.e., the Archegoniates and seed-bearing plants, would
not be inappropriate.

There is good reason to suppose that among plants, as among
animals, the most primitive forms were aquatic ; and it is highly
probable that many existing fresh-water algae are but slightly
modified descendants of these ancestral types. This is evinced
by the great uniformity shown by existing green algae all over
the world. Most genera and many species are cosmopolitan,
and they exhibit far less variety than is shown by their larger
and more specialized marine relations. Presumably the condi-
tions in fresh water have changed but little, and as the more
specialized forms have taken to the land, or have developed in
the sea, the few that remained in their primitive environment
have been subjected to much less competition, and have prob-
ably persisted with but little change from very remote times.

Leaving aside certain forms of doubtful affinities, like the
bacteria and blue-green algae, the existing forms which repre-
sent most nearly the ancestors of the higher plants are the
Volvocineae and Protococcaceae. The former are, as all bota-
nists know, free-swimming green organisms which show resem-

2 ' BIOLOGICAL LECTURES.

blances on the one hand to low animals, and on the other are
very like the free-swimming reproductive cells, or zoospores,
of many of the higher algae. The latter probably originated
from forms like the simpler Volvocineae by the loss of motil-
ity associated with the development of a continuous cellulose
membrane about the vegetative cells. This stage in the
evolution of the green algae is represented by some of the
simpler Protococcaceae. Above these unicellular forms are a
number of filamentous plants, single rows of nearly uniform
cells. Such plants are CEdogonium or Conferva, which well
represent the next step in the evolution of the plant body.

The differentiation of the reproductive cells in the algae,
while it offers one of the most interesting and instructive
examples of the evolution of plant structures, must be passed
over here. It may be mentioned, however, that the differentia-
tion of the sexual cells has evidently taken place quite inde-
pendently in several groups of algae.

In the further development of plant types two very important
factors are to be considered. First, the adaptation to a marine
existence, and, second, the exchange of an aquatic for a terres-
trial life.

The conditions in the ocean are markedly different from
those in fresh water, and most plants which have adapted
themselves to life in the sea have become decidedly changed.
The salinity of the water has no doubt been one of the factors
in these changes, but more important is probably the question
of light. The two most characteristic groups of sea plants,
the red and brown seaweeds, are provided with special pig-
ments in addition to the chlorophyll, and there is little ques-
tion that these pigments are developed, in part at least, in
response to changed light conditions.

As the conditions of light and temperature in a marine
environment are far more constant than those in fresh water,
we find, as a rule, that marine algae, especially those inhabiting
the deeper water, are more susceptible to changes of light and
temperature than are most fresh-water forms.

While certain seaweeds growing between tide-marks are
subject' to exposure to the air, it is only for a short time, and

THE EVOLUTION OF THE SPOROPHYTE. 3

we find in most such algae a development of mucilaginous
tissue which prevents their complete desiccation.

These seaweeds have adapted themselves perfectly to their
peculiar environment, and such highly specialized forms as the
great kelps and many red algae probably represent the highest
types of these marine plants. They have diverged widely
from the simpler fresh-water algae, and there is no reason to
suppose that they have given rise to any higher types.

The simple fresh-water green algae, which, so far as we
know, most nearly represent the ancestors of the terrestrial
green plants, differ much in their conditions of life from the
seaweeds. Most bodies of fresh water are subject to great
fluctuations of depth, often drying up completely for long
periods, or sometimes being frozen. It is obvious that plants
living under such conditions must be very resistant, and we
find that such is the case among most green algae. Not only,
as a rule, are they capable of enduring a great range of tem-
perature, but usually at the end of their vegetative period they
produce special cells, "spores," which can endure complete
desiccation without injury, and are also uninjured by freezing.
By means of these " resting-spores " the plant is carried over
from one growing period to the next, and when the conditions
are favorable, the spores germinate and give rise to a new
generation. This production of resting-spores is one of the
most striking differences between these fresh-water algae and
their red and brown relations in the sea, where there is usually
no necessity for such resting-spores.

Certain green algae, like some species of the common genus
Vaucheria, may be considered amphibious, as they do not
actually grow in the water, but are exposed to the air on the
surface of moist earth, from which they absorb the water
necessary for their growth. The ability to thus grow with a
diminished water supply is an evident advantage, and in some
such way as this it is probable that the first strictly terrestrial
plants originated from some originally aquatic algal ancestors.

We can imagine some such forms gradually becoming better
and better able to vegetate on the mud left by the subsidence
of the water, and finally becoming adapted to life on land. The

BIOLOGICAL LECTURES.

lower liverworts, which represent the simplest types of existing
land plants, probably originated in some such way. There are
still existing certain amphibious liverworts, species of Riccia,
which live for the most part as floating aquatics, but sometimes
settle down in the mud left by the subsiding water, or even
creep up the muddy banks and establish themselves as land
plants. This seems to be most commonly done before the devel-
opment of the reproductive organs. It is very likely that the
individual history of these liverworts illustrates the origin of

the typical forms from their
aquatic ancestors.

Among those algae which
approach most nearly to the
lower liverworts there is a
hint of the characteristic al-
ternation of generations so
conspicuous in the mosses
and ferns. The egg-cell in
such algae, on being fertil-
ized, produces a resting-
spore which may be invested
with a protective envelope
of cells. This resting-spore
(oospore) does not, on germination, produce a new plant at
once, but a globular cellular mass is formed (Fig. i, a), each
cell of which gives rise to a motile zoospore, which then grows
into a plant like the parent one.

This alternation of sexual and non-sexual plants becomes,
as is well known, very prominent in the mosses and ferns, or
Archegoniates as they are called on account of the peculiar
female organ, the archegonium.

While there is unmistakable evidence of relationship between
the lower liverworts and the green algae, it must be admitted
that at present the gap between the two groups is a very great
one, and we are not in a position to state positively just where
is the point of contact between Archegoniates and algae.

The structure of the archegonium (Fig. 2) is very constant
throughout the Archegoniates, but has very little in common

Fig. \. — a, Diagram of a section through the germi-
nated resting-spore of an alga (Coleochaete), showing
the mass of cells within. Each cell of the rudi-
mentary sporophyte produces a single zoospore.
b, Longitudinal section of the lower part of the
fertilized archegonium of a liverwort (Riccia),
showing the contained sporophyte, sp., developed
from the egg-cell. Each cell of the sporophyte,
except the outside ones, gives rise later to four
non-motile spores.

THE EVOLUTION OF THE SPOROPHYTE.

with the much simpler oogonium, or female organ of the green
algae, and its origin is still a matter of conjecture.

Aside from the character of the sexual organs, there is no
difficulty in homologizing the sexual plant (gametophyte) of the
lower liverworts and that of certain green algae.

As a result of the change from the aquatic to an aerial envi-
ronment, we find the tissues of the gametophyte in the liver-
worts more resistant to loss of water than is the case in the
algae. The outer cell-walls are cuticularized, and there is
more or less specialization
of the inner tissues, and this
specialization may be very
pronounced, as we find in
such large liverworts as
Marchantia. Finally, in the
leafy liverworts and true
mosses, the gametophyte
develops a definite axis and
special assimilating organs,
leaves.

It is the sporophyte, how-
ever, or non-sexual plant,
developed from the fertilized
^g'g within the archegonium,
that we wish especially to consider. We have already intimated
that in certain algae the oospore, on germination, gives rise to a
globular cell-mass (Fig. i, a), each cell of which produces a zoo-
spore. In the lowest liverworts, e.g., Riccia (Fig. i, /^), a similar
globular cell-mass is produced from the fertilized ^gg, but there
is a much shorter period intervening between fertilization and
the germination of the spore. This cell-mass, while directly
comparable to the corresponding structure in Coleochaete,
differs from it in two important particulars. In the first place,
a single external layer of cells in the sporophyte of Riccia
(Fig. 3, a) remains sterile, and, secondly, each cell of the interior
sporogenous tissues, instead of producing a single zoospore,
gives rise to four non-motile, thick-walled spores, which do not
germinate immediately, but are capable of becoming quite dry

Fig. 2. — a. The archegonium, or female organ, of a
liverwort (Targionia), seen in longitudinal section.
b, A similar section of the archegonium of the cin-
namon fern ; o, the egg-cell.

6 BIOLOGICAL LECTURES.

without losing their vitality, thus answering physiologically to
the single oospore of the algae. In short, the resting period is
here shortened, owing to the power of the gametophyte to grow
on land, and for the protection of the growing sporophyte within
the archegonium.

It must be remembered, however, that for fertilization to be
effected, the gametophyte must be covered with water, or, as it
were, go back to the original aquatic condition, since in all
Archegoniates water is necessary for the opening of the sexual
organs, as well as for conveying the ciliated spermatozoids to
the archegonium.

This amphibious habit of the gametophyte is retained
throughout the whole group of Archegoniates, and is even
found in the lowest seed plants, e.g., Cycas.

As the archegoniate type becomes better developed, we find
a change taking place in the relative importance of gameto-
phyte and sporophyte. While in certain groups like the leafy
liverworts and the true mosses there is a considerable amount
of specialization in the gametophyte, in other types, especially
those which seem to lead up to the ferns, the gametophyte
is much simpler in structure, and it is the sporophyte which
becomes the conspicuous phase. We must remember, how-
ever, that in these forms the sexual organs, archegonium and
antheridium, are essentially the same as in the lower liverworts
and true mosses, and like them require water in order that fer-
tilization may occur.

As every botanist knows, the fern spore on germination pro-
duces, not the fern as we ordinarily understand it, but the
inconspicuous liverwort-like gametophyte, or '< prothallium,"
which not only closely resembles in general appearance a
simple liverwort, but also bears reproductive organs of very
similar structure.

The change in the relative importance of gametophyte and
sporophyte has evidently been a very gradual one and is well
illustrated in the existing Archegoniates.

It is safe to assume that the most primitive Archegoniates
had a sporophyte in which all the cells were sporogenous as
they are in Coleochaete, but no such primitive forms are known

THE EVOLUTION OF THE SPOROPHYTE. 7

to exist at present. The simplest known type of sporophyte
is that of Riccia (Fig. 3, a), already referred to. Here, although
a thin peripheral layer of cells is sterile, much the greater part
of the sporophyte is composed of sporogenous tissue, all of
whose cells produce spores.

There are still existing a number of low liverworts which
very clearly show the mode of progression from the simple
capsule filled with spores, found in Riccia, to the more highly
organized sporophyte of the higher forms. The first step in
the development of the sporophyte is the separation of the

Fig. 3. — Diagram showing the evolution of the sporophyte in the typical liverworts, a, Riccia;
b, Sphaerocarpus ; c, Porella. In Riccia all but the single outer layer of cells produce
spores ; in Sphaerocarpus an absorbent organ or foot is present, and some of the sporoge-
nous tissue remains sterile, forming the sterile cells, j^.; in Porella a stalk is developed
between the foot and the capsule, and the sterile cells within the capsule develop into
elaters, el. ; j/., tetrads of spores.

upper spore-bearing portion from a lower part, which becomes
an organ of absorption, — the foot, — and thus the sporophyte
begins to assume the character of an individual plant with both
vegetative and reproductive tissues. This stage is well shown
by the little liverwort Sphaerocarpus (Fig. 3, b). Here the upper
part of the sporophyte becomes differentiated into a globular
capsule, with a well-developed wall enclosing the sporogenous
tissue, some of whose cells remain undivided and serve to
nourish the growing spores derived from the other sporoge-
nous cells.

In the higher liverworts (Fig. 3, c) the foot is well devel-
oped, and a stalk is usually formed between it and the cap-
sule. In the latter the sterile cells form the curious spirally

8 BIOLOGICAL LECTURES.

thickened elaters, whose principal function is to help in dis-
tributing the spores.

While the last-mentioned type of sporophyte is character-
istic of most liverworts, there is one order — sometimes con-
sidered to be a class, intermediate between the liverworts and
Pteridophytes, or ferns. In these Anthocerotaceae, of which
the genus Anthoceros is the most highly developed, the sporo-
phyte becomes far better developed than in any of the typi-
cal liverworts, although the gametophyte is exceedingly simple.
In these plants the sporophyte (Fig. 4, a) becomes relatively
very large, and is remarkable for the subordination of spore
production to the vegetative life of the sporophyte. The
sporogenous tissue is reduced to a single layer of cells — at
least in the beginning. Outside of this sporogenous layer
are several layers of green cells with large air spaces between
them, which communicate with the outside atmosphere by
means of stomata precisely like those on the leaves of the
higher plants. There is no question that the spongy green
tissue forms a very efficient assimilative apparatus, like the
green tissue of ordinary leaves, so that the sporophyte is
probably quite able to provide its supply of carbon from the
atmosphere. The axial strand of cells, the columella {col)^
corresponds in position to the young vascular bundle in the
fern sporophyte, and possibly may, like it, serve as a means of
conducting water. No root is developed, however, and the
sporophyte must get its supply of water from the gametophyte
by means of the very large foot.

Unlike all other liverworts, the sporophyte in the Anthocero-
taceae develops a zone of growing cells between the foot and
the upper part of the sporophyte ; and these active cells keep
adding new tissue to the base of the sporophyte, which may
thus continue to grow for several months. Were the sporo-
phyte connected with the earth by a root, it would be entirely
able to take care of itself and be quite independent of the
gametophyte.

In tracing the evolution of the sporophyte in the Bryo-
phytes, or mosses, the most marked characteristic is the gradual
reduction of the sporogenous tissue, and its subordination to

THE EVOLUTION OF THE SPOROPHYTE.

the purely vegetative part of the sporophyte, which finally
approaches an independent condition, most evident in such
forms as Anthoceros and the true mosses. In these plants,
by the development of a special assimilative green tissue, com-
municating with the outside atmosphere by means of stomata,
the sporophyte becomes entirely independent, so far as its supply
of carbon is concerned, as it can utilize the CO2 of the atmos-
phere. It remains, however, to some degree dependent upon

Fig. 4. — a. Diagram of a longitudinal section of the sporophyte of Anthoceros, the nearest
approach to the sporophyte of the ferns found among the Bryophytes. The sporogenous
tissue is confined to a single layer surrounding the central strand of tissue, or columella,
col. A large foot is developed, and the outer tissues develop chlorophyll. The tetrads of
spores are separated by sterile cells, st. b. Cross-section of the same, c. Diagram of a
vertical section of a young fern embryo showing its close resemblance to the correspond-
ing stage of the liverwort sporophyte. d, A similar section of an older fern sporophyte
showing the formation of the external organs — stem, s. ; leaf, L. ; root, R. ; and foot, F.
Each of these young organs contains an axial strand of elongated cells, the beginning of
the vascular bundle.

the gametophyte, from which it secures water and some nitrog-
enous food materials, very much as a parasitic plant does from
its host.

While the higher Bryophytes show considerable differentia-
tion in the sporophyte, so far as the tissues are concerned,
there is little or no external differentiation, and nothing com-
parable to the leaf and stem found in the sporophyte of the
higher plants. If, however, in such a form as Anthoceros, we
could imagine the large absorbent organ, the foot, to penetrate

lO BIOLOGICAL LECTURES.

through the gametophyte into the earth, so that the sporophyte
could obtain water and the mineral constituents directly from
the earth, we should then have the sporophyte an entirely
independent organism.

This completely independent sporophyte is first encountered
in the ferns or Pteridophytes — ** vascular cryptogams," as
they are sometimes called. The simple liverwort-like game-
tophyte of the fern, as we have already stated, bears sexual
organs of very much the same structure as those of the liver-
worts, and from the egg-cell within the archegonium is pro-
duced, as the result of fertilization, a sporophyte which in its
early history follows very closely the development of the
sporophyte of a liverwort (Fig. 4, c). As a result of the early
divisions, a nearly globular mass of cells is formed, but very
soon a difference is manifest. Instead of retaining the original
globular form, as in Riccia, or simply forming a cylindrical
body, as in Anthoceros, the young fern sporophyte very early
develops several growing points (Fig. 4, d), which soon cause
it to lose its original form ; and it is evident that each of the
growing points is the beginning of a special organ. One of
these shows itself to be the cotyledon or first leaf, another
becomes the apex of the stem, and a third develops into the
primary root of the young sporophyte. This external differen-
tiation of the sporophyte, which by virtue of the root becomes
a self-supporting plant, at once sharply separates the ferns
from even the highest mosses. This development of special
organs is also associated with far more perfect tissues than are
ever found in the mosses, the most noteworthy of these being
the characteristic vascular bundles.

Of the primary organs, the leaf or cotyledon — at least in
the true ferns — soon assumes a flattened form, and the green
tissue of which it is mainly composed indicates that its prin-
cipal function is that of carbon assimilation.

The root soon breaks through the tissues of the gametophyte
and grows downward into the substratum, thus putting the
young sporophyte into direct communication with the supply of
water. As soon as the sporophyte became thus entirely self-
supporting, a new type of plant was evolved which was destined

THE EVOLUTION OF THE SPOROPHYTE. \\

to become the predominant type of the future. The develop-
ment of a sporophyt-e with definite axis or stem, to which are
attached leaves and roots, is the characteristic of all the higher
plants, or the "vascular plants" as they are sometimes called.

In the ferns, spore production, which in the lower Archego-
niates is the sole function of the sporophyte, becomes entirely
subordinated to the vegetative existence of the sporophyte,
which may reach a very large size, some of the tree ferns hav-
ing stems ten metres or more in height, and living for many
years. Spore formation in these forms does not take place
until several years after the sporophyte is first formed. Finally,
however, the spores are developed from a sporogenous tissue,
exactly as in the liverworts and mosses. The spores are, how-
ever, borne in special organs — sporangia, which in the ordinary
ferns are borne in groups on the back of the leaf. The form
and position of the sporangia are quite constant and form one
of the best means of classifying the Pteridophytes. The first
hint of a segregation of the sporogenous cells is met with in
Anthoceros and its allies, where the continuous layer of spo-
rogenous tissue is imperfectly divided into fertile and sterile por-
tions (Fig. 4, a). This sterilization of a part of the sporogenous
tissue is probably the first step in the 'direction of the special
spore-bearing organs, or sporangia, to which the spore formation
is restricted in the Pteridophytes. The development of simple
sporangia probably occurred very early in the history of the
group.

With, the growing importance of the sporophyte is associated
the development of very perfect tissues of different kinds,
most of which we already find present in the ferns. Of these,
the vascular bundles already referred to are the most charac-
teristic. It is not unlikely that the central strand of elongated
cells, occupying the axis of the sporophyte in Anthoceros and
in most of the true mosses, may be compared to a rudimentary
vascular bundle, but none of the Bryophytes show the charac-
teristic tracheary tissue, the woody part of the bundle in the
higher plants.

In all the plants above the mosses it is the sporophyte which
attracts our attention. We cannot here go into details as to

12 • BIOLOGICAL LECTURES.

the extraordinary development of the vegetative organs of the
sporophyte, but must confine our attention to the evolution of
the spore-bearing organs, or sporangia.

The gradual development of the sporangium is well illus-
trated in the true ferns, or Filicineae, the predominant type,
at present, of the Pteridophytes.

We have already referred to the beginning of a segregation
of the sporogenous tissue in Anthoceros. The nearest ap-
proach to this condition in the Pteridophytes is found in the
peculiar genus Ophioglossum (Fig. 5, a). Here the spore-
bearing part of the leaf has the form of a spike with two rows
of sporangia which are hardly indicated at the surface, but
consist practically of large cavities filled with spores sunk
below the surface of the leaf and opening at the surface by
a transverse slit. It has even been claimed that here, as in
Anthoceros, there is a continuous layer of potential sporoge-
nous tissue which becomes divided into separate sporangia by
the sterilization of portions situated at regular intervals. A
very important difference between the sporangia of Ophioglos-
sum and the sporogenous tissue of Anthoceros is that in the
latter all the spores are discharged through a single opening at
the apex of the sporophyte, while in Ophioglossum each group
of spores has its own opening at the surface of the leaf.

Starting with the very simple type of sporangium found in
Ophioglossum, we find a very perfect series of forms still
existing, which lead up to the type found in the more special-
ized ferns. Thus within the genus Botrychium (Fig. 5, c, d),
which is closely allied to Ophioglossum, there are several
species which illustrate most beautifully almost every interme-
diate form. The simplest ones have a few large sporangia,
more or less sunken in the tissue of the leaf, and closely
resembling those of Ophioglossum ; while the highest mem-
bers of the series have numerous small sporangia which are
borne upon a short stalk or pedicel, and are very much like
those of the ordinary or leptosporangiate ferns, whose sporan-
gia are developed entirely from the epidermal tissue. The
genus Osmunda, which includes the common cinnamon fern of
the eastern United States, is in certain respects intermediate

THE EVOLUTION OF THE SPOROPHYTE. 13

between the typical leptosporangiate ferns and the type found
in Botrychium, altho-agh it must be admitted that the affinities
of Osmunda are perhaps quite as evidently with another
ancient group of ferns, the Marattiaceae, an order of tropical
ferns which are closely related to many extinct carboniferous
ferns.

It is probable, although this cannot be absolutely proved,
that such forms as Ophioglossum and Osmunda, which have
the whole leaf, or leaf segment, covered with sporangia, are

b

BC.

Fig. 5. — a, The upper part of a leaf of an adder-tongue fern (Ophioglossum), showing the spike
of simple sporangia, b. Diagram of a longitudinal section of a part of the spike, showing
the cavities containing the spores, c, Sporangial spike of Botrychium. simplex, with the
few large sessile sporangia, d, Part of the sporangial spike of B. Virginianum, with
numerous small, somewhat stalked sporangia.

more primitive than those in which the sporangia are borne
upon the backs of ordinary leaves.

Wherever the sporangia are borne upon specially modified
leaves (sporophylls) we have structures which may be legiti-
mately compared to the flowers of the seed-bearing or *' flower-
ing" plants, where the flower consists essentially of similar
modified spore-bearing leaves, or sporophylls.

In passing from the lower to the higher Pteridophytes, there
is manifested in several places a remarkable phenomenon
which is of the greatest importance as bearing upon the origin
of the seed-bearing plants. This is known as "heterospory,"
or the production of two sorts of spores.

With the increasing importance of the sporophyte in the
Pteridophytes, there is an accompanying reduction in the

14 BIOLOGICAL LECTURES.

gametophyte. While the latter in the lower ferns is a com-
paratively large liverwort-like plant, sometimes living for sev-
eral years, we find in the higher forms that it becomes much
smaller and lives but a short time, and it is not uncommon to
find that separate male and female plants are produced ; but
there is no apparent difference in the spores from which they
arise.

In the heterosporous Pteridophytes, however, some of the
sporangia produce a small number of spores, sometimes but
one. These ** macrospores " are many times larger than the
much more numerous "microspores" developed in the remain-
ing sporangia. From the large spore is produced the female
gametophyte, from the small one the male. These are both
much reduced, and their whole development may take place
within twenty-four hours. The development of the sporo-
phyte from the fertilized ^gg is in such cases very rapid, and
it soon becomes independent, and all trace of the gametophyte
disappears. In most of these heterosporous forms the gameto-
phyte never becomes self-supporting, depending for its growth
upon the materials already accumulated in the ripe spore. Such
gametophytes develop no chlorophyll, and scarcely exceed in
bulk the spore from which they arise, and within which they
are contained, there often being only sufficient splitting of the
spore coat to expose the reproductive organs.

In the club mosses of the genus Selaginella, heterospory is
carried still further. Before the spores are discharged from the
sporangium they have already undergone the first steps in ger-
mination, and by the time the spore is ripe the gametophyte
has begun to develop within it. In this case the growing
gametophyte does not depend entirely upon the food stored in
the spore, but draws its nourishment to some extent directly
from the sporophyte through the wall of the sporangium, which
thus serves not only to protect the spores but also, to some
extent, as a medium for the transfer of food to the develop-
ing gametophyte. In short, the relation of gametophyte and
sporophyte is the reverse of what obtains in the mosses, where
it is the sporophyte which is permanently dependent on the
gametophyte.

THE EVOLUTION OF THE SPOROPHYTE. 15

It is very certain that heterospory has developed quite inde-
pendently in all the chief groups of Pteridophytes, and there is
evidence from the fossil remains of these plants that several
heterosporous types have become extinct. The multiple origin
of this peculiar phenomenon makes it exceedingly probable that
the formation of seeds, which is simply a further development
of the same phenomenon, has also originated more than once ;
and it is highly improbable that all of the seed-bearing or
flowering plants, to use a common term, have descended from
a common stock. It is much more in accordance with the
known facts to assume a multiple origin for the seed-bearing
plants or Spermatophytes.

These highest of all plants do not possess any structures
which are peculiar to them. The flowers consist of one or
more sporophylls, or sporangial leaves, which are homologous
with those of the ferns. Upon these sporophylls — stamens,
carpels — are borne sporangia which are directly comparable
to those of the heterosporous Pteridophytes. The microspo-
rangia are here known as pollen sacs, and the macrosporangia
as ovules. The pollen sacs and spores agree in their most
minute particulars with the corresponding microsporangia and
spores of the heterosporous Pteridophytes. The spores in both
cases are shed, and the development of the very rudimentary
gametophyte takes place away from the sporophyte. The pol-
len spores always arise in tetrads, like the spores of all Arche-
goniates, and the structure of the ripe pollen spore is not
noticeably different from the spore of a liverwort or moss.

With the macrospores the case is somewhat different. The
macrosporangium in the flowering plants differs in some respects
from that of the heterosporous Pteridophytes. Within it is
borne a single spore (embryo sac) which may be one of four
sister-cells, but is not necessarily so. The macrospore never
becomes free, but is retained permanently within the sporan-
gium, where it germinates, usually while the sporangium is still
attached to the sporophyte, but occasionally after it has become
separated.

Within the macrospore, or embryo sac, is developed a rudi-
mentary female gametophyte, closely resembling that of some

1 6 BIOLOGICAL LECTURES. .

of the heterosporous Pteridophytes, and in the lower forms, like
the coniferous trees, archegonia are produced. The embryo
sporophyte is formed from the fertilized egg-cell, as in the ferns,
and is surrounded by the tissue of the gametophyte, here called
the '' endosperm." During the growth of the embryo the outer
tissues of the sporangium become hard, and finally the sporan-
gium with the enclosed spore and gametophyte falls away as a
seed.

The permanent retention of the macrospore within the spo-
rangium necessitates a somewhat different mode of fertilization,
and we find, therefore, that the pollen spore, instead of devel-
oping an antheridium and motile spermatoids, produces a long
tube into which the sperm-cells pass. The pollen tube grows
through the tissues overlying the apex of the embryo sac, much
as a parasitic fungus penetrates the tissues of the host plant.
When it reaches the archegonium the sperm-cells are dis-
charged, and one of them passes directly into the egg-cell.

One of the most important discoveries of recent years is the
presence of large ciliated spermatozoids in the lowest types of
seed-bearing plants ; ^.^., Zamia, Cycas. In these the pollen
tube becomes much distended with water, and finally bursts,
discharging the large spermatozoids with the fluid into a space
above the archegonium, which the spermatozoids enter just as
they do in the ferns. This discovery removes the last barrier
separating the ferns from the flowering plants, and there can
be no further doubt as to the close relationship of these
groups.

The Gymnosperms — Cycads, Coniferae, etc. — are undoubt-
edly much nearer the ferns than they are to the ordinary flow-
ering plants, or Angiosperms. It is usually taken for granted
that the latter originated from the Gymnosperms, but it is quite
as likely that they constitute an entirely independent develop-
mental line originating directly from the ferns, which has proved
itself better fitted on the whole to modern conditions than are
the simpler and more ancient Gymnosperms, which they have
largely superseded.

Although the sporophyte in some of the Gymnosperms, such
as the giant conifers, is often of gigantic size, it never shows as

THE EVOLUTION OF THE SPOROPHYTE. 17

complex structure as is found in most of the Angiosperms, or
typical flowering plants. The latter represent the highest
plant type, and the sporophyte exhibits almost endless variety,
especially in the flowers. The gametophyte, however, is
reduced to a microscopically small body consisting of a few
cells only.

The type of structure exhibited by the Angiosperms has been
remarkably successful, and the great bulk of existing terrestrial
plants are Angiosperms, which have to a great extent replaced
the ferns and Gymnosperms of earlier geologic times. In the
water, however, where conditions have changed less, the lower
types of plants still persist, and the Angiosperms are much less
important.

Summary.

In passing from the higher algae and lower liverworts to the
ferns and flowering plants, it is the sporophyte, or non-sexual
plant, which continues to increase in importance. Starting as a
minute globular capsule completely filled with spores, there is
first a sterilization of a part of the tissue which becomes devoted
to the vegetative needs of the sporophyte. There is soon
developed a special absorbent organ, the foot, and later a sys-
tem of green tissue for the assimilation of COg. The sporo-
phyte at last, in the ferns, by the development of a root,
becomes entirely self-supporting, and from this time the sporo-
phyte becomes the important phase in the plant's existence, the
gametophyte becoming more and more reduced. The complete
independence of the sporophyte is associated with the develop-
ment of special organs, — and a corresponding perfecting of the
tissues.

The sporogenous tissue, which at first constitutes practically
the whole of the sporophyte, finally becomes restricted to spe-
cial organs, sporangia, which are met with in all the higher
plants. The development of the spores themselves, however,
is extraordinarily uniform throughout, and is one of the strongest
arguments for the common origin of all these forms.

The remarkable development of the sporophyte in the higher
plants has been influenced by many factors, but probably the

1 8 BIOLOGICAL LECTURES.

most potent was the abandonment of the originally aquatic
habit characteristic of the primitive plants. Having assumed
the terrestrial environment, many other causes combined to
produce the marvelous structures which we see associated with
the sporophyte of these higher plants.

SECOND LECTURE.

THE NATURE OF THE EVIDENCE EXHIBITED

BY FOSSIL PLANTS, AND ITS BEARING

UPON OUR KNOWLEDGE OF THE

HISTORY OF PLANT LIFE.

D. P. PENHALLOW.

In directing attention to some of the more recent results in
the study of those strange forms of plant life which flourished
in past ages, and which are now known to us only through
their fossilized and very fragmentary remains, it is my purpose
to discuss some of the fundamental aspects of the question and
endeavor to indicate the direction along which palaeobotanists
are working in their efforts to discover those facts which at
present constitute the missing links in a phylogenetic chain, as
we know it through a study of existing species; and in so doing
I shall present a few typical cases in illustration of the nature
of the evidence it is possible to secure from organisms long
since extinct, and the remains of which have not only been
broken up and the parts widely scattered, but which, during
their entombment in the crust of the earth for enormous
periods of time, have been subjected to the destructive influ-
ences of decay, usually supplemented by the infiltration of
mineral matter, and those multiform alterations which follow
from heat and pressure, consequent upon the great and varied
changes in the material surrounding them.

Until recently the study of fossil plants has possessed few
attractions for the rising botanist of the new school, very prob-
ably because he saw little opportunity for the development of
the philosophical faculty and no great probability of discoveries

19 .

20 BIOLOGICAL LECTURES.

of a higher order than the possible identification of a new
species. The progress of botanical science in other directions,
however, notably in our knowledge of anatomy and a more crit-
ical appreciation of the significance which attaches to the devel-
opment of tissues and the growth of organs, has brought about
a radical change in the point of view from which we to-day
regard fossil plants. The very fact that we now apply to such
remains, so far as their nature admits, the same exact and
searching methods of study which are applied to living forms,
and that in them we recognize the possibility of discovering the
key to some of the most important problems in development
which confront us at the present time, imparts to this branch
of botanical work the same zest that is so marked a feature in
other departments of the science, while the many and often
great difficulties which surround the elucidation of any given
problem seem but to strengthen the energy and keenness with
which the pursuit is maintained. It is, therefore, a field which
offers abundant attractions to an enthusiastic and diligent stu-
dent; but for successful results it demands a broad foundation
in the science as a whole, and especially an intimate knowledge
of anatomy, both gross and minute.

The difficulties which surround the study of fossil plants are
so unlike those to be met with in a study of existing species,
that a brief consideration of their nature and bearings may
serve to give a clearer conception of some of the facts which
I shall later bring forward. These difficulties may be grouped
under two principal heads :

1. The nature and position of the material.

2. The character of the alterations it has undergone.

I. The Nature and Position of the Material.

It seldom happens that fossil plants are found in situ. It is
quite true that, in the overlying slates of coal beds, ferns and
other characteristic plants are commonly met with, and are thus
to be observed in the places occupied during growth. We also
have similar examples in the erect Sigillarias of the South
Joggins, Nova Scotia, or the more recent submerged forests

THE HISTORY OF PLANT LIFE. 2 1

to be met with at Leasowe and other places about the coast
of Great Britain — examples of plants which have been preserved
in place, and from which, therefore, we may draw important
inferences respecting their former environment and habits of
growth, but such examples are comparatively rare. The great
bulk of the material with which the palaeobotanist has to deal
has been displaced, chiefly through the action of water, as exem-
plified at the present day by such great rivers as the Amazon,
Mississippi, or Nile. The evidence of such transport is gen-
erally conspicuous and admits of little if any doubt. Thus, in
the woods so abundant in certain Pleistocene deposits, the usu-
ally small fragments show in their worn surfaces and rounded
angles all the characteristic features which can come only
through long immersion in water and the prolonged action
of waves, or transport over considerable distances. In the
banks of the Moose and Missinaibie rivers, which run from
near the north of Lake Superior to James Bay, extensive
deposits of moss, lying under a depth of fifty feet of soil, and
consisting chiefly of Hypnum and Distichium, occur. These
deposits are characterized by the compressed, flaky form of the
material, — the flakes being separated by silt consisting of sand,
— and commonly enclosing fragments of wood which have been
worn by the prolonged action of water. The evidence here
indicates that the material must have had its origin near the
head waters of the rivers, that it was carried down to a point
near the mouth, and there deposited in pockets, as now found.
As, at the present day, vast quantities of plant remains are
washed down the tributary rivers into lakes or into the ocean
itself, so must we consider that the same process has been oper-
ative during all the ages through which the earth's crust has
been undergoing continual change. Few examples offer more
instructive lessons relative to the general conditions under which
plant remains become buried by successive layers of sand and
mud, and eventually converted into fossilized forms, than those
which occur upon almost any ocean beach. Here we note in
one enormous mass the commingling of land and marine plants
which, in conformity to conditions of topography, wind, and
currents, are commonly localized in a particular place, and there,

22 BIOLOGICAL LECTURES.

through the repeated and long-continued action of the waves,
are beaten into fragments mingled in the most inextricable con-
fusion, until eventually covered by a deposit of sand or clay.
Through all the long period of this process, decay has wrought
destruction in the more perishable material, until that which
finally passes into the fossil state represents a mere fragment
only of the original material. The significance of such evidence
becomes apparent when it is desirable to employ such remains
as a test of climate, since the conclusions drawn must obviously
depend upon whether the material is in situ or not, and if not,
its original source.

Fossil plants are very rarely, if at all, found in anything like
a complete condition. This naturally results from the destruc-
tive action of the water which serves to transport them, as well
as from the unequal effects of decay and the operation of pres-
sure and heat to which they are so often subjected. From this
it follows that the mere association of parts, when organic union
is wanting, is altogether untrustworthy as evidence of common
origin or specific identity ; and it most commonly happens that
the restoration of a species can be accomplished, if at all, only
after an exhaustive study of numerous fragments from different
sources. Few such restorations are satisfactory ; most of them
are purely hypothetical with respect to many of their most
important features, and they thus serve no final purpose; they
are merely the expression of what we commonly term "work-
ing hypotheses." Large numbers of plants are known by their
leaves only. This is particularly true of the Tertiary, as also
of the Cretaceous flora, where, however, the fruit and sometimes
the silicified or calcified stems also aid in identification. In
the Devonian and Silurian they may be known by leaves, but
more commonly by fragments of branching stems, by fruits,
or by exceptionally well preserved stem structure. In recent
formations like the Pleistocene, the stem structure may be so
perfectly preserved as to admit of direct and exact comparison
with existing species, and thus the genus, and often the species
itself, may be determined beyond all reasonable doubt. It will
be observed, however, that in the preservation of those parts,
upon which the determination of relationship so largely depends.

THE HISTORY OF PLANT LIFE. 23

viz. J the more delicate structures of the gametophyte, we
encounter one of the most serious difficulties in the prosecution
of such studies. In the case of delicate plants or the delicate
parts of woody plants, complete disappearance of structure
commonly follows, and if any evidence of organic structure
remains it generally appears in the form of highly altered
carbon, as coal or graphite, or else, as usual in the case of
leaves, only a mere impression of the external features is pre-
served in the surrounding material. It is altogether exceptional
that, as in Parka decipiens, we meet with structures of so
delicate an organization as the prothalli, and when such evidence
is obtainable it acquires an unusual value.

2. Character of the Alterations.

Plant remains are rarely found in their original condition
except in the most recent deposits, such as the Pleistocene.
A very superficial examination of the floor of any well-grown
forest, or of a forming peat bog, will serve to disclose the
general conditions under which plants pass into the fossil state-
Here we note that as the various forms of vegetation attain
maturity and fall to the ground, they at once pass into a state
of decay. If it then becomes possible for this decaying
material to be covered by a fresh deposit of silt, secondary,
effects follow ; the decay is either arrested or its character is
altered, and very possibly the structure becomes infiltrated
slowly with mineral matter which is held in solution by the
surrounding water. That such conditions have surrounded
the fossil plant is evident from the form in which it is
found.

Decay, more or less pronounced, is an invariable antecedent
of the final preservation of plant remains. In many cases it
has been carried to an extremity, so that the structure is
unrecognizable, and then there remains nothing but a residue
of carbon, as so commonly occurs in the Eozoic formation. In
other cases it may be so limited as to permit a recognition of
all the original structural features preserved with great perfec-
tion. The fungi of decay are commonly met with, and are not

24 BIOLOGICAL LECTURES.

only represented by their mycelia, but, as shown by Renault
and others, even the bacteria are now regarded as well-known
forms among carboniferous plants.

Plants which are enclosed in a deposit of compact clay while
yet living, or before any essential decay has taken place, often
exhibit the most perfect preservation, since they have been
hermetically sealed, and thus have retained all their structural
features in an unaltered condition for an unknown number of
thousands of years. Notable examples of this occur in the
various interglacial deposits. In the Leda clays at Montreal,
when fresh cuttings are being made, one may see leaves of the
common Vallisneria spiralis as perfect in all their structural
features as if freshly gathered from the adjacent river. But
the delicate structure has been so far influenced by its long
enclosure in the moist clay that desiccation leaves nothing but
a mere impression of the former organ. In the same formation
near Ottawa, fragments of perfectly preserved wood are enclosed
in nodules of clay, while at Toronto, from the same formation
again, we obtain the Osage orange with all its structure perfectly
preserved, and red cedar, which is not only perfect as to struc-
ture, but which possesses the characteristic shreddy bark, red
color, and distinctive odor. Such woods are sectioned in the
microtome precisely as if taken from a living plant.

While woods from the Pleistocene are often preserved as just
described, it more commonly happens in older formations that
the mass has been wholly or in part carbonized, or that it has
become so charged with mineral matter that it has passed into
a petrified state. These two forms of alteration commonly
accompany one another.

(a) Carbonization. — Plant remains which are covered shortly
after decay arises, so that the latter continues under exclusion
of air, become converted into a mass of carbon which, accord-
ing to circumstances, retains the original structural features
more or less perfectly. This carbonization results from with-
drawal of the elements of water under the peculiar conditions
established, and if pressure and heat are subsequently brought
into operation, there results a compact coal, and possibly cer-
tain oily and gaseous products, as may be noted in the case

THE HISTORY OF PLANT LIFE. 25

of the Ohio shales, which are remarkable in some localities for
the vast number of spores of Protosalvinia, associated with oil,
and in such a state of preservation that they may be recognized
very readily. The extreme form of such alteration is to be
found in graphite, while the various forms of anthracite, bitu-
minous coal, lignite and peat indicate in inverse order the
successive stages through which the plant remains pass. It
follows from these considerations that, as a rule, we can expect
little structure to be exhibited in coaly masses, but we may
draw correct conclusions as to the general character of the
parts represented by taking into consideration the position in
the specimen and the original nature of the various tissues.
Thus we know that unmodified cellulose, such as constitutes the
soft parts of plants generally, contains carbon 44.55^, hydro-
gen 6.14^, and oxygen 49.51^. Lignin, or the essential basis
of all woody structures, contains C. 62.25^, H. 5.93^, O.
36.82^, while cutin and suberin, the basis of cuticle, the walls
of spores and of cork, contain C. 72>-7A% H. 10^, O. 16-17^.
A careful consideration of these figures leads to most important
and interesting conclusions, since the durability of parts, or
their ability to resist decay, as also the readiness with which
they pass into the form of carbonized remains, is in direct pro-
portion to the relative excess of carbon in the original structure.
Hence it is not difficult to understand why the more perishable
fundamental tissues of plants are so rarely preserved while the
woody and cortical parts are well preserved, and why the latter
particularly may remain as a shell of coal when all other parts
have been removed by decay or replaced by mineral matter.
In the light of these facts, also, the remarkable preservation of
many spores may be readily understood.

As carbonization proceeds, the material may become satu-
rated with water holding in solution small quantities of carbonate
of lime or of silica, in which case the entire structure becomes
gradually converted into a mass of calcite or of silica, as the case
may be, the displacement of the organic matter proceeding at
so gradual a rate that all the structural features are retained.
The structure is then represented by fine particles of carbon
disposed along the original lines of structure, while the

26 BIOLOGICAL LECTURES.

various cavities are filled with the calcite, silica, or other
infiltrated material.^

Plants from the early Palaeozoic are often most beautifully-
preserved in this way, but cases occur in which the recrystalli-
zation of the infiltrated material brings about a redistribution
of the carbon particles in such a way as to produce a false
structure, as found most notably in the Celluloxylon primaevum
of Dawson, which was originally supposed to represent a purely
cellular plant, the component cells of which were of gigantic
size. More recent studies of this material, and comparison
with other well-determined plants in various conditions of
alteration and petrifaction, have afforded ample proof that the
apparent structure of Celluloxylon is nothing more nor less than
a condition incident to petrifaction, in which the carbon of the
original structure has been redistributed through the influence
of crystallization and deposited upon the surfaces of, or other-
wise between, the adjacent crystals.

Two full centuries have passed since the first observations
upon the occurrence of fossil plants were made. Yet palaeo-
botany has only recently attained to a position commensurate
with its importance as a branch of botanical science from which
most important and necessary data respecting the gradual evo-
lution of the higher plants may be gained. During the seven-
teenth century, when men were still experiencing the influence
of the Middle Ages, the discovery of plant and animal remains
in the crust of the earth was well calculated to call forth in all
seriousness the most remarkable explanations of their occur-
rence. And among other things we are thus told that such
evidence gave proof " that the whole terrestrial globe was
taken all to pieces and dissolved at the Deluge, the particles

1 A most instructive example of the extent to which a given material may be
involved in the process of petrifaction is afforded by Osmundites skidegatensis
Penh, from the Cretaceous of Vancouver Island. An analysis shows it to contain

Calcite 70.46%

Silica 12.15%

Combustible matter .... 17.36%

iDO.00%

The last constituent which disappears upon ignition, leaving a calcined mass,
probably represents the amount of carbon residue derived from the original plant.

THE HISTORY OF PLANT LIFE. 27

>
of stone, marble, and all solid fossils discovered taken up into

the water and there sustained together with seashells and

other animal and vegetable bodies ; and that the present earth

consists and was formed out of sand, earth, shells, and the rest

falling down again and subsiding from the water."

We are also informed that ** The Deluge came forth at the
end of May when nuts are not ripe," because of the occurrence
of imperfectly formed hazel nuts in certain moss beds. Even
half a century later the idea that such remains were proof of a
deluge had in no way lost its force, since da Costa, who first
pointed out that Sigillarias and Stigmarias represented un-
known forms of life, and was, therefore, the first to indicate
the extinction of former types, firmly believed cones to be of
vegetable origin buried in the strata of the earth at the time
of the universal deluge recorded by Moses.

In these views we, no doubt, have an expression of the sur-
vival of primitive beliefs which are still current among certain
aboriginal people of eastern Asia.

And so for fully a century and a quarter the remains of
plants buried in the crust of the earth for millions of years
were matters of speculation without any adequate conception
of their real significance. Since the time of Witham, Sprengel,
and Goeppert there has been a constantly increasing interest in
the scientific study of plant remains and a correspondingly
greater appreciation of their true bearing upon the history of
plant life. Of the pioneers in this work, among English-speak-
ing people, probably no one has done more than Williamson in
England and Sir William Dawson in Canada to emphasize the
primary importance of the internal structure as a true guide to
relationship.

Our present knowledge of living forms leads us to the con-
clusion that there has been a more or less regular succession
of types from the most simple to the most complex ; and this
has been brought about, not by acts of special creation, but
by the gradual evolution or unfolding of continually higher
types in direct response to changed conditions of environment.
What those conditions were we are unable to say, but from our
present knowledge of plants in respect to their environment

28 BIOLOGICAL LECTURES.

we are permitted to draw certain inferences as to the general
nature of the controlling forces. That such changes may, in
particular cases, be brought about quickly, so that the results
become more or less apparent within the space of an average
life, and thus form the basis of broader generalizations, is well
known ; but in the main they have progressed but slowly, and
enormous intervals of time have been required for the transi-
tion from one type to another. We are not permitted to con-
ceive that this line of descent takes the form of a strictly lineal
succession. On the other hand, there is strong evidence in
support of the view that our mental figure must be that of a
deliquescent tree in which the main line of descent, or stem,
is dissolved into a series of great limbs or primary divisions.
These again branch repeatedly, until in the final division all
direct connection with the original stem is lost. As certain of
these lesser branches retain a full measure of vigor and persist
in their development to the very end, so others attain their
highest development at a comparatively early stage, enter upon
a period of decline, and shortly disappear, thereby introducing
a further important disturbance of the natural arrangement
whereby it becomes increasingly difficult to determine the pre-
cise order of development and relationship between any two
members. And so with plants as a whole. Certain side lines
of descent have not yet attained their full development ; others
are now far along in their period of decline, which may have
had its origin in very remote geological times ; while yet others
have long since disappeared altogether, leaving no visible sign
of their former presence. And thus the symmetry of the bio-
logical tree is disturbed to such an extent and in such a way
that the remaining members appear to have little in common,
and sometimes even stand forth as isolated groups which seem
to justify the ancient idea of special creation. Fortunately the
situation is saved, in the first instance, by an appreciation of
the fact that, within certain limits, each individual reproduces
in the course of its life history the history of the group.
From this it follows that, however divergent species or groups
may appear in their fully matured state, the relationship may be
ascertained through their embryological phases and through

THE HISTORY OF PLANT LIFE. 29

their most elementary members ; and thus it becomes possible
to ascertain at what points suppression has occurred and the
normal succession been broken.

We are perhaps not far wrong in the assumption that the
general line of descent began with the green algae as the first
clearly defined type. From these aquatic forms, which were,
no doubt, at first dominant, if not the exclusive forms, am-
phibious types appeared, leading eventually to terrestrial forms,
as represented by the mosses and liverworts, plants which
clearly show their derivation from an aquatic ancestry when
they enter upon their reproductive phase, and especially in the
development of an algoid protonema. But here we encounter
one of the so-called missing links, since, although the approach
of the alga to the moss, and of the moss to the alga, is well
defined, the intermediate stages are as yet unknown. And so
again the thallus of the liverwort reappears in the prothallus
of the fern, and as these latter lead on to higher types, we find
undoubted evidence of relationship, the exact bond of which is
as yet wanting. One of the most significant facts, however,
is to be found in the evidence of an aquatic ancestry, which
extends through all grades of development in terrestrial plants
until the Cycads are reached ; I refer to the occurrence of
motile spermatozoids, the significance of which can scarcely be
misunderstood.

It is highly probable that further light respecting these ob-
scure problems will be gained as our knowledge of living forms
advances ; but as many of the existing gaps in the biological
tree have resulted from suppressions which must have occurred
in remote periods of the earth's history, it is peculiarly within
the province of palaeobotany to discover the necessary data, and
thus supply the missing links in the chain of plant life, in order
that we may have a clear and complete explanation of the rela-
tions of the various great groups of plants. And it is among
those forms which are now extinct that our search must be
prosecuted with the greatest hope of success. That this is
within the limits of possibility has been abundantly shown by
the investigations of recent years, and possibly no better illus-
tration could be had than the evidence derived from extinct

30 BIOLOGICAL LECTURES.

forms as presented by Dr. Jeffrey in his recent admirable
memoir on " The Development, Structure, and Affinities of the
Genus Equisetum."

Among the most interesting and promising of all botanical
problems is the theory relative to the origin of the sporophyte
as propounded by Bower a few years since, and more recently
emphasized by Campbell, who has himself done so much to
advance our knowledge in this direction. While the evidence
to be derived from living plants points with great force to the
possible correctness of Bower's hypothesis, it is not yet so
complete as to justify us in regarding the law as fully estab-
lished. But it is in the solution of problems of precisely this
nature that palaeobotany would prove of the highest importance,
and it is probably not too much to expect that the study of fos-
sil plants, particularly of types now extinct, may eventually
enlarge our views upon this as upon other important problems.

If we now turn to plants as we find them in the rocks, it is
to be observed that the succession displayed by living forms is
there essentially repeated and thereby confirmed. A glance at
the geological succession of the earth's crust shows that there
are four great periods in which geological time may be reck-
oned, and that these periods conform in the main to epochs in
the development of plant life. In the earliest or Eozoic time,
-chiefly represented in northeastern America by the great
Laurentian formation which gives the dominant physical as-
pect to the northern watershed of the St. Lawrence River
throughout the greater portion of its length, there are few and
trustworthy evidences of former plant life to be obtained ;
that is to say, we find in those rocks no well-defined plant
remains. On the other hand, the occurrence in the Palaeozoic
of plants of a somewhat high degree of organization leads to
the inference that they represent a line of descent which must
have had its origin very early in Eozoic time. But if definite
remains are wanting, there is nevertheless evidence in the
abundance of graphite which occurs in the Laurentian forma-
tion, commonly interstratified with gneiss and attaining a ver-
tical depth upwards of six hundred feet or more, of former
vegetation ; for, as Prestwich very correctly observes, there is

THE HISTORY OF PLANT LIFE. 31

good reason for the belief that the various forms of carbon now
found in the crust of the earth must have been derived in the
first instance from the atmosphere through the agency of green
plants. Sir William Dawson has also expressed the opinion that
if the Laurentian graphite may be taken as representing former
plant life, it indicates its occurrence in vast profusion, and this
was no doubt the case. To these views we may also add the
opinion of the late Sterry Hunt, that the great Laurentian
beds of iron ore had their origin in plant decay, precisely as
they are formed at the present day.

During those early times the water resting upon the earth's
surface was of a rather high temperature. At the same time
the atmosphere was charged with a relatively high percentage
of carbon dioxide, in consequence of which its temperature was
considerably higher than at present. These considerations
lead to the conclusion that the earlier and aquatic vegetation
— of the general character of the Chlorophyceae with which we
are familiar to-day, but whose delicate structure did not admit
of permanent preservation — flourished under conditions now
fairly represented by hot springs, where vegetation thrives at
temperatures upwards of 93° C. If, furthermore, plants began
to emerge from their aquatic condition in later Eozoic time,
they must have been brought under an environment essentially
similar to that of the tropics of to-day. Certainly this was the
case until late in Palaeozoic time.

We are thus led to the belief that the flora of Eozoic time
was not only very abundant, but that it consisted, perhaps
wholly, of aquatic thallophytes comparable with the marine and
fresh-water algae of to-day ; that those plants were capable of
photosynthesis, and that along certain lines they attained a
somewhat high degree of development.

One cannot fail to remark upon the peculiar absence of plant
life during Huronian and Cambrian times, while evidences of
animal life are abundant. This may have its explanation in
the complete destruction of the more perishable plant struc-
tures, consequent upon the great disturbances which character-
ized the close of the Eozoic and the opening of the Palaeozoic
period. That plant life must have been maintained in full

32

BIOLOGICAL LECTURES.

vigor during that time is evident from the nature of the
remains which appear in later formations, and the first well-
defined occurrence of terrestrial forms in the Siluro-Cambrian
or Ordovician would go far to show that plants passed through
their amphibious stage at least as early as the Huronian, thus
preceding the corresponding phase in the progress of animal
life by a very long period of time, since in the latter case this
change was not effected until the later Carboniferous and
Permian. That the Eozoic and early Palaeozoic must have
witnessed the development of high types of plants is abun-
dantly evident from the abrupt occurrence, in the Upper Silu-
rian and Lower Devonian, of gigantic marine algae, of which
there are no equally large representatives in later formations ;
while in the Devonian, also, there occur for the first time plants
which are comparable in their reproductive structures with the
more modern Pilularia, and a brief consideration of the leading
features of some of these plants may assist us in gaining a more
complete appreciation of the statements already made.

In his Old Red Sandstone^ Hugh Miller described certain
peculiar discoid bodies found in the Devonian rocks of Scotland
at Blairgowrie, Myreton, and other localities. In 1831 these
objects were described by Dr. Fleming under the name of
Parka decipiens, and both by him and subsequent observers,
during a period of sixty years, they were regarded as represent-
ing the spawn of JVIollusca, the eggs of frogs or other animals,
but always without suspicion that they might have been derived
from plants ; and it was reserved for Dawson and Penhallow, in
1 89 1, to prove clearly their vegetable nature. These bodies
consist of carbonized discs, or their impressions, about 5-6 mm.
in diameter and grouped in oval masses varying in size from
3.5x5.3 cm. to 13x20 mm. In their more complete forms
these masses show the presence of an external covering,
and there is also some indication that a stalk may have been
present. No stems or leaves can with certainty be considered
to belong to these bodies, since as yet no organic union between
such organs has been observed, but there is reason to believe,
from the associated structures, that they belonged to plants with
a creeping stem and upright leaves, while the recent finding

THE HISTORY OF PLANT LIFE.

ZZ

of leaf-like bodies similar to those of Marsilia seems to suggest
the possibility of foliage of that type. Our present knowledge,
however, centers in the carbonized discs, from which important
data have been obtained. Boiling them out in nitric acid they
yield two kinds of spores. The macrospores measure 40 /-t in
diameter and are therefore slightly larger (34 \x) than the spores
of Lycopodium. The somewhat elongated microspores are 1 5 /x
in diameter. In addition, there are to be found numerous
compound cellular bodies with a central carbonized mass.
The obvious remains of spores, from which as a center there
extends a tissue in various stages of development, leave no
doubt as to the fact that these structures represent prothalli in
various stages of growth. From these facts the conclusion is
a direct and justifiable one that Parka was a heterosporous
plant with its separate male and female sporangia enclosed in
a common sporocarp, and therefore comparable among modern
plants with Pilularia. Beyond this nothing definite can be
stated, hence for the present the specific name decipiens, as
originally applied by Fleming, is retained.

In 1855 Sir William Dawson discovered, in the Devonian
sandstones of Gaspe, the remains of a gigantic alga, to which
he subsequently gave the generic name of Nematophyton. Since
then plants of the same type have been found in Germany,
England, Scotland, and various parts of the United States,
thus indicating the very wide distribution of plants having a
common ancestry. At the present time, for purposes of con-
venience, eight different species are distinguished, but it is
altogether probable that these may in reality represent only
four separate species at the most.

As represented by specimens of N. Logani in the Peter
Redpath Museum of McGill College, these plants were some-
times eighteen inches to two feet or more in diameter, and
specimens recently found in New York State show a recover-
able length of twenty-four feet. These facts seem to indicate
that the complete plant attained to great size, far exceeding
anything known among the arborescent forms of alga^ at the
present day. No foliage or fruit has been found, and our
knowledge of these plants rests entirely upon the details of

34 BIOLOGICAL LECTURES.

internal structures, which are often most beautifully preserved.
The main features are as follows :

The stem exhibits a concentric structure comparable with
that which is shown by the larger Laminariae, a feature which
in the first instance' led to a misinterpretation of their true
nature, as it appeared to indicate an exogenous stem. This
view was still further strengthened by the appearance of cer-
tain radial canals simulating medullary rays, but of very irregu-
lar width and indeterminate length, and wholly lacking definite,
limiting walls. Such spaces reappear in N. Ortoni, N. Crassum,
and others, where they take the form of nearly isodiametric
openings in the body of the structure, through which they are
scattered without any definite order or apparent connection
with the larger elements of the organism. They nevertheless
appear to be intimately connected with the smaller and second-
ary elements, as will be shown presently, and it is therefore
quite possible that, as suggested by Barber, they were designed
to perform a certain role in the internal aeration of the plant.

The principal structure consists of tubular, non-septate cells
having a diameter upwards of 6^ ft, but of indeterminate
length and loosely interlaced, so that their general direction
coincides with the axis of growth. These large cells of the
medulla branch into smaller hyphae which form an intercellular
plexus. The hyphae are chiefly from 4-6 ft in diameter, though
sometimes much less. They appear, in some cases at least
(N. Storriei), to be septate, and their derivation from the large
cells of the medulla appears to occur chiefly in the immediate
region of the medullary spaces which they occupy in the form
of a very loose web. The majority of the species exhibit
minor differences only, but in N. Ortoni we observe the occur-
rence of well-defined trumpet hyphae, which not only estab-
lish a certain degree of aflinity with algae of the type of
Macrocystis, but also serve to suggest its possible generic
separation from the other known forms.

Our interest centers in the fact that these plants have no
modern representatives in any way comparable with them in
point of size. Occurring as they did so early at the Upper
Silurian, their gigantic dimensions and somewhat highly spe-

THE HISTORY OF PLANT LIFE. 35

cialized structure suggest that they represent a line of descent
which must have had its origin very early in Eozoic time.

The limits of time and space will not permit of our consid-
ering the many problems suggested by our title more at length,
but, in conclusion, it may be well to point out that as, with the
progress of the ages, plants became less and less aquatic in
their habits, they were brought more and more completely
under the influence of conditions which became continually
less stable, in response to the lower temperature and changing
surface of the earth's crust. This eventually led to waves of
vegetation which passed over the great land areas, from the
direction of the poles toward the equator and back again, in
direct response to changing climatic conditions. Such migra-
tions are well known as having occurred during glacial times,
and the evidence of this fact, as presented by fossil plants,
especially those of the Pleistocene age, has an important bear-
ing upon our knowledge of the distribution of plant life, as well
as upon the general history of development.

THIRD LECTURE,

INFLUENCE OF INVERSIONS OF TEMPERATURE,

ASCENDING AND DESCENDING CURRENTS

OF AIR, UPON DISTRIBUTION.!

PROF. D. T. MACDOUGAL.

The soil and the air receive the same amount of heat from
the sun during the course of the day, but the former, on account
of its greater conductivity, becomes warmer than the air resting
upon it. Both the soil and the air begin to lose heat shortly
before sunset, but the former undergoes cooling much more
rapidly than the latter, for the same reason that it becomes
warm more rapidly. If the air is poor in moisture, showing
a low relative humidity, it will conduct heat very slowly. As a
consequence of this fact, the layer of air, a few meters in thick-
ness, nearest the ground will become cooled by conduction and
radiation to the cold surface of the soil, and soon falls to a
temperature many degrees below that of the air a few yards
above. This is termed inversion of temperature by the mete-
orologist, and the effects of inversions of temperatures are
well known to those engaged in horticultural and agricultural
operations.

The late spring frosts and the earlier frosts of the autumn
are generally due to such inversions of temperature. It is a
matter of common observation in these occurrences that the
lower branches of trees and shrubs will be killed, while the top-
most buds will be unharmed ; showing that the layer of cold
air was not deep enough to submerge the entire plant. As a

1 Delivered August, 1899. Based upon investigations reported to the U.S. De-
partment of Agriculture in 189S. Printed by permission from the chief botanist.

37

^8 BIOLOGICAL LECTURES.

recent example of such action, a description of the manner
in which plants were injured by a late frost in Florida illus-
trates the matter quite fully [Botanical Gazette l8, p. 417, 1893).
In this paper it is noted that all of the young shoots on the
limbs of the china tree, Melia Azedarach L., "less than eight
feet" from the ground, were destroyed, while those ** above
twelve feet " were left unharmed. The young shoots on the
lower part of shrubs of the prickly ash, Xanthoxylum clava-
herculis L., were destroyed, while those ''above twenty feet"
from the ground were uninjured. All the young leaves of the
mulberry, Morus alba, within ten feet of the ground were killed.
These figures show us the depth of the layer of cold air, with a
temperature below the freezing point, as plainly as if we had
tested it with a recording thermometer. Very naturally the
effects of inversions of temperature are felt only on clear, still
nights. A wind would prevent the accumulation of the layer
of cold air, and a shield of low-lying clouds would act as a
reflector in preventing the escape of the heat from the ground
at the usual rate. The device of the orchardist and farmer in
building smoky fires at intervals in his crops would both give
rise to local currents, which would prevent the accumulation of
a cold layer, and would also simulate the presence of clouds ;
constituting, in fact, a very effective device for the avoidance
of frosts. Coverings, however slight, such as paper or of the
thinnest and lightest cloth, would also prevent the greatest loss
of heat from the soil underneath the plants thus protected.

These inversions of temperature occur practically over the
entire temperate zone in a manner of importance to the plant,
and this phenomenon accounts for the low night temperatures
prevalent on all the elevated plains in western North America,
in which the effect is emphasized by th'e comparatively low
relative humidity.

The frosts resulting from this action are not to be confused
with the freezing which ensues when great waves of cold air
sweep over a section of the country.

In the summer of 1898 I carried out some ecological obser-
vations for the United States Department of Agriculture on
the plateau, inclusive of northern Arizona, and lying at an ele-

INVERSIONS OF TEMPERATURE.

39

vation of about 2300 meters. I found the nocturnal inversions
of temperature very marked and more or less constant — a fact
which must exercise a very notable influence upon the zonal
boundaries which cross this region. As will be shown presently,
northern plants would go farther south on a plain subject to
inversions of temperature than one free from such influences.

Quite early in the course of my investigations attention was
drawn to the fact that crops of oats, wheat, and vegetables in
the valleys around the San Francisco Mountains were frosted,
while those on uplands and ridges were unharmed. In order

6A.M

July 18-23.

Fig. I. — Temperature curves obtained at Flagstaff, Arizona, at elevations of about 2300 meters.
The dotted line is the thermographic record for the hilltop, and the unbroken line is the
thermographic record for the valley, 100 meters lower. The valley is warmer than the
hilltop during a period of only about three or four hours during the middle of the day, but
its temperature is lower than that of the hill during almost the entire night of twelve hours.
The daily average temperature of the valley is thus much lower than that of the hill.

to determine the temperature relations thus indicated, record-
ing thermometers were placed in the valley at Flagstaff and at
the Lowell Astronomical Observatory, lOO meters above it, and
continuous records of the temperatures were made through June
and a part of July. The record of a few days is shown in fig-
ure I. From the data thus obtained it is to be seen that the
valleys and canons are subject to much greater extremes of
temperature than the adjoining hills and mesas, and that the
average daily temperature of the valley is lower than that of
minor elevations near it. The temperatures of the hill and

40 BIOLOGICAL LECTURES.

valley begin to fall shortly before sunset, but continue with
such unequal rapidity that the difference is very great imme-
diately before sunrise. Thus, for instance, on the morning of
Sept. 1 8, 1898, the temperature of the hilltop was 62° F. at
6.45, while* that of the valley was 35° F., showing a differ-
ence of 27° F. between two points which differed only 100
meters in elevation, and which were actually so near each
other that one could walk from one instrument to the other in
ten minutes.

If the curves of the thermograph are examined, it will be
found that the valley becomes from 3 to 8° warmer than the
hilltops during the middle of the day.

These differences of temperature of the valleys and highlands
are due to the action of currents of air set up by the inversions
of temperature and by the direct action of the sun's rays.

If the surface of a region were perfectly level, the surface
layer of air cooled by inversion would remain in place, and a
fairly uniform temperature would prevail over the entire area.
The cooling of- the lower layer of air results in its contraction
and consequent increase in weight, with the result that in
regions with broken or irregular topography the cold air on the
mesas, ridges, and hilltops flows down the slopes into the
canons, valleys, or other depressions, forming a deep layer,
while a constant supply of warmer air settles down on the high-
lands, with the result that these minor elevations have a higher
average temperature and a more equable climate than the
valleys below.

The influence of this drainage of cold air has long been
recognized by the farmer and horticulturist in the northern and
New England states. Thus it is customary to select ridges and
uplands for the growth of vineyards, orchards, and small fruit
plantations in broken countries, since these places are less
subject to frost than the lowlands.

If the air cooled by inversion on the highlands descends by
long, continuous, steep slopes and drains into narrow valleys,
it may exercise exactly the reverse influence upon the tempera-
ture of the valley. As it descends it will increase in tem-
perature at the normal adiabatic rate, by compression, and

INVERSIONS OF TEMPERATURE. 41

actually reach the valley or plain as a warm wind, if it has had
a descent of a few thousand feet. As a fall of sixty meters
gives rise to an increase of temperature of 1° F., the total
amount may be distinctly noticeable, although but few actual
observations of this phenomenon are recorded. If the cooling
slopes descend gradually to the plain, the heat lost by radiation
in the slow downward progress of the current may entirely
compensate for the adiabatic increase, and the breeze will
reach the lowlands with no marked change in temperature.

A probable example of this adiabatic increase of temperature
of a descending current of air is to be seen in some of the ther-
mographic records made at Flagstaff, Arizona, by myself. This
place lies in the mouth of a great valley, enclosed on one side
by a mesa, on the other by a spur of the mountain, and on the
other the valley gradually ascends the long slopes of the San
Francisco Mountain, reaching the summit of Humphrey's Peak,
twelve kilometers distant, at a height of over 4000 meters. As
previously described, the cold air from the near-by mesa and
ridges pours into the valley, steadily reducing its temperature
through the night. The downward currents also form on the
long slopes of the mountain, but, descending very slowly, do
not show any great rise in temperature, and usually do not
occasion any marked disturbance in the cooling of the valley.
At certain times, however, the descending current gains such
strength and rapidity that its increase in temperature overbal-
ances radiation, and it reaches the valley in the latter part of
the night as a warm current which causes an upward bend of
the thermographic curve, which is steadily falling. The rise
due to this warm current may amount to 2 to 5° F, at some
time between 2 and 5 a.m.

The heating of the layers of air in contact with the soil in
valleys and canons during the daytime generally results in
ascending currents, which, cooling at the normal adiabatic rate
as they ascend, blow over the adjacent mesas and hilltops as
cool winds. This reduces the maximum of the highlands, and
in the curves taken on the Observatory hill at Flagstaff it is
seen to be 2 to 6° lower than that of the valley below. It is
thus to be seen that the nocturnal inversions of temperature

42 BIOLOGICAL LECTURES.

and the diurnal ascending currents complement each other in
reducing the extreme of the temperature of the highlands and
in increasing the daily variation of the canons and valleys.

The phenomena discussed in the preceding paragraphs are
well known to the meteorologist, and may be found more or
less thoroughly exploited in text-books upon that subject, and
the above lengthy consideration of them is given in order that
their application to the distribution of life and the limits of
zones may be better and more easily comprehended.

It is presumably a fairly established conclusion with bio-
geographers that the northward distribution of southern plants
and animals is governed by the sum of the positive tempera-
tures for the entire season of growth and reproduction, checked
of course by the low temperatures of the winter season. Posi-
tive temperatures are computed by adding the excess of tem-

FiG. 2. — Showing drainage of cold air into a valley.

perature above the point at which plants start into activity in
the spring. The season during which the positive temperature
would be of moment would embrace only a few weeks of the
hottest part of the summer (Merriam, " Life Zones and Crop
Zones," Bull. No. l6, U. S. Dept. of Agriculture, Div. of Biol.
Survey). The limiting low temperatures might occur at any
time in the winter season.

It is also generally accepted that the southward distribution
of northern plants and animals is governed by the mean tem-
perature of a brief period during the hottest part of the year.

Now it needs but a glance to bring us at once to the con-
clusion that the valleys have a lower mean temperature and a
lower minimum temperature than the adjoining hills, and that
they are therefore better suited to the southern advance of
northern species than the reverse. Again, the higher minimum

INVERSIONS OF TEMPERATURE.

43

temperatures of the ridges and hills, and the greater sum of their
positive temperatures would eminently fit them for the northern
advance of the southern species. Given then a great plain, cut
by valleys and canons, or the foothill region of a mountain
range, it may be definitely asserted that the southern plants
reach their northern limit on the highlands, and the northern
species reach their southern limit in the lowlands. Stated in
another form, it may be said that zonal boundaries are deflected
southward in valleys and northward on ridges and highlands.
This conclusion applies, of course, to topographical features
only which show differences in elevation of not more than 300
to 500 meters. Even with this limitation the configuration
of the country may be such as to deflect the zonal boundary
from its general course 100 kilometers or more; a fact of very
great importance both in biogeography and also in economic
operations.

The conclusions reached in the last paragraph are amply
supported by the observations of the writer in the San Fran-
cisco Mountain region in northern Arizona, also by numerous
facts of the distribution of plants and animals in an adjoining
region in New Mexico, which have been kindly cited me by
Professor Tyler-Townsend.

Large numbers of species of northern plants are to be found
in the valleys of the San Francisco Mountains, while many
southern forms are to be collected beyond the general outline
of their zone on the southern mesas, hills, and spurs. The most
marked illustration of the descent of the northern flora is to
be found in the crater of the San Francisco volcano. The
eastern wall of the crater has been broken away at some time
in the last stages of its activity in Tertiary times, and the crater
is now a U-shaped valley, ten kilometers long and three to
five kilometers wide, heading up on the inner slope of the
mountain at an elevation of 3500 meters, and opening out to
the eastward on the plain at an elevation of about 2500 meters.
According to current conclusions the species of the pine or
transition zone should advance up this valley, especially along
the face of the northern slope or side of the valley which has
the most favorable slope exposure. The effect of the slope

44 BIOLOGICAL LECTURES.

exposure should be such as to advance the transition flora
about 300 meters higher along the northern wall than along
the southern. As a matter of fact, the effect of slope expo-
sure is totally lost, and the boundary of the transition zone is
pushed backward and downward 300 meters below its normal
level.

Professor Tyler-Townsend states that similar illustration of
this principle is to be seen in the Mesilla valley, a portion of
the valley of the Rio Grande in New Mexico. The valley has
an elevation of 1200 meters and rises gradually on the east to
an elevation of 1700 meters, on the slopes of the Organ Moun-
tains fifteen kilometers distant. As a consequence of the
meteorological phenomena detailed above, Larrea tridentata
advances northward along the mesa above elevations of 1300
meters, while it cannot grow in the valley 100 meters below.
At the same locality he reports the presence of a peculiar fly,
Raphiomidas xanthos Towns., of the lower Sonoran fauna, not
found elsewhere north of Lower California. Vohicella opales-
cens Towns., a member of the subtropical group of the genus,
and Lecanium imbricatuniy a distinctly neotropical type, also
occur in this locality. Volucella Jiaagii Jacun. occurs on the
mesa in question, as well as in similar locations in the White
and Magdalena mountains, where it is beyond the general curve
of its zonal limits.

That this distribution is entirely the result of the meteor-
ological factors, chiefly inversions of temperature and cold-
air drainage, seems to be indicated by the fact that genera of
flies such as Gymnosoma^ Hyalomyiuy Trichopoda, Cistogaster,
Heniyda^ and Ocyptera, which range from the transition zone
to the tropics, are found both on the mesa, along with the other
subtropical forms, and also in the valley, showing that food sup-
ply or other conditions of environment do not account for the
northward advance of the southern forms enumerated.

The facts which have been brought before the writer in the
investigations described above lead to the belief that many of
the so-called aberrations of distribution which have been re-
corded in various parts of the world may be found directly
attributable to the influence of the ascending air currents,

INVERSIONS OF TEMPERATURE. ' 45

although it is well known that species of plants and animals
often form isolated colonies outside the natural boundaries of
their zones because of special or highly local conditions of
dissemination agencies or food supply.

Incidentally, it is believed that these meteorological factors
may have some connection with the variations in the depth and
tint of the autumnal colorings of foliage leaves, though the
apparent evidence may be entirely coincidence.

Aside from these phenomena which deflect the zonal bound-
aries of plants and animals, I now wish to call your attention
to certain movements of the air which may be said to have a
purely local selective effect.

Among these are the chinook or foehn winds, which are
prevalent in certain localities over the world, being especially
well developed in western North America, and the moist ascend-
ing currents which exercise their greatest influence in regions
with low relative humidity.

Chinooks are formed to the leeward of nearly all high moun-
tain ranges, and are heaviest near those which stretch their
barriers directly across the course of the prevailing winds. The
factors necessary to produce a chinook consist chiefly in the
difference in barometric pressure on the opposite sides of the
mountain ridge, a high-pressure center to windward and a low-
pressure center to leeward. The ascending current of air
which flows over the crest to restore the equilibrium loses much
of its moisture in the elevating, expanding, and cooling processes,
and then as it descends the opposite slope, undergoing compres-
sion, its temperature increases at the normal adiabatic rate of
1.6° F. for every 100 meters of descent, and reaches the lower
slopes as a warm, dry wind, comparatively speaking. The tem-
perature of such winds is rarely above 50° F., but they occur
at a time when the general temperature is near or below the
freezing point.

Under the influence of this high temperature many plants
are induced to start into activity, and as these chinooks last
only a few hours, the restoration of the general average tem-
perature results in the destruction of the partially opened buds
and active growing plants, while the direct desiccating action

46 BIOLOGICAL LECTURES.

of the warm wind itself is often very injurious. The presence
of the Chinook results in the elimination of the species which
are particularly susceptible to the warm, dry air from the locali-
ties affected.

A few years since, the leaves of coniferous trees on the lower
slopes of Pike's Peak were killed in this way, and the San Fran-
cisco Mountains offer marked examples of this action. On
the eastern slopes of the principal elevation of this group the
coniferous forest reaches an elevation of 3000 to 3500 meters.
On these slopes, in the localities which are exactly suited to
receive the strongest action of the chinook, are extensive areas
of pine and spruce trees which, so far as repeated examination
by myself and others shows, were not destroyed by fire or dis-
ease, and the supposition is fairly allowable that the chinook
was the destructive factor. This supposition is strengthened
by the fact that the small hills and volcanic cones which stand
directly to the eastward in the path of the winds lose their snow
before other places of the same altitude and geological forma-
tion elsewhere in the region.

In addition to the action of the ascending currents of air
on the temperature of highlands they often occasion serious
disturbances of the moisture conditions in a highly localized
manner.

The ascension of a current of warm air from the bottom of
a valley or canon necessitates the loss of some of its heat to
furnish the energy for expansion, and the consequent cooling
entails the lowering of the dew point. This action may pro-
ceed until the point of saturation is reached and precipitation
occurs. More generally, however, the current reaches the
highlands as an extremely moist wind. The margins of mesas
and hilltops adjoining deep valleys are most subject to such
moist winds, and as a consequence these localities offer suitable
conditions for moisture-loving species which thus may occur in
great abundance as oases in an arid region. As the moisture-
laden current traverses the high plain, it soon becomes heated
to the temperature of the air through which it is flowing, and
its relative humidity is decreased.

The moisture influence of ascending currents is strikingly

INVERSIONS OF TEMPERATURE. 47

illustrated in the region of the Coconino forest bordering upon
the famous Grand Canon of the Colorado River in northern
Arizona, where many moisture-loving species may be found
fringing the rim of the canon, which are not to be seen in such
abundance elsewhere on the mesa. Razournofskya robtista vagi-
nata (Willd.) Kuntze ig a loranthaceous parasite growing on
the branches of the bull pine {Piniis poiideivsa var. scopuloriim)
throughout the transition zone. It is most abundant, however,
along the margins of mesas, the rims of canons, and hilltops,
and reaches its greatest density along the Grand Canon on the
Colorado River. Here the heated air, rising from the river
under the influence of the subtropical sun, loses about 20° F. in
its ascent of 2000 meters, and as a consequence it pours over
the mesa much cooler and with its relative humidity increased
to near the point of saturation. The Razoiimofskya finds this
strip of territory, where the effect of the moist air is greatest,
most advantageous in the germination of its seeds and the
attachment of the seedlings to the host plant. It is therefore
most abundant in a belt one or two kilometers in width, run-
ning parallel to the rim of the canon, while it is comparatively
infrequent at greater distances. Within this belt it is estimated
to have gained a foothold on more than half of the pine trees.

In conclusion, I must again remind you that none of the
meteorological principles described are in any sense original
with myself ; but I am able to adduce some very striking ob-
servations in illustration of their influence upon vegetation,
and it is confidently believed that this work constitutes the
first systematic attempt to use such data in explanation of cer-
tain seeming aberrations of distribution and zonal boundaries.
Furthermore, it is most interesting to note that the effects of
cold-air drainage and inversions of temperature have been taken
into consideration by the horticulturist and farmer long before
this analysis of their relations to the facts of natural distribu-
tion of plants and animals was brought forward.

FOURTH LECTURE.

SIGNIFICANCE OF MYCORRHIZAS.

PROF. D. T. MACDOUGAL.

By reason of the great adaptability of the seed plants, any
underground member, roots, stems, branches, or leaves may
come to serve as organs of absorption of liquid nutriment from
the soil in different species. Furthermore, because of this
readiness to adopt any method which will enable them to acquire
food more economically, the absorbing organs of an extremely
large number of species have acquired the habit of forming
symbiotic unions or partnerships with the hyphae of soil fungi.
Such unions are termed " mycorrhizas." But mycorrhizas are
not confined to the absorbing organs of the higher plants. The
roots of the sporophytic generation of a number of ferns and
fern-like plants also form similar structures, while the wide
occurrence of fungi in the thallus-like forms of the gameto-
phytes of the ferns, lycopods, hepatics, and equisetums, in such
manner as to constitute mycorrhizas, is well known.

A mycorrhiza is classed as ectotropic or endotropic, according
to the manner in which the contact between the two plants is
made. Ectotropic mycorrhizas are those in which the fungus
forms a thick felt or sheath around the absorbing organs of the
higher plant. This form is always made by the union of a
fungus with true roots and never with stems, leaves, or pro-
thallia. The sheathing mass of hyphae may completely enclose
the root, or it may be absent from the apical region of this
member. In ectotropic forms the fungus does not actually
gain entrance to any of the living cells of the higher plant,
except in one or two families. The epidermal cells are papil-

49

50 BIOLOGICAL LECTURES.

lose and touch each other at the base only, leaving wide
interstices between their outer extremities, which are occupied
by masses of hyphae. As a result of this arrangement, the
fungus appears to occupy the epidermal cells when a cross-
section of the root is examined, but such is not the case except
perhaps in Pterospora^ Monotropa, Hypopitys, and Sarcodes,
which I have recently examined as to this particular. Populns
(poplar), Querciis (oak), and the conifers furnish examples of
ectotropic mycorrhizas in which the fungus does not enter the
epidermal cells.

In order to comprehend the other, or ectotropic type of
mycorrhiza, it will be necessary to recall that the typical struc-
ture of roots is one in which the central portion of the organ is
occupied by the stele, of which the fibro-vascular tissue is the
principal component in bulk, and that this central cylinder is
surrounded by a sheath composed of thin-walled cells of the
cortex, outside of which lies the cortex proper, made up of ten
to thirty layers of large parenchymatous cells. Covering
the root is the epidermal system of one to three layers, with
the outer walls of the outermost layer drawn out into tube-like
extensions — the root hairs. The hyphae of the fungus gain
entrance to the root when quite young by pushing into it from
the stem, or by piercing its tender epidermis. Once inside
the root, it generally shows a differentiation into three regions.
The hyphae which enter the root push forward through the
walls of the outer cortex toward the apex of the root, occupying
one to three layers of cells with hyphae which are fairly free
from convolutions or enlargements. This portion of the body of
the fungus may be termed the " vegetative mycelium." Numer-
ous branches from the vegetative mycelium are sent into the
middle region of the cortex, and these branches subdivide and
pierce the thin walls of the cortex, so that they may occupy
more than half of the cells. These internal branches appear
to be most delicately chemotropic. In some plants they are
attracted to the neighborhood of the nuclei, where they form
large vacuolated sacs, termed "vesicles," or agglomerations of
interwoven hyphae, which doubtless serve as organs of inter-
change between these two plants, as will be described further

SIGNIFICANCE OF MYCORRHIZAS. 51

on. In some instances the tips of the growing hyphae are
attracted to the plastids or to the centers of carbohydrate for-
mation. The presence of the fungus with the organs of inter-
change generally sets up disturbances in the cells occupied, of
which enlargement of volume, hyperchromatism, and sometimes
fragmentation of the nucleus are the most noticeable. It is
most notable, however, that the cells occupied by the vegetative
mycelium undergo almost no variation from the normal charac-
ters of such organs. The cells of the middle portion of the
cortex are generally very much enlarged, but this has been
shown to be due to the character of the food material received
rather than to the presence of the fungus. Branches of the
vegetative mycelium pass outwardly through the epidermis,
which may or may not traverse the root hairs, and after emer-
gence into the soil they are numerously divided and extend to
unknown distances in the soil. The mycorrhizal fungi of the
gametophytes of the lower forms usually occupy the external
or the lower layers of the thallus.

On account of the great elasticity of the various factors to
be considered in the determination of the nature of the inter-
change which takes place in mycorrhizas, the final solution of
the matter has been attended with great difficulty. The chief
features of the metabolism of the uniting cells of the two organ-
isms, the composition of the substratum, the anatomy of the
higher plant, with special attention to degenerations, and the
behavior of the two symbionts when grown alone or in pure
culture must all be taken into account, and the consideration of
all these points entails an enormous amount of observation,
and no single mycorrhiza has yet been carried through all stages
of the work. Our conclusions, which are fairly final as far as
they extend, are therefore derived from evidence more or less
fragmentary, so far as any single formation is concerned, yet is
to be taken as fairly and justly supplementary.

So far as the various theories concerning mycorrhizal plants
are concerned, it is to be said that Unger concluded that
Hypopitys {Monotropa hypopitys) was parasitic on the roots of
trees (1840). Pfeffer recognized the symbiotic relations of the
members of a mycorrhiza in 1877, ^"^^ this was demonstrated in

52" BIOLOGICAL LECTURES.

Hypopitys by Kamienski in 1881. Frank, in a series of studies
published between 1885 and 1891, originated the theory that
the higher plant is really parasitic upon the fungus, inducing it
to grow on and in its roots, and finally yield its components to
the higher plant, and he termed the arrangement a *' fungus
trap." Recently Janse has endeavored to prove that mycor-
rhizas sustain the same functions as leguminous tubercules
(1896). The theory of Janse has been found to hold true
of Podocarpiis and the peronosporous fungus with which it
forms mycorrhiza only, although it may be applicable to a few
other special forms.

So far as the mutual dependence of the mycorrhizal symbionts
upon each other is concerned, it may be stated, with a fair
degree of certainty, that all of the fungi are capable of more or
less extended periods of growth and endurance in the soil ;
though to what extent has not been ascertained. The higher
plants which habitually form mycorrhizas are capable of devel-
opment in sterilized soils, but according to Frank's experi-
ments they do not attain the size of specimens united with
fungi in the usual manner. This applies only to those species
which are furnished with chlorophyl, and Frank's experiments
were made with coniferous trees. The capacity of such chlo-
rophylless forms as Monotropa and Pterospora for growth in
sterile cultures has not yet been determined, though I hope to
have some results witliin the coming season. It is pertinent
to recall in this connection the fact that when spruces or other
coniferous trees with mycorrhizal roots are cut down, the stump
often forms a callus plate, which finally covers the entire in-
jured surface, and it continues to live without forming leaves
or other organs for constructing food from the air. This was
thought by Frank to demonstrate that the living stump was
parasitic on the fungus attached to the roots. This conclusion
did not take into account the comparatively enormous amount
of food stored in the roots of the tree, or of probable metabolic
capacities of the protoplasm of the tree not possessed by the
fungus.

Omitting the discussion of the anatomical details of the
union of the- two plants which form a mycorrhiza, I may say,

SIGNIFICANCE OF MYCORRHIZAS. 53

in general, that we are justified in the following conclusions as
to the relations of the symbionts ; the higher plant furnishes a
habitat for the vegetative mycelium of the fungus, and yields
to it certain carbohydrate foods, principally starch and sugar.
The fungus takes up humous compounds from the soil which
are poor in oxygen, conducts them to the branches inside the
body of the higher plant, and manufactures proteids which it
in turn yields to the higher plant. The actual food substance
may all come through the body of the fungus, yet it undergoes
elaboration in the higher plant to forms which are very advan-
tageous foods for the fungus, such as starch and sugar.

Ectotropic mycorrhizas are generally unmistakable, but in
many instances it is difficult to determine whether the fungus
inhabiting a root is parasitic or sustains symbiotic relations to
the higher plant. Indeed it may do both at different stages of
development perhaps. The chemical exchange between the
two plants may balance at one time^ while at another, one may
receive a preponderance of benefit. Then again a mycorrhizal
formation may enable the individual to obtain its food more
economically, yet such easy nutrition may set up degenerations
which will unfit the seed plant for living outside of the nar-
rowest conditions, and which may result in the extinction of
the species. Pterospora appears to be an example of such
degeneration.

As a result of the investigations of the last decade, it is found
that the habit of forming mycorrhizas is of very wide prevalence
among the seed plants, ferns, lycopods, liverworts, and equise-
tums. This and the determination of the fact that all plants
are capable of absorbing and using organic substances of some
complexity of structure must be accounted among the most
important results of research upon the nutrition of plants
during the closing decade of the century.

So' far as relationships are concerned, it may be said that the
conifers, the orchids, the heaths, oaks, poplars, beeches are
most thoroughly given over to the formation of mycorrhizas,
and a few of the orchids have developed this adaptation to such
extent that they have lost their chlorophyl and depend entirely
upon the fungus for carbon compounds. The pine saps {Mono-

54 BIOLOGICAL LECTURES.

tropacecB)^ formerly included with the heaths, comprise nine
genera, Allotropa^ Monotropa, Pterosporay Hypopitys, Sarcodes,
Newberrya, Sckweinitzia, Pletijicospora, and Ckeilotheca, all of
which have lost their chlorophyl and depend upon the fungi
symbiotic with their roots for organic nutriment. All of these
genera are inhabitants of the north temperate zone, arranged
in a belt which circles the globe, but with the greater number
of species and genera in western North America.

Hitherto all chlorophylless seed plants which were not para-
sitic have been classed as saprophytes, but in some recent pub-
lications I have been able to show that this term may be applied
only to one species of seed plants, Wullschlcegelia aphylla.
All other so-called saprophytes are found to be symbiotic with
fungi in such manner as to form mycorrhizas. The tendency
to form mycorrhizas is most marked among species living in
humous soils. The percentage of the flora with the mycor-
rhizal habit increases from the arctic circle to the equator
as a consequence of this fact. Hoveler describes mycorrhizas
on 24 of the 72 species, native and cultivated, which he exam-
ined in northern Germany. Schlicht found 70 of the 105
species native to northern Europe to be provided with mycor-
rhizal adaptations. The writer found 13 of the 15 selected
species growing in the United States to be mycorrhizal.
Janse found mycorrhizas on 69 of the 75 species growing in
Java, which he examined. Wahrlich examined $00 species of
orchids, chiefly from the tropics, cultivated in the greenhouses
at Moscow, all of which he reports as being furnished with
mycorrhizas.

The identity of the greater number of fungi which enter
into mycorrhizal combinations is unknown. Present informa-
tion leads us to believe that each species of the higher plants
forms mycorrhizas with one, or at most two or three fungi only
in the different parts of its range. On the other hand, it is
fairly probable that one species of fungus is capable of enter-
ing into such combinations with several of the higher plants.

Tulasne noted the fact that Elaphomyces forms the ecto-
tropic mycorrhizas on the roots of pines, to which Boudier
attributed a parasitic relation. Reissek determined the fungus

SIGNIFICANCE OF MVCORRHIZAS. 55

on Platanthera bifolia as Fiisisporium endorhizuni. Schacht
found that the fungus in the roots of Liinodortim developed
reproductive organs similar to Eurotium. Treub and Bruch-
mann conclude that a Pythiiim forms the symbiosis with the
prothallium of Lycopodium. Bruns found that Polysaccimi
formed coatings on the roots of pines. Woronin identified
the mycelia on conifers, willows, poplars, hazelnut, birches,
and grasses as belonging to the Boleti. Rees determined
the fungus on certain conifers as an Elaphomyces. Janse
named the one on Celtis, Celtidia duplicispora, to be included
with the Tiiberacece. Recently Rees and Fisch have exam-
ined the mycelia formed by E. granulatus and E. variegatus
on the roots of the forest trees, and conclude that the rela-
tion is not a parasitic one, but constitutes a mycorrhiza.
Noack made some attempts to form mycorrhizas by experi-
mental methods in 1 887-1 889, and found that G easier fimbri-
ates and G. fornicattis form mycorrhizas with the roots of
conifers ; Agariciis terreiis with beeches and firs ; Lactarius
piperatus with Fagiis sylvatica and Quercus pedimcidata ; Cor-
tinarius callistetis and C. coertilescens with beeches ; C. fiil-
niineus with oaks. Wahrlich determined the fungus of the
mycorrhiza of Vanda tricolor as Nectria goroschankiana^ and
that of Vanda siiavis as N. vanda. Lendner made some spore
cultures of the fungi of Platanthera and Vanda in 1895, and
his observations confirm the results of Wahrlich. Chodat and
Lendner found that the fungus of Listera cordata resembles
Nectria as originally described by Wahrlich on Vanda. Jen-
nings and Hanna conclude that the fungus symbiotic with
Corallorhiza innata is a "hymenomycete and commonly an
agaric." Clitocybe infiindibuliformis was found attached to
the coralloid formations in one instance, and Hysterangiiim
stolonifenim in another — indicative of their identity with the
fungal symbiont. Nobbe and Hiltner observed the perono-
sporous fungus in the mycorrhiza of Podocarptis, and the writer
has found reproductive bodies resembling Penicilliiim on the
mycorrhizal roots of Pterospora. Jeffreys regards the fungus of
the gametophyte of Botrychium as being intermediate between
a Completoria and a Pythium. These and a few other examples

56 BIOLOGICAL LECTURES.

pointed out by Sarauw in his review of the subject in 1893
{Rodsymbiose og Mykorrhizer^ Copenhagen, 1893) comprise our
information upon this phase of the subject.

All known species of mycorrhizal fungi may therefore be
included in the Odmycetes, Pyrenomycetes^ Hymenomycetes, and
Gasteromycetes. The independent culture and identification
of these fungi offers a most alluring field to the mycologist,
and the exploitation of this work will doubtless result in the
setting up of scores of new species, the erection of genera,
the addition of new chapters to the life history of polymorphic
forms, and the furtherance of our knowledge of the morphology
of this great group of organisms, all of which will be of vital
importance in the determination of the essential features of
the general physiological value of mycorrhizas and the humus
relations of the higher forms.

The widespread occurrence and distribution of mycorrhizas
render this adaptation of high importance in the nutrition of
the greater number of the perennial seed plants. At the pres-
ent time no reports are at hand of any attempt to use our
information on this subject in horticultural work, or in the cul-
tivation of economic forms, although I confidently predict that
it will ultimately be found of high value in the transplantation
and growth of woody plants, shrubs, and trees cultivated for
fruit, foliage, and other products ; a forecast easily compre-
hended when we glance at the vast agricultural importance of
the results of our study of the leguminous tubercles, which are
in fact a special form of mycorrhizas.

Since the above lecture was written, Bernatsky's paper has
come to hand, in which he transfers the species of Nectria seen
by Wahrlich in Vanda tricolor to HypomyceSy and finds that H.
Vandce (Wahrl.) Bernts. also unites with Vanilla aromatica and
Psilotum triqiietrum to form mycorrhizas, and that Hypomyces
Psiloti Bernts. is the fungal symbiont in the mycorrhizas of Psi-
lotum. (J. Bernatsky. Adatok az endotroph mykorhizak ismere-
tehez. Termeszetrajzi Fuzetek. 22 ; ZZ. 1899.)

FIFTH LECTURE.

INSTINCT.

EDWARD THORNDIKE.

I HAVE first of all a request to make of the reader. It is
that, in receiving and estimating what I say, he temporarily
discard the definitions or formulae for psychological phenomena
now in his mind. Many of the discussions and quarrels of
comparative psychology are about mere words, and are there-
fore fruitless. We can avoid all such if for the time being I
am allowed to name the phenomena, the facts about which we
are to think, as I please. I hope to make clear what facts I
am referring to in every case, and you will be at liberty to
rename them after your own preferences as soon as we part
company.

The facts to which I refer by the words "instinctive reactions"
or " instinctive activities" or " instincts " are any activities which
do not have to be learned, which the animal is capable of without
experience. Let us not mind whether the act be accompanied
by consciousness or not, whether it represent the inheritance
of some ability acquired by the animal's ancestors or not,
whether it involve emotional feeling or not. I shall, for the
sake of keeping in line with customary usage, deal with such
instinctive activities as physiology generally leaves out, e.g.^
instinctive fears, food preferences, motor control in running,
jumping, flying, etc., though breathing, defecation, sleeping,
etc., really deserve treatment of just the same sort as these
more complex activities. Under instincts, then, let us study
all the abilities to respond to different situations (more particu-
larly certain complicated external situations) which the animal

57

58 BIOLOGICAL LECTURES.

has to start with, has apart from parental or self tuition. An
instinctive act, then, will be the opposite of an act due to
experience. The sum of an animal's instinctive abilities plus
the habits taught him by his life struggle will be the total of
his store of ability.

It is clear that what an animal has to start with, due to the
organic structure it inherits, is a matter of prime importance in
the case of any animal whose consciously directed activities
you are studying. If we do not know its instinctive equipment,
we are likely to credit it with intelligent thinking for doing
something it really couldn't have helped doing. Suppose I say
to you, *' I have an intelligent chick who has learned so to adapt
his movements in the water that he can swim to shore," and
tossing him into the water, demonstrate that fact to you. It is
probable that many a one would say, " How remarkable ! Do
you suppose he reasoned out the way to act } How did he
learn to do it .? " The real fact would be that he did not learn
to do it at all ; that all chicks react to water by swimming out
of it the very first time they ever get in it ; that chicks swim
instinctively. How long error may persist about any animal
activity is well shown by the recency of our knowledge that
walking is instinctive in the human infant. So the first task
of comparative psychology is to find out the instinctive equip-
ment of any animal studied. Instincts are, however, well worth
study for their own sake. An instinctive fear of a certain
enemy may be as truly useful to an animal as sharp teeth or
protective coloration. Instincts are the expressions of struc-
tures and functions of the nervous system, and are as real and
as important matters for the biologist as are bones and blood
vessels.

It is outside the province of this lecture to enumerate or
describe the particular instincts of any group of animals, but we
should note that in spite of the tremendous number of instincts
that have been observed, the story has not been half told. A
rich field awaits the investigator. The humble chicken has
been under everybody's observation, and has been specially
studied by Spalding, Eimer, Preyer, Lloyd Morgan, and others,
yet I was able to record the following additional instinctive

/ . INSTINCT. 59

activities : swimming to shore when thrown into the water,
fighting (observed as early as the sixth day), appropriate reac-
tions to distance in jumping down from heights, avoidance of
open places, reacting to honey-bees by seizing, knocking them
against the ground, and then eating them (see Psychological
Review^ May, 1899).

Turning to our proper task, the statement of some character-
istics of instinctive activities in general, we should first recall
the familiar fact that to say that a certain ability or tendency
is in an animal apart from teaching or experience, is born in
him, is not to say that at birth he possesses it or that all
through life he keeps it. Scratching the ground in the case of
the chick, walking in the case of the human infant, are obvious
examples of true instincts which yet do not appear for some
time after birth. The instinct to follow and the instinctive
avoidance of loneliness which the chick manifests so markedly
in the earliest days of his life disappear to a large extent if
given no chance to harden into habits. Instincts, that is, may
be delayed and may be tra7isitory. Care must always be taken
by the student, however, not to interpret as delayed instincts
cases where the reason for delayed activity is really lack of
strength in some organ, or as transitory instincts cases where
the reason for loss of the activity is loss of strength or inhibi-
tion of the activity by unpleasant consequences.

Perhaps the most important general statement one can make
about instinctive reactions is that they are often indefinite and
inexact. The same situation may be reacted to differently by
different individuals of a species, or by the same individual on
different occasions. The Peckhams (Wisconsin Geological
Survey, Bulletin No. 2) have shown decisively this inexact-
ness, this vagueness of reaction in the case of the nest-building
and stinging habits of the solitary wasps. Among vertebrates
one sees it almost everywhere. If you slam a door in a room
where you have half a dozen chicks, one may run and crouch
for twenty seconds, another may squat where he is for an equal
time, another may also chirp, while the others may keep on
feeding undisturbed. The old theory of instinct, that it was a
sort of God-given power bestowed on animals to make up for

6o BIOLOGICAL LECTURES.

their lack of reason, which in mysterious ways directed the
animal's footsteps and aroused in him by direct inspiration the
appropriate act in any set of circumstances, left us an unfortu-
nate legacy in the shape of a tendency to expect an accuracy
and infallibility and unchangeableness in animals' reactions
such as supernatural inspiration might well give, but which are
usually signs of death in the natural world. The common
human interest in the marvelous and unusual cooperated with
this tendency by selecting for observation such instinctive
activities as web-spinning, honeycomb construction, etc., and
neglecting the more ordinary activities. Thus arose such
amusing expressions as " Failure of instinct." Furthermore,
so long as there survived vestiges of this teleological notion of
an entity, "instinct" which fitted acts to situations so as to
get desirable results, there was the less interest in the real
intermediary between situation and act, namely, the nervous
system, and less likelihood of men looking for the variations in
response to be expected in the results of the activity of any
living organ.

The recognition of the vague in instinctive activity not
only brings such activity into line with other biological phe-
nomena, and lends a healthier tone to investigation, but also
provides a useful warning to the observer. The naturalist who
studies a single case of such activity is almost sure to be
misled. *' In a multitude of witnesses there is strength."

The utility in the struggle for existence of an animal's
equipment of instinctive reactions is much increased by reason
of their ability to harden into habits. It is a wide if not
a general law that any act which has been performed in a cer-
tain situation and resulted in pleasure, or even indifferently, is
the more likely to be performed again in that same situation,
the reverse happening to any act resulting in discomfort.

The general manifestations of this law will be dealt with
in a later lecture, but its influence on instinctive reactions
deserves mention here. First of all, transitory instincts may
gain thereby equal value with permanent ones.

The chick that has followed the mother-hen for six or eight
days does not thenceforth need any instinctive impulse to

INSTINCT. 6 1

follow, for the act has become habitual and will remain in the
absence of any such impulse. Again vague responses may be
hardened into more definite and more successful forms. A cat
with a general instinct to jump at small birds might with
practice jump straighter and more quickly because the quick,
straight jumps would result in the pleasure of capture. One
is likely, however, to be misled if he argues from the presup-
position that acts always tend to assume a "perfect" form.
If the animal gets along very well with its instinct still "im-
perfect," there may be no change. Lloyd Morgan, for instance,
has chosen a dubious example of perfecting through habit in the
seizing of bits of food by chicks. They often do fail to seize
in their first experiences, as he observed, but they often, per-
haps just as often, fail even after long experience. I took nine
chickens from \o to 14 days old and placed them one at a time
on a level surface over which were scattered bits of cracked
wheat (the food they had been eating in this same way for a
week) and watched their pecking. Out of 214 objects pecked
at, 159 were seized, 55 were not. Out of the 159, only 116
were seized on the first peck, 25 on the second, 16 on the third,
and 2 on the fourth. This is far from a perfect record.

In the growth of the chick's discrimination between objects
as food we find a sure manifestation of our law. The chick
instinctively pecks at all sorts of objects of suitable size, e.g.y
tacks, match ends, printed letters, the eyes and toes of
his mates, his own excrement, etc. The pecks at bits of
food and small stones bring satisfaction, and the chick that
when first confronted by the situation, " grain of wheat, match
end, and excrement," was as likely to peck at one as another,
becomes a chick who almost inevitably pecks at the wheat.
Thus the vague instinctive response may be educated into a
lot of definite food preferences and avoidances. Not only thus
directly, but also in indirect and complex ways, instincts may
furnish the foundation of habits or, as we may better call them,
associations ; associations, that is, between certain situations or
circumstances and certain acts. A young animal instinctively
follows or keeps near its parent and thereby forms associations
which later will lead him to go independently to certain feed-

62 BIOLOGICAL LECTURES.

ing grounds, to eat certain sorts of animals or plants, to run
away from certain enemies, to sleep in certain lairs. A kitten
when confined in a cage reacts instinctively with squeezings,
clawings, bitings, etc., some one of which may happen to open
a thumb-latch on the door and give it freedom. The pleasure
consequent may so stamp in that particular act, in connection
with that situation, that after enough experiences the animal
will, when put in that situation, manifest nothing of all the
instinctive activities first observed save the one particular poke
at the thumb-piece. Its activity now would seem far enough
from instinct to many people, but it is really a consequence of
instinctive activity. It has come to neglect the unsuccessful
squeezings and bitings, to choose the successful clawing, in
just the same way that the chick comes to neglect the peck-
ings at excrement and match ends, and choose the food.

I have already implied that instinctive activities may be
inhibited just as truly as they may be confirmed and reinforced.
If the conditions in which an animal lives become such that
an instinctive act brings discomfort, that activity tends to dis-
appear ; or if the animal has, prior to the appearance of a
certain instinct, learned to meet otherwise a situation which
would normally call forth the instinct, he may continue to
meet it in that rather than the instinctive way. It would even
be fairly reasonable to interpret the transitory instincts as
instincts which were inhibited by mere lack of exercise. A
convenient account of the inhibition of instincts may be found
in James's Psychology, Vol. II, pp. 394-397. I may quote in
addition examples of inhibition (i) by virtue of the previous
formation of a habit, (2) by (i) plus actual abolition through
resulting discomfort.

An instance of the former sort is found in the history of a cat
which learns to pull a loop and so escape from a box whose top is
covered by a board nailed over it. If, after enough trials, you remove
a piece of the board covering the box, the cat, when put in, will still
pull the loop instead of crawling out through the opening thus made.
But, at any time, if she happens to notice the hole, she may make
use of it. An instance of the second sort is that of a chick which
has been put on a box with a wire screen at its edge, preventing it

INSTINCT.

63

A

B

D

C

from jumping directly down, as it would instinctively do, and forc-
ing it to jump to another box on one side of it and thence down.
In the experiments which I made, the chick was prevented by a
second screen from jumping directly from the second box also. That
is, if in the accompanying figure, A is a box 34 inches high, B a box
25 inches high, C a box 16 inches high, and D the pen with the food
and other chicks, the subject had to go A-B-C-D. The chick tried
at first to get through the screen, pecked at it and ran up and
down along it, looking at the chicks below and seeking for a hole to
get through. Finally it jumped to B and, after a similar process, to
C. After enough trials it forms the
habit and when put on A goes im-
mediately to B, then to C and down.
Now if, after 75 or 80 trials, you take
away the screens, giving the chick a
free chance to go to D from either
A or B, and then put it on A, the
following phenomenon appears. The
chick goes up to the edge, looks

over, walks up and down it for a while, still looking down at the
chicks below, and then goes and jumps to B as habit has taught it
to do. The same actions take place on B. No matter how clearly
the chick sees the chance to jump to D, it does not do so. The im-
pulse has been truly inhibited. It is not the mere habit of going
the other way, but the impossibility of going that way. In one case
I observed a chick in which the instinct was all but, yet not quite,
inhibited. When tried without the screen, it went up to the edge to
look over ni7ie times, and at last, after seven minutes, did jump
straight down. — Animal Intelligence^ an Experimefttal Study, etc.,
pp. 99-100. New York, 1898.

It remains to consider the so-called "perversions" of in-
stinct, that is, the cases where an instinctive activity appears
in response to a situation v^^hich it does not fit, or does not
ordinarily go with. The cat that nurses rats or puppies, the
pigeon that capers before a bottle, are stock examples. Most
of such are due to the essential indefiniteness of the instincts
in question. The cat does not have an impulse to nurse a
particular, definite thing, to wit, a young kitten, and to abhor
all else as subjects for nursing, and if her own young are only
taken away and other young nursing things happen to be

64 BIOLOGICAL LECTURES.

around, they may call forth the activity quite as well. Other
" perversions " may be brought about by naturally or artificially
caused inhibitions. There may also, of course, be in animals,
as there may be in men, abnormalities in the nervous system
which may occasion abnormal instinctive manifestations. These
would need separate treatment. The interest of the ordinary
*' perversions" is simply their witness, first, to the subordination
of instincts to the laws of nervous activity in general, e.g.^ va.ri-
ation, association, inhibition, and, secondly, to the coarse, rough-
hewn nature of many instincts.

So much for the important general facts about how animals
meet situations apart from experience. Before entering on the
tasks of justifying the definition of instinct, which I asked you
to accept provisionally, and of saying somewhat about the
origin and evolution of instincts, I wish to take this opportu-
nity to advocate the study of instinctive activities by students
of elementary biology. Their observation requires no knowl-
edge of psychological terminology, no ability to make psycho-
logical interpretations. The observer has simply to answer
the question, *' How does an animal react to a certain situation
the first time he is in it } " The demonstration of every matter
discussed so far in this paper may be effected at practically no
expense in any quiet place where the student can find twenty
square feet of space in which to keep a few chicks. Such
study will serve as a partial corrective to the tendency to make
biology a study, not of life, but of form. It may give the stu-
dent a taste of and for fact in this field of biology, which will
keep him from getting lost when he meets with the specula-
tions about instinct. It is, moreover, worth while for its own
sake.

Returning now to the question of definition, it is clear that
any one has a logical right to name things to suit himself and
to determine the particular things which he shall denote by a
certain word at his own sweet will, so long as he is clear and
consistent and does not reason from any other meaning of the
word than that he has given to it. One might, for example, in a
study of fission call Paramoecium " Hylobates McKinleyensis "
and commit no logical error, provided his figures showed clearly

INSTINCT. 65

the animal he meant to talk about, and provided he did not use
the name to argue, for instance, that we might, having found
fission in Hylobates McKinleyensis, expect it in the forms
ordinarily called "Hylobates." Names are not facts, and quar-
rels about them are generally quarrels about rhetorical expe-
diency. So if any one chooses to call instinctive only such
unlearned acts as have consciousness accompanying them,
leaving other such acts the name "reflex action " and lumping
in still other such acts under the words " processes of respira-
tion, digestion," etc., I, personally, have no theoretical objection.
It seems to me practically unwise to separate the discussion of
unlearned activities with consciousness from that of unlearned
activities without consciousness, so long as you are thinking
about the external form of the act ; and I have even chosen to
so define the word that even digestion and respiration could be
studied under the name. There is really no need for any name
and no cause for worry if we cannot invent any one definition
which will consistently apply to all the facts we treat. For
what is important is concrete information about particular facts.
My use of a name or definition is to point to certain actual
facts, and I trust that I have made clear by my illustrations to
what sort of facts my statements refer. To one who should
say, "By your definition the knee-jerk or the formation of
saliva is an instinct, for they are 'unlearned reactions,' — may
even be considered reactions to a 'complex external situation,'
— yet your statements about the formation of habits and inhi-
bition are not true of them," I should reply, "I plead guilty,
though I did describe further my facts as those which 'physi-
ology generally leaves out,' but you are a more stupid reader
than I expected to address."

The fact is that unlearned activities range from direct physi-
cal and chemical activities of single cells to most complicated
activities involving the nervous system ; that the series seems
to be continuous ; that if you try to separate off one lot
"instincts," another lot "reflexes," another lot "non-nervous
reactions," you find your classes running into each other. It
is because of my confidence in this continuity of nature that I
used my definition only to point to what I was talking about —

66 BIOLOGICAL LECTURES.

not to describe authoritatively the attributes of a fixed class of
facts. Practically it seemed worth while to bring out this con-
tinuity in the definition.

This rather tiresome explanation serves an additional purpose
by introducing us to a real question — the question of the origin
of instincts. For some thinkers have cheerfully begged this
question, by defining instincts as inherited habits, and by study-
ing those unlearned abilities which looked like inherited habits,
out of connection with all other unlearned abilities. In his
lecture on ** Animal Behavior" (Woods Holl Biological Lecttires,
pp. 285-338, 1898) Prof. C. O. Whitman has, in opposition to
these thinkers, defended with great care and force the theory
that such unlearned abilities as the spider's web-spinning,
chick's scratching, dog's pointing, etc., — such activities, in
short, as have been discussed in this paper, — (1) have the same
origin as the unlearned activities of reflex action, digestion,
circulation, etc., (2) are due, in fact, to organic features, which
again (3) are due to germ variations, and that (4), therefore, we
should look for and expect to find, so far as traces have been
left for us to follow, as continuous a development of such activi-
ties, as true an evolution of ** instincts " as of organs.

All that I have to say about Professor Whitman's first two
contentions is that they seem to me so true that I cannot under-
stand any one's doubting them. As one runs through animals'
unlearned activities, from those most to those least like prod-
ucts of learning, he can nowhere stop and say, '' hitherto inher-
ited habits, from hence chance variations." These first two
contentions would remain true whatever might happen to the
third. As to the third, I can only add my mite to the evidence.
The failure of chicks to avoid their own excrement and their
ability to swim seem hard to explain on Lamarckian principles.
For the avoidance of excrement must have been formed as a
habit in every individual for many generations, while the
swimming instinct has been unused for even more perhaps.
Again, the vagueness of response discovered the more we study
animals' unlearned activities is just what one would expect if
these responses were due to germ changes selected by reason
of their success in procuring survival, while it is not what we

iNSTijycT. 67

would expect from the inheritance of habits acquired in refer-
ence to definite objects. If the instinct to follow in chicks
came from the habit of following (acquired, of course, with a
hen as the object followed), we would expect it to require as
its object something at least like a hen. As a matter of fact
it does not.

Professor Whitman's fourth contention, that, since instinc-
tive activities are the results of gradual development, they
should be, not merely enumerated, described, and explained as
to their utility, but also explained as to their development and
relationships, comes as a timely piece of advice. Even if
students of instinct should never succeed in working out the
genealogy of one instinct out of a hundred, the genetic method
of study would be valuable by preventing mythologizing and
reminding the student that instinctive activities are expressions
of organic structure as truly as are the activities of digestion
or excretion. And, in closing this lecture, I wish to give some
samples of development among instincts. Professor Whitman
has traced the ancestry of some particular acts. Let us look
at some instinctive activities which persist, with modifications
of course, over a wide range of forms, which correspond in a
way to the notochord, or brain-eye, or arthropod appendages
among physical organs. The frog, lizard, chick, and cat all
react to irritation of the head by scratching with the hind leg
with a quick, repetitive motion that is startlingly alike in the
last three. Here we have an instinct which apparently ranges
nearly over a subkingdom. In the primates it is modified, the
monkeys (at least some of them) using either hind or front
limb for the purpose, while man uses only the front. Distaste
at confinement is another widely prevalent instinct, which may
be a foundation-stone laid by germinal variation once for all.
The instinct to follow is another. It might be that the grega-
rious instincts originated as exaggerations of it by the selection
of individuals in whom the impulse to follow varied in the line
of greater intensity.

SIXTH LECTURE.

THE ASSOCIATIVE PROCESSES IN ANIMALS.

EDWARD THORNDIKE.

In the previous lecture we studied the general nature of those
reactions which animals make to various situations instinctively,
or apart from experience. It is obvious that with many ani-
mals, in many situations, acts are observable which cannot be
so explained. The cat that comes when we call " Kitty, kitty,"
is not provided by her organic inheritance with any tendency
to respond to the situation, "hearing the sound kittyy kitty,''
by the act, "running toward the source of that sound." 'The
ten-days-old chick that, on coming out from the brooder, turns
round a corner and goes straight to the dish of water kept
always in a certain place, is not guided by any innate tendency
to meet the situation, "feeling thirsty while in brooder," by that
particular act. So, too, with the dog that sneezes when you
say, " Sneeze, Bowser ! " the chick that avoids pecking at excre-
ment, and with millions of animals performing all sorts of acts.
We ordinarily distinguish such activities from those purely
instinctive by saying that the animal has learned them. They
are the results, not of organic inheritance, but of individual
experience. The object of the present lecture will be to explain
such activities, to show at least one of the ways in which ani-
mals learn or profit by experience.

Some hints of how this happens were given in our discussion
of how instincts led to habits and how they were inhibited. But
the matter is so important that I may be pardoned for beginning
at the beginning. Let us watch some animals as they learn to
meet certain situations in appropriate ways. If we make a pen,

69

70 BIOLOGICAL LECTURES.

as shown in Fig. i, and put a chick, say six days old, in at A,
it is confronted by a situation which is, briefly, ''the sense-
impression or feeling of the confining surfaces, an uncomfort-
able feeling due to the absence of other chicks and of food, and
perhaps the sense-impressions of the chirping of the chicks
outside." It reacts to this situation by running around, mak-
ing loud sounds, and jumping at the walls. When it jumps at
the walls, it has uncomfortable feelings of effort ; when it runs
to B, or C, or D, it has a continuation of the feelings of the sit-
uation just described ; when it runs to E, it gets out, feels the
pleasure of being with the other chicks, of the taste of food, of
being in its usual habitat. If from time to time you put it in

again, you find that it jumps and runs to By Cy and D less and
less often, until finally its only act is to run to Z>, Ey and out.
It has, to use technical psychological terms, formed an association
between the sense-impression or situation due to its presence at
A and the act of going to E. In common language it has learned
to go to E when put at A — has learned the way out. The
decrease in the useless runnings and jumping and standing still
finds a representative in the decreasing amount of time taken
by the chick to escape. The two chicks that formed this par-
ticular association, for example, averaged one about three and
the other about four minutes for their first five trials, but came
finally to escape invariably within five or six seconds.

The following schemes represent the animal's behavior (i)
during an early trial and (2) after the association has been fully
formed — after it has learned perfectly the way out.

THE ASSOCIATIVE PROCESSES IN ANIMALS.

71

Situation.

As described
above.

To chirp, etc.

To jump at various

places.
To ran to B.

u « « ^

a » u j)^

(I)

Acts.

Corresponding to
impulses.

(2)

Situation.
Same as (i).

Impulses.
To ran to E.

Acts.

Corresponding to
impulse.

Resulting Feelings.

Continuation of sit-
uation.
Fatigue.

Pleasure of company.
" " food.
" " surround-
ings.

Resulting Feelings.
Pleasurable as above.

A graphic representation of the progress from an early trial
to a trial after the association has been fully formed is given in
the following figures, in which the dotted lines represent the
path taken by a turtle in his fifth (Fig. 2) and fiftieth (Fig. 3)
experiences in learning the way from A to his nest. The
straight lines represent walls of boards. Besides the useless
traveling, there were, in the fifth trial, useless stoppings. The
time taken to reach the nest in the fifth trial was seven minutes ;
in the fiftieth, thirty-five seconds. The figures represent typ-
ical early and late trials, chosen from a number of experiments
on different individuals in different situations, carried on at
Woods HoU by Mr. R. M. Yerkes, of Harvard University, to
whom I am indebted for permission to use these figures.

Now the process of learning here consists of the selection,
from among a number, of a certain impulse and act in connec-
tion with a certain situation. And our first business is to dis-
cover the cause of that selection. The result of such discovery
was dogmatically stated in the previous lecture : " Any impulse
to an act which, in a given situation, leads to pleasurable feel-
ings, tends to be connected more firmly with that situation ;
and any impulse to an act which, in a given situation, leads to
discomfort, tends to become weakened in connection with that
situation." I say tends because the pleasurable feelings must
follow the act within certain limits of time — must be important

72

BIOLOGICAL LECTURES,

enough to outweigh possible instinctive opposition to the act,
or possible discomfort in its performance. Any one may dem-

FlG. 2.

Fig. 3.

onstrate to himself the truth of this law over a wide range of
animal life by observing the nature of the acts selected and
of those eliminated by experience. He will find that the com-

THE ASSOCIATIVE PROCESSES IN ANIMALS. 'j t^

mon element of the former is resultant satisfaction, and of the
latter the opposite. Later on it will be worth our while to
examine more carefully the way in which pleasurable and pain-
ful consequences respectively stamp in and stamp out the
impulses which lead to them. For the present we had better
return to actual specimens of animal behavior.

In the first place, the selection may be of one impulse and
act from only one. If the chick had happened to go to E the
first thing he did, the resulting satisfaction might have so
stamped in that impulse that on the second, third, and later
trials he would never have done otherwise. If accident or
instinct furnishes the right impulse, it will naturally be con-
firmed just as readily as if picked out by its success from any
number of inappropriate ones.

It will be well now to examine a more ambitious performance
than the mere discovery of the proper path by a chick. If we
take a box twenty by fifteen by twelve inches, replace its cover
and front side by bars an inch apart, and make in this front
side a door arranged so as to fall open when a wooden button
inside is turned from a vertical to a horizontal position, we shall
have means to observe such. A kitten, three to six months old,
if put in this box when hungry, a bit of fish being left outside,
reacts as follows ^ : It tries to squeeze through between the bars,
claws at the bars and at loose things in and out of the box,
reaches its paws out between the bars, and bites at its confin-
ing walls. Some one of all these promiscuous clawings, squeez-
ings, and bitings turns round the wooden button, and the kitten
gains freedom and food. By repeating the experience again
and again, the animal gradually comes to omit all the useless
clawings, etc., and to manifest only the particular impulse {e.g.,
to claw hard at the top of the button with the paw, or to push
against one side of it with the nose) which has resulted success-
fully. It turns the button round without delay whenever put
in the box. It has formed an association between the situation,
^'confinement in a box of a certain appearance," and the impulse
to the act of clawing at a certain part of that box in a certain

1 Confinement alone, apart from hunger, causes similar reactions, though not
so pronounced.

74 BIOLOGICAL LECTURES.

definite way. Popularly speaking, it has learned to open a door
by turning a button. To the uninitiated observer the behavior
of the six kittens that thus freed themselves from such a box
would seem wonderful and quite unlike their ordinary accom-
plishments of finding their way to their food, beds, etc., but the
reader will realize that the activity is of just the same sort as
that displayed by the chick in the pen. A certain situation
arouses, by virtue of accident or, more often, instinctive equip-
ment, certain impulses and corresponding acts. One of these
happens to be an act appropriate to secure freedom. It is
stamped in in connection with that situation. Here the act
is "clawing at a certain spot" instead of "running to ^," and
is selected from a far greater number of useless acts.

In the examples so far given there is a certain congruity
between the impulse associated with the situation and the result.
The act which lets the cat out is hit upon by the cat while try-
ing to get out, and is, so to speak, a likely means of release. But
there need be no such congruity between act and result. If we
confine a cat and open the door and let it out to get food only
when it scratches itself, we shall, after enough trials, find the
cat scratching itself the moment it is put into the box. Yet
in the first trials it did not scratch itself in order to get out, or
indeed until after it had given up the unavailing clawings, etc.,
and stopped to rest. The association is formed in the same
way with such an "unlikely" or incongruous impulse as that
to scratch, or lick, or, in the case of chicks, to peck at the wing
to dress it.

The examples chosen so far show the animal forming a sin-
gle association, but such may be combined into series. For
instance, a chick learns to get out of a pen by climbing up an
inclined plane. You then so arrange a second pen that the
chick can, say by walking up a slat and through a hole in the
wall, get from it into pen No. i. After a number of trials
the chick will, when put in pen No. 2, go at once to pen No. i
and thence out. You then arrange a third pen so that the
chick, by forming another association, can get from it to pen
No. 2, and so on. In such a series of associations the "act"
of one brings the animal into the "situation" of the next,

THE ASSOCIATIVE PROCESSES IN ANIMALS. 75

thus arousing its act, and so on to the end. Three chicks thus
learned to go through a sort of long labyrinth without mistakes,
the " learning " representing twenty-three associations.

The next matter to examine is the set of feelings in the
animal which we have called ''feelings of the situation," or,
more simply, "the situation." The important thing about
them is their vagueness, indefiniteness. The kitten did not
have in mind every bar of the box confining it, and feel the
impulse to turn the button in connection with such a precise
sense-impression. It felt the whole environment in an extremely
hazy way, and would still have turned the button if you had
planed off one-quarter of each bar, or painted the button, or
put the box in a different place, or sprinkled the bottom with
sawdust, though any of these acts would have vastly altered
the situation from our point of view. Cats tha»t climbed up a
certain screen when I whistled three times and said, " I must
feed those cats," would climb up just as surely if I whistled
only or spoke the words only, or even used any short sentence
with the same voice and manner. If I took the button off in
the box referred to above, the animals that had formed the
association would often claw as before, though the (to us) essen-
tial part of the situation was absent. Cats that had often
obtained food in consequence of being made to go into a certain
box, from which they liberated themselves by pulling down a
wire loop, would go in and pull the loop and then come out,
though no food was there and the door was open all the time.
All these facts witness to the vague, indefinite character of the
animal's feelings of any of these situations, and to the conse-
quent fact that much may be added to or subtracted from the
external situation without essentially modifying the animal's
reaction to it. Thus the kitten that in the first place responded
with the act of approach to the situation, " smell and sight of
milk in a certain dish," might later respond similarly to the
sight of the dish, though it was empty; thus the kitten that in
the first place responded to the situation, " sight or smell of
food plus hearing of certain sounds," might later respond
similarly to the sounds alone, might come, that is, at our call.
It is by virtue of this power of a single element in a situation

76 BIOLOGICAL LECTURES.

to set off the reaction associated with the vague total situation,
that horses are trained to stop at ''Whoa ! " and to turn at the
slightest pull of the rein, etc. It is in any particular case a
concrete problem how far you can change the external situation
and still secure the reaction associated with it. By such grad-
ual modification as we make use of in training domestic animals
you can of course change it without limit. For our present
purposes we need not bother with any instances of this problem,
remembering only that the animal does not react to a well-
defined, hard-and-fast feeling or notion of its surroundings, nor
with an immutable act, much less does it react to a well-defined
feeling of the essentials (for its purpose) in the situation, and
that therein lies the explanation of a host of animal activities.
So much for the feelings which are the first element in the
association. S9mewhat must now be said about the " impulses."
I have all along spoken of impulse and act, because for various
reasons out of place in this discussion I believe that there is
in one of these associations, besides the feeling of the situation
and the resulting act, a feeling preparatory to that act, such as
we ourselves feel in connection with any non-automatic muscular
performance. Such a feeling is meant by the word " impulse."
It does not, in my usage, mean "motive" or "desire," or even
the idea of the act to be performed, but merely the, in our own
case, rather evanescent feeling of doing. Whoever wishes may
discard this feeling from consideration, and substitute for
"impulse and act" the word "act" alone, and refer what is
said in the rest of this paragraph about the former to the latter.
In the cases so far considered, the causes of the impulses and
acts manifested at the first experience of a certain situation
have been, apart from accident, the instinctive tendencies of
the animals in question. That was because the particular
situations chosen (the pen for the chick, box for the kitten, etc.)
were situations for which the animals' previous experiences had
given them no preparation. But suppose us to take the kitten
that had formed an association between confinement in that
box and the act of clawing at a certain place and put it in
another box. We would witness less biting and squeezing and
more clawing than we did in our previous experiment. For its

THE ASSOCIATIVE PROCESSES IN ANIMALS. yj

previous experience would have strengthened the association
between clawing and confinement, or at least weakened it less
than it had weakened the impulses to bite and squeeze. That
is, in any situation the impulses and acts which appear are due
not only to the animal's instinctive tendencies, but also to any
modifications of them which have been wrought by associations
already formed. The question of how an animal will act in
any situation may thus require, for correct prophecy, knowledge
not only of its inborn constitution, but also of its entire previous
history.

Let us now return to the simple cases of animal behavior with
which we started and interpret them in accordance with the
facts we have found in our observation of the formation of
associations under test conditions. The cat's coming when we
call, " Kitty, kitty," we have already explained as a case of an
act associated with a certain total situation first (because, of
course, the act brought the pleasure of eating), then with one
element of that situation. The chick coming out of the brooder,
turning round a corner, and going straight to the dish of water,
leads back to a history somewhat like this :

Situation. Impulses. Consequent Acts.

Being in yard. To walk around and

Feeling thirsty. to peck at things, e.g.,

at stones,

seeds,

worms,

specks of all sorts, e.g.,

specks of dirt in the \

water of water dish, C These lead to the further

the edge of the dish. ) ^cts of drinking.

The last part of this piece of experience leads by repetition
to the formation of an association between the situation, (A)
** presence near dish when thirsty," and the act, "drinking."
But as soon as it is partly formed there is also being formed an
association between the situation, {B) " presence near brooder
when thirsty," and the act, ''walking around the corner"; for
whenever in that situation the chick by accident does stroll
around the corner, it comes to feel situation A, and so to be led
to the pleasurable drinking. The chick's conduct (and other

'jS BIOLOGICAL LECTURES.

conduct like it), then, is the result of several associations, the
act of one of which brings the animal into the situation of the
next. Of such serial associations the chick in the labyrinth
showed us a good example. The dog sneezes at our command,
because in teaching him we held his nose so as to compel him
to sneeze at the same time that we gave the command. The
act of sneezing thus became associated with a total situation,
including the pressure, etc., at the nostrils and the sound of
our command, and was later on committed when the latter part
only of the situation was present.

The behavior of animals under test conditions has, then,
shown us a simple associative process following a simple law
which seems c(5mpetent to explain ordinary cases of animal
learning. This does not imply that other processes, such as
imitation or inference or comparison, may not in some cases
mediate between a situation and the animal's reaction to it, that
no more complex process than that we have described ever
does occur, that this way of profiting by experience is the only
way for animals. Decision as to the presence or absence of
such processes is not within the province of this paper. Our
task so far has been to get a clear idea of what happens in a
considerable number of cases of animal learning, and to apply
our hypotheses to certain other cases than those from which
they were discovered. The task left us is to make some deeper
inquiries into the psychology of the process and its possible
neural counterpart, to investigate the extent to which such a
method of learning prevails down through the animal phylum,
the delicacy, complexity, number, and permanence of such asso-
ciations, and their importance in keeping the animal alive.

We have hitherto remained content with the evident fact
that pleasant consequences do stamp in the connection between
the impulse and act which led to them and the situation in which
the impulse was felt. Biologists in general, have been ready to
accept this fact. In human life it seems indubitable. Yet it
compels one to face a very difficult problem, familiar enough
to psychology, but seldom thought out by biologists. For
pleasurable feelings of taste, freedom, sexual activity, or what
not, 2irQ feelings ; are not facts of motion in space, are not

THE ASSOCIATIVE PROCESSES IN ANIMALS. 79

possessed of physical energy, are not amenable to the laws of
physical or chemical change. Yet this strengthening of the
connection between situation and act means some actual
change in the nervous system, some actual physical or chemical
readjustment, some actual phenomena of motion in space under
the laws of physical and chemical change. If, therefore, one
says simply, " The pleasurable feelings are themselves the cause
of the neural change which manifests itself in the strengthening
of the association," he is saying that a change in the physical
world takes place because of the presence of a certain non-
physical fact, and has either to admit the transference of physi-
cal energy into mental phenomena and back again, or to suppose
constant exceptions to the law of the conservation of energy.
In the former case we have the marvel of foot-pounds becoming
feelings of taste, etc., while in the latter case we have the greater
marvel of multitudinous uncaused feelings and multitudinous
motions in the nervous system uncaused save by these feelings.
A few psychologists, notably Prof. William James, do say so.
If, on the other hand, one says, " All consciousness, all feelings
are but a fifth wheel, paralleling certain neural facts, but unin-
fluenced by and not influencing them. The pleasurable feelings
do not cause anything. It is their neural counterpart that does
it," as most psychologists do say, he finds certain facts hard to
reconcile with his hypothesis. The old arguments /r^ and con
can be found in any elaborate psychological treatise and need
not be even mentioned here, especially since the question has
been all tangled up with speculations about the freedom of the
will, mental activity, voluntary attention, and all sorts of other
things. But the question as we meet it in a study of animal
behavior is as truly a question of physiological as psychological
biology, and certain more or less new and pertinent facts seem
worth mentioning. First of all, it is not the impulse and act
which are stamped in, but the impulse to that act in connection
zvitk that sit7iation. No cat goes around clawing, clawing,
clawing, because clawing at a button has freed it from a certain
box. No child keeps continually chewing because chewing
candy has brought pleasure. It is the connection, not the act,
that is strengthened. Secondly, the pleasurable feeling may

8o BIOLOGICAL LECTURES.

not be contemporaneous with the act, but may come considerably
later, and the act itself and impulse thereto may not be in the
least pleasurable. Our parallelist, therefore, who thinks that
the neural counterpart of the pleasurable feeling stamps in the
connection, has to show how the nervous activities aroused
by nervous currents from the taste buds, due to the presence of
fish in the mouth, can find their way around to the cells con-
cerned in the transmission of certain stimuli which came from
the retina and went out ten seconds ago to the end plates on the
leg muscles, and how, having safely arrived there, they can pick
out just the right cells and confirm in them just the right struc-
ture, so that when next those currents come from the retina
they will be directed in the particular way which shall make
the fore leg claw the button. He has, then, further to explain
how the neural actions arising from the taste of meat, the feel-
ing of company, sexual activity, etc., all find their way to the
same place and all do the same work. If he answers this by
saying that the pleasurable element of the feelings is in all
cases identical, and that therefore its neural counterpart is
identical and acts identically, he has the still harder problem of
explaining how any one neural activity can thus set its seal on
all sorts of nervous structure anywhere in the nervous system,
first convincing us also of the truth of the unlikely proposition
that the neural activities corresponding to the taste of fish,
sight of other chicks, and sensations from sexual contact have
one identical element. A speculative psychologist might at-
tempt all this, but where is the neurologist who would dare }

Yet the problem is not about fine-drawn distinctions, nor
a matter for metaphysical speculation, but is a clear question
about facts, and facts ever recurring in animal life, and so impor-
tant in animal economy that they surely ought not to be hud-
dled out of sight. The taste of the meat, or the neural action
corresponding thereto, does somehow strengthen the association ;
but for such influence the animal would react the hundredth
time as he did the first ; by such influence education by experi-
ence is possible. The answer, " Feelings of pleasure are them-
selves true causes," leaves us with a mystery on our hands.
The other answer, '< The neural counterparts of the pleasurable

THE ASSOCIATIVE PROCESSES IN ANIMALS. 8 1

feelings are the cause," seems to lead straight to neurological
monstrosities and myths, I turn over the problem to the
neurologists with my blessing.

Apart from this one enigma, the nervous action concerned
in the formation of these associations is extremely simple, and
is right in line with the nervous action concerned in instinctive
performances. Certain stimulation is caused in certain cells
by a certain external situation acting on the sense organs. By
virtue of inherited organization or acquired habits or accidental
influences, this stimulation is reflected into one or two or two-
hundred motor impulses. From these one becomes chosen in
the mysterious way which we have been discussing. In in-
stinctive activities natural selection has in advance prepared
the neural mechanism so that the stimulus is reflected from the
start into a limited field, a field which under natural conditions
leads to the appropriate act.

I ought, perhaps, to justify my use of the word *< associa-
tion" instead of " habit," especially as Principal C. Lloyd Morgan
has used the latter to denote the same resulting phenomena.
Neither word is a very good one, for both have in previous usage
meant something else than the process I have described. Asso-
ciation has meant the association of ideas, which, we shall see,
is a different process in important ways. I try to avoid con-
fusion from this source by calling our process '' animal associa-
tion." Habit has meant the strengthening of the tendency to
certain activities by mere repetition, as folds are made in the
sleeve of a coat. Now the associations, or habits if you will,
which we have been studying are formed in spite of habit in
this sense. The cat squeezes at a hole ten times where it
claws at the button once, yet the act of squeezing in later trials
becomes less frequent, and the act of clawing more frequent.
The very essence of the phenomena is that the animal does not
repeat its acts, but selects from among them in utter disregard
of the relative number of times each has been performed. The
word ''habit" can at best describe the result attained, while
the word "association" describes both the result and process,
and with no more danger of mistake.

It is now high time to consider critically the psychological

82 BIOLOGICAL LECTURES.

aspect of these associations, the feelings which the animal has at
the time, if he has any. All along I have begged this question,
have refused to entertain the supposition that the animal's
activities were purely mechanical, automatic, unconscious. I
ought to defend this procedure, though probably most of you
agree with me that animals do in such cases have consciousness,
and have in your own minds good reasons for so thinking.
One good reason is that we infer consciousness in the animal
for just the same reasons that we do in men, viz.^ they act as
we act when we have thoughts and feelings. Another good
reason is that we know of no mechanism capable of changing
its reactions to situations, as animals do in the formation of
associations, save the human body ; and it does have pleasurable
feelings controlling its activities in just the way we supposed
such feelings to do in animals. Without stopping to argue on
this question, let me proceed with those who agree that con-
sciousness of some sort is present, to ascertain of just what
sort. The word ** association" has been used by comparative
psychologists in the phrase "association of ideas," and animal
behavior of the kind we have been studying has been explained
by referring it to the association of ideas, a process which is in
a way the A-B-C of human psychology. I see before me a
dynamometer, think of Professor Cattell, its inventor ; of the
words ''New York," then of the words ''New Amsterdam," then
of the Dutch colonists and their multitudinous pantaloons, then
of the fitness of all this as an illustration, and end up by the
act of taking up my pen and writing. I hear a step, think, "It
is the postman," then, " I have a letter to post," and go to get
it. These are cases of the association of ideas, and the com-
parative psychologists would explain our cat's behavior by say-
ing, " When after a number of trials you put the cat in the
box, it thinks of the food outside, thinks of going out, thinks
of having clawed the button before going out, and so claws
at it." I believe that this explanation is totally wrong, that
the mental process involved is in no sense the association
of ideas of human psychology. It would crowd out more
strictly biological subject-matter if I introduced here the data
which lead to the rejection of this view, and all these data

THE ASSOCIATIVE PROCESSES IN ANIMALS. 83

are accessible to the skeptical in a monograph entitled " Ani-
mal Intelligence, an Experimental Study of the Associative
Processes in Animals," Supplement No. 8 to the Psycholog-
ical Review, Let it suffice to repeat here the assertion
therein justified, that in forming these associations the animals
do not think about the food, box, or clawing, at all, and do not
ordinarily connect idea with idea in an associative train of
thought ; that the association is between a directly felt situation
and a7i impulse to act. In the association of ideas, ideas are
the essential elements ; in the animal sort of associations the
^^ impulse and act'' is the essential. They are homologous, not
to our trains of thought, but to our feelings in learning to
swim, dive, play golf, etc. They are direct practical associa-
tions leading immediately to external acts. From them to
associations of ideas is a long step, longer and more important,
perhaps, than the step from the latter to logical thinking.

It is hard to answer our next question — *'To what extent is
this way of learning by the formation of associations prevalent
throughout the animal kingdom .'*" For students of animal life
have not had this process clearly in mind and have not collected
data with definite reference to it. It will be remembered that
illustrations of it in the case of mammals, birds, and reptiles
have already been given. Experiments on one of the bony
fishes, Fundulus majalis, have demonstrated its presence there.
These fish try to get out of the sunlight into a shady corner.
If, then, you arrange their aquarium so that one end is shaded,
the rest sunny, put them in the sunny end, with some mechan-
ical obstacle between them and the shady end, and watch their
actions, you will find that their manner of dealing with the
situation in their twentieth trial is different from that in their
first, in just the same way that the chick's in the pen or the
kitten's in the box was different. They engage in fewer use-
less acts, perform the successful act much sooner, and so, of
course, get back to the shade much more quickly. Let the
fish, for example, be in an oblong aquarium ten inches deep.
I put a wire screen, which fits across the aquarium, in behind
the fish, who is now at the shady end, and move it down
toward the sunny end. The fish swims ahead of it, turning

84 BIOLOGICAL LECTURES.

often, however, and trying to swim through the meshes of the
wire screen back to his habitual haunt. As soon as the fish is
at the sunny end, I place across the aquarium, between him and
the shady end, a piece of wire screening, impervious save at the
upper right-hand corner, where there is a hole 2x3 inches (my
aquarium was 4 feet long, 2 feet wide, the space in which the
fish was now confined being 2 feet long, 10 inches wide, and 10
inches deep), and remove the screen used to push him down to
the sunny end. The situation is, of course, '* sight of shady
end, feeling of the sunlight and of the walls around and screen-
ing in front." The animal reacts by remaining still, swimming
around, swimming particularly up and down the screening
which separates him from the shady end, poking his head
against it repeatedly in attempts to swim to the shady end.
The swimming is, for the most part, along the bottom of the
aquarium, but occasionally the fish rises slowly up and pokes
his head against the upper part of the screen. If he does this
at the right-hand end, he of course pokes through the hole
and gets to the shady end. Sooner or later he is practically
sure to do so. If, after giving him fifteen minutes' enjoyment
of his preferred habitat, you repeat the experiment, and so
again and again, the remaining still and swimming back and
forth and poking the head against the bottom of the screening
will gradually decrease, until finally the fish, when confronted
with the situation, will go directly to the right end, rise up,
and glide through the hole without more ado. Moreover, we
shall find, when pushing him down with the first screen, that
he no longer makes the useless attempts to swim through the
screen, which only hurt his head, but swims down to the sunny
end as soon as the screen is put behind him. Experiments
with different sorts of obstacles give the same results. Evi-
dently the presence of a cerebral cortex is not a prerequisite to
the formation of these associations.

How far down in the invertebrate series this process exists
remains a question for observation and experiment. Any ani-
mal which finds its way to feeding grounds or home apart from
direct response to sense-impressions would seem to do so by
virtue of associations thus formed.

THE ASSOCIATIVE PROCESSES IN ANIMALS. 85

The Peckhams' study of the solitary wasps convinced them
that these insects were gifted with associative powers as well
as instincts. If we begin at the extreme end, — with the Pro-
tozoa, that is, — we find diverse opinion's ; but the most recent
and by far the most thorough observer, Dr. H. S. Jennings,
finds absolutely no signs of any capability to modify reactions
in accord with the pleasurable or painful results they bring.
At the other end of the scale, in man, associative processes of
the animal sort are surely present, though in the adult they
are often overgrown and interpenetrated by the ideational life
to such an extent as to require, so to speak, careful dissection
to make them demonstrable. The human infant, however,
learns to cry to obtain certain favors, to go to certain places
for certain things, and, up to a certain age, to react to certain
words, in just the same way that the animals learn. And in
learning certain sorts of novel accomplishments, e.g., to play
billiards or tennis, the adult uses to a large extent this same
trial-and-error method, progressing because satisfaction stamps
in the particular movement or stroke which caused it and
makes the player, the next time the same situation appears,
more likely to make it. It may be that this is really the basal
fact in human intelligence, and that out of it are developed all
the higher' processes which are, in adult life, so ubiquitous that
the ordinary psychological treatise has nothing to say about this
simpler sort. Surely the infant, for about nine months, seems
to learn in no other way than the animals do. A completer
cornparative psychology will, let us hope, one day tell us, among
other things, when and how this method of learning arose
among animals, and how much of a part it has played in the
development of the human mind.

Delicacy, Complexity, Number, and Permarience of Associations.

Granted that an animal possesses this associative power,
question arises as to the degree of delicacy and complexity
which the associations may reach, and as to the number of
such associations the individual can acquire and the length
of time they last. By delicacy we mean the animal's power of

86 BIOLOGICAL LECTURES.

reacting differently to situations varying very slightly; for
instance, running toward you when you hold up two fingers
and away from you when you hold up three. I may quote
what I have said elsewhere on this subject to give an idea of
the means of estimating this power in any animal.

Delicacy of Associations.

" It goes without saying that the possible delicacy of asso-
ciations is conditioned by the delicacy of sense powers. If an
animal does n't feel differently at seeing two objects, it cannot
associate one with one reaction, the other with another. An
equally obvious factor is attention ; what is not attended to will
not be associated. Beyond this there is no a priori reason why
an animal should not react differently to things varying only
by the most delicate difference, and I am inclined to think an
animal could ; that any two objects with a difference appreciable
by sensation which are also able to win attention may be re-
acted to differently. Experiments to show this are very tedious,
and the practical question is, ' What will the animal naturally tend
to do .'' ' The difficulty, as all trainers say, is to get the animal's
attention to your signal somehow. Then he will in time surely
react differently, if you give him the chance, to a figure 7 on
the blackboard from the way he does to a figure 8 ; to your
question, * How many days are there in a week } ' and to your
question, ' How many legs have you t ' The chimpanzee in
London that handed out 3, 4, 5, 6, or 7 straws at command,
was not thereby proved of remarkable intelligence or of re-
markably delicate associative power. Any reputable animal
trainer would be ashamed to exhibit a horse who could not do
as much 'counting' as that. The maximum of delicacy in
associating exhibited by any animal, to my knowledge, is dis-
played in the performance of the dog * Dodgerfield,' exhibited
by a Mr. Davis, who brings from four cards, numbered i, 2, 3,
and 4, whichever one his master shall think of. That is, you
write out an arbitrary list, e.g.^ 4, 2, i, 3, 3, 2, 2, i, 4, 2, etc.,
and hand it to Mr. Davis, who looks at the list, thinks of the
first number, says, 'Attention, Dodger !' and then, 'Bring it.*

THE ASSOCIATIVE PROCESSES IN ANIMALS. ^'J

This the dog does, and so on through the list. Mr. Davis
makes no signals which any one sitting even right beside or in
front of him can detect. Thus the dog exceeds the human
observers in delicacy, and associates, each with a separate act,
four attitudes of his master which to human observers seem all
alike. Mr. Davis says he thinks the dog is a mind-reader. I
think it quite possible that whatever signs the dog goes by are
given unconsciously, and consist only of some very delicate
general differences in facial expression or the manner of saying
the words 'Bring it,' or slight sounds made by Mr. Davis in
thinking to himself the words one or two or three or four. Mr.
Davis keeps his eyes shut and his hands behind a newspaper.
The dog looks directly at his face.

" To such a height possible delicacy may attain, but possible
delicacy is quite another thing from actual untrained and un-
stimulated delicacy. The difference in reaction has to be
brought about by associating with pleasure the reaction to the
different sense-impression when it itself differs, and associating
with pain tendencies to confuse the reactions.^ The animal
does not naturally, as a function of sense powers, discrimi-
nate at all delicately. Thus the cat that climbed up the wire
netting when I said, * I must feed those cats ! ' did not have a
delicate association of just that act with just those words. For
it would react just as vigorously to the words, 'To-morrow
is Tuesday,' or, ' My name is Thorndike.' The reaction
naturally was to a very vague stimulus. Taking a cat just
beginning to learn to climb up at the signal, * I must feed
those cats ! ' I started in to improve the delicacy by oppos-
ing to this formula the formula, * I will not feed them,' after
saying which, I kept my word. That is, I gave sometimes the
former signal and fed the cat, sometimes the latter, and did not.
The object was to see how long the cat would be in learning
always to go up when I gave the first, never to do so when 1
gave the second signal. I said the words in both cases as I
naturally would do, so that there was a difference in emphasis
and tone as well as in the mere nature of the syllables. The

1 It is interesting to know that Mr. Davis says that fourteen months' time, four
hours a day, was expended on Dodgerfield's training.

88 BIOLOGICAL LECTURES.

two signals were given in all sorts of combinations, so that there
was no regularity in the recurrence of either which might aid
the animal. The cat at first did not always climb up at the
first signal, and often did climb up at the wrong one. The
change from this condition to one of perfect discrimination is
shown in the accompanying curves, one showing the decrease
m failures to respond to the right signal, the other showing the
decrease in responses to the wrong signal. The first curve

is formed by a line joining the
tops of perpendiculars erected at
intervals of i mm. along the ab-
p^^ scissa. The height of a perpen-

dicular represents the number
of times the cat failed to respond to the food signal in 20
trials, a height of i mm. being the representative of one failure.
Thus, the entire curve stands for 280 trials, there being no
failures after 60 trials, and only i after the 40th.

" In the other curve, also, each i mm. along the abscissa
stands for 20 trials, and the perpendiculars whose tops the
curve unites represent the number of times the cat in each 20
did climb up at the signal which meant no food. It will be
seen that 380 experiences were necessary before the animal
learned that the second signal was different from the first. The
experiment shows beautifully the animal method of acquisition.
If at any stage the animal could have isolated the two ideas of
the two sense-impressions, and felt them together in comparison,
this long and tedious process would have been unnecessary. "^

By a complex association I mean an association where either
situation or act (or both) is a compound. A man can learn
that to open a door he has to put the key in the hole, turn it,
turn the knob, and pull the door. Or, conversely, a man can
learn that when the sound of a bell is followed by a whistle,
and that by a blue light, a certain act is in place, while if any
one of the three comes alone it is not. In the first case, the
act is a complex of several ; in the second case, the situatio7i.

1 Animal Intelligence, an Experimental Study of the Associative Processes
in Animals, pp. 87-90.

THE ASSOCIATIVE PROCESSES IN ANIMALS. 89

Between such complexity and the serial complexity of the asso-
ciation formed by the chick in the labyrinth there is a vast dif-
ference. In the latter case situation A causes act B, which
brings the animal into situation Cy which causes act Z>, etc.
This is really a matter of a mmtber of simple associations.

Associations that involve complexity in the first sense are
very hard for animals to form. If, for instance, you arrange
a box so that a cat or dog to escape has to step on a platform,
push down a bar, and claw down a string before the door will
open, you find that even if the animal is familiar with each of
these acts by itself, he will be slow to form the habit of doing
the three in combination ; will even, after fifty or more trials,
not have the association perfect ; will frequently step on the
platform or claw the string again and again, and will not per-
form the several acts in any fixed order. The animal does not
feel "doing this, doing that, doing the other, getting out," but
simply feels, more or less confusedly, intermittently, and in
various orders, the three impulses needed.

As regards the number of associations which animals are
able to form, I may best quote, with trifling changes, a para-
graph or two from the monograph already mentioned :

" The patent and important fact is that there are so few in
animals compared to the human stock. Even after taking into
account the various acts associated with various smells, and
exaggerating the possibility of getting an equipment of associa-
tions in this field which man lacks, one must recognize how far
below man any animal is in respect to mere quantity of associa-
tions. The associations with words alone of an average Amer-
ican child of ten years far outnumber those of any dog. A
good billiard player probably has more associations in connec-
tion with this single pastime than a dog with its whole life's
business. In the associations which are homologous with those
of animals man outdoes them and adds an infinity of associa-
tions of a different sort. . . .

" Small as it is, however, the number of associations which
an animal may acquire is probably much larger than popularly
supposed.

"■ My cats and dogs did not mix up their acts with the wrong

90 BIOLOGICAL LECTURES.

sense-impressions. The chicks that learned the series of twenty-
three associations did not find it a task beyond their powers to
retain them. Several three-days-old chicks, which I caused to
learn ten simple associations in the same day, kept the things
apart and on the next morning went through each act at the
proper stimulus. In the hands of animal trainers some animals
get a large number of associations perfectly in hand. The
horse Mascot is claimed to know the meaning of fifteen hun-
dred signals ! He certainly knows a great many, and such as
are naturally difficult of acquisition. It would be an enlight-
ening investigation if some one could find out just how many
associations a cat or dog could form if he were carefully and
constantly given an opportunity. The result would probably
show that the number was limited only by the amount of motive
available and the time taken to acquire each. For there is
probably nothing in their brain structure which limits the
number of connections that can be formed, or would cause
such connections, as they grew numerous, to become confused.

" In their anxiety to credit animals with human powers, the
psychologists have disregarded or belittled, perhaps, the possi-
bilities of the strictly animal sort of association. They would
think it more wonderful that a horse should respond differently
to a lot of different numbers on the blackboard than that he
should infer a consequence from premises. But if it be made
a direct question of pleasure or pain to an animal, he can asso-
ciate any number of acts with different stimuli. Only he does
not form any association until he has to, until the direct benefit
is apparent, and, for his ordinary life, comparatively few are
needed." ^

The very fact of the formation of associations is evidence
that the connection between situation and impulse and act is
permanent. If the influence of the first few successes did not
remain to work on the next, there would be no association
formed at all. But such permanence is even more clearly wit-
nessed by the behavior of animals who are put in situations
with which they have in times past associated certain acts, but

1 Animal Intelligence, an Experimental Study of the Associative Processes
in Animals, pp. 93-95-

THE ASSOCIATIVE PROCESSES IN ANIMALS. 91

which they have not encountered for a long while. Dogs, cats,
chicks, and fishes tested after an interval of from six to eighty
days, either manifested as perfect associations as ever, or formed
such after far fewer experiences than were needed originally.^
For instance, the cat that had learned to discriminate between
" I must feed those cats " and '' I will not feed them " was, after
eighty days, given twenty-five trials with the first signal and
fifty with the other (all being mingled together indiscrimi-
nately). She reacted correctly to every one of the twenty-five
signals and by the end of the fifty reacted to that signal as well
as she did after 350 trials originally.

The degree of permanence of these associations, once they
are formed, vastly increases their utility. It allows experiences
rarely met with to still, little by little, build up habits. It allows
the experience of certain localities, for instance, to be useful
again, even if the locality is not revisited for a considerable
time. It lets the animal's past influence its present conduct
in all sorts of ways.

It seems hardly necessary to make any statement about the
general usefulness of the power of association in securing sur-
vival. If our view of the process is correct, it is a process of
selection among reactions, not by eliminating the animal that
does not react suitably and so developing a stock with certain
instincts, but by eliminating the unsuitable reaction directly by
discomfort, and also by positively selecting the suitable one by
pleasure, and so developing certain associations in the individ-
ual. It is, then, selection within the individual that is the great
case of plasticity, and is of tremendous usefulness, in that it
definitely enables the animal to modify his acts and so meet
varieties and modifications of environment. New feeding
grounds, new foods, new friends, new enemies are dealt with
by virtue of it. ** He who learns and runs away, will live to
learn another day."

Teachers' College, Columbia University.

1 Only one case out of over forty was an exception to this rule.

SEVENTH LECTURE.

THE BEHAVIOR OF UNICELLULAR ORGANISMS.

HERBERT S. JENNINGS.

In recent biological writings there is manifest a growing
tendency to interpret the processes taking place within the
bodies of higher animals — especially the developmental proc-
esses — as a series of responses to stimuli. In the egg and
the developing embryo, masses of protoplasm migrate from one
position to another, cells and cell masses alter in form, changes
of the most varied character are continually occurring. To
explain such changes it is becoming usual to call upon chemo-
taxis, geotaxis, phototaxis, thigmotaxis, and other motor reac-
tions of similar character. The prevalent ideas of these
reactions, known usually under names terminating in -taxis
or -tropism, have been derived to a large extent from the phe-
nomena shown by the movements of unicellular organisms ; the
classic experiments of Pfeffer on the chemotaxis of bacteria and
flagellates and of Strasburger on the phototaxis of swarm spores
having opened a fountain from which all have felt entitled to
draw. To understand the migration of a cell or mass of cells
in the embryo we are referred back to experiments on unicellu-
lar organisms, wherein it is shown that the movements of the
latter are controlled by chemical agents, by heat, by light, and
the like. Here the vital processes are seemingly brought into
the closest relation with chemical and physical ones ; chemo-
taxis, for example, is frequently interpreted as the direct expres-
sion of chemical affinity or chemical repulsion between the
substance of the protoplasmic mass and some other substance,
or between two protoplasmic masses. There is thus established

93

94 BIOLOGICAL LECTURES.

an immediate direct relation between the movements of organ-
isms and movements characteristic of inorganic substances ;
a long step is taken toward that analysis of vital processes into
simple chemical and physical ones, which is deemed by many
the final goal of biological science. If these phenomena do
indeed establish such a relation, they challenge the attention of
every man interested in the fundamental phenomena of life ;
in any case, they invite complete and thorough investigation
of the claims made for them. Since it is largely from the
reactions of free unicellular organisms that our ideas of chemo-
taxis, phototaxis, and the like have been derived, it is impor-
tant to study carefully the reactions of these creatures and to
determine the laws which control them. We shall then be in
a position to decide whether the movements of these organisms
do furnish a key to the understanding of ontogenetic processes
or not. It is these considerations that have impelled the inves-
tigation whose main results I shall try to present.

In studying the behavior of single-celled creatures we are
forced into relation with the much debated question of the
nature and importance of the activities of unicellular organ-
isms as compared with those of higher animals and plants.
Some hold that the cellular standpoint is the fruitful one for
general physiology ; that we must first determine the laws of
action for single cells, then carry these over to the cell state,
understanding the latter only as a combination of the former.
Some go so far as to maintain that the reactions of unicellular
organisms are of an intrinsically different character from those
of higher forms, being of essentially the same nature as the
reactions of inorganic bodies ; this is, for example, the posi-
tion of Le Dantec.^ Others hold that the division of organisms
into cells is, physiologically at least, a secondary matter ; that
nothing more fundamental is to be expected from the study of
a unicellular organism than from that of one composed of many
cells. This question can be decided, of course, only by a
thorough study of both the classes of organisms thus con-
trasted, with a comparison of the results, to see if the study
of the simpler organisms does, as a matter of fact, clear up

1 La mati^re vivante, Chapters I and II.

BEHAVIOR OF UNICELLULAR ORGANISMS. 95

and simplify the phenomena exhibited by the many-celled crea-
tures. For one desirous only of getting at the real laws under-
lying the phenomena, the conflict on such points between high
authorities ^ is very confusing, and the only recourse is to a
first-hand study of the' facts.

In the hope of getting light on the problems proposed and
others of similar character, I shall set forth and discuss obser-
vations and experiments made upon a number of free-swimming
unicellular organisms. In the investigation it was found well
to begin with some single species and work out its activities,
and the laws governing the same, completely enough to reveal
their essential nature, then to make a comparative study of the
activities of other organisms in the light of the knowledge so
gained. The same method will be advantageous for the pres-
entation of results.

I give first, therefore, some of the results of a preliminary
study of the activities of Paramecium caudatum. This is one
of the commonest of the ciliate Infusoria, living by thousands
in vegetable matter decaying in water. It is a somewhat cigar-
shaped creature, having a broad groove passing obliquely from
one end (the anterior) to the mouth, which lies at about the
middle of the length of the body. The side on which the
mouth and groove lie may be called the oral side ; the opposite
one the aboral side. The entire surface of the animal is covered
with cilia, by means of which Paramecium moves.

In beginning a study of the activities of such an organism,
we are at once confronted with the question of its psychic
powers. If these unicellular organisms do, as a matter of fact,
possess so complicated and highly developed a psychic life as
Binet, in his book on the Psychic Life of Micro-Organisms^ has
attempted to show obtains among them, then indeed there is
little prospect of gaining light on simple migrations of proto-
plasmic masses during development, through a study of their
behavior. A study of the chick or the dog would perhaps be
as promising. The activities which Paramecium shows are at

1 See, for example, Verwom, " General Physiology," and Loeb, " Einige
Bemerkungen liber den Begriff, die Geschichte und Literatur der allgemeinen
Physiologic," Pfluger''s Archiv, Bd. Ixix, p. 249.

96 BIOLOGICAL LECTURES.

first view of great complexity, so that they might seem to
entirely justify Binet's views as to the height and variety of the
psychic powers of these organisms. These activities and their
explanation have been discussed somewhat fully by the writer
in a paper ^ devoted entirely to the psychological aspect of the
matter, so that only so much of this aspect will be taken up at
present as has a necessary relation to the questions proposed.

If we place a number of Paramecia, in the culture water in
which they are found, upon a glass slide, and cover with the
cover glass, we soon find that the animals, which were at first
scattered uniformly, have gathered into groups in one or more
parts of the preparation. Usually we find that a bit of bacte-
rial zoogloea forms the center of such a group ; as many of the
Paramecia as can do so have pressed their anterior ends against
the mass, the ciliary current carrying bacteria to their mouths ;
others press in from behind. It is well known, of course, that
Paramecia make no choice in the food which the current brings
to their mouths, taking in particles of all sorts indiscriminately.
The possibility may suggest itself, however, that they have
gathered about these masses of zoogloea because the latter
serve them as food. The choice of food would thus occur a
step sooner — the Paramecia choosing their food by gathering
about it, then taking whatever comes. But if we introduce into
the slide a bit of filter paper or a fine raveling of cloth, we find
that the Paramecia gather about it with the same apparent
avidity as about the zoogloea, pressing the anterior end against
it and remaining thus, quiet, for long periods.

This and other experiments show, therefore, that this gather-
ing about a bit of bacterial zoogloea or other substance is not the
expression of a choice of food, but is merely a manifestation
of the fact that the Paramecia react to contact with solids of a
certain physical texture by suspending active locomotion and
remaining against the solid. A similar reaction to solids is, of
course, a very common phenomenon among organisms of dif-
ferent sorts; it has received the name ^'Thigmotaxis," or
** Stereotropism."

1 " The Psychology of a Protozoan," Anier. Journ. of Psychology^ vol. x,
No. 4, 1899.

BEHAVIOR OF UNICELLULAR ORGANISMS. 97

We have in thigmotaxis one of the fundamental reactions of
Paramecium, not further analyzable into simpler component reac-
tions. As it seems to consist chiefly or entirely of a cessation
of a part of the usual ciliary motion, — only the cilia in the
oral groove continuing to strike strongly backward, — it may
be more philosophical to consider this partly resting condition
as the "normal" condition, the usual forward motion being
then considered a reaction to a stimulus, due to a change or
removal of the solid body against which the animal is resting,
or to some other change in the environment. There seems to
be no decisive reason for considering either the condition of
partial rest or of the usual forward motion as more "normal"
than the alternative condition ; taking either as a starting
point, the other may be considered a response to a stimulus.

If the Paramecia are placed upon the slide in pure water,
containing no bacterial zoogloea, or any other solid, they do
not even then remain scattered uniformly throughout the prep-
aration. On the contrary, it is usually not long before the
animals are gathered into one or more close groups in some
part of the slide. Paramecia are usually found in the culture
jars also aggregated into groups ; this, taken together with the
above experimental demonstration that Paramecia, at first uni-
formly scattered, will soon collect into close groups without
evident external cause, might be held to indicate the existence
of a " social instinct " among these creatures. Another possi-
bility suggests itself — that there may be some invisible chem-
ical substance in the region of these groups by which all the
Paramecia are attracted ; so that the fact that they come near
together would be a secondary result of the fact that all are
attracted by the same substance.

The main results of the extended study of the conduct of the
Paramecia toward chemicals, to which this possibility led, may
be given in a few words. It was found that Paramecia tend to
gather together and form collections in drops of weakly acid
solutions, and in solutions of some salts, while they avoid alka-
line solutions and solutions of the salts of the alkali metals.

Among the substances into solutions of which they gather
is carbon dioxide. If a bubble of carbon dioxide is introduced

98 BIOLOGICAL LECTURES.

into a preparation of Paramecia, they soon collect closely about
it and swim in circles around it without leaving it.

It was then proved, by introducing the Paramecia into a
solution of rosol, which is decolorized by carbon dioxide, that
these Infusoria excrete a distinctly appreciable amount of this
substance, which diffuses into the surrounding water. When-
ever, therefore, a very few Paramecia get together, an active
solution of carbon dioxide is soon formed, and the region be-
comes at once a center of attraction for the Paramecia. A
most complete correspondence was demonstrated between the
diffusion of the COg into the water and the distribution of the
Paramecia in groups, and all the phenomena exhibited by
the (apparently) spontaneous collections of Paramecia could
be exactly imitated by introducing COg into the slide.

Thus these collections of Paramecia give no indication of
'/social instinct," but are merely the expression of positive
chemotaxis on the part of the animals toward a certain sub-
stance. In the same way all the seemingly complex activities of
these creatures may be reduced to simple factors, so that there
seems no evidence to indicate the possession by them of psychic
powers of anything more than the most elementary character.

We may proceed then to a closer analysis of the apparent
attractions and repulsions — chemotaxis, thermotaxis, and the
like ; it is from a study of these that light is to be gained on
the problems first proposed. We shall first consider chemo-
taxis.

The fact that animals and plants are attracted by certain
chemical substances and repelled by others is of course well
known for a large number of organisms. As to the essential
nature of this phenomenon, opinions differ. As pointed out
above, some hold that chemotaxis is the direct expression of
chemical affinity or repulsion between the living protoplasm
and the chemical. Le Dantec {La matihe vivante, pp. 51, 52)
gives geometrical figures illustrating the action between the
surface of a free cell and a chemical substance diffusing in the
surrounding water, demonstrating in mathematical form that
as a result of this action the cell must move either toward or
away from the center of diffusion of the chemical. The motion

BEHAVIOR OF UNICELLULAR ORGANISMS. 99

of such a protoplasmic body would be passive in the same sense
as the movements of iron filings are passive when acted upon
by a magnet. Delage and Herouard^ actually state that the
Flagellata have, in addition to their usual active movements,
also a passive motion, due to the attraction of chemical sub-
stances. Perhaps the majority of biologists hold less radical
views than this ; yet the opinion seems widespread that in
chemotaxis we are dealing with a simple primary phenomenon.
Coming now to an examination of the phenomena as exhib-
ited by Paramecium, we will first take up positive chemotaxis,
or attraction toward chemical substances. The phenomenon
to be explained shows itself as follows. If into a slide of Para-
mecia a drop of some attractive substance (as a weak acid) is
introduced, the Infusoria soon collect in the drop, forming there
a dense assemblage. Now, what is the exact action of the
attractive substance on the Paramecia to cause them to turn
and enter the drop } Observing carefully the conduct of the
animals, we find, first, that they do not turn toward the drop.
Owing to its slow diffusion, the margin of a drop thus intro-
duced beneath the cover glass is evident, and the Paramecia,
swimming in every direction throughout the preparation, may
be seen in their random course to graze almost the edge of the
drop, without their motion being changed in the least ; they
keep on straight past the drop and swim to another part of the
slide. But of course some of the Paramecia in their random
swimming come directly against the edge of the drop. These
do not change their motion, but keep on undisturbed across the
drop. But when they come to the opposite margin, where they
would if unchecked pass out again into the surrounding me-
dium, a marked reaction is caused ; the Paramecium jerks back
and turns again into the drop. Such an animal then swims
across the drop in the new direction till it again comes to the
margin, when it reacts negatively, as before. This continues,
so that the animal appears as if caught in the drop as in a trap.
Other Paramecia enter the drop in the same way and are im-
prisoned like the first, so that in time the drop swarms with the
animals. As a result of their swift, random movements when

1 Traite de zoologie concrete, tome i, p. 305.

lOO BIOLOGICAL LECTURES.

first brought upon the slide, almost every individual in the prep-
aration will in a short time have come by chance against the
edge of the drop, will have entered and remained, so that soon
all the Paramecia in the preparation are in the drop.

Thus it appears that the animals are not attracted by the
fluid in the drop ; they enter it by chance, without reaction,
then are repelled by the surrounding fluid. The peculiar fact
that the animals, after entering the drop of the substance in
question, are repelled by the surrounding fluid in which they
were previously immersed will become more comprehensible
after the phenomena of repulsion are considered.

Turning, then, to the matter of negative chemotaxis or
repulsion, we have the following phenomenon to be explained.
If into a slide of Paramecia swimming at random a drop of some
repellent chemical (as NaCl) is introduced, we find that the
drop remains entirely empty, not a single Paramecium entering
it. Now, exactly how do the Paramecia succeed in keeping out
of such a repellent solution }

Careful observation shows that when the Paramecium, swim-
ming forward, comes in contact with the drop of repellent sub-
stance, it swims backward a short distance, then turns toivard
its own aboral side, then swims forward again. The essential
point in this reaction method is, that the Paramecia always turn
toward their own aboral side, without regard to the position of
the stimulating drop. If a Paramecium comes obliquely in
contact with the drop so as to touch it only on one side of its
body, it nevertheless gives the reaction above described with-
out modification, even though turning toward its own aboral
side after backing off may carry the animal directly toward the
drop, instead of away from it. In such a case the animal when
it comes again in contact with the drop simply repeats the
reaction. As it continually revolves on its long axis both when
swimming forward and when swimming backward, the aboral
side is nearly certain to lie in a new position the second time,
so that the animal turns in a new direction. If this is repeated
a sufficient number of times, the Paramecium is fairly certain,
by the laws of chance, to get started finally in a direction which
carries it away from the stimulating chemical.

BEHAVIOR OF UNICELLULAR ORGANISMS. lOI

It thus appears that the direction in which a Paramecium
turns after stimulation by a chemical substance is not deter-
mined by the position of the stimulating agent, nor indeed by
any external factor, but by an internal factor, — by structural
differentiations of the animal's body. This is demonstrated in
a striking manner by immersing the Paramecia directly into a
chemical solution of such a nature as to act as a stimulus.
The entire surface of the animal is then bathed by the chemi-
cal, so that there is nothing in the external conditions to deter-
mine in which direction the animal shall move. Nevertheless,
under these circumstances, it swims backward, turns toward
the aboral side, and swims forward, usually repeating the
operation indefinitely. Very striking is also the experiment of
causing the chemical to act first upon the posterior end of the
animals. This may be done as follows : A large number of
Paramecia are frequently observed with anterior ends pressed
against the surface of a bit of bacterial zoogloea (thigmotactic
reaction), so that the posterior ends are all pointed in the same
direction. Now, a capillary glass rod, coated with some chemi-
cal, is introduced into the water behind the Paramecia. The
chemical gradually diffuses through the water, of course first
reaching the posterior ends of the Paramecia. But these, when
they respond, react exactly as in the other cases ; they swim
backward some distance and turn toward the aboral side. It
often occurs that in thus swimming backward they enter the
densest part of the chemical and are killed by it.

These experiments indicate that not only the direction of
turning after swimming backward, but also the swimming back-
ward itself is determined by internal factors, and is independent
of the position of the source of stimulus. This conclusion
seems strictly true for chemical stimuli both in Paramecium
and in other Infusoria experimented with. As will be shown
later, other experiments throw a light upon the cause of this
uniform backward motion when stimulated by a chemical.

•Summing up, then, we may say that when Paramecium is
chemically stimulated it swims backward, turns toward its own
aboral side, then swims forward. As a rule, the anterior end,
moving forward, comes first in contact with the chemical, so

I02 BIOLOGICAL LECTURES.

that swimming backward does, as a matter of fact, usually carry
the animal away from the source of diffusion of the chemical, and
turning toward the aboral side before swimming forward again
will, as a rule, if repeated, finally carry the animal in such a direc-
tion that it does not again come against the source of stimulus.
But these are, from the physiological standpoint, matters of
accident ; the animal conducts itself in the same way whether
the source of stimulus has this usual position at the anterior
end of the animal or not. The direction of motion after a
chemical stimulus, then, has no relation to the position of the
chemical substance. We cannot say, therefore, that the Para-
mecia are repelled by any chemical substance — just as we
were compelled to conclude that they are not directly attracted
by any chemical substance.

We find, then, that the effect of chemicals on Paramecia is
not to attract or repel them, but simply to cause a certain set
formula of movements. Such a set formula of movements,
"touched off," as it were, by stimuli of various sorts, may be
called a reflex. In returning now to the question of how the
apparent attraction of the Paramecia toward certain substances
. — that is, the fact that they collect in drops of certain sub-
stances— can be brought about through such a reflex, it is neces-
sary to recall certain general facts in regard to the nature of
reflexes. First, any change in the environment that can be
perceived by the organism may '* touch off" such a reflex.
Second, the character of the reflex has no necessary relation to
the nature of this external change, so that of a given kind of
change it cannot be predicted beforehand whether it will cause
the reflex or not, and changes of opposite character may pro-
duce the same reflex.

The mechanism of the gathering together of the Paramecia
into a drop of some weak acid is then as follows : When the
Paramecium passes from the surrounding fluid into the acid
solution there is, of course, at the moment of crossing the
boundary of the drop a change in its environment. Whether
this change will cause the characteristic reflex or not is impos-
sible to predict, that depending upon the internal mechanism
of the organism ; as a matter of fact we find that it does not

BEHAVIOR OF UNICELLULAR ORGANISMS. 103

cause the reflex. Now, after passing across the drop it comes
again to the boundary where, if not stopped, it would pass out
again into the surrounding fluid. At this boundary there is, of
course, another change in the environment — a change in the"
opposite sense from that experienced in passing into the drop.
Whether this second change will cause the reflex is of course
likewise impossible to predict, since it depends upon the nature
of the organism ; as a matter of fact we find that it does cause
the reflex. The Paramecium is, therefore, returned into the
drop and kept there in the manner already described. It
seems probable that the physiological condition of the Para-
mecium is changed by immersion in the drop of acid, so that
contact with the culture fluid now acts as a stimulus, though it
before did not. It seems not impossible to conceive, however,
that even without such a change in physiological condition, an
environmental change from b X.o a might cause a reaction, when
the opposite change, from a to b^ would cause none. This has,
as is evident from the nature of a reflex, no necessary relation
to the comparative actual mechanical difficulty in passing in
one direction or the other.

The one effect of a marked chemical stimulus on Paramecium
is, then, to produce the characteristic reflex already described,
and the apparent attraction or repulsion is determined by the
fact that some chemical substances or chemical changes cause
the reaction, while others do not.

Now, experimentation with stimuli other than chemical leads
to the highly important observation that this same reflex is
produced by stimuli of the most varied nature. Substances
which seem to act upon Paramecium only through their osmotic
pressure, such as solutions of sugar, cause the same reflex ;
tonotaxis, then (to use the name employed by Massart), acts
through the same reflex as does chemotaxis. Mechanical stim-
uli, produced by jarring the preparation, cause the same reflex.
Heat and cold act in the same way, and the Paramecia avoid
hot or cold areas and collect in regions of optimum temperature
in exactly the same manner as they avoid certain chemicals and
collect in others.

We are driven, therefore, to the conclusion that chemotaxis

I04 BIOLOGICAL LECTURES.

is not an activity differing in kind from the other reactions of
these animals. Many sorts of changes in the environment
produce a certain characteristic reflex in Paramecia, resulting
in their collecting in regions of certain characters and leaving
other regions vacant. Among the changes that act thus are
chemical changes, and the characteristic groupings of the ani-
mals so caused are said to be due to chemotaxis ; they are,
however, produced in an essentially similar manner to the
groupings produced by other agents. There is a unity under-
lying the motor activities of the Paramecia — a unity expressed
in the fact that the different classes of stimuli produce identi-
cally the same reaction.

To be accurate, however, we must distinguish two less
important forms of reaction to stimuli that are not manifested
through the characteristic reflex above described. One is
thigmotaxis ; this is, however, not a motor reaction, but one
characterized chiefly or entirely by a cessation of a part of the
usual motion. Again, as previously set forth, it is possible to
consider the partially resting condition characteristic of thig-
motaxis as the primary condition ; then the ordinary forward
motion of the animal will be a motor reaction to a stimulus,
since it is induced by a change in the environment. As will
be shown, there is sufficient ground in certain other Infusoria to
compel us to consider this forward motion as at times a reaction
to stimulus ; this, then, is a motor reaction which does not take
place through the above-described characteristic reflex. It seems
possible that the following represents the true state of the case ;
very weak stimuli acting on the resting individual cause the
ordinary forward motion ; stronger stimuli produce the above-
described motor reflex.

In view of the means by which chemotaxis is brought about,
it becomes more intelligible why the Infusoria may at times
collect in regions of injurious substances and avoid at times
areas of harmless substances. It is not a matter of attraction or
repulsion at all. In the former case the injurious substance
merely does not act as a stimulus to cause the motor reflex ; in
the second case, the chemical in question, though not injurious,
does act as a stimulus. An extended investigation directed

BEHAVIOR OF UNICELLULAR ORGANISMS.

105

upon this point showed that the chief factor determining
whether a substance does or does not cause the motor reflex
of Paramecium is not the injuriousness of the substance, but is
of a chemical nature.

We are now prepared to sum up the main results on Para-
mecium. In this animal we find that chemotaxis, thermotaxis,
tonotaxis, reactions to mechanical shock, and the like, are not
distinct kinds of activity ; that in each case we have the same
movements, merely induced by different agents. When Para-
mecium is effectively stimulated by any substances acting
chemically or through osmosis, by heat or by cold or by
mechanical shock, it responds with a reflex, which consists of
the following activities : the animal swims backward, turns
toward its own aboral side, then swims forward. The result of
this method of reaction is that the Paramecia tend to leave
the sphere of influence of agents causing this reflex, and to
congregate in areas where this reaction is not caused. For
chemical substances at least it is proved that the position of
the stimulating agent has no influence on the direction of move-
ment after a stimulus ; the direction of movement throughout
the reaction is determined by internal factors.

Is this reaction method one that is common among unicellu-
lar organisms, or is it peculiar to Paramecium } To answer
this question I have studied the reactions of a considerable
number of unicellular organisms belonging to the Flagellata
and Ciliata. The essential point in the reaction of Paramecium,
the factor that gives character to the entire response, is the
circumstance that the animal after stimulation turns toward one
side which is structurally defined, without regard to the nature
and position of the source of stimulus. The point to which
attention was primarily directed in studying the other organ-
isms was, therefore, whether after stimulation the creature
turned always toward one structurally defined side.

The organisms studied included, among the flagellates : Chi-
lomonas Paramecium and Eiiglena viridis ; in the ciliates the
following Holotricha : Paramecmm can datum, Loxophy litem mele-
agris, Colpidhim colpoda, Microthorax sulcatus, Dileptus anser,
Loxodes rostrum, and a species of Prorodon ; the following Hete-

I06 BIOLOGICAL LECTURES.

rotricha : Stentor polymorphus^ Spirostomum ambigiium^ and
Biirsaria truncatella ; of Hypotricha, Oxytricha fallax and a
number of undetermined species. In several of these creatures,
on account of the large size or other favorable circumstances,
it was possible to use methods of investigation not available for
Paramecium ; in particular it was possible in a number of cases
to localize very precisely the action of stimuli.

In all of the organisms named, in spite of great variations in
the nature and complexity of the usual movements, the reac-
tion method was essentially similar to that of Paramecium. In
all, the direction of turning after a stimulus was toward a struc-
turally defined side, without regard to the nature and position
of the source of stimulus. With regard to the details of the
reaction, as might be expected, the greatest variety exists, but
the general reaction plan was the same throughout..

This method of reaction evidently has a close relation to the
usual asymmetry of the cell body exhibited by these organisms.'
This asymmetry of the Infusoria has also a close relation to the
normal method of progression through the water, as well as to
the method of reaction to a stimulus. Most of these organisms,
as they swim forward, also revolve on the long axis, and the
resulting path is usually a spiral. The form of the body has a
constant relation to the axis of the spiral, the same side being
at all times directed toward this axis. The unsymmetrical
structure of the body, the usual method of progression, and the
method of reaction to a stimulus are thus evidently closely
interrelated. In the case of a bilaterally or radially symmetri-
cal animal one would certainly not expect that one side would
be always preferred to the other in turning away from a source
of stimulus, as is the case in the Infusoria.

In the case of chemical stimuli it was found for all the organ-
isms studied that not only the turning to one side, but the
swimming backward after a stimulus, was independent of the
position of the source of diffusion of the chemical. The action
of chemical stimuli was localized by bringing a capillary glass
rod coated with some chemical compound near the anterior
end, one side, or the posterior end, of the quiet organisms. In
every case (except in Euglena viridis, which does not swim

BEHAVIOR OF UNICELLULAR ORGANISMS. 107

backward under any circumstances) the organisms reacted to
the chemical stimulus by swimming backward, turning toward
the usual structurally defined side, then swimming forward.
The swimming backward, of course, sometimes carried the
creature away from the densest part of the solution (when the
chemical was held in front); at other times, directly toward
and into the densest part (when the same chemical was held
behind). In the latter case the organisms were frequently
killed by swimming into the dense solution. Thus, in chemical
stimuli, without exception, the direction of motion after stimu-
lation has no relation to the localization of the stimulus.

In several of the organisms it was possible to use also very
precisely localized mechanical stimuli ; and the results so gained
tend to modify in some particulars the general conclusions that
might be drawn from a study of the action of localized chemical
stimuli. Localized mechanical stimuli were produced by touch-
ing under a powerful lens any desired part of the body of the
organism with a glass rod drawn to the finest hair in a flame.
For Paramecium itself this method of experimentation was not
satisfactory, owing to the minute size of the cell body. One
point of importance was brought out in Paramecium, however.
The anterior tip of the body was shown, to be incomparably
more sensitive than any other part. On bringing the glass hair
near the anterior tip, Paramecium leaps backward almost before
the hair is seen to have reached it, giving the entire typical
reaction already described. Any other part of the body was
so insensible that it was not possible to cause a reaction of any
sort by touching it with the hair. Paramecium could be pushed
about and made to alter its direction of movement mechanically,
of course, but there was no active response of any sort when
it was touched at any point except the anterior end.

In Spirostomtmt ambiguum essentially the same results were
reached with mechanical as with chemical stimuli. If any part
of the body was touched, whether anterior end, posterior end,
or side, the infusorian gave the typical reaction — swimming
backward, turning toward the aboral side, then swimming for-
ward. A slightly greater percentage of cases of the typical
reaction was obtained by touching the anterior end, but the

I08 BIOLOGICAL LECTURES.

difference was little ; it varied in Spirostoma from different
cultures.

In the other organisms on which the effects of localized
mechanical stimuli were tried, particularly Loxodes rostrum,
Dileptus ansery OxytricJia fallax, and one or two other Hypo-
tricha, the following results were obtained: (i) The side
toward which the animal turns after a stimulus is entirely inde-
pendent of the side which is touched. In every case the organ-
ism turns toward one structurally defined side. If that is the
side which is touched, the organism turns continually toward
the source of stimulus, no matter how many times the latter is
repeated ; if the other side is touched, the creature of course
turns away from the source of stimulus. The impression is
given that it is physiologically impossible for the organism
to turn otherwise than toward this one side. (2) But the for-
ward or backward movement of the animals after a stimulus is
not thus independent of the localization of the stimulus. If the
anterior end is touched, the organism darts backward, turns
toward one side, then swims forward. The posterior half of
the body is very insensible, so that as a rule there is no response
to a mechanical stimulus occurring here. If, however, a strong
stimulus is given here, as by thrusting the tip of the rod
strongly against the resting animal, the latter simply swims
forward; if already swimming forward, it merely hastens its
forward motion when thus stimulated.

Thus, in the case of mechanical stimuli in these organisms
the direction of motion after a stimulus depends, to a certai?t
extenty so far as backward or forward motion is concerned, upon
the localization of the stimulus. This introduces a greater com-
plexity into the psychology of these creatures than the results on
Paramecium alone, or on the reactions to chemical stimuli alone
would lead us to judge. The organisms do in certain respects
react with reference to the localization of a stimulus affecting
them. The differing results gained with chemical stimuli are
probably to be interpreted, in view of the facts shown by a study
of mechanical stimuli, as follows : When a chemical diffuses
from a point lying behind the infusorian, it of course comes
first in contact, as a very weak solution, with the posterior end

BEHAVIOR OF UNICELLULAR ORGANISMS. 109

of the animal. Now, as already stated, this posterior end is
not at all sensitive, so that no reaction is caused. The chemical
continues to diffuse until it finally reaches the very sensitive
anterior end, when at -once the typical reaction occurs, and the
animal swims backward into the strong solution. The reaction
to a chemical is perhaps then always due to stimulation at the
anterior end.

Psychologically considered, we seem to have here a remark-
able transitional condition toward a perception of the localiza-
tion of the stimulus by the organism — a reaction with reference
to the localization of the stimulus so far as motion along the
axis of the body is concerned, a blind reflex, without regard to
the localization of the stimulus, so far as motion to one side is
concerned.

We may now summarize briefly the essential facts in regard
to the reactions of the unicellular organisms studied. The
reactions of these organisms may be classified into three reac-
tion forms :

(i) One is the thigmotactic reaction. Starting with the
moving infusorian, we find that it reacts to contact with solid
bodies of a certain physical texture by suspending part of the
usual ciliary motion, so that locomotion ceases and the organism
remains pressed against the solid. Whether anything more
than this cessation of part of the usual ciliary motion is con-
cerned in the thigmotactic reaction is very difficult to say.

(2) If we start with the resting individual, the simplest reac-
tion to a stimulus is the resumption of the usual forward motion.
This is the reaction that is produced when the solid substance
against which the creature is resting is removed ; it is also pro-
duced in some Infusoria when the posterior part of the body is
stimulated mechanically.

(3) The third, and, for our purpose, most important reaction,
to which most of the so-called tactic or tropic phenomena are
due, may occur in either active or resting animals. It is a
reflex consisting of the following activities : the animal swims
backward, turns toward one structurally defined side, then
swims forward. This reaction is produced by chemical stimuli
acting upon any part of the body or upon the entire body at

no BIOLOGICAL LECTURES.

once, by osmotic stimuli, by heat, by cold, by mechanical shock.
Its general effect is to take the organism out of the sphere of
operation of the agent causing the stimulus, and to prevent it
from reentering. The fact that certain areas are left vacant is
because the agencies within these areas cause this reaction ; the
collecting of the organisms within certain areas is due to the
fact that here the reaction is not produced, while it is caused,
by the influences active in the surrounding regions.

Thus, chemotaxis, tonotaxis, thermotaxis, and the like are
unified ; they are not qualitatively different activities, but are
fundamentally one activity due to different causes. The tactic
phenomena of unicellular organisms are brought throughout
under the same point of view as the motor reflexes so well
known in the physiology of higher animals.

We may now return to a brief consideration of the problems
which formed the starting point of this investigation — the
relation of the phenomena studied to the growth processes in
the protoplasmic masses of higher organisms. Do the laws
of the motor reactions of unicellular organisms, chemotaxis
and the like, really give us a basis for the understanding of
protoplasmic migrations and other processes in growth and
differentiation .'*

We find that the tactic phenomena of these unicellular forms
are brought about through a reflex that is in all essential points
similar to the reflexes of higher animals. The nature of this
reflex is closely bound up with the physiological and structural
differentiations of the body of these organisms ; it has a
specially close relation to the asymmetry of the cell body in
these Protozoa, and to the manner of the usual forward motion.
These differentiations are absent in the masses of protoplasmic
substance that are moved about in the processes taking place
within the eggs and embryos of Metazoa. It is difficult to see
how the laws controlling the movements of such substance
masses can have any similarity to the laws above developed
for the reflexes of free unicellular organisms. Above all, it is
evident that the tactic movements of unicellular organisms are
not direct expressions of simple chemical and physical laws ;
chemotaxis, for example, is not a direct result of chemical

BEHAVIOR OF UNICELLULAR ORGANISMS. Ill

affinities and repulsions between the protoplasmic substance
and other chemical compounds. Like all the other tactic
phenomena, it is the result of a motor reflex, which may be pro-
duced by the most varied means. These tactic movements,
then, do not establish an immediate relation between the move-
ments of organisms and the movements characteristic of inor-
ganic substances. The organism reacts as an individual, not
as a substance. To my mind the facts above brought out in
regard to the movements of these creatures tend, if these facts
have a general validity, to deprive such movements of their
supposed value for explaining or illustrating the processes of
growth ; in . so far as the ideas of chemotaxis and the like
in growth processes have been derived from the phenomena
exhibited 'by unicellular organisms, these ideas require a revision.

Especially do the facts above brought out reveal the fal-
lacy of the statement so often insisted upon, that the growth
processes induced by chemical or physical agencies are " the
same as " or "■ identical with " the locomotor reactions induced
by the same agencies. This has been carried so far that stren-
uous objection has been raised even to the use of distinguishing
terms for these two sets of phenomena. We are told that to
distinguish as -taxis the motor reactions of a free organism from
-tropisvi or the growth reactions of a fixed organ or organism
is all wrong ; the two are *' identical." It is reasonably certain
that the growth phenomena of plants are not brought about
through a reflex that is identical with the motor reflex of Para-
mecium ; it seems exceedingly probable that the ways by which
movements are brought about as responses to stimuli in the
various classes of plants and animals will present great variety.
It is difficult to see what is to be gained except confusion of
ideas by applying the same names to two such dissimilar
activities as the motor reflex of Paramecium when stimulated
by a chemical, and the bending of a plant to or from a chemi-
cal in solution.

In regard to the second question touched upon in my intro-
duction,— the nature and importance of the activities of uni-
cellular organisms as compared with those of many-celled
creatures, and their value for explaining the phenomena shown

112 BIOLOGICAL LECTURES.

by the higher, — the general trend of the answer is, I think,
evident. I should be inclined to interpret the facts presented
somewhat as follows : The claim that the motor processes of
unicellular organisms form a connecting link between inorganic
processes and the vital phenomena of higher creatures clearly
receives no justification for the organisms studied. Every
influence coming in from outside passes, as it were, through a
sort of central station, where it is completely transformed to
appear as a reflex action, the nature of which is conditioned by
the form and structure of the organism ; and the steps in the
transformation are no more evident than they are in the higher
forms. The reactions of these creatures are indeed simple, but
not qualitatively of a different sort from those of higher organ-
isms, so that for motor reactions of the sort studied I do not
see that a knowledge of the conduct of these particular uni-
cellular organisms really adds to our insight into the causal
relations in the activities of higher animals.

On the other hand, if we dismiss any idea of getting from
them knowledge of a different kind from that gained by the
study of other groups, then the behavior of these Protozoa is
of the greatest interest from the standpoint of comparative
psychology. In these creatures we see, as nowhere else, how
activities that seem so complicated and varied as to require
psychological powers of a high order, are produced merely
through one or two simple reflexes ; it seems not impossible
that the phenomena exhibited in the conduct of these organ-
isms may in time furnish important points of support for the
general theory of the origin and development of psychic
powers.

EIGHTH LECTURE.

THE BLIND-FISHES.

CARL H. EIGENMANN.

" An investigation into the history of degenerate forms often teaches us
more of the causes of change in organic nature than can be learned
by the study of the progressive ones." — Weismann.

The Amblyopsidae are a small family of fishes of six or seven
species. Four or five of these live in the caves of the Missis-
sippi Valley. Amblyopsis is the largest of the species and has
the widest range, being found both north and south of the
Ohio River, and reaching a length of 135 mm. The structure
of these fishes will be dealt with in various journals. The
habits in general are given in the Popular Science Monthly for
1900. I shall consider here the reaction to light and the breed-
ing habits of Amblyopsis and the color of the Amblyopsidae.

I. Reaction to Light.

A long series of observations and experiments were made to
determine the reaction of Chologaster and Amblyopsis to white
and monochromatic light. Incidentally other characteristics
were brought out. The experiments need not be given in
detail here.

Some previous experiments on blind or blinded vertebrates
may be recalled. Dubois ^ and Semper ^ record that Proteus,
the blind salamander of Europe, is sensitive to diffuse light.
Graber ^ records that blinded salamanders prefer dark chambers
to light ones. Korange^ notes that concentrated light deprived

1 Compt. Rend., no, pp. 358-360. ^ sb. Akad. fVt'ss., 87, pp. 221.

2 Animal I.ife, p. 79. * Centralbl. of Phys., 6, pp. 3-6.

"3

114 BIOLOGICAL LECTURES,

of heat rays, thrown upon the leg of a frog whose brain had
been laid bare and covered with extract of beef caused it to
respond each time with reflex movements.

Amblyopsis seeks the dark, regardless of the direction of
the light. An aquarium about eight feet long was placed in
the open, with the darkened end toward the north. The fishes
remained in the dark area all day, coming out at night. A
smaller aquarium was divided into two compartments, at first
with but a small connecting opening ; later with a double
partition, with an opening at one end of one partition and
another opening at the opposite end of the second partition.
One compartment and the connecting tunnel were dark. At
first there was a constant change of individuals from the dark
to the light chamber, but after several days they remained in
the dark chamber during the day. These experiments were
conducted in my laboratory where usually only diffuse daylight
prevailed. Jarring the aquarium or changing the water was
usually sufficient to cause them to come out into the lighted
compartment. It may be stated that the fishes found no diffi-
culty in finding the opening connecting the compartments, and
that the fishes in the dark were continually swimming past the
opening without attempting to come out. Rarely one passed
out into the light chamber, and then invariably showed signs
of uneasiness, frequently turning sharply and reentering the
dark chamber.

Four blind-fishes which had been kept for a day in a vessel
painted black, and covered to exclude the light, were experi-
mented upon as follows : a ray of light from a mirror about two
inches in diameter was thrown on each successively. After
from one to five seconds the fishes became uneasy, the uneasi-
ness giving place to discomfort, finally resulting in vigorous
effort to get out of the way.

Another jar, not painted, containing both blind-fishes and
blind Cambarus, was placed where light could be reflected upon
them from the mirror of a microscope. The Cambarus, if in
motion, came suddenly to a halt ; if quiet, backed or moved off
at once. The fishes also responded to the reflected light, but
it took several times as long for them to do so.

THE BLIND-FISHES.

115

Bright sunlight appears to be irritating ; if exposed to it, the
fishes swim about uneasily. A shadow passed suddenly across
them, when in the diffuse light of a room, does not affect them ;
nor did they seem disturbed, when swimming through a ray of
light entering the dark chamber by a small hole in the paint
made for the experiment.

Two examples kept in a pail in my cellar were quietly float-
ing, but when a lighted match was held above them, the fishes
at once darted to the bottom and sides of the pail. The heat,
in this case, could not have been a factor ; the reaction to the
light of the match was quick and violent. The same observa-
tion was made on forty individuals in two aquaria. They were
captured one morning, and the observation made the second
night after. They had been kept in the dark during most of
the intervening time. A lighted match, held near the aquaria,
produced a very general and active movement among all the
individuals.

A series of observations was made to determine to what rays,
if any, they reacted most vigorously. Only the concluding data
need be given here.

For this experiment a glass jar, three feet long and eight
inches in diameter, was divided into six compartments by five
partitions. Each partition had a vertical slit extending half-
way up from the bottom to enable the fish to swim freely from
one compartment to another. The compartments were thus
all connected. A cap was screwed tightly over the end of the
jar, which was placed horizontally on a window sill where each
compartment would have an equal amount of light. The jar
was surrounded with bands of tissue paper in several layers of
violet, blue, green, orange, and pink, so that each compartmerft
was lighted by one series of rays.

Three blind-fishes were used for these observations ; they were
selected for their size, and numbered : i, the smallest ; 2, the
middle-sized ; 3, the largest. These fish had been in confine-
ment some time, but had been transferred from the cave, with
as little exposure to light as possible, to a dark room, where
they were very seldom exposed to the light.

It was found that certain compartments were visited by a

Il6 BIOLOGICAL LECTURES.

certain fish without any definite regard for color. During
January, for instance, fish 3 moved out of the pink and orange
compartments but once ; fish i remained almost excUisively in
yellow, visiting pink once, orange once, and green four times.
Fish 2, on the other hand, remained mostly in violet, visiting
blue seven times and green three times. From this we must
conclude either that different individuals react differently, or
that one color does not produce a stronger reaction than an-
other, and the latter seems the more reasonable conclusion.

To determine whether the apparatus had anything to do with
the distribution, and also whether widely separated elements of
the spectrum would cause the fish to react positively or nega-
tively, they were put into a rectangular aquarium impervious to
light, except at the ends, and divided by a median partition.
The ends were covered with translucent celluloid film, care
being taken, of course, to have each end equally light.

In these experiments there was again conflicting evidence.
In general, it may be said that there was no marked difference
in the reaction to one or another of the rays of the spectrum.

A similar series of experiments was conducted with Cholo-
gaster with the same reaction to white light, but with a marked
positive attraction by the red rays of the spectrum as against
the blue.

A very positive series of observations has been made since
this was written. An aquarium two feet deep, four feet wide,
and about eight feet long was divided by a median partition to
near the bottom. Two lids were hinged at the top of this par-
tition. One or the other or both sides of the aquarium could
be covered at pleasure. Invariably, within a reasonable time
after one of the compartments was uncovered, all the fish
migrated to the dark compartment.

Even more striking than this is the action of a colony of
Amblyopsis in an open pool. During the bright part of the
day, the fishes always remained under the rocks at the bottom.
In the morning and evening and at night, they could be seen
swimming about in various parts of the pool.

THE BLIND-FISHES. II7

II. Breeding Habits.

We owe the first observations on the breeding habits of
Amblyopsis to Thompson,^ who states that a fish '' was put in
water as soon as captured, where it gave birth to nearly twenty
young, which swam about for some time, but soon died. . . .
They were each four hnes in length." Little or nothing has
been added to our knowledge of this subject since that time ;
but the highly interesting supposition of Thompson, that they
were viviparous, has gained common currency, and it is there-
fore unfortunate that in this respect he was in error.

Putnam 2 adds to the above that, judging from some data in
his possession, the young are born in September and October,
and further along remarks that they are undoubtedly viviparous.

The first young I obtained were secured on May 9, 1896.
The little fish could move actively for a few moments, but as
they were encumbered with much yolk, they soon settled to
the bottom and remained quiet. There were a large number of
old ones in the water in which the young were found, and the
mother of this lot was not identified with certainty. Another
lot of young were obtained on September 5 of the same year.
These were much further along in their development than the
first lot secured. Some were preserved and others were placed
in various aquaria, where one lived to be ten months old. As
before, the parent was not with certainty determined, simply
because it was taken for granted that they were viviparous, and
the ovaries only were examined. Two other lots of young were
obtained on June 5, 1897. One of these lots was in the stage
of the first lot obtained, with a large amount of yolk still pres-
ent, while in the other lot the yolk had almost entirely dis-
appeared. These had been carried in the gill cavity of the
mother, and it became evident that either the fish were not
viviparous at all, or their viviparity was not nearly of the pro-
nounced character hitherto supposed.

On March 11, 1898, twenty-nine individuals were captured.
Four of these were females with eggs in their gill cavities.

1 Ann. Mag. Nat. Hist. 1844.

2 Am. Nat., p. 16, 1872.

Il8 BIOLOGICAL LECTURES.

The youngest stage among these was at the end of segmenta-
tion, the oldest was a gastrula covering but one-third of the
yolk. The eggs had not been developing more than five days,
probably not more than two at the utmost, and decided beyond
a doubt that the fish are oviparous, and not viviparous.^

The eggs are laid by the female in under her gill membrane.
Here they remain for perhaps two months, till the yolk is
nearly all absorbed. If at any time a female with young in
her gill pouches is handled, some of them are sure to escape.
This was observed, and gave rise to the idea that this fish is
viviparous. Eggs have been obtained as early as March 1 1
and as late as September, and the indications are that the
breeding season extends throughout the year.

The eggs are translucent amber, in various shades in differ-
ent individuals. They are large, measuring 2.3 mm. in diam-
eter from membrane to membrane. The yolk measures 2 mm.
in diameter and contains an oil globule 1.2 mm. in diameter.
The globule protrudes from the yolk. The eggs are heavier
than water, and lie with the oil sphere uppermost and the
germ lateral. In one individual sixty-one eggs were found, in
another seventy. The exact number in the other two I cannot
give, but the number does not differ greatly from the above.
From one side of one I took thirty-five eggs, from the other
individual an uncertain number. The remaining eggs were
left in the gills to develop, but all those that were not subse-
quently preserved finally died.

The female with eggs can readily be distinguished by her
distended gills, and since dead eggs become opaque, such can
readily be distinguished through the translucent opercles and
branchiostegals. Dead eggs are retained in the gill cavity till
they disintegrate.

I have never secured as many young from any female as eggs
enumerated above. This may have been either on account of
the dying of many eggs or the liberation of the young during
the struggle of capture.

1 This is the first of a series of embryos obtained through a grant of $ioo
from the Elizabeth Thompson Science Fund, made to enable me to continue my
work on the blind-fishes.

THE BLIND-FISHES. II9

The details of the embryology are now under consideration.
Emphasis need be laid only on the fact that Amblyopsis is 7iot
viviparous, and that the breeding period extends at least from
the first of March to November, and probably throughout the
year. A female with nearly ripe eggs was secured on Septem-
ber 9, and since these would have been carried, either as eggs
or young, for about two months longer, November is a safe
limit. Between November and March but one trip has been
made to the caves, and this does not permit of a final settle-
ment as to whether or not breeding is carried on at this
time.

Certain structures gain an entirely new significance in the
light of the breeding habits. These are the enlarged gill cavi-
ties, with the small gills, the closely applied branchiostegal
membrane, and the position of the anus and sexual orifices.
The latter are placed just behind the gill membrane, in such
close proximity to it that they can be covered by it. It is
probable, therefore, that the membrane is drawn over the
sexual orifice and the eggs deposited directly into the gill
cavity. In an individual 35 mm. long the anus is situated
between the origin of the pectorals ; in one 25 mm. long it lies
between the pectorals and ventrals. In the young it lies
behind the ventrals, as in other fishes.

In an aquarium containing six Amblyopsis two took a great
antipathy to each other and engaged in frequent contests.
Whenever they came in contact a vigorous contest began.
Frequently they came to have a position with broadside to
broadside, their heads pointing in opposite directions. At such
a time the fight consists in quick lateral thrusts toward the
antagonist to seize him with the mouth. The motion is
instantly parried by a similar move by the antagonist. This
blind punching may be kept up for a few seconds, when by
their vigorous motions they lose each other and jerk themselves
through the water from side to side, apparently hunting for
each other. At this time they are very agile and move with
precision. When the belligerents meet one above the other,
the snapping and punching are of a different order. While jerk-
ing through the water immediately after a round, if one of the

I20 BIOLOGICAL LECTURES.

belligerents touches one of the neutrals in the aquarium, it
frequently gives it a punch, but does not follow it up, and the
unoffending fellow makes haste to get out of the road, the
smaller ones doing this most quickly. If after an interval of
a few seconds a belligerent meets a neutral, they quietly pass
each other without paying any further attention, whereas, if
the two belligerents meet again, there is an immediate response.
Whether they recognize each other by touch or by their mutual
excitability I do not know. At one time, in another aquarium,
I saw one belligerent capture the other by the pectorals.
After holding on for a short time it let go and all differences
were forgotten.

The thrust is delivered by a single vigorous flip of the tail
and caudal to one side.

These fights were frequently noticed, and, as far as deter-
mined, always occurred between males.

The absence of secondary sexual differences in the cave
fishes is a forcible argument in favor of sexual selection as the
factor producing high coloration in the males. The absence
of secondary sexual differences in caves opposes the idea of
Geddes and Thomson that the differences are the external
expression of maleness and femaleness.

Attempts at acclimating Amblyopsis in outside waters have
largely failed. A few were placed in Turkey Lake, Indiana.
They were surrounded by a fine wire net to keep off other
fishes. They died in a few days as the result of attacks of
leeches, Saprolegnia, or fish mould, and from unknown causes.
Others were kept in an elongated box sunk into the ground
where fresh spring water flowed through it constantly. Sapro-
legnia sooner or later destroyed all of them. They live longest
in quiet aquaria where the water is rarely changed. The young
I have secured died, with one exception, within a few weeks.
The difficulty of rearing the young is not at all insurmountable.
They eat readily. Their aquaria must be kept free from green
plants and have a layer of fine mud with a few decaying leaves
in the bottom. They will feed on minute crustaceans and
other micro-organisms. When they have reached a sufficient
size, Asellus are greedily devoured. Fish mould is the bane of

THE BLIND-FISHES. 121

the larvae. Many of them were found with tufts of the hyphae
growing out of their mouths and gill openings.

A colony has been apparently successfully planted in an
open pool at Winona Lake, Indiana.

III. The Color of the Amblyopsidce,

The three species of Chologaster are colored with varying
intensity from C. cornutus, in Florida, which is darkest, to C.
agassizii, in Mammoth Cave, in which the color is faintest.
The color cells are in all cases arranged in a definite pattern.
These are determined by the underlying muscles. The pattern
consists of three longitudinal bands on the sides following the
line where the muscle segments are angularly bent, and cross
stripes along the line separating successive segments.

The lower side of the head and the abdomen of Chologaster
papilliferus are sparingly pigmented and translucent. The
underlying liver and gills give the parts a rosy tinge. On the
sides and top of the head, pigment is abundant. There is a
more densely pigmented area extending along the middle of the
back, beginning as a narrow stripe at the nape, and widening
to the dorsal fin behind, where it occupies the entire back. On
the sides are three narrow stripes, which, owing to the accu-
mulation of pigment in two layers, are quite dark. Each stripe
has a lighter central band, widest at the rhiddle of the sides.
The light band, without the conspicuous bordering dark stripes,
runs along the middle of the belly. The sides are thickly
covered with a layer of pigment, leaving usually colorless lines
where connective tissue separates successive myotomes. On
the sides- of the tail the pigment is dense on either side of these
colorless lines. A dark band extends along the sides of the
head through the eye. The top of the head is dark.

The pattern of Chologaster cornutus agrees with that of
papilliferus. The longitudinal bands are much darker and
wider and without the light central streak. The middle band
is much wider than the others, and is continued forward to the
tip of the snout. The amount of color present varies very
greatly with the locality from which the specimens come.

122 BIOLOGICAL LECTURES.

The general color of C. agassizii is light gray. The scales
are lighter than the area surrounding them.

The color pattern is more striking than in the other species
of the genus. Each somite is bordered by a dark line. The
lines of successive somites are separated by an almost imper-
ceptible colorless line. A broad, not sharply defined, band
extends along the sides. The middle of this is lighter than
the margin. Another one extends between the somites and
the ventral musculature, another from the nape between the
lateral somites and the dorsal muscles, and a diverging one from
near the nape to either side of the dorsal fin. Dark areas are
found along either side of the dorsal and anal fins. These areas
are caused by the accumulation of pigment along the borders
of the small muscles of the fins. Still another dark area is
found about the caudal. The ventral surface is white, except
the accumulation of pigment along the lines separating the
muscles.

The fins are uniformly light gray. A light area is found on
both the upper and lower part of the caudal peduncle, just
within the first short rays of the caudal.

The general color of Typhlichthys is cream and pink. It is
abundantly pigmented. In younger specimens the pigment
is arranged in definite areas about the head. In the old it
is more uniformly distributed, being, however, specially abun-
dant about the brain. The pigment pattern of the body is pre-
cisely as in Chologaster, except that the individual pigment
cells are minute and their aggregate not evident except under
the lens.

The retention of the color pattern of Chologaster in Typhlich-
thys is not less interesting than the retention of similar habits.
It is perhaps due to different causes. The color pattern in
Chologaster is determined by the underlying muscular struc-
ture, and the retention of a similar pattern in Typhlichthys is
due to the same underlying structure, rather than to the direct
hereditary transmission of the color pattern.

In Amblyopsis the color is much less marked than in Typh-
lichthys. They are flesh-colored, ranging to purple in the gill
region, where the blood of the gills can be seen through the

THE BLIND-FISHES.

123

overlying structures, and over the liver, which can be seen
through the translucent sides and ventral wall. About the
head and bases of the fins the color is yellowish, resembling
diluted blood. The surface of the body is slightly iridescent,
and the surface of the head has a velvety, peach-bloom appear-
ance.

The general pink color of Amblyopsis is due to the blood.
It is not due to any abnormal development of blood vessels in
the dermis. In the fins, where the blood vessels are near the
surface, the general effect is a yellowish color. The surface
vessels of the dermis also appear yellowish. It is only on
account of the translucent condition of all the tissues, permit-
ting the deeper vessels to show through a certain thickness,
that the pink effect is produced. Amblyopsis has always been
spoken of as white. The term '* white aquatic ghosts " of Cope
is very apt, for they do appear white in the caves, and their
gliding motion has an uncanny effect. All alcoholic specimens
are white.

The chromatophores in Amblyopsis are differentiated and
contain color before the yolk is absorbed. The black chro-
matophores are minute granules, few (fifteen or thereabout) to
the segment. In an older larva the pigment was much more
abundant. The eyes are pigmented early, — shortly before hatch-
ing, — and, owing to their pigment, they soon become conspicu-
ous and remain so till the fish has reached 50 mm. in length,
when the overlying tissues have become thick. The pigment
of the body is lost, or, what amounts to the same thing, does
not increase much with age. There is an abundance of pig-
ment cells in the adult, but they are very poor in pigment, and,
being in the dermis and covered by the thick layer of epidermis
rich in glands, are not apparent. Pigment cells are also abun-
dant in deeper tissues in the adult, so that, while no pigment is
visible on the surface, an abundance of chromatophores are
present in deeper tissues.

The pigment cells cannot be made to show themselves even
by a prolonged stay in the light. The old, if kei>t in the light,
will not become darker, and a young one reared in the light
until ten months old not only showed no increase in the pig-

124 BIOLOGICAL LECTURES.

mentation, but lost its pigment, taking on the exact pigment-
less coloration of the adult. Pigment cells are late in appearing
in Amblyopsis. When the young are two months old, pigment
is abundant. This pigmented condition is evidently a heredi-
tarily transmitted condition. It disappears with age. Primarily
this disappearance was probably individual. But, as in the
flounder, the depigmentation has also become hereditarily trans-
mitted, for even those individuals reared in the light lose the
color.

Numerous facts and experiments show that, while pigment
may be and is developed in total darkness, the amount of color
in an individual animal depends, other things equal, directly on
the amount of light to which it is habitually exposed.^

The lower and upper surfaces of the flounder, the one pro-
tected and the other exposed to the light, give the most striking
example, and the argument is clinched here by the fact, noted
by Cunningham,^ that a flounder whose lower side is for long

1 A number of apparently contradictory observations may be noted : a. The
absence of pigment in pelagic animals or their larvae, which depend on their color-
less condition for their existence, is evidently due to causes entirely different from
those preventing the formation of pigment in cave animals. Natural selection
has, in pelagic animals, eliminated the color, b. The expanding of chromatophores,
when an animal is over a dark background or in the dark, and the contracting over
a light background, which may take place instantly or at the expiration of several
days, is evidently also a different question. The observations of Cunningham,
Agassiz, and Semper along this line are of interest, c. Fischel {A. M. Anat., vol.
xlvii, pp. 719-734, PI- XXXVII, 1893) has noticed that larvae of salamanders reared
in water at 6-7° are dark, while others kept in water from 1 5-58° are light, d. Flem-
ming {A. M. Anat., vol. xlviii, pp. 369-374, 1896) found that with uniform temper-
ature in two vessels side by side, the one dark, the other light, the salamander
larvae in the dark vessel develop pigment cells rich in color granules ; the larvae in
the white vessels become pale, although the number and character of the pigment
cells is not otherwise changed. The difference is entirely due to the character of
the vessels, for, if the larvae are taken from the dark to the light vessel, they
become light-colored in a few days. e. Semper (Animal Life, p. 89) records that
"... in the tadpoles of our common toads and frogs the pigment is equally well
developed in yellow, blue, or red light, and in absolute darkness." This was to
be expected, for even in the young of cave animals pigment is, as a rule, well devel-
oped. / Pouchet {Arch, de Physiol, et d'Anat., 1876, and Rev. Scient., vol. xiii,
1897) has demonstrated that change in color cells, such as are mentioned under
b and d, is brought about by the reflex control from the eye. The section of the
great sympathetic nerve puts an end to the changes of color under the influence
of light. 2 philos. Trans., pp. 765-812. 1893.

THE BLIND-FISHES. I 25

periods exposed to the light takes on color. Loeb has shown
that, in the yolk sacks of Fundulus embryos, more pigment cells
are developed if the embryos are kept in the light than when
they are kept in the dark. However, in the body, and espe-
cially in the eye, the pigmentation was not affected by the
absence of light.

The general absence of color in cave animals is conceded.
Packard ^ states *' as regards change of color, we do not recall
an exception to the general rule that all cave animals are either
colorless or nearly white, or, as in the case of Arachnida and
insects, much paler than their out-of-door relatives." Chilton ^
has made the same observations on the underground animals
of New Zealand. Similar observations have been recorded by
Lonnberg,^ Carpenter,* Schmeil,^ and Vire.^

Hamann^ enumerates a number of species living both in
caves and above ground. In such cases the underground
individuals are paler than the others. This confirms similar
observations of Packard.

Poulton ^ has mentioned that Proteus becomes darker when
exposed to the light. This has been verified by others. In
Typhlotriton, larvae living at the entrance of a cave are dark,
while the adult living further in the cave are much lighter, but
with many chromatophores containing a small amount of color.
Epigaean fishes found in caves are always lighter in color than
their confreres outside.

We have thus numerous exam^^les of colored epigaean animals
bleaching in caves and also bleached cave animals turning dark
when exposed to the light. We have also animals in which the
side habitually turned to the dark is colorless, while the side
habitually turned to the light is colored. Finally we have cave
animals that are permanently bleached.

1 Mem. Nat. Acad. Set., vol. iv.

2 Trans. Linn. Soc. Land., 2d Ser., vol. vi, pp. 163-284.

3 Zool. Anz., vol. xvii, p. 125.

* Irish Naturalist, vol. iv, p. 25.

5 Zeitsch.f. Naturw., vol. Ixvi, p. 339.

6 Bitll. Mus. Natur. Hist. Paris, 1895, P- 143*
^ Europ, Hohlenfauna, p. 6.

8 The Colour of Animals, p. 91.

126 BIOLOGICAL LECTURES.

Natural selection cannot have affected the coloration of the
cave forms, for it can be of absolutely no consequence whether
a cave species is white or black. It could only affect the col-
oration indirectly in one of two ways : first, as a matter of
economy, but since the individual is in part bleached by the
direct effect of the darkness, there is no reason why natural
selection should come into play at all in reducing the pigment
as a matter of economy ; second, Romanes ^ has supposed that
the color disappeared through the selection of correlated struc-
tures, a supposition he found scarcely conceivable when the vari-
ety of animals showing the bleached condition was considered.

Panmixia cannot account for the discharge of the color, since
it returns in some species when they are exposed to the light,
and disappears to a certain extent in others when kept in the
dark. Panmixia, Romanes thinks, may have helped to discharge
the color. In many instances the coloration is a protective
adaptation and therefore maintained by selection. " Panmixia
might, in such instances, lower the general average to what
has been termed the birth mean." Proteus is perhaps such an
instance. But in this species the bleached condition has not
yet been hereditarily established, and since each individual is
independently affected, "the main cause of change must have
b)een of that direct order which we understand by the term
■* climatic' "

Since, however, the bleached condition, which in the first
instance is an individual reaction to the absence of light, has
become hereditarily established in Amblyopsis, so that the
bleaching goes on even when the young are reared in the light,
it is evident that in Amblyopsis we have the direct effect of
the environment on the individual hereditarily established.

1 Darwin and after Darwin.

NINTH LECTURE.

SOME GOVERNING FACTORS USUALLY

NEGLECTED IN BIOLOGICAL

INVESTIGATIONS.

ALPHEUS HYATT.

The object of this lecture will be fully attained if, by
fragmentary statements, I can attract the attention of my
audience to certain neglected or little understood facts and
theories.

There will be no attempt to prove any special theory, but
simply to make statements and refer interested readers, if I am
so fortunate as to have any, to works in which more extended
information can be found.

The terminology used in the following pages has been found
necessary to express more exact ideas of the relations of phe-
nomena which cannot be accurately described without such
aids. An old term used in a new sense is confusing, because
the accepted usage invariably clings to the word in the minds
of all but the inventor of the new meaning.

It may seem useless to dwell on that familiar word " evolu-
tion," but, as a matter of fact, misconceptions of the true
meaning of this term are by no means rare, and a correct point
of view is necessary, especially in considering the relations of
ontogeny and phylogeny.

Evolution is essentially continuity of matter moving through
time and space, and actually changing as it moves. The laws
of evolution are concise descriptive expressions of the modes
in which these movements and changes have taken place. The
causes of evolution and of the* accompanying changes are still

127

128 BIOLOGICAL LECTURES.

more profound and difficult, and fortunately not pertinent to
the immediate object of this lecture.

This definition of evolution is based upon the whole mass of
phenomena which has been shifting and modifying itself before
the eyes of observers from the earliest times to the present day.
The fact of perpetual change is acknowledged, and it has also
been generally recognized that different bodies change in vari-
able degrees. Some substances are so elemental and primitive
in comparison with those succeeding them in time or in their
own series that they can be spoken of as relatively unchange-
able in comparison with these. The inorganic body that we
now speak of as the ether and some of the chemical elements
may be rationally imaged as still existent in unchanged condi-
tion in some parts of the universe, and in those particular cases
there would be continuity without change, and consequently no
evolution in the sense that this is generally used.

It has also been quite commonly and reasonably supposed
that the most primitive of the existing forms of the animal
kingdom among Protozoa, Protamoeba, for example, or even
so comparatively complex an animal as the existing Amoeba,
may still retain most of the characters that they possessed when
near the beginning point in the evolution of organisms.

It is difficult, in view of such possibilities, to avoid the theo-
retical conclusion that the evolution of organisms was quite
distinct in its first period, and the earliest primitive organisms
necessarily more like inorganisms and comparatively unchang-
ing in their structures. In other words, evolution probably
began to work gradually, and there was necessarily a time
in which the laws of organic evolution, as stated above, were
comparatively inoperative. Biology, like mathematics, has there-
fore an imaginary zero point in time and space, so far as vital
characters are concerned, and an actual unit of departure only
in the simplest organisms.

Terrestrial organic bodies necessarily move forwards in time
and space with the earth, but they have also definite motions
of their own. There are two classes of these : (i) the move-
ments of, the single organism through its different stages of
growth and development. These ontogenic changes are divisi-

BIOLOGICAL INVESTIGATIONS. 1 29

ble into five periods — the embryonic or earliest usually encysted
or ovarian stages, the nepionic or baby stage, the neanic or
adolescent, the ephebic or adult, and the gerontic or senile
stage. These pass insensibly into each other and have two
phases — the progressive, leading to the full-blown adult, and
the retrogressive, leading by senile degeneration to the final
destruction of organic continuity in death.

The second class of movements (2) take place on a larger
scale, and have been designated Phylogeny. This is the life
history of the stock, but is more generally conceived of as
the evolutionary history of any series of genetically connected
forms.

If the ontogeny be granted as possessing capacity for change
and the power of reproduction, the phylogeny becomes neces-
sarily a function or product of the ontogeny. This is so obvi-
ous that it would need no explanation but for the fact that at
present the appearance of variations in the young is supposed
by some to indicate fundamental inherent tendencies of organ-
isms to depart from the parent stock.

The mass of evidence and all the obvious phenomena, how-
ever, show the identity of reproduction and heredity ; they are
both necessarily the production of like by like. The simplest
forms of reproduction, the division of the unicellular Protozoa,
the formation of colonies among Protozoa, and all the asexual
modes of reproduction have this character in the Metazoa.
The sexual modes of reproduction do not alter these funda-
mental relations, except in so far as the spermatozoa graft dif-
ferent elements upon the ovum, which influence the subsequent
characters of the organism. The phenomena of alternation
of generations in Hydrozoa and the hyper-metamorphoses of
insects cannot be cited against the old views of heredity and
reproduction, since the cycle of the ontogeny in these invari-
ably swings around to the same form, however great the devia-
tion of the intermediate generations or stages of development.

Variations in any given forms are, as a rule, at first infinitely
slight as compared with the vast mass of recapitulated charac-
ters. They are necessarily grafts upon stock structures until
they become incorporated with them, and then are subject to

I30

BIOLOGICAL LECTURES.

the same law of reproduction. If the old position is true, that
an organism when it evolves into another organism necessarily
gives rise to one like itself, then the evolution of a genetically
connected stock is an inevitable corollary of the existence of a
single organism endowed with the power of reproduction. If
phylogeny is a product or function of ontogeny, there is a nat-
ural basis for the correlations between these two classes of
organic movements, and their essential similarity is probably
due to this connection.^

While this is obviously true, and one can describe the
phylogeny in parallel terms to those used for the ontogeny,
as is done further on, we should nevertheless bear in mind
certain essential distinctions. Thus, the genetic stock has not
the absolute continuity of the ontogeny. When regarded in
any living or fossil fauna it is an aggregate of organisms,
a mosaic of single life histories as distinctly separated from
each other in time and space as the pieces of mosaic, except
during the brief stage of parental connection. All the changes
and movements in phylogeny are necessarily based on two con-
ditions : first, the aspects of the faunas on any one level in
time, and second, the relations of series of these on successive
levels in time. The successive faunas on different levels, as
shown by the geologic history of any genetically connected
stock of phylum, resolve themselves into very definite series
of pictures characterized by distinct structural and physio-
logical changes.

Both the ontogeny and phylogeny start from the same initial
point, and the divergencies necessarily take place first in the
ontogeny and are afterwards transmitted to the phylogeny.
This initial point for the organism is of course the Protozoan-

1 Jackson, in his " Localized Stages in Development of Plants and Animals,"
Mem. Boston Society of Natural History, vol. v, No. 4, 1899, ^^^ lately added another
chapter to the correlations of ontogeny and phylogeny in showing that the twigs
and leaves of plants and plates of the echinoderms, or inflection of the septal edges
(sutures) of an Ammonite, or bud of an Actinozoon, etc., although arising from
or in the body of an adult organism, repeat the ontogenetic metamorphoses or
changes which occurred in similar parts that first arose in the early stages. He
has also applied this newly discovered phase of the ontogeny to the translation
of the phylogeny and the natural classification of groups.

BIOLOGICAL INVESTIGATIONS. I 31

like form recapitulated in every ovum more or less ; but leaving
this remote region on one side, we find the initial point for
any branch, large or small, of the animal kingdom is very often
a primitive form, acquiring the characters of its successors late
in life ; or they may have them as yet in such indefinite shape
that they become perceptible only after this primitive ancestor
has been traced to better characterized descendants through
the gradations of intermediate forms. Several phyla have
been followed out by the painful and tedious process of track-
ing from species to species, and others by using the ontogeny
or life histories of different forms as indices of the natural
relations of groups.^

1 The following is a partial list of these :

Beecher. Rev. of the Families of Loop-bearing Brachiopoda. Trans. Conn.

Acad. Set. Vol. iii, 4th Sen, 1893, PP- 376-391-
Beecher. Outline of a Natural Classification of the Trilobites. Anier. Journ.

of Set. Vol. iii, 4th Sen, 1897, pp. 86-106 and 181-207.
Beecher. Trilobites in ZitteT's Text-Book of Paleontology, Am. Ed.
BucKMAN, S. S. Mon. Inf. Ool. Ammonites. Paleontogr. Soc. London, 1887,

and still in course of publication.
HiLGENDORF. Planorbis multiformis in Steinh. Siisswasserkalk. Monatab. der

Preuss. Acad, der IVt'ss., p. 474. Berlin, July, 1866.
Hyatt. Genesis of Tertiary species of Planorbis at Steinheim. Ann. Mem. Bost.

Soc. Nat. Hist. 1880.
Hyatt. Genesis of the Arietidae. Smith's Contr.} No. 673, 1889, and Mem.

Mus. Comp. Zool.
Hyatt. Genera of Fossil Cephalopoda. Proc. Bost. Soc. Nat. Hist. Vol. xxii,

1883, pp. 253-338.
Hyatt. Phylogeny of an Acquired Characteristic. Proc. Amer. Phil. Soc,

Philadelphia. Vol. xxxii. No. 143, 1894.
Hyatt. Cephalopoda in ZitteTs Text-Book of Paleontology, Am. Ed., in course of

publication.
Jackson. Phylogeny of Pelecypoda. Mem. Bost. Soc. Nat. Hist. Vol. iv, No. 8,

1890, pp. 277-400.
Jackson. Studies of Palechinoidea. Bull. Geol. Soc. Amer. Vol. vii, pp.

171-254.
Neumayer. Beit. z. Kennt. foss. Binnen-Fauna. fahrb. Geol. Reichs. Bd. xix.

No. 3, 1869, pp. 355-382.
SCHUCHERT. A Classification of the Brachiopoda. Amer. Geol. Vols, xi and

xiii, 1893-94.
ScHUCHERT. Brachiopoda in ZittePs Text- Book of Paleontology. Vol. i, Pt. I,

1896.
WuRTENBERGER. Studien uber die Stammesgeschichte der Ammoniten. Dar-

winistische Schriften. No. 5. Leipzig, 1880.

132 BIOLOGICAL LECTURES.

In these as well as by the general experience of paleobiologists
in tracing groups by more empirical methods, phylogeny has
been found to be at first progressive. It has, like ontogeny,
a period of differentiation, when it spreads out into its many
different modifications, called species, genera, and so on, and
thus finally attains an acme of progress. This acme is fol-
lowed by a phase of retrogression in genetic series that have
what may be called a complete cycle, and this ultimately ends
in extinction. All genetic series do not have complete cycles,
any more than all individuals have complete existences, and
some certainly possess introduced adaptive stages (Echino-
dermata, Insecta) that complicate the record. Reference is
here, therefore, confined to those genetic stocks that did pass
through both progressive and retrogressive series of changes.

There are other qualifications to such a brief statement that
cannot be discussed here, but in a general way it may be said
that this hypothesis of the correspondences of ontogeny and
phylogeny has been worked out by hard labor of minute com-
parison and the study of definite morphic modifications. The
method thus elaborated has been and is now in constant use
by a number of paleobiologists, and has proved in application
more helpful than its most enthusiastic advocate had hoped.
The correspondences are not simply between the youngest
stages and the remote ancestors, like those traced by embry-
ologists, but between all the stages of the ontogeny and all the
periods and phases of the phylogeny. The recapitulations of
the embryonic stages are in fact of no greater importance in
these researches than those of the epembryonic or later stages,
meaning by this term to include not only the young but also
the gerontic or senile metamorphoses. The nepionic stages
have characters derived from less remote ancestors than the
embryonic ; the neanic are similarly related to still nearer
progenitors; the ephebic stages give the differentials elabo-
rated in the ontogeny at the acme of the evolution of the
stock. ^ Definite periods in the phylogeny characterized by

1 " Bioplastology and the Related Branches of Scientific Research," Proc. Bost.
Soc. Nat, Hist., vol. xxvii ; *' Phylogeny of an Acquired Characteristic," Proc.
Amer. Phil. Soc, Philadelphia, vol. xxxii, Nos. 143, 144, pp. 59-125.

BIOLOGICAL INVESTIGATIONS. 133

Structural details have been worked out and found to corre-
spond to those of the ontogeny, and several authors now de-
scribe these different stages of the phylogeny by the use of
compound terms derived from phylum and the name of the
corresponding stage in the ontogeny. They are in this way
enabled to designate the primitive types of the phylum as
phylembryonic, the nearer ancestors as phyloneanic, and the
full-blown acme as phylephebic.

But, as stated above, this does not finish the history of
correlations. Taking up next the retrogressive phase of the
ontogeny, similar methods have shown that the retrogressive
stages in the evolution of the phylum correspond to those that
occur in old age or the gerontic stage, and by a similar termi-
nology we have been able to describe the retrogressive stages
of the phylum as being phylogerontic.

There are, however, very essential differences, in the beha-
vior of the correspondences of the ontogeny and phylogeny,
between the progressive and retrogressive phases. It will at
once strike most thinkers that progressive changes in the on-
togeny take place, as a rule, before the reproductive period, and
have consequently a more direct influence upon the phylogeny.
Their recapitulation in the ontogeny may therefore be caused
by heredity, while the recapitulation of the retrogressive char-
acters, which occur in the phylogeny after this period is passed,
may perhaps not be caused by heredity, but may be due to the
necessarily similar character of degenerative changes in similar
organisms having a common origin.

Biologists have not as yet investigated the limits of the
reproductive period in the different stages of development of
animals, and it is therefore not practicable to solve this ques-
tion. Who knows, for example, what the limit of the reproduc-
tive stage is in a clam or a lobster or any of the invertebrates .•*
That these have senile changes is easily seen, but whether the
gerontic stage begins before or after the power of reproduction
is lost, if it ever is in the ontogeny of some of these animals,
no one seems to know. Certainly in many invertebrates, and
in as high a type as Mammalia, senile degeneration sets in long
before the reproductive power has disappeared.

134 BIOLOGICAL LECTURES.

Whether phylogerontic characteristics are inherited or simply
have a common origin, the facts of correspondence remain the
same, and this enables us to speak confidently of the whole
cycle of the ontogeny or ontocycle as more or less exactly par-
allel with the cycle of the phylogeny or phylocycle.

Even in such general statements it is perhaps necessary to
caution readers that the retrogressive phases above mentioned
are entirely those that apparently occur first in the gerontic
stages of the ontogeny of more or less progressive forms of the
same genetic stock. There are in some groups, especially
among parasites, also introduced, adaptive stages of a retro-
gressive nature that are not in any sense gerontic, and these
are not included in these statements.

Connected organisms, according to such views, are looked upon
as complex associations of series diverging like the branches of
a tree, each branch or phylum passing through a stage of
phyletic rejuvenation represented by the more or less primitive
forms with which it begins. These radicals are more or less
closely related or divergent as compared with the common
radical stock or trunk, according to their position in time or
their grade in natural classification, but become in turn starting
members of new series. These series, if they take a normal
course, first progress in structure and then retrogress, but ever
proceed onwards in time and space until extinction closes their
career.

All naturalists, even at the present time, have not acquired
the habit of looking upon organisms as in continuous motion
and unceasingly changing while moving onwards through time
and space. Farlow treats the errors due to this omission in
his paper, read before the Boston meeting of the American
Association for the Advancement of Science in 1898, on ''The
Conception of Species as affected by Recent Investigations on
Fungi." This high authority writes as follows: "Our so-
called species are merely snapshots at the procession of nature
as it passes along before us. In any case it represents only a
temporary phase, and in a short time will no longer be a faith-
ful picture of what really lies before us, for we must not forget
that the procession is moving constantly onwards, and at a

BIOLOGICAL INVESTIGATIONS. 135

more rapid pace than some suspect." The errors caused by
the want of a full realization of this point of view consist
largely in sins of omission, as shown by the absence of accurate
records. For example, every observation on a species should
be dated, location, temperature given, etc., and all the relations,
as far as practicable, described, such as habits, food, number of
specimens collected, etc. The observations of marine natural-
ists, under the leadership of Verrill, have reached a high degree
of accuracy in these respects, while those of terrestrial faunas
have left much to be desired in some departments, especially
Mollusca.

Observers in all departments, as a rule, have also neglected
the obvious manifestations of senility, and treated all animals
except man as if they were exempt from growing old. Ap-
parently, so far as current descriptions go, the invertebrates
and most vertebrates are not affected by senile changes, but
simply cut off at the end of their uneventful lives by sudden
death, unheralded by any preliminary metamorphoses. Thus
conchologists have, for the most part, collected and studied the
largest shells with so-called perfect apertures, and described
these without any suspicion that they were often dealing solely
with extreme senile substages. In fact they have often insisted
that these alone were the full-grown representatives of the
species, and even in many cases have discouraged the collection
and study of the real adults and the younger stages, calling
them '* immature specimens."

Such views also enable the observer to see at once the
fallacy of classifications and diagrams of the relations of organ-
isms that represent them as converging. They diverge from
their points of origin, but by no stretching of facts or of the
imagination can they be accurately represented anywhere as
converging. The series may be parallel, and different genetic
series may contain closely representative forms that may have
been once considered as belonging to the same genus or even
species,^ but the term <* convergence" conveys a false impression.
From the genetic point of view no convergence is possible

1 See Cope, Origin of the Fittest (New York, 1887), and other essays; also
Hyatt, " Arietidae," op, cit.

136 BIOLOGICAL LECTURES,

except in cases of actual '^hybridity." This use of "conver-
gence" is an excellent example of a well-meaning but mistaken
attempt to convey new ideas by the use of old terms, since I
imagine that the authors using this term never meant to imply
that such forms were genetically convergent.

Any form of diagram or expression that describes organisms
as masses or lines radiating in every direction from a center in
order to emphasize the fact that many of them are reversions
or returns of preexisting forms, can be objected to on the same
grounds. Suppose the author of such a diagram or description
to be correct, and that his so-called reversions are real rever-
sions and not simple parallelisms, nevertheless his reversionary
forms sprang necessarily from a given level of time and space,
and it is as erroneous as it is unnecessary to represent them
by lines or names which are placed below this level. Whatever
diagrammatic or descriptive form may be selected it must have
a base level in time or space upon which the primal point or
points or line or base rests, and below which lines or figures
representing genetic stocks should not be drawn.

The serial nature of all organisms and genetic stocks is also
thrown into proper prominence by this view of the phenomena
of evolution. All series begin necessarily with an imaginary
starting point, to which the series may be traced by its grada-
tions, but in which its essential characters do not exist, this
being the zero point of that particular series. Such organisms
are well known to all phylogenists, animals that would cer-
tainly be placed in different series from those to which they
are referable through intermediate gradations, if those grada-
tions had not been traced. Such forms often give rise to end-
less controversy unless they occur associated in the same fauna
with the forms into which they grade. Next to this first more
or less indefinable beginning comes the simplest unit of the
morphic series in which the type is distinctly recognizable.
This originates quickly and is followed by a comparatively
sudden expansion into a number of distinct and often widely
separated forms having very different morphic characters.
Thus, the foundations of the various stocks or branches of the
phylum are laid by quick processes of evolution, not slowly, as

BIOLOGICAL INVESTIGATIONS. 137

generally supposed. ^ Not only is this true for the smaller
phyla of animals, so far as known, but the now well-ascertained
fact that most existing types arose during earlier Paleozoic time
shows that it is probably a general law for the whole of the
animal kingdom.

This excessively plastic period of evolution is followed by a
phase of slow changes in which the different branches of the
phylum acquire their slighter and less marked divergencies.
Thus, evolution is not invariable, but moves with extreme slow-
ness while each branch or phylum is in the middle age or acme
of phylogeny. Finally, when retrogression begins, the pace is
again quicker, and the differences greater, owing to the abrupt
losses of progressive characters, and sometimes to the sudden
introduction of new and startling modifications.

Many types, even at the present time, are still in the pro-
gressive stage of advancement, but fortunately the most spe-
cialized and complex of all types, the Mammalia, has nearly, if
not quite, reached an advanced phase of retrogression, and' is
an admirable example of the quicker elaboration of differences
and the change of law in highly retrogressive types.

The record of the immediate ancestor of man is either lost
or so difficult to trace in the embryo and younger stages that
observers have failed to come to any agreement, and yet he is
plainly descended from some simian predecessor. He is not
distinct from the most specialized of this group, the chimpan-
zee or gorilla, either by his bones or soft parts, so that in the
absence of his works a paleobiologist would not separate him
widely from them. All efforts to demonstrate adequate struc-
tural characters for the separation of man from the anthro-
poids have failed. The size of the brain cannot be admitted
as adequate, or any other differences of mass. The wide gulf
between man and his nearest relatives nevertheless becomes
plainer with every effort to close it up, and it is obviously
physiological. Even if all the intermediate links between him
and some anthropoid or other simian ancestor were to be found,
as they probably will be in time, the differences would remain
just as marked as before.

1 " Phylogeny of an Acquired Characteristic," op. cit., pp. 366-371.

138 BIOLOGICAL LECTURES.

The evidence of his vast physiological powers, as shown in
his works, are in obvious correlation with his progressive char-
acters, such as the upright position and all the correlated modi-
fications like the sigmoidal outline of the back, the huge bipedal
le2:s, the differentiation of the hands and feet. The obvious
correlation of his works also with the size of the brain and
shortening up of the lower face, and other retrogressive char-
acters, as they are correctly diagnosed by several authors, nota-
bly Cope and Minot, make up a paradox which is insoluble as
long as evolution is looked upon as uniform in its action.

In types terminating phylogenetic series, however, such inter-
mixtures often occur on a smaller scale. These are not general-
ized, but highly specialized creatures, and notwithstanding their
retrogressive modifications, they are necessarily the highest in
position or grade, as determined by the facts of succession in
time and genetic position. Man, like one of these terminals,
is retrogressive in essential structures, and, like one of these
also, is separable from his nearest congeneric relatives, through
the change of law which has converted morphic retrogression
into physiological progression. The brain, although larger, is,
when compared with that of the human embryo and those of
the young of other mammals, plainly arrested in development,
as demonstrated by Cope and Minot. It maintains, in its
proportions to the mass of the body, the general characters of
the embryos of the Mammalia, but has nevertheless differen-
tiated in intimate structure and evolved into an organ of such
enormous and novel powers that its functional work has
become distinct qualitatively, as well as infinitely greater quan-
titatively, than that of the animal brain.

The disregard of researches in comparative ontogeny and
phylogeny has also led most investigators of the Darwinian
school to believe in the uniformity of action of the laws of
evolution, an error that has had far-reaching consequences.
Thus Poulton, recognizing that the sudden appearance of ani-
mal types in geologic strata was opposed to the Darwinian
theory of the slow accumulation of differentials through natural
selection, and in order to reconcile the facts with this theory,
has assumed that there must have been in existence a vast

BIOLOGICAL INVESTIGATIONS. 139

series of faunas which preceded the known Paleozoic, and in
which the earliest known fossil faunas were slowly brought
into being. The objections to this hypothesis are, not only
that evolution probably proceeded with greater rapidity at first,
and that in every period the first coming in of types was sud-
den and in accordance with the law of phylogenesis stated
above, but, what is still more convincing, the researches of
explorers do not sustain this position. The great thickness
of unaltered rocks found in the lowest Paleozoic of the Colorado
Canon, and hunted over step by step by C. D. Walcott, whose
capacity for finding fossils in the field has perhaps never been
surpassed, and the researches of C. F. Matthews, an equally
careful field worker, in the Cambrian of Nova Scotia, and Wal-
cott again in Newfoundland, as well as all other observers in
other lands, have failed in discovering any confirmations of the
supposed former existence of preliminary faunas of the extent
required by Poulton's theory. If Darwinians had not only
looked upon animals theoretically as genetic series, but put
their views into practice and worked out examples of phylo-
genesis, they would long since have ceased to regard the uni-
formitarian hypothesis as tenable.

The disregard of such researches will, I venture to suggest,
be found to affect all work upon the histology and embryology
of living animals. This must be the result if the theoretical
position taken in this paper be true, or even true in so far as
the nature of series is concerned. If series, for example, have
not been elaborated by uniform law, but by a process that is
forever changing its rate, and tending, as in the case of highly
retrogressive forms, like man, to depart so widely from the
parent type as to make their connections very difficult to
determine, it follows that all inferences with regard to phy-
logeny not based upon the comparative study of genetic series
are more or less likely to lead to erroneous conclusions.

Perhaps the only researches in the embryology of existing
animals that may be said to have been followed out in accord-
ance with this view are those that have described the serial
relations of cells in the ontogeny and have refrained from draw-
ing phylogenetic conclusions, except perhaps the most general.

140 BIOLOGICAL LECTURES.

from the data thus afforded. There is, for example, no con-
clusive evidence of the inference that the bodies of all Metazoa
are to be considered as a connected mass, and not as a colony
of cells, because the cells of the bodies of the more complex
invertebrates and vertebrates have been found to have such
close protoplasmic connections that no cellular independence
of action is possible. The highly differentiated tissues of com-
plex structures which stand high in the gradations of any
genetic series might possess in every case examined the char-
acters that have been relied upon to disprove the cell theory,
and yet the simple primitive forms of the same genetic series
have less closely connected and more independent cells even
in their full-grown stages.

Who knows, also, what the closer study of the ontogeny may
bring out with regard to the connections of cells, especially
in the building up of the earliest stages among the simplest
forms of each great group t There is, it appears to me, no
solid objection in embryology to the old theory of the rise of
the Metazoa out of Protozoa by a series of gradations that are
more or less distinctly traceable in the embryology of the
Metazoa, especially if the position taken by the lecturer is
true, that Volvox and Eudorina of the fresh waters is an inter-
mediate type, a real mesozoon, standing between the Protozoa
and Metazoa, and representing in their permanent aspect the
transient, single-layered blastula of the true Metazoa.^

Even leaving this out as inadmissible does not vitiate the
view taken by the advocates of the cell theory so far as the
obvious morphic phenomena are concerned. The evidence
that the earliest stages of the ovum in the most primitive
forms of the different classes of the animal kingdom, as a rule,
recapitulate the essential characters of the more primitive Pro-
tozoa, or, in other words, possess a reminiscent protozoanal,
unicellular stage, seems macroscopically and microscopically
demonstrated. In order to destroy the evidence of this obvi-
ous morphic basis, it would be necessary to prove that the ova

1 " Larval Theory of the Cellular Tissue," Proc. Bost. Soc. Nat. Hist., vol. xxiii,
1884, p. 151. Also "Values in Classifications of Stages of Growth and Decline,"
etc.. Am. Nat., 1888, p. 872.

BIOLOGICAL INVESTIGATIONS. 141

of selected primitive forms were not homologous with the uni-
cellular bodies of the simplest Protozoa.

I say selected, meaning thereby that an ovum taken at hap-
hazard anywhere throughout the animal kingdom would be as
likely to confuse the observer striving to solve the question of
the origin of metazoanal structures as to throw light upon this
problem. If the students of comparative ontogeny are correct,
an organism taken from among the primitive forms of any
genetic series has within itself a record of historical value pro-
portioned to its place in its own genetic line and also to the
position of its genetic stock in the whole animal kingdom, and
its evidence may be, and practically is, dependent upon its grade
or position. If it lie near the genetic base of origin, it contains
a more complete record, and recapitulates more ancestral char-
acters than if it is drawn from the acmatic period of the phy-
logeny ; and finally, if it come from the other extreme of the
phylogeny, where animals are passing through retrogressive
phases, its ontogeny may be so deficient as to fail to recapitu-
late, as in the case of the young of man, the hitherto essential
characters of the forms from which it was evolved.

This is the result of a law of development of the ontogeny,
which Cope during his life was in the habit of demonstrating
and using among recent animals, and the lecturer among fossils.
It is particularly well shown in such complete series as may be
found in the living Batrachia and in fossil Cephalopoda, and has
been independently discovered by several observers. ^

1 The following seems to the author a fair statement of the history of this law.
Haeckel, in his Morphologic der Organismen, vol. ii, 1866, p. 184, gave, under the
title of " Gesetz der abgekiirsten oder vereinfachten Vererbung," what seemed to
me for some years to be a definition of this law, but in writing my Bioplastology
I came to the conclusion (p. 78) that this was merely an emended general statement
of the law of correlation between the ontogeny and phylogeny. An abbreviated
statement of the literature seems to be as follows : Hyatt, " Parallelism between
Stages of Life of the Individual and Group of Tetrabranchiata," Meni. Bost. Soc.
Nat. Hist., vol. i, Pt. II, read Feb. 21, 1866, pubHshed 1867 (see Rep. of Custodian
of B. S. N. H., May i, 1867, and May, 1868) ; Cope, "Origin of Genera," read
October, 1868, published 1869 (^^e Account of Cope's Origin of the Fittest, p. vi.
Preface) ; Wiirtenburger, op. cit., p. 28, Leipzig, 1880 ; Buckman, op. cit., vol. for
1891, p. 290, also " Some Laws of Heredity and Their Application to Man," Proc.
Cotteswold Nat. Club, vol. x, Pt. Ill, 1891-92, p. 258 ; Ganong, W. F., " Contr.
Morph. and Ecol. of the Cactacae," read before Society for Plant Morphology and

142 BIOLOGICAL LECTURES.

It has been called the ** law of Tachygenesis, or accelerated
development," and is known as an isolated fact to most embry-
ologists under the name of " abbreviated development." It is
constantly noted by investigators, but, in consequence of their
failure to follow out their observations in serial succession
according to the natural order of phylogeny, they have not
realized that extreme cases of abbreviation are only exag-
gerated phases of a law of heredity acting more or less in all
forms.

In tracing phylogenesis by the correspondences of the
younger stages and by the gradations of the adult stages, it
has been foiind that successive, genetically connected forms
tend to inherit the characters of their ancestors at earlier
stages than those in which they first appeared in those ances-
tors. The law as observed by breeders and formulated by
Darwin for existing animals and plants is that characters of
parental forms tend to be inherited at the same time or earlier
in descendants. In phylogeny the result of this is cumulative,
and in favor of continuously increasing acceleration. Thus, the
latest acquired characteristics of any ancestral form appear ear-
lier, and ever earlier, in its descendants. In this way room is
made for still other acquired characters that are introduced later
in the lives of these same forms. In its normal course in phy-
logeny a characteristic takes up in succession different places
in the early ephebic, then passes into the neanic substages,
and finally comes into actual contact with the more invariable
nepionic characters. Then, according to the law of Tachygen-
esis, it is, as a rule, crowded out of the ontogeny and ceases to
be recapitulated. Sometimes, as shown in a yet unpublished
paper by Grabau, among Gasteropoda an apparently trivial char-
acter may pass this ordeal and actually invade the later substage
of the embryo. As a rule, however, the characters that arise
in the smaller phyla perish out of the ontogeny when they meet
the more persistent nepionic and embryonic characters.

Physiology, Dec. 29, 1897, published in Ann. of Botany, vol. xii, December, 1898 ;
and Jackson, "Localized Stages," op. cit., p. 138 (read Nov. 2, 1898, published
April, 1899), notices this law in plants. These all, except Jackson's, represent
independent rediscoveries of the law of Tachygenesis.

BIOLOGICAL INVESTIGATIONS. 143

The nepionic or baby stage has been found to have characters
derived from the more remote ancestors of the group or general
phylum of the class. It may be characteristic of a large genetic
division or order, or sometimes, perhaps, what many systema-
tists would call a class or subclass. These nepionic characters
often refer us back to earlier geologic time, or even to forms of
still unknown faunas. This brief statement is of course very
defective, but it is true in a general way and is in accord with
the common belief in the comparative invariability of embryonic
characteristics which immediately precede the nepionic in the
ontogeny. Whether this comparative invariability of earlier
ages, which forms the basis of most embryonic conclusions, be
considered as due to the similar conditions that may be sup-
posed to guard the earliest stages of every embryo among
Metazoa, or whether it be due to heredity, which necessarily
recapitulates the processes of rejuvenation derived from the
remotest ancestors more faithfully than those later acquired,
cannot be discussed here. It is, however, quite plain that
embryonic characters, as a rule, possess persistence in direct
proportion to their age, the most ancient being usually the
most persistent.

The few examples of later acquired characters that have been
traced back through earlier and earlier inheritance to the embry-
onic stages are positive so far as they go, and enable us to see
that the variations that do take place in these earlier stages are
not the result of a different law from that which governs the
inheritance of later acquired characters, but probably the same.
It is well known that in the same class, or even order, embryos
of the same genetic stock may pass from the detailed represen-
tation of many successive stages in the formation of the three
layers, and of the body cavity and other parts (ex. Amphioxus),
to a comparatively complex abbreviation of what are supposed
to be the same processes, since they have identical results in
the development of adults. But so many details have been
lost, the blastula and gastrula for example, and other charac-
ters added, the food yolk as an illustration, that it becomes dif-
ficult to see that they are really modifications of identical
modes of development. If one is prepared to try the law of

144 BIOLOGICAL LECTURES.

Tachygenesis as a working theory, it is obvious that these
exceptional or abbreviated processes can be considered as due
to that law, since it works in the same peculiar way — some-
times with simple and straightforward exactness, but often also
with a complex of introduced characters that make its transla-
tion very difficult.

A large amount of time has been wasted in useless efforts
to solve problems of phylogenesis by the aid of the embry-
ology of highly specialized and highly tachygenic forms.
These may sometimes throw light astern on their own imme-
diate ancestry, but are often too deficient in genetic recapitula-
tion to be used even in this limited way, and are worse than
useless in the solution of problems of phylogenesis in which
other genetic series are concerned. Examples of this class are
numerous, and almost every embryological work abounds with
misdirected efforts to use the early stages of such animals
either for or against opinions and theories where their bear-
ing has little or no value. This is particularly evident when
they are used to show that there are no positive general cor-
relations between the ontogeny and phylogeny, or when they
are spoken of subjectively as '* falsifications " of the embryo-
logical record, as if nature were playing a game of bluff with
scientific observers.

This law also forbids the drawing of general conclusions
from haphazard investigation, however numerous, broad, and
widespread the facts may be, and is consequently opposed to
prevailing statistical methods. These assume that, if a large
number of embryos be selected at random, the conclusion
drawn from their investigation would correct the discrepancies
and errors due to what is called insufficient data. The error
in such an assumption is that, while statistics apply, if care-
fully used, to a multitude of living forms of the same stock
{and one may safely trust them perhaps in the general meaning
of a character among animals of the same age or the same
grade), it does not apply to phylogenetic relations of animals of
different grades. The relations of these being serial, all obser-
vation of a statistical kind covering animals of different grades
must be arranged and compared upon the same basis. It seems

BIOLOGICAL INVESTIGATIONS. 145

to the lecturer, for example, that the real relations of the blas-
tula to the gastrula stage should be solved first by observations
made strictly within the limits of some one type, and these be
checked off and corrected by reference to the accepted or deter-
minable position of each animal in its own genetic stock. Also
that no conclusion could be accurately drawn from these data
until the different branches of the animal kingdom had been
reviewed in the same way, especially those in which, like the
Cephalopoda and Vertebrata, the earliest stages are highly
tachygenic and also complicated by the addition of a food yolk
or other introduced characters.

So far as the position taken in this paper would be likely to
affect such questions, a very good argument can be made for
the opinion that the gastrulas formed by the bending in of one
pole of the blastula are primitive, and other processes are
more or less tachygenic forms of this one. Whether this be
accepted or rejected does not, however, invalidate the assump-
tion made here, which is that the search for the solution of all
questions relating to the study of phylogeny can hope for suc-
cess only upon the basis of serial comparison, since this is the
only method strictly in accord with the principles of evolution
and with the methods that have been so efficient in the investi-
gations of the laws of development of the ontogeny.

All of the above remarks tend one way when considered
from a general point of view. First, in the study of organ-
isms we are dealing with plastic substances showing definite
gradations which can be arranged in their natural radio-serial
divergent lines of classification, either as they appear on the
present surface, or upon a succession of levels in past faunas.
Second, there are. notable and essential differences in the
behavior and structure of the beginnings of any single genetic
series as compared with its middle terms, and marked differ-
ences also between these same middle terms and the final or
aberrant terms, or branches of the same series. Third, it is
also asserted that the difference between the extremes and the
means of any one genetic series is proportional and usually so
great that the grade and position of an organism must be
taken into account in every phylogenetic investigation. The

146 BIOLOGICAL LECTURES.

probable penalty of neglecting this is failure in the attainment
of theoretically correct opinions, either through absence of data
that can be gathered only in this way or by the misleading
nature of the unclassified positive evidence.

Another direction in which legitimate criticism can be made
of many modern researches has still more fundamental meaning.

Cope always insisted upon the necessity of regarding the
immediate reactions of the organism as the proximal cause of
modifications. One cannot read his masterly analysis of the
phenomena of Bathmism in the American Natnralist for 1893,
under the head of "The Energy of Evolution," without realiz-
ing that he is accurately translating the two categories of the
obvious phenomena of evolution in the following passages :

" The term * energy ' is used to express the motion of matter,
and the building of an embryo to maturity is evidently accom-
plished by the movement of matter in certain definite direc-
tions." Then, after stating the wide differences between this
sort of energy, so far as it can be estimated by its results,
and other sorts of energy, known by their physical results
upon inorganisms, he divides the organic results into two
classes. "The anagenetic class tends to upward progress in
the organic sense, that is, towards the increasing control of
its environment by the organism, and towards the origin and
development of consciousness and mind. The catagenetic
energies tend to the creation of a stable equilibrium of mat-
ter in which molar motion is not produced from within, and
sensation is impossible. In popular language one class of
energies tends to life, the other to death."

The term "Catagenesis" is to me objectionable, and para-
genesis would be better ; but leaving this aside. Cope has hit
the exact explanation of the cycle in the ontogeny and phy-
logeny, and given us a true, broad picture in these few words
of the phenomena of life, and made plain that it is distinct
from inorganic phenomena because it possesses internal factors
that are the direct and immediate causes of modifications.

How many biologists keep constantly in mind this double
relationship of an organism to its surroundings, and the possi-
bility that most morphic modifications are complex phenomena,

BIOLOGICAL INVESTIGATIONS. 147

which necessarily take place through action that proceeds from
within outwards. The fact that it takes an appreciable time
for an organism to receive an impression, and to begin to react
by its own efforts in the presence of an irritating substance, or
to begin to change in response to new surroundings, and that
its corresponding movements and modifications are suitable,
shows this with a simple directness not characteristic of other
explanations of the origin of variations. It is not intended to
defend this theory, which originated with Lamarck, and was
subsequently, but independently, restated by Cope and the
author of this lecture, but simply to point out some of the
inevitable failures of reasoning due to the neglect of the con-
sideration of the internal reactions of organisms.^

Experimental biologists are very apt to neglect this factor in
their conclusions, and consequently either invert the natural

1 The general impression that neolamarckism implies belief in action of the
surroundings is true only so far as these act as stimuli, or may be proved to act
directly upon organisms. Neolamarckism, or, better named, the Dynamical School
of Biology, does not reject the Darwinian hypothesis when the struggle for exist-
ence or selection of any sort is recognized as a result of evolution, as a secondary
cause tending to the preservation and perpetuation of differences after these have
been originated by fundamental causes. The dynamical hypothesis can be shown
to be inoperative only by proving that the internal mechanism of motion in
organized beings is ineffectual to produce variations. It deals with the imme-
diate origin of variations through internal reactions, and is really open to any
explanations of the subsequent conduct and preservation of these that can be
proven.

Since this lecture was delivered, I have ha4 the pleasure of receiving a very
profound and interesting printed lecture of Dr. C. O. Whitman, on Animal
Behavior, biological lecture at Woods Holl Laboratory, 1898. Although he takes
up a position adverse to those advocated in this lecture, this author really does
seriously consider some of the neglected theories usually omitted by students of
existing biological phenomena, or neobiologists, and so far his lecture meets the
general criticisms offered here. It appears to me, however, after reading this
and other essays, that a large part of the differences between the schools that
believe in the inheritance of acquired characters lies in the different definitions
of what is meant by acquired. Again it is, perhaps, as stated by Whitman, very
difficult to find a character, or instinct or habit, which will meet his requirements
among the habits of existing organisms. Habits "having no hereditary basis
predisposing to them" (Whitman, p. 314) are as hard to find in existing complex
or fixed organisms as characters having no similar basis. My own definition of
an acquired character is that it is not present in the more primirive forms of the
same phylum, and tends to occur first in a comparatively late stage of the ontogeny
(see " Phylogeny of an Acquired Characteristic," op. cit).

148 BIOLOGICAL LECTURES.

order of causation in their conclusions, or give only more or
less partial statements of their results.

The most famous case is that of Schmankewitsch's experi-
ments on Artemia and Branchipus, which you will, I know,
pardon me for sketching over hastily in order to give weight
to this criticism. This Russian worker on one occasion, after
the accidental breaking down of the dams that defended the
salt pans near Odessa from the uncontrolled influx of the adjoin-
ing brackish waters, observed that the flood brought with it Arte-
^nia mulhauseni in great numbers, and also destroyed A. salina
of the dense salt water previously occupying the pans. It
struck this keen observer that the conversion of A. ^mdJiauseni
into A. salina might be determined by following closely the
changes that would take place after the dam was repaired, and
the conversion of the outside brackish water into dense salt
water was begun by evaporation. He not only followed these
changes successfully, but he also carried on a parallel series of
experiments in aquaria.

In both the natural and artificial experiments A. mulhau-
seni of the natural brackish waters changed into A. salina of
the denser salt waters of the pans after evaporation. Then
Schmankewitsch, by inverting the process and diluting the salt
water in his aquaria, succeeded in not only producing A. mul-
hauseni again, but proceeded farther, and by decreasing steadily
the amount of salt in the solution, evolved artificially Branchipus
of the fresh waters. The modifications of these crustaceans were
mainly produced in the swimming organs, especially the form
and hairiness of the end of the abdomen, and also the final
appearance of an additional segment in the abdomen of Bran-
chipus. Schmankewitsch at first attributed these modifications
wholly to the effects of the proportionate amount of salt in the
water as determined by its density, and subsequently appears
to have modified this opinion, allowing that temperature and
proportionate amount of oxygen contained in water of different
densities might have also been factors in producing the varia-
tions observed.^

1 Zeitschr.f. wiss. ZooL, vol. xxv, Supplement, 1875, P- ^^^5 ^^^ Ibid., vol. xxix,
1877, p. 429.

BIOLOGICAL INVESTIGATIONS, 1 49

Brauer of Vienna, accustomed to the study of the adaptation
of organs among insects, and the author of this lecture, accus-
tomed to look for internal factors from a more theoretical point
of view, agree in criticising these results as defective, because
they did not take into account the fact that the amount of salt
may have had no direct effect, except in so far as it increased
the density of the fluid in which the animals moved. The fact
that the modifications took place mainly in the abdomen, and
that the increase in hairiness, and in length and size of this
part, and variations of the appendages as the water decreased
in density, can be just as probably attributed to the extra
growth caused by increased efforts necessary for breathing
and swimming in a more rarified medium. Whether this
objection be true, or the reverse, does not matter here, since
it is simply introduced to show the need of considering the
organism as a vitalized body which is capable of reacting from
within appropriately in response to external stimuli. Experi-
ments that do not take this essential condition into serious
consideration are open to similar objections.

In order to prove a physical cause acting wholly and directly
from the outside, the internal reactions of the organism should
be shown to be passive, as in the cases of chemical reaction. As
in Schmankewitsch's investigation, the conclusions are other-
wise open to the objection that they have entirely disregarded
the consideration of fundamental elements, and that the physi-
cal causes they cite have only set in motion a chain of internal
reactions, which are the immediate causes of the observed
modifications. Under these conditions, also, strong doubts
arise as to the efficiency of the physical causes thus demon-
strated, since the mode of experimentation does not exclude
the possible existence of other causes that might set the chain
of internal reactions going in the same or similar directions.

Dr. H. S. Jennings's masterly studies "On the Reactions to
Stimuli in Unicellular Organisms " is one of the very few inves-
tigations not affected by such criticisms. These clearly estab-
lish the fact that a multiplicity of causes may bring about
precisely similar reactions on the part of several of the ciliated
Protozoa, like Paramecium, and that it is the organization and

150 BIOLOGICAL LECTURES.

not the stimulus that determines the functional results. " Cer-
tain agents set up a reaction in the animal, the directive features
of which depend entirely upon the structure of the organism." ^

This author's conclusion is particularly interesting both in
its originality and its bold denial of the conclusions of other
observers, and the support it gives to the opinions expressed in
this paper with regard to the necessity of considering evidence
of all kinds from the point of view of the position or grade of
the animal. Speaking of one of the most highly organized of
the Protozoa, Paramecium, with which he has become very
familiar by long-continued study, he says : "■ An animal that
learns nothing, that exercises no choice in any respect, that is
attracted to nothing and repelled by nothing, that reacts en-
tirely without reference to the position of external objects, that
has but one reaction for the most varied stimuli, can hardly be
said to have made the first step in the evolution of mind, and
we are not compelled to assume consciousness or intelligence
in any form to explain its activities." If Jennings is correct,
this highly differentiated protozoon is a zero point in the evolu-
tion of consciousness. Whether, however, this is due to the
essential character of its protozoanal or unicellular structure
or to the loss of functional power cannot be asserted until
the simpler forms of Protozoa have been studied in the same
thorough way and with corresponding tests of their internal
reactions.

The importance of these researches, if confirmed and estab-
lished, can hardly be overestimated, since they open the way
for experimental observation upon the evolution of intelligence,
and, if found to be true in simpler organisms, will show that
automatic habits come first in the evolution of consciousness,
and are not, as often supposed, invariably the results of repeated
conscious actions.^

A number of other recent papers and books have shown that
psychology has its obvious place in the study of organisms, but

1 «« Psychology of a Protozoon," Am.Journ. of Psychol.^ vol. x, No. 4, 1899, p. 513.

2 This does not imply that automatism is not a retrogressive result of the
constant repetition of conscious actions, but simply that it is also a primitive or
antecedent stage in the progressive evolution of consciousness.

BIOLOGICAL INVESTIGATIONS. 15 I

the fundamental importance of this science is not suspected by
most investigators. Nevertheless, it has been placed in its
proper relations by a number of writers, beginning, perhaps, in
definite modern form with Ewald Hering's paper, entitled,
" On Memory as a General Function of Organized Matter,"
an address given before the Imperial Academy of Science at
Vienna in 1870. Ribot took substantially the same ground
in his work on "Heredity," in 1876, and Haeckel also in "Die
Perigenesis der Plastidule," 1876, both being ignorant of Her-
ing's previous publication and of each other's results. These
were followed by an equally independent restatement by Cope
in the American Naturalist in 1889, and the lecturer came near
making a fourth in this list. Speculations upon the supposed
mysterious and indefinable nature of heredity might have had
a far more objective basis for the past twenty-nine years had
naturalists, whether in opposition or in approval, considered
Hering's theory worthy of their attention.

Confusion exists in the minds of many naturalists, arising
from the ordinary conception of memory as a function of con-
scious intellectual effort and as something essentially distinct
from automatic action and habits of all kinds. No essential
distinctions actually exist, however, since the two grade into
each other.

The essential common characteristic of both is repetition ;
and to represent this and avoid the risk of employing an old
word whose application had to be greatly extended, I have in
previous publications used the term "mnemism," or"mnemo-
nism," to express the nature of heredity, and " mnemogenesis "
for the theory. Before proceeding further in the argument of
this question it is necessary to glance hastily at the supposed
modes of transmission of characters.

These can be divided into two classes : (i) those theories
that depend upon some peculiar substance supposed to have
always been distinct from the surrounding forms of protoplasm.
These necessarily assume some sort of corpuscles or organic
molecules to be the vehicles of transmission. (2) Those theo-
ries that regard the germ plasm as not transmitted from body
to body but as arising out of the ordinary protoplasm of the

152 BIOLOGICAL LECTURES.

young by perpetual rejuvenation and subsequent differentia-
tion, and therefore in no sense a peculiar substance.

Every purely corpuscular theory, meaning thereby the phys-
ical transmission of gemmules, biophors, pangenes, or any other
supposed organic bearers of characteristics, must not only
account for a difficulty as great as that of the camel and the
needle's eye, but must also account for putting the numberless
characters, derived from the entire caravan of its immediate
progenitors and remote wild ancestors and their progenitors,
back to the origin of the phylum through the same narrow
tunnel.

One has to imagine the corpuscles and all this active circula-
tion and concentration taking place invisibly, and yet requiring
visible vehicles of transmission in the minute spermatozoon and
nucleus of the ovum. Then he must picture their redistribution
over the body of the offspring, the larger number remaining
latent until the proper time arrives for them to develop, and
then locating themselves and coming out in exactly the right
place, or repeating at the right time some tendency or habit of
the ancestors.

One is naturally led to ask how these particles acquired their
peculiar tendencies which force organisms along predetermined
paths of development, reproducing similar successive stages of
development with automatic regularity and repeating with more
or less precision the permanent stages of their ancestors } In
what way did they acquire their wonderful ability to remain
latent until they had reached the proper place, and then measure
time so as to wake up and go to work at the right moment }
The corpuscular hypotheses provide supposititious vehicles of
transmission, but they do not explain how or why these move
in their apparently predetermined cycles of change.

Nussbaum's and Minot's^ theory of panplasm or the con-
tinuity of germ plasm seems to be a necessity, if the transmis-
sion of germ particles or of organic atoms is rejected. Minot's

1 Nussbaum, " Differenzining d. Geschlechts in Thierreich," Arch. f. mikr.
Anat., Bd. xviii, 1880, pp. 1-121.

Minot's " Vererbung u. Verjiingung," Biol. Centralb., Bd. xv, 1895; and ibid.,
American Naturalist, vol. xxx, 1896, pp. 1-9, 89-101.

BIOLOGICAL INVESTIGATIONS. 153

decision that this substance is not a peculiar plasm but a dif-
ferentiated product of each organism undergoing a process of
rejuvenation as each link in the chain of being is forged, seems
more in accord with the facts than that which demands a wave-
less linear transmission of some peculiar substance.

Neither of these theories excludes the conclusion that heredity
of the transmission of like by like is a function of the automatic
repetition of the same act, or, in other words, identical with
what we call unconscious memory ; but it is obvious that if
heredity can be expressed by the term '' mnemism," the greatest
difficulties are removed from the path of the theory of pan-
plasm. This last cannot provide of itself any reasonable
explanation of the successive stages of development and their
recapitulations of ancestral characters, but with the aid of
mnemogenesis it can fully explain these phenomena.

Professor Cope has insisted upon the truth of this theory in
two essays. The following are quotations from his article
"On Inheritance in Evolution":

If the doctrine of kinetogenesis be true, this energy (the build-
ing energy) has been moulded by the interaction of the living being
and its environment. It is the expression of the habitual movements
of the organism which have become impressed on the reproductive
elements. It is evident that these and the other organic units of
which the organization is composed possess a memory which deter-
mines their destiny in the building of the embryo. This is indicated
by the recapitulation of the phylogenetic history of its ancestors dis-
played in embryonic growth. This memory, perhaps, has the same
molecular basis as the conscious memory ; but, for reasons unknown
to us, consciousness does not preside over its activities. The energy
which follows its guidance has become automatic, and it builds what
it builds with the same regardlessness of immediate surroundings as
that which is displayed by the crystallific growth energy. It is inca-
pable of a new design. It appears to me that we can more readily
conceive of the transmission of a resultant form of energy of this
kind to the germ plasma than of material particles of gemmules.
Such a theory is sustained by known cases of the influence of mater-
nal impression upon the growing foetus. Going into greater detail,
we may compare the building of the embryo to the unfolding of the
record of a memory which is stored in the central nervous organism

154 BIOLOGICAL LECTURES,

of the parent, and impressed in greater or less part on the germ
plasma, in the order in which it was stored. The basis of memory
is reasonably supposed to be a molecular (or atomic) arrangement
from which can issue only a definite corresponding mode of action.
That such an arrangement exists in the central nervous organism is
demonstrated by automatic and reflex movements.^

Any naturalist who has studied the sudden recurrence of
habits in animals, usually accounted for by the use of the mean-
ingless term "instinct," will find that this is no figment of the
imagination, but a working hypothesis that may be used and
tested.

Although Ribot's work on "Heredity" has not been widely
read or accepted by scientists, perhaps because he admits many
facts as substantial evidence which have not been verified, it is
nevertheless the most profound work on this subject that has
been published. This learned authority, after careful investi-
gation, came to the same startling conclusion that memory was
the only function of organisms which could be closely compared
with heredity, and he brings forward a large number of facts,
to which many others can be easily added. He says " heredity,
indeed, is a specific memory; it is to the species what memory
is to the individual. Facts will hereafter show that this is no
metaphor."

The authors quoted above have not discussed memory, as
has been commonly supposed by naturalists, in a metaphysical
sense, but as an organic function arising from voluntary or
involuntary repetitions of conscious or unconscious actions.
Ribot indulged in general speculations and was hampered by
them to a notable extent, but he struck the keynote of the
dynamical theory of evolution. Thus he say^ : " Every act
leaves in our physical and mental constitution a tendency to
reproduce itself, and whenever this reproduction occurs the
tendency is strengthened ; and thus a tendency, often repeated,
becomes automatic. This automatism is the link between
memory and habit, and gave rise to the saying that memory is

1 American Naturalist, vol. xvi, 1882, pp. 454-469; reprinted in The Origin of
the Fittest, pp. 405-421, and " On Inheritance in Evolution," American Naturalist,
vol. xxiii, 1889, pp. 1058-1071.

BIOLOGICAL INVESTIGATIONS. 155

only a form of habit — a proposition which, with some restric-
tions, is true." ^

Repetition, or the reproduction of parallelisms, is equally
characteristic of memory and of heredity, nor can either be
conceived of as having a tendency to produce variations. It is
entirely reasonable that any novel acquired habit, due to con-
scious effort or to involuntary reactions of organisms in the
presence of external stimuli, may be regarded as one of the
products of memory. It follows from this that any structural
modifications which may result from the repetition of the same
acts or habits can, with equal reason, be attributed to memory.
The tendency of descendants to perform the same action, i.e.,
to manifest the same habit in the presence of similar stimuli,
when no special structure has been originated, can thus be
readily accounted for if one considers heredity to be one form
of organic memory, or mnemism.

When special tendencies or structures have arisen, their
reappearance in descendants at the same time, or earlier, does
not present the same difficulties that it seems to place in the
path of other theories, but becomes one of the strongest
confirmations of the mnemogenic hypothesis.

In mnemonics it is the machine-like regularity of the succes-
sion of cause and effect, of one word begetting the next, that
surprises the student, the recurrence in the mind of long-
forgotten words, languages, and scenes, either through the
recurrence of some inciting cause or upon the removal of inter-
fering causes, as in the recapitulation of youthful reminiscences
in aged persons. Mnemotechnics, as embodied in the various
systems which have been devised to cultivate the memory, con-
sists essentially in teaching the habit of forming a chain of
associated ideas or words leading up to the word or thought to
be recalled.

Even reversions that may be supposed to be purely sporadic
do not oppose any serious obstacle, since there must always be
latent mnemism in the cells and organs ready to manifest itself
whenever more recently acquired characters are prevented from
being developed in their proper succession. This is the most

1 Heredity. Translation. New York, 1876.

156 BIOLOGICAL LECTURES.

frequent form in which reversions are found among Ammonitinae.
There are, of course, difficulties in the path of this hypothesis,
such as the so-called premature development of characteristics
and some other rare phenomena interfering with the regular
recapitulation of ancestral characters in the ontogeny. These
at present can only be explained by assuming an irregularity in
the action of mnemogenesis, due to disturbing causes, but this,
even as a provisional suggestion, is not satisfactory without
adequate statement of the nature of these causes in specific
cases.

This subject has been treated of at greater length in other
publications, and these remarks are repeated here only to call
attention to a theory of heredity which has obvious claims to
be considered by embryologists. I have, however, looked in
vain in the numerous works upon embryology and containing
also remarks upon heredity, as well as most of those treating
more directly with heredity itself, that have come under my
eyes in the last twenty-three years, and this is certainly not an
omission that tends to make such investigations more valuable
to students.

TENTH LECTURE.

ON THE DEVELOPMENT OF COLOR IN MOTHS
AND BUTTERFLIES.

ALFRED GOLDSBOROUGH MAYER.

If one examines the wing enclosed within the wing-case of
a young lepidopterous pupa, it will be found to be transparent
and of a glassy clearness. A few days before the insect is des-
tined to emerge, however, the wings become pure white. This
white soon deepens into a dull yellow or ochre color, and then
the mature colors begin to make their appearance. They are
faint at first, but gradually deepen into the color of the mature
insect. They first appear in places near the center of the
wings, upon the interspaces between the nervures, not upon
the nervures themselves. Indeed, the nervures are the last
places to acquire the mature coloration.

If we wish to comprehend the meaning of these color changes
which affect the pupal wing, we must first study the develop-
ment of the wing from its condition in the young larva up to
the time when the insect is about to emerge from the pupa.
In the caterpillar the future wings consist of small infolded
hypodermal pockets, which are situated upon the sides of the
second and third thoracic segments. The walls of these little
bag-like pockets are composed of closely crowded spindle-shaped
cells, and their inner cavities are penetrated by tracheae. When
the larva changes into a pupa, the wings expand to about sixty
times their former area, and, as a consequence, the cells which
compose the wing-bag, being no longer crowded, flatten out

157

158 BIOLOGICAL LECTURES.

into a pavement epithelium. The inside of each wing-bag is
hollow and contains tracheae, blood cells, and fat cells. At
first all of the cells composing the wing-bag are similar, each
to each, but soon we notice that at regular intervals some of
the cells are becoming larger than the others and are begin-
ning to send out each one a process, which extends outward
beyond the general level of the wing -surface. These cells
are destined to form the scales. The process which arises
from each scale-forming cell becomes larger and flattens out
into a spatula shape which exactly resembles that of the future
scale. Then the cell secretes a layer of chitin over its outer
surface, and thus the scales are formed. The scales are now
completely filled with protoplasm and are as transparent as
glass ; but soon the protoplasm begins to shrink away, leav-
ing them little hollow, chitinous bags, which diffract the light
and are therefore white in color. Soon after the scales have
become hollow the blood or haemolymph of the pupa enters
them, and they become dull yellow in color ; and then the
mature color begins to appear, and gradually deepens until it
acquires the tint of the imago. The pigment, then, is formed
within the scales during the time when they are filled with
haemolymph, and the natural inference is that the color is
derived from the haemolymph, or from some of its constitu-
ents, by unknown chemical processes. It is interesting to
notice that Friedmann ('99) finds that in Vanessa urticae the
scale-forming cells, and hypodermal cells, and also the blood
cells within the haemolymph, are filled with small fatty gran-
ules which stain deeply in Hermann's fluid. These fatty gran-
ules enter the scales and contribute toward the formation of
the color. No such granules were observed by me in 1896,
when studying the scales of Danais plexippus and other forms ;
but all of my material was killed in Perenyi's fluid or in corro-
sive sublimate, and Friedmann observes that corrosive sublimate
fails to demonstrate the fatty globules. It is probable, there-
fore, that my observations are wrong in this respect.

If it be true that the colors of the mature wing are derived,
by various chemical processes, from the haemolymph, or from
some of its constituents, then one ought to be able to manu-

COLOR IN MOTHS AND BUTTERFLIES.

59

facture various colors from the haemolymph by means of
chemical reagents. Also if the color so manufactured be simi-
lar to some color upon the mature wing, it may be expected to
present reactions to chemical reagents similar to those of the
color on the wing. As far as my experiments go I find this to
be the case. For example, if one treat the haemolymph of
Samia cecropia with warm concentrated HNO3, it congeals
into a deep chrome-yellow mass. If, now, ammonia be added
to excess, it changes to a reddish-orange, which is very similar
in color to the reddish-orange band that crosses the upper surface
of the hind wings of the moth. Now this reddish-orange band of
the moth is changed to chrome-yellow by HCl or HNO3, and
then if ammonia be added, the original color reappears. Exactly
the same sequence of reactions is produced with the reddish-
orange pigment derived from the haemolymph. A number of
other tests of this nature were tried, and all gave confirma-
tory results. The details of these tests may be found in the
author's ('96) paper on the development of wing-scales and
their color.

In 1864 Landois observed that when the blood of insects is
allowed to dry in the air it is very apt to assume a color which
is similar to the ground-color of the wings of the mature insect
from which the blood is drawn. It is well known that the
most universal colors of the more lowly organized moths are
the drab-gray and yellow-drab tints, and these colors may be
derived from their haemolymphs by mere exposure to the air.
The brilliant reds, yellows, etc., are the result of more or less
complex chemical processes, which have been slowly effected,
probably through the agency of natural selection. It is inter-
esting to notice that Hopkins finds that the white pigment
found in the scales of Pieridae is uric acid, and that the yellow
and red pigments of the same butterflies are due to the deriva-
tives of uric acid. Also Griffiths ('92) has shown that the green
pigment of several species of butterflies is due to a complex
derivative of uric acid. Indeed, the haemolymph of the pupa
is itself very complex, and it seems probable that various
substances, some of them highly colored, might be derived
from it by chemical processes. In the Saturnidae, for example.

l6o BIOLOGICAL LECTURES. ■

the haemolymph consists of egg-albumen, globulin, fibrin, xan-
thophyll, orthophosphoric acid, iron, potassium, and sodium.

When we consider the remarkable complexity of form and
diversity of color found in the scales of Lepidoptera, it appears
that some very potent factor must have operated to maintain
them so universally upon the wings of these insects. In the
first place, it is interesting to know whether the presence of
scales upon the wing serves to aid the insect in its flight. In
the act of flying it is obvious that the wing has two chief func-
tions to perform, i.e., it must beat downward against the air,
and in the performance of this action a relatively high coeffi-
cient of friction between the air and the wing may be of advan-
tage. But in addition to this, the wing must glide through the
air, and in doing this a small coefficient of friction is desirable.
There must, therefore, be an optimum coefficient of friction
which lies somewhere between these two. I tested the coeffi-
cient of friction between the wing and the air in the case of
several species of moths and butterflies, the wings being
mounted upon a short, light pendulum, and the damping of its
oscillations carefully observed. By this method it was found
that the coefficient of friction of the wings remained unaltered,
even when all the scales were removed from the wing mem-
branes. We may therefore conclude that the presence of the
scales does not aid the insects in flight, and it therefore seems
probable that the wings of the scaleless ancestors of the Lepi-
doptera had already attained to the optimum coefficient of
friction before the scales appeared, and that this coefficient
remained unaltered during the time when the scales spread
over the wing membranes.

As the scales do not aid the insects in flight, it appears that
they must have been maintained and developed because their
presence granted some other advantage to the insects, probably
mainly because they displayed colors which may have served to
aid the insects in their struggle for existence.

If one examine the colors of the wings of Lepidoptera by
means of Maxwell's color disks, it is found that nearly all
of the colors are quite impure ; that is to say, they contain a
considerable percentage of black. For example, the white

COLOR IN MOTHS AND BUTTERFLIES. l6l

ground-color of the wings of Peiris rapse consists of sixty per
cent white, seventeen per cent black, thirteen per cent emerald
green, and ten per cent lemon yellow. Furthermore, a spectro-
scopic examination of the colors shows that they are generally
compound, i.e.^ composed of several distinct colors. It appears,
then, that natural selection has not been very severe with color
in Lepidoptera, for the colors are neither very pure nor intense.

When we examine the color patterns displayed by the wings
of butterflies and moths, it appears that the spots show a strong
tendency to be bilaterally symmetrical, both as regards form
and color ; and the axis of symmetry for each spot is a line
passing through the center of the interspace in which the spot
is found parallel to the trend of the nervures. A good illus-
tration of this law is afforded by the case of "eye spots,"
where the center of the spot is always at the center of the
interspace in which the spot is found. This was pointed out
by Bateson in 1894. In addition to this law we notice that
spots of similar form and color tend to form a row by appear-
ing in homologous places in a series of adjacent interspaces.
A study of the color patterns shows that they follow these sim-
ple laws quite rigidly, and that there is surprisingly little
diversity among them. Natural selection has not been severe
with color patterns in these respects, and as a consequence
many patterns which we could easily picture to ourselves are
never seen in nature. For example, a very striking color effect
might be produced were a series of adjacent interspaces to dis-
play alternate colors, say red, yellow, red, yellow, red, yellow,
etc. ; but this is never seen in nature.

Considering the color patterns from another point of view,
however, we cannot fail to be struck with wonder and admira-
tion at the remarkably close resemblance between the wings of
many moths and their surroundings. And the reason for this
appears to be that such resemblance affords protection to the
insect possessing it, and that natural selection has perfected it
until now we find moths whose outspread wings almost exactly
match the colors of the bark of the tree upon which they rest,
etc. Even more remarkable are the numerous cases where an
edible species resembles almost exactly the form and color of

1 62 BIOLOGICAL LECTURES.

an immune species, which may belong to a different order or
family. Bates, in his classic paper upon the " Heliconidae of
the Amazon Valley," first called attention to this remarkable
fact and designated it very aptly as ''mimicry." A still more
remarkable mimicry is seen where one immune species resem-
bles another immune species. For many years the explanation
of this curious form of mimicry remained a puzzle to naturalists,
until Fritz Miiller in 1878 showed that, if it be true that young
insectivorous birds have to find out by experiment what insects
are good to eat and what ones are not, and if they remember
to avoid evil-tasting species after once having tried them, then
it becomes of advantage to both species of butterflies if one dis-
tasteful insect comes to resemble another. But we have not time
to discuss this interesting topic. Suffice it to say that a series
of careful and elaborate experiments, carried out by Mr. Frank
Finn upon insectivorous birds of India, demonstrate that birds
do remember to avoid retasting a noxious insect after once
having had experience with it, and that most birds soon learn
to beware of brightly colored butterflies which have a disa-
greeable taste.

In many species of moths and butterflies the males are quite
different in color from the females of the same species. Dar-
win sought to explain this upon the hypothesis that the females
exercised selection in choosing their mates, only males of a cer-
tain shade of color being popular, and that thus the males came
to depart more or less widely from the females in coloration.
With this hypothesis in view I have very recently carried out
a series of experiments upon our common Callosamia prome-
thea. In this moth the female is ferruginous, while the male
is black. As is well known, the males of this species display
a remarkable aptitude in seeking out the females, and they
always fly up against the wind to the place where the female
is resting. My experiments show that the males are not at all
attracted by the mere sight of the female, but that they are
attracted by some substance which emanates from the abdomen
of the female. Females about six hours old are not as attract-
ive as are females thirty hours old, and dead or dying females
are not visited by males. Also, if the female be deprived of

COLOR IN MOTHS AND BUTTERFLIES. 163

her eggs, she no longer attracts males to her. A male will,
readily mate with the females when his eyes are blinded by a
covering of black glue or pitch. The male will also fly toward
the females with normal eagerness even when his entire abdo-
men is cut off. If his antennae be covered with balsam, glue,
pitch, or flour paste, he no longer seeks the females and will
neither mate with them nor display any excitement when
brought into their immediate presence. If the glue or flour
paste be dissolved off, however, in water, he immediately flies
towards the females, as would a normal male. The males will
readily mate with a female upon which male wings have been
glued so that she presents the appearance of a male. They
will also mate with normal eagerness with a female deprived
of wings and having all the scales removed from her body.
The males will fly to a piece of raw cotton within which a
female is hidden, and will display intense excitement and grasp
the cotton eagerly with their abdominal claspers. Conversely,
the females will readily favor a male upon whom female wings
have been glued so that he presents the external appearance of
a female. Also they display no objection to a male who has
been deprived of wings and has had all of the scales removed
from his body.

As far as this moth is concerned, we are apparently justified
in concluding that the male is merely positively chemotactic
toward the female, being attracted by some substance which
emanates from her. He is not attracted by the sight of the
female, nor does the female care anything for the appearance
of her consort. In this case, therefore, Darwin's hypothesis
of sexual selection is not supported, and we must attempt to
explain the melanic appearance of the male moth on other
grounds.

In this connection it is interesting to notice that F. Plateau
maintains that insects are attracted far more by the odor of
flowers than by their color.

164 BIOLOGICAL LECTURES.

LITERATURE.

'62 Bates, H. W. Contributions to an Insect Fauna of the Amazon

Valley. Trans. Litin. Soc. London. Vol. xxiii, pp. 495-566,

Pis. LV, LVI.
'89 Bemmelen, J. F. van. Ueber die Entwicklung der Farben und Adern

auf den Schmetterlingsfliigeln. Tijds. der Nederlandischer Dierk.

Vereen. Ser. 2, Deel ii, pp. 235-247.
'95-97 Finn, F. Contributions to the Theory of Warning Colours and

Mimicry. Journ. Asiatic Soc. Bengal. 1 895, vol. Ixiv, pp. 344-356 ;

1896, vol. Ixv, pp. 42-48; 1897, vol. Ixvi, pp. 528-533; vol. Ixvii,

pp. 613-668.
'99 Friedmann, F. Ueber die Pigmentbildung in den SchmetterHngs-

fliigeln. Arch.f. mikr. Anat. Bd. liv, pp. 88-95, Taf. VI, Figs. 1-5.
'92 Griffiths, A. B. Recherches sur les couleurs de quelques insectes.

Compt. Rend. Acad. Sci. Paris. Tome cxv, p. 958.
'93 Haase, E. Untersuchungen iiber die Mimicry auf Grundlage eines

natiirlichen Systems der Papilioniden. Bibliotheca Zoologica. Heft 8,

Theil i, Taf. VI, 120 pp.
'96 Hopkins, F. G. The Pigments of the Pieridae. Philos. Trans. Roy.

Soc. London. Vol. clxxxvi, pp. 661-682.
'98 Linden, M. von. Untersuchungen iiber die Entwicklung der Zeichnung

des Schmetterlingsfliigels in der Puppe. Zeitschr. f. wiss. Zool.

Bd. xlviii, pp. 1-49, Taf. I, II.
'96, '97 Mayer, A. G. The Development of the Wing Scales and their

Pigment in Butterflies and Moths. Bull. Mus. Comp. Zool., Harvard

College. 1896, vol. xxix, pp. 209-236, 7 pis. On the Color and

Color Patterns of Moths and Butterflies. Ibid., vol. xxx, pp. 169-256,

10 pis.
'79 MiJLLER, F. Ituna und Thyridia: Ein merkwiirdiges Beispiel von Mim-
icry bei Schmetterlingen. Kosmos. Bd. V, pp. 100-108, Figs. 1-4.
'97 Plateau, F. Comment les fleurs attirent les insectes. Bull. Acad.

Roy. Sci. Belgique. Tome xxxiv, pp. 601-644, 847-880.
'89 ScHAFFER, C. Beitrage zur Histologic der Insecten. Zool. Jahrb.,

Morph. Abth. Bd. iii, pp. 61 1-652, Taf. XXIX, XXX.
'57 Semper, C. Ueber die Bildung der Fliigel, Schuppen und Haare bei

den Lepidopteren. Zeitschr. f. wiss. Zool. Bd. viii, pp. 326-339,

Taf. XV.
'91 Urech, F. Beobachtungen iiber die verschiedenen Schuppenfarben und

die zeitliche Succession ihres Auftretens. Zool. Anzeiger. Bd. xiv,

pp. 496-473-

ELEVENTH LECTURE.

THE PHYSIOLOGY OF SECRETION.

A. MATHEWS.

When a substance contained in protoplasm leaves it, it is
said to be secreted, or separated from it. As all forms of pro-
toplasm are constantly giving off matter of one kind or another,
all cells are true secreting cells. Such secretion is an attribute
of all living matter.

The word " secretion," as it is ordinarily used, however, is
made to designate something more than this simple process of
separation ; it is intended to include also the process of manu-
facture by the cell of the substances thrown off. It is unfor-
tunate that the same name has been given to two such different
processes as those of secretion and manufacture, and it is not
surprising that this has resulted in great confusion. In the
present paper we shall deal solely with the separation of sub-
stances from the cell. I shall use the word '^ secretion " to mean
this and nothing more, unless, as will be specially stated, it be
used to cover the discharge from the gland duct as well. The
process of the manufacture of substances by the cell I have
elsewhere proposed to call "hylogenesis," literally meaning
the formation of substance. This process will not be specially
considered here.

How do the substances formed by protoplasm escape from it ?
This is* accomplished in a variety of ways, depending upon
whether the substances are of small molecular size and soluble,
or whether they are complex bodies. Soluble substances of
small molecular size escape from the cell by osmosis. Such is
the case, for example, for all gases, such as hydrogen, sulphu-

i6s

1 66 BIOLOGICAL LECTURES.

reted hydrogen, carbonic anhydride, and oxygen ; for urea,
nearly all inorganic compounds, the so-called internal secretions
of organs of the body, such as sugar, hippuric acid, ammonium
lactate, and all the simpler products of the cell's activity.
Substances of this kind leave the protoplasm because their
tension in the protoplasm is higher than their tension in the
surrounding medium. Such substances are secreted by all
cells. Since they are of such small molecular size, they readily
pass from the cell as rapidly as they are formed. They never,
under ordinary circumstances, accumulate in the cell. We
never find any great accumulation of urea, carbonic anhydride,
the internal secretions of such bodies as the supra-renals, or
free sugar in the cells themselves, unless the tension of these
substances in the surrounding medium is also high. Osmosis
is hence the commonest mechanism of secretion. It is found
in all cells of all kinds.

But besides these substances of small molecular size, the
cell also produces substances of a much more complex kind,
such, for example, as the albumins, mucin, glycogen, starch,
proteids, fats, etc. These substances are either altogether or
almost incapable of osmosis. Special mechanisms must hence
be devised to eject such substances from the cells, if it be
desirable that they be ejected. To accomplish this end, special
devices have been developed, differing in different cells. Cells
which possess these special mechanisms are those generally
designated as "secreting" or "glandular-cells."

The commonest method to eliminate such substances is to
break them up first into soluble and osmosable substances.
They then escape from the cell by simple osmosis. This is
the method by which starch and glycogen are removed. They
are first converted into a simple sugar. Probably the stores of
albuminous substances laid up in the fleshy cotyledons of the
leguminous plants are removed in the same way, and trans-
ported in the form of these simple fragments to tissues needing
them. Observation shows that during germination large quan-
tities of the fragments of proteid molecules are set free, such as
asparagin, leucin, arginin, etc. In the case of muscle the complex
muscle substance appears to be broken down to such bodies as

THE PHYSIOLOGY OF SECRETION. 167

carbonic anhydride and ammonium lactate. Just how the fats
escape from such cells as those of adipose tissue is unknown,
but from analogy with the process of absorption of fats, it is
not impossible that they are first saponified or broken into
soluble glycerines and soaps. Even the ferments appear to be
formed in the same manner. They exist normally in the cell in
a much less soluble form, and are apparently broken up into
active ferment fragments only at the time of secretion. Be-
sides these cases in which the fragments are small enough to
be removed by osmosis, there is hardly a doubt that in all gland
cells this process of splitting up complex molecules into simpler
ones precedes secretion, even if the process does not normally
go far enough to produce an osmosable product. This process
of decomposition increases the number of soluble molecules in
the cell, and increases thereby its osmotic equivalent. The
cell, in consequence, grows by the absorption of water and
takes on the swollen appearance of many so-called loaded gland
cells. This incidental result of increased cellular turgor is
utilized, as we shall see, in certain processes of secretion.

The second manner in which these complex substances may
be eliminated from the cell is by the budding off of a portion
of the cell containing them. Actinophrys not infrequently
detaches small portions of its protoplasm, which crawl about
for a time but ultimately go to pieces. This is a true process
of secretion. The excretophores of leeches, according to Graf,
undergo a similar process. The cells loaded with secretory
products approach the funnels of the nephridia and portions of
them gradually break off, disintegrate, and the debris is swept
out of the body. We might call this secretion by budding.
Secretion here passes by a simple gradation into cell division,
the sole difference being that in such secretion the bud does
not contain a nucleus, and, in consequence, the budded off por-
tion soon goes to the ground. But even this is not a hard-and-
fast line of demarcation, because in many instances of cell
division the cells budded off, even though containing a nucleus,
undergo degeneration and are converted into a so-called secre-
tory product. Such cases occur, for example, in the mucous
glands of the skin of Petromyzon, in which the cells thus budded

1 68 BIOLOGICAL LECTURES.

off are transformed into a curious mucin and bodily extruded;
this happens also in the sebaceous glands, where they are
converted into sebum ; in the oil gland in the tails of birds ;
in the skin itself, in which the outer cells are converted into
keratin. In fact, nearly all the appendages of the skin, such as
hairs, feathers, and nails, are secretions of this sort. It will be
seen that it is impossible to draw a line between such simple
processes as the budding off of a portion of the cell and the
well-characterized cell division. As a general rule, however,
only those processes of this kind are properly designated as
secretions in which the cells budded off, or given off, do not
divide again, but are soon transformed into materials plainly
lifeless.

A third way in which these complex substances are removed
from the cell is by a compression of the cell by contractile
tissue. The substances are here squeezed out and the cell
temporarily disrupted. This is a fairly common method. One
of the most striking and elaborate examples is furnished by the
nematocysts of the Hydrozoa. In this case the hylogen to be
secreted takes the form of a spirally wound thread forming a
dart. This secretion is expelled violently from the cell by the
contraction of a specific part of the cell's protoplasm around
the cyst containing the dart. Here we have a secretion used
for defensive and offensive purposes. Other clear examples of
secretions of this sort are furnished by the large unicellular
glands of the edge of the mantel of Aplysia and the unicellular
glands of the carp-louse, Argulus. These are emptied of their
watery secretion by means of a special muscular sheath sur-
rounding the cells in Aplysia, and apparently by the body
musculature in Argulus. Many of the secretions from the
multicellular glands are produced in the same way. This is
the case in the livers of the mollusks and arthropods, the skin
glands of Amphibia, the poison glands of spiders, the sweat
glands of mammals, and, as will be pointed out later, the
salivary glands of mammals and cephalopods. The muscular
movements of the intestinal wall are not improbably a con-
siderable factor in most animals in the secretion of the glands
and cells lining this tract.

THE PHYSIOLOGY OF SECRETION. 169

A fourth method largely employed for removing these sub-
stances from the cell is by bursting the cell through internal cell
tension. We have seen that the hydrolytic splittings taking
place in protoplasm, particularly during vaso-constriction in the
absence of oxygen, render the cell contents more soluble and
of smaller molecular size. The absorption of water by the cell
and its turgor is thereby greatly increased. A typical picture
of such a cell is afforded by any mucous cell, particularly
those of the submaxillary gland. The whole cell is enormously
distended, and the nucleus, reduced to a granular mass, is driven
against the bottom of the cell. Now, if the cell membrane be
resistant enough to hold in these substances and prevent their
escape by filtration, obviously there must come a time when
the cell will burst. When this happens, the cell contents rush
forth often with such violence that they drag the nucleus with
them. More frequently, however, the nucleus remains behind
and builds up a new cell. This is the manner in which the
mucin appears to escape from the goblet cells of the intestine ;
from' the intestinal epithelial cells of Ptycoptera larvae; from
the cells of the submaxillary and sublingual glands. Probably
the secretion from the pancreas cells is of a similar kind, though
this is by no means so certain.

Finally there is the method employed by the cells of the
mammary glands in which the inner ends of the cells are dis-
organized and break off, apparently losing all coherence after
their fatty transformation. This process is also seen in the
calcareous glands of the earthworm.

In these different ways, then, do these non-osmosable sub-
stances escape from the cell. Each of these methods consti-
tutes a mechanism of secretion, and it is readily seen that there
are many mechanisms. It is hence absurd to speak of a special
secretory activity of cells, as Heidenhain and many other phys-
iologists do, as if there was one process common to all. To
which of these secretory processes does such a term have refer-
ence ? We shall come back to this in treating of the question
of the existence of secretory nerves.

The secreting cell does not, as a rule, secrete equally well in
all parts. It discharges always in one given direction. How

170 BIOLOGICAL LECTURES.

is this brought about ? It is clearly due to the fact that the
cell membrane is more pervious or less resistant in one part
than another. This is a matter that must necessarily be prop-
erly arranged, or the cell would discharge first in one direction,
then in another. In nearly all glands the cells are so arranged
that the weakest spot is toward the gland lumen and away from
the basement membrane. In many cases, however, nearly the
whole surface except that lying against the basement membrane
permits the escape of the secreted material. This is doubtless
the explanation of such pictures as one often sees in Golgi
preparations of the pancreas and salivary glands, in which the
impregnation of the so-called ** secretory capillaries" extends
between the cells nearly to the basement membrane. It is
probably owing to this differentiation of the cell and the uneven
resistance of the limiting membrane that cells will secrete even
osmosable substances in one direction only. Consider, for
example, the soluble bile salts of the liver cells which leave the
cells always by the biliary capillaries. Probably where the bile
capillary enters the cell is the point of least resistance, whither
the bile salts must journey as fast as they are formed. The
cell leaks, as it were, at this spot.

Nothing is more obvious than that the processes of secretion
are under some sort of a control. Many cells and glands do
not secrete constantly, but only when needed. They secrete
intermittently. How is this intermittence brought about }
How is secretion controlled .-* We shall find it necessary to
distinguish in the case of multicellular glands between the dis-
charge of secretion from the gland duct and its discharge from
the cell. It by no means follows that when the cell is secret-
ing, the secretion is escaping from the duct ; nor, on the other
hand, that when the gland is secreting, the cells of the gland
are secreting also. The two processes often coexist, but this
is by no means invariably, one might say even generally, the
case. A great deal of confusion unfortunately existing in our
conceptions of secretion arises from a failure to take account
of this fact. In a great many cases, perhaps the majority of
cases, the secretion from the cells is continuous, while the dis-
charge from the duct is intermittent. In other well-recognized

THE PHYSIOLOGY OF SECRETION. 171

instances, however, the discharge from the cell is also inter-
mittent.

Let us consider the simpler secretions first. The secretion
of carbonic anhydride is evidently controlled by its production,
since the gas is normally secreted as fast as it is produced.
The production is largely dependent on the food and external
conditions. In the case of muscle the secretion of carbonic
anhydride is largely dependent on the contraction of muscle,
large amounts of this gas being formed when muscle contracts.
In this case it is under the direct control of nerves. But let it
be remarked that these nerves do not control the secretion;
they only control the production. They have no action on any
hypothetical secretory function of the muscle cell at all. They
cannot, hence, be called secretory nerves in the proper sense.
They act on a very special substance in a very special cell.
These simpler secretions, then, so far as they are intermittent
at all, are rendered so by the intermittence of their production.
This in turn is largely affected by many different circumstances,
chief among which is the food supply of the cell. This is quite
beyond the cell's control, and any nerve affecting the secretion
of these substances, beyond the very special case of muscle,
affects it through controlling the food supply. This is, for
example, the case in the liver. Whenever the oxygen supply
of the liver is reduced either by vaso-constriction, insufficient
ventilation of the lungs, or in any other way, the production
of sugar by the liver cell is increased and its secretion increased.
Just as in the muscle, where nerves control the secretion of
carbonic anhydride by controlling the production, so in the
liver the secretion of sugar is controlled by its production.
And if we may speak of the muscle nerves as secretory nerves,
so must we equally well speak of nerves which indirectly con-
trol the secretion of sugar as secretory nerves. The vaso-
constrictor nerves going to the liver or to the blood vessels of
the intestinal area become, hence, secretory nerves. But they
do not directly affect any secretory power of the cell ; they only
affect the production of sugar. They do this, moreover, by
their action on blood vessels and not, as far as is known, by
their action on the liver cell. All vasomotor nerves, in fact,

172 BIOLOGICAL LECTURES.

as they control the food supply, including here oxygen, control
also the rapidity of formation and secretion of certain sub-
stances of small molecular size. They become, hence, true
controllers of secretion. They are secretory nerves in one
sense. The secretion of all simple osmosable substances is
controlled, then, by their production, which is in its turn
modified by the food and oxygen supply.

How is the secretion of the more complex substances con-
trolled } One of the most clearly defined ways in which such
secretions from the cells themselves are rendered intermittent
is by the action of muscle. The cells, after accumulating in
themselves the hylogens to be secreted, are periodically com-
pressed by the action of contractile matter, either within the
cell itself, or immediately about it. A fine example of this
sort is the nematocyst of the Hydrozoa. The hylogen elabo-
rated lies coiled in the distended cell. When the projecting
irritable point of the cell is properly disturbed, the contractile
substance, lying about the cyst in the same cell, suddenly con-
tracts and violently ejects the dart. It is not improbable that
the cell then forms another dart, which in its turn when fully
ready is ejected. If a secretory nerve controlled this secretion,
which is probably not the case, it will be seen that it must
effect the expulsion of the dart by acting on contractile tissue.
It would not produce its action by altering the elaboration of
the dart. A similar muscular mechanism is to be found in
the unicellular glands of Aplysia, in which the enormously
distended cells are surrounded by a well-marked contractile
sheath. That these cells are emptied by the contraction of
this sheath with a possible coincident dilation of a restraining
muscular sphincter can hardly be doubted. If this secretion
is under nerve control, and it probably is, as nerves have been
traced to the muscular sheath, the nerves which control it must
act on the contractile substance about the cell, not upon the cell
itself. The large unicellular glands of the carp-louse, Argulus,
are also emptied intermittently either by the contraction of the
cells themselves or the surrounding musculature.

A second means by which such cellular secretions are ren-
dered intermittent is by the periodic explosions of the cells.

THE PHYSIOLOGY OF SECRETION, 173

In such cells, for example, as those lining the gut of Ptycoptera
larvae, a sort of mucin is formed. This absorbs water, distends
the inner end of the cell until finally the turgor becomes so
great that the cell actually explodes. As the nucleus remains
behind, the cell is reformed, again breaks down into mucin,
again explodes, and so on. The cause in this case is perfectly
constant, the result only is intermittent. The secretion of
the goblet cells of the intestine is probably rendered inter-
mittent in this manner, although in this case the variation in
blood supply hastens or delays the final explosion.

Finally the secretion of the cells of the vertebrate glands is
controlled beyond doubt, in part at least, by the blood supply.
In the vertebrate we have a highly elaborated, carefully coor-
dinated, closed vascular system which enables an organism to
turn food, oxygen and liquid, on or off from a given area at
will. The cells may be thrown at once into an anaemic state or
overwhelmed with oxygen. In all glands of vertebrates which
are strikingly intermittent, the glands during rest are almost
white ; during secretion, pink, from the dilated blood vessels.
That these secretions are controlled by the blood system the
following facts will show : in the kidney and liver the rapidity
of secretion rises and falls with blood pressure and the rapidity
of flow through the glands ; it is very dependent on oxygen ;
in the pancreas every agency which induces vaso-constriction
stops secretion; every agency which induces dilation in the
normal gland causes secretion ; in the intestine and stomach
similar relations prevail ; in all vertebrate glands the principal
secretory nerves are dilator nerves, the inhibitory, constrictors.
Evidently, in the vertebrate, one of the chief ways of controlling
secretion is through the vascular system.

Let us now consider the manner in which the discharge of
the secretion from the gland is brought about. This is the
process which is incorrectly called secretion by most authors.
We will call it glandular secretion in distinction from cellular
secretion. Let us consider the invertebrate glands first. Prac-
tically the only manner in which the glandular secretions of
the invertebrates are rendered intermittent is through muscle.
This muscle acts in either of two ways ; first, as a sphincter

174 BIOLOGICAL LECTURES.

around the duct, holding the secretion in so that it accumulates
under high pressure. By the dilation of the sphincter the
secretion pushed from behind rushes forth. Or, second, the
gland or its duct is provided with a muscular sheath, and the
secretion is driven out by the contraction of the sheath. In
both these cases the secretion from the cell is constant ; it is
only the discharge from the ducts which is intermittent. In
the second method mentioned above, two modifications exist ;
the muscular duct may be dilated into a bladder, or the mus-
cular sheath may be around the gland.

Among the secretions which are rendered intermittent by
the dilation of a sphincter may be mentioned the secretion of
silk in the Lepidoptera. In this case three chitinous jaws
press the thread, holding it firm ; the thread accumulates under
pressure in the rear. When the jaws are drawn apart by
muscle the thread rushes forth. Other instances of inter-
mittence produced by sphincters are the segmental organs
of leeches, in which the vesicle is emptied by pressure, after
dilation of a sphincter ; the odoriferous glands of the Forfi-
culidae; the poison glands of insects.

The second method in which a constant secretion accumulates
in a bladder with muscular walls provided with a sphincter is
also common among invertebrates. Examples of such secretion
are exhibited by the green glands of lobsters, the salivary glands
of certain mollusks, the salivary glands of bees. Among the
cases in which the gland itself is siupplied by a muscular tunic,
by the contractions of which it is emptied, may be mentioned
the salivary glands of cephalopods, the poison glands of spiders,
the pericardial glands of lamellibranchs, the livers of all Crusta-
cea and mollusks. It may be pointed out that the nerves going
to the musculature in all these cases will control the discharge
from the gland and become in this sense secretory nerves.

In the vertebrates the cases in which muscle controls the
intermittence are also numerous and unmistakable. Let us
consider first the simplest and most obvious case — the kidney.
The secretion of the kidney cells, the real secretion of the kid-
ney, is continuous. It is, however, directly under the control
of the vascular system. The secretion from the duct of the

THE PHYSIOLOGY OF SECRETION. 175

kidney is intermittent. This intermittence is brought about,
as every one knows, by the interposition of a muscular bladder,
or reservoir, and a muscular sphincter. The intermittence of
flow from the duct is brought about, not at all by an intermit-
tence of true secretion, but, just as in many invertebrates, by
the action of a bladder. Now in the kidney no secretory
nerves have ever been found ; that is, no nerves acting on
the kidney cells and thus controlling secretion. Nerves which
control secretion, however, are undoubtedly there. Those con-
trolling the true secretion from the cells are the vasomotors
of the kidney. Those controlling the intermittence of secre-
tion from the gland duct are the nerves governing the bladder
and its sphincter. These are the secretory nerves of the kid-
ney. Place the kidney in the bladder and the sphincter close
to the pelvis, and secretory nerves to the kidney cells would
have been found without difficulty. They have not been found
because the relations are too plain.

The next simplest secretion is that of the liver. Here again
the relationships are perfectly clear, and here again is the strik-
ing coincidence that no secretory nerves can be found. The
secretion of the cells of the liver, like that of the kidney, is
continuous, the secretion accumulating in the gall bladder or
finding its way to the blood. Here, also, as in the kidney, the
secretion of the cells is closely dependent on the blood flow.
Not only is this the case in regard to the sugar secretion, as
already mentioned, but the rapidity of bile secretion rises and
falls, while other factors remain constant, with the blood flow.
The secretion issues from the duct intermittently. This inter-
mittence is produced by the action of the gall bladder and its
sphincter, just as it is produced in the kidney and various
invertebrate glands. Where are the nerves controlling this
secretion .^ Those which control the intermittence of the
gland's secretion act on the musculature of the bladder, the
gall ducts, and the sphincter; those which control the true
secretion of the cells act on the blood vessels either of the
liver or the intestinal area.

In the stomach we have a slightly different condition. In
this case the gland is inside the bladder. The relationships

176 BIOLOGICAL LECTURES.

are no less clear. The secretion of the cells into the cavity of
the stomach or the bladder is inconstant. The intermittence
is controlled by the blood supply. As long as the stomach is
secreting slowly, the blood supply is almost cut off the glands ;
when the blood vessels dilate, the glands secrete. The flow
from the gland duct, in this case the duodenum, is still more
intermittent. It is rendered intermittent by the action of the
pyloric sphincter and the muscular stomach wall. The secre-
tory nerves controlling the ejection of the stomach's secretion
are the nerves governing the stomach's movements ; those gov-
erning secretion proper are the vasomotors. The conditions
are quite the same in the intestine, which may be regarded as
a gland on a large scale. We have here a constant secretion
rendered intermittent from the duct by the anal sphincter and
the musculature of the intestine.

In the pancreas we come upon a somewhat simplified case,
since here the muscular elements appear to be lacking, or, at
least, of small importance. In the rabbit and other rodents the
discharge from the duct and cells is continuous. In the higher
mammals, for example the dog, the secretion is intermittent.
The intermittence is produced by the intermittence in blood
supply. The resting pancreas is very white, the blood is
almost wholly cut off from it; the secreting pancreas is
exceedingly red ; it has a profuse blood supply. Any agency
which causes vaso-dilation, provided that the gland be unpoi-
soned, causes secretion. For example, the gland may be
caused to secrete by cutting the spinal cord, leading to vaso-
dilation ; by cutting the vagus, allowing it to degenerate four
days and then stimulating it, a proceeding necessary to give it
a dilative effect ; by the drugs pilocarpine, chloral, or other vaso-
dilators. On the other hand, secretion is stopped by any inter-
ruption of this blood supply ; by constricting the arteries ; by
clamping the aorta ; by any drug which causes vaso-constric-
tion, such as digitalis or strychnine. The secretory nerves of
the pancreas are the nerves governing the blood vessels. The
evidence that there are any others to the gland cells we shall
shortly examine.

Let us now take up one of the skin glands — the amphibian

THE PHYSIOLOGY OF SECRETION, 177

skin glands. Some of these glands may be watched living
under the microscope. They are simple glands with an incom-
plete contractile sheath lying between the cells and the base-
ment membrane. The glands are provided with a sphincter.
If these glands be watched during secretion, it may be seen
that on stimulation of the sciatic nerve (for the glands of the
web) the sphincter dilates, the gland membrane contracts, and
a secretion ensues. After ejecting the secretion from the
gland the cells slowly secrete again, the glands become large
and are under high pressure, not sufficient, however, to force
the sphincter. The gland, in other words, is inside the bladder.
It is exactly a case of a small stomach. The secretion from the
cells appears to be constant, but it is doubtless dependent, like
all other vertebrate secretions, upon the blood supply ; the inter-
mittence of the ejection is caused here, as in many other glands,
with but few exceptions, i.e., the pancreas, by the dilation of
the sphincter and the compression of the gland. In the glands
of the nictitating membrane the sphincter and the musculature
appear to be innervated by nerve fibers of different origin.
Drasch found that he could produce a secretion either by
stimulation of the trigeminal, which caused a compression of
the gland's body by contraction, or by stimulating a sympa-
thetic spinal nerve, which dilated the orifice. Drasch himself
believed that the latter nerve acted on the gland cells, because
these became higher during stimulation ; but it is obvious that
if the sphincter opened, as he states it did, the diminution of
pressure inside the gland would undoubtedly tend to increase
the height of the cell and diminish slightly the diameter of the
gland. Here we have, at any rate, an intermittent secretion
from the ducts produced, not by intermittent secretion of the
cells, but by muscle. The secretory nerves to this gland inner-
vate the sphincter and muscular sheath. It will be noticed that
here the relations are somewhat less clear and more difficult to
understand, and here we begin to have evidence, though still of
a very dubious kind, of the existence of nerves going to the
gland cells.

The mammalian sweat glands offer the next step in the
series. These are homologous with the amphibian skin glands.

I 78 BIOLOGICAL LECTURES.

Like the latter, the sweat glands are provided with a muscu-
lature between the cells and the basement membrane. There
is no evidence of the existence of a sphincter. The intermit-
tence of secretion of the sweat glands is produced in two ways.
The easiest perceived is that produced by the vascular system.
Dilate the blood vessels of the skin by any means, and sweat
secretion, under normal conditions, always follows. It makes
no difference how the dilation is brought about ; we may pro-
duce it by cutting off the circulation from an area entirely for
fifteen to twenty minutes. On readmitting the blood, vaso-
dilation ensues, and profuse sweating. Cut the cervical sym-
pathetic nerve in the neck of the horse, and vaso-dilation,
accompanied by heavy perspiration, results. Here, as in all
other vertebrate glands, then, other factors being equal and the
gland in a normal state, vaso-dilation invariably causes secre-
tion. It causes a discharge from the duct as well as from the
cells, because there is no bladder and no sphincter.

There is another way, however, in which sweat secretion is
rendered intermittent. This is by the actions of the muscu-
lature surrounding the gland. This drives out the fluid in the
gland and its duct. This is the secretion which takes place on
stimulation of the sciatic, which can go on after death, in the
absence of blood supply or during vaso-constriction. If the
sciatic nerve of the cat be stimulated, and during stimulation
the skin plunged into some powerful and rapid fixative, the
muscular sheath is found in strong contraction. This secretion
generally coexists with vaso-constriction, since the same agen-
cies which cause contraction of . the sheath of the gland cause
also contraction of the arterioles. Nerves which act on this
sheath will be secretory nerves. Such sweat secretions are
caused also by constricting drugs, such as strychnine, picro-
toxine, physostigmine, and so forth. The secretion of sweat is
controlled, then, in the same two ways as are nearly all other
vertebrate intermittent secretions — by the action of the nerves
on the muscular sheath and by the action of nerves on the
blood vessels.

Another interesting skin secretion is that of the mammary
glands. Here we have a constant secretion by the cells, at

THE PHYSIOLOGY OF SECRETION. 179

least nearly constant in some cases. So far as it is intermit-
tent, it is controlled by the vascular system. After parturition
the glands are gorged with blood, and secretion from the cells
begins. The intermittence of the discharge of the secretion is
brought about by muscles, just as it is in the other secretions
examined. In this case it is largely the action of a sphincter,
though doubtless the muscle of the ducts and about the alveoli
comes into play. The cells secrete constantly against the
sphincter, until, like the frog-skin glands or the poison glands
of spiders, a high pressure is generated. The gland is put on
the stretch. This pressure may even become so great that the
sphincter can no longer restrain the milk which issues from
the duct. The intermittence is caused by the dilation of the
sphincters and the contraction of the muscle around the
alveoli. In the mammary glands, also, evidence of the exist-
ence of nerves acting on the gland cells is lacking.

We come now to the most interesting glands of all — the
vertebrate salivary glands. They are the most interesting, be-
cause they are the most complex, the most obscure, and possi-
bly on account of their obscurity they have furnished nearly
the whole of the evidence for the existence of special secre-
tory nerves acting on the gland cells themselves. How are
these glands rendered intermittent } In the light of the com-
parative study we have been through I have no hesitation in
saying that they are rendered intermittent in just the same
two ways as are all the other glands, i.e.^ by the action of the
vasomotors and by the action of muscle. It is another case of
the gland being inside the bladder.

That the submaxillary gland is regulated in its secretion by
the blood flow, not the slightest doubt can exist. The resting
gland, like the pancreas, is white ; the active gland, red ; the
chief secretory nerve is a dilator; vaso-dilation, as long as the
gland is unpoisoned, always causes secretion. Thus, if the blood
be cut off for fifteen or twenty minutes and then readmitted,
vaso-dilation and secretion ensue. If one extirpate the sub-
lingual ganglion leading to dilation in that gland, spontaneous
secretion begins and continues. These glands are rendered
intermittent, then, by the vascular changes going on in them.

l8o BIOLOGICAL LECTURES.

There is, however, here as elsewhere, a second mechanism —
that of muscle. This contractile tissue exists, like that of the
skin, sweat, lachrymal, and milk glands, between the cells and
the basement membrane, but it is present in much smaller
amounts. The sympathetic nerve innervates this muscle, and
the secretions following stimulation of this nerve, secretions
coincident with vaso-constriction, secretions caused by picro-
toxine, supra-renal extract, physostigmine, are caused by this
muscle.

From this short survey, necessarily very incomplete, it may
clearly be seen that there is not one but many different
mechanisms of secretion ; that secretion proper — that is, the
discharge from the cell — is generally constant ; its intermit-
tence, when it is intermittent, is caused first by contractile
tissue compressing the cells, second by the action of sphinc-
ters in the case of some unicellular glands, and third by the
vascular system. The intermittence of discharge of the secre-
tion from the gland's duct is due first to sphincters, second
to the muscular contractions of the gland or of its ducts
(bladder) ; or, third, to vasomotor changes in the gland.
Nerves which produce this intermittence act, therefore, upon
one of these three mechanisms. They do not act on the gland
cell.

But besides this indirect control of secretion, do special
nerves exist which control secretion by action on the secreting
cells themselves } As every one knows, it is generally believed
by physiologists that such nerves do exist, and that they con-
trol secretion quite apart from any condition of the vascular
system, and quite apart from muscle action. It is impossible
to go into this matter here at any length, — something which I
have already done elsewhere, — but we can at least consider one
of the main evidences for the existence of such special secre-
tory nerves. The theory of their existence is, I believe, quite
erroneous.

According to Ludwig, the originator of the theory, these
nerves did not affect any peculiar secretory power of the cell at
all. They caused only a decomposition of substances in the
cell, which thus affected its turgor and hence secretion. In the

THE PHYSIOLOGY OF SECRETION. l8l

sense originally used by Ludwig the nerves ought not properly
to have been called secretory nerves ; they correspond exactly
to Heidenhain's "trophic" nerves. If we mean, when we
speak of secretory nerves, secretory nerves in Ludwig's sense,
as is, I believe, generally meant, then all talk of secretory
nerves affecting a special secretory power of the cell is wrong.
They had no direct influence whatever on any hypothetical
secretory power of the cell. Heidenhain, who put the Ludwig
theory on its present pedestal, adopted a new sort of nerves.
Those which acted on the cells, rendering their contents solu-
ble and thus affecting osmosis, he called '^ trophic " nerves ; he
postulated an entirely new sort of fiber which did act on the
secretory power of the cell. These were the secretory fibers.
They had no influence at all on the chemical processes of the
cell, but only affected the resistance to filtration of the inner
end of the cell. If we use the term "secretory" nerve in
Heidenhain's sense, we can speak of their affecting the secre-
tory power of the cell. Some physiologists use the term in
one sense, some in another. This last theory of Heidenhain's
is one of the most extraordinary, to designate it by no harsher
term, of all physiological theories. It necessitates the conclu-
sion that each single cell of the submaxillary gland, for exam-
ple, has acting on it at least four different nerve fibers, i.e.,
the trophic of the sympathetic, the trophic of the chorda tym-
pani, the secretory of the sympathetic, and the secretory of the
chorda tympani. With such consequences the Heidenhain
theory is undoubtedly false. But are we in any better position
if we follow Ludwig } Suppose we relinquish the trophic
fibers and adopt the customary fashion of speaking glibly of
secretory fibers. How do these nerves act } Do they act, as
Ludwig thought, by producing chemical changes in the cell, in
which case they do not directly affect any secretory function
of the cell, or do they act in the Heidenhain sense on the inner
border of the cell.? It is all very well to speak of secretion,
secretory activity, secretory nerves. In science definite mean-
ings must be assigned the terms used. Just which of the
secretory mechanisms are to be governed by such nerves }
Certainly not the secretion of milk, in which the whole end of

1 82 BIOLOGICAL LECTURES.

the cell goes to pieces, nor the secretion by budding, nor the
explosion of the cell by turgor. Indeed, the secretory nerves,
like the epicycle, are hypothetical existences devised to explain
certain facts. What are those facts } If we can explain them
without assuming the existence of such nerves, the hypothesis
loses its value.

The most important fact is the action of atropine. If atro-
pine be injected into a dog, vaso-dilation will still ensue on
stimulation of the chorda tympani nerve, but no secretion from
the submaxillary. This shows, says Heidenhain, that vaso-
dilation by itself is not able to cause secretion. This inference
is, as will be seen, entirely unwarranted, since it is quite possi-
ble, and is, I believe, undoubtedly true that vaso-dilation always
causes secretion in the normal gland, but not in that poisoned
by atropine. Atropine might have acted on the capillary
walls, preventing the exudation of lymph. That it does so act
there can be hardly a doubt. It might act on the gland cells,
preventing them in some way from secreting in spite of vaso-
dilation. Why was the latter, the natural explanation, not
adopted } Because it was assumed that there was but one
mechanism of secretion. There is but one mechanism of
secretion, says Heidenhain. The sympathetic, hence, causes
secretion in the same manner as the chorda. But the sympa-
thetic is not paralyzed by atropine ; therefore, as the sympa-
thetic innervates the gland cells, these have not been paralyzed.
But if the gland cells have not been paralyzed, and the dilator
function of the chorda is unaffected, and atropine does not act
on the nerve fiber, then the drug must paralyze the nerve ends
of the hypothetical secretory nerve endings. The final false
inference from three erroneous, or unproven, assumptions !
This one example will suffice to show the character of the evi-
dence of the existence of separate secretory nerves. Outside
of the salivary glands of mammals there is no evidence worthy
of the name of the existence of such nerves. It is only in
'those glands in which the muscle and gland cells are most
closely intermingled, and relationships the most obscure, that
such evidence can be found. It is to the writer almost incred-
ible that so far-reaching and fundamental an hypothesis should

THE PHYSIOLOGY OF SECRETION. 183

have been so widely accepted on evidence of such an extraor-
dinary character. If we only assume that there is but one
mechanism of glandular secretion, — a false assumption to be
sure, but one which seems theoretically probable and hence
answers the prime requisite of a good hypothesis, — if we make
this erroneous assumption, then the facts of secretion become
so completely obscure that we are compelled to assume secre-
tory nerves and secretory vital activities to explain them. If,
however, the fact be recognized, which has stared us in the
face for years, that there are several ways in which secretions
are driven from glands, several of which mechanisms may
coexist in one gland, then the facts of secretion, with one or
two unimportant exceptions, become plain.

In closing, let us point a moral to this physiological tale.
An explanation in science, whatever it may be elsewhere, is
the statement of a phenomenon in simpler known terms. It
is no assistance to science, but rather a scientific misdemeanor,
to propose as an explanation something far more complex and
difficult to understand than the facts explained. There is, to
be sure, a superficial simplicity in the assumption that there is
but one secretory mechanism common to all cells. But when
it becomes necessary, to make new assumptions to bring the
theory into accord with observed fact the theory may be safely
regarded as false. It was so with the geocentric theory of the
universe — new epicycles had constantly to be added to account
for the planetary motions ; it was so with the Weismannian
theory of inheritance — units of a new order had constantly
to be invented; it is equally true of the secretory nerve
theory.

TWELFTH LECTURE.

REGENERATION: OLD AND NEW
INTERPRETATIONS.

T. H. MORGAN.

The great interest that was awakened in the last century
in the study of regeneration was the result of the experiments
of Trembley, Reaumur, Bonnet, and Spallanzani. The interest
aroused through the work of these four naturalists has not
decreased, although from time to time other problems have
come to the front and attracted the attention of investigators.
More particularly in our own time has attention been directed
to problems connected with the egg and its development. But
it is becoming clearer, I think, that development by means of
an egg and development by means of regeneration cannot be
considered as separate and different phenomena, but at bottom
have many factors in common. There can be little doubt that
the results in one of these fields of study will throw light on
the other. It will be possible for me to consider at this time
only the suggestions and hypotheses advanced in connection
with the problems of regeneration. At another time I shall try
to compare these interpretations with those connected with the
development of the egg.

Trembley's experiments on Hydra were the starting-point for
the three other naturalists. Spallanzani occupied himself mainly
with collecting new facts, while his friend Bonnet, who also
made many new observations of great value, seems to have
been more interested in the theoretical side of the problem.

Bonnet supposed that regeneration was brought about through
the development of preformed germs. These germs exist in the

185

1 86 BIOLOGICAL LECTURES.

animal solely for the purpose of replacing lost parts, and since
in some animals the same part could be replaced time after
time, Bonnet assumed that on each occasion a new set of germs
was awakened. He pointed out that, since some animals are
more subject to injuries than others, they are supplied with as
many sets of germs as the times the animal is liable to be injured
during its natural life.

Bonnet seems to have been especially impressed by the fact
that from the same region of Lumbriculus a head or a tail may
arise according to whether that region happens to lie at the
anterior or posterior end of the cut surface. For instance, if
the worm is cut into two pieces, a new tail will develop from the
posterior end of the anterior piece, and a new head from the
anterior end of the posterior piece. If, however, the cut had
been made a little further in front of or behind this level, the
same result would have followed ; hence it is clear that at every
level a head or a tail may develop. Which of these develops is
determined by the position of the region, i.e., whether it lies
at the exposed anterior or posterior end of a piece. Bonnet
interpreted this to mean that there are throughout the worm
head germs and tail germs. He saw that it is necessary to give
some further explanation of why the one rather than the other
kind of germ is aroused to activity, and made, therefore, a
further assumption. The fluids of the body that pass forward
carry nourishment for the head. When the worm is cut in two,
these substances are, in the posterior piece, stopped in their
forward movement by the cut surface, and accumulate at this
place. They act especially on the head germ, and, nourishing
it, bring about the development of the new head. Similarly,
fluids to nourish the tail are assumed to flow posteriorly; hence
in an anterior piece of a worm they will accumulate at its poste-
rior cut surface, and, acting on the tail germ, there bring about
the development of a new tail.

In another species of fresh-water annelid Bonnet found (1745)
that when the worm was cut in two a new tail developed at the
anterior end of the posterior piece, and not a head. He sup-
posed that in this worm only tail germs are present throughout

REGENERATION. 1 87

the body ; hence only a tail can develop, even at the anterior
end of the piece. ^

In one passage Bonnet states that the fluids that flow towards
the head are there used up, and we must infer that these head-
nourishing fluids are being continually made somewhere else in
the body of the worm. It may be pointed out, in passing, that
this idea of Bonnet's, that the fluid passing towards the head
(he seems to have had the blood in view) is a special kind of
fluid laden with head-nourishing substances, is not in agree-
ment with what we know of the function of the blood or of
other fluids in the body. The tissues of the head may take
out of the blood those substances in it that they use in their
life processes, but the blood itself going to the head is not
specialized in a particular direction, and is the same fluid that
flows posteriorly in other vessels.

Bonnet advanced three ideas : preformed germs ; head- and
tail-nourishing stuffs ; and the flow of these latter in definite
directions. I shall return later to these views and consider
them more fully.

The process of regeneration has been often compared to the
completion of a broken crystal ; just as the growth of an animal
or of a plant is sometimes contrasted with the growth of a
crystal in a saturated solution. Herbert Spencer, in particular,
has elaborated this view. In his book on the Principles of
Biology ■^ he says : "What must we say of the ability an organ-
ism has to re-complete itself when one of its parts has been cut
off 1 Is it of the same order as the ability of an injured crystal
to re-complete itself } In either case new matter is so deposited
as to restore the original outline. And if in the case of the
crystal we say that the whole aggregate exerts over its parts
a force which constrains the newly integrated molecules to
take a certain definite form, we seem obliged in the case of
the organism to assume an analogous force." Starting here
with an hypothesis that is no longer held, viz., "that the whole

1 Bonnet does not tell us how, in this case, the germs are awakened, since tail-
stimulating fluids are assumed to flow backwards. Perhaps, only tail germs being
present, he did not think it necessary to apply his subsidiary hypothesis.

2 Chapter IV, Waste and Repair. First published in 1863. I quote from the
last edition, 1898.

1 88 BIOLOGICAL LECTURES.

aggregate exerts over its parts a force," etc., Spencer follows
this up with the 7t07i seqtultir, '' we seem obliged, in the case of
the organism, to assume an analogous force."

After showing that this property, i.e., '* this tendency to
aggregate into specific forms," cannot reside in the chemical
compounds, because , if it resided in *'the molecules of albumen
or fibrin or gelatine or other proteid," there would be noth-
ing to account for the unlikeness of different organisms ; and
after showing further that it cannot reside in the morpho-
logical units or cells, because the same power is shown by uni-
cellular organism, Spencer concludes that this proclivity is
" possessed by certain intermediate units which we may call
physiological," etc.

Striking as is the comparison between the growth of a broken
crystal and the regeneration of an injured animal or plant, the
emptiness and superficiality of the comparison are at once appar-
ent on closer examination. The looseness of Spencer's argu-
ment is equally evident, and his rejection of the idea that the
chemical substances composing protoplasm cannot account for
the facts leaves him only his imaginary physiological units to
bring about the results. The latter, after all, could only be
formed by combinations of the chemical substances, and if so,
why introduce into the argument unknown ** units" having the
property of bringing about regeneration } To my mind nothing
could confuse the whole subject more surely than reasoning
and arguments like those advocated by Spencer.

The recent work of Rauber on the " regeneration " of crys-
tals gives us now a basis of fact on which to rest any com-
parison we may make. Rauber's results show that during the
growth of a broken crystal the typical form may be assumed
and the broken surfaces obliterated. In some cases the growth
may be more rapid over the broken surface, since this rougher
surface presents a greater area for the deposition of new mole-
cules. When a piece is broken off, the closing in of the
exposed part presents no phenomena that are in any way
different from the growth of the crystal everywhere else. The
position of the new molecules that are added is determined by
the condition at the points to which they are applied. It is

RE GENERA TION. 1 8 9

misleading to speak of the '' whole aggregate exerting over
its parts a force," etc.

Rauber himself speaks with much reserve in his comparison
between the "■ regeneration of the crystal " and the regenera-
tion of animals and plants. His results show, it seems to me,
very clearly that the comparison is only a general analogy; and
the moment we attempt to press the comparison further it
breaks down at every point.

It will, no doubt, be admitted by every one at the present
time that the form of the crystal is somehow determined by
the chemical composition of the substance of which the crystal
is made up, and likewise that the form of an animal is also
determined by the substances of which it is composed ; but the
broken crystal regains its original form only when surrounded
by a saturated solution containing the same substance as the
crystal. New substance is then added over its entire surface
as well as over the part broken off. On the other hand, the
process of regeneration is entirely different in an animal or
plant. The new material, if any is formed, comes from within
the animal or plant. Further, an animal that is slowly starv-
ing and decreasing in size will regenerate a missing part.
Again, several of the lower forms regenerate by changing over
the entire piece into the typical form. Can any one suppose
for a moment that such a process is comparable to the re-com-
pletion of the form of a crystal ?

If further evidence is asked, attention may be drawn to
those cases in which an organ, different in kind from the one
cut off, is regenerated. A striking case of this sort is the
regeneration of an antenna in place of an eye, or the develop-
ment of a tail at the anterior end of a posterior piece of an
earthworm.

It would not be difficult to multiply at length these points
of difference, but what I have said will suffice, I hope, to show
how little we can gain by Spencer's comparison.

Pfiiiger ^ has given a brief outline of his conception of the
process of regeneration. He says, since there is always

1 Pfluger, E. Ueber den Einfluss der Schwerkraft auf die Theilung der Zellen
und auf die Entwicklung des Embryos, pp. 64-67. 1883.

190 BIOLOGICAL LECTURES.

replaced only as much as is lost, it is clear that the new limb
(of a salamander) does not arise from a preexisting germ of a
limb. The wounded surface of the limb draws food material
to itself, and these new food molecules are organized into a
new limb. The arranging force is a molecular force that does
not work at a distance from the living substance of the stump
of the limb, but acts only in so far that the force draws the
food molecules into the sphere of influence of the molecules
in the stump itself, bringing them to a definite place, and in
this way laying down a new living layer over the cut end of
the limb. The way in which this new layer is organized
depends on the law of organization of the part, i.e., on the
chemical condition of the surface layer on which the new layer
is deposited. In a word, the condition of the new layer is the
necessary mathematical consequence of the condition of the
older regenerating layer.

On the surface of the wound a knob of indifferent tissue
arises, but long before the first layer of the new material has
become fully formed (differentiated }) it has in turn acted on
the succeeding layer, and this on the next, and so on, so that
all the layers appear almost at the same time ; but the proxi-
mal ones are somewhat nearer the definitive form than the
more distal ones.

Pfliiger also points out that certain exposed surfaces can
organize new material only in one direction. In other words,
the organized surfaces of the body show a polarization, since
one side of a surface shows peculiarities not present in the
other. The direction of the polarization of the regenerating
surface is the cause of the direction of the new growth. For
this reason a portion of an animal cannot produce the entire
animal.

Pfliiger has not taken into account some of the most con-
spicuous facts of regeneration — facts known even at the time
at which he wrote. His explanation fails completely to account
for those cases where a piece of an animal, hydra, for example,
changes its form into a new small hydra. Moreover, Bonnet
had shown that when Lumbriculus is cut in two, only a few
new segments are added at the anterior end of the posterior

REGENERA TION, 1 9 1

piece. The same thing is true for the earthworm, in which, if
the piece be cut off behind the fifteenth segment, no new
reproductive region is ever developed. Pfliiger's view leaves
the absence of this region unaccounted for.

Weismann, in his book on the Germ Plasm published in
1893, elaborated an hypothesis of regeneration. Weismann's
central idea is not different from Bonnet's (1768). Both be-
lieve in preformed germs. The differences in their views
result largely from the application of modern cell doctrines
to Weismann's hypothesis. Regeneration is supposed to be
brought about by latent cells containing preformed germs
which exist in the chromosomes of the nucleus in the form of
determinants. There are supposed to be cells of this sort in
the leg of a newt, for instance, at every level and in all the
parts. At each level the latent cells are slightly different, and
each contains germs of such a sort that all the distal part and
only the distal part is represented. This germ stuff, coming
into action after a series of qualitative nuclear divisions, influ-
ences the part in which it is found. Further, since the new
limb will itself regenerate if cut off, the further assumption is
made that during regeneration new subsidiary germs are laid
down at each level in the new limb. This is supposed to take
place by a division of each germ into like parts (a quantita-
tive division) after it has reached its proper position in the
new leg.

This host of invisible germs, moving at the command of
Weismann's imagination, is supposed to carry out the process
of regeneration. No one can fail to see that the difficulty is
only shifted into a region where fancy can have free play and
a scientific, experimental test cannot be applied. At one blow
the difficulties are overcome, and the array of mystical germ's
is summoned to explain how regeneration takes place.

Since regeneration occurs in some animals and not in others,
and better in those forms, Weismann thinks, that are liable to
injury, additional hypotheses are added. Weismann combats
the idea that regeneration is the result of the inherited effect
of injuries to the part. The Lamarckian conception cannot
apply in this case, since it is not the use or disuse of a part

192 BIOLOGICAL LECTURES.

that is in question, but its power to regenerate after injury.
In this I agree with Weismann, for it is no more evident how
a series of injuries in succeeding generations could finally bring
about complete regeneration than that the result should follow
after one injury. If after the first injury, then, there is no need
of any theory of inheritance.

Weismann believes, however, that the power of regeneration
is under the guidance of ''Natural Selection." His argument,
as far as I understand it, is this : Of all the animals injured in
each generation, those that regenerate better are more likely
to survive ; and since in each species certain organs are more
liable to injury than others, the selection will take place mainly
in respect to these parts ; hence they possess the power of
regeneration. Other organs of the body not subject to frequent
injury do not show the power of regeneration, either because,
having once had it, it has been gradually lost (through panmixia),
or because the process has never been acquired in these organs.
It may be pointed out, in passing, that since the limb of a newt
and the tail of a tadpole regenerate at every level, and regen-
erate the kind of limb peculiar to that particular species, we
must conclude, on Weismann's hypothesis, that this power has
been acquired through selection for every possible level. The
demand made on our credulity is enormous.

Weismann has made other statements in his book on the
Germ Plasm, and in a later paper ('99) entitled "Regeneration :
Facts and Interpretations," that are worth examining. In the
former he says : " It may, I believe, be deduced with certainty
from those phenomena of regeneration with which we are
acquainted that the capacity for regeneration is not a primary
quality of the organism, but that it is a phenomenon of
adaptation^ Again: "Hence there is no such thing as a
general power of regeneration; in each kind of animal this
power is graduated, according to the need of regeneration, in
the part under consideration." "We are, therefore, led to infer
that the general capacity of all parts for regeneration may have
been acquired by selection ^ in the lower and simpler forms,
and that it has gradually decreased in the course of phylogeny,

1 The italics are my own.

REGENERATION. 1 93

in correspondence with the increase in complexity of organiza-
tion ; but that it may, on the other hand, be increased by special
selective processes in each stage of its degeneration in the case
of certain parts which are physiologically important and at the
same time frequently exposed to loss."

There are many points in these quotations that are, I think,
open to criticism, but it would lead me too far were I to attempt
to discuss them. Yet I cannot let this opportunity go by with-
out calling attention to another point raised by Weismann in
his later article. He says : " It may not have occurred to
Morgan that the changes in the structure of a species may
have kept pace with the changes in the conditions of its life ;
yet this is a presupposition of the hypotheses of natural selec-
tion, and is, indeed, its conditio sine qua non. Hermit crabs
have certainly possessed the power of regeneration ' from the
beginning' ; but may they not have inherited it from their ances-
tors, the long-tailed forms, which possess it to this day, and have
need of it for all their appendages, since all are liable to injury }
And cannot, nay must not, these in their turn have inherited it
from their ancestors, the sessile-eyed crustaceans, and so on,
through the whole crustacean pedigree, back to the unknown
annelid-like ancestors of the class .-*... It seems almost as if
Morgan ascribed to me the view that the capacity for regenera-
tion must be built up anew for each species — must be inscribed,
so to speak, on a tabula i^asa; my view, however, is that here,
as in all transformations, nature started with what was already
present, and by modifying it brought about adaptation to new
conditions. The assumed general power of regeneration in the
lowly ancestors of the crustaceans would thus gradually have
adapted itself to the changes in the body and to the new con-
ditions resulting from these changes as well as from other
causes ; it would have become localized and specialized. . . .
As in the course of time the appendages of the different
body-segments became more widely differentiated in adaptation
to different functions — giving rise to antennae, jaws, walking-
legs, or swimmerets — the predisposition to regenerate in certain
parts of the body slowly varied also ; and thus, not indeed at
the same rate, but not lagging very far behind, the adaptation

194 BIOLOGICAL LECTURES.

of the capacity for regeneration followed the adaptation of a
limb to a new function."

Weismann had, in his book on the Germ Plasm, committed
himself to the statement that the germ plasm for regeneration
is different from that which brings about the development of
the ^gg. He believed that it was necessary to make this
hypothesis to account for the appearance of the so-called ''an-
cestral organs " that are sometimes (?) regenerated. Combin-
ing this statement with what has been quoted above, it will be
clear that as a new species evolves from an older one its regen-
erative germ plasm must also change if the animal regenerate
a part like the one lost. Hence each species must acquire the
power of regenerating its own particular kind of structures.
This opinion I did ascribe to Weismann, and still suppose he
holds to it. I did not imply that regeneration " must be
inscribed, so to speak, on a tabula rasa!' The capacity to
regenerate the parts of the old species would be present in the
new one, but the new species must acquire through natural
selection of favorable variations the power to regenerate the
new structures that have arisen through ^gg variations. I have
quoted at length this argument of Weismann to show where we
are landed by the results of his speculation. He argues for a
double process of natural selection for each species that can
regenerate, and is led into this position by the assumption that
seems quite necessary on the preformation hypothesis that
regenerative germs exist in the ^%g independent of the germs
of embryonic development.

I find in this whole argument only an attempt to shift the
difficulties of the problem back to the unknown ancestors of
present forms, just as the difficulties of other parts of the prob-
lem are also shifted back upon the unknown germs that exist
preformed in the Qgg or in the parts that can regenerate.
Weismann does not, perhaps, realize the difference between
himself and those whom he somewhat scornfully calls '' the
younger investigators." The problems that they are trying to
solve are those that Weismann also tries to answer, but " the
younger investigators " base their interpretations on the assump-
tion that when a change takes place a sufficient cause for the

REGENERA TION.

195

change is to be sought in the organ itself and in the external
conditions surrounding that organ. They are not content to
rest their " explanations " on *' the phyletic origins " of the
changes. It is not necessary to deny the theory of descent,
but it is unsafe and in many cases unscientific to base '* causal
explanations " on an imaginary line of ancestors. It is cer-
tainly unprofitable to shift our difficulties back to these historic
forms, and most unfortunate to find our "explanations" also
resting on the same shadowy past.

In a masterly essay, entitled " Stoff und Form der Pflanzen-
organe," Julius Sachs has considered the question of regenera-
tion in plants, and has outlined an hypothesis to account for
the phenomena. Sachs bases his view on the conception that
the form of a plant is the outcome of its chemical structure ;
that whenever the form changes there has been an antecedent
change in the material (Stoff). Sachs vigorously combats Voch-
ting's idea that there exists in the organism a polarization of
every part, and that this polarization is a directive agent that
determines the kind of regeneration that takes place. No less
earnestly does Sachs protest against the metaphysical concep-
tion of many morphologists, expressed or implied, viz.^ that for
each species of organism there is a form that tends to express
itself and controls the development of each part. According
to Sachs there are no such formative forces in the organism,
but all changes are brought about by differences in the chem-
ical composition of the "■ Stoff," and this leads to the develop-
ment of the form peculiar to that material. For example, the
flower-buds of plants are produced, not because of some innate,
mystical force that causes the plant to complete its typical form,
but because some substance is made in the leaves that, flowing
into the undifferentiated growing point, there acts on the ma-
terial of the growing point and changes it into that sort of stuff
from which a flower develops.

Applying this idea to regeneration, Sachs supposes that in
the plant two substances are being produced ; one of these is
a leaf-forming substance, the other produces roots. When a
piece of the stem is cut from a plant these two substances con-
tained in it flow in their respective directions, and bring about

196 BIOLOGICAL LECTURES.

the production of new leaves and new roots. Sachs made
numerous experiments which showed, he thought, that the
direction of the flow of these substances is determined by
the action of gravity ; the leaf -forming substances flowing up-
wards, and the root-forming downwards ; the direction of
the flow being thus determined by some factor outside of the
plant itself. It had been shown that if twigs of the willow be
planted with the distal end in the ground, new roots arise from
the end in the ground, and leaves from the free end. This
result follows, Sachs thinks, from the direction of flow of the
root-forming and leaf-forming substances.

In considering Sachs's view, it will be well, I think, to keep
apart the two ideas, that specific substances produced in the
plant bring about the change, and that these substances may be
transported from one part to another in definite directions.
We might think of the transportation taking place in a given
direction as due to some peculiarity of the substance itself,
or of the tissues of the organism, or, as Sachs supposed in this
case, as the result of some outside influence.

It is interesting to notice that this idea of a transportation
of specific stuffs goes back to Bonnet. The latter imagined
head-nourishing and tail-nourishing stuffs to flow, respectively,
forwards and backwards, and, acting on the germs in those
parts, determine the kind that develop. Sachs does not intro-
duce the idea of preformed germs, and correspondingly sim-
plifies his hypothesis.

The idea of specific substances determining the regeneration
of a part is, in my opinion, one deserving of very careful con-
sideration. We have seen that the idea first suggested itself
to Bonnet when searching for an explanation of the development
of a new head or tail from the same region of Lumbriculus.
A head developed if the exposed part lay at the anterior end
of a piece, and a tail if the exposed part lay at the posterior
end. Bonnet said that something must awaken the one or the
other kind of germ, since both were assumed to be present at
every level. Hence the idea of two specific stuffs. It is evi-
dent, of course, that the difficulty is only shifted from the
germs to the stuffs, for no such stuffs were known, much less

REGENERA TION.

197

their migration in definite directions. Nevertheless, the assump-
tion gave a formal explanation of the phenomena.

I shall next consider, in the light of the hypothesis of trans-
portation of specific stuffs, the results of certain experiments
that bear on the question.

When an oblique piece is cut from a planarian, as indicated
in Fig. I, A, I find that the new head develops not from the
center of the piece but far up on one side, Fig. i, B; and
always on that side that lay nearer to the head of the plana-
rian. The new tail develops out of that part of the posterior
side that was originally nearer to the tail of the planarian, Fig.

Fig.

Fig. 2.

I, B, C. The new pharynx appears in the middle of the piece,
with its long axis at first in the direction of the old long axis ;
but as the old part changes its form to produce the body of the
new worm, the pharynx comes to lie symmetrically with respect
to the new head and tail. Fig. i, D.^

If a piece is cut from the planarian by making two oblique
cuts that form an angle with each other, as shown in Fig. 2,
A, a new head develops, as before, at the most anterior part of
the anterior cut surface, and a tail at the most posterior part.
Fig. I, B. In this case, both head and tail lie on the same side
of the piece. The new pharynx appears in the middle, but not
in line at first with either the new head or tail.

These experiments show that the position of the new head

198 BIOLOGICAL LECTURES.

and of the new tail is determined by the obliquity of the exposed
part. The nearer the cut approaches a plane at right angles to
the long axis, the nearer the new head and tail will be in the
middle line.

If we assume that specific substances bring about the develop-
ment of the new head and tail in these pieces, and also assume
that the one substance flows towards the most anterior end of
the piece, and the other towards the most posterior end, we
should have an apparent explanation of the results.

A further application of this hypothesis may be made to
certain phenomena in the regeneration of the earthworm. If
the worm (Allolobophora fcetidd) is cut posterior to the middle
into two pieces, the posterior piece produces in many cases, at
its anterior end, a new fail and not a head. If we assume that
in the tail region only tail-forming substances are present, and
little or no head-forming substance, we might offer this suppo-
sition as an explanation of the result. But a difficulty arises
in connection with the idea of transportation of this stuff, for
on the hypothesis it should flow backwards instead of forwards.
However, the unusually long time before this sort of regener-
ation begins might be utilized to save the transportation theory.
We might assume that the tail-forming material flowed posteri-
orly, but after a time so much will have accumulated that it may
extend forward even to the anterior end, and there start the
regeneration. It is obvious that this is a forced interpretation
and that the result is not in accordance with the transporta-
tion idea.

This example of heteromorphosis in the earthworm is due, it
appears, to influences within the piece itself. This seems true
also for those cases, described by Herbst and myself, in which
an antenna is regenerated in place of an eye in some of the crus-
taceans. This case also does not harmonize well with the trans-
portation hypothesis, although, as I shall try to show later, it
might be explained on the assumption of specific materials in
the parts themselves.

There are other cases of heteromorphosis in which, as shown
by Loeb, the influence that determines the kind of regener-
ation comes from the outside. In certain hydroids regener-

REGENERA TION. 1 99

ation is influenced by the orientation of the pieces with respect
to gravity ; in others to light ; in others to contact. In those
cases in which gravity is the determining factor, we could
readily imagine that the transportation of head-forming or tail-
forming substances is brought about by the action of gravity.
In the other cases in which light or contact has an influence
on the regenerating part, it is not easy to see how a stimulus
of this sort could bring about the transportation of material,
although we can readily imagine that either factor acting
locally might cause the production of some substance which, in
turn, might act on the new tissues.

These illustrations would seem, in several cases at least, to
harmonize well with the stuff-hypothesis, less well, perhaps,
with the idea of its transportation. On the other hand, one
does not have to look very far to find other cases to which
the stuff-transportation hypothesis does not apply, or applies
badly.

In the first place, in many of the lower animals regeneration
does not take place by the development of new tissues, but by
a remoulding — morpholaxis — of the entire piece into a new
form. This side of the question has been almost entirely neg-
lected by those who have proposed hypotheses of regeneration,
and yet it seems to me that just here we find some of the most
important phenomena. A protozoon cut into pieces makes as
many new individuals of small size as there are nucleated
pieces. The size of the new individual is, within certain limits,
in proportion to the size of the fragment, and it develops —
regenerates — not by forming new material at the cut surfaces,
but by remoulding the entire piece into the characteristic
form. If a piece is cut from a hydra, it bends together and
makes a sphere or cylinder out of which a new hydra is formed.
There is no evidence of the formation of new tissue, unless the
tentacles are formed in that way; but in other hydroids, viz.,
Tubularia and Parypha, even the tentacles are known to
develop out of the old cells. In planarians a small amount of
new tissue forms at the cut ends, and out of this a new head
and a new tail develop, but the old piece changes over to form
the body of the new individual. It is in the higher forms alone

200

BIOLOGICAL LECTURES.

— echinoderms, molluscs, annelids, and vertebrates — that we
find regeneration taking place only by the addition of new
tissue at the cut ends.

In those forms that regenerate by morpholaxis we must find
a theory that can account for the changing over of the old
piece into a new whole. To assume two formative substances
might account for the changes at the
two ends, but not for the rest of the
piece.

There are still other difficulties for
the stuff -hypothesis. If a planarian be
cut in two longitudinally, each half
forms a new individual. Along the
entire length of the cut edge new tissue
appears. The new pharynx appears
along the border between the old and
the new tissue (Fig. 3, A, B). It lies
at first quite unsymmetrically in the
new worm, but the new side continues
to grow and the old side to get nar-
rower. The result is that although the
new worm is only as broad as the half from which it devel-
oped the pharynx lies at last in the middle plane of the body
(Fig. 3, C),

Shall we assume that side-forming substances are present,
which, being transported laterally, bring about the development
of the new side } If so, we should have to assume that at each
point the substance must be different from that elsewhere,
for a different structure is formed. I believe that few persons
would like to assume the responsibility for such a view.

One further objection. If the foot of a newt is cut off, only
the foot develops ; if the cut is made through the forearm,
then the forearm and foot are both regenerated ; if the limb is
cut off through the humerus, then all distal to that point is
renewed. If we assume a leg-forming substance, the assump-
tion is insufficient to account for the differences in the result
at each level. If we assume a different kind of substance flow-
ing from the body for each level, then the hypothesis becomes

Fig. 3.

REGENERA TION. 20 1

extremely complicated, and we might as well fall back at once
on the Bonnet- Weismann theory of preformation.

There is another experiment on planarians that has a direct
bearing on Bonnet's hypothesis. If a planarian be almost
entirely split in two, leaving the halves connected only at
the anterior end (Fig. 4), two new heads may develop at the
most anterior end of the cut edges (Fig. 4). Van Duyne,
who first carried out an experiment of this sort, found two
heads developing, and he interpreted their development as due
to a process of heteromorphosis. I have repeated this experi-
ment a number of times with the same results, but I think
there is a simpler and more obvious way to account for the
development of the new heads. They
appear at the sides of each half, as
they would .do were a long piece cut
from the side of the body ; but in the
latter case the result is not due to
heteromorphosis. In the former case
the two new heads are, after their for-
mation, prevented from being carried
forward by the presence of the old
head. This interpretation is in har-
mony with the results of several other
experiments. The bearing of this experiment on our present
examination is obvious. Two new heads develop, although the
old head is present. If the development of the new heads is
due to the presence of head-forming substances, as Bonnet
supposed, how could they develop as long as the old head is
present to use up these substances } The objection might not
apply with as much force if the transportation theory did not
include the using up of head-forming substances in the old
head.

Other examples might be cited, but those given above will
suffice, I think, to show the improbability of the stuff-trans-
portation theory, or, at least, the results show that it cannot be
universal in its application to the phenomena of regeneration.
The assumption of* head-forming and tail-forming stuffs is too
general to explain the results. We have, however, clear evi-

202 BIOLOGICAL LECTURES.

dence of the fact that the chemical composition of the various
organs of the body is different, and I think it is not going too
far to claim with Sachs that in a sense their form is the out-
come of their composition. We have also many observations
to show that during regeneration like parts form like. In the
earthworm, for instance, the new ectoderm comes from the old,
the new digestive tract in large part from the old one ; the new
nerve cord comes, in part at least, from the old one and in part
from the ectoderm. In other words, the specific character of
the old cells may be handed over to the new ones. If we
choose to think of a formative substance in each kind of cell,
we could make this the basis of our interpretation of the results ;
but since we do not know of such formative substances, it is
safer to rest our claim simply on differences in the chemical
substance of the cell itself. In other words, we only complicate
our view by assuming formative substances acting on an indif-
ferent medium — the protoplasm — and determining its changes.
We need not deny that this might sometimes happen, and, in
fact, several cases that Sachs has cited seem, to be due to some
such action ; but, in general, it is not necessary to make this
distinction, at least not in most cases where regeneration takes
place. ^

The assumption, in those cases in which regeneration takes

1 By composition of a cell or of any part of the body, we may mean either of
two things. The fundamental substance may be thought of as everywhere the
same, i.e., indifferent, and its action is directed by grosser substances (formative
substances), or we may mean by composition the entire substance without making
any distinction between the substances of which it is made up. For the latter alterna-
tive we may use the terms " chemical " or " molecular " structure or " constitution,"
and retain the term " formative substance " for the former alternative. For example,
we may think of a growing point as made up of indifferent protoplasm, and its fate
determined by the kinds of formative substance carried to the growing point (Sachs).
When new substance is added during regeneration at a cut surface, we may think
of the new cells as at first indifferent, and their fate determined by the formative
substances transported to or acting at the cut surface, or we may think of the new
cells as having a molecular structure like that of the cells (or parts) from which
they have come. This molecular structure might be thought of as changing later,
at least so far that at different levels in the new part it changes in its relation to
what is around it, and forms a new whole. The change would be due to some
other factor or factors, such, for instance, as the form of the new part or its
relation to the old part. -

REGENERA TION.

203

place by the development of new tissue, that the new cells have
inherited their specific nature from the cells from which they
have arisen, meets with no serious objection. But can we
explain all the results as the outcome of this inheritance ? I
think not. For instance, the assumption will not explain why
the new part assumes a form that is often different from that
part of the body from which its cells have been derived. For
example, if the first five segments be cut from the anterior end
of the earthworm, five will come back ; but if more than five
are cut off, istill only five come back. Now the new cells will
in the latter instances be derived from parts of the worm that
are quite different in their structure from the fifth or sixth
segment, and yet from any level of the anterior end of the body
only the five segments develop. The assumption of formative
substances, or of a molecular structure in the new cells, might
account for the fact that the new ectoderm comes from the old,
the new muscles from the old ones, endoderm from endoderm,
etc.; but that would leave the main problem unexplained, viz.^
the development of a new head. This example shows also that
Pfl tiger's hypothesis is insufiicient to account for the result ;
for, according to his hypothesis, we should expect that at every
level the whole of the missing part should be replaced, and not
merely five segments, at whatever level the cut is made.

We see that other factors are also at work. Let us see if we
can account for some of these. The digestive tract, at first end-
ing blindly in front, pushes out into the new part until it comes
in contact with the ectoderm. Its further growth forward is
prevented, not simply by meeting a mechanical obstacle, but
by what we may call, for want of a better term, a stimulus
received from the point of contact. At this point the ectoderm
now turns in to form, as Hescheler has shown, the buccal
chamber. We might assume that the endoderm acted in turn
as a stimulus on the ectoderm, and the invagination then
followed ; but it is probable, from certain results of Rievel
and of Hazen, that the invagination would take place, even
if the endoderm did not come in contact with the ectoderm.
The new nervous system extends forward from the old one.
In coming to the anterior end, its further growth forward is

204 BIOLOGICAL LECTURES.

prevented by the ectodermal covering, and the cord then
divides and turns upwards around the digestive tract to form
the commissures and brain. Still the most essential point is
unaccounted for, viz.^ that the cord divides at the anterior end,
for were the new part a tail and not a head, the division
would not take place. These examples will suffice to show that
the phenomena are too complicated to be accounted for by the
form of the new part ; however, the form of the new part may
be one of the factors, but not the only one. The obvious
fact that cannot be left out of account is that the new part is a
head long before the cells undergo their definitive differentia-
tion ; and it seems to be owing to the new substance being
from the start a head, that many of the results take place.
Nothing is easier at this stage of the analysis than to drop
into metaphysical ideas and speak of vital factors or formative
forces, etc. I have not escaped this pitfall altogether, for I
pointed out on one occasion that the process appeared as
though guided by intelligence. *^ I mean that what we call
correlation of the parts seems here to belong rather to the
category of phenomena that we call intelligent than to physical
or chemical processes as known in the physical sciences. The
action seems, however, to be intelligent only so far as concerns
the internal relations of the parts, etc." ^ The reactions to
stimuli that are amongst the most common phenomena of living
things is what I had in mind, and it is true that at present we
cannot explain them as the result of known chemical or physical
properties of matter, but I do not think that therefore I was
justified in calling them intelligent processes, even in the broad-
est use of the word, for we thereby fall into the error of attempt-
ing, to explain simpler processes by more complicated and less
well-understood ones.

We are confronted, then, with the question : What is it that
gives to the new part the structural character or chemical com-
position to form a new head } If we say its form, i.e.^ its closed
dome-shaped figure, we meet with the objection that this same
form is almost universally present where the new part is present
as a knob. However, since in each such knob of new cells the

^ " Some Problems of Regeneration," Biol. Led., 1898.

REGENERA TION. 205

cells have a different molecular structure from those derived
from other regions, we might, by combining the two assump-
tions, escape from the dilemma. On the first assumption, that
the cells of each new layer or new organ have received their
molecular character from the old part, and in each kind of
tissue are all more or less alike, we are driven to assume, on
the second assumption, that the form of the new part deter-
mines subsequent changes in the molecular composition, and
the material is so changed at each point that its arrangement
produces the foundation of a new head.^

Before we follow further this line of thought let us examine
those cases in which the entire piece is changed over into a new
organism. If a piece is cut from the middle of the body of
hydra, it closes in at both ends and a cylindrical form is the
result. If the piece owed originally its characteristic form {i.e.,
as a piece of the body of hydra) to its molecular structure, it
would have this same character after separation ; and in giving
rise to a new hydra it does not develop new tissue at the two
ends like that lost, but, on the contrary, after a short time the
old piece itself assumes the form of a new hydra. If we hold
to Sachs's view, that wherever a new form arises there has been
an antecedent change in the material, we must conclude that
in the piece of hydra the material has rearranged itself into a
new whole. The cells do not materially change their position,
but each goes over into a different part of the body from that of
which it formerly made a part. Since, however, as in the
case of Lumbriculus, each cell may develop into a part of either
the anterior or posterior end of the new hydra, according to the
level at which the piece was cut off, we must account for the
one or the other change. What factors can we suppose bring
about this result } In the first place, however the piece is

"'■ The specific substances that the new cells have brought with them from the
old parts may in some cases determine the kind of new organ rather than the rela-
tion of the new organ to the rest of the body (through, of course, its area of con-
tact with the old body). In this way a new eye may develop in the hermit crab
when the distal end of the old eye stalk is cut off, and not an eye but an appendage
when the stalk is cut off at its base. The development of a tail at the anterior
end of a posterior piece of the earthworm would be due to the specific character
of the new cells dominating the development.

2o6 BIOLOGICAL LECTURES.

cut out, the material at one end will have been more anterior
in position in the original piece than the material at the other
end. This difference is sufficient, theoretically, to account for
the different results at the two ends. I do not assume that a
polarization is present, because we do not know enough about
any such principle, if it really exists at all, to make it of any
service. My assumption rests simply on the basis that at each
level the material is somewhat different from that at every other
level. These considerations give us a sufficient basis to build
up an idea of how the development of the piece may be thought
of as taking place. As soon as the piece has been removed and
its ends have closed in, it is possible to think of a rearrange-
ment of the molecular structure taking place throughout the
whole mass. Since the two ends are different, we can imag-
ine their differences to become greater and greater in the same
directions in which they differed at first. If the expression
is pardonable, the anterior end becomes more anterior, and the
posterior more posterior ; and this influence extending to the
intermediate regions, they too change in their respective direc-
tions. As a result, the material of the new piece assumes the
molecular arrangement characteristic of a hydra. Then the dif-
ferentiation begins that changes the entire piece into the new
individual.

Let us not hesitate to push this view to its logical conclu-
sions and ask in what part of the material does this change take
place. Does each cell change and through simple contact with
its neighbors bring about a change in them, and so from cell to
cell } In other words, is the result an intercellular reaction }
While it might be possible to look at the result as brought
about in this way, still, I think there are several important
reasons why we must regard the change as more fundamental
than that involving only the cells as units. I cannot take up
at this time my reasons for so thinking, but I may point out
that since in the Protozoa and in the eggs and embryos of other
forms changes similar to those I have described take place, we
seem to be forced to the conclusion that the change is in the
whole protoplasm and is probably a molecular change of the
protoplasm.

REGENERATION. 207

To repeat ; when a piece is cut out of the body of hydra a
molecular change takes place in the protoplasm of such a sort
that the entire mass is changed over into a structure that rep-
resents in its structural basis a new hydra. It is this molecu-
lar change that, dominating the subsequent development, seems
to control it, and gives us the impression of formative processes
at work. On my view, the formative processes are only the
expression of the physical, molecular structure that has been
assumed by the piece.

It will be seen at once how the same conception may be
applied to those cases of regeneration in which a knob of new
tissue appears and the missing part, or a portion of it, is regen-
erated. After the new cells have been formed they will have
the same relations to each other and to the old part that the
cells have in the isolated piece of hydra. The new part does
not form a new whole, but only a part, simply because it is con-
nected with the old part by the same kind of molecular union
as are the cells of the new part to each other. Whether all the
missing parts, as in the limb of the newt, or whether only a
part, as in the head of the earthworm, will develop out of the
new material, will be determined by the volume of the new
material that forms, and its relation to the structural pecul-
iarity of the new part. This sounds vague, perhaps mystical,
but I think it can be given a real meaning. If, for instance,
five segments are cut off from the anterior end of the earth-
worm, the new material that forms suffices to make five seg-
ments, but if ten be cut off, the new material is still only
sufficient, owing to some molecular peculiarity, to make five.^
In the limb of a newt we must suppose that at every level
enough new material is formed at first to make possible the
formation of any part of the limb.

This analysis of one of the problems of regeneration has
been undertaken in order to see if it is possible at the present
time to construct an hypothesis that can bring under one point
of view many isolated observations. It is offered as a working

1 There is a difficulty, perhaps, in accounting for only two segments coming
back when only two are cut off, and not five, but if it can be shown that less new
material is formed the difficulty is avoided.

2o8 BIOLOGICAL LECTURES.

hypothesis or as a possible point of view, and not as an elabo-
rated theory of regeneration. In fact, I think it would be a mis-
take at the present time to attempt to construct a final theory,
for, if I have been in the least successful in this discussion of
the problem, I hope to have made it clear that the process of
regeneration involves many factors. It is obvious that until we
have analyzed the problem into its component parts it would
be ridiculous to attempt to formulate a theory of regeneration.
It is not difficult to show that there are in reality many factors
in the process quite different in kind. For instance, the clos-
ing in of the exposed surface seems to be due to some sort of
■cytotropism in the cells as well as to other factors ; the produc-
tion of new cells from the old ones must be due to another set
of processes, as well as their migration out over the exposed
end ; the form of the piece and its size are also factors ; the
specific character of the cells derived from the old ones is still
another problem ; and finally, if my analysis is sound, the sub-
sequent molecular arrangement taking place throughout the
new part is one of the most important changes that take place.
I have laid emphasis only on this latter characteristic because
it seems to me that it is just this change that comes nearer to
what we mean by regeneration than any of the others that I
have named ; they all enter into the problem, however, and
none of them can be neglected, but to find a theory to account
for them all at once would be, I think, extravagant to attempt
and probably disastrous in its results.

THIRTEENTH LECTURE.

NUCLEAR DIVISION IN PROTOZOA.

GARY N. CALKINS.

Mitosis, or indirect division of the nucleus, with its compli-
cated processes, is almost as characteristic of the cell as the
nucleus itself. The general agreement of the separate com-
ponent parts of the mitotic figure, in most cases down to the
finest details of structure, makes it appear as difficult to trace
its evolution as to describe the evolution of the cell itself. Yet
the cell has evolved from the general to the special, and the
mere fact that mitosis is strikingly similar in the most special-
ized of animal and plant tissues should not weigh against the
view that these mitotic structures have had a history and that
some stages in this history may be paralleled by structures to
be found in the lowest forms of life.

All nuclei of higher plants and animals pass, during mitosis,
through similar pro-, meta-, and telophases, and a similar divi-
sion, or mitotic, figure is formed. But minor variations and
differences in what may be homologous structures are seen in
the so-called achromatic and chromatic portions of the spindle
fi-gure, differences which may be accounted for by the supposi-
tion of divergent modification of a common ancestral form. A
few of the more important of these variations may be briefly
considered before turning to the more generalized structures
and processes in Protozoa which show, I believe, a more primi-
tive condition.

Broadly speaking, the variations in mitotic figures may be
reduced to three types: (i) Forms with centrosome, central
spindle, mantle fibers, and astral rays. (2) Forms with centro-

209

2IO BIOLOGICAL LECTURES.

some and astral rays but no distinct central spindle. (3) Forms
without centrosome and central spindle.

In all of these cases the granular chromatin of the resting
nucleus is arranged in the form of a reticulum upoji a linin
network, and the prophases of division are similar in that the
granules are brought together and concentrated in the form of a
much wound or coiled thread, — the spireme, — while the stain-
ing reactions become much more intense. This stage is univer-
sally followed by segmentation of the spireme into chromosomes
of definite number, shape, and size for each species. Finally
the nuclear membrane disappears and the chromosomes are
left naked in the cytoplasm, but connected by spindle fibers
with the two poles of the mitotic figure.

As the variations in the three types mentioned have mainly
to do with the achromatic structures, the chromatin changes
may be omitted. In the first type considered, the approach of
division is signalized by the division of the centrosome into
two daughter-centers connected by fibers which form a small
spindle, called the ''central spindle " by Hermann ('89). 1 This
continually enlarges as the centrosomes diverge, other fibers
(mantle fibers) meanwhile growing from the centrosomes and
pushing in the nuclear membrane, which finally disappears,
leaving the chromosomes in the cytoplasm. The mantle fibers
then become attached to the chromosomes, and the latter
finally surround the central spindle like a ring. The origin of
these fibers is differently interpreted by different observers.
The mantle fibers, according to some, arise from cytoplasmic
material ; according to others, from linin substance in the
nucleus. The central spindle fibers, on the other hand, arise
from archoplasm (Boveri), or from the substance which sur-
rounds the centrosome (centrodesmus material of Heidenhain).

Several transitional stages between the first and the second
types have been described by different observers. In these
the central spindle appears first as very delicate fibers between
the dividing centrosomes, but these break later and no connec-
tion remains. The complete spindle, in such a case, consists
of fibers which pass, apparently, from pole to pole, with chro-

1 Arch. f. mikr. Anat., Bd. xxxiv.

NUCLEAR DIVISION IN PROTOZOA. 21 I

mosomes strung upon them, while central spindle fibers, if
present, must be intermingled with the others, and both sets
must have the same origin (Wilson, '95), Toxopneitstes} In
many cases of this nature, especially in the huge mitotic fig-
ures of the first cleavage nucleus of many eggs, the spheres
are relatively enormous, while the centrosomes are small and
often difficult to find [Toxopneustes and Arbacia). Finally the
third type of mitosis differs from the first in the absence of
central spindles and centrosomes, and from the second type
by the absence of centrosomes (higher plants).

While opinion differs as to the presence or absence of cen-
trosomes in plants higher than the fungi, some observers deny-
ing, others affirming, its presence in the same species, the
balance of opinion at present appears to be towards the nega-
tive side, and evidence is certainly accumulating to support
this view. Nevertheless, the difficulty of proving a negative is
nowhere more apparent than here, and one positive affirmation
throws down the entire conception. On the other hand, posi-
tive assertions are excessively dangerous, many plastids, gran-
ules, or other products of the cell being easily mistaken for
centrosomes, and only the most far-reaching experiments, with
the best technical skill, can safely determine their presence.
According to numerous observations by Strasburger, Mottier,
Osterhout, etc., the spindle in plant mitoses arises by the grad-
ual convergence of rays which make their appearance tangential
to the nuclear membrane. Arising, as it were, from the sub-
stance of the cytoplasm, and converging to a bipolar mitotic
figure, the spindle fibers are supposed, by Strasburger and his
followers, to be composed of a definite and distinct substance
to which he gave the name ** Kinoplasm." The nuclear mem-
brane here, as in the other types, always disappears before the
nuclear plate is formed, and nuclear division proceeds in the
usual way.

Leaving out of consideration for the present the discussion
of the many intermediate forms of resting nuclei which, among
the Protozoa, approach the type of nucleus of higher forms, we
find in this primitive group so many variations of mitosis,

^ Journ. of Morph.y vol. xi.

212 BIOLOGICAL LECTURES.

some complex, some remarkably simple, that it is possible to
pick out a series which represents the stages through which
the complicated mitotic figures of the higher plant and animal
nuclei may have passed in their evolution to the present
state. It must be distinctly understood, however, at the out-
set, that these variations among the Protozoa give absolutely
no clue to the phylogeny of the Metazoa and the Metaphyta ;
all that they show, at the most, is that the various forms here
considered have arrived at a certain grade of differentiation
where they have stopped, although it may be logically inferred
that similar stages have been passed through by the higher
forms.

Beginning with the division which most closely resembles
mitosis for the first type as described above, we shall see that
the process in the cystoflagellate Noctiluca can be described
in almost the same terms as that of the Metazoa.^ The nucleus
is of large size, with a distinct membrane, and with chromatin
in the form of large reservoirs (Binnenkorper of Rhumbler,
Chromatosphere of Doflein) from eight to eleven in number.
The remainder of the nucleus is filled with large granules,
which have a distinct affinity for the acid dyes. There is no
trace of linin network or ''achromatin " other than the large
granules, which may be identical with Reinke's oedematin
granules. Upon the outside of the nucleus, in the cytoplasm
and close against the nuclear membrane, is a large, faintly
staining mass, which corresponds with the sphere in higher
cells, and which during division acts as a kinetic center.

During the early stages of mitosis the chromatin reservoirs
break down, by repeated division, into an immense number of
minute chromatin granules, which, at first collected in groups
in the region of the reservoirs from which they were severally
derived, later become distributed about the nucleus until the
latter is apparently filled with them. The granules are then
collected again in lines which radiate inwards from the nuclear
membrane at the point where the sphere is in contact (Fig.
I, A)\ these lines form the chromosomes, which soon become
double. The chromosomes thus formed are later connected

1 Cf. Calkins, Mitosis in Noctiluca miliaris. Boston, Ginn & Company, 1898.

NUCLEAR DIVISION IN PROTOZOA.

213

with the centrosomes in the spheres, and then separate in the
lines of previous longitudinal division. At no time is there
a spireme, as in ordinary mitosis.

During early stages of nuclear activity the sphere divides
into two similar halves connected by a strand composed of
fibers which are formed from the substance of the sphere.
These fibers compose the central spindle and are homologous
in every way with the central spindle fibers of the first type of
mitosis given above (Fig. i. A). The nucleus then elongates in
a direction at right angles to the central spindle, and at the
same time it bends in the center in such a way that the cen-
tral spindle sinks into a depression in the nucleus which sur-

Fig. I. — Stages in mitosis of Noctihica miliaris. A, amphiaster and chromosome formation ;
B, metaphase showing central spindle in the hollow of the nucleus ; C, section through
long axis of the central spindle ; D, section through one pole of late anaphase in spore-
forming mitosis. Double centrosome connected with the longitudinally divided chromo-
somes by mantle fibers.

rounds it upon three sides (Fig. i, B). In this way the nuclear
plate is finally wrapped about the central spindle in the form
of an incomplete ring, a condition, it will be observed, which
obtains in all higher mitotic figures where the central spindle
is present. The nuclear membrane then disappears in that
part of the nucleus which is turned towards the central spindle,
while it is retained unbroken in all other parts of the nucleus
(Fig. I, Cy D). Thus the chromosomes again, as in the higher
type, are brought in contact with the central spindle fibers.
They then split longitudinally, and, through the agency of
mantle fibers, are separated into two equal groups, each group
drawn towards its own daughter-sphere. Within the sphere
the mantle fibers are focused in a centrosome which at this
period can be demonstrated with the greatest ease (Fig. i, D),
The division is finally completed by the separation of the

214 BIOLOGICAL LECTURES.

remainder of the nucleus and the re-formation of the daughter-
nuclei, while the chromosomes disintegrate into granules which
again form the large chromatin reservoirs characteristic of
Noctihica.

It is hardly necessary to point out further the similarity
between this division figure and the mitotic figure of the first
type given above. The centrosome within the sphere, the
mantle fibers and their insertion in the chromosomes, the ori-
gin of the central spindle from the substance of the sphere, are
features which are obviously common to the two types.

The mitotic figure, in all cases, can be easily divided into two
portions, of which one consists of chromatin, the other of cen-
trosome and spindle fibers. The two parts have more or less
separate antecedent phases, and are brought in contact only
during the metaphase and later stages of division. It is possi-
ble, therefore, to conceive the two processes unequally devel-
oped ; a high grade of chromatin and chromosome development
may accompany a relatively slight achromatic specialization, and
vice versa. In the Protozoa, specialization of the two parts is
not synchronous, and the types which are used to demonstrate
the successive stages in chromosome formation cannot be used,
in the same order at least, for the successive stages in achro-
matin differentiations. It is, therefore, more satisfactory to
consider the two lines separately.

Beginning with the chromatin changes, it has been shown
that in Noctiluca the chromatin reservoirs disintegrate into
smaller and smaller pieces prior to division, and that these
small parts are secondarily united into chromosomes. In many
of the simpler Protozoa, notably in the Phytoflagellates, in Fo-
raminifera (Schaudinn) and Sporozoa (Siedlecki), the chromatin
material is aggregated into a single chromatin reservoir which
can be compared with one of the eight or ten similar structures
in Noctiluca. Rhumbler ('94) ^ described a varying number of
chromatin reservoirs (Binnenkdrper) in the foraminifer Sacca-
niina spkcerica, from one in the smaller forms to three hundred
in the larger nuclei. Gromia and other reticulates have a vary-
ing number of similar masses (F. E. Schultze). In many cases

1 Zeitschr. f. wiss. Zool., Bd. Ivii.

NUCLEAR DIVISION IN PROTOZOA. 215

the chromatin reservoir is known to break down into minute
granules before division, as they do in Noctiluca. This was
early made out by Buck ('78) for Arcella, by Gruber, Hertwig,
and others for various rhizopods, and is therefore a widely
distributed phenomenon among the lower forms.

This disintegration of the chromatin masses is a significant
and suggestive process and may be regarded as universal in
cell life. In the majority of cases the granules are secondarily
united into chromosomes of more or less definite shape and
size for each species. Among the Metazoa, Brauer's ('93) ^
description of the granulation of chromatin before spireme for-
mation shows, if not identical, a closely analogous process, and
the numerous observations, by various zoologists and botanists,
of spireme formation and synapsis stages, apparently belong in
the same category. Among the Protozoa, Brauer ('94),^ Hert-
wig ('98),3 and Gruber ('93)* have each described the granule
formation in ActinosphcBriuni Eichhornii, Schaudinn ('96) ^ in
Actinophrys sol, Schewiakoff ('88)^ in Euglypha alveolata, Hert-
wig, Biitschli, Plate, and Balbiani in Spirochona gemmipara,
Schaudinn, Wolters, Labbe, Clarke, and Siedlecki, in various
Gregarinida and Coccidiida. Whatever may be the cause of
this disintegration, possibly an expression of the "limit of
growth " which Spencer postulates in connection with the rela-
tionship of surface and mass to growth, it is apparently a sec-
ond stage in chromatin changes, the primary stage being the
single chromatin reservoir.

The disintegrated chromatin or granules which represent the
second stage in the specialization of chromatin are in reality
the preliminary stage in chromosome formation ; for, by their
union into new aggregates, the definite bodies known as chro-
mosomes are formed. In many Protozoa the granular stage is
permanent throughout resting and active phases of the nucleus,
neither the aggregation into a chromatin reservoir nor forma-

1 Arch.f. mikr. Anat., Bd. xlii.

2 Zeitschr.f. wiss. Zool., Bd, Iviii.

3 Abhand. d. K. Bayer. Akad. d. Wiss., Bd. xix.

4 Zeitschr. f. wiss. ZooL, Bd. xxxviii.
6 Sitz.-Ber. Ak. Wiss. Berlin, 1896.
^ Morph.,Jahr. 13.

2i6 BIOLOGICAL LECTURES.

tion of chromosomes taking place. Representatives of this
stage are found chiefly in the Flagellata, although some Rhizo-
poda and some Ciliata show the same thing. Forms with this
permanently granular chromatin, again, are found in two con-
ditions. In one type the granules are scattered throughout
the entire cell, and are never confined by a nuclear membrane
(so-called *' distributed " or ** diffuse " nuclei). In nuclei of the
other type the granules are confined in a definite, more or less
spherical space, which may or may not be bounded by a nuclear
membrane. Examples of the first type were described by Gruber
in certain Rhizopoda and in a number of Ciliata, and, as he sug-
gested, it is highly probable that many, if not all, of Haeckel's
Monera will be found to possess nuclei of this type. Among
flagellated forms it has been described by Biitschli ('96) i for
Chromatium okenii and Ophidomonas jenensis^ and by myself
('98)2 fQj. ^ species of Tetramitus. In the latter form the gran-
ules of chromatin, which at first are scattered throughout the
entire cell with no apparent order, come together to form a
loose aggregate prior to division. In this condition the aggre-
gate is divided into halves, an equal portion going to each
daughter-nucleus (Fig. 2). It is important to note here, how-
ever, that another element comes in to complicate the process.
In the resting condition of the cells, when the chromatin is
distributed throughout the cytoplasm, a faintly staining body
can be found somewhere near the center of the cell (Fig. 2, A).
This body becomes more definite as the chromatin granules
come together for division, and it divides into two equal por-
tions before the nucleus is halved. During the process of
division the chromatin granules become heaped about this
partly divided body, one-half of which remains in the center
of each daughter-heap of granules until the end of the division
(Fig. 2, Z>, E). After division the granules again separate, form-
ing the distributed nucleus. The central body, therefore, has
the attributes of an attraction sphere.

The condition represented by this temporary aggregation of
chromatin granules about the sphere is permanent in the ma-

1 Weitere Ausfiihrungen iiber den Bau der Cyanophyceen und Bacterien.
Leipzig, 1896. 2 Ann. New York Acad. Sci. 1898.

NUCLEAR DIVISION IN PROTOZOA.

217

jority of the Flagellata, and may, perhaps, be regarded as the
usual condition of protozoan nuclei. Among some Flagellata
the aggregation of chromatin granules about the sphere, al-
though permanent throughout resting and active phases, resem-
bles the loose aggregation of the division period of Tetramitus
in having no nuclear membrane {Chilomonas paramcecmnt^
Trachelomonas lagenella, and T. hispida). In other cases, how-
ever, a nuclear membrane is present either as a faint and
extremely delicate outline (Syniird) or as a well-defined mem-
brane {Eiiglena^ Phacus^ etc.). In the latter forms, accordingly,

Fig. 2. — Tetramitus species. A, individual with scattered chromatin granules and central
cytoplasmic sphere ; B and C, the chromatin is aggregated in the vicinity of the sphere ;
D and E, stages in the division of the sphere and of the chromatin. Optical sections.

the nucleus has become a definite structure, consisting of chro-
matin granules and a kinetic center or sphere (the " nucleolus
centrosome " of Keuten and other authors), and nuclear divi-
sion takes place without rupture of the membrane. The chro-
matic and achromatic elements of the dividing nucleus are thus
brought together permanently.

A third stage in the differentiation of the chromatin is seen
in the secondary union of the chromatin granules to form chro-
mosomes. In Noctiliica, as described, the process of chromo-
some formation is probably as simple as in any case where
longitudinal diyision of the resultant chromosomes takes place.
The union of the granules into larger bodies is found in many

2l8 BIOLOGICAL LECTURES.

of the Flagellata with intranuclear sphere where, as for exam-
ple in Euglena, the granules become welded together into
short, thickened chromosomes, but there are no observations
to show that longitudinal division takes place. A somewhat
similar process of chromosome formation is described by Hert-
wig ('98) for Actinosphcerium. In the latter case, as in Nocti-
luca and Etigtena^ the number of chromosomes is very great,
and it is impossible to state whether the number is constant
or not.

In none of these cases is there a preliminary arrangement of
the chromatin into a definite spireme, as in the higher forms.
In Noctiluca an occasional thread-like arrangement of the
granules is observed, but it does not appear to be constant.
In numerous other Protozoa, however, spireme formation may
become as complex as in Metazoa, and, as in the latter, the
thread may be segmented into chromosomes of definite shape
and size, although usually of large number {Etiglypka, Opa-
linay Actinopkrys sol, Acanthocystis, Ceratiuniy etc.).

Turning now to the so-called achromatic structures of the
cell, we find that the origin of the metazoan centrosome and
attraction sphere from simpler elements in Protozoa has been
the subject of a number of interesting theories. Most of these
undertake to work out the morphogenesis of the centrosome as
it appears in Metazoa. Butschli was the first to seek its homo-
logue in the micronucleus of Ciliata, a comparison which has
found its way in nearly all of the subsequent theories. It was
quickly caught up by Hertwig and by Heidenhain, and the
latter, especially, elaborated it into a complicated theory of
phylogeny, the main thesis being that the central spindle, as
described by Hermann, is derived from the micronucleus by
loss of chromatin, and becomes secondarily connected with the
chromosomes of the other nucleus. Lauterborn improved the
theory by assuming that micronucleus and centrosome may
have had a common ancestor in some primitive binucleated
protozoon, such as Amceba biniicleata (Schaudinn), and that
intermediate stages may be seen in certain existing Protozoa,
such as Paramceba Eilhardi (Schaudinn) and Noctiluca. He
considered the micronucleus to be a differentiation of one of

I

NUCLEAR DIVISION IN PROTOZOA. 219

these nuclei, but along a different path from that taken by
the centrosome. A different theory has appeared elaborately
worked out in several of the recent publications of R. Hertwig.
In its latest form Hertwig's theory may be reduced to the
following three stages in evolution: (i) The achromatic sub-
stance is at first uniformly distributed in the resting nucleus,
but appears during division as pole plates, the equivalent of
centrosomes.^ (2) The achromatic substance becomes perma-
nently an intranuclear centrosome. (3) It is protruded from the
nucleus to form an extranuclear centrosome. The centrosome
{i.e,^ spheres) thus extruded becomes differentiated for other
functions in the micronucleus of Ciliata, while the centrosome
in Metazoa is defined as follows : " From a morphological stand-
point the centrosomes are to be considered as portions of
the achromatic substance of the nucleus which have become
extruded ; they are nuclear in origin. One can call them, in
a certain sense, nuclei without chromatin." Thus Hertwig has
introduced a new conception by regarding the centrosome as
coming originally from nuclear substance. Doflein substan-
tially follows Hertwig.

Still another point of view was suggested by Schaudinn,
after his very remarkable observations and experiments upon
Acanthocystis and Oxyrrhis marina? He postulated two possi-
bilities — either the centrosome is a structure formed within the
nucleus, and subsequently becoming cytoplasmic, as it does in
the buds of Acanthocystis, and as the " nucleolus-centrosome "
does from the nucleus of Oxyrrhis marina when immersed in
dilute sea water, or else the centrosome is normally cytoplasmic
in position and becomes secondarily intranuclear. The latter
alternative is rejected by Hertwig as far less probable than
the former ; his criticism being based upon his own observa-
tions and experiments with sea-urchin eggs, and upon the
nuclear division of Actinosphcerium, in which he describes the
centrosome as formed from bits of chromatin. Brauer also
described centrosomes in Actinosphceriumy supposing them to

1 Hertwig uses the term " centrosome " in the same sense that morphologists
in this country have used the term " sphere."

2 Verh d. deutsch. zool. Ges., 1896.

2 20 BIOLOGICAL LECTURES.

come from the substance of the pole plates, and hence from the
nucleus, although his observations were somewhat incomplete
and his account obscure. A chromatin origin of a centrosome-
like body was also described by Balbiani in the nucleus of
Spirochona gemmipara, while a nuclear origin of the centro-
some among the Metazoa has been described by Brauer for
the germ cells of Ascaris niegalocephala iinivalens, and by
Riickert for the germinal vesicle of Copepoda.

In the last two cases, however, it seems obvious that such
an origin among higher metazoan cells can have little value in
problems relating to the origin of centrosomes, for here dif-
ferentiation is fully as complete as in any other cell, and the
origin of such centrosomes can give no light on the problem.
To a certain extent, Hertwig's view also is open to the same
criticism, although to a far less degree. It must not be over-
looked that ActinosphcBrium is a highly differentiated proto-
zoon ; its nuclei are much more like the nuclei of Metazoa
than like most other Protozoa, and its reproductive processes
are far advanced along the line of sexual development. It
appears quite probable, therefore, that Hertwig's theory, al-
though exceedingly ingenious and important in the light of
certain facts to be mentioned later, does not go to the bottom ;
he picks up the thread at a point some distance from the spool,
but from that point he gives a satisfactory explanation, at least
from a tentative point of view, of many highly specialized
processes.

Schaudinn's other alternative has opened the way for another
view which explains more primitive conditions in more primi-
tive organisms. This view, which I advanced in a previous
publication, is based upon the lower flagellates and rhizopods.

It is not improbable that the condition of distributed nucleus,
as in Tetramitus, is widely spread among the lower forms of
life. Biitschli ('96) described it for many of the lower algae
and for Bacteria, where the nucleus corresponds closely to that
of ordinary yeast cells, Saccharomyces cerevisecs} In Tetrami-

1 In these forms the chromatin granules lie scattered throughout the cell, or
aggregated in various ways in a cytoplasmic vacuole. Like Tetramitus, they also
have a peculiar structure (the " nucleus " of Wager) in the cytoplasm, which may

NUCLEAR DIVISION IN PROTOZOA.

221

tus the sphere becomes more compact and distinct during the
preparatory division stages, while the chromatin granules col-
lect in a small aggregate in its immediate vicinity. The sphere
then divides, and the chromatin aggregate separates into two
portions. This stage in the division probably corresponds to
the stage described by Schaudinn for the swarm spores of
Paramceba, where the chromatin granules form a ring about the
divided sphere, for, in a later stage, the chromatin granules are
closely aggregated about the daughter-spheres {cf. Fig. 2, E),
Here, also, Schaudinn describes the sphere (his Nebenkorper)

Fig. 3. — Development and division of the swarm spores of Paramoeba Eilhardi (after Schaudinn).

as a portion of a similar body in the cytoplasm of the parent
Amceba, from which it arises by repeated division. The swarm
spores increase by binary fission, the sphere dividing first, as
in Tetramitus. The nucleus, which is a permanent aggregate
of granules, then moves around the connecting strand of the
daughter-spheres, until, like the central spindle of Noctiluca,
it is entirely surrounded (Fig. 3). The connecting strand of
the daughter-spheres of Paramceba is homologous, therefore,
with the central spindle of Noctiluca and of the Metazoa.
After division the sphere lies upon the outside of the recon-
structed nucleus, as in the resting cell of Noctiluca.

correspond in function to the sphere, although I have not been able to follow
this body in division.

222 BIOLOGICAL LECTURES.

It is a temporary stage like this in Tctramitus or Paramceba
that lends support to Schaudinn's alternative. Intermediate
stages between this condition and the permanent intranuclear
sphere may be seen in numerous flagellates, such as Chilomonas
and some species of Trachelomonas, where no nuclear mem-
brane surrounds the chromatin granules. On the other hand,
it cannot be denied that there is very strong evidence for Hert-
wig's view in the observations of Schaudinn upon the nuclei
of Foraminifera,^ and of Siedlecki upon Coccidiida.^ In both
cases the nuclei first appear as solid masses of chromatin, with
(Coccidiida) or without a nuclear membrane. These become
vacuolated in Foraminifera, and membrane and nuclear reticu-

PiG. 4. — Formation of nuclear membrane and reticulum from the chromatin of an originally
homogeneous chromatosphere in Calcitubu (after Schaudinn).

lum, as well as the definitive chromatin, are all derived from the
homogeneous chromatosphere (Fig. 4). There is no mitotic
figure, however, the chromatin fragmenting by what Schaudinn
calls "multiple nuclear division," and the sphere, as such, does
not appear in either case.

Even in the lowest forms the sphere apparently exerts some
force of attraction, perhaps chemotactic, upon the chromatin,
and this force may or may not be strong enough to keep the
granules permanently aggregated ; if not, the distributed nucleus
results ; if so, the intranuclear condition of the sphere is the
outcome. In Paramceba, Noctihcca, in diatoms, and in the
majority of the Metazoa and plants, a nuclear membrane is
formed, and the sphere remains outside of the nucleus. As a

1 Biol. Centralbl., 1894, Bd. xiv.

2 Ann. d. rinst. Pasteur, tome xii, 1898.

NUCLEAR DIVISION IN PROTOZOA. 223

cytoplasmic body (archoplasm), or as a kinetic substance (kino-
plasm), it may undergo further differentiation leading to the
complicated mitotic figures of higher animals and plants. In
all cases, however, the nuclear membrane disappears during
mitosis (save perhaps in Paramoeba and diatoms), and the origi-
nal or primitive condition is thus repeated. By the disappear-
ance of the membrane, the substance of the sphere again comes
into direct contact with the chromatin.

Among the differentiations which the extranuclear sphere
undergoes is the formation of spindle fibers, of centrosomes
(or centrioles of Boveri and Hertwig), and of astral or cyto-
plasmic rays. The spindle fibers in their simplest form arise
as a fibrous arrangement of the sphere or archoplasm. In
Paramceba there is no indication of fibers ; the mass of the
sphere merely pulls out until the ever-thinning connecting
strand is broken. In Noctiluca the connecting strand becomes
distinctly fibrillated, the fibers consisting of a linear arrange-
ment of the microsomes which compose the sphere. In this
condition the strand may be called the central spindle, for its
subsequent position in the mitotic figure shows that it has the
same relations with the chromatin as the central spindle of the
Metazoa, where, according to numerous observers, it arises in
a similar way from the material about the centrosome (archo-
plasm according to Boveri, from substance of the centrodesmus
according to Heidenhain). In many cases, however, the central
spindle of the Metazoa is only a transient structure, the fibers
of which it is composed sooner or later breaking across, so
that it does not penetrate the nucleus as a spindle (many ^gg
cells).

In plant cells, on the other hand, quite a different process
seems to have taken place. Instead of a growing concentra-
tion of the sphere into a definite cell structure, there has been
apparently an ever-increasing diffusion of its substance, until,
as Strasburger states it, the cytoplasm is everywhere penetrated
by kinoplasm, which, prior to nuclear division, comes together
in lines to form the extranuclear spindle.

In those Protozoa where the sphere has become an intra-
nuclear structure a similar history may be postulated. ' In

224

BIOLOGICAL LECTURES.

some forms (simple flagellates) it retains its definite shape,
although Schaudinn's observations on Oxyrrhis marina (a
flagellate with intranuclear sphere like that of Euglena) indi-
cate that the size and condition of this body are dependent upon
the density of the surrounding medium. When this flagellate
is placed in diluted sea water, the intranuclear body swells to
many times its normal size, and may even be forced out of the
nucleus, where, in the cytoplasm, it forms an abnormally large
spindle ; in concentrated sea water, on the other hand, it shrinks
to an abnormally small intranuclear granule. This important
observation opens the way for an explanation of many hitherto
unexplained structures. The nucleus of Amoeba proteus^ for
example, contains chromatin in the form of minute granules,
which are arranged about the periphery of the nucleus, while
the central portion is occupied by a large homogeneous mass,
which can be explained as an enlarged or diffuse intranuclear
sphere. The pole plates also, which are widely distributed
throughout the Protozoa, may be explained as a temporary
accumulation of this ordinarily diffuse archoplasmic substance,^
and thus homologous with the **nucleolus-centrosome" or sphere
of Etiglena and with the centrosphere of Metazoa.

The origin of the centrosome is far more complicated, and
cannot be conceived in so clear a way as the central spindle,
owing to the diverse observations and the complex theories
which have been made to explain it. Certain data, neverthe-
less, have been obtained which may serve as a basis for further
considerations. The process of mitosis in Actinosp/icsrmm is
extremely interesting in this connection, and, observed by
Gruber, Hertwig, Brauer, and recently made the subject of an
extensive re-investigation by Hertwig, it has become perhaps
the most familiar of protozoan mitoses. Both Brauer and
Hertwig describe centrosomes as transient structures in this
cell, appearing, according to the former, during the anaphase
of the mitosis of encysted animals ; according to the latter,
before and during the two maturation divisions by which the
cells prepare for conjugation. Brauer's view as to the origin
of the centrosome and its fate is obscure and unsatisfactory,

1 Cf. Hertwig.

NUCLEAR DIVISION IN PROTOZOA. 225

while Hertwig's appears to be somewhat better founded. Both
observers find that the resting nucleus consists of a basis of
achromatin in the form of a network, which thickens to form
a peripheral zone, while chromatin granules and plastin mate-
rial lie within its meshes. When the nucleus is preparing for
division, two heaps of homogeneous cytoplasm appear at oppo-
site sides against the nuclear membrane (cytoplasmic Kegel).
The nucleus becomes pressed together between these two
masses, until it forms a lens-shape body, while within the
nucleus, homogeneous masses — the pole plates — appear at
each pole. Between the two pole plates the nuclear network
becomes arranged in spindle fibers, upon which the chromo-
somes gather, at first along the entire course of each fiber,
but later concentrated as distinct chromosomes in a central
nuclear plate. Hertwig definitely states that the pole plates
are formed from the same substance as the connecting fibers,
viz.y the network within the nucleus, while the curious cyto-
plasmic and extranuclear heaps against the nuclear membrane
are considered by both Hertwig and Brauer to come from within
the nucleus. The substance of both pole plates and sphere,
therefore, they think to be the same. In the maturation
mitoses a complication is brought about by the presence of
centrosomes. The formation of centrosomes is first indicated
by protoplasmic radiations formed from the homogeneous extra-
nuclear substance (Hertwig). The chromatin, meantime, collects
at the pole of the nucleus turned towards this external mass,
and from it a small portion is budded off to become the cen-
trosome. This chromatin bud rapidly enlarges, assumes a
spongy structure, and two granules appear within it (Hertwig
calls these the *'centrioles" of Boveri). After these are formed,
the spongy mass (Hertwig calls it archoplasm) disappears, leaving
the centrioles isolated within the cytoplasmic heap. Hertwig
did not make out the origin of the other centrosome. Here,
as well as in the vegetative mitoses, the nuclear membrane
remains intact during the entire process ; pole plates, less
conspicuous than in the ordinary mitoses, are developed, and
spindle fibers are formed within the nucleus.

This somewhat bizarre account of the origin of centrosomes

226 BIOLOGICAL LECTURES.

from chromatin cannot be discarded as altogether too fanci-
ful, for numerous observations have indicated that some similar
process takes place. Thus Balbiani ('95) describes in the ciliate
Spirochona the origin of a body which he considers to be the
centrosome from the fusion of numerous granules of chromatin.
Ishikawa and I had grounds for thinking that the centrosome
in Noctiluca comes from the nucleus, while in Metazoa numer-
ous accounts have been given of the origin of the centrosome
from nucleoli and other intranuclear structures. The account
of the origin of the centrosome in the buds of Acanthocystisy
which are formed after direct division of the mother-nucleus, and
which therefore contain no portion of the original centrosome, is
very suggestive. Schaudinn^ ('96) asserts that, until the fifth day
after the bud is formed, there is no trace of a centrosome, but
during the fifth day a centrosome appears within the nucleus
and then makes its way into the cytoplasm, or rather the
nucleus moves away from the centrosome and assumes an
excentric position. Schaudinn does not say from what the
new centrosome is formed.

While there is plenty of room for scepticism on Hertwig's
observations on the origin of the minute body which he calls
the centriole, the later history of this body is strikingly similar
to many recorded observations upon the sphere and centrosome
in Metazoa. The enlargement of the sphere and the persist-
ence of the centriole, as recorded by Lillie, Mead, Van der
Stricht, Sobotta, etc., may be described in almost similar terms.
But all of these cases are different from ActinospJicerium in that
the substance of the sphere is reported as forming a central
spindle between the centrosomes, and this structure, according
to Hertwig, is absent in the heliozoon. What then is the sig-
nificance of the intranuclear spindle fibers in this and in so
many other Protozoa where pole plates are present t An
answer is immediately suggested by the relation which these
fibers have to the chromatin granules. Hertwig and Brauer
maintain that the granules are strung upon them into chromo-
somes, while Schewiakoff describes a similar connection in
Eiiglypha^ and Schaudinn in Actinophrys. Central spindle

^ Loc. cit.

NUCLEAR DIVISION IN PROTOZOA. 227

fibers have nothing to do with chromosomes. The intranu-
clear fibers of forms with pole plates thus correspond to
mantle fibers of the Metazoa and Noctihicay and the evidence
is strong that they, like the central spindle fibers, are formed
from the substance of the intranuclear sphere. The origin
of the extranuclear cytoplasmic masses from the intranuclear
sphere {Aciinosphcerium) lends support to the theory, first
advanced by Hertwig, that the centrosome arises by emergence
from the nucleus, while it also explains the origin of mantle
fibers from linin substance of the nucleus as described by
numerous observers upon different forms of Metazoa.

Summary and Conclusions.

The component parts of the mitotic spindle in Metazoa may
be thus traced back to the generalized condition found in many
Protozoa. Although many points are obscure and need verifi-
cation, one fact, at least, appears well supported, vis., the
so-called achromatic and chromatic portions of the mitotic
figure were originally more or less independent bodies which
came together only during mitosis. The view advanced by
Lauterborn, that the original form of the kinetic center was in
some binucleated ancestor such as Amoeba binucleata, meets
with many obstacles. It must be assumed, according to this
theory, that one nucleus not only changed its form, but also
became completely changed in function. Again, it must be
assumed that differentiation of this ** kinetic" nucleus must
have resulted in the well-differentiated cytoplasmic body
(sphere) of Tetramitus or Paramoeba as well as the intranu-
clear centers of many flagellates, forms which are probably as
low in the scale of living things as Amoeba. It seems hardly
necessary to regard this body, in its original condition, as a
nucleus which later becomes differentiated into a well-defined
kinetic substance with degeneration of the original nuclear
contents. It seems much more simple, and, in the light of
facts, much more probable, that the kinetic substance was as
definite a portion of the original primitive cell as the nucleus
itself, although formed possibly from metamorphosed chroma-

2 28 BIOLOGICAL LECTURES.

tin. This practically restates Boveri's archoplasm theory as he
gave it out twelve years ago — a theory which has been virtually
unaccepted so far as the structures for which it was developed
are concerned, but which finds its greatest support in the Pro-
tozoa. Originally cytoplasmic, this substance, after taking the
form of a definite kinetic body, became intranuclear during
division of the cell — a condition which is permanent in the
majority of the Protozoa, and which is repeated, during mitosis,
by every cell throughout the animal and plant kingdoms.

Within the nucleus the archoplasm body, at first definite
and compact (Flagellata), became secondarily diffuse, appearing
indefinite in structure {Amoeba proteus) or evenly distributed as
linin throughout the nucleus (Actinosphcermm, Euglypha, Acii-
nophrysy and nuclei of higher animals and plants), while, during
division, it now re-collects to form the pole plates {Infusoria
and many Sarcodina), or spindle fibers and *' archoplasm "
(Metazoa).

The origin of the extranuclear sphere may be interpreted in
two possible ways. It may be regarded as a direct continua-
tion of the primitive cytoplasmic condition through forms like
Tetramittis, Paramceba, and Noctiluca, where it never becomes
intranuclear more than the central spindle does in the latter
form ; or it may be regarded as nuclear in origin, being sec-
ondarily extruded from the nucleus, as Hertwig describes in the
case of Actinosphcerium, or Schaudinn in the case of Acantho-
cystis. In the former case both Brauer and Hertwig are agreed
that the cytoplasmic mass {Kegel) comes from the achromatic
substance of the nucleus, hence from the intranuclear archo-
plasm, and as this substance becomes the matrix for the later
appearing centrosomes, it is not incorrect to speak of it as the
sphere.

Whatever its primitive relation to the nucleus may have been,
the cytoplasmic sphere is large and permanent in Noctiluca, but
it becomes smaller in Metazoa; and just as it becomes diffused
throughout the nucleus in Protozoa and Metazoa, so it may be-
come diffused throughout the cytoplasm in Metazoa — a condition
which Boveri postulated in his remodeling of the archoplasm
theory ('95). In higher plants, also, it may be regarded as per-

NUCLEAR DIVISION IN PROTOZOA. 229

manently in this diffused condition, as Strasburger practically
postulates in his kinoplasm theory. In plants, however, no cen-
trosphere is formed, the spindle fibers alone representing the
archoplasm. The increasing diffusion of the substance of the
sphere (archoplasm) explains the variations in the mitotic figures
as given at the outset of this chapter, the central spindle gradu-
ally becoming fainter and fainter in a series of forms, until it
finally is lost. It also explains a number of anomalous cases,
such as the formation of asters in poisoned echinoderm eggs,
described by Hertwig, * or the artificial asters recorded by
Morgan.

The mantle fibers, or fibers which connect the kinetic cen-
ters and the chromosomes, also have an archoplasmic origin,
and are primarily nuclear in position. In this primitive con-
dition they are seen in ActinosphcBriicm (Gruber, Hertwig,
Brauer), in Euglypha (Schewiakoff), in Spirochona (Biitschli,
Plate, Hertwig, Balbiani), and in micronuclei (Hertwig, Mau-
pas). In all of these cases they run from pole to pole, or from
pole plate to pole plate. With the extranuclear position of the
kinetic center, the rupture of the nuclear membrane becomes
necessary, for only then can the union of central spindle and
mantle fibers take place. With this in mind the controversy
over the nuclear or cytoplasmic origin of mantle fibers gains an
added interest ; for in either case they are formed from the
same substance, which may be either nuclear or cytoplasmic
in position.

A comparison, I believe, may be drawn between chromatin
and archoplasm. In its earliest form the chromatin is massed
in one or several large reservoirs, which later break down into
granules distributed throughout the nucleus. The granules
may be permanent or temporary, but in all cases they come
together again as chromosomes. Similarly, archoplasm, in its
primitive form, is a single mass which becomes diffused through-
out the nucleus and cytoplasm, but in all cases comes together
again in the form of spheres or spindle fibers. As chromatin
is considered a definite substance or modification of protoplasm,
so also may archoplasm be regarded in the same light.

The centrosome question, finally, is made no clearer by study

230 BIOLOGICAL LECTURES.

of the Protozoa. Recent observations lead to the view that it is
not a permanent organ of the cell (Brauer, Hertwig, Schaudinn,
Morgan, etc.), while its complete absence in higher plants and its
appearance in certain mitoses while absent in others of the same
species (Actinosphcerium) justify Flemming's view that centro-
somes may be present in mitosis, but not necessary for it. The
origin of the body, which in various cases is called the centro-
some, is equally unsatisfactory. Hertwig's and Balbiani's view,
that it comes from chromatin which may {Actinosphceriuni) or
may not [Spirochona) become extranuclear, is opposed by the
views, often hypothetical, of many observers upon centrosomes
in Metazoa. Further and guarded observations, especially
among the Protozoa, must be made before the various con-
flicting views can be reconciled.

Columbia University,

December i, 1899.

FOURTEENTH LECTURE.

THE SIGNIFICANCE OF THE SPIRAL TYPE OF

CLEAVAGE AND ITS RELATION TO THE

PROCESS OF DIFFERENTIATION.

C. M. CHILD.

During the sixty years since the cell theory was propounded
by Schleiden and Schwann (1838-39) it has come to dominate
almost completely the various departments of biological re-
search. In the introduction to his book, The Cell ('97), Wilson
remarks that "it has become even more clearly apparent that
the key to all ultimate biological problems must, in the last
analysis, be sought in the cell." Notwithstanding the wide
acceptance of the cell theory, a voice of protest has from time
to time been heard, from both the botanical and the zoological
side, based upon actual observation as well as upon theoretical
considerations. Prominent among those who have upheld
what Whitman ('93) designates as the ''organism standpoint"
are Sachs, Rauber, Adam Sedgwick, and Whitm.an himself.
It is not my intention to enter into a historical discussion of
the views of the various authors, but to direct attention to one
phase of the problem, viz,^ the relation of the process known
as differentiation or organization of the cell.

As is universally recognized, differentiation may occur within
the limits of a single cell, as in the Protozoa, or within an
organism consisting of many cells, and, therefore, cannot
necessarily be connected with the process of cell formation.
This fact is most clearly expressed by Whitman as follows :
" It is not division of labor and mutual dependence that con-
trol the union of the blastomeres. It is neither functional

231

232

BIOLOGICAL LECTURES.

economy nor social instinct that binds the two halves of an
^ZZ together, but the constitutional bond of individual organ-
ization. It is not simple adhesion of independent cells, but
integral structural cohesion.

*' That organization precedes cell formation and regulates it,
rather than the reverse, is a conclusion that forces itself upon
us from many sides. In the Infusoria we see most complex
organizations worked out within the limits of a single cell.
We often see the formative forces at work, and structural fea-
tures established before fission is accomplished. Cell division
is here plainly the result, not the cause, of structural dupli-
cation. The multicellular Microstoma behaves essentially in
the same way as the unicellular Stentor or the multinucleate
Opalinopsis of Sepia. The Microstoma organization duplicates
itself and fission follows. The chain of buds thus formed bears
a most striking resemblance to that of Opalinopsis, and the
resemblance must lie deeper in the organization than cell
boundaries.

**We must look entirely behind the cellular structure for
the basis of organization."

It is the continuity of organization, whether the organism
be composed of one cell or many, which is insisted upon
throughout this paper, and in agreement with Sachs ('82), cell
formation is regarded as of only " secondary significance."

Wilson ('93, p. 595) stated in regard to Amphioxus that
" the unity of the normal embryo is not caused by mere juxtaposi-
tion of the cells. They (the facts) indicate that this unity is not
mechanical but physiological, and point toward the conclusioft
that there must be a structural continuity from cell to cell that is
the medium of coordination and that is broken by mechanical dis-
placements of the blastomeres.'' ^

McMurrich ('95), in a lecture of this series delivered in 1894,
after a discussion of the early development of centrolecithal
ova, concludes *' that in embryological development the differ-
entiation which occurs is a differentiation of the entire organ-
ism and not of the constituent parts of which it is composed ;
physiologically, if not morphologically, every organism is a

1 Italics in this and in all succeeding quotations are as in the original.

THE SPIRAL TYPE OF CLEAVAGE. 233

syncytium, and future theories of heredity must take this
into consideration."

Even more directly, I believe, do the facts point to the
existence of such a structural continuity in the eggs of anne-
lids, mollusks, etc., where the division of a single cell at other
than the proper time would in many cases disarrange the
whole complex. The theory of prearranged harmony makes
too great a demand upon the cell as distinguished from the
organism to be within the limits of probability in this case.

We have, moreover, as a basis for our belief in the continuity
of organization, not merely the results of comparative study of
the unicellular and multicellular forms and the inferences from
experiment, but there are a rapidly increasing number of obser-
vations showing the presence of actual protoplasmic continuity
between cell and cell. Observations of this kind upon plants
have been very numerous and include a wide range of forms.
(For the literature upon this subject see Zimmermann, '93, and
A. Meyer, '96). In animal tissue intercellular connections have
been found to be of wide occurrence among epithelial cells,
and from time to time, especially during the last few years, our
knowledge of these structures has been increased by their dis-
covery under many other circumstances. A brief resume of
the state of our knowledge at the time his book appeared is
given by Wilson ('97), concluding as follows: *'As the subject
now lies, however, the facts do not, I believe, justify any gen-
eral statement regarding the occurrence, origin, or physiological
meaning of the protoplasmic continuity of cells ; and a most
important field here lies open for future investigation."

It is interesting to contrast with Wilson's somewhat conserv-
ative view the opinion expressed by A. Meyer ('96) in a paper
published almost at the same time that Wilson's book appeared.
After a resume and discussion of the observations along this line
upon both the botanical and the zoological side, Meyer sums
up his views as follows : " So far as I can see at present, all
our knowledge upon this subject favors the view that proto-
plasmic connections occur between practically all cells of every
individual, that a characteristic of both the animal and vege-
table individual is the possession of one continuous mass of

2 34

BIOLOGICAL LECTURES.

cytoplasm. Whether the individual consists of a uninucleate
cell, a multinucleate cell, or a system of cells, the cytoplasm
forms a continuous whole."

And in another paragraph he says : "■ They (the protoplas-
mic connections) do not appear to be, like the nerve fibers,
alloplasmatic organs adapted for special purposes, but strands
of normal cytoplasm which connect the cytoplasm of neighbor-
ing cells. Their significance is probably to be explained as

'^'^^^^^^m^W'£^-y:^t:' '•"

Fig. I . — A small part of the blastula of the starfish, showing the intercellular connections
(after Andrews).

follows : All activities and material changes in that part of the
multicellular system (the individual) which we designate as a
protoplast have a direct effect upon the constitution of these
protoplasmic processes, and the alterations which occur in the
organization of these exercise, in their turn, a direct influence
upon the activity of the neighboring cells."

Hammar's observations upon sea-urchin eggs ('96) were made

THE SPIRAL TYPE OF CLEAVAGE.

235

upon fixed material, but since these were made a very inter-
esting series of observations upon living cells have appeared
(G. F. Andrews, '97 ; A. E. Andrews, '973, '97b, '98a, '98b, '98c),
which must, I believe, lead us to the conclusion that intercel-
lular protoplasmic continuity between animal cells is at least
of very wide occurrence, if not universal, during developmental
stages (Figs, i and 2). This being the case, we have, it seems
to me, very strong evidence upon both the botanical and zoologi-
cal side for the correctness of
the so-called organism stand-
point, as distinguished from
any cell theory. And I be-
lieve we must admit that
intercellular protoplasmic
continuity is of the greatest
importance in our interpreta-
tion of the organism.

Although my study of
cleavage was begun from a
point of view coinciding fully
with the cell theory, my observations have led me to the belief
that any explanation of the spiral cleavage is impossible upon
that basis. The developing Qg<^ of the turbellarian, annelid,
and mollusk is, I believe, more than a cell mosaic ; it is dis-
tinctly an organism at every stage.

I desire, then, to discuss the spiral type of cleavage, not as
a series of self-differentiations or cellular interactions, but as a
series of processes, each of which has its cause in the organism,
as distinguished from this or that cell. From this point of view
alone, as I believe, is it possible to explain the various' phenom-
ena observed, and especially the lack of correspondence in ori-
gin and fate of the blastomeres, which is not explicable by any
strictly cellular theory of -development.

In his classical paper upon Nereis^ Wilson ('92) has given us
a most valuable discussion of the origin of the spiral form of
cleavage, and it is desirable at this point to review this briefly.
In general, his view is as follows : ^^ The fundamental forms of
cleavage are primarily due to mechanical conditions, and are only

Fig. 2. — Two-cell stage of the egg of a fresh-water
nemertean, showing the ectoplasmic activity in
the cleavage furrow and the rounded form of the
cells immediately after division.

236 BIOLOGICAL LECTURES.

significant morphologically in so far as they have beeii remodeled
by processes of precocious segregation'' As regards the spiral
cleavage, he believes that it '* is a secondary derivative of the
true radial type,'' being the result of ^' mutual pressure among
the cells." Furthermore he says: ''But the remarkable fact,
and one which does not seem to be very clearly recognized, is
that the effect of these mechanical conditions has become hered-
itary!' And again : '* We now come to the main point, which
is that the rotation of the cells is in the spiral type in many
cases predetermined in the parent cells , as is proved by the posi-
tion of the spindles aiid by the form of division!' And lastly :
"To sum up, I conclude that the spiral form of cleavage is
owing to a precocious appearance of the alternation of the
cells, which, in its turn, is a result of mutual pressure."

The question now arises as to the significance of the mutual
pressure and the reason for the establishment of the spiral
cleavage as hereditary. Discussion along this line must, of
course, be of a purely speculative nature, and, although there
is a certain truth in Driesch's contention that explanations of
this kind do not explain ('99), I still believe that they are of
value as showing, perhaps, some of the steps in the process, and
a possible or probable motive for these steps, even though the
why and the how of the reaction of the organism is a mystery.

It is clear that the alternation of blastomeres that would
result from mutual pressure, etc., produces a form of cleavage
in which the blastomeres, like soap bubbles, conform to the law
of minimal contact surfaces, and that therefore each blastomere
is in contact with the greatest number possible of the others.
From the so-called organism standpoint such a relation must
unquestionably be advantageous, unless we assume that the
peculiar property of protoplasm which binds the cells into an
organism is capable of some mysterious ''Fernwirkung," which
does not seem probable. Roux's ** Cytotropismus " appears,
perhaps, to be such a '' Fernwirkung," but I would like to
suggest the possibility that this so-called cytotropism may
resolve itself in some cases into the establishment of extremely
fine protoplasmic connections.

From the standpoint of the cell theory this arrangement of

THE SPIRAL TYPE OF CLEAVAGE.

237

Fig. 3. — Two-cell stage of a fresh-water nemertean
after the attainment of the resting stage, showing
the flattening together of the cells and the tem-
porary intercellular cavity.

the blastomeres would appear distinctly disadvantageous, for
here a much larger portion of each blastomere is in contact
with others than in the looser cell aggregate, so that respiration
and excretion apparently cannot occur as freely as in a loose
mass of rounded cells. The periodic appearance of the large
intercellular cavities filled with liquid, as described by Kofoid
('95) for Limax, and which
occur in other forms as well
(Figs. 3 and 4), is probably
a provision for aiding the
excretion, some such provi-
sion having become neces-
sary under certain conditions
(fresh-water or terrestrial life
in moist localities), in conse-
quence of the extremely inti-
mate unionof the blastomeres.
Moreover, we do find that
in many forms, and espe-
cially in those in which the
spiral type of cleavage oc-
curs, there is a distinct
condensation of the proc-
ess of development, which
finds expression in the
early appearance of dif-
ferentiation and morpho-
genesis for forms in which
any such tendency existed ;
it must be clear that a con-
formation to the principle F^^- 4. -Four-cell stage of a fresh-water nemertean,
^ ^ showing the temporary intercellular cavity.

of minimal contact sur-
faces among the blastomeres will afford the most direct and
intimate communication possible between them.

It is then, I believe, on account of the necessity for just this
intimate communication between the component parts of the
organism that the spiral cleavage, which in its simplest form
fulfills these conditions in the most perfect manner possible.

238

BIOLOGICAL LECTURES.

Fig. 5. — Arenicola. Fifth cleavage viewed
from upper pole. 2d, the "first somato-
blast" already formed.

has become so firmly established, even in spite of the fact that
the free respiratory and excretory surfaces of the cells have
been reduced, in consequence of its presence, to such an extent
that in some eggs a compensatory arrangement in the form

of the intercellular cavities is
necessary.

How this process has oc-
curred we cannot explain at
present, but that it has occurred
I believe we must admit, and
that there is sufficient motive
for its occurrence I trust I have
shown.

The next point which requires
investigation is the fact that, in
the great majority of cases, the
direction of corresponding cleav-
ages is the same, i.e.^ the second cleavage is laeotropic and
the third dexiotropic, and so on (Fig. 5). Why should this
constancy occur } Directly correlated with this problem is the
question as to the cause of the few cases of reversed cleavage
in gasteropods.

If we suppose the cleavage to be hereditary, the direction of
the early cleavages is no longer a matter of chance, and the
direction of succeeding cleavages would follow according to the
principle of regular alternation. This being the case, the assump-
tion is not at all unwarranted, that, of the two possibilities, one
has become established and the other has disappeared. What
factor determined which one of the two possibilities should
become the rule is beyond our knowledge, but it is certainly
a fact that in nearly all spiral cleavages the second division is
laeotropic and the third dexiotropic, .etc. ; and it is just as
certain that this constancy is not merely the result of mutual
pressures in groups of four, eight, or more blastomeres.

The occasional appearance of reversed cleavage is explicable
as a variation from the normal type, — the only variation possi-
ble with continued conformation to the spiral type, — or in some
cases it may be that this variation for some special reason has

THE SPIRAL TYPE OF CLEAVAGE. 239

become the rule. According to this view the reversed spiral
cleavage has arisen from the more common type.

Our knowledge of the cleavage of reversed forms is very
incomplete as yet. As regards Planorbis, Rabl ('79) shows
clearly a reversal of the usual direction of cleavage, and the

work of Holmes ('97) affords con- ^ -.^^

firmation (Fig. 6). Crampton ('94) >^- -.^ ^,^~^

discovered that the cleavage of /; /^-^ j / \\
Physa is reversed. The case of /\\ / ;^ \ \

Janthma is doubtful. Haddon / \\——J /V_y I \
has figured an apparent left spiral / >- — -V^J / I

in the third cleavage, hvXjanthina \ -1 ^^^ /^ \ /
is, at least usually, dextral. Conk- \ \ J / /^ ^

lin ('97) points out the possible ^Jx J ^^^[/

causal relation between reversed ^\^^^_2!^:^^'^

cleavage and reversed asymmetry yig. e.—pianorbis. Eight-ceii stage. An

in the adult, but the point I de- -ample of reversed cleavage (after Holmes).

sire to emphasize is the probable secondary nature of this
reversal, its occurrence as a variation somewhat of the nature
of a sport, which, at least in some cases, was inherited and
became the rule. If we accept this view, the occurrence of
sinistral gasteropods must be regarded as without phylogenetic
significance.

We find among the gasteropods some species with occasional
sinistral individuals, others in which the sinistral individuals are
more common, and still others in which all are sinistral. A
study of the cleavage in species where sinistral forms are com-
mon is greatly needed. I am inclined to believe, however, that
reversed cleavage precedes reversed asymmetry, and further
that the cause of the reversed asymmetry is the reversal of the
cleavage and not vice versa; and, finally, that cases of reversal
are secondary modifications, and have arisen after the relation
between cleavage and adult asymmetry, of which Crepidula
affords an example, had been established.

Having reached this point in our consideration of the spiral
cleavage, it is pertinent to inquire whether we find in this type
of cleavage, as it exists at present, any evidence of the conti-
nuity of organization, or whether we must admit, as appears to

240 BIOLOGICAL LECTURES.

be generally believed, that the cells possess a high degree of
independence.

And first we turn to the evidence furnished by experimental
work upon the spiral type of cleavage. Evidence of this kind
is limited to Crampton's experiments upon Ilyanassa ('96),
Wilson's upon Nereis ('92), and Fujita's upon Aplysia ('97),
which latter form, according to Blochmann's account of the
cleavage ('83), adheres strictly to the spiral type.

Crampton succeeded in isolating various blastomeres of Ily-
aiiassay some of which continued to develop for a longer or
shorter period. As I showed in 1897, he is in error in regarding
the larvae resulting from these isolated blastomeres as strictly
partial larvae, since, as he himself states, the entodermal blas-
tomeres are completely enclosed by ectoderm, which would be
impossible without modification of the form of cleavage. In
order to accomplish this enclosure the ectoderm cells must cover
a greater amount of surface than that normally covered by them,
for, in addition to covering the outer side of the large entoder-
mal blastomeres, they must also cover that portion of these cells
which is normally in contact with the other entodermal blasto-
meres. This modification in form after the injury and the
complete closure of the blastopore indicates that the mutilation
has affected the whole organism and that a partial rearrange-
ment has occurred. And even as regards the earlier cleavages
of isolated blastomeres, which Crampton has studied in detail,
there is evidence in his figures of some degree of modification
in the form of cleavage. The fact that the earlier cleavages do
continue so nearly as if the rest of the Qgg were present, even
though the blastomeres assume a spherical form, seems to me
to indicate simply that the form and direction of cleavage are so
fixedly determined by heredity that the shape of the cell has
little influence upon them. Intercellular protoplasmic con-
nections would also tend to keep the form of cleavage constant.
But gradually a rearrangement does occur, or complete closure
of the blastopore would be impossible. The fact that the pro-
totroch is only partial is due to the early differentiation of these
cells. It is not proven, moreover, that it would not have been
completed if the larvae had lived long enough.

THE SPIRAL TYPE OF CLEAVAGE. 24 1

A very interesting result obtained by Crampton is the failure
of those eggs in which the yolk lobe had been removed to form
a cell corresponding to the mesoblast in structure, although a
cell which is packed with yolk spherules, like the other entoder-
mal cells, is formed. According to Crampton the yolk lobe
appears to consist almost wholly, or wholly, of yolk spherules ;
and in any case it cannot contain the bulk of protoplasm repre-
sented by the mesoblast, so that it is apparently not the loss of
mesoblast material, but some other effect of the mutilation which
results in the defect. The result of this experiment indicates,
it seems to me, a considerable amount of rearrangement of the
^gg substance.

But perhaps the most interesting result of all these experi-
ments is the extreme sensitiveness of the ^gg to injury, as in-
dicated by the very small proportion of mutilated eggs which
continued to develop. This fact points most unmistakably to
the existence of a very high degree of interdependence among
the blastomeres. In most cases the portion of the ^gg observed,
although itself apparently uninjured, is so crippled by the loss
of the part removed that death ensues. Even in those cases
where development continued for a time the embryo never suc-
ceeded in completing it, but sooner or later succumbed.

Wilson succeeded in modifying the cleavage of Nereis by
means of pressure, so that the micromeres were not formed till
the fourth division, and then eight of them appeared above
eight more, larger, yolk-laden macromeres. These eggs devel-
oped into trochophores with double the usual number of ento-
dermal blastomeres. Here again the insignificance of the cell
as compared with the organism is apparent.

I have not had access to Fujita's paper upon Aplysia^ but in
a brief abstract of it, it is stated that he found that isolated
blastomeres were able to restore the lost portions and to reas-
sume the normal shape. Whether the rearrangement occurs
directly or in later stages is not stated. In any case the effect
of the mutilation upon the remaining portions is evident. If
complete rearrangement does take place, it may be because the
Q,gg of Aplysia is more hardy than that of Ilyanassa.

Turning now to the normal cleavage, we find there a number

242 BIOLOGICAL LECTURES.

of phenomena which, as I believe, can be satisfactorily inter-
preted from the organism standpoint alone.

And first I am inclined to believe that the flattening of the
first two blastomeres and of later ones against each other, a
phenomenon of very wide occurrence {cf. Figs. 2 and 3),
may be the result of material connections between the cells,
which bring the cells into contact with each other over the
largest possible portion of their surface after the intracel-
lular pressure accompanying mitosis has begun to diminish.

In many cases the reversal of the direction of spiral cleav-
ages, especially when occurring in one or two quadrants, is
indicative of interrelation; for, as Conklin states ('97, p. 186),
the ultimate cause of reversal is, in most cases, the precocious
appearance of certain organs or planes of symmetry.

The relative time of differentiation of various organs, and
especially of the early larval organs, such as the prototroch,
affords to my mind a .most striking example of the interrelation
of all parts of the developing ^gg. For instance, in Arnphitrite
and Lepidonottis the primary trochoblasts become ciliated before
the sixty-four-cell stage (Mead, '97). In Nereis the cilia appear
at about the tenth hour, at a stage when the closure of the
blastopore is nearly completed. In Arenicola the first traces
of ciliation appear when the embryo is from seventeen to nine-
teen hours old, after the blastopore has closed completely and
the embryo consists of hundreds of cells. In Crepidula the
trochoblasts are very small and remain apparently quiescent
up to a late stage, then increasing in size, and still later
attaining their definitive functional condition. In IscJmochiton
(Heath, '99) the velum becomes ciliated much earlier than in
Crepidula^ — thirty-one hours, — and there is no marked in-
crease in size of the cells. Now in all these cases the so-called
primary trochoblasts arose in a very similar manner and at
essentially the same time in the history of the ^gg, yet for
some reason the time of formation appears not to be related
directly to the time at which they become functional.

It is important to note that in each case the differentiation
occurs at such a time that the trochoblasts shall be prepared
to perform their function when called upon by the environment.

THE SPIRAL TYPE OF CLEAVAGE. 243

The simplest expknation would seem to be that the energy of
the Qg^ is used where most needed. In cases where the larva
swims at a very early stage, the cilia appear correspondingly
early, but in other cases the trochoblasts remain apparently at
rest, perhaps for a long time, till some time before they are
needed the differentiation appears. To me this appears to be
simply a case of the saving of energy. The energy of the ^^^
is so exactly distributed that none is wasted in the development
of organs before they are needed. The distribution of the
energy must be connected with some physical or chemical
change, and in this way intrinsic differences may arise between
the various cells, but the differences between the quiescent
trochoblasts and the other cells do not necessarily signify
that the former contain a special substance which makes them
distinctively trochoblasts from the time of their formation. Of
course at some time they do become distinctively trochoblasts,
but simply because of their relation to the whole.

Again, if the course of the cleavage in different forms be
followed, the different time relations in the division of the
various cells indicate the nicest adjustment to prevailing con-
ditions. I cannot see how such adaptations could arise, nor
how others can arise in the future without the closest relation
between the parts.

The close relation of many cells to their environment, even
after differentiation, is evident from many of the facts of regen-
eration, and experiments on the early developmental stages of
many eggs lead us to the same conclusion. In those cases
which seem to point to a directly opposite conclusion, it is
sometimes perfectly clear that the apparent independence of
the cell or cells in question is really due to differentiation, but
there is also reason to believe that in many cases the essential
features of the environment are not recognized, and that what
appears to be independence is in reality dependence upon
factors of the environment whose existence is unsuspected.

It is difficult to understand how ''precocious segregation"
could have appeared at all without a very intimate interrelation
of parts ; but, on the other hand, it seems clear that the exist-
ence of a high degree of continuity would favor it. Once estab-

244

BIOLOGICAL LECTURES.

lished, however, as we find it, there is little reason to suppose
that, since its establishment, the blastomeres have lost their
connection — physiological or otherwise — with each other,
unless the existence of a« high degree of self-differentiation
is actually proven.

Now the spiral cleavage is, as shown above, the most favor-
able form of cleavage possible for the establishment of physical
continuity of organization, for each cell is brought into contact
with the greatest possible number of cells. Is it not a justifi-
able conclusion that its wide occurrence, firm establishment, and
invariability, especially in those forms in which a condensation

f/2Z

Fig. 7.

Fig. 8.

Fig. 7.^ — Arenicola. Formation of cross by radial divisions.

Fig. 8. — Arenicola. The terminal cells of the ventral arms of the " cross " show spindles which
are distinctly spiral.

of development — a precocious segregation — is found, are due,
at least in some measure, to the fact that it affords the least
possible physical obstacle to intercommunication between the
component elements ? For, if a tendency to condensation be
present, the possession or acquisition of the spiral form of
cleavage must be of great advantage. Or it may be that the
presence of this form of cleavage has led to that form of con-
densation known as precocious segregation. Whichever alter-

1 In this and following figures of cleaving eggs the annelid cross (" rosette-
series " of Conklin) and its descendants, and the somatic plate arising from 2d
are enclosed by a heavy line; and the cells of the prototroch and paratroch, so far
as formed in each case, are stippled.

THE SPIRAL TYPE OF CLEAVAGE,

245

native is the correct one, it seems clear that the spiral form of
cleavage and precocious segregation are very closely connected
in annelids and mollusks.

In the forms under consideration — forms possessing spiral
and determinate cleavage — there is imposed upon the spiral
cleavage proper a strictly bilateral form of cleavage, in which

Fig. 9.

Fig. 10.

Fig.

Fig. 12.

Fig. 9. — Arenicola. Somatic plate consists of three cells which have arisen by spiral divisions.

Fig. id. — Arenicola. Somatic plate after spiral divisions have ceased.

Fig. II. — Arenicola. Somatic plate. The cell dp gives rise to the two dorsal cells of the para-

troch, the cell xl to the two lateral paratroch cells of the left side.
Fig. 12. — Arenicola. Somatic plate. The cell fl^ dividing spirally.

the axes of symmetry are the same as in the adult {cf. Fig. 5
with Figs. 7 and 8, also Fig. 9 with Fig. 10). This bilateral
type of cleavage is evidently connected with the process of
morphogenesis, and I desire to call it, for convenience, a
moi'phogenetic cleavage. A morphogenetic form of cleavage

246

BIOLOGICAL LECTURES.

^<^'

Fig. 13. — Arenicola. Somatic plate, dp'^ and
dj>^, cells resulting from division of dp ; at the
top of cut the four " intermediate girdle cells "
■which are formed in front of the prototroch, but
come to lie behind it.

is not merely determinate, as Conklin has defined the term,
but the various processes of cleavage are so arranged that the
interaction of the component parts leads directly to the estab-
lishment of form. The factors of morphogenesis in cleavage
are direction and time of division and size of the products,

accompanied, of course, by
more or less differentiation,
according as morphogenesis
proceeds more or less rapidly;
but differentiation, as I be-
lieve, is a process wholly dis-
tinct from cleavage, whether
cleavage be indeterminate, de-
terminate, or morphogenetic.
Morphogenetic cleavage, of
course, possesses in general a
real phylogenetic significance.
A sufficient cause for the
establishment of this form of
cleavage is, I believe, the
saving of energy which must
occur. If the energy of cell
division can be utilized to
bring material to the point
where it is needed, and at the
time when it is needed, it is
evident that less energy will
be required than if extensive
migrations were necessary
after the formation of the cells.
A study in detail of the growth
and concrescence of the somatic plate in Arenicola (Figs. 9-18),
or in any other form in which it has been fully worked out,
affords one of the most beautiful illustrations of the perfect
coordination in morphogenetic cleavage, and I think that no
one who follows it can doubt that each and every cell division
plays its part in bringing about the concrescence with the least
expenditure of energy.

Fig. J.4. — Arenicola. Somatic plate,
cells of paratroch formed.

Dorsal

THE SPIRAL TYPE OF CLEAVAGE.

247

Fig. 15. — Arenicola. Portion of so-
matic plate, showing position of para-
troch when first formed, x x, the
first cells which meet in the median
line during concrescence.

Since every cell plays a definite part,
it is evident that morphogenetic cleav-
age is necessarily accompanied by a
certain degree of segregation, i.e.^ the
material which is to form the various
organs is more or less completely
segregated into definite cells or groups
of cells, though they may not be dif-
ferentiated, or, rather, the differentia-
tion may be simply one of position.

It appears probable that this bilat-
erally symmetrical, morphogenetic
cleavage has encroached upon the
spiral period during the course of
phylogenetic development and short-
ened it, and that, in addition, inde-
pendent modifications have arisen
in the different groups, thus altering
the relations of the constituent ele-
ments. A study of the differences piate
in cleavage in various forms brings to
light a number of facts which appear
to justify these conclusions.

In the following paragraphs I have
brought together some of the facts
which appear to me to furnish evidence
upon these points.

I. The spiral period proper ends
with different cell generations in d\i-Y\o.\T.—Arenicoia. Portion of so-

r . c matic plate. The cells x xva contact

ferent forms. in median Une.

Meager as are the data for a comparison
between the different larger groups, they
indicate what I believe will be amply
demonstrated in the future, viz.^ that as
the process of development becomes tnore
and more highly modified^ the strictly spiral
cleavage^ whe^i originally present, gives way
at an increasingly early stage to a morpho-

FlG.

6. — A renicola . Portion of somatic
Concrescence proceeding.

Fig

18. — A renicola. Para-
troch just before ventral ends
are brought into contact by the
concrescence of somatic plate.

248

BIOLOGICAL LECTURES.

genetic form of cleavage^ which finally attains its most pro-
nonnced expression in teloblastic growth.

A comparative study of the cell lineage of various Poly-
chaeta shows a considerable degree of difference within the
limits of this group as regards the stage at which the spiral
form of cleavage disappears {cf. Figs. 7, 19, 20), but it is
impossible to discuss this point in detail here.

2. Cases of what maybe called "reversion" to the spiral
type of cleavage occur, i.e.^ cells which have departed from the
spiral type may return to it in later stages for one or more gen-
erations. Numerous examples of this " reversion " occur. I

Fig. 19.

Fig.

Fig. 19. — Podarke. Cross formed by spiral divisions (after Treadwell).

Fig. 20. — ChcBtopterus . Cross does not appear because of extreme spiral character of divisions
which commonly give rise to it (after Mead).

wish to mention here some of my own observations on Areni-
cola without reference to other forms, for, where special at-
tention has not been paid to this point, figures and general
statements are usually too indefinite to serve in this connec-
tion. The cross in Arenicola is formed by radially symmetrical
divisions (Fig. 7). In the division following this the four cells
terminating the arms of the cross often show traces of spiral
divisions. This is especially noticeable in the ventral arms of
the cross (Fig. 8, i<2"% i^'"^), which often curve at the tip
toward the right side of the Q,gg. Both the spindle and the
position of the cells after division indicate in many cases the
spiral nature of the cleavage. The direction is laeotropic, as it

THE SPIRAL TYPE OF CLEAVAGE. 249

would be if the regular alternations of the spiral period had
continued.

In the formation of the two dorsal paratroch cells (Fig. 14,
^'^ dp"^), the last two divisions which they and their sister
cells undergo are, so far as direction and alternation are con-
cerned, true spiral divisions, although they givq rise to a bilat-
erally symmetrical group of cells. The first of the two is
laeotropic (Fig. 12, dp, Fig. 13, dp^, dp"), the second dexio-
tropic (Fig. 14, ^", dp'\ dp'\ dp"'). The cell (Fig. 12, dp)
from which the four are formed is a cell of the tenth gener-
ation, and if all the preceding divisions had been spiral it
should divide laeotropically as it does. But in the ninth gener-
ation the spindle was inclined sinistrally, though approaching
horizontal (Fig. 11, dp, xl), while in the eighth and seventh
generations the spindles were both inclined dextrally.

Under this head the fact may also be mentioned that in
Arenicola some spindles, which are meridional or horizontal in
later stages of mitosis, are sometimes at first distinctly spiral.
One of the best examples of this kind is the first somatoblast
after its first two divisions. The third cell (Fig. 10) lies,
when formed, directly anterior to the large stem cell, and the
spindle in later stages is perfectly meridional, but often, even
up to the equatorial plate stage, it is distinctly dexiotropic, as
it normally would be in spiral cleavage.

Conklin has given a table of the reversals of cleavage in
Crepidida, of which there are a considerable number. Some of
these, however, are peculiar and possess especial interest with
respect to the relation of spiral and bilateral cleavage to each
other. These are described as follows : " In a few cases which
are classified here as reversals, the nuclear spindle does not
indicate that the cleavage is to be reversed, and even the
daughter-nuclei may occupy the same relative positions as in
the quadrants in which there is no reversal, while at the same
time the lobing of the cytoplasm and the subsequent rotation
shows that the cleavage is reversed ; the first division of ^d
(Figs. 25, 26, 29) is a case in point. In such cases the con-
ditions which influence the direction of the cleavage are not
manifested until after the nuclear division is completed,

250 BIOLOGICAL LECTURES.

whereas they are usually shown in the direction of the nuclear
spindles and in the earliest stages of cleavage." These obser-
vations, which, so far as I know, have not been duplicated in
the study of any other form, indicate, it seems to me, that the
reversal is of later date than the original spiral and has been
imposed upon the older form, but without removing all traces
of it as yet. I have sought very carefully for similar phenomena
in Arenicola, but without results except those mentioned above.
Reversals occur but rarely in Arenicola.

3. The spiral form of cleavage ceases first in regions which
are to play important rdles in the later development (Figs. 7,
10), while the regions which do not represent large areas in the
later stages continue to divide spirally for some generations.
In many of these cases it is impossible to say where the spiral
form of division has given place to a bilateral form, for there is
no sudden change, but the spiral direction simply becomes less
and less marked in succeeding divisions, until it is no longer
distinguishable as such. The cells of the third quartette of
ectomeres in Arenicola afford an excellent illustration of this
point. Conklin states that in Crepidtda the bilateral divisions
appear very slowly and that it is difficult to determine in some
cases whether the divisions are really bilateral or not. This, I
think, will be found to be the case in many forms. It is cer-
tainly true of Arenicola as well as of Crepidula. A few cells
contrast sharply with the rest of the ^g^ as regards the initia-
tion of bilateral cleavages, and these are the cells whose divi-
sions distribute the material for the embryo, including the
other cells themselves to a large extent.

4. The probability that independent modification of the
bilateral form of cleavage has appeared in the different groups
or species is amply supported by the fact that in many cases
cells of similar origin have different fates, and vice versa. It
is sufficient here to mention a few of the more striking cases
of this sort. Perhaps the most striking case of all is that of
the annelid and molluscan cross. In each group the cells of
the cross form the larger portion of the pre-trochal region,
though the exact relation of the various cells to the apical
plate is not sufficiently well known at present to afford a basis

THE SPIRAL TYPE OF CLEAVAGE.

251

Fig. 21

A renicola. Descendants of
cross cells.

for close comparison. The cross does not consist of the same
cells in the two groups. As Conklin has shown, the cross of
Crepidula arises mostly from the cells corresponding to the
so-called "intermediate girdle cells" of the polychaete, which
there form a relatively small portion of the pre-trochal region.
On the other hand, the cross
of the polychaete is composed
of the cells which Conklin calls
the " rosette series," and these
certainly do not form as large a
part of the pre-trochal region in
Crepidula as in the polychaete
{Figs. 21, 22).

Another good illustration of
the lack of correspondence be-
tween origin and fate is found
in a comparison of the proto-
troch formation in different forms, e.g., in Nereis, Arenicola,
Capitella, Sternaspis, Crepidula, and Ischnochiton. In all these
forms except Sternaspis, which has no prototroch, the details
of the prototroch formation are different, although the cells

which form the prototroch arise
from the same general region
of the ^g^, and, where the divi-
sions are spiral, of course by
similar divisions. In Sternaspis,
cells corresponding to the so-
called primary trochoblasts in
AmpJiitrite, etc., are formed,
but are simply a portion of the
ectoderm. Again, in Nereis
^ , ,, , , and Capitella the stomatoblasts

Fig. 22. — Crepidula. Upper pole at late stage, -^

showing space occupied by cells correspond- (•< OCSOphagoblastS ") are mCUl-
ing to annelid cross (after Conklin). , r - 1 i ...

bers of the second quartette,
while in Arenicola they belong to the third, and in Ischnochiton
to both second and third. The difference in position of larval
mesoblast, or its absence in some cases, indicates independent
modification. Numerous other examples of the same kind

252 BIOLOGICAL LECTURES.

might be cited, but these, I think, are amply sufficient for the
present purpose.

The facts given above appear to afford conclusive evidence
in support of the view that there has been an increasingly early
appearance of morphogenetic cleavage and considerable inde-
pendent modification in the phylogenetic history of the annelids
and mollusks and their ancestors. Nevertheless, it must be
admitted that there is a high degree of resemblance in the
cleavage even of widely separated forms. The tendency has
been to emphasize these similarities, even to the extent of
neglecting whatever differences might be found. Comparison
of details shows that many and considerable differences do
exist, and it shows further that the resemblances that occur
concern the more general features of the development. Two
facts seem to me to point the way to an explanation of these
similarities. First, there is in all cases the absolutely identical
(or reversed) spiral form of cleavage in the earlier stages to
serve as a basis for later modifications. And, second, the larval
form, when present, is similar in the different groups. Where
the larval form is not present, as in the oligochaetes and leeches,
the cleavage departs at an early stage from the plan adhered
to in the other groups. Whether the trochophore be regarded
as palingenetic or coenogenetic, its similarity of form in the
various groups must have been closely connected with the pre-
cocious appearance of resemblances in earlier stages of develop-
ment. Conklin remarks : *' The * reflection ' of similar larval or
adult characters would produce similar effects upon different
eggs, and consequently the similarity of the prelarval stages of
annelids and mollusks may be held to be due to the similarity of
their larvce ; but there is no reason for supposing that this par-
allel precocity has been independently acquired by annelids and
mollusks, since it may well have been produced before the phy-
logenetic separation of these groups." Conklin has not recog-
nized the influence of the strictly spiral form of cleavage of the
early stages in producing similar later stages, probably because
he regards the whole cleavage from the beginning as morpho-
genetic. I agree with him in regarding it as quite possible
that the ** reflection " or most of it occurred before the separa-

THE SPIRAL TYPE OF CLEAVAGE. 253

tion of the groups, and indeed I think it quite probable of
annelids and mollusks. The trochophore, reduced to its sim-
plest terms, doubtless arose from a larval form which occurred
in the life history of the common ancestor of the two groups,
and which I believe was probably not greatly different from the
polyclad larva. But the question as to whether the trochophore
is an ontogenetic recapitulation of an ancestral form which was
introduced into the ontogeny before the separation of the
groups must not be confused with that of its homology within
the groups. I do not desire to discuss this matter in detail at
this point, but simply to remark that even the high degree of
coenogenetic modification involved in the precocious appear-
ance of morphogenetic cleavage need not necessarily result in
a coenogenetic larval form. I am, however, inclined to regard
the original of the trochophore as distinctively a larval form and
not as a stage in the phylogeny of the adult forms of the groups
in which it belongs.

We have seen above that the spiral period is followed by a
distinctly morphogenetic form of cleavage, and that this mor-
phogenetic period has been encroaching upon the earlier spiral
period. We find, moreover, that the morphogenetic period
shows in large part a distinct bilateral symmetry, and we now
turn to a consideration of the significance of this symmetry,
i.e.^ the significance of the direction of cleavage during this
period. Notwithstanding some differences, a comparison of
various forms shows that the more important bilateral divisions
occur with a considerable degree of regularity and in the same
cell generation as a rule. Moreover, these are the first evidences
of bilaterality. In no case have distinctly bilateral divisions
been found at an earlier stage. Wilson regarded this fact as
one of the most striking features of the cleavage. In Crepidula
the change is not so abrupt, according to Conklin, though it
appears in almost exactly the same cell generation as in Nereis^
at least in certain parts of the &gg. Wilson believed the appear-
ance of bilaterality was due to the reduction of the " left pos-
terior macromere " to the same size as its fellows as the result
of the formation of the first somatoblast and the mesoblast.
Conklin has shown, however, that this rule does not hold good

2 54 BIOLOGICAL LECTURES,

in the case of Crepidula and many other gasteropods, since here
the " left posterior macromere is not appreciably larger than the
right, and in some {e.g., Umbrella, Heymons, '93) it is smaller,
and . . . the mesoblast {^d) is only one member of a quartette
which is separated in a left spiral from the macromeres, each
of the other members being quite as large, or even larger than
the cell 4</." And finally, after consideration of the manner in
which bilateral cleavages arise, he says : " The conclusion, there-
fore, is unmistakable that bilaterality first appears in processes
which lead to the formation of the trunk and the elongation of
the future animal, while the primitive radial symmetry of the
anterior quadrants is correlated with the fact that these quad-
rants give rise largely to larval organs, most of which bear
traces of radial symmetry." And again it must be said of this
conclusion, as Conklin said of Wilson's, that it is not applicable
here, or applicable only in modified form, for the anterior quad-
rants show a high degree of bilaterality in many forms.

I believe the facts as known thus far indicate that the func-
tion of bilateral cleavage is the symmetrical distribution of the
material for the formation of the adult or larva. It is evident
that in the adaptation of means to end symmetrical divisions
would be useless at a stage when the material was not in a posi-
tion to be distributed. Suppose the radial divisions forming the
cross (Fig. 7) appeared before the cross cells were surrounded
by a complete ring of cells. The formation of the cross would
not result in forcing the ectodermal cap down over the ^gg,
but would simply produce four lines of cells extending outward
from the anterior pole as far as or beyond the other ectoderm
cells. Again, suppose the first somatoblast 2d were to divide
bilaterally at its first division or its second : the growth of the
somatic plate in the direction which is to result in its concres-
cence would not be begun ; but, as it is, two cells are formed,
lying on either side of, and posterior to, the stem cell (Fig. 9),
and then a cell is given off anteriorly in the median line (Fig.
10), and thus the whole mass is forced somewhat posteriorly
and the symmetrical distribution of the material is begun.
Then in succeeding divisions of the stem cells (Fig. 11-17)
how exactly the relation between the parts is preserved, every

THE SPIRAL TYPE OF CLEAVAGE.

255

division balancing others and aiding in the accomplishment of
the ''desired" end, viz.^ the concrescence of the plate and the
formation of the " growing tip " ! Moreover, the posterior move-
ment of the descendants of the first somatoblast which is ini-
tiated by the third division — the first symmetrical division —
begins the covering over of the mesoblast. The mesoblast
divides bilaterally at its first division, for it is formed in such a
position that it must be covered over by the somatic plate and
thus forced into the interior of the ^gg, and its fate is to fur-
nish the double " Anlage " for the mesoderm bands. I think
these examples are sufficient to show how exactly the appear-
ance of the bilateral cleavages is timed for the accomplishment
of the purpose of development in the most perfect manner.
And, to my mind at least, they furnish ample explanation of
the fact that the bilateral cleavage does not appear in the
earlier stages.

It may be urged as an objection to this view that in various
forms {e.g., ascidians) the first cleavages are bilateral, and the
distribution of the material is accomplished equally well. This,
of course, is perfectly true, but there is no evidence that the
bilateral form of cleavage has displaced a spiral form in this
case. In the annelids and mollusks the spiral form of cleavage
is undoubtedly older than the bilateral form, and has been in
part displaced by it, the change occurring first in later stages,
and proceeding to the earlier. The point made here is simply
that there is no adequate cause for the displacement of the
spiral by the bilateral form of cleavage in stages earlier than
those at which the latter now appears, unless of course the
spiral period could be eliminated at one blow, as it were, and a
bilateral cleavage could begin at once. This may perhaps be
a more complete modification, but there are no indications dis-
cernible of its appearance, unless, indeed, the cleavage of such
forms as Teredo and Cyclas is to be regarded as such. The
spiral form of cleavage seems to serve perfectly in plotting out
the material of the ^gg in these forms.

In general, the direction of each cell division in morpJiogenetic
cleavage plays a perfectly defifiite role iii the accomplishmefit of
the 7norphoge?tesis of the species concerned. Bilaterally symmet-

256 BIOLOGICAL LECTURES.

rical cleavage is necessary for the symmetrical distribution of
the precociously segregated material^ so that it is to be expected
that the axes of symmetry in the cleaving egg should correspond
with the axes of symmetry in the adult. It is probable, however y
that the establishment of these axes is significant as an incident
connected with the mechanics of this form of cleavage y rather than
as indicating differentiation.

The time of division in the different blastomeres is also
a factor in the result attained. As mentioned above, the time
at which the spiral form of cleavage ends is closely connected
with the process of gastrulation, etc. All through the bilateral
period of cleavage the divisions seem to occur at the right time
to serve as a factor in morphogenesis. In the case of symmet-
rical divisions on the two sides of the ^^'g this fact is especially
noticeable, and the best illustration is found in the later divisions
of the somatic plate (Figs. 11-17). The symmetrical divisions
occur at the same time, or nearly, although they may be almost
on opposite sides of the ^g'g. Sometimes one side is slightly
ahead of the other, but in any case the variation is not great,
and the arrangement of the material is the same in every case.
In general, it may be stated that the larger cells divide more
rapidly than the smaller, but this rule does not hold good in

all cases, and does not apply of
course to yolk-laden cells. In-
deed, as Conklin ('97) and Lillie
('95) have suggested, the rapidity
of division is apparently regulated
in some cases by the time at which
the portion of material concerned
is to become functional, but
neither does this rule apply to all
cases. For instance, the primary
trochoblasts in Arenicola are rel-

FiG.23. — C/^;w«^//«. Small size ofcross cells -.I <,^^11p^ \\\^r\ in Al^'hln-
and large size of primary trochoblasts (after ^"^IVeiy SmaUCr inaU m JimpUl-

'^^^^)- trite and Clymenella {cf. Figs. 7

and 23), and become ciliated at a much later stage. Yet in
the two forms the trochoblasts pass through all of their divi-
sions with equal rapidity, i.e., relatively to the other blasto-

THE SPIRAL TYPE OF CLEAVAGE.

257

meres. The early division of these cells in Arenicoluy where
the cells do not become ciliated for a long time, must, I think,
be considered as without significance for the prototroch itself,
but as directly connected with the overgrowth of the entomeres
by the ectomeres. In Sternaspis the divisions of the cells
which correspond to the trochoblasts occur at almost exactly
the same time relatively to
the other cells, but here there
is no prototroch, and these
cells apparently form merely
a part of the ectoderm.

The mesoblast in Arenicola^
though larger than most other
cells in the ^gg^ divides very
slowly. This is not wholly
because of the fact that it
becomes differentiated only
at a late stage, for the ante-
rior ends of the mesoblast
bands form muscles in the
trochophore. But there is no room for the products of
its division in the blastocoele in early stages (Fig. 24). If it
divided rapidly, invagination of the entomeres would be impos-
sible, unless the ectoderm were stretched to a considerable
degree. In cases where the mesoblast is smaller, the rapidity
of division may be more closely connected with the time at
which it becomes functional as mesoblast.

The size of the various cells is another factor. It is appar-
ently correlated with several features of the process of morpho-
genesis, vis.^ the final fate of the cells in question, the stage at
which the material is used, and the relation to other cells in
the complex. Lillie and Conklin have called attention to the
first two of these factors, but the third seems fully as important
as either of the other two. The size of the various cells is
an essential factor in the accomplishment of the processes of
growth. Examination of the growth of the somatic plate gives
the impression that not only is the direction of division
adapted to the form of growth, but the size of each cell also.

Fig. 24. — Arenicola. Mesoblasts MM nearly fill-
ing the blastocoele . Optical section.

258 BIOLOGICAL LECTURES.

A bit of material is given off at one point, serving to fill up a
chink, as it were; a large cell appears at another point, etc., and
every cell in the complex aids in the posterolateral growth
and concrescence of the plate, not simply because it was
formed in a certain position, but because it is of a certain
size. Some cells, like the turret cells of Crepidula, are small
when formed, but later increase greatly in size. In these cases
it seems probable that the appearance of a large cell at the
time when these divisions occur would interfere with the dis-
tribution of the ectodermal material. The later increase in
size may be connected with the role which the cell itself is

to play, or may be another
means of aiding in the distribu-
tion of the material.

But in many cases a great
difference in size is found among
the cells at the very beginning
of cleavage, as evidenced by
the extremely unequal division
in many forms (Fig. 25). This
difference continues to be ex-
pressed at various points during
Tig. 2s:— Arenicoia. Unequal two-cell thc Spiral pcriod, notably in the
stage rom upper po e. formation of thc first souiato-

blast, 2d (Fig. 5), and the mesoblast, ^d, and, in a less degree,
in many other cells. These are the differences which have
especially attracted the attention of nearly all those who have
studied this form of cleavage. They have been quite generally
regarded as differential in significance. There is here undoubt-
edly an anticipation of, and provision for, the later stages of
cleavage, but, to me at least, it seems that, so far as morpho-
genesis is concerned, it is, at least usually, rather quantitative
than qualitative. Certain amounts, rather than certain kinds,
of material are stored up in certain cells just where they will
be in position to produce by coordinated action the ''desired
result." Thus a still further ''condensation" of the process
of development is effected. Cells which are to serve as centers
of distribution of material, e.g.y the first somatoblast, may exceed

THE SPIRAL TYPE OF CLEAVAGE. 259

all other cells of the ^gg in size, as is the case in Arenicola.
It seems at least unnecessary to suppose, however, that this
cell contains an entirely different ''organ-forming substance"
from that contained in, for instance, the four cells, products of
the dorsal intermediate girdle cell of the first quartette, which
pass through the dorsal gap in the prototroch, and lie just in
front of the somatic plate (Fig. 13), or from that contained
in the cells of the second and third quartettes which form the
ectodermal region about the mouth. Indeed, Dr. Treadwell
tells me that in Podarke a large portion of the dorsal ectoderm
of the trochophore is formed by the descendants of the dorsal
first quartette cell, which have passed through the gap in the
prototroch.

Of course, difference in size between yolk-bearing and non-
yolk-bearing cells is an expression of difference in constitution.

Notwithstanding the fact that the spiral form of cleavage
remains throughout the earlier stages of development, a form
of modification does occur which in a way anticipates the morpho-
genetic character of the later cleavage. The determinate char-
acter of the morphogenetic period establishes the fate of each
cell under normal conditions ; and, moreover, it renders possible
a further degree of modification in the cleavage, viz.^ '' preco-
cious segregation." As the result of precocious segregation,
the blastomeres not only have a definite fate, but each contains
a definite amount of material, according to the part which it is
to play in the cleavage and irrespective of the yolk which it
may contain. This modification extends back to the earliest
stages of cleavage, but without altering the spiral form, not-
withstanding the great differences in pressure that may exist
between the blastomeres. It has become the custom to speak
of a protoblast as containing from its earliest appearance the
substance for the organs which are to be formed by its descend-
ants, and the term ''differential cleavage" has been employed
to denote the divisions which give rise to the protoblasts. Pre-
cocious segregation is undoubtedly connected with differentia-
tion in so far as the perfectly exact distribution of the material
renders an earlier differentiation possible ; but, as regards the
separation of various organ-forming substances by the planes

26o BIOLOGICAL LECTURES.

of cleavage, it is very difficult for me, at least, to believe that
it occurs, especially as I cannot see that there is evidence to
warrant the conclusion. There are, of course, cases where cer-
tain substances have been found to pass always into definite
cells, e.g., the granules of the " Urvelarzellen " of Neritina
(Blochmann, '8i) and certain granules in AsplancJma (Jennings,
'96). Evidence is lacking, however, to show that these sub-
stances are necessary to the formation of the organs to which
the cells give rise. Conklin ('99, p. i8) admits that there is
abundant evidence to prove the absence of any necessary rela-
tion between cell formation and differentiation, but believes
that in certain cases the two processes are related. He distin-
guishes three categories of relation : (i) ** Cell formation follow-
ing the lines of preceding differentiation, e.g., certain cleavages
of ctenophores, mollusks, and ascidians ; or (2) cell formation
and concomitant differentiation, e.g., many cleavages of turbel-
larians, nematodes, annelids, and mollusks; or (3) differentia-
tion following the lines of preceding cell formation, eg., many
cleavages in the eggs of annelids, mollusks, and probably many
other animals." The significance of these so-called relations
becomes apparent when we observe that every differentiation
which occurs in multicellular material must be related in one
of these three ways to cell division. In other words, cell divi-
sion and differentiation are just as independent of each other
here as elsewhere, but the process of differentiation is occur-
ring to a certain extent during a series of very conspicuous cell
divisions, i.e., the earlier ones. A comparison of the process
of formation of a single organ, the prototroch, in different ani-
mals seems to me to show how absolutely distinct cell formation
and differentiation are. The prototroch may consist of eight
or of sixteen cells, of twenty-five, of thirty, or of many ; the
series of divisions leading to its formation are almost always
different in different forms, and where they are not they are
similar simply because they are purely spiral cleavages ; it may
become functional almost immediately after the divisions are
completed, — indeed in Capitella (Eisig, '98) divisions continue
after ciliation, — or it may remain a long time before becoming
ciliated ; and, finally, cells perfectly equivalent to the "primary

THE SPIRAL TYPE OF CLEAVAGE. 26 1

trochoblasts " in origin may be formed without ever forming a
prototroch.

It is hardly necessary to remark that the differentiation of
an ^^^ into protoplasmic and deutoplasmic portions, which so
often accompanies cleavage, undoubtedly has, in most cases at
least, a comparatively simple explanation.

Differentiation is then^ I believe, ^* a function of position'' {cf.
Driesch). Whether the material is contained within one cell or
many is a mere incident so far as the result is concerned. The
comparative study of the spiral type of cleavage leads irresistibly
to this conclusion.

The appearance of definite protoblasts in cleavage does not
necessarily imply that they contain a specific material which is
necessary for the formation of the organ in question. Proto-
blasts are, I think, to be regarded, in general, as centers of dis-
tribution of the material of the ^^%\ and their formation is
probably due to a condensation in the process of development,
or a saving of energy, as Conklin suggested. As I understand
Conklin's position, however, he regards the saving of energy as
occurring in processes of differentiation merely, for, as he
remarks, *' it is possible to see that in an organ which reaches
functional activity after a dozen divisions less energy has been
expended than in one which reaches this stage only after one
hundred divisions." This is undoubtedly true, but how does it
explain the formation of the whole trunk ectoderm from the
first somatoblast, which occurs in some annelids, or the forma-
tion of the mesoblasts, for the descendants of these cells do
not, for the most part, become ''functional" in the definitive
sense for a long time, and then chiefly with reference to the
adult body and not the larval.-* There is surely no reason for
believing that the number of cells in the organs of the adult is
less in those forms in which precocious segregation has occurred
than in others. Lillie ('95) has pointed out the fact that "almost
every detail of the cleavage of the ovum of Unio can be shown
to possess some differential significance." It is undoubtedly
significant as regards the distribution of the ^gg material, but
that is not necessarily differentiation. The material separated
as the result of precocious segregation may, I believe, be per-

262 BIOLOGICAL LECTURES.

fectly indifferent material, except as regards position, and the
true significance of the segregation may lie in the fact that a
certain amount of material is present, which is to be distributed
through the means of a series of perfectly definite cell divisions.
By ''distribution " is meant the arrangement of the material in
such a manner that the various morphogenetic processes are
directly accomplished thereby, e.g.^ formation of the somatic
plate, gastrulation, which in Arenicola and some related forms
is probably accomplished very largely by the growth of the
somatic plate, formation of the mesoderm bands, beginning
elongation of the larva, etc. The important point in this view
lies, I believe, in the conclusion that these results are accom-
plished, at least very largely, directly by the energy of cell divi-
sion itself. Given a certain degree of cohesive power in the
protoplasmic material, a certain degree of surface tension, and
a certain definite series of cell divisions and the result must
always be the same, and the same cells will always occupy the
same relative positions at corresponding stages and will form
the same parts of the whole, i.e.^ will have the same fate under
normal conditions. This would be true even if the segregated
material should remain totipotent long after its segregation.

Since the fate of the various cells is definitely established
and can be foretold, it is possible that differentiation might
begin at an earlier stage than otherwise, but there is no reason
to suppose that the process of cell division is connected with it
even here. To take an example : why should the two lateral
cells of the paratroch on each side in Arenicola pass through
one more division than the dorsal cells before they form the
paratroch (Fig. 1 5) } If cell division is connected in these
eggs with differentiation, it must be either that one of these
divisions is "non-differential," according to Conklin's distinc-
tion, or else it is necessary to assume that the lateral cells
are more highly differentiated than the dorsal cells. As
regards the first assumption, there seem.s to be no definite cri-
terion of differential divisions, so that no decision is possible.
As regards the second, there is no reason to believe that it is
true. Taking up the matter from a different point of view, viz.,
the relation of the constituent parts with regard to the result

THE SPIRAL TYPE OF CLEAVAGE.

26

to be accomplished, an explanation appears possible. This divi-
sion is a necessary step in the concrescence of the somatic plate.
It does not bring *' paratrochal material" into the proper posi-
tion or into the proper degree of differentiation, but it does aid
in accomplishing the end toward which each of the symmetrical
divisions in the somatic plate contributes its part. In another
case, e.g., AmpJiitrite, where some of the preceding divisions
are different and the material is
somewhat differently grouped,
no such division occurs, and
the paratroch consists of four
cells instead of six.

The extreme precocious seg-
regation, which extends back
even to the first cleavage and
in some forms {Aretticold) is
indicated even before cleavage
by the shape of the ^^g (Figs.
26, 2^), is rendered possible by
the fact that the whole cleav-
age is strictly determinate in
character. A determinate form
of cleavage is, as I believe,
the product of a considera-
ble degree of modification in
development in the direction
of condensation. Thus the ap-
pearance of precocious segre-
gation in the earliest stages

Fig. 26. — Arenicola. Egg before division,
seen from upper pole.

Fig. 27. — Arenicola. Egg before division,
seen from side.

must indicate a still further

departure from the original primitive forms of cleavage.

Precocious segregation reaches its highest expression at
present in the oligochaetes and the leeches, especially in the
latter, where the development is very largely teloblastic, the
material for almost the whole body being segregated into a few
cells which form the germ bands. Here is not only segrega-
tion, but regulated cell division in the highest degree.

Teloblastic development must undoubtedly be regarded as
the last term in a long series of modifications, all leading to a

264 BIOLOGICAL LECTURES.

shortening or condensation of the process of development,
although this, as Conklin has remarked, does not necessarily
signify a shortening of the time of development. I have at-
tempted in the preceding pages to follow the various steps of
the process, as I believe they must have occurred historically.

It appears probable that a large amount of yolk affords a
condition favorable to the teloblastic form of development, for
it renders possible the very early localization of the embryo
upon certain portions of the ^gg. In the annelids and mollusks
the amount of yolk in the ^gg is not sufficient to render any
such localization of material necessary in the interests of con-
densation of development, and a partial localization of material
at various points is the most favorable arrangement for rapid
development.

In any form of cleavage which is morphogenetic, it is incor-
rect to regard certain cells as becoming functional at certain
times. Each cell is functional at every stage of the cleavage
and directly connected with morphogenesis. Since this is the
case, it is necessary in seeking for the explanation of its vari-
ous qualities, such as structure, size, direction, and time of
division, etc., to consider, not simply the part which it is to
play in the adult organism or in the larva, but the part which
it plays at the moment under consideration and at each moment
before and after it. Each stage of development is the result
of the preceding stage and in some manner the cause — not of
the adult or larval organization — but simply of the next suc-
ceeding stage. Each stage is a whole, not a series of inde-
pendent parts, and, moreover, modifications may occur in any
stage which are not connected — except, of course, indirectly,
since development is continuous — with the final organization,
but relate only to some detail of some particular stage. Thus
we may expect to find numerous differences in detail in the
solution of the same general problem, i.e., the production of
the same general organization.

I believe that, without an adequate appreciation of the
extreme plasticity of the cell, a true conception of organiza-
tion is impossible. The cell may be likened to a bit of plastic
material, which may be modeled according to the needs of the

THE SPIRAL TYPE OF CLEAVAGE. 265

organism, and which, even when once modeled, may be
brought back, like a bit of clay, to the original condition and
remodeled into a different form if necessary. And further,
the material connection between cell and cell, which occurs so
widely, is of the utmost importance, for through this, as I
believe, the cell becomes simply a part of the organism, in
touch with the whole, influenced by the whole, and exercising
in its turn an influence upon the whole. Whether that pecul-
iar property cf protoplasm which we know as organization be
merely the expression of its chemical and physical structure,
or, as Driesch believes ('99), the expression of a mysterious
force wholly different from any in the inorganic world, it is the
organism — the individual^ which is the unit and not the cell.

University of Chicago.

LIST OF LITERATURE CITED.

'97a Andrews, A. E. Some Activities of Polar Bodies. Johns Hopkins

Univ. Circ. Vol. xvii, No. 123.
'97b Andrews, A. E. Spinning in Serpula Eggs. Ainer.Nat. September.
'98a Andrews, A. E. Activities of Polar Bodies of Cerebratulus. Arch.

f. Entwicklungsmechanik d. Organismen. Bd. vi, Heft 2.
'98b Andrews, A. E. Filose Activities in Metazoan Eggs. Zo'dl. Bull.

Vol. ii, No. I.
'98c Andrews, A. E. Some Ectosarcal Phenomena in the Eggs of Hydra.

Johns Hopkins Univ. Circ. Vol. xviii, No. 137.
'97 Andrews, G. F. Some Spinning Activities of Protoplasm in Starfish

and Sea-Urchin Eggs. Journ. of Morph. Vol. xii, No. 2.
'81 Blochmann, F. Ueber die Entwicklung von Neritina fluviatilis.

Zeitschr.f. wiss. Zool. Bd. xxxvi.
'83 Blochmann, F. Beitrage zur Kenntniss der Entwicklung der Gas-

teropoden. Zeitschr. f. wiss. Zool. Bd. xxxviii.
'97 Child, CM. A Preliminary Account of the Cleavage of Arenicola

cristata, etc. Zool. Bull. Vol. i, No. 2.
'97 CONKLIN, E. G. The Embryology of Crepidula. Journ. of Morph.

Vol. xiii, No. I.
'99 CoNKLiN, E. G. Cleavage and Differentiation. Biol. Led., delivered

at Woods Holl, Sessions of 1897 and 1898.

i

266 BIOLOGICAL LECTURES.

'94 Crampton, H. E. Reversal of Cleavage in a Sinistral Gasteropod.

Ann. New York Acad. Set. Vol. viii.
'96 Crampton, H. E. Experimental Studies of Gasteropod Development.

Arch. f. Ent'wicklungs7nechanik d. Organismen. Bd. Hi, Heft i.
'99a Driesch, H. Die Methode der Morphologie. Biol. Centralbl.

Bd. xix, Nr. 2.
'99b Driesch, H. Die Lokalisation morphogenischer Vorgange. Arch.

f. Efitwicklungsmechanik d. Organismen. Bd. viii, Heft i.
'98 EisiG, H. Zur Entwicklungsgeschichte der Capitelliden. Mittheil.

a. d. zool. Stat. Neapel. Bd. xiii. Heft i and 2.
'96 Hammar, J. A. Ueber einen primaren Zusammenhang zwischen den

Furchungszellen des Seeigeleies. Arch.f. mikr. Anat. Bd. xlvii.
'99 Heath, H. The Development of Ischnochiton. Zool. Jahrb.., Abth.

f. Anat. u. Ont. Bd. xii.
'93 Heymons, R. Zur Entwicklungsgeschichte von Umbrella mediter-

ranea. Zeitschr. f. wiss. Zool. Bd. Ivi.
'97 Holmes, S. J. Preliminary Account of the Cell Lineage of Planorbis.

Zool. Bull. Vol. i, No. 2.
'96 Jennings, H. S. The Early Development of Asplanchna Herrickii

De Guerne. Bull. Mus. Comp. Zool., Harvard College. Vol. xxx.

No. I.
'95 KoFOiD, C. A. On the Early Development of Limax. Bull. Mus.

Comp. Zool., Harvard College. Vol. xxvii. No. 2.
'95 LiLLiE, F. R. The Embryology of the Unionidae. Journ. of Morph.

Vol. x.
'95 McMurrich, J. P. Cell-Division and Development. Biol. Led.,

delivered at Woods Holl, Session of 1894.
'97 Mead, A. D. The Early Development of Marine Annelids. Journ.

of Morph. Vol. xiii, No. 2.
'96 Meyer, A. Die Plasmaverbindungen,*etc. Bot. Zeitschr. Bd. liv.
'79 Rabl, C. Ueber die Entwicklung der Tellerschnecke. Morph. Jahrb.

Bd. V.
'82 Sachs, J. Vorlesungen iiber Pflanzenphysiologie. Leipzig.
'93 Whitman, C. O. The Inadequacy of the Cell Theory of Develop-
ment. Journ. of Morph. Vol. viii. No. 3.
"92 Wilson, E. B. The Cell Lineage of Nereis. Journ. of Morph.

Vol. vi.
'93 Wilson, E. B. Amphioxus and the Mosaic Theory of Development.

Journ. of Morph. Vol. viii.
'97 Wilson, E. B. The Cell in Development and Inheritance. Columbia

Univ. Biol. Ser.
'93 ZiMMERMANN, A. Sammel-Referate aus dem Gesammtgebiete der

Zellenlehre. Beihefte z. Bot. Centralbl. Bd. iii.

FIFTEENTH LECTURE.

THE AIMS OF THE QUANTITATIVE STUDY OF

VARIATION.

C. B. DAVENPORT.

There are two fields of investigation possessed of well-
defined methods whose rise and development have character-
ized the advance of zoological morphology in the present
decade. The first of these is experimental morphology, or (to
use a more restricted term) experimental embryology. Experi-
mentation is not new, even in morphology, but there has been
a decided advance upon the methods in vogue a century ago.
The second new field of investigation is the precise, quantita-
tive study of variation. The method is here also in part old,
having long been used in anthropology ; but in its application
to zoology in the strict sense it is quite new. I wish to point
out what are the aims of this study.

The aim of quantitative variation study may be stated gener-
ally to be the investigation of the specific ^ form, including the
laws of its development and maintenance in the individual, the
laws of intermingling of specific forms in the progeny of
parents having dissimilar specific forms ; and the laws of evo-
lution of new specific forms in nature; i.e., the method of
origin of species. I will treat of these three subdivisions of
the field of quantitative investigation in inverse order.

In studying specific differentiation in nature we shall find it,
first of all, necessary to investigate species as we find them in

1 " Specific " is here used in a broad sense to cover the form-peculiarities of
families, communities, races, species, genera, in short, peculiarities constant to a
group of related individuals of whatever extent.

267

268 BIOLOGICAL LECTURES.

nature ; not indeed the species of the systematist, but rather
those homogeneous lots of individuals taken from one locality
to which Duncker has given the convenient name '' form-units."
The quantitative method of studying the form-unit consists in
counting certain numerically repeated organs, or of measuring
one or more organs, on each of many individuals. The quan-
tities resulting from these measurements are then ** seriated"
or arranged in classes according to size. More quantities will
fall in certain of these classes — usually the middle classes —
than in the others. The distribution of "frequencies of occur-
rence" in the several classes will give those characters of the
form-unit for which we are seeking — its rigidity or plasticity,
its stability or instability, its independence of or subjection to
selection, and its unity or its tendency to break up into several
forma. From a comparison of measurements of two or more
organs of the same individuals we may learn additional facts
concerning the form-unit ; such as the degree of correlation
existing between its different parts, from which we may infer
directly the morphogenetic kinship of the different organs,
and, more indirectly, get light on homologies. The quanti-
tative study of the form-unit thus bears on many biological
topics.

The possibility of getting such biological data from statistics
of organisms depends upon the fact that biological statistics —
more particularly the distributions of "frequencies" — follow
definite laws capable of mathematical expression and analysis.
The mathematical analysis necessary to deduce the biological
results is not very difficult, but is quite within the range of
any one who has studied trigonometry and logarithms.^

The objection has been raised that the quantitative method
of investigating the form-unit is limited to organs which can
be counted or measured with a millimeter scale. It is, how-
ever, applicable to the study of all quantities which can be

1 For an account of the methods of dealing with biological statistics, see a
paper by Dr. Georg Duncker, on " Variations-Statistik," in Roux's Arch. f. Ent-
wickelungsmechanik, 1899, sold separately by W. Engelmann, Leipzig; and my
little book, Statistical Methods with Special Reference to Biological Variation,
published by John Wiley & Sons, New York, which contains also all the tables
necessary for the mathematical analysis of frequency distribution.

THE QUANTITATIVE STUDY OF VARIATION. 269

measured ; practically to all qualities of form and color which
an organism possesses. The thorough application of the quan-
titative method often takes time, practice, and ingenuity, but
the results are worth the cost.

As stated above, the distribution of frequencies in the several
classes gives a measure of the. stability or variability of an
organ. Let us consider what this single result enables us
to do. It enables us to express the laws of normal variation
quantitatively instead of loosely, as heretofore. Thus we can
now compare the relative variability of the sexes ; we can
determine whether wide-ranging species are especially varia-
ble ; whether specific characters are more variable than generic
ones ; whether parts developed to an extraordinary degree are
especially variable ; whether in serially repeated organs those
standing at the ends of the series vary most ; and whether
secondary sexual characters are more unstable than other
somatic ones. All these and other ''laws" can be established
only by quantitative methods.

Again, in systematic zoology this method will be of great
use. I know that systematic zoology is much decried in some
schools, and unquestionably it has suffered from the adherence
of some unphilosophic workers. On the other hand, many of
the workers in this department of zoology have been brought
face to face with the great problem of the origin of species,
have seen their opportunity, and have contributed valuable data.
If they have failed to do more, it is because the methods they
have used — the methods of adjective descriptions — have been
wholly inadequate to the task. What headway can one make
on the classification of the American Unios by saying that the
individuals from one river differ from those of another by being
more elongated, or rounder, or browner, or by having a lower
beak, or sharper radiating lines. Not until these qualities and
others are reduced to numbers can we hope to disentangle
this maze of varying communities. After we shall have a
number of studies made in this fashion upon wide-ranging
species or genera, we shall have gained some adequate notion
of the relation of the origin of species to geographic distri-
bution.

270

BIOLOGICAL LECTURES.

It has been urged by some naturalists that the quantitative
study of animals will take away that peculiar charm which
attaches to biological study. This peculiar charm, as I under-
stand it, is due to the contact with out-of-doors, into which the
biologist is brought, and to the fact that the phenomena which
he studies of growth, reproduction, variation, inheritance, and
differentiation of form and activities, of structure and habits,
form an important part of human experience. Now the stu-
dent of variation also is concerned with these same phenomena,
only he uses a different method in his study of them. More-
over, the necessity of collecting large numbers of individuals
for study will be especially apt to take the student to the field.
After all, it is the search for truth, particularly a truth which
seems worth while, which inspires us, and the quantitative
method gives us a new tool in this search.

There seems to be an impression that all this quantitative
work is very difficult. But this is not at all the case. Many
important pieces of work require only abundant material and
the ability to count and record. Thus, by counting the ray
flowers of the field daisy, Ludwig discovered that in the distri-
bution of their frequencies they exhibit several " modes " occur-
ring in a mathematical series — the series of Fibonicci. By
counting the grooves in scallop shells from two Long Island
localities, I have found a significant difference in the variability
of the upper and lower valves, and also a difference in the aver-
age number of grooves from the two localities. With the aid of
a hand lens only, a series of studies has been made on the vari-
ability of the rostral teeth of prawns from different countries and
from the same country, but subjected to various conditions. Of
course, with the aid of a millimeter scale, dividers, planimeter,
map measurer, color top, balances, and other apparatus, one
can go farther. As to the number of individuals required, at
least from two hundred to five hundred should be employed
when possible.

Not only the qualities of the form-unit, but the causes which
have determined those qualities are to be investigated quanti-
tatively. Among these, selection, the direct action of environ-
ment, and internal tendencies to change are reckoned the most

THE QUANTITATIVE STUDY OF VARIATION. 27 1

important. Now how can we know anything definite concern-
ing the causes of specific change unless we can express exactly
the present condition of a form-unit and express exactly again
its condition after the working of a factor ? Descriptions based
on adjectives are inadequate for a precise expression either of
the present condition of a character or of a slight change in
it. The beautiful results on the change in form of the crabs
at Plymouth, which Weldon obtained by the quantitative
method, could hardly have been obtained without that method.
It has long been a reproach to biology that it could give so
few demonstrations of race differentiation going on in nature.
It has, indeed, been pointed out that we should not expect in
man's brief lifetime, nor even within historic times, to find such
changes ; but this defense is valid only if we rely upon adjec-
tives for specific descriptions. If specific descriptions are
given quantitatively, on the other hand, we can hope soon to
prove that many species are undergoing change and to deter-
mine the cause of the change.

Again, correlation is an important biological conception ; it
is especially important for the determination of homologies,
for homologous organs exhibit, as is well known, a tendency to
vary correlatively, e.g.^ the right and left sides of the body ; the
series of teeth or ribs. Accordingly within certain limits one
can infer homology from correlation. A precise measure of cor-
relation consequently will serve as a measure of morphological
kinship. Such a precise measure is given by ''Galton's
function."

Intimately associated with the study of correlation between
organs of the body is the quantitative study of heredity, or the
correlation between parent and offspring or between other
related individuals. In this field there are numerous matters
demanding thorough investigations. There is the confirmation
of Galton's law of Ancestral Inheritance, by which all the par-
ents of any (;?th) degree of ancestry (the parents being of first
degree, grandparents of second, etc.) contribute together —^
of the total heritage. There is the matter of prepotency in
breeding — prepotency of one race or one sex. Prepotency is
the preponderating transmission from one parent. The very

272 BIOLOGICAL LECTURES.

definition implies a quantitative difference in the power of
transmission. A great difference may indeed be detected and
expressed without resorting to measurements ; but it will often
happen that there is true prepotency even if slight, and such
prepotency must go unobserved without the use of the quanti-
tative method. Then there is the problem, so important for
the question of species, of the relative potency of sports and
the normal form when mated. Even a slight prepotency, too
small to be detected qualitatively, may suffice to prevent the
sport from being swamped by crossing with the normal form.
Again, there is the question of telegony. Whether the first
sire influences subsequent offspring can be afBrmed or denied
only by comparing the correlation between offspring and pre-
viously cross-mated mother with the correlation between off-
spring and previously pure-mated mother. If there is less
correlation in the first case than in the second, telegony exists.
The facts of xenia are to be investigated in the same way.
Still further, the phenomena of blending of parental characters
in hybridization, or of their alternative heritage, or their patch-
work intermingling — all these phenomena demand careful
quantitative study.

Finally, even the development of the specific form in the
individual demands quantitative investigation. Many investi-
gators have already studied embryonic variation, but few quan-
titatively. The precise comparison of variability in the embryo
and the adult will be of importance on account of its bearing
on theories of selection and self-adaptation. The most impor-
tant results in this field are, however, clearly gained by the
quantitative study of variation in embryos subjected to varying
•environmental conditions. The exact comparison of change of
environment with change in form will lead to a still more pre-
cise knowledge of the effect of external agents. So likewise
the quantitative studies made on the result of experiments to
produce race change will be the most striking. In fact, in the
application of combined experimental and statistical methods
to genetic problems, zoology will reach its highest development.

SIXTEENTH LECTURE.

ON THE NATURE OF THE PROCESS OF
FERTILIZATION.

JACQUES LOEB.
I.

Leeuwenhook demonstrated in 1677 the existence of sper
matozoa. It was about one hundred and sixty years before
biologists convinced themselves that these spermatozoa were
no parasites. In 1835 K. E. von Baer was still of the opinion
that the spermatozoa had nothing to do with the process of fer-
tilization. The parasitic conception of spermatozoa was finally
done away with by Wagner's demonstration that only those
animals are capable of fertilizing eggs whose sperm contains
spermatozoa. Very soon afterwards histologists showed the
origin of spermatozoa from cells.

Leeuwenhook was of the opinion that the spermatozoon repre-
sents the future embryo. On the other hand, it was not diffi-
cult to notice that the embryo in fishes, amphibians, and birds
develops from an egg furnished by the female. The question
arose as to what was the homologous element in the female of
mammals. In 1672 De Graaf discovered the follicle in the
ovary of mammals, and in 1827 von Baer discovered the mam-
malian egg cell.

The next problem that was solved in this branch of science
was the relation of the sperm to the egg. Many scientists,
among them De Graaf, had assumed that no direct contact
between egg and sperm was necessary, that something volatile
in the sperm, the aura seminalis, was sufficient for the act of
fertilization. Contrary to this, Jacobi showed (1764) that fish

273

2 74 BIOLOGICAL LECTURES.

eggs can only be fertilized by bringing the sperm into direct
contact with the eggs, and Spallanzani showed the same for
the frog. He succeeded in producing artificial fertilization of
mammals by introducing sperm into the vagina. But even
Spallanzani did not realize that the spermatozoa were the
essential element in the sperm. In 1824 Prevost and Dumas
proved this by filtering the sperm, and demonstrated that the
sperm whose spermatozoa had been retained by the filter lost
its power of impregnating the ^gg. These observations estab-
lished the fact that the spermatozoon has to come in contact
with the Qgg in order to bring about fertilization.

The next step was the observation made by Barry in 1843
that the spermatozoon actually enters into the ^gg. This
observation was confirmed ten years later by a number of
authors, Meissner, Newport, Bischof, etc. It is rather remark-
able that it was one hundred and sixty years after the discovery
of the spermatozoon and the follicle before the fact was recog-
nized that the spermatozoon has to enter the ^gg in order
to bring about fertilization. Had the biologists during these
one hundred and sixty years lost their interest in the investiga-
tion of this problem } This was certainly not the case, but they
spent their energy not in fruitful research, but in speculations
and controversies which were admired by their contemporaries
and made their authors famous, but which were a mere waste of
time. History has since repeated itself in other fields of biology.
The outcome of the facts gathered concerning the process of fer-
tilization was four apparently different theories of fertilization,
which, however, have much in common.

The first theory of fertilization is a morphological one.
According to this theory, it is the morphological structure
of the spermatozoon which is responsible for the process of
fertilization.

The second theory is a chemical one. According to this
theory it is not a definite morphological or structural element
of the spermatozoon, but a chemical constituent, that causes the
development of the ^gg. Against this second view Miescher
has raised the objection that his investigations showed the
same compounds in the ^gg and the spermatozoa. I do not

THE PROCESS OF FERTILIZATION. 275

think that this objection is valid. We know that simple varia-
tions in the configuration of a molecule have an enormous
effect upon life phenomena. This is shown among others by
the work of Emil Fischer on the relation between the molecu-
lar configuration of sugars and their fermentability. When
Miescher made his experiments he was not familiar with such
possibilities. Moreover, Miescher was not able to state whether
the spermatozoa contain enzymes or not.

A third theory was a physical theory (Bischof ). This theory
assumes that a peculiar condition of motion exists in the sper-
matozoon which is transmitted to the ^g^^ and causes its devel-
opment. It should be said, however, that this idea is not so
very different from the chemical conception, because it assumes
exactly the same for the spermatozoon that Liebig assumes for
the enzymes. Liebig thought that the enzymes owed their
power of producing fermentation to the motions of certain
atoms or groups of atoms.

The fourth conception is the stimulus conception, which was
originated by His. According to this conception the Q,gg is
considered as a definite machine which if once wound up will
do its work in a certain direction. The spermatozoon is the
stimulus which causes the Qg'g to undergo its development. It
is to be said in connection with this stimulus conception that
the main point at issue is omitted, as to whether the stimulus
carried by the spermatozoon is of a physical or a chemical char-
acter, and in this way, of course, the stimulus conception is
nothing but a disguised repetition of the chemical or physical
theory of fertilization.

All these theories are so vague that we do not need to be
surprised that none of them has led to any further discovery.
If we want to make new discoveries in biology, we must start
from definite facts and observations, and not from vague spec-
ulations. Among these observations the most important are
those on parthenogenesis. It had been observed for a long
time that the unfertilized ^gg of the silkworm can develop par-
thenogenetically. It was, moreover, known that plant lice can
give rise to new generations without fertilization. The most
impressive fact concerning the parthenogenesis of animals was

276 BIOLOGICAL LECTURES.

contributed by Dzierzon, who discovered that the unfertilized
eggs of bees develop and give rise to males, while the fertilized
eggs give rise to females. Similar conditions seem to exist in
wasps. It is, moreover, certain that a few crustaceans show
parthenogenesis.

A beginning of parthenogenetic' development had been
observed in the case of a great many marine animals which
develop outside of the female in sea water. It was found that
such eggs when left long enough in sea water may develop into
two or three cells, but no further. (2) On the other hand, in
ovaries of mammals now and then eggs were found that were
segmented into a small number of cells. These facts and the
occurrence of a certain class of tumors in the ovary, the so-called
teratomata, suggest the possibility of at least partial partheno-
genesis in the eggs of mammals. But all these phenomena
were considered to be of a pathological character. It must be,
however, admitted that we cannot utilize these facts with any
degree of certainty for the theory of fertilization, as in this
case certainty can only be obtained by the experiment. It
was not until very recently that such experiments were made.

II.

Eight years ago I observed that if the fertilized eggs of the
sea urchin were put into sea water whose concentration was
raised by the addition of some neutral salt they were not able
to segment, but that the same eggs, when put back after they
had been in such sea water for about two hours, broke up into
a large number of cells at once instead of dividing successively
into two, four, eight, sixteen cells, etc. Of course it is neces-
sary for this experiment that the right increase in the concen-
tration of the sea water be selected. (3) The explanation of
this fact is as follows : The concentrated sea water brings about
a change in the condition of the nucleus which permits a division
and a scattering of the chromosomes in the ^g'^. As soon as
the ^%'g is put back into normal sea water it breaks up into as
many cleavage cells at once as nuclei or distinct chromatin
masses had been preformed in the ^gg. Morgan tried the same

THE PROCESS OF FERTILIZATION. 277

experiment on the unfertilized eggs of the sea urchin, and found
that the unfertilized ^gg, if treated for several hours with con-
centrated sea water, was able to show the beginning of a seg-
mentation when put back into normal sea water. A small
number of eggs divided into two or four cells, and, in a few
cases, went as far as about sixty cells, but no larva ever devel-
oped from these eggs. (4) Morgan had used the same concen-
tration of sea water as Norman (5) and I had used in our
previous experiments. I had added about 2 grams of sodium-
chloride to 100 c.c. of sea water. Norman used instead of this
3/^ grams of MgCU to 100 c.c. of sea water, and Morgan used
the same concentration. Mead made an observation somewhat
similar to Morgan's upon Chaetopterus. He found that by
adding a very small amount of KCl to sea water he could force
the unfertilized eggs of Chaetopterus to throw out their polar
bodies. The substitution of a little NaCl for KCl did not have
the same effect. (6) While continuing my studies on the effects
of salts upon life phenomena, I was led to the fact that the
peculiar actions of -the protoplasm are influenced to a great
extent by the ions contained in the solutions which surround
the cells. As is well known, if we have a salt in solution, say
sodium-chloride, we have not only NaCl molecules in solu-
tion, but a certain number of NaCl molecules are split up into
Na ions (Na atoms charged with a certain quantity of posi-
tive electricity) and CI ions (CI atoms charged with the same
amount of negative electricity). When an egg is in sea water,
the various ions enter it in proportions determined by their
osmotic pressure and the permeability of the protoplasm. It
is probable that some of these ions are able to combine with
the proteids of the protoplasm. At any rate, the physical
qualities of the proteids of the protoplasm (their state of matter
and power of binding water) are determined by the relative
proportions of the various ions present in the protoplasm or in
combination with the proteids. (7) By changing the relative
proportions of these ions we change the physiological properties
of the protoplasm, and thus are able to impart properties to a
tissue which it does not possess ordinarily. I have found, for
instance, that by changing the amount of sodium and calcium

278 BIOLOGICAL LECTURES.

ions contained in the muscles of the skeleton we can make
them contract rhythmically like the heart. It is only necessary
to increase the number of sodium ions in the muscle or to reduce
the number of. calcium ions or both simultaneously.^ On the
basis of this and similar observations I thought that by chang-
ing the constitution of the sea water it might be possible to
cause the eggs not only to show a beginning of development
but to develop into living larvae, which were in every way
similar to those produced by the fertilized ^gg.

There seemed to be three ways in which this might be
accomplished. The first way was a simple change in the con-
stitution of the sea water without increasing its osmotic pres-
sure. The second way was to increase the osmotic pressure
of the sea water by adding a certain amount of a certain salt.
The third way was by combining both of these methods. The
first way did not lead to the result I desired. All the various
artificial solutions I prepared had only the one effect of causing
the unfertilized ^gg to divide into a few cells, but I was not
able to produce a blastula. I next tried the effects of an
increase in the sea water by adding a certain amount of mag-
nesium-chloride. In this case I had no better results than
Morgan. Very few eggs began to divide, but these did not
develop beyond the first stages of segmentation. I then tried
the combination of both methods. The osmotic pressure of
ordinary sea water is roughly estimated to be the same as that
of a I nNaCl solution or a 3J> nMgCb solution. I found, after
a number of experiments, that by putting the unfertilized eggs
of the sea urchin into a solution of 60 c.c. of y nMgCls solu-
tion and 40 c.c. of sea water for two hours the eggs began to
develop when put back into normal sea water. Such eggs
reached the blastula stage. I do not think that anybody has
ever seen before such blastulae as resulted from these unferti-
lized eggs. As these eggs had no membrane, the amoeboid
motions of the cleavage cells led very frequently to a discon-
nection of the various parts of one and the same Qgg^ and the
outlines of the ^^g became extremely irregular. The blastulae

1 It is due to the Ca ions of our blood that the muscles of our skeleton do not
beat rhythmically like our heart.

THE PROCESS OF FERTILIZATION. 279

showed, as a rule, the same outline as the ^g^ had in the
morula stage. It was, moreover, a rare thing that the whole
mass of the ^^g developed into one blastula. The disconnec-
tion of the various cleavage cells led, as a rule, to the formation
of more than one embryo from one ^gg. The results were in
a certain way similar to those I had obtained when I caused the
fertilized eggs of sea urchins to burst. In such cases a part
of the protoplasm flowed out from the ^gg but was able to
develop. These extraovates had no membrane, and of course
showed some irregularity in their outlines, but the irregularity
in this case was far less than that observed in the unferti-
lized eggs of my recent experiments. But although I had thus
far satisfied my desire to see the unfertilized eggs of the sea
urchin reach the blastula stage, I was not able to keep these
eggs alive long enough to see them grow into the pluteus stage.
They developed more slowly than the normal eggs, and died,
as a rule, on the second day.

It was my next task to find a solution which would allow the
eggs to reach the pluteus stage. I found that this can be done
by reducing the amount of magnesium-chloride and increasing
the amount of sea water. By putting the unfertilized eggs for
about two hours into a mixture of equal parts of ^^-- nMgCb
and sea water, the eggs, after they were put back into normal
sea water not only reached the blastula stage, but went into
the gastrula and pluteus stages. The blastulae that originated
from these eggs looked much healthier and more normal than
those of the former solution with more MgCl2. Of course as
these unfertilized eggs had no membrane it happened but rarely
that the whole mass of an ^gg developed into one single embryo.
Quadruplets, triplets, and twins were much more frequently
produced than a single embryo. The outlines of each blastula
were much more spherical than in the previous experiment.
These eggs reached the pluteus stage on the second day (con-
siderably later than the fertilized eggs do). Thus I had suc-
ceeded in raising the unfertilized eggs of sea urchins to the
same stage to which the fertilized eggs can be raised in the
aquarium. I have not yet succeeded in raising the fertilized
eggs in my laboratory dishes beyond the pluteus stage.

28o BIOLOGICAL LECTURES.

Though I do not wish to go into the technicalities of these
experiments, I must mention a few of the precautions that I took
in order to guard against the possible presence of spermatozoa
in the sea water. The reader who is interested in this tech-
nical side of the experiments will find all the necessary data
in my publication in the American Journal of Physiology (8).
Here I only wish to mention the following points : —

I. These experiments were made after the spawning season
was practically over. 2. Bacteriological precautions were taken
against the possibility of contamination of the hands, dishes,
or instruments with spermatozoa. 3. The spermatozoa con-
tained in the sea water lose, according to the investigation of
Gemmill (9), their fertilizing power inside of five hours if dis-
tributed in large quantities of sea water.

4. We have a criterion by which we can tell whether the
^^^g is fertilized or not in the production of a membrane. The
fertilized ^g^ forms a membrane and the unfertilized Qg<g has
no distinct membrane. None of the unfertilized eggs that
developed artificially had a membrane.

5. With each experiment a number of control experiments
were made. Part of the unfertilized eggs were put into the
same normal sea water that was used for the eggs that did
develop. None of these eggs that remained in normal sea
water formed a membrane or showed any development, except
that a few of them were divided into two cells after about
twenty-four hours. 6. I made another set of control experi-
ments by putting a lot of eggs of the same female into a
solution which differed less from the normal sea water than the
one which caused the formation of blastulae or plutei from the
unfertilized eggs. In this case it was shown, that although
these eggs received the same sea water as the ones which devel-
oped, and although they were injured less than the ones which
developed, yet not one single ^gg formed a membrane or reached
the blastula stage. If the sea water had contained any sper-
matozoa these eggs should have reached the blastula stage. ^

1 Through other control experiments I convinced myself that a treatment of
eggs or spermatozoa with equal parts of a "^f- nMgCla solution and sea water
diminishes the impregnability of the eggs and annihilates the fertilizing power of
spermatozoa in a very short time.

THE PROCESS OF FERTILIZATION. 28 1

Hence, as in nine different series of experiments these results
were confirmed, we may assume that by treating the eggs for
two hours with a solution of equal parts of a ^-^- nMgCb solu-
tion and sea water we can cause them to develop parthenoge-
netically into plutei.

III.

What conclusions may we draw from these results ? If we
wish to avoid wild and sterile speculations, I think we should
confine ourselves to the following question : What alterations
can be produced in an ^gg by treating the same for two hours
with a solution of equal parts of ^- nMgCU and of sea water ?
Even in this regard we can only give a very indefinite answer,
which, however, will have to be in the following direction : The
bulk of our protoplasm consists of colloidal substances. This
material easily changes its state of matter and its power of bind-
ing water. It seems probable that changes of these two quali-
ties are mainly responsible for muscular contraction and perhaps
amoeboid motions. Among the agencies that cause changes of
these physical qualities we know of three that are especially
powerful. The one is specific enzymes (trypsine, plasmase, etc.).
The second is ions in definite concentration. The concentra-
tion varies for various ions. The third agency is temperature.
In our experiments it is obvious that only the second possibility
can have been active. I do not consider it advisable to enter
into theoretical discussions beyond these statements. The
next question that should be raised would be whether the
spermatozoa act in the same way. It is true that the sperma-
tozoon contains a considerable proportion of salts, especially
K3PO4, but it may contain enzymes or it may contain substances
which have similar effects upon the physical qualities of the
colloids, like the three agencies mentioned above.

In the last volume of these lectures I pointed out that it is
impossible to derive all the various elements that constitute
heredity from one and the same condition of the ^gg. (10) Our
recent experiments suggest the possibility that different con-
stituents of the ^gg are responsible for the process of fertiliza-
tion and for the transmission of the hereditary qualities of the

282 BIOLOGICAL LECTURES.

male. While we are able to produce the process of fertilization
by a treatment of the unfertilized ^g'g with certain salts in
certain concentrations, we cannot hope to bring about the trans-
mission of the hereditary qualities of the male by any such
treatment. Hence, the inference must be that the transmis-
sion of the hereditary qualities of the male and the agency that
causes the process of fertilization are not necessarily one and
the same thing. I consider the chief value of the experiments
on artificial parthenogenesis to be the fact that they transfer the
problem of fertilization from the realm of morphology into the
realm of physical chemistry.

BIBLIOGRAPHY.

1 . The historical data are taken from Hensen's Physiologie der Zeugung.

(Hermann's Handbuch der Physiologie. Bd. vi.)

2. Hertwig, O. Die Zelle und die Gewebe. p. 239. Jena, 1893.

3. LOEB, J. Experiments on Cleavage. Journ. of Morph. Vol. vii.

1892.

4. Morgan, T. H. The Action of Salt Solutions, etc. Arch. f. Ent-

wickelungsmechanik. Bd. viii. 1899.

5. Norman, W. W. Segmentation of the Nucleus without Segmentation

of the Protoplasm. Arch. f. Entwickelungsmechanik. Bd. iii.
1896.

6. Mead, A. The Rate of Cell-Division and the Function of the Centro-

some. Woods Holl Biol. Led. 1898.

7. LoEB, J. On lon-Proteid Compounds and their R61e in the Mechanics

of Life-Phenomena. Amer. Journ. of Phys. Vol. iii. 1900.

8. LoEB, J. On the Artificial Production of Normal Larvae from the

Unfertilized Egg of the Sea Urchin. Amer. Journ. of Phys.
Vol. iii.

9. Gemmill. The Vitality of the Ova and Spermatozoa of Certain

Animals. Journ. of Anat. and Phys. 1900.
0. Loeb, J. The Heredity of the Marking in Fish Embryos. Woods
Holl Biol. Led. Boston, 1 899.

RETURN TO the circulation desk of any
University of California Library
or to the
NORTHERN REGIONAL LIBRARY FACILITY
BIdg. 400, Richmond Field Station
University of California
Richmond, CA 94804-4698

ALL BOOKS MAY BE RECALLED AFTER 7 DAYS

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

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

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

DUE AS STAMPED BELOW

MftY 0 0 2003

12,000(11/95)

:N9 518664

QH301

Woods Hole, h^ss.

M3

Marine Biological

V.7

Laboratory.

Biological

lectures de-

livered at the

Marine Biological

Laboratory of

Wook's Hole.

LIBRARY

UNIVERSITY OF CALIFORNIA

DAVIS