;
:■■..■ ■
AN INVESTIGATION OF EVOLUTION IN
CHRYSOMELID BEETLES OF THE
GENUS LEPTINOTARSA
7k. p. J.
-juu
c^
^c
a*- /*£-
BY
WILLIAM LAWRENCE TOWER
Instructor in Embryology, University of Chicago
Associate, Station for Experimental Evolution
WASHINGTON, D. C.
Published by the Carnegie Institution of Washington
1906
Carnegie Institution of Washington, Publication No. 48.
papers of the station for experimental evolution at
Cold spring Harbor, New York, No. 4-
3 4-of
PSE38 OF
JUDD & DETWEILER, INC.
WASHINGTON, 0. C.
PREFATORY NOTE.
The investigations, the results of which are herein set forth, have been
carried out by aid from the Elizabeth Thompson Science Fund, in the collec-
tion of the data of distribution and migration of L. decemlineata and in some
of the earlier experimental work upon the same species, and by certain grants
from the University of Chicago and from the Carnegie Institution of Wash-
ington, which have made possible experimental study in the tropics. I
desire to express my deep sense of obligation to all these for the aid rendered.
The investigations have been conducted in the Museum of Comparative
Zoology in Cambridge, in the Biological Laboratory of the Brooklyn Insti-
tute of Arts and Science, at Cold Spring Harbor, and in the Hull Zoological
Laboratory of the University of Chicago ; and to the directors of these
institutions and their staffs I am under great obligation for much kindly
advice and valuable criticisms.
The experimental studies presented in this contribution, with the excep-
tion of three small experiments, were brought to an end in July, 1904.
Through the carelessness of workmen engaged to attend to the heating plant
of the university, on the hottest day of the season steam was turned on in
full force in the greenhouse where my experiments were, terminating at
once the series of experiments and cultures which had been carried with
great labor and expense through several years. These cultures have been
started again and are now being continued at Chicago and in the tropics of
southern Mexico.
CONTENTS.
Chapter I. — Geographical Distribution and Dispersion of Leptinotarsa.
Page
The genus Leptinotarsa I
Data of the distribution 4
Geographical distribution and dispersion of the genus 9
The center of origin and dispersion 9
The distribution and migrations of the lineata group 14
Distribution and oecology of the species 14
Distribution of L. intermedia and decemlineata and probable movements
within historic time 21
Chronological history of dispersal of L. decemlineata, 1859 to 1904 25
Occurrence in Europe 39
The highways of dispersal and rate of movement 39
Other factors which have controlled the direction of movement . 44
Distribution of L. defecta and L. juncta and retreat of L. juncta 49
Distribution of the other groups 50
The relation of the distribution of the genus Leptinotarsa to natural environ-
mental complexes 52
Chapter II. — Variation in Leptinotarsa.
Introduction 59
Individual variations 62
Observations and data of continuous variations 62
Color characters 62
Individual variation in the color pattern of the pronotum 62
Individual variation in the color pattern of the epicranium 72
Individual variation in the color pattern of the elytra 76
Individual variation in the color pattern of the ventral surface 84
Individual variation in the color pattern of the legs 86
Individual variation in the color pattern of mature larvae S6
Structural characters 87
Individual variation in the puuctation of the elytra. ... 87
Individual variation in the punctation of the pronotum 90
Individual variation in the glands of the elytra 90
Individual variation in size and shape 90
Observations and data concerning extreme variations 91
The laws of variation in the genus Leptinotarsa 92
Place variation 93
Geographical variation .... 105
Distribution and variation and the evidence they afford concerning evolution ... 113
Chapter III. — Coloration in Leptinotarsa.
Color phenomena in insects 121
Color classes 121
Chemical or pigmental colors 121
Physical or structural colors 123
Chemico-physical or combination colors 123
v
VI CONTENTS.
Color phenomena in insects — Continued. Page
Development of coloration 124
Purity of insect colors 124
Evolution of the color patterns of insects 125
Laws of coloration 1 25
Phylogeny of coloration 126
Effect of environment upon coloration 12S
Protective resemblance, warning coloration, mimicry 128
The problems of insect coloration to-day 128
The colors of the genus Leplinolarsa 129
Chemical or pigmental colors 129
Physical or structural colors 131
Chemico-physical or combination colors 131
The development of colors in Leptinotarsa .... 132
Chemical or pigmental colors 132
Cuticula colors 132
The cuticula, its structure and composition 133
Chemical examination of the cuticula colors 135
Development of the cuticula colors 136
Experiments pointing to the existence of enzyme action 137
Hypodermal colors 140
Chemical examination of the hypodermal colors 140
Microscopical examination of hypodermal colors 141
Development of hypodermal pigments. . . 141
Diffuse hypodermal pigments. . 141
Subhypodermal colors 142
Physical colors 142
Chemico-physical colors 142
The ontogeny of color and color patterns 1 43
In the larvae 143
In the imagines 152
The ontogeny of color on the epicrauium and pronotum 152
The ontogeny of color on the wings 154
On the elytra 154
On the hind wings 157
The ontogeny of color on the abdominal and thoracic segments ... 157
General aspects of color and color-pattern ontogeny 158
Localized stages in ontogeny of coloration 163
The experimental modification of color and color patterns 168
Modification of color in L. decemlineata 168
Temperature experiments 168
Moisture experiments 1 88
Experiments with temperature and moisture both variable 195
Experiments in which temperature and relative humidity vary together
in the same direction above or below the normal 195
Experiments in which L. decemlineata was reared during successive gen-
erations, with both temperature and moisture varying together, or
in opposite directions above or below the normal 201
Experiments with different foods 205
Experiments with soils 206
Experiments with light 20S
Experiments with atmospheric pressure 208
CONTENTS. VII
The experimental modification of color and color patterns — Continued. page
Modification of color in L. signaticollis 209
Temperature experiments 209
Moisture experiments 211
Modifications of color in other species of Lepiinotarsa 211
General results derived from experimental modification of color 212
Adaptation in coloration 215
Warning coloration 215
Protective resemblance ... 220
The evolution of coloration 222
Chapter IV. — Habits and Instincts in Leptinotarsa.
Habits and instincts connected with reproduction 229
Mating habits 229
Assortive mating '. 238
Number of generations 243
Habits and instincts connected with hibernation and aestivation 245
Habits and instincts connected with self-preservation 252
Chapter V. — Production in Experiment of Races, New Characters,
and Species in Leptinotarsa.
Pedigree breeding of L. decemlineata 261
Experiments in race production and modification by artificial selection 261
Selection experiments with color characters 261
Selection experiments with structural characters 26S
Selection experiments with physiological characters 270
Experimental pedigree breeding of new characters and species 271
Pedigree breeding in L. muttitceniata . 283
Experiment in production of new characters and species 286
Experiments with L. decemlineata 287
Experiments with L. muttitceniata. 292
Experiments with L. undecimlineata 294
Conclusion 294
Chapter VI. — The Problem of the Origin of Species — Discussion,
Summary, and Conclusion.
The method of evolution 298
The nature, cause, and control of variation 29S
Mutation 307
The origin of species 310
Bibliography 315
INTRODUCTION.
Although we hold that the general proof of organic evolution is abundantly
self-evident, it is unfortunately all too true that its method is still an open
question. The discovery of the method of evolution and the question of
the origin of species is not, if we may judge by the prevailing diversity
of opinion among biologists, much nearer a solution than it was when Dar-
win left it. This does not mean that no advances have been made ; it simply
indicates that the question is deeper, broader, and more difficult to solve than
was at first supposed.
The current hypotheses of the method of evolution, each supported by
evidence more or less convincing, and each capable of satisfactorily explaining
some phenomena, fail utterly to account for all the phenomena in the origin
of species. In other words, current hypotheses seem to be partial truths
only, and we are probably far from discovering the final truth concerning
the method of evolution. Since the publication of the ' ' Origin of Species, ' '
the amount of literature concerning the general facts and method of evolution
has been stupendous. When we examine this literature, however, we find,
unfortunately, that it is far too much the writings of advocates of one or the
other of the current hypotheses and all too little that of the investigator.
On the question of the method of evolution biologists have grouped them-
selves into different schools, each strongly maintaining that his method is
the real one. With the facts put forward by Darwin and his contemporaries
and such new ones as have since been accumulated, these advocates have
built up ingenious pleas in favor of their hypotheses, but have had little
time left in which to attempt to penetrate farther into the unknown and get
new facts and more conclusive evidence.
Biologists have gradually come to see, however, that any further advance
rests, not upon controversial and argumentative writings, but upon new
investigations so planned and executed as to bring to light new facts and
evidence, whatever they may be, and regardless of how they may affect cur-
rent hypotheses. Already we are beginning to have the fruit of this new
line of work, such as the well planned, clearly executed research of De Vries,
in which is presented a large body of new facts and evidence concerning the
origin of species in plants. Investigations of this character are, however,
the work of years, and a long time must necessarily elapse before any great
body of new data and evidence from these studies can be accumulated.
Eleven years ago I began the study of evolution in the chrysomelid beetles,
especially in certain genera which are confined entirely to America, with the
idea that one ought by stud}', sufficiently long continued and properly con-
IX
X INTRODUCTION.
ducted, to obtain new facts and evidence concerning the origin of species.
This investigation has continued to occupy the larger share of my time dur-
ing these years and is in no wise complete, but I venture to present in this
contribution a first report of progress and a statement of some of the results
obtained in some directions up to the present time.
In this contribution have been brought together data concerning evolution
in the genus Lcptinotarsa Stal as gathered from various sources and in as far
as it applies in the origin of species. In general, the evidence herein pre-
sented has been derived from three sources: (i) its natural history, including
distribution and cecology, variations, habits, and instincts; (2) development;
(3) experiment. In the account herein presented I have in general followed
the order in which the investigation has evolved during the past decade, and
the conclusions drawn at the end rest upon data and evidence converging
from diverse sources and obtained as a result of long observation. The
correctness of my conclusions can be determined only by time and added
investigation.
W. L,. Tower.
The University of Chicago,
Hull Zoological Laboratory,
January 30, 1906.
CHAPTER I.
GEOGRAPHICAL DISTRIBUTION AND DISPERSION OF LEPTINOTARSA.
The Genus LEPTINOTARSA.
The genus Leptinotarsa Stal is one of a number of closely allied genera
which constitute the major portion of the Chrysornelid fauna of America
north of the Isthmus of Panama. In the Mexican and Central American
regions 13 genera and about 225 species are known. Of these all but
Phcedon, which is a genus of cosmopolitan distribution, are entirely confined
to the American continents, and several of them to North America. The
genera and their distribution in Mexico and Central America are given in the
following table :
Table i. — Distribution of the Genera and Species of Chrysomelidas in
Mexico and Central America.
Genera.
No. of
species.
Mexico.
Guatemala.
Nicaragua.
Costa
Rica.
Panama.
Phcedon
Plagiodera
Melasoma
Calligrapha.. . .
Zygogramma. .
Stilodes
Leptinotarsa. . .
Labidomera . . .
Prosicela
Doryphora
Desmogramma
Elytrosphcera..
8
18
3
38
36
14
43
2
3
48
1
9
1
8
10
3
29
30
6
33
2
1
16
9
1
1
9
I
13
8
6
S
2
17
I
3
1
6
4
4
6
14
I
3
9
5
4
4
13
39
3
3
1
6
2
17
1
33
Total
224
148
65
39
With the exception of Phcedon, all of these genera are so closely allied that
only trivial characters are used as generic differentials, and these are often
highly variable. Within the genera the species form groups, or series,
about some central type, and the same type is often found running through
several genera. The slight generic differences, the still less pronounced
specific characters, and the continuity of the series are strongly suggestive
of a recent origin, or at least of a recent rise of species, which may still be
going on.
2 DISTRIBUTION AND DISPERSION OF LEPTINOTAKSA.
The genus Leptinotarsa is confined almost entirely to the continent of North
America,1 extending southward to the Isthmus of Panama and northward to
the northern United States. Forty-three species are known, of which three
have crossed the Rio Grande into the United States, and an equal number
have been found as far south as the Isthmus of Darien. By far the greater
portion of the species is found in southern Mexico ; only a few are known
from Guatemala, and a few from the Mexican Plateau. Within the genus
the species are arranged in groups, in each of which some one type of
coloration prevails to the exclusion of all others. Between these groups of
species there are no intergrades ; these are to be sought for in the parent
genus Zygogramma, where they are often clearly discernible. Specific dif-
ferentiation is almost entirely based upon coloration, is by no means easy,
and is especially difficult in dead specimens. The life histories are almost
entirely undescribed, and the figures of the species that have been published
are misleading, as in Jacoby's work in the Biologica Centrali Americana.
Many of the species are exceedingly beautiful in life on account of their
delicacy of coloration, but at death this is lost.
The following species are known:
Leptinotarsa calceata Stal. Leptinotarsa juncta Guer.
novemlineata Stal. lacerata Stal.
dilecta Stal. heydeni Stal.
flavitarsis Guer. puucticollis Jacoby.
nitidicollis Stal. modesta Jacoby.
obliterata Chev. chalcospila Stal.
lineolata Stal. dahlbotui Stal.
pudica Stal. hogei Jacoby.
typographica Jacoby. haldemani Rogers,
distinguenda Jacoby. libatrix Suffr.
cacica Stal. violescens Stal.
undecitnlineata Stal. chlorizaus Suffr.
diversa n. sp. litigiosa Suffr.
angustovittata Jacoby. tlascalana Stal.
signaticollis Stal. rubiginosa Rogers,
raultita^niata Stal. zetterstedti Stal.
oblongata n. sp. stali Jacoby.
rnelanothorax Stal. evanescens Stal.
rubicunda n. sp. dohrni Jacoby.
intermedia n. sp. belti Jacoby.
defecta Stal. flavopustulata Stal.
decemlineata Say.
1 A few species of Leptinotarsa are described from the northern and western portions
of South America. There seem to be southward extensions of the northern forms. Little
is known concerning them. The references to Leptinotarsas from South America are
usually incorrect, or misprints (as in several instances). In other cases there is uncer-
tainty as to the source of the described material.
DATA OF DISTRIBUTION.
The species of this genus, like the rest of the Chrysomelidse of North
America, show a marked tendency toward the development of groups about
some central type. In Lcpthiotaisa this is carried to the extent that all
of the species fall into well-marked series not connected by intermediate
conditions. Each of these is traced with ease backward into the genus
Zygogramma, which is the ancestor of the genus Leptinotarsa.
The species fall naturally into the following groups :
Flavopustulata Group.
Leptinotarsa stali.
evauesceus.
dobrni.
beiti.
flavopustulata.
Haldemani Group.
Leptinotarsa dablbomi.
hogei.
haldemani.
libatrix.
violescens.
chlorizans.
litigiosa.
tlascalaua.
Lacerata Group.
Leptinotarsa lacerata.
beydeni.
puncticollis.
ruodesta.
chalcospila.
LineaTa Group.
Leptinotarsa undecimlineata.
diversa.
angustovittata.
signaticollis.
multitseniata.
oblongatB.
rnelanothorax.
rubicunda.
intermedia,
decemlineata.
defecta.
juncta.
Dilecta Group.
Leptinotarsa calceata.
novemlineata.
dilecta.
flavitarsus
nitidicollis.
obliterata.
lineolata.
pudica.
typographica.
distinguenda.
Rueiginosa Group
Leptinotarsa rubiginosa.
Zetterstedti Group.
Leptinotarsa zetterstedti.
The distribution of these groups over North America shows a strong
development in numbers and in species in southern Mexico. Only i reaches
the northern United States and Canada, 3 reach the southern United States,
and 6 northern Mexico, while 30 are found in southern Mexico only, 4 in
Guatemala, and 2 as far south as Costa Rica and Panama.
4 DISTRIBUTION AND DISPERSION OF LEPTINOTARSA.
DATA OF THE DISTRIBUTION.'
FLAVOPUSTULATA GROUP.
Leptinotarsa stali Jacoby.
Known only from the headwaters of the Rio Atoyac and the Rio Coetzala, in the
region of the southern escarpment ; recorded from Puebla,* Matamoros de Izucar* (Sall£),
and Atlixco,* all in the State of Puebla.
Leptinotarsa evanescens Stal.
A rare species recorded by Stal from Guatemala and Costa Rica. It has not been
recorded since it was first described. Exact localities are not given.
Leptinotarsa dohrni Jacoby.
A rare species described from a single specimen collected by Salle at Yoltepec. Sall£
does not state whether the locality given by him is Yoltepec in the State of Hidalgo or
Yoltepec in the State of Oaxaca. In the latter State there are two places of the same
name, one in the northern part, 8 miles west of Huajuapan, and another in the south-
western part near Istayulta.
Leptinotarsa belti Jacoby.
Closely allied to L. flavopustulata and perhaps only a variety of it ; is confined to the
Guatemala Plateau and its extension southward into Nicaragua. It is recorded from
Chacoj and Panima in Guatemala (Champion), and from the province of Chontalesin
Nicaragua (Belt).
Leptinotarsa flavopustulata Stal.
Known only from Guatemala, where it is recorded from San Juan in Vera Cruz, and
Chacoj (Champion).
HALDEMANI GROUP.
Leptinotarsa dahlbomi Stal.
A wide-ranging species which has been recorded from the Rio Grande region of Texas
southward to Nicaragua. It is known from the following localities : Texas ;* Mexico —
Yolos, Atlixco,* and Puebla,* in the Stateof Puebla (Salld) ; Yucatan (Pilate); and
Nicaragua — Chontales and Grenada (Salle), Cuernavaca,* State of Morelos.
Leptinotarsa hogei Jacoby.
A curious species known only from one locality. It is clearly a Leptinotarsa , but
bears'a striking resemblance to Zygogramma ornata, with which it occurs. (Mexico —
Cerro de Plumas.*)
Leptinotarsa haldemani Rogers.
A highly variable species confined to the southern end of the Mexican basin and the
upper portion of the escarpment on the south and west. It is recorded from Ventanas
(Forrer), Puebla,* Cuernavaca* (Salle), and Almolonga*2 (Hoge). I have found it in all
stages at Cuernavaca in Morelos, and at Puebla,* Atlixco,* and Matamoros de Izucar *
in the State of Puebla.
Leptinotarsa libatrix Suffr.
Closely allied to the preceding and almost undistinguishable from it. The two do not
occupy the same regions and probably represent geographical races of the same species.
It is recorded from the following localities: Mexico— Cordoba,* Amatlan,* Jalapa,* and
Almolonga,* 2 in the State of Vera Cruz (Salle\ Hoge) ; Xantipa, Mescala,* Savana Grande,
1 Localities where I have collected or known these species to occur are marked with
an asterisk (*).
3 Probably this locality is in the canton of Jalapa in Vera Cruz.
DATA OF DISTRIBUTION. 5
Dos Arroyos, and Chilpancingo, in the State of Guerrero (H. H. Smith); Huetamo, in
the State of Michoacan; Tapatchula, in the State of Chiapas (H. H. Smith); and Guate-
mala— CerroZumil, El Reposo, Volcan de Atitlan, Zapote, and Las Mercedes (Champion).
Leptinotarsa violescens Stal.
Closely allied to L. libatrix and probably only a modified form of that species, which
is found higher up on the eastern and southern slopes of the plateau. It is recorded from
Maltrata,* Orizaba,* Cordoba,* Almolonga,* ' and Tuxtla,* all in the State of Vera Cruz
(Salle).
Leptinotarsa chlorizans Suffr.
A Pacific coast species not seen since it was first described. Recorded from Maz.atlan,
Sinaloa.
Leptinotarsa litigiosa Suffr.
Evidently closely allied to the preceding. Recorded from Mazatlan, State of Sinaloa,
Mexico.
Leptinotarsa Uascalana Stal.
Known only from the original description given by Stal, which would place it as a
near ally of L. dahlbomi. It is quite possible that it is a variation of this well-defined
species. The only locality given by Stal is Mexico.
LACERATA GROUP.
Leptinotarsa lacerata Stal.
Common in the northern and eastern part of the State of Oaxaca and also in the State
of Chiapas. Recorded from Oaxaca* (Hoge) and Etla (Salle) in Oaxaca, and at Playa
Vicente* (Hoge), Paras,* and La Parada2 (Sall£) in Vera Cruz; Cuernavaca,* Barranca
de San Anton,* State of Morelos.
Leptinotarsa heydeni Stal.
Stal gives this species as coming from Brazil, but as all known specimens are from
Mexico it is probable that Stal was mistaken in the locality. It is known from Tanetza
(Salle) and Almolonga*3 (Hoge) in Oaxaca,* and Tremax in northern Yucatan.
Leptinotarsa puncticollis Jacoby.
In many ways resembles L. heydeni. Known only from Ventanas in the State of
Durango (Forrer), and from Sonora (Jacoby).
Leptinotarsa modesta Jacoby.
Described from three specimens from Guanajuato,* Mexico (Sall£); Ocotlan,* Jalisco.
Leptinotarsa chalcospila Stal.
Mexico.
RUBIGINOSA GROUP.
Leptinotarsa rubiginosa Rogers.
A peculiar species not related to any of the other groups in the genus. Recorded from
Mazatlan (Suffr.), the Alvarez Mountains (Palmen), Ciudad Milpas (Forrer), Guanajuato
(Salle), and Puebla* (Salle), Tlalpan,* Sante Fe,* Federal District.
LINEATA GROUP.
Leptinotarsa undecimlineata Stal.
The distribution of this species is, with one exception, the widest in the entire genus.
According to Stal (1S58), this insect is a native of Mexico, Costa Rica, Colombia, and Bolivia.
1 See foot-note 2 on page 4.
2 A locality " La Parada " often given by Salle' has not been located.
3 A name given to a mountain and an accompanying tract of forest-covered hills and
areas of bushy savanna in the valley of the Rio Cosolapa.
6 DISTRIBUTION AND DISPERSION Of LEPTINOTARSA.
Curiously enough, the records of Stal are the only oues known from South America, but
they are probably accurate, as I have specimens from the Rio Sucio * on the Isthmus of
Darieu. In Panama * it seems to be generally distributed, especially in the lower portions
of the country. In Costa Rica, Rogers and Van Patten record it from the slopes of Volcan
de Irazu, and in Nicaragua, Jameson and Belt have found it abundant in Chontales. It
does not seem to have been recorded from Spanish Honduras, although Blancaneaux
found it near the Rio Sarstoon, in British Honduras. In Guatemala numerous records
are known — El Reposo, Purula, Tamahee, Duenas, Capetillo, San Geronimo, Cubliquitz
(Champion), and Guatemala (city); in Mexico — from Vera Cruz (city),* Cordoba,* Tuxtla,
Orizaba,*Misantla,*Cerrode Plumas (Hoge,Sall6),Tierra Blanca,*Motzorongo,*Achotol,*
Guadalupe,* Tesonapa,* ' Vista Hermosa,* Los Changos,* Jalapa,* in the State of Vera
Cruz; at Oaxaca* (Sall£), Mitla,* Tlacolula,* Rincon Antonio,* Ubero,* Mato Ouamado,*
Obispo,* Agua Fria,* Perez,* San Marcos,* Juanita,* and Tomellen,* in the State of
Oaxaca; Guanajuato (Duges, Salle). Its distribution is wide, although it is nowhere a
common species and is always local. At no place does it seem to have reached the Pacific
slope of the Central American or Mexican region, and it is not common along the Atlantic
coast. In Guatemala it is apparently a common species and in Mexico, where it is best
known, it is abundant locally. In Mexico it is far from having the general distribution
given by Stal; in fact, it is absolutely limited to southern Mexico, especially to the
States of Oaxaca, Chiapas, Tabasco, and Vera Cruz.
The distribution of this form is shown on the map (plate I) to be confined entirely
to the Atlantic slope and the savannas of Vera Cruz and Tabasco, but appears tc be
absent from Campeche and Yucatan. To the westward it has occupied the head of the
Rio Balsas Valley, finding entrance through the pass afforded by the Rio Ouiotepec, and
by the Oaxaca plateau.
Lcptinotarsa diversa.
This species is known only from the western portion of the valley of the Rio Grande de
Santiago,* La Barranca,*2 near Guadalajara, Jalisco, Mexico, and near-by barrancas.
Leptinotarsa signaticollis Stal.
This species, also a close relative of L. undecimlineata, is confined to the valleys of the
Rio Balsas system. It has not been found outside of this area, and, indeed, not over the
entire system. It is recorded from Tacambaro and Morelia (Hoge), in Michoacan;
Amula and Xucumauatlan (Smith), in Guerrero; Cuernavaca* (Smith), Alarcon,*
Cuautla,*and Jojutla,* in Morelos; and Puebla,* Matamoros delzucar* (Salle), Atlixco,*
and Tatetla,* in the State of Puebla. The distribution is plotted on plate I, along with
that of L. undecimlineata. The habitat of this species appears to be a belt of country
upon the eastern side of the Rio Balsas Valley.
Leptinotarsa angustovittata Jacoby.
Recorded from Guanajuato (Sall£); from Morelia and Tacambaro, in Michoacan (Hoge);
and from Xucumanatlan (Smith), in Guerrero. I have reared it from L. undecimlineata
fromTierra Blanca,* in Vera Cruz.
Leptinotarsa oblongata n. sp.
This species, previously confounded with L. multitaniata and L. decemlineata, has a
peculiar distribution. It is confined to the valleys of the Rio Balsas, Rio Verde, and Rio
Tehuantepec, and the headwaters of the Rio Quiotepec, in southwestern Mexico. I have
taken it in all its stages at Cuernavaca,* Cuautla,* Yautepec,* Jojutla,* Trienta,* and
Puente de Ixtla,* in the State of Morelos ; Iguala,* Naranjo * Los Ainates,* and Kilo-
1 Also spelled Tepextempa, an old Indian town.
2 The correct name of this locality is La Barranca de Oblatos. A name applied to part
of the canon of Rio Grande de Santiago near Guadalajara, Jalisco, Mexico.
z
•
•
>
o
h
3
:
0
m
•
<
•
•
1:
o
*
4
0
I
•
•
<
o
z
□
4
4)
z
4
4
O
z
3
z
u
Q
•
*
•
1-
M.
Z
<3
•
•
>
°
•
•
i|:
0
■
4
4
o
z
<
•
*
I
111
1-
z
4
4
2
u
4
4
4
DATA OK DISTRIBUTION. 7
meter 191 * (Cuemavaca Division of the Mexican Central Railroad), Balsas,* and along
the Rio Balsas,* State of Guerrero ; Puebla,* Cholula,* Atlixco,*Tatetla,* Matarnoros de
Izucar,* and Tlaucualpican,* in the State of Puebla;* Oaxaca,* Mitla,* Tlacolula,* Tula,*
San Juan,* El Parian,* Tecotnavaca,* and Venta Salada,* in the State of Oaxaca ; and
at Guadalupe * and Texcoco,* in the Federal District. It is common at all of these places
excepting the last two. It is' probable that Hoge's record of L. decemlineata from
Tacambaro, in Michoacan, and Smith's record at Chilpancingo and Venta de Zopilote, in
Guerrero, are really records of this species. The distribution is shown on plate I.
Leptinotarsa multitaeniata Stal.
This species is confined to the Plains of Apam, Puebla, the Valley of Mexico, and the
headwaters of the Rio Lerma. It is recorded from Puebla,* Atlixco,* Matarnoros de
Izucar,* and San Marcos,* in the State of Puebla;* Apizaco * and Tlascala,* State of
Tlascala ; * Irolo,* State of Hidalgo ; Toluca* and Ornetusco,* State of Mexico ; Mexico
City* (Smith), Guadalupe,* Tacubaya,* Tacuba,* Santa Fe,* San Angel,* Tlalpam,*
Tlalnepantla,* Texcoco,* and Mixcoac,* in the Federal District; Cerro de Plumas (Hoge),
and at La Parada (Sall£). The latter place I am unable to locate. The distribution is
shown on plate 1.
Leptinotarsa melanothorax Stal.
This rather uncommon species is recorded from Mexico City * (Hoge) and Guadalupe,*
Federal District; Toluca* (Hoge, Salle), Puebla* (Hoge), Guanajuato* (Duges), and
Morelia (Hoge).
Leptinotarsa intermedia n. sp.
A form transitional between L. multittzniaia and decemlineata, known from the Rio
Lerina Valley at Ocatlau,* La Barca,* Guadalajara,* Zopopan,* the Rio Panuco Valley at
Rio Verde,* and at Chihuahua.*
Leptinotarsa rubicunda n. sp.
Known only from the high plateau at Toluca* and the basal slopes of Mount Toluca*
(altitude 8,500 to 9,500 feet).
Leptinotarsa decemlineata Say.
Although recorded from Mexico, I have not yet been able to find this species iu that
country, and am convinced that it is strictly confined to the United States east of the
Rocky Mountains. It is common in the Mississippi Valley and its tributaries, and east-
ward to the Atlantic coast, northward in southern Canada, and southward to the Gulf of
Mexico.
Leptinotarsa juncta Stal.
This species originally had a much wider distribution than at present. It is now limited
to a narrow strip along the Gulf of Mexico and the lower Mississippi Valley.
Leptinotarsa defecta Stal.
A close ally of L. juncta, which inhabits Texas, Colorado, Missouri, and Indian Terri-
tory. It is recorded from Dallas and Galveston, Texas, and from Pueblo, Colorado. It
has been collected by Salle in Yucatan, but no localities are given. Stal likewise records
it from the same regions from collections made by Chevrolat and Deyrolle. I have not
seen it from these regions and know nothing further concerning it.
DILECTA GROUP.
Leptinotarsa calceata Stal.
Common in the lowlands of the State of Vera Cruz and probably also in the region to
the south ; Misantla* (Hoge), Vera Cruz* (Salle), Orizaba,* and Motzorongo,*all in the
State of Vera Cruz.
8 DISTRIBUTION AND DISPERSION OF LEPTINOTARSA.
Leptinotarsa flavitarsis Guer.
Closely allied to the preceding and probably only a local modification of it in southern
Mexico, Guatemala, and southward. Baly records it from Mexico and Hoge from
Tapachula, in Chiapas. It is recorded in Guatemala from Guatemala City (Salle), Zapote,
and Cerro Zumil (Champion) ; from Nicaragua, in the province of Chontales (Jameson
and Belt), and from Costa Rica (Sall£), but no localities are given.
Leptinotarsa nitidicollis Stal.
Differs only in slight details from flavitarsis and calceata, and as it is known from only
one locality, it is probably only a local variety of one or the other of these species. It is
recorded from Tuxtla (Salle1), in the State of Chiapas.
Leptinotarsa novemlineata Stal.
A close ally of the preceding, recorded from Oaxaca* and Jaquila, in the State of
Oaxaca.
Leptinotarsa dilecta Stal.
Closely allied to the preceding, but known only from the southern end of the plateau
and along the escarpment on the south and west sides. It is recorded from Puebla,*
Yoltepec,* and Yolos,*inthe State of Puebla; Jaquila, in the State of Oaxaca; Cuernavaca,*
in the State of Morelos, and La Parada (?) (Sall£).
Leptinotarsa obliterata Chevr.
This species differs in no respect from nitidicollis except in the reduction of the stripes
to spots. It is a form characteristic of the lowlands, recorded from Toxpam (Salle),
Almolonga * (Hoge), and Cordoba* (Sall£), in the State of Vera Cruz.
Leptinotarsa pudica Stal.
Smaller than the preceding, but separated from it by constant characters. It is recorded
from Cordoba * in Vera Cruz and Jaquila in the State of Oaxaca.
Leptinotarsa Iineolata Stal.
A small species allied to flavitarsis and nitidicollis. It is limited to the northern part
of the Mexican Basin and to the southern portion of Texas,* New Mexico,* and Arizona.*
It is recorded from Chihuahua City * (Hoge) and Piiios Altos (Buchan-Hepburn) in the
State of Chihuahua ; northern Souora, Texas, New Mexico, and Arizona.
Leptinotarsa distinguenda Jacoby.
Known only from Teleman and Chacoj (Champion) in Guatemala and the Province
of Chontales in Nicaragua (Janson).
Leptinotarsa typographica Jacoby.
Closely allied to Iineolata, but of more restricted distribution. It is recorded only from
Piiios Altos in the State of Chihuahua and Ventanas* in the State of Durango (Hoge).
Leptinotarsa cacica Stal.
A common and variable species from the savannas and foothills of Vera Cruz, Ta-
basco, and Chiapas. It is recorded from Cordoba,* Orizaba* (Salle), Misantla* (Hoge),
Cerro de Plumas, Tesonapa,* Mellan,* Guadalupe,* and Amatlan,* in the State of Vera
Cruz, and from Tuxpam, in the State of Chiapas.
ZETTERSTEDTI GROUP.
Leptinotarsa zetterstedti Stal.
Mexico (Stal), La Barranca,*1 Guadalajara, Jalisco, Mexico.
1 See foot-note 2, p. 6.
CENTER OF ORIGIN AND DISPERSAL. 9
GEOGRAPHICAL DISTRIBUTION AND DISPERSION OF THE GENUS.
THE CENTER OF ORIGIN AND DISPERSION.
In any study of the evolution of organisms it is important to know the
origin of the more recent species and to trace as far as is possible their his-
tory. By this means one is better able to determine whether the variations
and modifications which are observed to arise or which produced experi-
mentally are new or simply repetitions of stages in the phylogeuy of the
material studied.
To determine the point of origin in a group of organisms already well
established over a wide area is b)' no means an easy task, and only an
approximately correct conclusion can be arrived at by indirect methods.
With the animals under discussion all evidence points strongly to the con-
clusion that there has been but a single center of origin for the genus. The
place of origin, or " center of dispersal" (Adams, 1902), or "center of
adaptive radiation" (Osborn, 1900), can best be determined by the use of
the criteria given by Adams. Aside from the evidence afforded by fossils,
evidence wholly lacking in this case, ten criteria are given by this author
for the determination of the center of origin, which are more effective than
any method of solution hitherto proposed. We will discard at once the
notion that because these insects are tropical they necessarily came from South
America. This theory of the origin of the fauna of North America deserves
less general application. Other areas, as those in Central America and
Mexico, have been dry land as long as much of South America and were
equally well-adapted to become centers of dispersal. In fact, I suspect that
when we know better the fauna of Mexico and Central America we shall
find abundant evidence that these regions have been strong centers of origin
and dispersal, and that they have supplied both North and South America
with many of their characteristic forms.
The ten criteria for the determination of the center of origin are:
(1) Location of the greatest differentiation of a type.
(2) Location of dominance or greatest abundance of individuals.
(3) Location of synthetic or closely related forms. (Allen.)
(4) Location of maximum size of individuals. (Ridgway, Allen.)
(5) Location of greatest stability and productiveness in crops. (Hyde.)
(6) Continuity and convergence of lines of dispersal.
(7) Location of least dependence upon a restricted habitat.
(8) Continuity and directness of individual variation or modification
radiating from the center of origin along the highways of dispersal.
(9) Direction indicated by biogeographical affinities.
(10) Direction indicated by annual migration in birds. (Palmen.)
IO DISTRIBUTION AND DISPERSION OF LEPTINOTARSA.
Obviously some of these are of more value than others, and some are of no
use whatever for our purposes (No. 10). Others are based primarily upon
the unproven assumption that evolution has been by certain methods (No. 8).
The first criterion is easy of application in the present case. Of the 43
known species of Leptinotarsa, 3 have been found north of the Rio Grande,
6 in the North Mesa, 25 in the South Mesa and the escarpment, 2 1 in northern
Central America, and 2 reach as far south as Panama and the Isthmus of
Darien. Clearly the area of greatest specific differentiation for the genus is
southern Mexico, or the region occupied by the southern end of the Mexican
division of the North American region and the northern end of the Central
American region. This area, situated in the main between the plateaus of
Mexico and Guatemala, is now maturely dissected and immensely diversified.
Three-fourths of the species of the genus are found in this region, and of
these nearly the entire number are confined entirely thereto. In the lineata
group, of the 12 known species 3 are found in the country to the north of
the Rio Grande, 1 both north and south, 1 in the North Mesa, 8 in southern
Mexico, and 1 in Guatemala and southward to the Isthmus of Darien. In
this group the same proportions hold, i. c, three-fourths of the species in
the region in which the genus as a whole shows its dominance, with a stronger
representation to the north than to the south. By this first criterion the
center of origin is clearly shown to be in southern Mexico.
The second criterion is not as easy of application as the first, largely
because the data from many regions are meager or altogether wanting.
First, to consider the genus as a whole : North of the Rio Grande, one
species, L. dccemlincata, is exceedingly abundant, surpassing all of the other
species in the genus in point of numbers of individuals. Dominance in this
area is not of the kind that would have significance in the object of our
search, because the records show that this species in the region where it is
most abundant is an introduced species, a new member of the fauna, and is
at present subject to the laws governing such cases. If, however, this fact
were not known, if the introduction had taken place 100 or 200 years ago,
or even if the introduction had passed unnoticed, we should be obliged to
place the dominance of individuals in the Mississippi Valley and eastward.
Dominance, or great abundance, of individuals is a point carefully avoided
in his notes by the professional collector, the records of the collectors who
have worked in Mexico and Central America being silent on this point.
Excepting in the case of L. decemlineata , my observations indicate a region
of great abundance of individuals in southern Mexico. One species, L. unde-
cimlineata , appears, however, to be common and widely distributed in
Guatemala. Large series have been recorded from there and extensive
collections exist in several museums. We conclude in regard to the second
criterion that it shows the center of origin of these beetles to be in southern
Mexico, but that the evidence derived from this criterion may be seriously
CENTER OK ORIGIN AND DISPERSAL. 11
distorted or even entirely falsified by recent migrations and introductions
that can not be easily detected.
The third criterion is not different to any extent from the first. None of
the species found north of the Rio Grande are closely related. In southern
Mexico, however, the species are as a rule closely related, as, for example,
L. undecimlineata, signaticollis, angustovittata, and diversa. This rather unim-
portant criterion also points to southern Mexico as the center of origin.
The fourth criterion is not, in my opinion, one of any great value.
Size is a character dependent upon nutrition and specialization. Thus L.
decemlineata is larger in New England, New York, and along the southern
shore of the Great Lakes than it is in its original area of distribution, in
Colorado, Kansas, and Nebraska, the difference in size in the two being
about 20 per cent. Likewise Pieris rapes is larger in America than in
Europe, and Passer domesticus lays larger eggs in America than it does in
Europe (Bumpus). So cases might be multiplied in which it is known that
the maximum size of a species" is attained, not in the original habitat, but
in the area later occupied by it, as in the case of decemlineata in this genus.
It is not possible to apply this criterion to the genus Leptinotarsa.
The fifth criterion, like the preceding, is of comparatively little value.
No one would claim that the wild potato of the tropics would under any
circumstances give as many tubers per acre as the cultivated potato of Europe
and America, or that the wild jungle-fowl in India is as prolific as its domes-
ticated races, or that the hare of Europe is as prolific in its native home as
it has proven to be in Australia, or that wild grains are as productive as
cultivated varieties. Productiveness is a character of high specialization in
so many cases, such as cultivated varieties, often, if not always, accompany-
ing introduction into a new habitat, that it seems to me to be of little use
for the purpose for which Hyde proposed it. Stability, however, is quite
another matter, for we may have stability as the result of rigorous selection,
as in cultivated varieties, or stability as a specific property of the species, as
it occurs in nature.
The sixth criterion is of prime importance and one which, when the evi-
dence is good, is sufficient to determine beyond question the center of
adaptive radiation. Rarely can the highways of dispersal be determined
from records or from the correlation existing between the movements of ani-
mals. In the majority of cases this data can be determined only by a study
of the continuity and directness of individual variation and modification
along certain lines. Adams's eighth criterion is in reality a highly neces-
sary part of his sixth, it being merely a method of determining the impor-
tant data in his sixth. This criterion presupposes, however, that all evolu-
tion has been by continuous variation. It is quite probable, however, that
if the theory of the origin of species by mutation should prove to be of wide
application this criterion would be of little use. If the mutation theory
12 DISTRIBUTION AND DISPERSION OP LEPTINOTARSA.
is true, then continuity and directness of variation are no longer of any great
value in the determination of relationships and the phylogeny of animals.
In Lcptinotarsa the lines of dispersal are short and not any too clearly
marked. This is due to the fact that the habitats of the various species are
only a few miles apart, and that they have been produced, not by wide sepa-
ration, but by unlike topographic and climatic conditions situated in the
main very close together. It is not possible, therefore, to say with certainty
that lines of dispersal exist, except in one or two cases. The lines of dis-
persal which can be determined, those of intermedia and decemlineata, point
southward to southern Mexico as a center of origin.
The seventh criterion, "location of least dependence upon a restricted
habitat," does not appear in this case, at least, to be of any great weight.
For example, L. decemlineata in the eastern United States is less dependent
upon a restricted habitat than is any other member of the genus. That this
is true is shown by the fact that it regularly feeds upon some dozen or fifteen
species of food plants and has been found upon about forty. No other mem-
ber of the genus has so many food plants, nor does decemlineata in its origi-
nal habitat in Colorado, western Kansas, and Nebraska. Likewise it occurs
on dry or wet or sandy soils, on clay or rich alluvium, on plain, hillside, or
mountain slope, in the open, and even in the forest. In the case of this
species least dependence upon a restricted habitat is a new character (habit)
and not an original one. In its original home it feeds almost exclusively upon
,S. rostrahim and lives in the open country (savanna and steppe formations) ;
never upon mountain slopes nor in wooded areas.
Other members of the genus known to be nearer to the point of adaptive
radiation, as nndecimlineata, have only one species of food plant and are
restricted to a very narrow habitat — the open portion of the country. The
genus is never found upon mountain sides or in woodland formations.
Likewise mnltitceniata, the direct ancestor of decemlineata, feeds only upon
5. rostratum and its allies, and does not live upon any other. In this case
' ' least dependence upon a restricted habitat " is a late acquisition of the
species and not an original character.
That Adams's seventh criterion is often one of value is certain, but the
case cited above, and others equally good that can be brought forward, would
serve to weaken seriously one's confidence in this criterion as one of general
value. It might be useful in specific cases, when the other evidence and
the history of the group are so fully known that the recent acquirement of
this character is beyond question, but when we know this much about the
animals in question, the criterion is no longer needed in the search for the
center of adaptive radiation. It would be already known from more reliable
evidence.
The eighth criterion, "continuity and directness of individual variation
or modifications radiating from the center along the highways of dispersal,"
DISTRIBUTION AND MIGRATION OF UNKATA GROUP. 1 3
is no criterion at all, although it has been shown in the discussion of the
sixth criterion that continuity and directness of individual variation or
modification may be of value in the determination of highways of dispersal.
The eighth criterion assumes, if I understand it correctly, a knowledge of
the center of origin and highways of dispersal previous to the beginning of
the study of the data to determine that center of origin. It certainly is
impossible to apply this to the genus Leptinotarsa, although it does assist
very directly in the solution by criterion No. 6.
The ninth criterion, "direction indicated by biogeographical affinities,"
seems to be a loose reiteration of the ideas expressed in No. 3, as applied
to a somewhat broader area. It is, however, a criterion of no value from
the standpoint of this group of animals.
The tenth criterion, "direction of annual migration in birds" (Palmeu)
has no bearing upon the present case.
To sum up the evidence derived from the application of these criteria, it
appears that 1,2, and 3 point conclusively to southern Mexico as a center
of origin; 4 is shown to be of no value, and 5 is based, seemingly, upon
wrong conceptions and data. The important criterion, 6, points unmis-
takably to the same area as Nos. 1,2, and 3, while from criteria 7, 8, 9, and
10 we derive nothing of value. Our evidence clearly indicates that southern
Mexico is the center of origin or adaptive radiation of the genus Leptinotarsa.
There is nothing to even suggest an origin in South America, and we may
well surmise that like results will be obtained concerning other groups of
animals when adequately studied.
The following criteria are adequate for determining the center of origin
or adaptive radiation without the introduction of any of doubtful value:
(1) Location of greatest differentiation of a type. (Adams.)
(2) Continuity and convergence of lines of dispersal. (Adams.)
(3) Location of synthetic or closely related forms. (Allen.)
(4) In some cases, location of dominance or great abundance of individ-
uals. (Adams.)
These criteria cover fully all cases and are not open to the very serious
objections that may be advanced against the others.
The distribution and dissemination of these beetles can be best considered
by groups of species, as we shall then be able to follow closely one line of
species differentiation at a time and examine the effect of the different environ-
mental complexes in controlling and directing distribution and dissemination.
14 DISTRIBUTION AND DISPERSION OF LEPTINOTARSA.
THE DISTRIBUTION AND MIGRATIONS OF THE LINEATA GROUP.
Distribution and CEcology of the Species.
LEPTINOTARSA UNDECIMLINEATA Stal.
It is highly probable, although it can not be absolutely proven by actual
observation, that the Guatemala-Chiapas Plateau, when perhaps it had a
much lower altitude, was the original habitat, the center of origin, and first
center of dispersal for this group of beetles. That L. undccimlineata is the
most primitive of all the species in this group and close to the parent form
is reasonably certain. The data of its distribution, which has been given,
shows that, as far as is known, it is the only species of the group in Guate-
mala. It is evident that it did not come from South America, but is strictly
a Central American form.
Of its habitat or habits in Guatemala I know nothing. In Chiapas,
Oaxaca, and Vera Cruz I have been able to study it in nature and to
discover some of the factors which control its range of distribution and
confine it to a restricted habitat.
In a very general way the distribution is shown on plate i, where L. unde-
cimlineata is shown to be distributed over the American Continent to the foot
of the great escarpment on the southern end of the Mexican region. It is
distinctly a tropical species. The distribution is irregular, it being confined
to the valley floors and the low coastal plains of Vera Cruz, Oaxaca, Tabasco,
and Chiapas, andsouthward, and it is not known from the Pacificcoast regions.
Throughout the entire area in which I have studied this species it is always
found in open grassland formations (plate 2), never in the forests, and is
local in its distribution, being limited to isolated colonies in narrow and
precise habitats, from which it does not wander far. These colonies are
maintained from year to year in almost exactly the same spot. The situa-
tion in which the colonies thrive best is where there is a growth of their food
plant upon the edge or within a few feet of the edge of a small, deeply-cut
stream, so deeply cut that the chances of overflow are slight, or on a slightly
elevated mound in the meadow or savanna, but there must be either an
abundant supply of telluric water or the superficial soil must be kept moist by
frequent precipitation. The soil most favorable to it is a fine residual or allu-
vial soil sufficiently free from clay and cementing materials, so that it does not
become too hard during dry times. These are the general conditions best
adapted to this insect — conditions which can be realized only in rather
restricted areas. The immediate edges of the rivers in the country in which
it lives are subject to extensive overflow, and hence are impossible sites, as
are also the open, level flood plains. Neither a hillside nor the talus slopes
at the foot of a hill provide the ideal conditions. In fact, only that portion
of the valley situated between the upper limits of the rainy season floods
and the talus slopes of the hillsides provides the conditions necessary for
-1 c
x -
S S
DISTRIBUTION 01-" I.. UNDECIMLINEATA. 15
habitat formation — a region which, in the mountainous parts of the country,
is frequently wanting altogether, and when present is usually narrow and
level and deeply cut by tributary streams. The food plant of this beetle
grows from the edge of the river to the talus slopes, and even far up on the
hillsides; and, while stray beetles are often found in locations where envi-
ronmental conditions are adverse, the characteristic, well-developed colonies
are never thus located, because any extended existence and development in
these places is not possible.
The distribution of L. undccimlineata is in isolated colonies and it does
not occupy the wide range of habitats that its food plant does. The isola-
tion is strikingly apparent to one searching for these beetles, wide stretches
of country often existing between the colonies. When found the insects are
usually numerous, and I have often been able to collect from 100 to 300
specimens in a short time within a radius of a few feet.
From these colonies dispersion occurs by flying aud not by any means of
transportation that I have been able to discover. This flight is observed to
begin at the opening of the rainy season, when the insects emerge from
aestivation, to continue slightly during the rainy season, and to increase
greatly toward its close.
An instructive illustration of the manner in which the dispersion of this
beetle takes place was afforded by the recent building of a railroad through
a perfectly flat, frequenth- flooded savanna near Tierra Blauca. The food
plant grows generally over the savanna, but the beetle is entirely absent,
excepting at a few points along the road where the work of constructing
ditches to keep the road-bed intact has created new localities with favorable
conditions for their existence. Over a distance of about 18 kilometers there
are now located flourishing colonies at each place where the work of the
railroad builders has made existence possible, while on the unmodified
savanna I have not been able to locate a single colony, and doubt if there
are any. In this instance the advance into a new area has occupied two
years and has been rapid. That transportation did not bring about the
starting of these colonies is certain, as the work of railroad construction
was entirely suspended during the rainy season, when the beetles are active
aud dispersion takes place. It is perfectly clear that in this case the distri-
bution was brought about by some few individuals from a colony happening
by chance to discover the newly created habitat, proper for aestivation and
for the breeding of the next generation. In each generation many will
perish b\' not being able to reach the proper habitat after once having aban-
doned the parent colony, but the fact remains that some do discover proper
habitats, and when such are found new colonies are established. In this
way, in regions where previously existence was impossible, as in the case of
the savanna described above, new colonies are established. In the case of
the savanna near Tierra Blanca the production of favorable locations was,
1 6 DISTRIBUTION AND DISPERSION OF LEPTINOTARSA.
it is true, due to the processes of railroad building, but if the same condi-
tions had been produced by floods or other natural agencies which may act
rapidly to change local conditions by cutting new waterways, the newly
created possible habitats would in all probability have been seized upon with
equal suddenness.
We see how definitely undecimlineata is limited in its distribution to a
particular portion of the locality in which it lives, and it is probable, there-
fore, that its present distribution through the valleys of Vera Cruz, Oaxaca,
Tabasco, and Chiapas has been a growth dependent upon the development
of appropriate habitats at successively higher and higher levels in the river
systems, so that the beetle has been brought by a series of steps nearer and
nearer to the Mexican region. Its food plant everywhere precedes it, and
as fast as appropriate conditions of soil and moisture are developed at any
place, the beetle occupies it. This process of reaching out for new sites is
easily seen on the eastern slopes of the Mexican highlands at Orizaba or
Jalapa, in the State of Vera Cruz, where well-established colonies are found.
From these colonies numerous individuals find their way to points higher
up on the valleys, where food, temperature, and moisture are favorable,
but where permanent colonies are not established, because the conditions of
existence at these points are unfavorable to aestivation and pupation ; and as
long as the conditions demanded are not fulfilled no permanent colony can
be established. It is not necessary that the soil should be of a special chem-
ical composition or temperature and rainfall of special amounts, but it is
essential that during pupation and aestivation the beetle shall not be sub-
jected to excessive desiccation or moisture, and that the soil shall be porous
enough to admit of an abundant supply of air.
The southward distribution of this insect is not well known. The data is
meager, consisting of records of capture, and I have not been able to visit
these southern regions to study its habits and distribution there. However,
from the data which we have, it seems probable that it occupies the same
habitat there as elsewhere, and that it is limited by the same determining
factors as in the northern portion of its range. It has extended south to
the isthmuses of Panama and Darien. It is everywhere a Caribbean or Gulf
coast form, never, as far as our knowledge goes, occurring upon the Pacific
slopes.
LEPTINOTARSA DIVERSA.
The narrow range and the few observations that have been made upon this
species do not permit of any extended discussion of its distribution or the
control of its habitat. It is, as I have pointed out, a close ally of undecim-
lineata, and, like it, is confined to a narrow habitat. In the barrancas of the
Rio Grande de Santiago it occurs upon clumps of Solatium sp.?, which grow
near the edges of the cliffs and upon their faces and the steeper sides of the
n o si
DISTRIBUTION AND MIGRATION OF I.INEATA GROUP. I"
barrancas. It does not seem to extend outside of these limits, but is dis-
tinctively a cliff and talus-slope form. Just whatrelatiou this species has to
the preceding from the standpoint of distribution it is impossible at present
to saw
LEPTINOTARSA SIGNATICOLLIS.
Apparently this species and the two preceding ones nowhere occupy the
same locality. The data of its distribution would indicate this, and, although
they are closely related, I have nowhere been able to find them living together.
Like the two preceding species, this one is an inhabitant of a particular
portion of the country. On plate I is shown the general distribution, where
it is represented as limited to the western side of the Rio Balsas Valley, and
on plates 3 and 4 are shown photographs of the habitat and its relation to
the general physiographic features of the country. It occupies the northern
and eastern tributary valleys of the Rio Balsas system, between the altitudes
of 4,000 and 5,500 feet. This distribution embraces the lower portion of
the Mexican escarpment, many of the great washes, as at Cuernavaca, and
the upper portion of the lower country leading up thereto.
L. signalicollis, like undecimlincata, is confined to a narrow local habitat.
Its food is Solatium sp. , and it lives in colonies which continue in the same
place from year to year. The general habitat best adapted to this species is
similar to that occupied by undecbnlineata, and is shown in plate 4. The
favored locality for a colony is usually a growth of its food plant upon or
near the edge of a small stream or ditch, cut somewhat deep into the soil
of the flatter portiou of the valley floor, or on the upper slopes of the larger
barrancas (plate 3). It is rarely found near the edge of the smaller barrancas,
or at the foot of the hills.
The rigid restriction of both the food and the beetles to a narrow habitat
tends to produce a band or zone upon either side of the valle3'S in which, at
points where the proper conditions have beeu developed by the side streams,
colonies of signaticollis are found. This band or zone is in some cases rather
sharply drawn, as at Cuernavaca, and less rigidly in others, as in the region
of Atlixco and Matamoros. The difference is, however, one of topography.
At Cuernavaca the topography is sharp and decisive ; at Atlixco and Mata-
moros it is more open, less rugged, and the zone of possible habitat formation
is much broader and correspondingly less distinct. This restriction to so
narrow a habitat has resulted in isolation, often of great intensity ; but this
has not brought about local variation, so that each valley shows the same
peculiarities seen in adjacent ones. The intense isolation and segregation
has not been of any apparent utility in the production of local forms or
races of L. signaticollis.
As far as I am able to observe, the same method of dispersal is found in
signaticollis as in the preceding species, but I have not had an opportunity
to study its method of spreading over a new area.
2— T
1 8 DISTRIBUTION AND DISPERSION OF LEPTINOTARSA.
LEPTINOTARSA ANGUSTOVITTATA.
The relation of this species to environmental complexes can not at present
be discussed. It is known from only a few localities, is never abundant, and
I have never been able to study it to any extent in nature. The fact, however,
that I have reared it in experiment from undecimlineata and that it is inter-
mediate between undecimlineata and sigjiaticollis, is suggestive, but its full
significance must await experimental elucidation.
LEPTINOTARSA MULTITAENIATA.
This species, like undecimlineata, is a central one about which closely
related species are grouped. The data of its distribution is given on page
7, and on plate i is plotted its distribution. It is limited to the Valley of
Mexico, the Plains of Apam, and the flat mesa southward to Puebla and
Matamoros in Puebla ; eastward it extends to the foot of the Sierra Madre
Orientale, and northward to about 20° 30" north latitude. It is everywhere
confined to the level country, never being found in the lowlands nor upon
the mountain sides.
The area which is the habitat of this species is distinct topographically,
being sharply marked off from the rest of the Mexican Plateau. It is level,
but is surrounded by high hills and mountain chains with volcanic peaks to
the east, south, and west, and cut off by a rugged range to the north. It is
highest in the Valley of Mexico (7,400 feet), sloping southward to Puebla
(7.091 feet), and rising slightly towards the Sierra Madre Orientale. It is
practically level, with a variation of only a few hundred feet in altitude.
Over it are scattered low hills, often of volcanic origin, small volcanic cones,
eroded necks, and lava flows with steep sides covered with talus. The
streams, which are all of small size, have beds deeply cut from 10 to 50 or
100 feet in depth, with perpendicular walls in a stiff, sticky deposit of modified
volcanic ejecta. The drainage is largely to the south, through the Rio
Atoyac into the Rio Balsas system, and to a limited extent to the northeast
and east through the short Gulf coast streams.
The rainfall is least in the lee of the Sierra Madre Orientale, where pine
barrens and semi-arid grassland conditions are extensively developed ; it
increases gradually to the westward, and is largest in the Valley of Mexico.
The northeast trade winds, striking the high eastern mountains, lose the
greater part of their moisture upon the eastern slopes and pass over the
mountains as cold dry winds, bringing but little moisture to the eastern
portion of the country, which is always semi-arid. To the westward these
winds become warmer and more moist, and the rains more abundant, until
in the Valley of Mexico the rainy season has copious daily showers.
This region is a rather old land area which had stood for a long time at a
much lower altitude and had been worn down to base level with residuals
about it. Subsequent volcanic action and mountain building produced
/* -
£ 1
DISTRIBUTION OF L. MULTIT.UNIATA. 19
basins like the Valley of Mexico, where were located great Tertiary lakes
which have been filled with sediment and volcanic ejecta until only remnants
remain, as, for example, the shallow lakes in the Valle5r of Mexico. In these
lakes various Tertiary land vertebrates have been preserved. Subsequently,
the whole area was elevated and the precipitation lessened, and the lakes
began to dry up, the volcanic ejecta and the sediment brought down from
the surrounding mountain ranges was too abundant to be removed by the
streams, and the old valleys became filled up with a thick covering of adobe
soil, largely of modified volcanic debris. The area now presents all of the
characters of a mountain park or basin.
This region has served as a second center of dispersal for these beetles,
and one in which new forms have originated and are probably now being
produced. Beyoud doubt multit&niafa is the parent form of this group.
In its habitat multitceniata is a typical grassland form, where it feeds upon
Solatium rostratum or its allies, and has a much wider habitat than any
of the preceding species. It is most abundant about Puebla and in the
Valley of Mexico, especially upon the western side ; and it becomes less and
less common to the eastward and to the north until it finally vanishes. The
food plant is generally distributed, especially about places where cattle,
horses, or burros are pastured or in waste places, the seed being carried to
these places in the hair of the animals. The general aspect of its habitat is
shown in the photograph on plate 5, which shows an almost pure steppe.
As in the preceding species, the control of the distribution within the
habitat is dependent primarily upon moisture which is regulated by soil and
the precipitation. The soil conditions demanded are a fine, moist adobe earth,
so situated as regards drainage, either by streams, ditches, or a porous subsoil,
that it shall not remain saturated with water for any length of time. If the
soil remains saturated, or if it dries out too quickly, fatal results will rapidly
follow to pupae or aestivating imagines and will produce a high rate of mor-
tality, if not complete extinction. Over the Valley of Mexico and about
Puebla the optimum conditions are extensively developed, but even in these
places the beetle thrives best near a ditch or the edge of a small stream
(plate 6), which is cut deep enough so that the danger of overflow is slight,
insuring adequate drainage and an abundant supply of telluric water. Over
the Plains of Apam the soil conditions are more variable, and, consequently,
the distribution is much less uniform than in other portions of the habitat.
The hillsides, talus slopes, or water-washed sands are tabooed, as also, to a
less degree, the low, marshy areas. The condition demanded for existence
by L. multitceniata are much like those of the preceding species, but the dis-
tribution in isolated colonies is not found. The continuous distribution so
characteristic of this species is due to the topographic development of the
area in which it lives, which has produced over wide areas the conditions
requisite for proper habitat formation. It is of interest to note in connec-
20 DISTRIBUTION AND DISPERSION OF LEPTINOTARSA.
tion with the horizontal extent of its habitat that this species shows the
greatest range of individual variation of any in the group, a condition
dependent upon fluctuating differences in its habitat which are sufficient to
produce marked individual variation, but are not of enough magnitude to
act as barriers to its distribution. From the optimum conditions there is a
continuous series of gradations in the environment to the minimum, whereas,
in the case of the species previous^ described, the changes in the habitat are
rapid, often abrupt, and serve as effectual barriers.
Closely associated with this species are the forms or species mclanothorax ,
oblongata, intermedia, rubicunda, and decemlineata, which are the direct
descendants of this ancestral stock. In some cases the difference between
the species is so slight that separation is difficult, and is especiall)' so in
museum material. In life, however, the differences are abundant and defi-
nite, so that the various species can be recognized easily.
LEPTINOTARSA OBLONGATA.
This species, clearly distinct from the preceding, has been by previous
writers confounded with multitceniata. The food plant and the conditions
which govern its position in any particular place are the same as for multi-
tczniata, and need not be restated. In the upper part of the Rio Atoyac
Valley the two species live together upon the same plant, but are, as I have
determined by breeding experiments, physiologically isolated. The same is
true in the Valley of Mexico, where this species also occurs, although rather
sparingly. From the Plains of Apam this species extends southward into
the Rio Balsas Valley and the Oaxaca-Guerrero Highlands, but does not
reach as far south as the Isthmus of Tehuantepec. Everywhere its habitat
is narrower than that of the preceding species, and its range of variation is
less. This form is a direct descendant of multitaniata, which has become
established in the region where it now lives.
LEPTINOTARSA MELANOTHORAX.
Associated with multitceniata is a form described under the above name.
It is never common only a few specimens appearing at any one time. Its
food is Solatium rost ration , and the controlling factors of its habitat are the
same as those for multitcrniata. Its distribution is irregular, as is also its
occurrence in any locality. I have reared this form from eggs laid by
multitaniata. It is an elementary species which appears at various localities
at frequent intervals, but which has not yet been able to become established
as a member of the fauna of the habitat into which it is born. The point
to be noted in this connection is the fact that the appearance of this ele-
mentary species occurs at several points rather constantly, as is shown in
the data of its distribution. We shall have occasion to consider this case
again in a later portion of this paper.
ft
V
I
Q
5 'J
^ 8
■^MMMH^VHHSH
DISTRIBUTION OF LINKATA GROUP. 21
LEPTINOTARSA RUBICUNDA.
At Toluca, Mexico, at an altitude of 8,500 to 9,000 feet, is found another
form, derived from multitaniata, with which it lives. This form {rubicunda}
is, however, able to maintain itself only within a narrow area as a member
of the fauna. Its distribution is shown on plate 1. As far as I am able to
determine, the habitat of this form is limited to the flat plain about Toluca,
where it lives upon Solanum rostratum in physiological isolation from mul-
titceniala. It is able to reproduce itself without change, as I have proven
by breeding it for six consecutive generations. This form is a developing
species, as is shown by the fact that my specimens of multitceniata from
Toluca have given me as an extreme variation a form exactly like the
rubicunda which I obtained at Toluca. These extreme variations were
almost completely sterile when crossed with the parent species, but freely
fertile when crossed with the rubicunda strain from Toluca. This experi-
mental evidence I shall present in detail in another chapter of this paper,
but the facts above noted are strong enough evidence to serve as a basis for
asserting that rubicunda is a species which is able to maintain itself as an
independent member of the fauna of its habitat. Whether it can do this to
the extent of widening its boundaries and adapting itself to slightly changed
conditions only time, careful observation, and experiment can determine.1
Distribution of Leptinotarsa intermedia and decemlineata and Probable
Movements within Historic Time.
leftinotarsa intermedia.
Over the central table-land in the valleys of the Rio Panuco, the Rio
Grande de Santiago y Lerma, and the Rio Grande systems, and throughout
southern and western Texas, New Mexico, and Arizona occurs, sparsely
distributed, another form, to which I have given the name intermedia. Its
food and habits and the conditions of its habitat are the same as those of
multitcEniata, excepting that it occupies a region of much less rainfall. Like
multitcsniata, it shows a wide range of individual variation corresponding to
its horizontally extensive habitat. The distribution is decidedly discontinu-
ous, due largely to the conditions of climate, and more remotely to topography.
In the southern portion of its range this form merges with multitaniata
and to the north with decemlineata. Over most of its range it is in life dis-
tinguishable from either of these species by constant characters in both larva
and adult.
The distribution of this form within the area occupied by it is not to any
great extent controlled directly by topography, because everywhere proper
conditions of topography and soil are present. The relative humidity and
1 The distribution of L. rubicunda was in September, 1906, limited to a tract of about
10 acres in extent, where only a very few specimens were found. In August, 1903, it
was found over an area of some 7 or 8 square miles. Apparently it is being exterminated.
22 DISTRIBUTION AND DISPERSION OF LKPTINOTARSA.
rainfall are the controlling factors. Whenever and wherever there is suffi-
cient moisture this species and its food plant thrive ; wherever moisture is
lacking for a year both fail to develop and lie dormant in the earth in
aestivation until such time as the rains come to awaken them. This
direct control by the precipitation is productive of an irregular distribution
and complete absence from many otherwise favorable localities, and is due
to its living in the transitional zone between grasslands and desert. The
photograph of its general habitat (plate 7) shows the nature of this zone.
A rather striking feature of the distribution of this species is the fact that
it is most abundant along highways of human traffic, and often present there
only. Since the beginning of the occupation of northern Mexico by the
Spanish two chief routes of travel to the north, those along the eastern and
western edges of the plateau, have been in almost constant use. Along
these — the eastern leading into Texas, the western into New Mexico and
Arizona — Solatium rostratum is distributed as a weed, growing largely, not as
a native plant, but as one introduced through transportation by beasts of
burden. It everywhere grows best about corrals, stables, and watering
places where cattle congregate. Mr. Pringle, who for twenty years has
studied and collected extensively the flora of Mexico, tells me he is fully
convinced that these plants, as well as others, have been carried from places
in southern Mexico northward into the regions where we now find them by
Spanish pack-trains and caravans. From the study of the distribution of
intermedia, I had also come to the conclusion that these insects had been
able to spread northward only by means of some outside help, and Mr.
Pringle's observations concerning the food plant is strong evidence in sup-
port of this conclusion. It is highly probable that previous to Spanish occu-
pation and exploration these beetles and their food plant were limited to the
area at present occupied by multitecniata .
From the conquest of Mexico, in 1521, until the close of the sixteenth
century numerous exploring expeditions traversed the country north, west,
and south. Early in the seventeenth century colonies were established in
northern Mexico, Texas, New Mexico, and Arizona, and the final conquest
of these regions was completed in 169 1. It was during this period that the
main routes of travel to the north became established, and it is probable, if
the above hypothesis be true, that it was during this period that Solatium
rostratum and L. multitczniata were able to extend their habitats into the
northern part of the Mexican area through aid in transportation. L. inter-
media, therefore, has probably occupied its present area of distribution since
early in the eighteenth century.
Whether the above hypothesis be true or not one fact is certain, that inter-
media exists to-day most abundantly along the three chief routes of Spanish
travel and commerce, and often there only. No feature of topography or
climate could be productive of such a distribution. Moreover, I have found
*"->
\» > I
«*-:
DISTRIBUTION OF h. DECEMLINEATA. 23
in several instances that as new routes of travel become more and more
commonly used, as, for example, those leading to irrigated arid lands, 5. ros-
tra turn aud L. intermedia soon increase along these lines and enter areas not
hitherto occupied by them.
LEPTINOTARSA DECEMUNEATA PREVIOUS TO 1859.
Greater doubt exists concerning the original distribution of this species
than of that of any other. I have shown it to be highly probable that the
movements of Spanish travel and commerce resulted in the dispersal of
Solatium rostratum , an essentially tropical plant, into Texas, New Mexico, and
parts of Arizona. We know that this plant was and is nowfound farther north
than the most northern limits of Spanish activity, and we may now inquire
into the factors which were possibly active in its dissemination over the Great
Plains as far as the boundary between the United States and Canada.
Solarium rostratum grows best about places where cattle congregate,
especially about drinking places and wallows, although it is found else-
where. Its seed pods are well adapted to being transported by large mam-
mals, being armed with hooks and spines, which provide efficient means
whereby they are held in the hair of the cattle and thus carried often for
long distances. That this method of transportation has been the means of
spreading this plant over the country to the east of its original habitat is a
fact known from many records. I have often observed burro trains, cattle,
and horses which had gathered at places where Solatium rostratum was grow-
ing to leave these places with its burs entangled in the hair of their coats
so firmly that a considerable journey might be completed without their
becoming dislodged. This plant is therefore liable to be carried in any
direction by cattle, but principally in the lines of travel followed by pack-
trains. The heavy seeds, devoid of any means of dispersal, are largely
dependent upon this means for their dissemination. If we grant that Spanish
caravans brought this plant into Arizona, New Mexico, and Texas, and that
without outside assistance it would be impossible for it to spread long dis-
tances to the north of the range of Spanish activity, what agencies were
present in this portion of North America at this time that would be able to
bring about this result ?
The agents most likely to do this seem to be, first, migratory mammals ;
second, wandering bands of men, and third, possibly, birds. Other agencies,
such as wind, water, and storms, are clearly of little effect in this case. The
third agency suggested is also of such slight importance, if of any, that we
may as well discard it at once. Of the two remaining possibilities it seems
to me that the first has in this case by far the stronger hold upon our atten-
tion and the greater chance of being a true cause of dispersal. It is known
that huge herds of bison on the Great Plains1 migrated south into Texas
•Although the bison ranged much farther east at this time, the huge herds of the Great
Plains seem to have been distinct from the eastern bison.
24 DISTRIBUTION AND DISPERSION OF LEPTINOTARSA.
and New Mexico in the winter and north in the summer, and further, that
the area throughout which this plant was found before it began to spread
eastward was coincident with the range of the Great Plains herds of bison
at the end of the eighteenth century. Now, while I have often seen the
burs of this plant clinging to the hair of cattle, I have not seen them on
bison and know of no record of it ; but if they will cling tenaciously to the
hair of the cattle of the Great Plains to-day, there is no reason to believe
that the thick, curly hair of the bison would form a hold any less secure.
On the contrary, the bison would be a far more efficient agent of dissemina-
tion because of the ease with which the seed pods would become firmly
entangled in its hair and be transported perhaps to great distances.
We know from actual records that this plant has advanced steadily east-
ward in the last fifty years, chiefly by the means of transportation described
above, and if this has been its method of dispersal in recent years, in the
absence of evidence that conditions have changed we must regard the same
method as the essential one in the time before records were made.
Wherever the food plant goes the beetles go also, provided the soil and
climate are favorable, and over the area now occupied by Solatium rostratum
conditions for the existence of L. intermedia were favorable. Into this
region these beetles must have spread gradually, and on the eastern slope of
the Rocky Mountains have undergone modification which resulted in the
production of L. decemlineata. The original distribution of dccemlineata
was on the eastern slope of the Rocky Mountains northward to the Canadian
boundary, eastward into western Kansas and Nebraska, and southward into
Texas and New Mexico. In this habitat it was found by Say in 1823.
Then, as now, it was probably sparsely distributed over the area, feeding
upon Solatium rostratum. I know of no record of its having been found at
an elevation of over 8,000 feet, or in the Great Basin.
In this habitat it remained in stability until 1845 or 1850, when the western
extension of human colonization introduced into its habitat a new factor by
the addition to the flora of a new plant, Solan?im tuberosum, which proved
an acceptable food.
About 1845 or 1850 the settlers in the Mississippi Valley, in making their
advance westward, brought the cultivated potato into the edge of the habitat
of dccemlineata, where the beetle soon learned to use it as a food. This
extension of the area wherein S. tuberosum was grown into the habitat of
dccetnlincata resulted in the removal of a previously existing barrier to
further eastward dispersion. This barrier was the wide stretch of country
in which no food plant had hitherto been available and into which it could
not go without food. The advent of this new food, however, completely
removed the barrier, and there lay open to the eastward an expanse of
territory where optimum conditions of existence were developed. Into this
= f,
c -
- i
DISPERSAL OF L. DECEMLINEATA. 25
area it began to spread — at first slowly and in small numbers and unnoticed-
then in increasing numbers and more rapidly, making itself felt as an
economic factor in agriculture. Onward it advanced yearly, in increasing
numbers and speed, until in twenty years from the time of its start it had
reached the Atlantic Ocean, a barrier beyond which it has not been able to
successfully pass. Fortunately the history of this advance is known from
extensive records, so that it can be traced in detail from year to year.
CHRONOLOGICAL HISTORY OF THE DISPERSAL OF LEPTINOTARSA DECEMLINEATA.
1859 TO 1904.
Change or extension of habitat is apparently a common phenomenon. I
say apparently common, because we suppose that this has happened and is
going on continuously in the life of all species of animals ; but those that we
can trace and study from year to year are few. It is true that cases of
which we have records enabling us to trace the dispersal may not add exten-
sively to our ideas or speculations as to what the results of dispersal and
change of habitat have been ; they are, however, of prime importance in
judging the results derived from less certain cases and in keeping speculation
within bounds. Hence is presented in this place the history of the dissemi-
nation of decemlincata as it has been worked out from records and data of
diverse kinds.
It has already been shown from whence this beetle came and how it
probably attained its center of dispersal along the eastern slope of the Rocky
Mountains, where, for a time at least, it was in a state of stability. It has
also been shown how a very slight change in the flora of the habitat was
sufficient to upset the stability of the species and to start it upon a prolonged
and extensive dispersal. I shall now present the history of this dispersal
year by year and then consider it for information concerning the effects of
climate, topography, and other factors which have influenced or controlled
this dissemination.
sSj9. — The first recorded appearance of this beetle east of its original
habitat was for the year 1859 by Hazen (1865). According to this author
"the potato-bug {L. dccemlineata) was first discovered in 1859 about 100
miles west of Omaha City, Nebraska, whence they have been marching east-
ward annually." No other record of this beetle for this year is known to
me. I do not know on what evidence Hazen made his statement, but Walsh
and Riley (1868) accepted it as correct. The delaying of the publication
of this record for six years detracts somewhat from its value. However, it
is of interest to note that this first discovery was in the line of the old
emigrant trail from Omaha to Denver, Colorado.
1861. — The first authentic records are found in the annals of this year.
At Atchison, Kansas, it was numerous and destructive. Colman (1862),
26 DISTRIBUTION AND DISPERSION OF LEPTINOTARSA.
editor of the Valley Farmer, published the following letter from Mr. Thomas
Murphy of that place :
Atchison, Kansas, May 22, 1862.
Editor Valley Farmer : . . . . My object in sending them is to see if you person-
ally or through your valuable journal can find out any method or remedy by which I
can get rid of them. I cultivate about 10 acres of land for the purpose of raising potatoes
for my hotel ; it is situated on high prairie land. Last August, soon after a heavy shower
of rain, these bugs suddenly made their appearance in large numbers on the potato vines
They were so numerous that in many instances they would almost cover the whole vine.
It is no exaggeration when I tell you that we have often in a very short time gathered as
many as 2 bushels of them. When cold weather set in they disappeared. Early this
spring I was setting out some apple trees, and away down in the hard yellow clay I found
these bugs apparently dead, but put them in the sun and they immediately came to life.
They have again made their appearance in my garden in large numbers. Last year they
ate up everything green on the potato vines, then commenced on the tomatoes, and so
on, on everything green. Strange to say, they trouble no one else.
Yours, etc., Thomas Murphy.
This occurrence some 200 miles from the point of "first discovery," in
1859, is of interest. There is not the slightest doubt as to the accuracy of
this record, because specimens received by Colman were sent to Walsh for
identification ; it, however, throws grave doubt upon the statement made by
Hazen as to the ' ' first discover}'. ' ' In the sparsely settled condition of
Kansas, Nebraska, and Iowa at that time the beetle would be hardly able to
travel 200 miles in two years, when in the more thickly settled portions of
the country it was able to advance only 50 or 80 miles in a year. It is also
recorded from Gravity, Taylor County, Iowa, by Edgerton (1861), who
states " they made their appearance upon the vines as soon as the potatoes
were out of the ground, and there being a cold, wet spell about that time
they devoured them as fast as they were up. ' '
These two occurrences, so widely separated from the record of 1 859, are very
conclusive evidence that the beetle must have been much farther east in 1859
than Hazen's record places it. The records at Gravity, Iowa, and Atchison,
Kansas, are not those of first introduction, but of first ravages, which always
occur from two to four years after the date of actual introduction. It is
therefore certain that the beetles were in western Iowa and eastern Kansas
in 1858 or 1859, if not at an earlier date. This date places the front of the
advance for the year 1861 somewhat to the east of Gravity, in Iowa, and
Atchison, in Kansas (plate 8), and it is certain that the advancing hordes
had reached the two points mentioned by i860, if not before, and I have
drawn the yearly lines of advance according to this interpretation of the data.
1862. — In this year Emery records the presence of the potato beetle in
Crescent City, Pottawattamie County, Iowa, where it "first appeared" in
1862. Specimens were sent to Walsh for identification, hence the record is
accurate. In this case the first record of appearance is also one of first
DISPERSAL OF I,. DHXEMLINEATA. 27
ravages, as is evident from Emery's statement — "they infest the potato,
egg-plant, tomato, and horse nettle."
The editor of the Wisconsin Farmer (April 13, 1867) states that he found
this beetle abundant in Grant County, Wisconsin, as early as 1862, although
no specimens were seen from that locality in that year by any entomologist.
The publication of this record five years after the date when the insect was
first noticed, and then from " I remember" data, might throw some doubt
upon its validity were it not true that in the near-by State of Iowa all of the
records for the year are of first ravages, so that we are warranted in con-
cluding that the State was almost completely overrun with this insect as
early as 1861, and in that event its presence in southwestern Wisconsin in
1862 is most natural.
1863. — -Additional evidence that this beetle was generally distributed over
Iowa is furnished by the record from New Sharon, Nebraska County, where
its ravages were so severe "that some were discouraged from planting
potatoes " (Fitch, 1863). This bit of evidence from Iowa, indicating exten-
sive damage to the potato crop, further confirms the opinion that the beetle
was in Iowa in i860 or earlier. In any event, the great numbers and the
widespread damage to crops clearly indicate a residence of at least two or
three years, and also add greatly to the probability of the correctness of the
record from Wisconsin in the year 1862.
186 j.. — In the records of this year we find that the beetles had crossed the
Mississippi River into Illinois at several points. In August of the year
1864 Walsh received specimens and the following data from the State
geologist ( Worthen) :
It is committing the most destructive ravages on the potato crop in the vicinity of
Warsaw, Illinois. I find that the same insect in the same year appeared near New Boston,
Illinois. In the autumn I captured two specimens in Rock Island, Illinois, but it has
not yet reached a point lying 30 miles east of us in such numbers as to be noticed by
farmers. I hear that it has swarmed this year at Mount Carroll, Illinois. It does not
appear to have advanced any considerable distance into the State, but it must have
crossed the river into Illinois in 1864-65 at five different points. The northernmost lies over
200 miles from the southernmost. These records, all settiug forth the damage done,
indicate an earlier introduction into Illinois than that given by Walsh, certainly as early
as 1S63, so that in the year 1S64 the eastern limits of the beetle were certainly a consid-
erable distance east of the Mississippi River.
On plate 8 the eastern limits of the beetles for this year have been drawn
on the basis of these conclusions.
1865. — Abundant reports of the potato beetle in Iowa attest its general
distribution in large numbers over the State. So great was the damage
done that the potato crop was reduced to half the usual amount. Timble
(1865) records placing on the table of the New York Farmers' Club " a large
handful of letters, boxes, bottles, and packages from Iowa, all of them con-
28 DISTRIBUTION AND DISPERSION OF LfiPTINOTARSA.
taining a repetition of the same sad story." We learn from this data that
it required four years in which to disperse generally over the State of Iowa,
to become well adapted to its habitat, and to increase sufficiently in numbers
to be at the end of that time a dangerous foe to agriculture. In this year
we have the first records from Calloway and Putnam Counties, Missouri, of
the beginning of a southward advance. Neither record is of first introduc-
tion, because the beetle had been abundant for some years in the neighboring
counties in Iowa. The record in Calloway County, Missouri, which is one
of first ravages, marks the beginning of the southward extension along the
Missouri-Mississippi Valley. It is probable that a large portion of the State
of Missouri north of the Missouri River was quite as generally inhabited by
the beetles as was Iowa.
In Wisconsin, the editor of the Wisconsin Farmer (April 13, 1S67) states
that the beetle was abundant in the St. Croix River Valley in 1865, and
that it existed in small numbers in Marquette County in the same year.
It was also reported by Priest, from Mosinee, Marathon Count}', Wisconsin,
1865, where it did great damage to the potato crop. No specimens from
either locality were seen by entomologists. Walsh ( 1 8b6i>), for reasons which
he does not state, doubts the accuracy of the Mosinee record, but accepts
the statement of the editor of the Wisconsin Farmer, which is based upon
less certain evidence. In Illinois, at Mount Carroll, it was found in large
numbers ("swarms," Shimer, 1865; Walsh, 1865). At Warsaw, Rock
Island, and Alton it was abundant and did much damage to potatoes and
egg-plants. It was reported by Riley near Chicago in small numbers
(Walsh).
For the first time in the history of the advance of this beetle the records of
the year are sufficiently complete to enable one to draw the eastern boundary
with some degree of exactitude. The front of the distribution begins at
the north in the St. Croix River Valley (plate 8), passes in a southeasterly
direction across Wisconsin to Marathon County, and onwards in the same
direction to the shore of Lake Michigan at Chicago. From Chicago the line
runs almost straight in a southwesterly direction through Alton, Illinois, on
the Mississippi River, and across eastern Missouri to the Missouri River,
which it follows to the region of Kansas City, Kansas, and then runs north-
ward across northeastern Kansas. In Wisconsin, Iowa, and Illinois the
distribution of the beetle west of this line was general, while in Missouri
and Kansas it was still local in character.
Walsh ( 1 865) , calling attention to the rate of advance of this insect during
six years, says: "From Omaha, Nebraska, to Rock Island, Illinois, is over
360 miles. If the above statement is correct (Hazen), the insect has traveled
360 miles in six years, at the rate of 60 miles a year. At this rate it will
reach the Atlantic in fourteen years" (*. e. , 1879).
W. L. TOWER
PLATE 8
■
MAP OF EASTERN UNITED STATES AND LOWER CANADA SHOWING YEARLY LINES OF ADVANCES OF
L. DECEMLINEATA DURING ITS EASTWARD MIGRATION, 1859-1904
Important records are indicated by red dots, advance introductions by a uniform neutr;
The prevailing direction ot the wind during summer is shown by a blue arrow
DISPERSAL Of L. DECEMUNfiATA. 20,
1866. — -During this year the beetle spread slowly eastward, but not as far
as in the preceding 3^ears. It seems to have been mainly occupied in com-
pletely overrunning the area east of the Mississippi River, which it had
already covered, so that in this year it is recorded as injurious to a greater
or less extent throughout Illinois and Wisconsin. In the trans-Mississippi
region it does not appear to have been more than ordinarily abundant, nor
to have done any great amount of damage. In the southern counties of
Wisconsin considerable injury was done to the crops (Wisconsin Farmer).
In Illinois it was recorded from Bloomington by Walsh (1866$) , from Wood-
ford County, and from Athens, Menard County, by Dodge. It was abun-
dant in Marshall, Bureau, Champaign, Coles, and Lake Counties (Walsh),
and in Winnebago County (Dodge, 1866). It was found near St. Louis,
Missouri, but had not been found in the southern part of the State (Walsh).
In July it was abundant in Missouri at Hannibal (New York Semi-weekly
Tribune, August io, 1866), and at Florisant, in St. Louis County (Cultivator
and Country Gentleman, vol. xxviii, p. 16, July 5, 1866). The only eastern
advance of any moment recorded was that into western Indiana: "As I
predicted, this insect has now spread into .... western Indiana. Accord-
ing to Dr. Worthen it occurred in this locality even in 1866." It was also
recorded from Hardin County, Kentucky, and from Indiana County, Penn-
sylvania (Dodgei, but no specimens were seen, and the records are clearly
inaccurate.
The eastern front of the recorded area of distribution at the end of this
year follows the same general course as the boundary for 1865, excepting
that it has advanced from 50 to 70 miles. The southward extension does
not appear to have been great, but the lack of records indicates only a freedom
from injury to crops. Throughout most of the infested area the insect was
moderately abundant and destructive, but in the following year there was a
marked decrease in the number of individuals and in the consequent ravages.
186 j. — From Indiana the beetle spread eastward into southwestern Mich-
igan, where it was reported by Gage (1867) at Dowagiac as having appeared
for the first time. It had covered the entire northern and central part of
Illinois and northern and western Indiana, and had spread into southwestern
Michigan (Rep. Ind. Hort. Soc. , Jan., 1868). In the same year Dodge (1867)
records "fearful" ravages in De Kalb and Putnam Counties, Illinois, and
its rapid spread and general destructiveness in Brown County, Wisconsin.
According to Walsh (1867a), it invaded the southern part of Illinois, being
recorded from Union, Marion, and Effingham Counties. Over the infested
area as a whole the beetle caused little damage to crops, and, as a result, the
records of this year are few and show no great change in the area of distri-
bution. The eastern boundary was advanced about 50 miles, excepting on
the south in Illinois, where there was a rapid progress to the lower part of
30 DISTRIBUTION AND DISPERSION OF LEPTINOTARSA.
the State. The recorded dates and occurrences make the spread seem rapid,
while in reality they are merely notices of first damage to crops. In the
following year the insect was more numerous and covered more new ground
to the east.
1868. — Several important and interesting records are found in the annals
of this year. In the north Gilman (1870) states : " Last summer (1868), to
my sure knowledge, it had reached the south shore of Lake Superior and the
northwest corner of Michigan, where it abundantly manifested its presence."
In Minnesota it was found in Houston County (Am. Agr. , vol. 27, p. 248,
1867), and in southwest Michigan it was reported as numerous by Gage,
who kept careful records of its numbers, ravages, habits, and foes. Gage
states that it was abundant at Dowagiac, Cass County, and appeared for the
first time at Decatur, Van Buren Count}', 12 miles northeast of Dowagiac.
We know nothing of its distribution in the southern counties of Indiana,
but in the northeastern it continued to move slowly eastward, reaching nearly
to the Ohio State line. Dodge records the beetle as injurious in Columbia,
Jackson, Green, Lake, Brown, Vernon, Douglas, and Bayfield Counties,
Wisconsin, in Appanoose County, Iowa, and in Ford County, Illinois.
In this year is recorded the first unquestionable introduction in advance of
the general horde. The editor of the Ohio Farmer (July, 1868) writes as fol-
lows: "We have now to record the actual presence of D. 10-lineata
in the southwestern corner of Ohio, a very few specimens of this pest having
been taken within the past week in Hamilton County." The main body of
the beetles was still o,Ter 100 miles away. This introduction seems to have
come from the region of Cairo, Illinois, or St. Louis, Missouri, by means of
the steamers or barges which are constantly passing up and down the Ohio
River. I am sure that the coal barges which go down the river from Pitts-
burgh, Pennsylvania, have carried into the States along the Ohio River, and
perhaps also into the lower Mississippi Valley, many of these beetles from
Pennsylvania and Ohio. In August, 1900, at McKeesport, Penns3'lvania, I
counted 52 of these beetles on one barge of a " tow " loaded and on its way
to New Orleans, Louisiana, and as there were eleven barges in this " tow,"
it is a low estimate to say that altogether it was probably carrying 300 or 400
beetles south into the lower Mississippi Valley. The vicissitudes of such a
journey are great, but it would seem that at least 1 or 2 per cent of these
beetles might stand a fair chance of reaching the lcwer part of the river valley
1869. — The river barges are probably also responsible for the occurrence
of the beetle so far ahead of the general horde, not only in Hamilton County,
Ohio, but also in the region about Louisville, Kentucky, where it was
recorded in the following year (F. J. Key, letter to Riley in 1871). The
beetle was less abundant in this year than during the previous one, doing
but little damage and attracting but little attention. In the following year
DISPERSAL OF L. DECEMUNEATA. 31
the numbers increased, and the data affords many interesting cases of rapid
advance in various directions.
/cS'70. — In the trans- Mississippi region the beetle was numerous, but the
injury done was local and due almost entirely to ignorance or to indolence
on the part of those farmers who had not yet learned how to deal with the
pest. In subsequent years, although the beetle was abundant, few records
occur of ' ' great ravages. ' '
In Michigan the advance was definitely in a northeasterly direction. The
concise reports for the years of the migration of this beetle over the State
enable us to trace with precision the spread from county to county. It had
covered about half the southern peninsula of Michigan, being numerous in
most of the counties and doing considerable damage. The returns of the
county agricultural societies report it as present in Barry, Cass, Eaton,
Genesee, Hillsdale, "Ingham, Lapeer, Shiawassee, and Tuscola Counties (Rep.
State Bd. Agr. Mich., 1870), and in Marion County (Dodge, 1870). In Ohio,
Richmond (Rep. State Bd. Agr. Ohio, 1871, p. 542) records the beetle from
the garden of J. H. Klippert, in Erie County, and Dodge records it from
Van Wert and Mercer Counties. It was also reported by others from Butler,
Champaign, Henry, Hocking, Miami, Morgan, Pickaway, Preble, Putnam,
and Shelby Counties (Ohio Agr. Rep., 1870). In Missouri it was recorded
from the vicinity of Springfield by Holman, and from Greene, Webster,
Phelps, Reynolds, Dent, and Texas Counties by Riley (187 1). Shimer
reports that it was present in the vicinity of St. Paul, Minnesota, and that
in Indiana and Illinois it was abundant in some places and not in others.
South of the Ohio River it was recorded from Covington, Kentucky (Riley).
The front of the recorded distribution for this year is materially different
from that of 1868 or 1869. Beginning in the north near Bigstone Lake,
Minnesota, the line runs eastward to Minneapolis and northeast to the head
of Lake Superior, then southeast and by east across two-thirds of the north-
ern half of Wisconsin, and then south aud southeast to Grand Rapids,
Michigan. From Grand Rapids the line runs easterly, and then northeast-
erly to the head of Saginaw Bay, then turns south to Detroit and follows
the lake shore to Erie County, Ohio, and then southeasterly to Morgan
County. Thence it follows the Ohio River Valley into the south, crosses
Missouri to Springfield, and passes west into Kansas, where we have no data
of its distribution. The distribution shows for this year a remarkable
advance in two directions — in Michigan to the northeast, and to a great
excess over the usual 70 or 80 miles covered annually ; and in Ohio to the
southeast, where the rapidity of advance was so great that the horde along the
Ohio River outstripped all others in the race towards the Atlantic Ocean.
The rapid advance in Michigan, as also that in Ohio, was due to the fact
that the axis of the advance was in the track of well-developed winds, i. e.,
the prevailing westerlies.
2,2 DISTRIBUTION AND DISPERSION OF LEPTINOTARSA.
i8ji. — Over the whole of the infested area the beetle appears to have been
more numerous this year than during the two or three preceding. Riley
(1871) gives the following account of an occurrence in the Lake Superior
region :
Juno Hurlburt, who has been engaged in surveying and prospecting in that part of the
country, . . . found them in immense quantities in a potato field belonging to some
Indians on the Menomonee River, yet this potato patch was on a clearing of about 20
acres, and to his certain knowledge there could not have been another potato patch
within 150 miles.
Riley attributes this occurrence to their being carried down stream to that
point. Dodge records the beetle as injurious in Columbia, Ozaukee. Outa-
gamie, Portage, Richland, Dane, Fond du Dae, Green Dake, Iowa, Juneau,
Kenosha, St. Croix, and Sheboygan Counties, Wisconsin ; in Fillmore,
Carver, Houston. Kandiyohi, Meeker, and Ramsey Counties, Minnesota ; in
Barry, Bay, Cass, Kent, Kalamazoo, Monroe, Newaygo, Ottawa, and Van
Buren Counties, Michigan, and in several counties of Ohio, Indiana, Illinois,
and Iowa. It was also found in Livingston, Manistee, Saginaw, Shiawassee
(crop a failure), and Tuscola Counties, Michigan (Rep. Sta. Bd. Agr. , 187 1),
while south of the lakes it had spread from eastern Ohio into western Penn-
sylvania and New York. In Missouri the southward spread was slight. ' ' It
has not yet reached the southern counties" (Riley 1.
The chief event in the history of this year's spread is the invasion of Canada.
The beetle had been abundant in eastern Michigan and Ohio, especiall}' along
the shore of Lake Erie, but before the end of the year it had crossed the lake
and overrun a large portion of the Province of Ontario. Riley (1 871) describes
its methods of crossing the lake thus : "In the spring the Detroit River was
swarming with the beetles and they were crossing Lake Erie on ships, chips,
staves, and any floating object. ' ' The following agricultural societies reported
its presence in their section of the country : Both well, Brandt (S.), Elgin (E.),
Essex, Huron, Kent, Lamberton, Middlesex (E.), Oxford (N.), Simcoe (N.),
Wentworth (N.), and York (N. and W.). The insect did but little damage
in this year (Ann. Rep. Comm. Agr. and Pub. Work, Prov. of Ont., on Agr.
and Arts, 187 1). They were most numerous on the western frontier between
Sarnia, Lamberton County, on the north, and Amherstburg, Essex County,
on the south, and inland from 20 to 40 miles (Sanders and Reed). They
were found along the north shore of Lake Erie in small numbers as far east
as Toronto, York County.
The front of the area occupied this year is bounded by a line which,
beginning at the western end of Lake Superior, runs southeasterly across the
middle of Michigan to Saginaw7 Bay ; thence along the shore of Lake Huron
to Port Huron, across the St. Clair River to Sarnia, in Lamberton County,
through Stratford, in Perth County, to Toronto, in York County ; thence
DISPERSAL, Ol-* I.. DECEMWNEATA. 33
back along the southern shore of Lake Erie to western New York and Penn-
sylvania ; thence in a southwesterly direction to the Ohio River Valley, and
having followed closely along the valley to Cairo, Illinois, passes westward
across the lower part of Missouri into Kansas (plate S).
rSj2. — Over the whole of the area occupied by this beetle there was a
decrease in numbers this year, which, in Wisconsin, amounted to over
50 per cent. Dodge also states that in Michigan, Ohio, and Illinois the
beetles were not as abundant as in 1871 ; that in Missouri they were
numerous in Phelps, Perry, and Franklin Counties, and were spreading
south, but had not yet reached the southern counties, and that in Kansas
they were found in small and constantly decreasing numbers. Along the
Ohio River they extended east to Harrodsburgh, Mercer County, Kentucky
(J. B. Clark, letter to Riley, 1872). In Ohio they were found in every
county in the State (Klippert, Rep. Sec. Sta. Bd. Agr., 1872), while S. S.
Rathborn reported to Riley that the beetle had appeared in Lancaster County,
Pennsylvania. This latter introduction, so far ahead of the main body of the
beetles, probably took place through the agency of some cargo, and was
important in determining the spread of the beetles to the south and east, and
in enabling them to reach the seacoast a year or two earlier than had been
predicted by Walsh.
In Canada, dementi records the beetle from North Duro, in Peterboro
County, Ontario. The following agricultural societies also report its presence:
Addington, Bothwell, Brandt (S.), Durham (W.), Elgin (E.), Essex, Fon-
tenac, Grey (S.), Hastings (N. and E.), Lamberton, Middlesex (N., W.,
and E.), Niagara, Norfolk (N.), Northumberland (W.), Oxford (N. and S.),
Perth (S.), Peterboro (E.), Simcoe (S.), Victoria (S. ), Welland, and Welling-
ton (N. and S.) (Rep. Comm. Agr. Ont., 1872). It gradually spread eastward
through Ontario, reaching Kingston in September (Rogers, 1872). It was
abundant in most of the western counties of the Province of Ontario, but did
comparatively little damage.
The boundary, which in Wisconsin and Minnesota changed but little from
that of the previous year, crossed Lake Michigan in the region of Manistee-
and passed due east across the State and through the northern part of the
Province of Ontario to Kingston, whence, turning back, it retreated along
the northern shore of Lake Ontario to the Niagara River, crossed over, and
then followed the southern shore of the Lake and inland to the region of
Syracuse, New York ; thence southeast in a tongue-like projection to El-
mira, New York ; then back along the shore of Lake Erie. Here there is
another tongue- like projection into western Pennsylvania, and then the line
turns back and westward along the Ohio River Valley, with the beginning
of tongue-like projections into West Virginia and Kentucky.
3— t
34 DISTRIBUTION" AND DISPERSION OF LEPTINOTARSA.
The feature of most interest in this year is the rapid spread over Ontario
and the advance down the St. Lawrence River Valley. Of interest also is
the rapid pushing out of projections along the natural highways which
penetrate the Appalachian highland and afford easy passages to the Atlantic
coast area.
1873. — In this year, over most of the area in which the beetle was now a
resident, it was found in only moderate numbers, although it was abundant
in some counties and did more or less injury to crops. In Minnesota, Wis-
consin, Michigan, Iowa, Illinois, Indiana, and western Ohio almost no
damage was done, and consequently it passed almost unnoticed. In north-
eastern Ohio it was a pest, one newspaper describing their abundance in the
following words : ' ' Clouds of Colorado potato-beetles .... The scene
suggested to Bible readers locusts in ancient days." Riley remarks that
he has good authority for the statement that at best only a few dozen could
be seen at one time by a single person. In this year the insect covered the
whole Province of Ontario (C. J. S. Bethune), fifty-nine agricultural societies
reporting its presence, but in most instances only a small amount of damage
was done.
During the two years following the advance introduction into Lancaster
County, Pennsylvania, which was first noticed in 1872, the beetle advanced
eastward and southward with great rapidity. It spread east into Salem
County, New Jersey, and into Caroline, Dorchester, Frederick, Queen Anne,
Cecil, Baltimore, and Washington Counties, Maryland (Glover). It was
found in the District of Columbia, where living specimens were taken to the
United States Department of Agriculture for identification. It was also
found in Prince William and Page Counties, Virginia, and in Butler, Clear-
field, Franklin, Northampton, Perry, Bedford, Huntingdon, Cambria,
Juniata, Armstrong, and Indiana Counties, Pennsylvania (Glover). In
New York it was reported from Erie County, and in West Virginia from
Monroe, Preston, Hardy, Jefferson, Marion, Brooke, Tyler, Fayette, and
Pleasants Counties (ibid.). No specimens were seen from these West Vir-
ginia localities. It was also recorded from Nicholas, Anderson, Clarke,
Lexington, and Fayette Counties, in Kentucky, and from Fentress County,
in Tennessee.
The boundary of the invasion in Minnesota and Wisconsin does not change
in this year ; in Michigan it has moved northward, but is parallel with that
of the preceding year. From Michigan the line crosses Lake Huron to
Georgian Bay, passes along the northern border of the Province of Ontario
and the St. Lawrence River Valley to near Montreal, where it turns south-
ward in a tongue along the Lake Champlain Valley, then westward along
the southern side of the St. Lawrence Valley to western New York. In
New York, Pennsylvania, and West Virginia the tongue-like projections
DISPERSAL OF L. DECEMUNEATA. 35
noticed in the preceding years have now almost completely penetrated the
Appalachian highland, and in Pennsylvania and Maryland they have nearly
united with the Lancaster colony.
i8j-i.. — From the region west of New York and Pennsylvania but few
records are found in this year's reports, although the beetle was numerous
and destructive in almost all of the counties of Ontario. The center of
interest was along the Atlantic coast, where in many places it was abundant
and did much damage. It was found in Quebec in small numbers and in
western Vermont (C. J. S. Bethune). It was also reported from Williams-
town. Massachusetts (J. S. Kingsley), and from western Connecticut (New
York Weekly Tribune, August 26, 1874). In New York it was recorded
from Allegany, Chautauqua, Delaware, Erie, Madison, Tioga, Wayne, and
Wyoming Counties ; in New Jersey from Burlington, Gloucester, Monmouth,
and Salem Counties ; in Delaware from Kent County ; in Maryland from
Allegany, Baltimore, Caroline, Cecil, Carroll, Dorchester, Frederick, Har-
ford, Montgomery, Prince George, and Queen Anne Counties ; and in
Virginia from Culpeper, Fauquier, Greene, Highland, Page, and Prince
William Counties (Mo. Rep. U. S. Dept. Agr., 1874). Almost every county
in Pennsylvania was invaded by the beetles, and at many places they were
numerous and did considerable damage. At Germautown they swarmed
(Riley, 1874), as also at Cape May, New Jersey (New York Weekly Tribune,
July 22, 1874).
1875. — At the beginning of this year the beetle was distributed along the
seacoast from New York to Chesapeake Bay, and by the end it had overrun
most of the remaining territory of the coast States. It reached Boston,
Massachusetts, in the autumn. It penetrated farther into Vermont and was
reported from New Hampshire and Maine. It was found in the central
portion of the latter State, which led Fernald to believe that it was taken
there by rail. It was reported from Skowhegan, Somerset County, from
Saco, in York County, and Temple, in Franklin County, Maine 'Packard), and
it seems to have penetrated as far as the Kennebec River (C. H. Fernald).
In New York, Pennsylvania, New Jersey, Delaware, and Maryland it was
generally distributed ; but in Virginia, West Virginia, and Kentucky it had
not extended any farther south.
1876. — Eastern Connecticut, Rhode Island, and Cape Cod do not seem to
have been occupied until 1876. The insects were abundant and did consid-
erable damage in many of the counties of the New England States. It is
related that they were washed ashore in many places in such numbers as to
poison the air with the "noxious vapors" arising from their decaying
bodies. The captain of a New Loudon vessel relates ' ' that while at sea
(Long Island Sound ) they boarded him in such numbers that the hatches had
to be closed." At Cape May, Long Branch, Rockaway, and Newport they
proved a great nuisance to the pleasure seekers, being crushed and killed in
36 DISTRIBUTION AND DISPERSION OF I.EPTINOTARSA.
large numbers by the continual promenading along the beach ; while in
New York they are reported to have stopped a train upon the New York
Central Railroad ' New York Times). They were abundant everywhere
and by the end of the year had overrun the entire northern and eastern
part of the United States, excepting northern Maine. In Canada they were
still spreading to the northeast along the St. Lawrence River Valley. In
Virginia the boundary was pushed south into the highlands, but remained
stationary at the seacoast ; while in Kentucky the beetle was recorded from
the southern counties of the State. In the Mississippi Valley it was slowly
spreading in the south, but how fast or how far it is impossible to determine
because of the lack of data from that region.
From the time the beetle had completely overrun the Northern States
fewer and fewer records of it are found as the years go by, and such as are
found are limited to notes and data of no value. The State reports and
agricultural papers contain no more notices of its spread, and as a result
the history of this beetle in the northeastern portion of the St. Lawrence
Valley or the migration into the Southern States can not be followed in any
detail. Almost all of the data of the southern spread for the last twenty
years would be wanting were it not for the careful collecting of records
concerning it by the Division of Entomology of the United States Depart-
ment of Agriculture.
i8jj. — In this year the beetle was found in Knoxville, Tennessee (Nichol-
son, Can. Ent., vol. ix, p. 174, 1877). This author believes that it was
brought in with the seed from the north, but it is highly improbable that
this method of spreading the beetle ever occurs. Seed potatoes are always
sorted at frequent intervals, and I know from actual observation and experi-
ment that the beetles do not stay in potato sacks or barrels, but crawl out at
the first opportunity, and, moreover, it is not possible for them to pass the
winter in such situations, as death would certainly intervene.
i8yS. — The continued spread of the beetle to the northeast resulted, in
1878, in the almost complete overrunning of the Dominion of Canada. It
was found at St. Johns, New Brunswick, and had spread through the
Maritime Provinces (Saunders).
18J9. — In this year, in the northwest, it was recorded from Manitoba
(Comstock, Can. Ent., vol. xi, p. 196, 1879), where it seems to have become
well established ; but whether it came from the east or from the original
area of distribution in the west we have no means of determining. It was
also reported in May from Lynchburg, Virginia (Cor. U. S. Dept. of Agr.).
1880. — In this year the beetle was recorded from South Carolina (Riley,
Can. Ent., vol. xn, p. 173, 1880), and from Manitoulin Island, Lake Huron,
where it had been for a number of years, but had not thrived or made any
headway (Saunders, 1880). In the correspondence of the United States
DISPERSAL OF J,. DECS M UN RATA. 37
Department of Agriculture it is reported from Jefferson, Marion County,
Texas, in April, 1880, and St. Agatha, Manitoba, in August.
//. — In 1884 the beetle was recorded in Manitoba from Manchester and
Dufferin Counties and from Portage la Prairie, which would indicate that
it was spreading rapidly in the northwest. In the St. Lawrence Valley,
especially near Quebec, its numbers were said to have shown a marked
decrease during the four years between 1880 and 1884 (Can. Ent., vol. xvi,
p. 207, 1884).
1885. — Although from the south the records are meager, the spread in
that direction has shown many features of interest. From 1874 or 1875,
when the beetle had reached the middle of Virginia along the Allegheny
highlands, until this year, 1885, it had spread very slowly southward, and
even more slowly along the coast than elsewhere ; but in this year, May 10,
it was found at Savannah, Georgia, on Wilmington Island (Cor. Div. Ent.,
U. S. Dept. Agr. ). It was certainly imported into this place, probably in
the hold of some vessel, where it had become imprisoned while the vessel
lay at her dock in some northern port. Although it was fairly abundant
there this year, no further records of it are available until March, 1887, nor
from that date until five years later.
1888. — For the Mississippi River Valley notes of occurrences are even
more rare than for the Atlantic slope. F. M. Webster records the beetle in
this year from Jackson, Mississippi, on hearsay evidence, although the
record is probably valid. It was also found in Tyler, Smith County, Texas
(Cor. Div. Ent., U. S. Dept. Agr.).
i88g, — In 1899 Phares sent specimens to the Division of Entomology,
United States Department of Agriculture (Insect Life, vol. 11, p. 22), from
Madison Station, Madison County, Mississippi. At that time this was the
southernmost point from which the beetle had been recorded.
i8go. — Earle reported it from the middle of Mississippi in this year, but
said that it had not reached the southern part of the State (Insect Life, vol.
in, p. 84, 1890).
1891. — In the extreme southwest it was found in Las Cruces, Donna Ana
County, New Mexico (Townsend, Insect Life, vol. iv, p. 26, 1891), where
it was not uncommon. In the correspondence of the Division of Entomology,
United States Department of Agriculture, it is recorded from Birmingham,
Alabama, in July, and from Fayetteville, Cumberland County, North Caro-
lina. Concerning the latter occurrence the following facts are on file : Hon.
William G. Le Due, in reply to questions from L. O. Howard, states that
the beetle first appeared there in abundance in 1891, so that he places 1890
as the date of introduction. They were plentiful therein 1892, 1893, aQd
1894, but less so in 1895. On the authority of Colonel Broadfoot, " Potato-
bugs appeared here four or five years ago" (July 1, 1895).
38 DISTRIBUTION AND DISPERSION OF LEPTINOTARSA.
1892. — In the northwestern part of Alabama the insect was reported as
abundant (Insect Life, vol. v, p. 356, 1892). It was also recorded from
Yemassee, Hampton County, South Carolina, in May, and from " Hay Field
Farm," Charleston Neck (Cor. Div. Ent., U. S. Dept. Agr.). The following
information concerning this latter record was placed on file by H. M. Simons :
The beetle, in the opinion of Mr. Simons, first appeared there in 1892, on a
nearby farm, where it was abundant on guinea squash. The bugs were
hand-picked, but appeared the next year (1893) in a field across the road,
where they covered an area of about 2 acres. A few bugs were found on
his own farm, but they came into his fields late in the season. In 1894
they appeared on March 6, on the potatoes, and gradually increased in num-
bers. It is evident from this account and from that of the appearance at
Fayetteville, North Carolina, that the insect is far less active there as a
migrant than in the States farther north.
1893. — Hubbard (Insect Life, vol. vi, p. 282, 1893) reported the occur-
rence of L. dece?)ili?ieaia at Fort Assinniboine, Montana, where it had not
come in contact with the cultivated potato. This record represents probably
the northern limit of its original habitat. It was also recorded from Jack-
sonville, Calhoun County, Alabama (Cor. Div. Ent., U. S. Dept. Agr.).
1895. — In this year it was again recorded from Charleston, South Caro-
lina, where it seems to have been introduced a third time (ibid.}.
1896. — The only record of this year that is of importance is that at Seneca,
Oconee County, South Carolina (ibid.).
1897. — More records for the beetle are known this year from the south,
and they mark a considerable extension of its bounds. It is reported in May
from Victoria, Marshall County, and Oxford, Lafayette County, Mississippi ;
from Dye, Montague County, Texas (ibid.), and from fifteen counties in the
northern half of Mississippi (Weed, 1897).
1898. — Recorded from Lenoir, Caldwell County, North Carolina, in May.
1899. — Recorded from Sheridan and Silver City, Grant County, Arkansas,
in August (ibid.).
1900. — Prof. A. L. Quaintance informs me that in this year the beetle was
found in large numbers at Experiment, Georgia, and that it has entirely
replaced or crowded out L. juncta, which, in 1899, was one of the most
abundant insects in that locality.
1901 io 1905. — Since the year 1900 no records of any value concerning the
farther advance of this beetle have been found, and, as it now covers nearly
the entire eastern part of the United States, further records of this nature are
not to be expected.
The dissemination of this insect over the eastern United States, now com-
pleted, has occupied a period of fifty years, and of these we have accurate
HIGHWAYS OF DISPERSAL. 39
records for the last forty-six — i. c, 1859 to 1905. During this period it has
become a common resident in all of the country wherein it has been at all able
to gain a foothold. It has gone from habitat to habitat, from one climatic or
topographic area to another, without undergoing any marked change; it
everywhere has shown itself to be a plastic form, well fitted to undertake the
extensive migrations that it has so successfully accomplished.
Occurrence in Europe.
During the years in which L. decemlineata was spreading eastward over
the United States and lower Canada, everywhere proving a serious pest to the
potato crop, the people of western Europe were intently watching its prog-
ress and dreading the time when it might be transported to their lands, where,
in the thickly populated countries, its presence would prove even more disas-
trous than in the United States and Canada. So great was the apprehension
felt that many German States published popular information about this pest,
and German schoolmasters were required to instruct the school children con-
cerning it. The French government likewise published an elaborate bulletin
describing this beetle and its ravages.
As anticipated, it was transported to Europe in 1875 or 1876, probably
in the holds of vessels, appearing in England, Sweden, and Germany ; but
such prompt and effectual measures were taken to suppress it that it was
soon exterminated. Since then, although frequently introduced, as far as
I can learn, it has not been able to gain a foothold in any continental
country. It has been introduced at various times into England and Ireland,
and it was reported in 1900 to have become established in the Pyrenees
Mountains. None of these introductions, however, have been able to spread
over more than a few acres, and have all been promptly exterminated.
The dissemination of this beetle has ceased for the time being. The
Atlantic Ocean has proven a hard barrier to cross, but it has crossed it in
the past and will do so again and again, until finally it will probably become
established as a member of the fauna of Europe. When this happens, it will
spread, as it did in this country, until it is found in all the countries of
Europe in which it is possible for it to live.
The Highways of Dispersal and Rate of Movement.
This dispersal, covering a large area of country of diverse climatic and
topographic conditions, affords an opportunity in a specific case for examin-
ing closely some phases of migration.
There is no reason to suppose that the dispersion of organisms from their
center of origin is ever uniform in all directions. We know, in fact, that it
is more pronounced in some directions than in others, owing to various
directive factors — winds, temperature, and topographic features — such as
those natural highways which afford lines of least resistance to migration.
40 DISTRIBUTION AND DISPERSION OF LEPTINOTARSA.
The center of dispersal for L. decern faieata was on the eastern slope of the
Rocky Mountain Plateau in Colorado, Nebraska, and Kansas, between the
altitudes of 7,000 and 3,000 feet. The dispersal was started by the intro-
duction of a single species of plant into the habitat of decemlineala, which,
taken as a food, furnished a base of supplies in the country to the east.
This opportunity was seized, and decemlincata advanced steadily eastward,
increasing its numbers and extending its habitat. The line of advancement
from its original habitat was at first narrow, being dependent upon the
earliest westward extension of civilization along the Platte River Valley and
through northern Kansas.
Having passed its original bounds, the whole eastern half of North America
lay open to it. Which way should it go? What factors should determine
the direction of its advance ? Two highways were already developed — one
leading directly in the route of emigrant trains, stage lines, and settlers'
farms, along the North Platte River to eastern Nebraska and Iowa ; another,
less important, along the route followed by certain stage lines from Atchison,
Kansas, to Denver, Colorado. That the beetle probably followed both of
these lines of travel and settlement is indicated by its simultaneous appear-
ance at both termini. These lines of dissemination were dependent upon
the fact that it was along the routes of travel that settlements had been made
and a series of bases of food supply had been created.
However, about i860, after the}' had crossed the Missouri River, a more
thickly settled portion of the country was encountered, with wider distribu-
tion of their new food plant, which removed all further necessity of following
paths of human travel. The beetles were now free to respond naturally to
the control of the climatic and topographic features of the country into
which they were migrating. Throughout the remainder of this dissemina-
tion we find them following well-defined trends of movement, frequently
along paths or natural highways where human movements are also active.
They have not, however, followed these because of the presence of man ;
rather, both have followed the same highway because topographic conditions
and climatic causes have determined for all organisms the highways of
movement.
The chief trends of movement found are as follows : (1) Northward, up
the Mississippi River Valley ; (2) northeastward, along the Great Lakes-St.
Lawrence River Valley ; (3) southward, down the Lake Champlain and
Hudson River valleys ; (4) southeast and east, along the Missouri and Ohio
River valle3?s ; (5) seven distinct movements along natural highways from
the western region into and through the Appalachian highlands to the
Atlantic coast ; (6) northward and southward along the Atlantic coast, and
(7) southward along the Mississippi River Valley (plate 9). As the result
of transportation by man, the following centers of dissemination were estab-
W. L. TOWER
Is
< H
O uJ
< CQ
HIGHWAYS OF DISPERSAL. 4 1
lished : Hamilton County, Ohio, 1868 ; Lancaster County, Pennsylvania,
1872; and Charleston, South Carolina (1892), and Savannah, Georgia (1885)
(plate 8).
None of these, however, changed in any material way the general move-
ment. The seven trends of migration found all follow natural highways of
dispersal or paths of least resistance. These highways are absolutely inde-
pendent of organisms, but are the results of geologic and geographic forces
which are at work upon the continent. In the main these highways are
associated with river valley developments, and also, to some extent, with
coastal plains. The existence of these natural highways has been recognized
by many writers, but Webster (1898, 1903), who has studied the trends of
insect migration, does not seem to have correlated them in any degree with
natural topographic highways.
L. deccmlineata, on its escape from its habitat, entered an area devoid of
natural barriers, across which it was able to spread in any direction, with
the results shown on plate 8, where the yearly lines of advance show a
nearly even expansion in all directions. What irregularities there were
were produced by winds and moisture, which we shall discuss presently.
As soon as it reached the Mississippi Valley, the first great natural highway
that was encountered, there was a marked change in its behavior. Previous
to 1863-64 it had crossed wide stretches of country each year, but after
entering the Mississippi Valley its advance was checked for two years. In
this period we find it extending northward with great rapidity, reaching
nearly to the headwaters of the river in 1865, and also southward, although
much more slowly, down the valley. Thus this natural highway served
for movement in both directions at the same time, but much more rapidly
in one direction than in the other. Webster (1898, 1903) has thought of
the Mississippi Valley as a highway only for northward migration of southern
forms, but Adams (1902) and others regard it as a pathway for both north
and south migration, and it is here shown that movement in both directions
occurs. The difference in the rate of movement in the two directions is due
largely to winds, temperature, and moisture, climatic factors whose directive
influence I shall discuss later.
On passing eastward out of the Mississippi Valley the beetle again spread
out, this time in a broad wave over Illinois, Wisconsin, and western Indiana.
This wave extended from near the head of the Mississippi Valley southward
to the junction of the Mississippi and Ohio rivers and eastward to the
southern end of Lake Michigan. Over this territory natural barriers of any
size or importance were absent. On passing the southern end of Lake
Michigan it entered the great natural highway of the Great Lakes-St.
Lawrence system. Along this it now spread with great rapidity in the years
1866 to 1880 northeastward across the State of Michigan and through
42 DISTRIBUTION AND DISPERSION OF LEPTINOTARSA.
southern Ontario, on the northern shore of the lakes, covering great stretches
of the country each year (plate 8). In this rapid and direct advance the
advantages of the natural highway were not entirely responsible for the
accelerated rate of movement. The strong southwesterly winds which sweep
down this basin aided very materially in this dissemination, especially in
the early years. On the southern side of the lakes the St. Lawrence Valle}-
advance was less rapid, the difference in the rate of movement being due to
the fact that the southern side is broad, and there was a large area to cover,
whereas upon the northern side the habitable area was narrow, and, as a
result, there was greater concentration of the animals along the front, and
therefore a greater chance for marked advance each year.
The southern column, working eastward and southward through Indiana,
Ohio, and adjacent States, soon encountered the Appalachian highlands.
There resulted a turning of the column northeastward, with a concentration
in the region of the lakes, and the formation of several independent columns
which moved eastward along natural highways across the Appalachian
Mountains. That portion of the southern column which had concentrated
near the lakes, moving rapidly eastward along their southern shores and
down the St. Lawrence Valley, finally turned south along the Lake Cham-
plain and Hudson River valleys, penetrated the Appalachian barrier in the
north, and gained entrance to eastern New York and New England. The
most prominent of the minor columns which were crossing this barrier moved
along the valleys of the Mohawk, Delaware, Susquehanna, and Potomac
rivers, while still smaller lines of advance were developed in the valleys of
the Cheat, Kanawha, Licking, Kentucky, Cumberland, and Tennessee rivers.
To the north the barrier was penetrated, first by the Delaware, Susquehanna,
and Potomac columns, and later by the Mohawk (plate 9).
The establishment of the colony at Lancaster, Pennsylvania, in 187 1 or
1872, and the arrival of the beetles from the west over the Delaware, Susque-
hanna, and Potomac highways produced a strong center of dispersal from
which movement along the Atlantic coast began. The northward dissemi-
nation along the Atlantic slope during the years 1872 to 1876 was rapid, it
being accelerated by climatic factors, principally winds, which retarded the
movement to the south. There is evidence also of a strong local control of
the direction and rapidity of migration by small features of the topography,
as, for example, the movement up the Connecticut River Valley and in the
river valleys in Maine.
In the south the available habitat of the beetle was limited, and the culti-
vation of its food plant by no means as general as to the north and west ;
consequently, advance was generally quite slow. Over the Piedmont belt
the retardation of the southward movement between the years 1872 and
1900 was less marked, and here the beetle became fairly common and gener-
HIGHWAYS OF DISPERSAL. 43
ally distributed ; whereas along the Atlantic Coastal Plain its movements
were irregular and uncertain, and although its advance still continues, com-
plete occupation of the region has not thus far been effected.
In 1872 and 1873 there branched off from the general horde advancing
eastward through Illinois, Indiana, and Ohio several columns which moved
southward along the river valleys of the Allegheny Plateau, and in the years
following showed a well-marked trend of migration. About 1891 or 1892
these columns, having worked slowly down the western side of the Appa-
lachian barrier and extended southward beyond it, united with the southward
trend of migration along the Piedmont belt, thus completely encircling the
Appalachian Mountains. These two united columns have since 1894 or 1895
been working slowly southward, but have not as yet been able to reach the
Gulf coast.
The topographic area comprised within the Ozark Plateau and the Ouachita
Mountains, situated between the Great Plains on the west, the prairies on the
east, and the Coastal Plains area on the south, is one well adapted to retard
migration, if not to serve as an effectual barrier. Unfortunately, almost
nothing is known concerning the dissemination of L. decemlineata in this
region. We know that from middle Missouri it extended very slowly to the
southern counties (Riley), but this may well have been due to other than
topographic causes. Over much of this area it is not common, and it is
probable that in many places it does not exist. A possible trend is indicated
to the east of the Ozark Plateau, which passes the eastern end of the Ouachita
Mountains through Arkansas, into northeastern Texas, where decemlineata
is found. This trend arose from the Mississippi Valley movement and may
be considered to be a part of it.
It may be of value at this point to compare the dissemination of L. decem-
lineata with that of other forms that have spread over the same regions and
have been influenced by the same natural causes. The data of the dissemi-
nation of Pieris rapce L. has been gathered and published by Scudder (1S87),
and although he does not place any stress upon the relation of the spread of
this insect to topographic and climatic conditions, a study of the published
data and lines of yearly advance shows the same general control of dissemi-
nation as that found in L. decemlineata. It is of interest to compare the
spread of these two insects, one a native of the Great Plains, which began its
advance eastward at about the same time that the other, a native of Europe,
was introduced into the east and began its march westward. Although the
advance of Pieris rapce has been against the action of powerful climatic factors,
the lines of its most rapid and successful advance have been along such natural
highways as those afforded by the St. Lawrence-Great Lakes system. Other
species of insects have responded in like manner to these same influences.
Webster, in his study of the distribution of the chinch bug, has come
to the conclusion that it entered the United States from the south through
44 DISTRIBUTION AND DISPERSION OF LF.PTINOTARSA.
natural gateways, and has followed in its advance many of the highways
that have been mentioned in the preceding pages.
In following the history of the spread of L. decemlineata it is evident that
when a natural highway is encountered the beetles respond at once to the
change and the control of their direction of movement. They either go or
do not go along a certain path, and the choice is made at once and not after
generations of indecision and futile trials. It is clear that in this dispersion
the routes followed were natural ones, and were almost entirely independent
of man, notwithstanding the frequent statements to the contrary. Human
advance nearly always follows natural highways, but this in nowise alters
or invalidates the conclusions regarding the part played by these topographic
features in the dissemination of L. decemlineata.
Other Factors Which Have Controlled the Direction of Movement.
We shall now consider some of the habits of L. decemlineata and the various
climatic factors of its environment which have been of importance in directing
the lines of its most rapid advance and aiding in its general dispersal.
The life cycle and habits of this beetle are such as to aid greatly in any
migration in which it may be engaged. All through its range it is double-
brooded in the summer, the second brood passing the winter as an imago.
In Minnesota, Lugger (1895) records three broods in one year, which I believe
to be incorrect. All through its habitat the beetles appear on the first warm
days of spring, flying about or crawling along the sunny sides of fences.
They are thus active for some time before their food plants are up and egg-
laying can begin, and during this period they undoubtedly do a considerable
amount of traveling. During the eastward migration this period between
the emergence from the ground and the time for egg-laying to begin was
occupied in extending its bounds. When the number of beetles which hiber-
nated was large, so that the mortality resulting from the vicissitudes of the
winter would leave a goodly number to emerge alive in the following spring,
a considerable and important advance would result. The best example of
this occurred in the spring of 1871. The beetle had been numerous and
destructive in Michigan and Ohio in 1870, when there is recorded a rapid
spread northeast into the Province of Ontario. There was a similar condi-
tion in 1872 and 1873 in the Lancaster County, Pennsylvania, colony, when
there followed a rapid spread down the shores of Chesapeake Bay.
The spring dissemination, however, is not of as great importance as that
which comes later in the season. As soon as the food plants appear above
the ground the beetles cease their wandering habits, and late in May begin
to lay their eggs. The first brood reaches maturity in about 35 days, or
about the first of July, and after feeding for a few days they pair and deposit
the eggs for the second brood, which reaches maturity about the middle of
August. The last brood does not pair until the following spring, although
EFFECTS OF HABITS IN DISPERSAL. 45
they remain active for about three weeks, or even longer, in the autumn,
before seeking a place in which to hibernate. It was in this interval
between the emergence of the second brood and the beginning of its hiberna-
tion that most of the area of advance was covered. During this time both
sexes are active ; the females are not laden with ripe ova, and are therefore
able to fly with greater ease than in the following spring, or than are the
females of the first summer brood. Riley notes that the fall brood has a
tendency to migrate. He records seeing them ' ' swarming in the air or
traveling on foot," and he believes that " most of the advance ground was
covered in the latter part of the growing season." Under the best of condi-
tions this insect is a poor flyer without outside aid ; it can be driven some
distance, but when undisturbed it never flies far without settling down to rest.
The existence of this active period was of great importance in the advance
eastward. The females, when laden with ripe ova, fly with difficulty, and
only small distances are covered before the eggs are deposited. Hence, if
the second brood had paired and deposited its eggs within a few days after
emergence, as did the first, it is evident that there would have been but little
flying about, and as a result its advance would have been far less rapid and
the history of its dissemination quite different. If we compare L. decem-
lineata with Ocneria dispar in its spread over Massachusetts, the importance
of a period of free and easy movement in the life cycle as an aid in rapid
dissemination is at once apparent. In O. dispar the spread has been exceed-
ingly slow, owing in part to the inability of the females to fly, but as much
or more to the short time the female imago lives. The female is fertilized
almost immediately upon hatching from the pupa, deposits her eggs within
a few feet of the spot where the larval and pupal stages have been passed,
and then dies. If the female moved about for three or four weeks and
then hibernated, and did not deposit her eggs until the next spring, all
other conditions remaining the same as at present, the spread of this insect
would have been much more rapid and its partial subjection a much greater
task than the State has found it. The existence and duration of this period
of free and active movement is a factor of prime importance in determining
the rate of dissemination of any insect, and must, I think, in the case of
introduced insect pests, receive due consideration in the discussion of the
probability of suppression or the possibility of extermination.
In the eastward spread of the beetle it early became evident that the
advance was more rapid in some directions than in others, and this accelera-
tion and retardation have been found throughout its whole history. I have
shown that there was little transportation by human agencies, and further
that the insect is a poor flyer ; hence the covering of the wide stretches of
country that were overrun year by year required some outside aid. This
aid I have found to come largely from winds, as is demonstrated by the
following experiments : On a clear, calm day in August or early September
46 DISTRIBUTION AND DISPERSION OF LERTINOTARSA.
put mature, active beetles in a shallow tray mounted on a tripod or other
support. In an hour or so all of the beetles will have crawled out and the
larger part will have taken flight from the edges of the tray ; but the
directions of these flights will be as many as there are beetles. Under such
conditions there is no one direction in which they show a tendency to travel.
In fact there is not, as has been stated by Riley, Walsh, and others, any
instinct or tendency that causes them to migrate in a certain direction.
This experiment has been tried many times in Massachusetts, on Long
Island, New York, in the Ohio River Valley, and on the shores of Lake
Michigan, and always with the same result as that given above. But let
the same experiment be tried on a day when a steady breeze is blowing 6 to
8 miles an hour, but with other conditions similar, and a strikingly different
result is obtained. The beetles all start out from the tray as before in as
many different directions as there are beetles, but when the direction is
against or even across the wind they soon become tired with their exertions
and drift with the wind, often for a considerable distance. These experi-
ments and close observation of the beetles as they start of their own accord
from their food plant have uniformly given the same result — that while the
beetle is able to fly some distance against a strong wind, it almost invariably
turns and drifts back with the wind for a much greater distance than it has
flown against it, the actual or net movement or migration being in the
direction of the wind, i. e., with the wind.
The influence of the wind is well shown along the Atlantic coast or the
shores of Lake Michigan when there is an " offshore " breeze. The beetles
are then carried out over the water and, becoming exhausted, fall to the
surface, where they float, and not infrequently are washed ashore in such
numbers that they form a small windrow along the beach. In 1876, when
they were abundant along the Atlantic coast, they were blown out to sea in
large numbers, and, being washed back upon the beaches, were so numerous
that they were obnoxious to the pleasure seekers there.
The ease with which the wind determines the direction of flight of this
beetle is well known on the shores of Massachusetts, where an on-shore
breeze is often well developed for days at a time. Then the beetles are
found flying away from the shore or crawling about on objects to the leeward
of the potato patches from which they started, and if a nearby beach be
searched very few living or freshly killed beetles will be found. But let
this sea breeze be overpowered by the prevailing westerlies, as it often is,
and the beetles are all found to be flying with the wind or driven by it
toward the sea, and if the beach be searched a day or two later there will
be found numbers of freshly killed or living beetles.
On plate 8 is plotted by means of arrows the prevailing wind direction
for the months from May to September, inclusive. It is evident that the
lines of most rapid advance are correlated with the prevailing wind direction
EFFECTS OF WINDS IN DISPERSAL. 4/
for the grooving season. This acceleration by the wind is especially note-
worthy when the wind happens to correspond in direction with a natural
highway, as in the Great Lakes-St. Lawrence Valley. In this region the
lines of distribution from 1866 to 1872 became wider apart, until in 1871,
187;, and 1873 great stretches of country were overrun each year. While
other factors helped to produce this result, such as the beetles being carried
on the waters of the lakes, the major part of the advance ground covered
each year was due to the aid given by the wind to the natural migratory
power of the beetle. Immediately after crossing the Missouri River and
entering upon the prairie country the influence of winds upon the dissemi-
nation of L. decemlineata was clearly shown in the rapid advance to the
northeast into Wisconsin. This advance was, as shown on plate 8, in
general coincident with the direction of the wind. Topographic barriers
and natural highways were absent, the wind serving almost, if not entirely,
as the controlling factor in that region. In the Mississippi Valley, also, we
find that the rapid northward advance was in part due to the existence of a
natural highway and in part to the influence of the wind.
It is not always that winds have been favorable to the migration of this
beetle. Whereas in the above cases aid of a substantial kind was given, in
others it has acted as a strong retarding agent, the best examples of this
being found in the history of the southward advance. Tiie migration down
the Mississippi Valley was, as we have seen, slow, owing almost entirely to
the fact that there is a strong indraft of air northward during the summer
months. How effectually this wind served to retard the migration may be
indicated by the fact that it required about 32 years (1867-1900) to cover
the territory from Cairo, Illinois, to New Orleans. The slowness of this
advance appears still more striking if we compare it with the migration
from the southern end of Lake Michigan to the Atlantic coast, a distance
about equal, which required seven years. In other words, fourteen genera-
tions were required for the beetle to advance from Lake Michigan to the
Atlantic coast, and more than sixty to cover the distance between Cairo and
the Gulf of Mexico. Likewise, the dissemination over the Atlantic Coastal
Plain and the Piedmont belt has been retarded to a considerable extent by
adverse winds.
Cyclonic disturbances often transport birds and strong-flying insects long
distances from their native habitats, and these, coming to rest upon a distant
land, might under favorable conditions be able to start a colony. But while
numerous cases occur every year of the flight for hundreds of miles of strong-
winged Lepidoptera and Orthoptera, when aided and controlled by the
wind, we do not yet know of a single instance in which this flight has
resulted in a permanent dispersion. For example, Erebus odora, a strong-
winged noctuid, which breeds in southern Mexico and southward, is found
in numbers every year in the United States, and to some extent in Canada.
4— T
4S DISTRIBUTION AND DISPERSION OF LEPTINOTARSA.
Although more or less worn specimens showing the effects of long flight are
found year after year in the same localities, it does not breed in these places,
and hence this dissemination is futile, as there can be by this means no
extension of its bounds. These isolated and sporadic cases of transportation,
always accompanying the development of the tropical cyclones, which in
the latter part of the summer sweep up through Central America, Mexico,
and the United States, have only a remote bearing upon the true dissemina-
tion of animals. In the dissemination of L. decemlineata we find a true
example of the aid afforded by winds in the dispersion of animals, and not
one of the sporadic cases usually mentioned as such, nor one in which the
conditions for starting a new colony are too difficult to be realized, excepting
in rare instances.
Other factors also have aided in the dissemination of this beetle — temper-
ature, moisture, sunshine, food, and soil — but of these none have played any-
conspicuous part in directing its movements. On the north low temperature
and in the south high temperature and high humidity have established
seemingly effectual barriers to prevent the beetle from extending beyond its
present limits. Between these bounds, however, temperature and moisture
are too nearly uniform to be of any moment in the control of dissemination.
Food has had a greater influence than the preceding, especially in the south,
where the cultivation of the potato is by no means general, and, consequently,
the distribution of the beetle has been somewhat retarded and irregular.
To sum up, in the distribution of L. multiiceniata, intermedia, and decem-
lineata over North America the chief directive factors have been: first,
habits — i, e. , easy flight at the period of its life cycle best adapted for the
extension of its bounds ; second, direct response in flight to the direction
of winds ; third, climatic barriers to the north and south due to temper-
ature and moisture ; and fourth, the amount of its food available. Of these
the first is of the greatest importance. If, as has already been pointed out,
a period of easy flight had not previously been developed, probably no dissem-
ination could have taken place. The development of this habit dates far
back in the history of the species, certainly to the time when midtitccniata
began to spread out over northern Mexico, if not earlier. Previous to the
beginning of this dissemination it was of no very general importance in the
economy of the species, but changes in the environment resulted in the
raising of this habit into one of prime importance as far as ability to undergo
rapid dispersal was concerned. Now that the dissemination is over for the
present, this habit again sinks into a position of relatively small importance
in the economy of the species. While emphasizing the importance of this
habit, I do not mean to underestimate the influence of environment upon
the spread of this beetle, but I do believe that we must understand better
the habits and constitution of an organism before we decide concerning the
relative potency of the various factors in its distribution and evolution.
W. L. TOWER
RETREAT OF L. JUNCTA. 49
DlSTiUBUTION OF Iv. DEFBCTA AND L,. JUNCTA AND RETREAT OF Iv. JUNCTA.
Leptinotarsa juncta Guer. and L. dejecta Stal are two closely related species
with a peculiar distribution. L. dejecta is known from Texas, Arkansas,
and Missouri, and Stal has described it from Yucatan. As far as I am able
to learn, neither juncta nor defccta occurs in Mexico or Central America, and
as careful search along the coast belt of Mexico has failed to reveal any
traces of these beetles, I strongly suspect that Stal's record from Yucatan is
incorrect.
L. juncta had originally a much wider range than at present. In 1865 it
w is found throughout the Atlantic Coastal Plain from Maryland south to
the Gulf of Mexico, over the entire Piedmont belt, along the Gulf coast into
eastern Texas, and up the Mississippi Valley to southern Illinois. It appar-
ently entered the Allegheny Plateau along several river systems, especially
the Coosa-Alabama or Allegheny River. Westward it spread into southern
Missouri, Kansas, Arkansas, and Texas. Over this entire area it was a
common form feeding upon Solatium caroliniense (plate 10).
But with the advent of L. decemlineata into the eastern United States Juncta
very soon retreated or was exterminated. Evidently the two are unable to
live in the same habitat, for, although they have different food plants, Juncta
disappears in any locality soon after the arrival of decemlineata. For example,
it was formerly abundant about Washington, District of Columbia, and in
parts of Maryland, but Dr. Howard states that since the advent of decem-
litieata into that region juncta has retreated southward, and the same is true
in southern Illinois, where it is no longer found. Professor Ouaintance
informs me that in 1899 and previously juncta was perhaps the most common
insect about the station grounds at Experiment, Georgia, but that in 1900
it was found in only very small numbers, its place having been taken by
decemlineata ; nor is it longer found in Virginia, the Carolinas, Kansas,
Missouri, and the upper Mississippi Valley.
Why juncta should retreat before decemlineata is a question of considerable
interest, as the two species have entirely different food plants, and it can hot
be thaX. juncta is crowded out by the larger numbers of decemlineata. I am
informed by Professor Quaintance that the two species hybridize freely in
nature, although the eggs that are laid are not fertile, at least in so far as his
observations go. Concerning this crossing in nature and its effect upon
juncta I shall have more to say in a later paper. The full explanation of
the extinction of juncta is to be found in the fact that the two species cross
freely in nature, and that this natural crossing has resulted in a most inter-
esting and peculiar case of prepotency in one species and of submergence in
the other.
5°
DISTRIBUTION AND DISPERSION OF LEPTINOTARSA.
DISTRIBUTION OF THE OTHER GROUPS.
The distribution of the other groups presents on examination much the
same history as that found in the lineata group, with the exception that none
of them are as widely distributed or as variable, or present the interesting
series of migrations and the interrelation of species, as does the lineata group;
hence the history of this group has been presented in full, while the others
may be passed over with only a few comments.
In the lineata group it was shown that the various species are restricted
in a striking manner to natural environmental complexes. In the other
groups also this same limitation is seen, and the extent to which the develop-
ment of the different groups has given rise to species in the natural divisions
of the country can best be shown by the following table :
Table 2. — Distribution oj Leptinotarsa by Groups in the Different Faunal Areas.
Faunal area.
Lmeata
group.
undecimlineata.
Dilecta
group.
Haldemanii Lacerata
group. group.
Flavopustulata 1
group.
CENTRAL AMERICA :
Guatemala - Chiapas
Plateau.
flavitarsis.
nitidicollis.
distingueuda.
libatrix.
dahlbomi.
evanescens.
belti.
flavopustulata.
O a x a c a-Gu errero
highlands.
undecimliueata.
oblongata.
novemlineata.
dilecta.
obliterata.
libatrix.
dahlbomi.
lacerata.
undecimlineata.
angustovittata.
pudica.
obliterata.
cacica.
hogei.
libatrix.
violescens.
chalcospila.
NORTH AMERICA :
Mexican escarpment:
Humid eastern
slope.
calceata.
obliterata.
cacica.
Moist southern
slope.
signaticollis.
oblongata.
dilecta.
haldemani
lacerata.
dohrni.
Dry western slope.
signaticollis.
oblongata.
dilecta.
Pacific lowlands
diversa.
chlorizans.
litigiosa.
Gulf lowlands
undecimlineata.
rubicunda.
multitaeniata.
melanothorax.
dilecta.
dahlbomi.
tlascalana.
modesta.
stali.
intermedia,
melanothorax.
lineolata.
typographica.
dahlbomi.
puncticollis.
1 decemlineata.
defecta.
juncta.
dahlbomi.
Atlantic Coastal Plain
decemlineata.
juncta.
M2
DO
Q.O.
OC3
-*
4
+
■f
+
z
«
a
z
>
0
0
o
z
■A
X
o
a
4
4
4
4
VOPUSTULA
LACERAT
GROUP
<
rr
"J
<
•
•
•
Z
o
>
I
o
o
o
z
3
*
•
•
01
+
on
+
K
n°
•
+
\-
(DO
•
n
<
•
1-
+
3
.,
«
#
I
^
«.
z
■
«
<
<
•
X
+
o
4
z
•
u
4
u
3
«
i
•
•
>
o
0
z
•
•
a
+
+
H
4
2
•
E
C
3
4
B
Z
1
■4
2
+
£
a
z
1
+
+
5
x
z
.■
4
CO
+
+
<
•
3
□
a
O
o
o
o:
1-
<
O
<
o
+
<
*
*
*
~
I
'
cr a
5 a
DISTRIBUTION OF THE OTHER GROUPS. 5 1
In the lincata group it is evident that there are two chief centers of distri-
bution, one Central American and one North American, and that in each of
these a series of species has developed along lines independent of the others.
In the dilccta group there are also two distinct sets of species, one Central
American and one North American, both of which have followed the high-
lands and moved northward over the Mexican Plateau. In the hdldemani
group, however, the species, with the exception of L. dahlbomi, a few rare
species from the Pacific slope and lowlands of the Mexican area, and one
doubtful species from the north end of the mesa, all develop on the Central
American area. The lacerata group shows an advance eastward from the
center of origin into Yucatan, then northward along the Atlantic slope,
and finally up into the mesa from the eastern side, while the flavopustulata
group barely reaches the North American area. (See plates n, 12, and 13.)
From the foregoing table may be drawn additional confirmation of the
correlation between species formation in the genus and natural physio-
graphic areas. What information we can derive from their geographical
distribution is strongly in favor of the view of a close correlation of species
development with topography and climate.
The general cecological relations of each of the species in the Ii?ieata group
has been given, in as far as it is concerned in the distribution of the species.
In the other groups in the genus Leptinotarsa the same factors have the
controlling influence in the distribution of the various species and in the
determination of habitats. In some of the groups certain species are found
with a general distribution over a wide area of country. Most conspicuous
of these is L. dahlbomi (plate 13), which, of all the species in the genus, is
the least dependent upon a restricted habitat for its existence. In its wide
and varied range of distribution, from Guatemala and Yucatan northward
across the Rio Grande into Texas and New Mexico, it encounters a greater
variety of topographic and climatic conditions than any other species in the
genus and is one of the least variable. Others, as, for example, L. libatrix
violescetis, lacerata, icndedmlineata, and others, are restricted to a ven^ narrow
local habitat by similar environmental conditions. Some of the species in
this genus, L. rubiginosa, for example, are extremely limited in their dis-
tribution, owing to the rarity of their food plants and the conditions of
existence which they seem to demand. The main cecological factors which
control the distribution of all the groups of this genus are, first, moisture ;
second, temperature ; third, soil ; and fourth, altitude.
52 DISTRIBUTION AND DISPERSION OF LUPTINOTARSA.
THE RELATION OF THE DISTRIBUTION OF THE GENUS LEPTINOTARSA
TO NATURAL ENVIRONMENTAL COMPLEXES.
The geographical distribution of animals, or animal geography, is usually
considered from one of two viewpoints — the static or the dynamic. Con-
sidered from the static standpoint, the facts of distribution are taken and
arranged in order according to some empirically chosen standard, and zones,
subzones, or other unnatural areas of distribution are established. The
study of animal distribution from this standpoint is a dead and profitless
pursuit. Dynamically considered, animal geography seeks to explain the
facts of animal distribution as we now find them in terms of the relation of
the animals to each other and to their environmental complexes. It is the
dynamic aspect of the distribution of these beetles that I shall consider in
the remainder of this paper.
One of the most striking facts of animal geography is the correlation
which exists in any large group of animals between the development of its
various species and groups of species and the environmental complexes of
the country which forms the habitat of the animals in question, specific
differentiation following directly upon changes in the environmental com-
plex. This correlation has long been recognized, but whether the relation
is cause and effect or onty apparently so, and how the various lines of specific
differentiation came into being, are open questions.
The groups of species in Leptinotarsa occupy in a marked way one or more
of the natural complexes into which the country in which they live is
divided. All the groups have their origin in the same area," and from this
center have spread out and developed in different lines of species differentia-
tion in the different topographic and climatic areas of the country in which
they live.
On examining the table on page 50 it is evident that the two chief areas
into which the territory inhabited by these beetles is divided — the Central
American and the North American — have almost no species in common.
Only one species, L. dahlbomi, can be said to be common to both areas, while
another, L. dilecta, is found to be an inhabitant of both along their boundary.
It is probable, however, that as the number of local records is increased more
species will be found to have crossed over from one area to the other and
penetrated for short distances inward. This sharp demarkation between
the species that are Central American and those that are North American
is characteristic of the genus as a whole as well as of each group of species
in the genus — a fact that may be seen by an examination of the table. All
of the groups excepting rubiginosa and zetterstcdti have species to represent
them in both areas.
In order to examine more closely into the distribution of the species accord-
ing to topographic areas, let us look for a moment in some detail at the
RELATION TO ENVIRONMENTAL COMPLEXES. 53
Central American area. Although many of the species present are common
to two of the three areas into which the Central American region is naturally
divided, only three, L. dah/bomi, undccimlineata, and libatrix, are distributed
throughout. The number of endemic species in each of the different areas
is of considerable interest. Thus the oldest land area from the standpoint
of the geologist, the Guatemala-Chiapas Plateau, has but two endemic species,
L. belli and cvanescens ; the next oldest, the Oaxaca- Guerrero highlands,
has two, L. lacerata and novem/ineata, while the youngest of all, the Atlantic
slope and lowlands, has four, L. Ziogei, violescens, ftudica, and chalcospila.
In the North American area the distribution of these beetles shows a
similar series of facts. There is no one species that is common to the entire
area, and although the escarpment and the south mesa have four specimens
in common — L. oblongata, viullitanata, dahlbomi, and dilccta — the different
subdivisions have, in general, but few species in common. This area is rich
in endemic forms, there being four on the escarpment, six on the South
Mesa, three on the North Mesa, one on the Great Plains, and one in the
Atlantic Coastal Plain.
From whatever standpoint we examine the distribution of these beetles,
whether of large or restricted areas of the country, it is evident that the
groups of species in the genus Leptinotarsa and the species themselves are
confined to particular habitats, which habitats are natural topographic and
climatic areas or combinations or subdivisions thereof, and that the various
groups and species are in some way very closely correlated with the physical
conditions of their habitat. Similar correlations are known to exist in many
other groups of animals, but not always to the extent found in this genus.
The question is how this distribution and the limitation of the various species
to climatic or topographic areas came to exist as we now find it. Were the
various species produced as the genus was dispersed from its original center
by the action of the environment upon a plastic race so that each group and
each species represents the effect of some particular environmental complex
encountered? Or were the species produced essentially as we now find them
in some ancestral habitat, and from this disseminated, each seeking uncon-
sciously the habitat best adapted to its particular needs and, having found
it, establishing itself therein ? Can we account for the present distribution
on the hypothesis of direct modification in response to environmental stimuli
accompanied by natural selection? or must we invoke the aid of unknown
factors which will operate to produce species having structures and consti-
tutions adapted to various remote environmental complexes, and then, by.
chance dissemination, have the newly modified organisms and the appro-
priate habitats brought together, with the final result that the entire genus
shall present a uniform and perfect correlation and adaptation to its envi-
ronment ? Are the phenomena of animal geography to be explained on
54 DISTRIBUTION AND DISPERSION OF LEPTINOTARSA.
the basis of variation and natural selection or on the basis of " mutation "
and " segregation in the fittest environment " ?
It is almost impossible to obtain direct evidence upon these question.": ;
hence we must depend to a large extent upon the indirect and circumstnntinl.
Aside from experimentation, which is usually not possible, evidence from
two sources only is available in the attempted solution of this question — that
from distribution and migration and that from the study of geographical
variation. Herein only the facts from distribution and migration are con-
sidered, the data and arguments from geographical variation being presented
in the next chapter.
If the distribution of these species has been produced by variation in direct
response to environment and natural selection, we should find evidences of
continuity of species differentiation, as the different races or groups were
disseminated over the country; but if mutation and "segregation in the fittest
environment ' ' has been responsible for the present distribution of the genus,
continuity of distribution could hardly be expected, and its general occur-
rence would be improbable ; hence continuity of specific differentiation in
a given group of species in its distribution over a country should give us
some evidence as to what the evolution of the group in question has been.
In the distribution of the species in this genus evidences of continuity in
specific differentiation, correlated with natural environmental complexes, are
found, the series of species in the lineata group presenting a striking exam-
ple of this phenomenon. The most generalized species and the probable
ancestor of the group, L. imdecimlineata , gives rise, as the result of certain
environmental conditions in nature and in experiment, to a form, angus-
tovilfata, which is a close approximation to signaticollis. In this series of
three species undecimlineafa inhabits the low, hot, and moist country to the
south and east of the Mexican highlands, angustovittata occurs as a regular
member of the fauna of the lower portion of the escarpment of the Mexican
Plateau on the east, south, and west, while higher up on the southern and
western sides of the escarpment signaticollis is found. Between these three
species there are variations of an intermediate character, both in nature and
experiment. It is true that angustovit/ata arises as an extreme variation in
experiment, and also, as I have observed, in nature, and that intermediate
stages are rare ; but this fact is indisputable, that this species arises in
direct response to the stimulus of a certain environmental complex, and
that it is a stable species both in nature and in experiment. From the
.distribution of this series of three species — undecimlincata, angustovittata,
and signaticollis — in three closely placed environmental complexes, we have
strong evidence for the orthogenetic evolution of species as the result of
response to changes of environment. If they had originated as the result
of mutation and if their present distribution were the result of segregation
RELATION TO ENVIRONMENTAL COMPLEXES. 55
iu a favorable habitat, it would be a remarkable coincidence that these
species, which represent steps in an orthogenetic specific differentiation,
should occupy habitats which are also steps in habitat differentiation.
Additional evidence along this line is afforded by other species in this
group. Distributed over the South Mesa, the Escarpment, the North Mesa,
the Great Plains, and the eastern portion of the United States, is a series of
species that show the most perfect continuity, not only in their distribution,
but also in their geographical variations. We are also able to follow the
migrations of these species over different habitats. The species multiUsniata,
intermedia, and decemlhicata present continuity in distribution and specific
differentiation over a wide area to an unusual degree. On pages 18 to 49
the data of the distribution of these beetles is given and discussed at length,
and it seems only too evident that the dissemination and specific differentia-
tion of the race in response to environment have gone hand in hand. So
continuous is the distribution and the geographical variation in these species
that their separation is difficult. In this group, wherein continuity in dis-
tribution is indisputable, another process is also at work. At Toluca, Mexico,
there exists a species (riibicunda) which arises as an extreme variation
from multitceniata. This species is able to maintain itself as a member of the
fauna at Toluca, and, in as far as can be determined from observation, it is
far less successful in the struggle for existence than the parent form with
which it lives. It has a very restricted distribution, owing to its depend-
ence upon a particular environmental complex, in direct response to which
it has perhaps arisen. At this same place, as also at others, another form
(species) arises as an extreme variation of multitceniala. This is the species
melanothorax , which is found in almost every generation of multittzniata at
Toluca, although it is notable there nor at any other point in the habitat of
its parent species to become established as a member of the fauna. Evidentl v
it is ill adapted to the environment into which it is born. There are environ-
ments, however, to which it is adapted, but to reach these requires a numerical
strength sufficient to enable it to spread out ; and a species so ill adapted to the
habitat into which it is born that it can not maintain itself will not develop the
numerical and physical strength that is necessary to enable it to seek new and
more favorable habitations. The theory of ' ' segregation in fittest environ-
ment " could be of no use in the explanation of the early stages in specific dif-
erentiation and dissemination, as it is impossible to see how an " unadaptive
mutation ' ' (Davenport) is able to find its proper place in nature. If it is
unadapted, as in the case of melanothorax , does not the very fact of its
unadapted condition so hamper it in the development of that numerical and
physical strength necessary for any extensive movement that dissemination
would never occur ? It must always remain an ill-adapted member of the
fauna into which it has been born, unless by chance it is carried into some
56 DISTRIBUTION AND DISPERSION OF LEPTINOTARSA.
more favorable habitat, and accidental causes are insufficient for the explana-
tion of natural phenomena. A species must possess a certain amount of
adaptation to its original habitat in order to live and gain strength to spread
out into one to which it is better adapted, as did deccmlincata by migration.
In the group of beetles to which decemlineata belongs the development of
the present distribution and specific differentiation has certainly been pro-
foundly influenced by factors of the environment in which it lived. At
least no other conclusion is warranted from the data at our command.
If we should examine the data concerning the other groups in the genus
Leptinotarsa in the same manner as we have the lineata group, a similar set of
observations and the same arguments would be given. As far as I am able
to discover, they all show the same exact limitation to and dependence upon
a particular environmental complex. All of the species of the genus are
limited to grassland habitats — that is, to areas in which there is a frequent
atmospheric precipitation of greater or less amount, which keeps the surface
of the ground and the lower stratum of air moist during the summer or
growing season and with a distinct dry or winter season. The temperature
is of little moment, as long as it is not low enough to inhibit growth processes
during the moist season. The essential vegetation is perennial grasses grow-
ing in bunches with annual, herbaceous, or woody plants, often of consid-
erable size. The environmental complex, which I here term " grasslands,"
is a natural combination of physiographic, climatic, and plant conditions,
which together form an environmental complex outside of which Leptinotarsa
can not live. Within this general type of habitat there are, however, exten-
sive and often striking differences, and the grasslands are further divided by
local or edaphic conditions into grasslands that are physiologically moist and
physiologically dry. The physiologically dry areas are differentiated by
local or edaphic conditions which differ in individual cases but produce in the
end the same general type of environmental complex — savanna, steppe, or
semi-desert. We recognize, therefore, four chief divisions of the grasslands —
meadow, savanna, steppe, and semi-desert. Each of these is modified locally
by edaphic conditions and between them there are all possible transition
stages, and the change from one to the other is never abrupt, although it may
be rapid ; there is never any discontinuity.
Some idea of the distribution of the species of Leptinotarsa in the four
chief types of grassland habitats may be gained from the table on page 58.
From this table it is evident that the species of Leptinotarsa are limited to
the physiologically dry grasslands, only one being found in physiologically
moist areas. By far the larger portion of the species are found in the
savannas, a few in the steppe, and also in the semi-desert. Although lim
ited almost entirely to the savanna type of grassland habitat, it differs
greatly in its edaphic factors even in nearby areas, and it is the changes in
RELATION TO ENVIRONMENTAL COMPLEXES. 57
the edaphic factors with which the species are most closely correlated.
Each of the groups in the genus represents a line of specific differentiation
which has in the main been brought about in a savanna habitat, but as far
as anything that can be discovered in the distribution of the beetles is con-
cerned, it is impossible to say whether this specific differentiation is due to
the direct influence of the environment, to selection, to the effects of isola-
tion, or to segregation. It would beau easy matter to build up from the distri-
bution of the species of this genus a good argument in support of the idea
that the species have arisen through the action of isolation and segregation or
by selection. Either or both may be responsible for the conditions found, but
as far as I am able to discover there is not one whit of evidence to indicate
which.
The distribution of the species does, however, indicate in the strongest
terms that there has been no discontinuity in the evolution of the genus,
aud in the absence of positive evidence to the contrary we may safely hold
that the theory of mutation and segregation in fittest environment can not in
this instance be used in the explanation of the distribution and dissemina-
tion of Leptinotarsa. It matters little whether specific differentiation has
been rapid or extremely slow, and of this we have no means of gaining
information, but of the continuity of the dispersion and species formation
there is in this case not the least doubt.
The present geographical distribution of a group of animals is the result
of a long series of processes, of action of organism upon organism, and of
environment upou organism. From our study of the distribution of these
beetles we have gained information as to what the geographical history of
the genus has probably been, what forces were most active in the distribution
of the group, and the extended correlation existing between distribution,
specific differentiation, and environment. From this we get a general view,
a clue here and there, of what may perhaps be of vital importance in evolu-
tion. At no point are we able to put our finger upon any one fact and say
with any certainty that this is the result of that factor in the environment
or of any method of evolution. The facts presented in this chapter, those of
geographical distribution, become more significant when considered in rela-
tion to the facts to be presented in the next chapter, on "Variation."
5»
DISTRIBUTION AND DISPERSION OF LEPTINOTARSA
Table 2a. — Distribution of Leptinotarsa by Groups in tlie Four Chief Types of Grassland
Habitats within the Topographic Areas given in Table 2.
Faunal areas and l.ineata
habitats. group.
Dilecta
group.
Haldemani ' Lacerata
group. group.
Flavopustulata
group.
CENTRAL AMERICAN.
Guatemala - C li i a pa s
Plateau :
1
undecimliueata.
Exact habitats not known in this area.
Oaxaca-Guerrero
highlands :
uudecimliueata.
oblongata.
dilecta.
novemlineata.
obliterata.
dahlbomi.
dahlbomi.
lacerata.
Pacific slope :
Atlantic slope :
uudecimliueata.
pudica.
obliterata.
cacica.
violescens.
hogei.
libatrix.
dahlbomi.
dahlbomi.
dahlbomi.
dahlbomi.
Yucatan :
heydeni.
NORTH AMERICAN.
Mexican escarpment :
Humid eastern slope:
calceata.
obliterata.
cacica.
uudecimlineata.
Moist southern slope:
signaticollis.
oblongata.
dilecta.
haldemani
lacerata.
dahlbomi.
haldemani
dahlbomi.
dahlbomi.
Dry western slope :
signaticollis.
oblongata.
dilecta.
dahlbomi.
1 Not (
j known. 1
Pacific lowlands :
diversa.
Gulf lowlands :
uudecimlineata.
South mesa :
multitceniata.
melanothorax.
multiteeniata,
melanothorax.
rubicunda.
multi taenia ta.
melanothorax.
dilecta.
dahlbomi.
modesta.
o
stall.
North mesa :
dahlbomi.
puucticollis.
intermedia.
Great Plains:
dahlbomi.
decemlineata.
Atlantic Coastal Plain:
juucta.
defects.
CHAPTER II.
VARIATION IN LEPTINOTARSA.
The characters chosen for the study of variation in these beetles are color
characters and structural characters. The color characters are of two kinds,
special and general. The special color characters are the elements and com-
binations of elements of the color patterns ; the general are the effects of
the combined color patterns in the production of melanism or albinism or
tendencies toward other general colors. The first have to do largely with
continuous or individual variability ; the second, with place and geograph-
ical variation. The structural characters which are considered are form,
size, puuctation, and glandular openings.
Special color characters :
(a) The color pattern on the pronotum : In the species not unicolorous,
as in the lineata group, the color pattern is composed of a series of bilaterally
symmetrical spots, designated a, I?, c, d, e, and / on the right side, and a',
b' , d , d', e' , and/' on the left side, and am and pm for the two median spots
(text-fig. i). The unicolorous prouota we shall discuss from the standpoint
of their intensity of color (melanism) and their development of metallicism.
(6) The color pattern on the epicrauium : In the species not unicolorous
the epicrania have color centers or areas which we shall designate on the
right side as g, h, i, and on the left asg', h! , i' (text-fig. i).
(c) The color pattern of the elytra : On the elytra, excepting in uni-
colorous species, there are stripes of color related morphologically to the
veins of the wing (text-fig. i), with certain others presently to be described.
The areas of color development parallel to the veins are the costal {cos.),
subcostal (s. cos.), ramous (ra.), medius {med.) , cubitus {cub.), and anal
{an. ). At a right angle to the veins at the base of the wing is an area of
color development, the basal, or first band {bas.)\ a corresponding area at
the apical portion of the wing, the marginal, or fifth band {mar.), and
between these three areas, also at right angles to the veins, the proximal,
or second band {prox.), the middle, or third band (?«.), and the distal, or
fourth band {dist.) (text-fig. 3). It will be shown in the third chapter
that these color areas (fig. 3) are related to structures, and are invariable in
position excepting as the structures vary.
{d) The color pattern of the ventral abdominal surface : In the lineata
group only is there a color pattern on the ventral surface. The elements of
59
6o
VARIATION IN LEPTINOTARSA.
this are designated according to the areas and segments thus : First segment,
/ os, i ms, i is on the right side, and / os', i ms' , i is' on the left side ;
second segment, 2 os, 2 ?ns, 2 is, and 2 os1 , 2 ms' , 2 ?V ; third segment, 3 os,
3 ms, 3 is, and 3 os' , 3 ms' , 3 is' .
In the account of the special color characters use will be made of formulae to
represent given conditions. Thus the formula for the pronotum of text- fig. 1
would be f , c', d' . d ', b' , a': a, b, c, d -+- e, — , where the sign plus indicates
Fig. 1.
Fig. 2.
Fig. 1. — L. dectmlineatd. To show the areas of cuticula color which in this genus enter into the forma-
tion of the color pattern, i, eye ; g' , left anterior lateral epicranial spot ; h' , left posterior lateral
epicranial spot; a', left inner tergals (united); b' and d' , left middle tergals ; c and e, right outer
tergals ; /and /', spots arising from the division of the posterior half of the middle tergals : s. cos i
and s. cos 2, ou-er and inner subcostal stripes ; la. ramous stripe ; »«. medius stripe ; at + a, cubitus
and anal stripes fused. (The two last stripes were wanting in the specimen drawn, but their position
is indicated.)
Fig. 2. — L. dccemlincata. To show the elements of the color pattern formed by the cuticula pigment
on the ventral surface, especially on the abdominal surface, os, outer sternal areas; ms, middle
sternal areas : is, inner sternal areas.
fusion of spots and the sign minus absence of spots ; while for text-fig. 2
the formula for the ventral abdominal surface is :
1 .
os!
ms ,
is
is.
ms,
OS
2
OS'
ms1,
is'
is,
ms,
OS
3-
os'
ms',
is'
is,
ms,
OS
4-
os'
ms',
is'
is,
ms,
OS
5. os' -\- ms? -\- is1 ; is + ms -f os.
General color characters :
(a) Melanism, or the tendency to a dark color exhibited in relation
either to geographical distribution or to environmental changes.
{b) Albinism, or the tendency opposite to melanism.
(c) Xanthism, or the racial tendency toward a yellowish condition.
(d) Rufism, or a racial tendency toward a reddish condition.
() Metallicism, or a racial tendency toward a metallic coloration.
CHARACTERS EXAMINED. 6 1
Structural characters :
(a) The form-index, which is determined by multiplying the greatest
anterior-posterior length by the greatest transverse diameter and the division
of this result by the greatest dorso-ventral diameter of the thorax.
(6) Size ; determined by measuring the greatest anterior-posterior length.
(c) Punctations on the pronotum and elytra.
(d) Glandular openings on the elytra and pronotum.
The nature of my material has made measurements almost impossible, espe-
cially in the color characters, hence I have in the main seriated it into classes
by inspection, and this method has on actual trial given exactly the same seri-
ations as were obtained by measurements. It seemed, therefore, inadvisable
.^^tfSSB^fc^. bas.
bas-prox. m-^ ^HkUmi s' cos'
eu + an ~B&HSB Wm----Prox-
vied. B-^B Wf ^BjK&ym
prox-m. - - — ■-- * -iii'il^^B
n-dist.
dist-mar. . _ .
. mar.
Pi*™.. 3.— .£.. lacerata. To show areas of cuticula color in banded species.
das., basal band ; prox., proximal band ; m., middle band ; dist., distal
band ; mar., marginal band. Other abbreviations as in text-fig. 1.
to employ the complicated methods of statistical investigation elaborated by
Pearson. I have, moreover, proven to my own satisfaction by actual test
that the highly complicated methods of biometry are no more reliable in their
results as far as this material is concerned than other methods far less cumber-
some and slavish. It has been sufficient to determine the range of variation,
the empirical mode and the per cent of variates therein, and the mean.
The color characters have been treated as though the material were all of
one sex. This I am warranted in doing, owing to the fact that there is no
dimorphism, nor even a distinguishable difference in the color pattern of the
two sexes. That no error was introduced by ignoring sex is shown by the
production in every case of clearly defined monomodal curves. As seriation
of the color characters of the two sexes gives identical results, this material
may be treated as though it were homogeneous, and as if sex were non-
existent.
62 VARIATION IN LEPTINOTARSA.
INDIVIDUAL VARIATIONS.
OBSERVATIONS AND DATA OF CONTINUOUS VARIATIONS.
color characters.
Individual Variation in the Color Pattern of the Pronotum.
Variation in the color pattern of the pronotum in the lineata group. — The com-
position of the color pattern is shown (text-fig. i) to consist of spots of dark
pigment upon a lighter ground. On plate 14, figs. 1 to 12, is given a series
of figures of pronota in L. undccimlineata to illustrate the directions of modi-
fication which the color pattern undergoes. In figs. 1 to 9, inclusive, is
shown the most common and almost the only series of variations exhibited
upon the pronotum in these beetles ; i. e. , fusion or apparent fusion of the
color areas by the extension of the spots peripherally or in certain directions,
so that continuous color areas are produced. From any given condition,
excepting the extremes, variation proceeds in two opposite directions toward
reduction of the pigmented area and toward increase thereof.
If in this series we consider fig. 6 as the modal type, then figs. 5, 4, 3.
2, and 1 show stages in the reduction of pigment and also where sepa-
ration of spots first occurs. Thus in fig. 5, d, sand d', e' become separated
by an area of hypodermal color, and this is broadened in figs. 4, 3, 2, and
1. In figs. 1 and 2 is seen the last process of reduction which occurs in
this species, namely, the separation of a and a' and the loss of c and d ',
With this reduction, however, two spots hitherto merged in a + a', am and
ptn, are brought into view. In figs. 3 and 4 the separation of a + b and a' -f b'
is shown. In the opposite direction fig. 7 shows the union of d + e -\- /
and d' + e1 -\-f, and in fig. 8 a further fusion of spots, so that the formula
am
for the color pattern is now d , /' + d + d' -\- b' + a' + a -f- b + d -f e + /, c,
pm
while in fig. 9 this process of increase in the area of individual spots is
further extended.
In this series the variations are definite in two chief directions, and other
lines are not common, although they do exist. These less frequent varia-
tions consist in irregularities in the fusion, as, for example, where we find
a pattern which may be represented by the formula r', /', d' -f e' , b' ,
am am
a'+a + 6 + d+e+J, c, or c' — ', d' , c' , b' + a' + a + b, d, e 4- /, c. These
ptn pm
irregularities are due, however, to the failure of a given spot to develop
symmetrically with its fellow of the other side. Spots are sometimes wholly
absent, most commonly/ and/', the others in this series being about alike
in the frequency of their failure to develop.
Explanation of Plate 14.
Figs. 1 to 12 represent variations in color pattern on the pronotum of L. undecimlineata.
Figs. I, 2, 3, 11, and 12 are from Panama. Figs. 4, 5, and 6 are from Tierra Blanca,
Vera Cruz, Mexico. Figs. 7, 8, 9, and 10 are from Orizaba, Vera Cruz, Mexico.
Figs. 13 to 22 illustrate variations in color pattern of the pronotum in L. multitamata.
All the figures given are taken from beetles collected at Guadalupe, Federal District,
Mexico.
Figs. 23 to 30 illustrate variations in color pattern on the pronotum in L. decemlineata.
Figs. 23, 24, and 29 are from specimens collected at Cold Spring Harbor, New York.
Figs. 24 and 26 are from specimens collected at Dalhart, Texas. Figs. 27 and 28 are
from specimens collected at McPherson, Kansas. Fig. 30 is a variation obtained
in Georgia and also produced experimentally.
Figs. 31 to 35 illustrate variations in elytral color pattern in L. undecimlineata. Figs.
31 and 32 are from specimens taken at Tierra Blanca, Vera Cruz, Mexico. Fig. 33
is a variation of undecimlineata found at Tierra Blanca. Figs. 34 and 35 are from
specimens taken at Orizaba, Vera Cruz.
Figs. 36 and 37 illustrate the transverse fusions found in the elytral color pattern in
L. decemlineata, from specimens taken at Chicago, 111.
Figs. 38 and 39 illustrate variations in elytral color pattern in L.junda, from specimens
taken at Baton Rouge, Louisiana.
Fig. 40 illustrates color-pattern condition in elytra of L. de/ecla, from a specimen taken
at Dallas, Texas.
Figs. 41 to 45 illustrate variations in color pattern of elytra of L. lineolaia. Figs. 41, 42,
and 43 are from specimens collected in Chihuahua, Mexico. Figs. 44 and 45 are
from specimens taken at Deming, New Mexico.
Figs. 46 to 48 illustrate variations in elytral color pattern of L. dilecta, from specimens
taken at Atlixco, Mexico.
Figs. 49 to 52 illustrate variations in elytral color pattern of L. zettersledli, from spec-
imens collected at La Barranca, near Guadalajara, Jalisco, Mexico.
Figs. 53 and 54 illustrate variations in L. pudica, from specimens taken at Peras, Vera
Cruz, Mexico.
Figs. 55 to 57 illustrate variations in elytral color pattern of /.. lacerata, from specimens
collected at Oaxaca, Oaxaca, Mexico.
1
2
•:y:»
*V*
*VS*
9
\/\ t\/*» *Vv
m
■ « \i •
/.'■
tfVS .»V
i; :«:
#
7 //>
17
23
•A/7 -j
99i
12
18
♦ : .*•
30
i |
•I]
MS!)
tf
fill. TO) M
/?*/' WW *///
^
42 43 44 45 46
To illustrate the vai lor pattern of some Leptinotarsa:
COLOR PATTERN OJ? PRONOTUM.
63
Besides the variations described, certain others which are abmodal occur
from time to time. These, while of no great number or apparent impor-
tance in evolution, are of value in the interpretation of certain phases of color
variation. Some of these are found on plate 14, figs. 10, 11, and 12.
In divcrsa, angustovittala, and signaticollis, which are the closest allies of
undecimlineata , the same kind and direction of variation are found. I have
not drawn figures of these species, but the variations appear in the tables.
The color patterns of L. multitceniata and decemli7ieata show a series of
variations of the same general nature as those of L. undecimlineata, as will
be apparent from a study of figs. 13 to 22 on plate 14, in which are shown
the variations of these two species. In this series there is found the same
definite tendency toward reduction of the elements of the color pattern on
the one hand and increase on the other, the extremes being represented by
figs. 22 and 30. Irregularities and abnormalities are of the same nature as
those in L. undecimlineata and are to be explained on the same basis.
With the exception of melanothorax, all of the species in the lineata group
have this same type of color pattern.
The frequency and direction of fusion of the spots is of interest and shows
in the different species variation along definite lines which is apparent in
the following tables :
Table 3. — Direction and frequency of the fusion of spots on the pronotum in L,. unde-
cimlineata (Mexico, Guatemala, Costa Rica, Panama), 1,200 individuals.
Order of
frequency of
occurrence.
Formulae showing color pattern.
Individuals
showing
fusions.
Per cent.
f
a
)
■ i
pm 4- am
a'
\
94
• {
a + b
pm 4- am
a' + b>
}
92
3 {
d 4- e
d' 4- e'
}
80
' {
d +e + f
d' + e' +/'
}
63
■ (
a + t> + d + e+f
pm 4- am
a' + b> + d' 4- e' +f
}
20
• {
a+b+d+f+e+c
pm 4- am
a' + b' + d' +f + e' 4- C
}
1
64
VARIATION IN I.EPTINOTARSA.
Table 4. — Direction and frequency of the fusion of the spots on the pronotum of
L. diver sa (Mexico), 100 individuals.
Order of
frequency of Formulae showing color pattern,
occurrence.
Individuals
showing
fusions.
f "
1 \ pm + am
\ a'
1 * + b
2 < pm + otn
I a' + b'
I d+e
3 j d' + e'
( a+6+d+e+f
5 1 ■ /»« + am
( , a' 4- 6' 4- rf' 4- «?' +/'
Per Cent.
> IOO
[ 93
} 93
1 74
[ 62
•1
Table 5. — Direction and frequency of the fusion of the spots on the pronotum of
L. signaticollis (Mexico), 1,000 individuals.
Order of
frequency of
occurrence.
Formulae showing color pattern.
Individuals
showing
fusions.
Pi?»- c»/ 4" "'"
a> + b>
[ 99
3 {
d+e
d> + e>
} 7&
4 !
f
d+e+f
d' 4- C + /'
a + b + J + e + f
I 6l
5 1
pm + am
a' + b> + d' + c' +f
[ 40
6 \
a + b + d + c+\h.
pm 4- "'"
1
1
a> 4. b> 4. d> 4- e> + {■£
!
a + b + d+e + U
1
1
7
1
pm + am
a' + b' + d> 4- <■' +{ £
1
L
J
In the fusions of the spots indicated in table 5, a and d are united by a lateral-ward
extension of « and a medianward extension of d. (See plate 14, fig. 19.) The fusion is
indicated by the sign 1 1 1.
COLOR PATTERN OF PRONOTUM.
65
Table 6— Direction and frequency of the fusion of the spots on the pronotum in L. mul-
titmiiata (Mexico), 4,200 individuals.
Order of
frequency of Formula; showing color pattern.
occurrence.
Individuals
showing
fusion.
I
put 4- "'"
a'
( a +6
■t 1 pm + am
I j a' + b>
< d+e
I d> + e'
I d+e+S
\ d'+e'+f
a + b+.d+e+f
pm -\- am
a' + b' + d' + e ' +/'
Per cent.
75
7i
52
4i
2S
r
a + b + d + e +
pm -\- am
\f
a'+b'+d' + e' +
If
( Hypodernial color exposed as 1
\ a uarrow lateral margin. (
Total melanism
Table 7.— Direction and frequency of the fusion of the spots on the pronotum in
L. oblongata {Mexico), 1,000 individuals.
Order of
frequency of
occurrence.
Korinulee showing color pattern.
Individuals
showing
fusion.
Per cent.
f
a
1
■ 1
pm. 4- am
a'
:
21
- 1
a
pm + am
a'
[
62
■ I
d+e
d' 4- e'
}
4i
« {
d + e+f
d> 4- e' +/'
}
19
■ (
a+b+d+e +/
pm + am
a' + b' + d' 4- ' +/'
!
3
« {
c + e+f
C + e' +/'
1
i
1
66
VARIATION IN I^EPTINOTARSA.
Table 8. — Direction and frequency of the fusion of the spots on the pronotum in L. rubi-
cunda (Mexico), 200 individuals.
Order of
frequency of
occurrence.
Formulie showing color pattern.
Individuals
showing
fusion.
Per cent.
(
a
}
■ 1
pin + am
a'
V 100
* \
a + 6
pm + o»t
a' 4- y
] 96
> :
e + d
e> 4- d'
« *
* {
a+6+d+e+f
pm 4- am
a' 4- b' 4- d> 4- e> + f
r
Table 9. — Direction and frequency of the fusion of the spots on the pronotum in
L. intermedia (Mexico), 300 individuals.
Order of
frequency of
occuirence.
Formulae showing color pattern.
Individuals
showing
fusion.
■ {
■ !
* (
< {
* {
* \
- 1
a
pm, 4" am
a'
a
pm 4- am
a'
a + b
pm 4- am
a' 4- b'
d + e
d' + e'
e+J
e> +f>
d+e+f
d' + e' +f
a+b + d+e+f
pm + am
a' + b' + d> + e' + /'
Pet cent.
[ 40
I 52
[ 50
} 32
} 3<5
I 6
f 2
COLOR PATTERN OF PRONOTUM.
67
Table 10. — Direction and frequency of fusion of the spots on the pronotum in L. decem-
lincata (United States and Canada), 6^00 individuals.
Order of
frequency of
occurrence.
Formulae showing color pattern.
Individuals
showing
fusion.
pm, -\- am
pm -\- , am
a'
pm -f- am
a'
a + b
pm -\- a vi
a' + b'
a + e
d' + e'
e+f
e> +/'
a+b+c+e+f
pm + ai/i
a' + b' + d' + e' +f
1 + ' ecies of the lineata group.
Species.
L. undecimlineata .
diversa
signaticollis . . . .
multitEeniata . . ,
oblongata
rubicunda
intermedia
decemlineata. . .
juncta
defecta
a' and a.
d and e .
Per cent.
Per cent.
94
80
100
100
93
76
75
62
100
52
4i
62
52
27
32
16
2.2
0.2
0
0
The illustrations on plate 14 and the figures given in tables 3 to 11 show
conclusively that the fluctuating variations of the pronotal color pattern in
the lineata group do not occur promiscuously and in all directions, but in two
well-defined lines. Rarely abnormal conditions occur, but these are the
results of pathological conditions operating during pupation. Individual
variation is definite.
Variation in the proportion between the areas of dark and light pigment
gives interesting facts along the same line. Twenty classes illustrative
of this color relation have been established, of which fig. 30, plate 14, rep-
resents one extreme, with the value of 1, and fig. 32 the other, with the
value of 20. Somewhere between the extremes in this series all the spe-
cies of the lineata group find their proper places. The distribution of the
classes into which each species falls can best be shown by tables.
The proportion of light to dark color on the pronotum of these beetles
shows in table 13 a tendency toward melanism on the one hand and albinism
on the other. Thus some of the species have their mode at the darker end
of the series (L. diversa), while others have theirs at the lighter end. In all,
however, the variations of this character obey Quetelet's law of individual
variability. With one exception no two of the species have exactly the same
modal condition as regards this character, a fact clearly shown in table 14.
COLOR PATTERN OK 1'KoXoTUM.
69
Table 13. — Variation in the proportion between the pigmented and unpigmcnted areas
on the pronotum of L. undecimlineata, L. diversa, L. signaticollis, L. multitamiato, L.
intermedia, L. decemlineata, and L. oblongata.1
Classes.
Per cent of frequency.
/. undecim-
lineata.
*■ <*-*»•■ L^?"
L. n tu I lit (V-
riiata.
L. inter-
media.
L. decemli-
neata.
L. oblongata.
Class 1
....
....
O.I503
2
....
0.1503
3
1
::::
;;;; ;;;;
0.4509
0.6012
5
.... ; 1
I.5030
6
.... 2
2.2545
7
0.7098 4
4.2084
S
0 712
1.8928 6
7-9°59
9
2.436
3-3124 18
28.6675
I
IO
3-348
1 6.6248 36
24.9498
3
II
7- 3°8
2 10 3664 21
16.S938
7.2
12
4.872
4 23.2688 10
7.6051
16.4
r3
S.i 20
S 19 6678
3-7575
39 0
14
22.S60
5
16 14-4326
0.7515
18.4
15
38.164
25
56 9.^972
0. 1503
11. 6
16
T2.180
70
13 4.9686
3-4
17
2.6026 j
....
IS
2.0196
19
0.2366 '
....
20
•
....
L. undecimlineata : N. = 1,230 individuals. Mean of classes = 12.92892.
L. diversa : N. = 100 individuals. Mean of classes = 15.65.
L. signaticollis : N. = 1,000 individuals. Mean of classes = 14.67.
L. multitaniata : N. = 4,220 individuals. Mean of classes = 12.762124.
L. intermedia : N. = 200 individuals. Mean of classes = 9.92.
L. decemlineata : N. = 6,650 individuals. Mean of classes = 9.647779.
L. oblongata : N. = 500 individuals. Mean of classes = 13.024.
Table 14. — Position of the empirical mode of the species of the lincata group upon the
scale of color proportion used in the foregoing tables, and the per cent of variates in
the modal class, which indicates the variability of each species.
Class.
Species.
Per cent
in modal
class.
Class.
Species.
Per cent
in modal
class.
Class 3.
5-
9-
10.
12.
L. decemlineata.. . .
L. inultitseniata. . . .
60
28.6675
36
23.2688
Class 13.
16.
20.
J L,. undecimlineata
L. melanothorax. .
39
38.164
70
100
The numbers of individuals used in the above tables are large, so that the
probability of inaccuracy in the modes and percentages is slight, and the fact
that the determinations in each case are made from material taken from
several generations further enhances the probability of accuracy.
The modes shown in table 14 and the variations shown in the preceding
tables are for the species. The modes for different generations and for
'The empirical mode in this and all following tables is indicated by being set in bold-
faced type.
7o
VARIATION IN LEPTINOTARSA.
single lots of variates show deviation from these species modes and values,
the result of place variation, a phenomenon which will be considered later.
In the individual variation of the color pattern of the pronotum two facts
appear — first, that the variations, while numerous, are not promiscuous, but
in definitely directed lines; second, that the species tend to segregate their
variations about modes arranged along the chief directions of individual
variation.
Variation in the color of the pronotum in the dilecta group. — The pronotum
in this group has a uniformly colored surface, which, however, shows in
some species variations of interest, especially in the intensity of its metal-
licism. Thus, in the series of species L. calceata, L. dilecta, and L. lineolata
there are variations in the color of the pronotum from reddish-brown to
dark metallic green. The range of color is divided into six classes, as
follows: (i) reddish-brown; (2) dark reddish-brown; (3) dark reddish-
brown with a greenish sheen ; (4) light metallic greenish-red ; (5) bright
metallic-green, and (6) dark metallic-green. In table 15 is given the fre-
quency of the variations seen in the different species.
Table 15. — Variation in the color of the pronotum
L. dilecta, and L. lineolata.
in L. calceata,
Per cent of frequency.
Classes.
L. calceata.
L. dilecta.
/.. lineolata.
Class 1
8
2
1 9
3
2
2 12
4
5
22 IO 29
72 85 41
6
4 2 1
L. calceata: N. = 50 individuals. Mean of classes = 4.78.
L. dilecta : N. = 300 individuals. Mean of classes = 4. 85.
L. lineolata : N. = 300 individuals. Mean of classes = 3.89.
I have not been able to procure enough material of the other species of
this group to give satisfactory seriations or data of color variation. In the
three species seriated above the variations are plainly limited to very narrow
lines ; one series of color changes — reddish-brown to metallic-green. These
are the only variations known. All three species show a mode on class 5,
bright metallic-green, although the range of variation is greatest in lineolata
and least in calceata.
Variation i?i the color of the pronotum oj the haldcmani group. — None of the
species of this group possess color patterns, and the only variations are those
of intensity of color. The pronotum of L. dahlbomi, the most widely ranging
COLOR PATTERN OF PRONOTUM. 7 1
member of the genus, is black or deep brown, appearing black by reflected
light, and is almost invariable in color. L. hogei has the thorax dark brown
with a greenish iridescence, which in the few specimens at hand appears to be
slightly variable toward a stronger development of the metallic-green portion
of the coloration. The species haldcmani , libatrix, and violescens present a
series of color variations from black to deep metallic-blue, a range which
may be divided into the following classes: (i) Black, (2) polished black.
(3) black with faint blue iridescence, (4) light metallic blue-black, (5)
metallic blue-black, (6) deep blue-black, (7) violet. In the following table
is given the frequency of the variations seen in the different species :
Table 16. — Variation in the color of the pronotum in L. haldemani,
L. libatrix, and L. violescens.
L. haldemani : N. = 200 individuals. Mean of classes = 1.68.
L. libatrix: N.= 100 individuals. Mean of classes = 4.o8.
L. violescens : N. = 100 individuals. Mean of classes = 6.87.
No other lines of color variation in the pronotum of this group are known
to me. However, some of the species of this group I have not had an oppor-
tunity to examine. The variability of the three species studied is low, a fact
plainly indicated by the high percentage of variates in the modal class, and,
like the preceding groups, is limited to two directions.
Variation in the color of the pronotum i?i the other groups. — In the lacerata
group, as far as is known from the specimens available, the individual varia-
tion of the pronotum is extremely slight. In the rubiginosa and zetterstedti
groups the species are constant, presenting only the slightest oscillations, too
slight to be measured with any degree of success, while in the flavopustulata
group, as far as I know, the color of the pronotum is invariable.
The study of the individual variation of the color and color pattern of the
pronotum warrants the following conclusions :
(1) Variatiofi is definite in opposite directions.
(2) No trace of indeterminate variation exists anywhere.
(3) The variations follow exactly Quetelet's law.
J2 VARIATION IN LEPTINOTARSA.
Individual Variation in the Color Pattern of the Epicranium.
Variation in the color pattern of the epicranium in the lineata group. — The
color pattern of the epicranium, while simple, is no less variable than that
of the prouotum. It consists of two spots placed just laterad of the median
line and two placed at the outer posterior angle behind the eyes (text-fig. i).
In the variations of this area a range is exhibited from complete melanism
to total xanthism and albinism. On plate 15, figs. 1 to 7, is shown the range
of variation found in L. multitceniata and L. me/anot/wrax (fig. 8). This
species shows an almost complete series from one extreme to the other.
Only at the xanthic and albinic end of the series are stages wanting to make
a complete series in this one species. These are supplied, however, by its
northern modifications, intermedia and dccemlineata, which complete the
series (plate 15, figs. 12 to 16). The range of variation found in L. decem-
lineata, as shown in these figures, is considerably less than that found in
multitceniata, but greater than in the other species of the group. The varia-
tions of undccimlineata are shown in figs. 9 to 11, plate 18.
The direction of the fusion of the spots on the epicranium is well shown
in figs. 1 to 6. In fig. 1 all of the spots g, h, i and g' , //', i' are distinct
pigment areas. In fig. 2 are shown the first steps in the fusion, the union
posteriorly of g and g' to form a V-shaped spot and the medianward exten-
sion of h and //'. This process is continued in fig. 3 until g and g' have
formed a heart-shaped spot in the center of the epicranium, which becomes
more pronounced in fig. 4. In fig. 5 is shown the caudalward extension of
this spot until it meets the medianward extension of // and //, and all combine
in the posterior median border of the epicranium. Further variation con-
sists in the lateralward and cephalward spread of the spots g and g' and //
and kl, as shown in figs. 5 to 7. No other variations of the color pattern of
the epicranium of the lineata group are known, excepting rare abnormal
conditions of asymmetrical variation caused by unequal development upon
the two sides. These, however, are due to accidental causes acting during
pupation, and are of no interest in this connection.
The proportion of light to dark color in the epicranium of the lineata
group is of interest, showing, as it does, a series of gradations much like
that found in the pronotum. It has been convenient to establish ten classes
to represent these gradations, class 1 (plate 15, fig. 12) indicating total
xanthism or albinism and class 10 (plate 15, fig. 8) total melanism. The
variation in this character of the different species in the lineata group has
been studied by sedations, which can best be presented in the form of tables.
Explanation of Plate 15.
Figs. 1 to 8 illustrate variations in color pattern on the epicranium of L. multiiceniata .
All from specimens taken at Guadalupe, Federal District, Mexico.
Figs. 9 to 11 illustrate variations in color pattern on the epicranium of L. undetimlineata.
All from specimens taken at Tierra Blanca, Vera Cruz, Mexico.
Figs. 12 to 16 illustrate variations found in color pattern of epicranium of L. decemlineata.
All from specimens taken at Chicago, Illinois.
Figs. 17 to 24 illustrate the variations found in color pattern of ventral surface of abdom-
inal segments in L. multitceniata. All from specimens taken at Guadalupe, Federal
District, Mexico.
Figs. 25 to 29 illustrate variations in color pattern of ventral surface of abdominal seg-
ments in L. decemlineata. All from specimens taken at Chicago, Illinois.
Figs. 30 to 32 illustrate reduction of pigment on the ventral abdominal surface in L. un-
detimlineata. All from specimens taken at Tierra Blanca, Vera Cruz, Mexico.
Figs. 33 to 40 illustrate variations in color pattern of ventral surface of thoracic seg-
ments in L. multittzniata. All from specimens taken at Guadalupe, Federal District,
Mexico.
Figs. 41 to 44 illustrate variations in color pattern of ventral surface of thoracic segments
in L. decemlineata. All from specimens taken at Chicago, Illinois.
Figs. 45 to 52 illustrate variation in color pattern of the legs in L. multitceniata. All
from specimens taken at Guadalupe, Federal District, Mexico.
Figs. 53 to 57 illustrate variation in color pattern of mature larvse of L. multiiceniata,
from Guadalupe, Federal District, Mexico.
Figs. 5S to 62 illustrate variation in color pattern of mature larvae of L. decemlineata,
from McPherson, Kansas.
Figs. 63 to 67 illustrate variation in color pattern of mature larvst of L. signalicollis, from
Cuernavaca, Morelos, Mexico.
... tt £-> A sSjl
12 3 4 5 6 8
&k AX* VU ^> f ' # ?^> W T*£
» ■ - - - » •*— — - # *^«^«
-'" '--~^i
15 16
W^
25 26
20 21
28 29
23 24
|*>
V tu
33 34
« •
' ♦w^' 'ite^' |jm*i |feg*}
35 36 57 38
42 43
45 46 47 48 49 50 5 1 52
^ &«?gfr .--::;-# ^::>$
^flBfe* ,^'^
-••"*^ *'":w
63
64
*iu ^ ^5U
Afi. c t
I I lust rate the variation in color pattern in some species of Leptinol
COLOR PATTERN 01- EPICRANIUM.
73
Table 17. — lariation in t/ie proportion between the. pigmented and unpigmented areas in
the epicranium of L. undecimlineata, L. diversa. L. signatieollis, L. multiteeniata,
L. oblongata, L. decemlineata, and L. intermedia.
Classes.
Per cent of frequency.
L. undecim-
L. diversa.
L. signati-
L. mnllittr-
L. oblon-
L. decem-
L. inter-
lineata.
collis.
niala.
gata.
lineata.
media.
Class 1
2
2
O.472S 06308
5
4
3
o-9456 ! 3 7848
12
16
4
3.5461 14.8218
54
60
5
O.90
I
17.2577 ' 67.5158
20
15
6
15-3°
s
2
10.4019 10.4082
7
4
7
24 32
26
12
13.4754 1.8924
1
8
47.75
66
25
22.9314 0.6308
9
"•73
52
27.8959 0.3154
10
....
8
3-0732 . ..
L. undecimlineata : (Panama, Costa Rica, Nicaragua, Guatemala, Mexico). N.= 1,230
individuals. Mean of classes = 7.5447.
L. diversa : (Jalisco, Mexico). N. = 100 individuals.
L. signatieollis : (Mexico). N. = 1,000 iudividuals.
L. multitirniata : (Mexico). N. = 4,230 individuals.
9; less, class 5. Mean of cl asses = 7.2733.
L. oblongata: (Mexico). N. = 3,170 individuals. Mean of classes = 4. 9303.
L. decemlineata : (United States and Canada). N. = 6,650 individuals. Mean of
classes = 4.000.
L. intermedia : (Mexico). N. = 100 individuals. Mean of classes = 4.0200.
Mean of classes = 7.58.
Mean of classes = 8 45.
Empirical modes = greater, class
Table 18-
-Position of the empirical mode of the species of the lineata group upon the
scale of color proportion used in the foregoing tables.
Classes.
Species.
Per cent
in modal
class.
Classes.
Species.
Per cent
in modal
class.
Class 2
• 1
82
68
54
60
17.2577
67.5I53
Class 8 |
10
L undecimlineata .
L. diversa
L. multitseniata . . .
L. signatieollis
47-75
66
27.8959
52
100
L. juncta
L. decemlineata.. . .
L intermedia
L. multitseniata . . .
L. oblongata
Comparison of the epicranium and pronation in the lineata group. — The color
pattern of the epicranium shows, as far as the proportion of pigmented to
unpigmented areas is concerned, essentially the same phenomenon as that
exhibited by the pronotum — that is, the modes for the different species are
located at various points along the scale of proportion used. If we compare
table 18 with table 14, which gives the position of the modes of the pro-
notum, we see that in the main the position of the modes for each species is
the same in both tables. In both me/anot/iorax occupies the highest place
and defecta the lowest, while undecimlineata and diversa are in both cases in
the same modal class. Signatieollis changes its position in table 18 to one
6— T
74
VARIATION IN LEPTINOTARSA.
above that of undecimlineata, and multitczniata shows two modes, the high
mode being the same as that of signaticollis and the low one the same as
that of oblongata.
The percentage of individuals found in the modal classes is shown in the
following table :
Table 19. — Comparison of the per cent of individuals in the modal classes of the dif-
ferent species of the lineata group.
Species.
Per cent in modal class
in seriation of—
Species.
Per cent in modal class
in seriation of —
Pronotum.
Epicrauium.
Pronotum.
Epicranium.
L,. angnstovittata. .
Per cent.
78 ' '
60
85
100
26.6675
Per cent.
54
58
90
100
54
L. oblongata
multitaeniata . .
signaticollis . .
diversa
undecimlineata
Per cent.
39
25-5
23.2688
56
70
38.164
Per cent.
67.5158
29
27.8959
52
66
47-75
juncta
inelanothorax . .
decemlineata . . .
In this table it is apparent that where the percentage of individuals in
any modal class is high when the pronotum is taken as the standard, the
per cent of individuals in the corresponding class of the epicranium is also
high and in most of the species higher than in the pronotum. In only a
few cases does the percentage of individuals in the modal class of the epicra-
nium fall below that of the corresponding class of the pronotum, and even
in these cases the differences are slight, being only 4 per cent in L. signati-
collis and L. diversa.
The range of variation in the color values in these two parts of the body
is also of interest. It is shown briefly in the following table :
Table 20. — Comparison of the range of variation of color proportion in the pronotum
and epicranium of the species of the lineata group.
Species.
Range in
classes.
Per cent occu-
pied of pos-
sible range.
Species.
Range in
classes.
Per cent occu-
pied of pos-
sible range.
Prono-
tum.
Epicra-
nium.
Prono-
tum.
Epicra-
nium.
Prono-
tum.
Epicra-
nium.
Prono-
tum.
Epicra-
nium.
Iy. defecta ....
juncta . . .
rubicunda. .
melanotho-
rax ...
decemline-
ata
oblongata. .
4
6
4
1
15
8
3
4
3
1
7
8
20
30
20
5
75
40
3°
40
30
IO
70
80
L. intermedia.
multitseni-
ata
sisrnaticol-
'lis.
diversa . .
undecimli-
neata ....
7
13
7
3
9
4
9
6
3
5
35
65
35
15
45
40
90
60
33-3
50
COLOR PATTERN OF EPICRANIUM.
75
From this table it is apparent that the range of variation is in every case
greater in the epicranium than in the pronotum ; but in most of the species
the variability of the pronotum is greatest — a fact easily seen by comparing
the per cent of variates in the modal class ; that is, the epicranium shows
variates in more of the classes between the possible extremes of variation
than does the pronotum, but on the average in the group has the variates
less widely distributed.
The color patterns of the pronotum and epicranium represent, the one a
complex and the other a simple condition, and it appears from the com-
parison of data in tables 19 and 20 that the simpler color pattern has the
greater range of variation, but lower variability.
Variation in the color of the epicranium of the dilccta group. — The epicra-
nium in the dilecta group, like the pronotum, shows only a uniformly colored
surface varying in intensity and in the same directions as the pronotum — that
is, from red-brown to bright metallic green. These variations are not of
great moment, but such as occur are presented in the following table :
Table 21. — Variation in color of epicranium in L. calceata, L. dilecta, and L. lineolata.
Classes.
L. calceata.
L. dilccta.
L. lineolata.
Class 1
6
2
I
I
11
3
3
3
15
4
18
9
26
5
73
84
40
6
5
3
2
L. calceata: N. = 100 individuals. Mean of classes = 4. 78.
L. dilecta : N. = 100 individuals. Mean of classes = 4.85.
L. lineolata : N. = 300 individuals. Mean of classes = 3.89.
Variation in the color of the epicranium of the haldemani group. — As in the
dilccta group, a color pattern on the epicranium of the species of this group
is wanting, there being found only changes in color from a dull black to a
violet. The classes into which the material falls are the same in number as
are those of the pronotum given on page 71.
Table 22 shows the variation found in this group. From a comparison
of this table with table 16, it is evident that there is a strong correlation
between the variations of the epicranium and pronotum. Other than this
correlation and the fact of the limitation of its variation to narrow lines and
very low variability, this group presents in the color pattern of these parts no
points of interest from the standpoint of individual variation.
76
VARIATION IN LEPTINOTARSA.
Table 22. — Variation in color of epicranium in L. haldcmani. L. libalri.v. and
I,, violescens.
Classes.
L. halde-
£. //»a-
L. vio-
mam.
//■!;*■.
lescens.
Class i
61
2
12
2
3
25
5
4
2
80
2
5
8
3
6
4
4
7
i
91
L. haldemam : N. = 200 individuals. Mean of classes = 1.62.
L. libatrix : N.= 300 individuals. Mean of classes = 4. 10
L. violescens : N. = 100 individuals. Mean of classes = 6 84.
Individual Variation in the Color Pattern of the Elytra.
The elytra present more interesting and significant variations in their color
pattern than do any other portion of the body, and the fact that their color
characters are used as specific differentials adds much interest to a study of
their variation.
An elytron has six veins (text-fig. 1), between which there are six inter-
spaces where pigment is laid down. This deposition of pigment gives either
stripes or rows of spots as found in the lineata and dilecta groups (plate 14,
figs. 31 to 48), or in other members of the genus transverse bands, or rows of
spots in transverse series, situated at definite points upon the wings (figs.
49 to 57). In the whole genus the transverse markings occur at definite
and homologous points upon the elytra, and in their modifications follow
certain well-defined laws.
Variation in ike elytral markings in the lineata group. — The elytra of this
group all present a color pattern composed of alternating stripes due to the
deposition of cuticular and hypodermal pigments. These dark and light
stripes are situated, the one between the veins, and the other over the veins
(text-fig. 1 ), excepting that between the subcostal and ramous veins there
are two stripes of dark color which behave very differently from the others,
and rarely there are found elements of transverse bauds producing transverse
fusions. The extreme reduction of the anal portion in the elytra in Coleoptera
has compressed the anal and cubital areas into a narrow zone along the inner
border of the elytra.
In the twelve species of this group the variation consists of an increase or
a decrease in the dark color. The extremes of variation in color are repre-
sented by L. signaticollis (plate 14, fig. 33) and L. melanothorax or L. unde-
cimlineata (plate 14, figs. 31 and 34), while the extremes of lateral fusion
between the longitudinal stripes are represented by deecmlineata and undecim-
lineata (plate 14, figs. 34, 36, and 37). The variations of increase and decrease
COLOR PATTERN OF ELYTRA. JJ
of the longitudinal stripes are the normal individual variations, while the
lateral unions between stripes are rare and sporadic in their occurrence.
The stripes are most variable at their ends, especially at the distal end,
where the major portion of the variation occurs. The most common varia-
tion is the reduction in the length of the stripes, so that they do not reach
the posterior angle of the wing (plate 14, figs. 31 to 37). In the different
species of the group the variation produced by a reduction in the length of
the stripes is always more marked in juncta and defccta than in the others.
It is to be noted that reduction in the stripes begins first near the anal edge
of the wing and proceeds most rapidly there, and that when stripes are
wanting it is the anal and cubital stripes, or those that are morphologically
the posterior members of the series, that drop out first, while the subcostal
stripes are the least variable and the ones which are last to go in a reduction of
pigmentation. At the proximal end the variation is slight and unimportant.
The two subcostal stripes differ from the others in their behavior in that in
most of the species they are fused at both ends of the elytron, and in juncta,
and often in melajiolhorax and rubicunda, along their entire length f plate 14,
fig. 38). Variations in these stripes consist in separation, first in the middle,
then at the posterior end, then at the anterior end, then the reduction of the
posterior member of the pair, and finally of the anterior member ; or this
variation may be reversed.
The transverse fusion of stripes is most common in decemlineata, melano-
thorax, and undecimlineata (plate 14). In almost all species this consists in
the union of the cubital and anal areas and their further fusion with the
medius from the region of the first transverse band through the rest of their
length. A common variation is the union of the remaining stripes, the sub-
costal and ramous, posteriorly in the region of the fourth transverse band.
These variations, which result from the reduction of the anal and marginal
portion of the wing, are common in all of the species where fusion occurs.
Other transverse fusions of sporadic occurrence are figured on plate 14,
figs- 34, 36, and 37. In these we have the union of the ramous and posterior
subcostal stripes by means of a transverse band in the region of the first
transverse bar (plate 14, fig. 36), of the subcostals in the region of the
second transverse bar, and of the ramous and posterior subcostal in the region
of the third transverse bar. In the figures on plate 14 it is evident that the
transverse fusions, both common and sporadic, occur only at the region of the
formation of transverse bars of dark color. This is common in other species
of the genus, the one exception being the two subcostal stripes, which,
unlike all the other stripes on the elytron, fuse for their entire length.
In forms like signatico/lis and angustoviltata, where the dark markings are
absent, we occasionally find that the position of the stripes is indicated by
small deposits of pigment at the proximal and distal ends, especially in the
region of the anterior subcostal stripe.
78
VARIATION IN LEPTINOTARSA.
Variation in the color pattern of the group proceeds in two directions, one
antero-posterior, the other from right to left, and between these there are no
intermediate conditions. All variations are restricted definitely in position
by controlling forces which we shall study later. In these two directions
variations of fluctuating character following Quetelet's lawr abound.
Quantitative study of the variations of the elytra in the lineata group. — The
quantitative study of these variations has been made by seriations into classes.
We shall examine first, the frequency of certain types of variation, and, sec-
ond, the proportion of light to dark in the color of the elytra.
Table 23. — Frequency of variation in the elytra in L. undecimlineata, L. multitaniata,
L. dccemlineata, and L. juncta, expressed in percentages.
Classes of variations.
L. undecim-
lineata.
No fusions
co. + (cu. -f- a.) p
(sc. a. + sc. p.)p
(sc. a. -\- sc. p.) a
{sc. a. + sc. p. ) a. to p
(ra. + sc. p.) a
{ra. + sc. p. ) p
(m. + ra.) a
(m. + ra.) p
(All fused) a
(All fused \p
Transverse fusions between sc. a
and sc. p
Transfusions sc. p. and ra
Transfusions ra. and m
Stripes absent :
c.
sc. a
sc. p
ra
m
c.
a
12.
20.
40.
45-
8.
10.
12.
15-
11.
o.
o.
0.5
0.01
0.02
L. multitce-
niata.
II .
12.
80.
3°-
I .
4-
29.
18.
3-
6.
i-3
0.1
0.09
o.
o.
o.
o.
o.
o.
0.2
L. decern-
lineata.
31-
3-
65-
'9'
o. 1
24-
2.
18.
70.
5-
1.5
2.
14
1.1
05
o.
o.
o.
0.2
1.6
4-
L. juncta.
»5-
95
76.
1.
o.
o
o.
o.
o.
78.
o.
o.
IOO.
o.
o.
o.
2.
18.
98.
The frequency with which certain types of fluctuating variations occur
is of interest, especially when we consider them with reference to the possi-
bility of their being of use in the production of new species or modifica-
tions. In table 23 the variations are designated by classes and formulae
similar to those used in the discussion of variation in the pronotum. In
these formulae c. = costal stripe; sc. — subcostal stripe ; r. = ramous; cu. =
cubital ; m. = median ; a. = anal ; sc. a. = anterior subcostal ; and sc.p. =
posterior subcostal. Fusions are indicated by +, and the absence of a
stripe by the abbreviation for it followed by minus in parentheses ( — ).
a
(Sc. a. + sc. p.) p. and (sc. a. + sc. p.) a. indicate that the two subcostal stripes
P
COLOR PATTERN OF ELYTRA.
79
are fused posteriorly or anteriorly, respectively. Union at other than the
anterior or posterior ends is indicated by formulae like sc. a. + sc. p. /, which
shows fusion at the first transverse band.
This table shows a most decided race tendency in each species in the
frequency of certain variations. Thus, in undecimlineata, almost no stripes
are wanting, the anal stripe being absent in only o. i per cent of the cases ;
but in multiUeniata, deccmlineata, and juncta is found an increasing tendency
for stripes to be missing, until in juncta and dejecta a maximum is reached.
In the various types of fusion also we find the same racial tendency for a
particular variation to increase or decrease as we follow it through a group
of species. These racial tendencies are of importance in geographical varia-
tion and species modification, which we shall discuss in the latter part of
this chapter.
Variation in the proportion between the pigmented and tmpigmented areas in
the elytra of tlie lineata group. — The variation in the proportion existing
between the light and dark color of the elytra of the group is significant,
indicating, as it does, racial tendencies or trends of variation. These ten-
dencies have been studied by seriations along a scale of values in which o
indicates total albinism or xanthism, or complete absence of dark color,
and 20 a uniform black or dark color over the whole elytral surface. The
data derived from this source is presented in the form of a table giving the
frequency which each species shows in any given class.
Table 24. — Variation in the proportion of light to dark color in the elytra of L. unde-
cimlineata, L. signaticollis, L. diversa, L. multitaniata, L. rubicunda, L. oblongata,
L. decemlineata, L. juncta, and L. defecta.
L. unde-
L. sig-
L. di-
L. multi-
L M
L. rubi-
L. ob-
long-
ata.
L. dtcem-
L.
L. de-
ata.
collis.
versa.
ttzmata.
cunda.
hneala.
juncta.
fecta.
Class 1
15
3
2
..
45
0.5
0.5
8
3
40
I
1-5
24
4
.
1.5
3
60
5
I
..
I
2
15
3
6
I
°-5
I
2
66
2
7
3
°-5
2
3
10
.
8
5
1
2
21
2
.
9
7
2
2
4
44
2
10
21
3
6
52
19
11
41
22
11
2
21
3
12
14
65
23
3
10
2
13
6
5
29
20
6
1
14
: 1
2
21
60
1
15
1
4
14
16
Individua
"1
.
2
1
Is 1,000
100
100
4,000
100
500
5,000
200
100
8o
VARIATION IN LEPTINOTARSA.
In the preceding table are shown conditions similar to those found in the
epicranium and pronotum— that is, variation in the position of the modes of
the different species upon the scale of values, and the wide difference in the
percentages of the scale covered by the range of each species. When pre-
sented in tabular form these conditions become more evident.
Table 25. — Position of the modal condition of the species of the lineata group in the
scale of values used for proportion of light to dark color in the elytra; also the per
cent of total range covered by each species and per cent of variation in modal class.
Position of mode of species upon scale of value.
Per cent of
range.
Per cent of
value in modal
class.
4. L. defrcta
'5
20
3°
60
50
5°
35
55
3°
25
45
60
66
44
52
41
65
29
60
75
10. L, oblongata.
Comparison of the above table with tables 14, 18, 19, and 20 reveals the
fact that there is in the variation of the elytra the same racial tendency to
proceed in a definite direction along a given line of modification, and that
each species of the group is segregated about a modal condition which is
placed in the main at similar positions along the scale of value used for the
different parts. Among the various species the only conspicuous difference
between the modal states of the elytra and those of the pronotum and
epicranium is found in L. signaticollis, which is at the lowest end of the scale
in the elytra and high upon the scale in the pronotum and epicranium.
Variation in the elylra of the dilecta group. — The elytral color pattern con-
sists of longitudinal stripes, as in the lineata group, and certain spots belong-
ing to the second, third, and fourth transverse bands. The longitudinal
stripes show, moreover, a striking tendency to reduction and to division
between the second and third transverse bands, and the entire absence of
the costal, anal, and cubital stripes. Only the two subcostals, the ramous
and the median stripes, are found, and these are often greatly reduced. On
plate 14, figs. 41 to 48, are shown some of the variations which the color
patterns of the elytra of this group undergo. The same two directions of
variation are found, the longitudinal and the transverse, but in this group
the variation in the transverse direction is much more pronounced, all modi-
COLOR PATTERN OF F.LYTRA.
8l
fication tending to the production of transverse markings, or rows of spots,
caused by the breaking up of the longitudinal stripes in the spaces between
the transverse bands (figs. 41 to 45). The addition of these new spots to the
transverse bands (figs. 41 to 43) complete, as in fig. 45, a color pattern of trans-
verse rows of spots. Fusions between these different color areas are so rare
that they may be considered to be non-existent.
The process of the reduction of the longitudinal bands, beguu in the lineata
group and carried to completion in L. signaticollis, takes in this group a some-
what different trend, iu that the stripes are broken up into spots by their
reduction in the spaces between the transverse bauds. This takes place first
in the spaces between the second and first transverse bands, and also first in
the posterior stripes. Sometimes, however, the reduction is first seen in the
anterior subcostal stripe (fig. 46), and this is later followed by a reduction
in the space between the third and fourth bands. If after the reduction of
the stripes in the space between the bands there should occur an expansion
of the spots, a pattern of transverse bands of solid color would be produced,
such as occurs in other species of the genus.
Variation in the percentage of light to dark color in the elytra of the dilccta
group. — The seriations in this character show it to be one of the least vari-
able of any in the genus. This is explainable by the fact that as reduc-
tion of the stripes goes on development of the spots belonging to the
system of bands as a rule goes on also, and thus the balance of color propor-
tion is remarkably well preserved. The variations found as the result of
seriation iu this character are presented in the following table :
Table 26. — /aviation in proportion of light to dark color in elytra of the dilecta group.
Species.
Class.
Remarks.
6
7
8
9
10
2
8
8
2
81
90
8
9
2
84
6
N. = 100 individuals.
N. = 300 individuals.
N. = 50 individuals.
Variations in the elytra of the haldemani group. — The color of the elytra of
the beetles in this group presents only the variations of the uniformly colored
surface, to which is added in most cases a modification of physical origin.
The colors are largely physico-chemical. As is the pronotum and epi-
crankim, the color of the elytra is only slightly variable. Seriations on the
basis of the same classes used in the case of the pronotum gave the results
shown in table 27.
82 VARIATION IN LEPTINOTARSA.
Table 27. — Variation in color of elytra in L. haldemani, L. libatrix, and L. violesrens.
Species.
Class.
Remarks.
1
2
3 4
5
6
7
2 8
3 5
68
6
1
22
82
5
4
94
N. = 500 individuals.
N. = 200 individuals.
N. = 100 individuals.
L. libatrix.
L. violescens. . .
Variations in the elytra of other groups. — The beetles in these groups
comprise species of which I have not been able to obtain a sufficient num-
ber for quantitative study, but which, nevertheless, present interesting and
suggestive color-pattern conditions. In these species the pattern consists
largely of bands, stripes being almost entirely wanting. When stripes do
exist they are made up of the coalesced anal, cubital, and often medial stripes,
and are found only along the anal edge of the elytron. These coalesced
stripes are compressed into a narrow sutural line of dark color, from which
bands extend across the wing (plate 14, figs. 49 to 57). These figures explain
themselves and show clearly the way in which the color pattern of bands of
dark color, bands connected by longitudinal stripes, and spots of hypodermal
color on a dark background are produced by the development of the band and
stripes, or both. In plate 14, figs. 55 to 57, are shown, in L. lacerata, the
transition from a pure banded condition to one in which the bands are con-
nected by the partial development of certain of the stripes ; and finally by the
further coalescing of bands and stripes, is shown the production of a pattern
with light spots. In L. pudica (plate 14, figs. 53 and 54) we see the opposite
process in the formation of black spots on a light background by a reduction
of the bands, which leave spots at the points of crossing of the bands and
stripes. In L. zctterstcdti is shown much the same process at work.
The color patterns of the elytra of the genus Leptinotarsa show certain
important and interesting phenomena. The rarity of abmodal cases and
of sports, or extreme variations, is striking. Where these exist they are
limited entirely to certain species of the lineata group. The variations are
almost always continuous and fall in every instance about a single modal con-
dition which is, as far as the evidence goes, constant for the character and
species. We shall see later that the modes of the different species vary geo-
graphically and from year to year, but the modes for the entire species seem
to be a constant quantity.
By far the most interesting phenomenon exhibited by the color patterns of
the elytra in these beetles is the rigid directing of variation along certain well-
defined lines. In no species is there the slightest evidence of promiscuous or
indeterminate variation .
COLOR PATTERN OK ELYTRA. 83
In the lineata group, e. g., the evolution of species has clearly followed
narrow, fixed lines in the color pattern of the elytra, and that this has resulted
in the production of species in certain directions only. It is in the directions
in which the species have evolved that we find the individual variations fluctu-
ating on cither side of the mode toward other closely related species until,
in very many cases, they overlap. It is also to be noted that individual
variation does not extend in other directions than those mentioned. We can
explain this condition only on the basis that the variation of the species is con-
trolled and directed by the constitution of the organism independent of the
conditions of its existence. The study of variation can not give us the
reasons for these phenomena ; it can simply discover the facts concerning the
direction and extent of variations and the laws which they follow. A.s to whv
they follow these laws variation studies must remain forever silent. Embry-
ology and experimentation only can offer aid in the solution of these
difficulties.
In the elytra the following laws hold for the variations exhibited by the
color pattern :
(1) There exist two primary sets of color-pattern markings: (a) bands
across the wing and (b) stripes along the length of the wing. All color pat-
terns in the genus consist either of stripes or modifications thereof, bands or
modifications thereof, or combinations of the two. Unicolorous conditions,
either melanic, albinic, xanthic, or rufic, result from complete fusion of bands
and stripes, or complete reduction thereof.
(2) We recognize six stripes related to the veins, and five bands. These in
all species occupy homologous positions. (See third chapter, on the develop-
ment of color patterns, for causes of these conditions.) The point of crossing
of a band and stripe is an area of pigment production more resistant to reduc-
tion than any other area.
(3) The edges of the wing are more variable than the middle, and the anal
edge less variable than the costal in banded species and more variable in
striped species. The distal end of the wing is more variable than the proxi-
mal. These laws are in accordance with Bateson's (1894) law that the
extremes of a series are more variable than the middle.
(4) Individual variations which are limited to definite directions are either
along stripes, by the reduction at the extremes, by the breaking up into spots,
or by fusion ; or it is along bands by reduction at the extremities, by breaking
up into spots, or by fusion. Variation in a given species may go along one or
the other or both of these lines at the same time, but for a given species the
direction of variation is always constant
84 variation in leptinotarsa.
Individual Variation in the Color Pattern of the Ventral Surface.
The ventral surface of these beetles presents variations of two kinds — first,
modifications in the elements of a color pattern, and second, variation in the
color intensity of a unicolorous surface. The second is the predominating
one in the genus. The first, limited to a part only of the lineata group, is
more complicated than the others and of far greater interest.
In the lineata group all the species present a highly variable color pattern
excepting undecimlincata, angustovittata, signaticollis, diversa, and melano-
tliorax. In these the ventral surface is unicolorous. The variation of the
color pattern of the abdominal segments is shown on plate 15, figs. 17 to 32,
and that on the thoracic segments in figs. 33 to 40. As may be seen, the color
pattern is made up of a series of spots arranged upon the abdominal and
thoracic segments. These spots, which are located over muscle attachments,
consist of three sets on either side of the median line. It is from these spots
as centers that coloration spreads out, or to which it contracts, as the case may
be — that is, color appears here first in development, and disappears here last
in reduction. L. inultitcciiiata is the most variable species in the group, exhib-
iting the series of stages shown on plate 15, figs. 17 to 24, inclusive. An
examination of these figures makes plain the manner in which the color
spreads from these segmentally placed spots. Fusion moves first laterally,
extending until all the spots are united into bands on either side of the median
line ; next these bands begin to extend across the median line at the anterior
end of the series of abdominal segments until they finally fuse, first anteriorly,
and then posteriorly. The increasing pigmentation next extends between the
bands, first until the entire anterior border of the segments is covered, and
then posteriorly until the condition shown in fig. 24 is reached. In figs. 25 to
29 are given stages in the process of reduction, as observed in L.deccmlineata,
the series starting with a condition like that of fig. 17 in L. multitccniata.
First the middle rows of spots on each side are reduced (figs. 25 and 26),
then the median rows (fig. 27), and lastly the outer rows (fig. 28). In all
the pigment begins to disappear first on the segments at the posterior end and
remains longest at the anterior end. In figs. 30 to 32 (plate 15) are shown
stages in the reduction of the pigment of the ventral surface of L. undecim-
lincata from Panama, in which it is evident that the color disappears first pos-
teriorly and then anteriorly. On the thoracic segments is shown the same
processes at work in L. multitccniata (figs. 33 to 40) and /,. undecimlincata
(figs. 41 to 44).
These variations in the color pattern of the ventral surface are found in
the genus only in multitccniata, rubicunda, oblongata, intermedia, undecim-
lincata, dejecta, and juncta, and all present the same strongly marked tendency
to modification along the definite lines designated above. Indeed, from the
variations figured on plate 15, conditions could be selected to illustrate the
COLOR PATTERN OF VENTRAL SURFACE.
85
variation of all of the species mentioned above, no other variations being
known excepting asymmetrical abnormalities.
The variation of the color pattern upon the ventral side is a metameric one,
dominated by the principles of metameric control of variation in such struc-
tures, as is shown in the beginning of increase of color at the anterior end
and decrease at the posterior end. The variations are in complete accord-
ance with the general laws of progressive development in a metameric series
which begins anteriorly, and of regressive development which begins poste-
riorly. Upon individual segments the variations move distalward and caudal-
ward in increase, or proximalward and cephalward in reduction, and these
directions of modification are, as far as this material is concerned, absolutely
inflexible, being most rigorously controlled by the structure and constitution
inherent in the race.
Quantitatively, the variations of the ventral surface present points of some
interest. These have been seriated, as have the variations on other parts of
the body, on a scale of value indicating the proportion of light to dark color.
Table 28.— Variation in the proportion of light to dark color on the ventral surface of
L. multitaniata, L. oblongata, L. intermedia, L. rubicunda, L. decemlineata, L. juncta,
and L. defecta.
Classes.
L. multi-
tee niata.
L. oblon-
gata.
L. inter-
media.
L. rubi-
cunda.
L. decem-
lineata.
L. juncta.
L. defecta. I
I
6
2 . . . .
2
10
3
I
3
72
4
I
8
9
5
I
15
2
6
3
42
1
7
I
7
19
8. ..
2
3
16
6
9
I
3
6
35
2
I
4
25
20
2
11
2
20
42
I
10
4
40
17
I
3
13
22
21
5
2
2
....
14
42
5
1
12
1
15. ...
19
2
...
60
16. ..
4
20
17
2
4
18
1
•■
19. ..
1
20
Individuals
1
4,000
2,000
1,000
300
100
200
100
This table shows clearly the range of variation in this color pattern. L.
multitceniata is the most variable and rubicunda the least. The ventral sur-
faces in species where this part is unicolorous are not as variable as the uni-
colorous dorsal parts of the body. As far as is known, they show no points
of interest.
86
VARIATION IN I.EPTINOTARSA.
Individual Variation in the Color Pattern of the Legs.
In the majority of the species of this genus the legs are unicolorous, and are
only slightly variable. Such variations as occur on them are minute fluctua-
tions, too slight to be seriated. In several of the species of the lineata group,
however, the legs show a color pattern of some interest. On plate 15, figs.
45 to 52, inclusive, are given the variations found in L. multitceniata, which is
the most variable of the species in respect to the color pattern on the legs.
Here we find a series from total melanism to total xanthism, with all possible
intergradations. The stages in color modifications are in this case perfectly
evident and need not be described. Of the other species with a color pattern
on the legs the condition in L. intermedia is represented by figs. 48 to 51, in
undecimlineata by figs. 48 to 52, in oblongata by figs. 47 to 50, in rubicunda
by figs. 45 to 49, in juncta by figs. 50 to 52, and in defecta by figs. 51 and 52.
As in other color characters in this genus, variation in this is limited to cer-
tain directions, as shown on plate 15. Some species show wide variability,
others slight ; but none go beyond that indicated in the figures.
The frequency of the variation in the different species, together with the
range and the per cent of variates in the modal classes for the different species,
are shown in the following table :
Table 29. — Frequency of variation in the color pattern of the legs in the lineata group.
Fig. on
plate 15.
Frequency of variation in the color pattern.
Classes
L. multi-
L. oblon-
L. inter- L. decern-
L. rubi-
L. juncta.
L. de-
teg niata.
gata.
media, j lineata.
cunda.
fecta.
Per cent.
Per cent.
Per cent. Per cent.
Pet cent.
Per cent.
Per cent.
Class 1
45
1
.... .... ....
3
2
46
12
IS
3
47
28
15
....
52
....
4
43
42
58
1° 5
21
5
49
13
20
72 12
6
6
5°
2
7
17
61
5
7
Si
1
1
20
75
IO
8
52
1
2 20
90
Individual Variation in thk Color Pattern of Mature Larvae.
The larvae of these beetles also show marked variations, especially in the
later instars. The newly hatched larvae show few variations in the dark
markings, but as they grow modifications in the color markings appear and
culminate in the final instar. These modifications are purely of an individual
character and are due to acceleration or retardation in growth or to extension
PUNCTATION OF ELYTRA. 87
or reduction in pigmented areas, caused by changes in temperature, moisture,
and food supply.
On plate 15 are given figures to show the variations observed in the mature
larvae of L. multitceniata, decemlincata, and signaticollis. In figs. 53 to 57 of
multitceniata is shown reduction of the spots and loss of many of them, and in
figs. 58 to 62 of decemlincata is shown the variation found in that species.
The two species form a rather complete series from the most spotted condition
(fig- 53) to the least (fig. 62). In figs. 63 to 67 of signaticollis is another
similar series of color variation. In all of the larvae the variations are in one
direction — toward increase or reduction of spots. The young larvae are
hatched with more spots than those shown in fig. 53, and the constant
tendency in all of the species is to reduce the number of these areas during
the ontogeny of the larvae. That some succeed better than others is shown
in the figures on plate 17. When seriated these characters follow exactly
Quetelet's law for fluctuating variation.
structural characters.
Individual Variation in the Pdnctation of the Elytra.
In Leptinotarsa, Zygogramma, and Stilodes, as well as in Calligrapha, as
observed by Jacoby, the color markings are surrounded by deeply impressed
punctations, which lie above the columns of chitin and hypodermal cells con-
necting the two lamellae of the wing. These punctations are so correlated
with the color markings that they may be considered a part of the general
scheme of ornamentation found upon the elytra.
The plan of the punctations found in the genus Leptinotarsa is shown in
dccemlineata. There are two rows of these punctations between each two
veins, and four in the subcostal space ; and they correspond exactly in position
to the edges of the stripes. Besides the stripes of punctations there are bands
which lie at right angles to the stripes. These bands of punctations are well
developed in some species, are absent in others, and appear in still other spe-
cies as rather infrequent variations.
Thus there are two systems of punctation — stripes and bands. The wing
exhibits one or the other of these, either as predominant or, more rarely, both
equally developed. In addition to these larger punctations there are also
smaller ones which are distributed without any discernible scheme of arrange-
ment.
In the lineata group the punctations show much the same series of variation
as was found in the dark stripes already considered. There is a tendency to
fusion at the ends, toward reduction distally and posteriorly, and toward the
union of the stripes by transverse rows of punctations similar to the variations
found in the dark markings. Variations of the punctations in undecimlineata
88 VARIATION IN LEPTINOTARSA.
and in dccemlincata show similar conditions. In species like signaticollis and
angustovittata the punctations become more variable. In all but two of the
species of the group the punctations are arranged in irregular double rows, as
in undccimlincata, dccemlincata, and signaticollis. These irregular double
rows allow of considerable latitude in variation. In signaticollis the lateral
displacement is often so extreme that the rows of punctations are almost
indistinguishable. In two of the species, however, juncta and defccta, the
punctations occur in a single even row, without any observable lateral devia-
tion. In these two species, juncta and defccta, the variation occurs almost
entirely at the distal and posterior portions of the wing. In reduction the
distal and anal punctations disappear first, later the proximal and costal, and
last of all the subcostal.
The arrangement of the punctations and some of the variations in L- cal-
ceata, dilecta, and lincolata, in the dilccta group show, as in dilccta, a decided
tendency to reduction at the interbands, and this system of ornamentation,
like that of the coloration, is much reduced distally and in the anal area of the
elytron. The purely fluctuating variations in the dilccta group are not as
numerous nor as extensive as in the lincata group. In the dilccta group,
however, the variations tend to reduction of the stripes and bands and to pro-
duction of spots or rings of punctations which surround the color areas. In
species like L. zettcrstcdti this reduction of the stripes and development of the
bands is more pronounced, although the stripes are still recognizable. In
lacerata, pudica, and modesta are found variations in which the system of
bands is prominent, and the stripes are largely obscured. In the unicolorous
species, such as haldcmani and violcsccns, the system of stripes is variable to
a considerable extent.
It is found that the punctations vary in some directions and not in others.
Moreover, the stripes and bands vary most in the apical and anal portions of
the wing, to some extent at the proximal end, and least at the center. Finally,
they follow the same line and the same direction of variation as do the color
areas of the elytral color pattern. There is apparently a strong correlation
between the two.
In the quantitative study of the punctations the plan has been to seriate
them into classes by inspection, and to some extent by counting. The fre-
quent variations of this character, found on seriating, while not showing
much that is of interest, confirms the conclusions reached concerning indi-
vidual variation in the color character of the elytra by evidence derived from
structural characters. The seriated variations shown in table 30 are com-
pletely in accordance with those of the color characters, and, like them,
follow Quetelet's law. The smaller punctations, in their variations, also
follow the same laws as do the larger ones.
PUNCTATION Ol- PKONOTUM.
89
Tabu; 30. — Variation in the elytral punctation in L. signaticollis, L. decemlineata,
L. nmltitceniata, L. juncta, L. lineolata, L. dilecta, L. haldemani, L. violescens, and
L. zettcrstcdti.
Classes.
L. sr'gna-
ticollis.
L. decemli-
neata.
L. inulti-
tceniata.
L. juncta.
L. lineo-
lata.
L. di-
lecta .
£. Aaftfe-
mani.
lescens.
L. zetler-
sledti.
Class I
2
3
4
5
6
7
3
9
10
2
4
5
8
18
41
17
5
2
4
8
60
19
6
1
I
5
7
20
52
10
3
2
2
16
71
10
1
2
19
54
21
3
i
"S
18
46
21
6
4
....
9
22
35
20
10
3
2
24
37
22
12
3
4
21
50
J9
3
2
r
L. signaticollis : N. = 1,000 individuals. Mean of classes = 7.76.
L. decemlineata : N. = 4,500 individuals. Mean of classes = 5.12.
L. multit&niata : N. = 2,500 individuals. Mean of classes = 5.69.
L. juncta : N. = 100 individuals. Mean of classes = 2 92.
L. lineolata: N. = 300 individuals. Mean of classes = 4.07.
L. dilecta : N. = 200 individuals. Mean of classes = 4.17.
L. haldemani : N. = 100 individuals, Mean of classes = 5.06.
L. violescens : N. = 200 individuals Mean of classes = 5.25.
L. zetterstedti : N. = 100 individuals. Mean of classes = 3.86.
Table 31. — Variation in the punctation of the pronotttm in l,. signaticollis, L- decem-
lineata, L. nndccinilincata, L. lineolata, L. dilecta, L. haldemani, and L. violescens.
Classes.
L. signa-
ticollis.
L. decem-
lineata.
/.. unde-
[cimh?ieata.
L. lineo-
lata.
L. dilecta.
L. halde-
mani.
L. violes-
cens.
Class 1
2
....
3
I
8
2
I
4
4
I
20
8
8
8
5
I
15
4
29
25
26
27
6
5
26
15
21
32
42
45
7
26
30
20
15
21
16
14
3
36
21
24
4
10
6
6
9
22
2
16
2
2
1
10
7
1
10
I
11
2
6
...
12
1
4
....
13
....
14
15
L. signaticollis : N. = 1,000 individuals. Mean of classes = 8.07.
L. decemlineata : N. = 3,900 individuals. Mean of classes = 6.56.
L. undecimlineata : N. = 1,400 individuals. Mean of classes = 8.04.
L. lineolata : N. = 200 individuals. Mean of classes = 5.40.
L. dilecta : N. = 200 individuals. Mean of classes = 6.00.
L. haldemani : N. = 100 individuals. Mean of classes = 5.76.
L. violescens : N. = 200 individuals. Mean of classes = 5.77.
7-t
90 VARIATION IN LEPTINOTARSA.
Individual Variation in the Punctation of the Pronotum.
On the pronotum the punctations are most abundant on the posterior and
lateral margins. They are highly variable, both individually and geographic-
ally. When seriated, as in table 31, they give results that are in complete
accordance with those derived from the seriations in other parts.
Individual Variation in the Glands of the Elytra.
The beetles of this genus have in the elytra rows of compound hypodermal
glands situated between the dark stripes, which secrete a yellow, oily fluid of
strong odor and acrid taste. This fluid is strongly repugnatorial, and serves
the purpose of protecting these beetles from the attacks of insectivorous ani-
mals. As far as is known, these structures exist only in rows beneath the
stripes. The number of glands does not vary greatly in any given species,
but between the different species there are wide differences. Thus, in L. sig-
naticollis there are relatively few glands, while in multitcrniata and in decem-
lineata they are numerous. As these glands are strongly protective, species
possessing them in abundance would have an opportunity to multiply in num-
bers, whereas those with fewer glands would be less protected and more ac-
ceptable as food for insectivorous animals, and hence less abundant. L. multi-
tceniata, deccmlincata, and oblongata possess large numbers of these glands ;
consequently these species are active and the most common and widely dis-
tributed ones in the genus ; while signaticollis, dilecta, haldemani, and others
have fewer and less active glands, and are much less common and widely dis-
tributed. The correspondence between the number of glands and the numer-
ical abundance of certain species is suggestive of the part which these glands
may play in enabling a species to become a common and widely distributed
form. Natural selection would in this case be an important factor in the
increase of the species possessing the glands and in the elimination of those
with few of these protective structures.
The variations of these glands when seriated give results completely in
accordance with those derived from other characters already studied.
In table 32 there is a marked difference in the position of the modes
of the species, and also in the range of variation. Thus, in the species of the
lineata group, multitcrniata and decemlineata show high modes and high
ranges of variation, while juncta and signaticollis have modes lower in the
scale and a much smaller range of variability. In other groups the modes are
low and the range of variation limited.
Individual Variation in Size and Shape.
Individual variations in size and shape are of interest to a limited extent
in geographical and place variation. Both the variation in size and in shape
follows exactly the laws of the distribution of fluctuating variations found in
other characters, and further discussion of them is unnecessarv.
GLANDS OF ELYTRA.
91
Table 32. — Variation in per cent of frequency in the glands of the elytra in L. undecim-
lineata, L. signaticollis, L. multitaniata, L. oblongata, L. decemlineata, L. juncta, L.
rubicunda, L. lineolata, L. dilecta, L. calceata, L. haldemani, L. violescens, L. libatrix,
and L. rubiginosa.
Classes...
1
2
3
4
5
6
7
8
9
10
II
0
to
25
to
50
to
75
to
100
to
125
to
150
to
175
to
200
to
225
to
250
to
24
49
74
99
124
149
174
199
224
249
274
L. undecitnl ineata
16
74
8
2
signaticollis
multitaeniata
9
90
I
I
6
82
8
I
oblongata
decemlineata
••
2
4
14
9
70
71
12
14
2
2
juncta
21
60
19
26
49
24
1
lineolata
14
7
1
72
84
11
10
8
80
4
I
6
2
10
76
12
2
7
79
10
4
libatrix
2
74
14
8
2
rubiginosa
•
4
21
60
14
1
■•
L. undecimlineata : N. = 1,000 individuals.
L signaticollis : N. = 1,000 individuals.
L. multitceniata : N. = 2,200 individuals.
L. oblongata : N. =600 individuals.
L. decemlineata : N. = 5,200 individuals.
L. juncta ; N. =200 individuals.
L. rubicunda : N. = 100 individuals.
L. lineolata : N. = 300 individuals.
L. dilecta : N. = 100 individuals.
L. calceata : N. = 100 individuals.
/.. haldemani : N. = 200 individuals.
L. violescens : N. = 100 individuals.
L. libatrix : N. = 100 individuals.
L. rubiginosa : N. =100 individuals.
OBSERVATIONS AND DATA CONCERNING EXTREME VARIATIONS.1
Observations upon variations of this class must of necessity be scanty,
excepting in experiment, and even there they are not abundant. In nature
several extreme variations have been found. The best and longest known
of these is the species L. melanothorax , which occurs as a variation of
L. multitaniata, but does not become a permanent species. It differs from the
parent in general appearance, shape, and color, in the black pronotum and
ventral surface, and in the fusions of the two subcostal stripes on the elytra.
On plate 16 is represented in fig. 4 the parent species, multitaniata, in life,
and in fig. 5 melanothorax. A second extreme variation from multitaniata
1 1 have used " extreme variation " or " extreme variate " in place of " discontinuous
variation " and " mutant " because there is no real discontinuity in these variations, but
a rapidly developed extreme deviation from the parental state.
92 VARIATION IN LEPTINOTARSA.
is the form rubicunda, shown in fig. 3, plate 16. This I have found in nature
and reared in experiment. It breeds true to type.
An interesting extreme variation is that represented in fig. 2, plate 16,
described by Jacoby as L. angustovittata. I have found this form in nature
and reared it in confinement from L. undecimlineata (plate 16, fig. 1), so
there can be no doubt of its character. It breeds true to type.
L. decemlineata has proven a species from which several of these variations
have arisen, some of which I have shown on plate 16. Six undoubted cases
have been observed to arise from this species — L. pallida (plate 16, fig. 7),
defectopitnctata (plate 16, fig. 6), minuta, tortuosa (plate 16, fig. 9), melan-
icum, and rubrivittata. These, excepting tortuosa, have arisen several times
in nature and in experiment and are known to breed true to type.
It is highly probable that additional variations of this class will be found as
observations are extended, and I suspect that many of the species of this
genus, which are known only by one or two specimens from one or two
restricted localities, are in reality variations of this character and not perma-
nent species.
THE LAWS OF VARIATION IN THE GENUS LEPTINOTARSA.
From the study of variation in this genus certain rules or laws have been
found, which are followed most rigorously by the different species and groups
of species. The study of variation can not give us data as to why a given
species obeys certain laws, or how modifications are produced therein ; it can
furnish only exact data as to their existence and extent of influence. It is
important that such information should be obtained concerning a form if
further investigation of the problems of evolution is contemplated.
(1) Direction of variation. — In all of the species and characters examined
in Leptinotarsa variation is determinate, and in few directions, no case of
indeterminate variations being found.
(2) Certain special laws appear, which apply to this genus and perhaps to
others as well:
(a) Coloration shows variation in two general directions only, toward the
dark and the light, the dark being melanic or metallic and the light albinic,
xanthic, or rune. In any lot of variates modification exists in not more than
three directions, two of which are light, are closely related, and intergrade,
while the other, the dark, is diametrically opposite.
(b) In the elements of the color pattern there is a tendency for the spots
to spread out or contract peripherally and the stripes and bands to extend or
contract at their ends. The spots, stripes, and bands are most variable in the
posterior or distal portions of the structures on which they occur, and least
variable in the anterior and proximal portions thereof. Increase of pigmenta-
Explanation of Plate 16.
Figs, i and 2. — L. undecimlineaia (i) and its extreme variate atigustovittata (2).
Drawn from life. It will be noted that the color in life is quite different
from the dead specimens.
Figs. 3 to 5. — L. multilcrniata (4) and its extreme variates rubicunda (3) and
melanothorax (5). Drawn from living specimens. The colors here repre-
sented change greatly in death.
Figs. 6 to 9. — L. decemlineata (8) and its extreme variates totiuosa (9), pallida
(7), and defedopuvctata (6). Drawn from life.
Figs. 10 and 11. — Larva; of L. mvllitirniata (11) and of its extreme variate
rubiatnda (10).
W. L. TOWER
PLATE 16
t n
—U
MUTATIONS" OR RAPIDLY DEVELOPING LARGE INHERITABLE VARIATIONS IN
LEPTINOTARSA WITH PARENT SPECIES.
PLACE VARIATION. 93
tion or modification of color pattern moves caudalward or distahvard, while
decrease moves cephalward or medianward.
(c) On the wings the markings follow the same rules which were found
by Mayer (1902) to hold for the wings of Lepidoptera.
(d) Large or extreme variations are determinate and always occur in direc-
tions corresponding to the maximum lines of fluctuating variations.
(3) Range of variation. — In all cases for any given character variation is
limited in both directions, but there seems to be no limit to the combinations
of characters in the formation of species. The species show differences in the
extent to which they exhibit variation in any character, in some the percent-
age of the total range of variation being high and in others low, and this is
a constant quantity, a characteristic of the species.
(4) Correlation of variation.- — All variations of color and structural char-
acters are strongly correlated. On the dorsal and ventral surfaces, and
between them, correlations in variation are equally strong, so that causes
which produce a variation in one part bring about either directly or indirectly
corresponding variations in other parts.
PLACE VARIATION.
The term "place variation" is used to designate a phenomenon more or less
general in plants and animals. It has long been recognized that animals and
plants vary from generation to generation, from season to season, from year
to year, and perhaps in longer periods. Fluctuations in environmental factors,
in the intensity of the struggle for existence, in fecundity, and other causes
bring about the condition that succeeding generations of the same animals are
not even in one restricted locality alike. As frequently as this place variation
must have been noted, it has been passed over as unimportant, and has not
been studied in animals and only slightly in flowering plants. The researches
of Burkhill (1895), MacLeod (1899), Ludwig (1901), Tower (1902), Shull
(1902), Yule (1902), Pearson (1903), Reinohl (1903), and Shull (1904)
have shown the existence of this phenomenon to an unrealized degree, and its
importance where studies of evolution and variation are in progress. As has
been pointed out by several authors, and especially emphasized by Shull
(1904), workers in the field of biometry are all too prone to ignore this
phenomenon.
Shull states (p. 372) :
The interpretations which students have based upon the assumption that seasonal
fluctuations do not occur will have to be greatly revised or discarded altogether, and
before we can appreciate the exact bearing of any case of variation upon the great
problems of evolution it will be necessary to know the laws governing that variation.
94 VARIATION IN LEPTINOTARSA.
In a former paper ( 1902) I have used the term "secular variation" to desig-
nate this phenomenon, and have advanced certain views regarding the study
of variation. It was shown that biometric work must take into account this
fluctuation in the population, and the term "place mode" was expanded to a
broader significance. Pearson (1902), however, objected to my use of the
term "place mode" in that "it might refer to any constant whatsoever of the
frequency — to the mean, mode, variability, or indeed to the whole frequency
distribution itself"- — and I believe that the point is well taken. Shull (1904)
proposes the term "place condition" or "place habit" to designate the phenom-
enon for which I used the term "place mode." This is, I believe, a good term,
not open to biometric or other objections, and I shall use it in this paper. In
the interest of uniformity and for certain philological reasons, I use the term
"place variations" to designate what I called "secular variations" in 1902.
We recognize, then, (1) place variation, or the variation in any given species
in the same locality from generation to generation, or from season to season,
or year to year; (2) place condition, as the state prevailing in the population
of a given species at a particular locality during one generation; and (3)
place constant, determined for the population of a given species at a given
locality for one generation.
, The existence of the phenomenon of place variation was discovered early
in the study of this material and was at first interpreted as indicating rapid
change in condition. But in succeeding generations, when the variations
moved back to the old original condition, the phenomenon with which we now
have to deal was recognized. In the genus Leptinotarsa not all species show
this phenomenon alike, but all do to some extent.
In L. decemlineata the place variation is extreme, and differs in degree in
different places. We can best illustrate this by data from specific localities.
Hence I shall give the place variations of decemlineata for a locality in West
Bridgewater, Plymouth County, Massachusetts, and for Chicago, Cook
County, Illinois, for several generations at each place.
At the first locality collections of L. decemlineata were made during the
years 1895 to I9°2» or during a period of 16 generations. They were collected
from potato fields in which there were about 12 inches of soil, 18 inches
to 2 feet of clay subsoil, and a stiff blue till clay beneath. Various characters
of the beetles collected at this locality were studied. The table of the distri-
bution of frequencies (table 33 a) indicates the proportion of light to dark
color on the epicranium and the variations that have been found from genera-
tion to generation. Table 33 b shows the corresponding variation in the pro-
notum ; table 33 c that in the elytra ; table 33 d that in the abdominal seg-
ments, and table 33 t the variation of males and females in size. In table 33 F
are given the polygons of place condition for the species taken as a whole.
It is evident that in these 16 generations there is abundant place variation.
place; variation.
95
At Chicago I have observed this phenomenon during the years 1902, 1903,
and 1904, or for 6 generations. In the tables of distribution I have given the
seriations obtained from this material. In table 34 A is shown the variation in
the proportion of light to dark color in the epicranium, and in tables 34 b,
34 c, and 34 d the frequencies for the same relation of the pronotum, elytra,
and ventral surface. In table 34 E are given the polygons of variation in size
of the two sexes.
Comparing the place condition polygons for decemlineata at the two local-
ities, it appears that at both there is marked place variation, although less at
Chicago than at West Bridgewater. However, I know of no reason why
continued observation might not show as great place variation at Chicago as
at other localities where it has been studied.
That the phenomenon is not limited in the genus to L. decemlineata is
shown by the data obtained from the study of L. multitmiiata during 1903,
1904, and 1905 at Guadalupe, Federal District, Mexico. The soil is a deep,
stiff adobe, well drained by ditches. I have shown the variations in the pro-
portion of light to dark color on the epicranium, pronotum, elytra, and ventral
surface for these years in tables 35 a, 35 b, 35 c, and 35 d. In the polygons
of distribution of this species we see again a large amount of place variation,
even more than in decemlineata.
Table 33. — Place variation in the epicranium, pronotum, elytra, ventral surface of ab-
dominal segments, size of males and females, and for the species as a whole of
L. decemlineata at West Bridgewater, Plymouth County, Massachusetts.1
(A)
PLACE VARIATION IN THE EPICRANIUM.
Class.
Year.
Gene
alior
r-
.
1
2
3
4
5
6
7
8
9
10
11
12
13
M
IS
1895. {
I. .
I
4
15
56
18
4
2
2. .
1
3
t7
20
42
10
4
2
1
1896. j
1. .
2. .
7
4
60
16
18
20
8
49
5
9
2
2
1897. {
I . .
2. .
4
19
2
52
20
20
61
4
12
1
4
1
1898. |
1 . .
2 .
1
2
3
6
17
25
49
36
19
22
6
5
3
2
2
2
1899. {
I. .
1
2
20
4b
29
3
2 .
1
3
8
22
37
17
8
4
1900. j
I. .
• 9
36
27
16
7
3
2
2. .
14
39
3i
11
3
1
1
|
I..
7
19
50
18
6
1901 :
2. .
5
22
46
21
3
■>.
1902. j
I . .
11
52
21
11
4
1
2. .
2
10
48
28
8
3
1
1 In all of these tables the frequencies are given in percentages.
96
VARIATION IN LEPTINOTARSA.
Table 33— Continued.
(B) PLACE VARIATION IN THE PRONOTUM.
Year.
Geuer-
atiou.
Class.
1
2
3
I
I
4
1
2
I
5
I
I
3
1
1
6
I
I
4
4
1
2
1
7
I
3
1
2
11
12
7
4
5
3
8
5
2
1
6
20
2
2
6
1
1
34
41
19
5
11
8
9
14
2
3
61
50
22
'9
25
3
4
20
30
40
16
44
48
10
54
14
16
19
22
59
49
36
20
7
15
12
15
41
18
26
11
20
24
18
7
4
14
17
22
47
23
8
2
6
21
10
10
12
3
41
51
4
2
7
5
26
36
2
1
2
5
4
3
'3
3
9
7
3
1
3
2
2
16
2
1
1
3
2
1
14
5
4
1
1
9
1
1
2
1
•5
3
1
4
1
1
1895. {
1896. |
1897. {
1898. |
1899. {
1900. 1
igoi.j
1902. <
I. . . .
1. . . .
2. . . .
I....
(C) PLACE VARIATION IN THE ELYTRA.
I895- {
1. ...
2
3
14
45
19
10
5
1
1
2. . . .
4
17
20
42
13
4
1896. |
1
2. . . .
2
5
1
7
19
60
20
18
49
8
11
1897. {
1. . . .
2. . . .
I
4
2
20
20
51
59
18
14
5
4
1
1
1898. |
1
2 . . .
I
2
2
6
17
26
53
35
18
20
6
7
3
2
1
1
1899. {
1
2. . . .
1
1
4
2
7
21
23
45
39
28
18
3
8
1900. <
1
2. . . .
1
2
2
s
4
7
9
39
3b
31
29
10
15
4
4
1
1
1
1901 {
1
2
7
T.H
49
18
6
2. . . .
2
S
20
46
21
^
1
1
1902. •{
1 . .
2. . . .
3
2
10
11
50
47
19
27
11
9
4
4
3
(D) PLACE VARIATION IN THE VENTRAL SURFACE OF THE ABDOMINAL SEGMENTS.
1895. {
1896. {
1897. {
1898. 1
1899 {
1900. <
1901 . 1
1902. -J
I
I .
2....
I
I
I
I
2
I
I
I
3
4
1
1
4
5
1
1
4
8
2
5
5
3
5
1
11
18
3
4
7
1
1
42
8
5
11
9
9
14
2
4
51
51
19
16
24
2
3
20
18
18
42
44
48
55
16
16
17
21
60
50
36
20
9
16
50
46
20
20
28
19
21
20
8
3
13
18
20
46
21
7
17
21
11
10
8
4
40
50
5
1
5
6
7
25
35
4
7
3
4
5
4
2
12
9
2
1
5
4
6
8
1
2
1
2
7 •■
1
9 •■
1
1
1
PLACE VARIATION.
97
Table 33— Continued.
(E) PLACE VARIATION IN SIZE OF MALES AND FEMALES.
:iass.
Year.
Gener-
ation.
Sex.
1
2
14
3
61
4
18
s
4
6
3
7
8
9
10
11
12
JL
f
■{
Male. . .
I895- \
Female .
1
S
14
55
19
6
■{
Male . . .
S
17
19
44
12
3
[
Female.
4
14
2.3
42
14
3
r
■{
Male. . .
s
16
20
42
10
7
1896.
Female .
I
4
1.3
24
40
12
.5
1
H
Male. . .
6
61
19
7
4
3
Female.
2
S
60
18
8
7
*
■{
•1
Male . . .
7
16
50
20
S
2
1897. •
Female.
Male. . .
9
1
14
21
SI
60
19
13
6
S
1
Female .
3
21
60
12
4
*
■{
Male. . .
4
17
50
18
6
.S
1898
Female.
1
3
18
49
19
S
4
1
•{
Male . . .
8
23
38
21
6
4
Female.
1
10
20
39
20
7
3
'
Male. . .
1
3
20
45
27
4
1S99. ■
Female .
Male . . .
1
1
4
4
9
20
20
44
34
28
17
3
IS
Female.
2
6
31
36
IS
10
■{
Male. . .
8
37
25
18
7 .
.S
1900 •
Female .
14
36
23
LS
9
3
'{
Male. . .
12
39
31
14
4
Female.
11
40
28
17
3
1
,'
■{
Male. . .
1
6
19
50
18
6
Female .
1
8
17
50
17
7
-{
Male . .
2
4
21
45
22
3
2
1
Female .
1
21
49
20
S
1
'
■{
Male. . .
11
52
21
10
6
1902. ■
Female.
2
9
50
22
11
s
1
■1
Male . . .
3
8
44
33
6
S
I
*■
Female.
3
9
47
29
9
3
IF) PLACE VARIATION FOR THE SPECIES AS A WHOLE.
Year.
Gener-
ation.
Class.
I
2
5
1
"e
21
2
4
6
1
1
37
40
6
4
9
9
3
14
3
4
50
50
22
16
24
1
4
27
30
20
23
50
45
4
58
16
11
20
51
40
37
22
9
17
11
40
46
23
31
5
16
20
21
8
5
20
19
20
40
20
6
5
28
21
10
7
6
7
43
40
4
2
5
15
7
29
30
3
6
6
5
4
7
8
9
10
11
12
'3
14
15
I895- {
1896. j
1897. j
1898. j
1899. j
1900. |
1901. |
1902. j
I.
2.
I.
2.
I.
2.
I.
2.
I.
2.
I .
2.
I .
2.
I .
2.
••
2
I
2
.
8
14
2
3
10
18
2
2
2
3
21
2
1
1
4
2
1
2
1
2
7
2
1
1
1
5
1
1
3
98
VARIATION IN LEPTINOTARSA.
Table 34. — Place variation in the epicranium, pronotum, elytra, size of males and fe-
males, and ventral surface of abdominal segments of L. decemlineata at Chicago, Cook
County, Illinois.
(A)
PLACE VARIATION IN THE EPICRANIUM.
Year.
Gener-
ation.
Class.
1
I
I
2
9
4
19
4
3
1
4
3
38
19
45
19
2
6
5
17
4
41
50
26
40
10
10
18
48
5
9
21
8
21
17
20
52
19
6
2
6
2
6
35
39
20
8
7
4
21
19
4
3
8
3
11
2
1
9
2
10
1
II
12
13
u
IS
1902. <
I9°3- {
1904. J
I905- {
1. .
2. .
I..
2. .
1. .
2. .
I .
2. .
(B) PLACE VARIATION IN THE PRONOTUM.
1902. 1
1903- {
1904. 1
1905- {
I....
2. . . .
I
2
2
I
2
I
6
2
8
3
10
4
3
1
3
20
20
45
19
2
6
5
18
49
40
21
41
10
10
17
40
17
29
8
20
15
19
52
27
4
7
4
6
37
30
21
8
2
4
20
19
4
2
1
3
12
4
1
2
4
3
1
1
1
(C) PLACE VARIATION IN THE ELYTRA
1902 j
I
1
9
29
40
16
4
2. . . .
. .
I
4
19
39
30
6
j
I903- {
1
I
2
S
II
4b
20
9
^
3
1
2 ...
1
3
18
42
21
5
4
3
2
1
1904. {
2
9
14
37
20
10
6
2
2. . ..
I
I
I
2
4
5
9
18
32
18
5
3
I
1905- {
1
6
i"i
53
20
4
1
1
2
18
47
20
7
3
1
1
(D
PLACE
VARIATION IN
SIZE OF
MALES AND
'EMALES.
Year.
Gener-
ation.
Sex.
Class.
1
2
8
7
3
1
18
10
5
1
2
3
1
3
3
3°
31
20
3
46
40
18
4
1
5
6
5
16
4
4
49
48
50
18
20
3°
36
17
II
14
11
10
18
I
49
17
5
8
7
20
51
14
14
25
39
15
13
19
18
50
23
18
54
6
3
5
7
19
2
4
6
24
37
32
40
40
22
45
9
18
7
7
2
3
12
20
24
16
12
3
27
2
6
8
I
3
2
12
12
5
6
1
4
1
1
9
4
I
4
5
5
1
10
11
••
12
13
1902.
1903.-
1904. ■
I9°5-
■{
'{
■J
'{
•{
'{
■{
'{
Male. . .
Female.
Male. . .
Female.
Male. . .
Female.
Male. . .
Female.
Male . .
Female.
Male. . .
Female.
Male . . .
Female.
Male. . .
Female.
]
I
!
!
[
PLACE VARIATION.
99
Table 34 — Continued.
(E) PLACE VARIATION IN VENTRAL SURFACE OF ABDOMINAL SEGMENTS.
Year.
Gener-
ation.
Class.
1
2
4
5
1
2
10
1
1
3
2
6
2
20
46
3
3
4
17
7
S
40
20
15
6
6
1
45
8
2S
22
s
45
14
8
6
23
9
36
l\
21
35
20
14
8
10
20
1
2
6
22
30
54
2
11
3
1
5
19
21
1
12
1
1
1
13
4
3
1
13
3
5
2
1
14
1
1
15
1902. |
I9°3- {
1904. |
I9°5- {
I
I . . .
2. . . .
2. . . .
I
I
2
I
I
I
I
5
1
Table 35. — Place variation in the proportion of light to dark color in the epicranium,
in the pronotum, in the elytra, and in the ventral surface of abdominal segments of
L. multitaniata at Guadalupe, Federal District, Mexico.
(A) PLACE VARIATION IN THE PROPORTION OF LIGHT TO DARK COLOR
IN THE EPICRANIUM.
Year.
Gener-
ation.
Class.
1
2
3
4
5
6
7
8
9
10
11
12
13
•4
15
16
17
18
19
20
'9°3{
1904 J
1905 |
1. . .
2. . .
I . . .
2. . .
. .
I
I
4
I
2
5
IS
12
3
19
21
51
49
1
9
55
46
19
19
5
21
21
22
11
II
20
53
3
4
1
3
31
12
1
I
40
1
3
(B) PLACE VARIATION IN THE PROPORTION OF LIGHT TO DARK COLOR
IN THE PRONOTUM.
1903 {
1904 {
1905 {
I
I
I
2
I
7
4
2
3
15
2
10
7
5
8
'7
9
17
12 13
i°| 15
11 14
20: 27
10 14
17 30
26
22
25
II
25
12
14
28
15
5
16
4
12
10
14
2
13
3
1
3
5
1
6
3
2
2
2
1
1
1
1
1
1
1
(C) PLACE VARIATION IN THE PROPORTION OF LIGHT TO DARK COLOR
IN THE ELYTRA.
1903 {
1904 {
1905-}
I
3
3
1
1
8
5
8
4
5
4
11
10
I
9
11
14
15
16
30
33
'4
25
20
25
21
20
30
15
31
15
14
17
23
11
11
14
10
11
10
8
3
5
!
4
4
2
4
3
1
2
2
1
1
1
1
100
VARIATION IN LEPTINOTARSA.
Table 35 — Continued.
(D) PLACE VARIATION IN THE PROPORTION OF LIGHT TO DARK COLOR IN THE
VENTRAL SURFACE OF THE ABDOMINAL SEGMENTS.
Year.
Gener-
ation.
Class.
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
•9
20
1903 {
1904 |
1905 {
I
I
4
2
3
1
1
5
10
4
8
2
4
7
14
5
11
3
£
27
II
16
6
12
15
17
17
21
IO
15
20
II
20
14
11
35
16
8
16
11
22
16
12
5
10
8
31
8
7
3
8
4
6
3
5
6
2
4
2
1
1
4
1
Table 36. — Place variation in proportion of light to dark color in L. signaticollis at
Cuernavaca, Morelos, Mexico, by classes.
Year.
Gener-
ation.
In elytra.
In epicranium.
In pronotum.
On ventral sur-
face of abdom-
inal segments.
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
I903 ■•
1904 |
1905 {
2. . . .
I....
2. . ..
I
2. . . .
IO
20
21
21
II
55
60
34
55
71
35
20
45
24
18
I
3
1
3
1
1
8
2
11
15
5
12
15
26
21
31
21
33
52
60
54
50
45
7
3
9
6
4
I
3
2
2
3
6
3
3
8
8
10
11
9
20
17
20
17
21
58
56
50
57
55
II
13
14
12
IO
I
I
I
I
I
3
4
1
96
99
94
100
98
Table 37. — Place variation in piopoiiion of light to dark color in the epicranium, prono-
tum,elytra, and abdominal segments of L.undecimlineata at Orizaba, Vera Cruz, Mexico.
(A) VARIATION IN PROPORTION OF LIGHT TO DARK COLOR IN THE EPICRANIUM.
Year.
Gener-
ation.
Class.
1
2
3
4
5
6
7
8
9
10
11
12
13
H
15
16
17
18
19
20
1903 . .
1904 |
1905 {
2. . . .
I
6
2
IO
18
17
7
3
29
43
21
28
27
50
26
49
53
62
IO
7
11
12
8
(B) VARIATION IN PROPORTION OF LIGHT TO DARK COLOR IN THE PRONOTUM.
1903 ••
1904 1
1905 {
I
3
2
8
7
3
7
2
10
11
14
13
5
26
3°
28
39
28
41
38
49
36
51
12
11
6
5
14
PLACE VARIATION.
Table 37 — Continued.
(C) VARIATION IN PROPORTION OP LIGHT TO DARK COLOR IN THE ELYTRA
IOI
Year. GaZT
ation.
Class.
1
2 3
4
5 6
- ■
9
10 11
12
13
h
15
16
17
18
19
20
1903 . . 2
1904 { | 2
1905 1, 2....
6 12
I
21
5
4
4
4
45 12
23 55
2S 50
32 40
36; 39
3
14
15
21
20
r
1 . .
ID) VARIATION IN" PROPORTION OF LIGHT TO DARK COLOR IN THE VENTRAL ABDOMINAL
SEGMENTS.
1903 ..
1904 {
1905 {
2
2...
I
I
1
I
I
5
1
1
100
98
93
99
98
Table 38. — Place variation in the proportion of light to dark color in the elytra of
L. dilecta at Pitebla, Puebla, Mexico.
Year.
Gener-
ation.
Class.
t>
7
8
9
1903
2. . . .
2. ...
I
7
84
72
S
I9°4
6
In some of the species of the genus, as, for example, in L. signaticollis,
there is relatively little place variation. In the individuals of this species
which were obtained at Cuernavaca in the years 1903, 1904, and 1905,
almost no variation is found from year to year in the color proportion of the
epicranium and pronotum, and no change of mode. (See table 36.) For L.
undecimlineata, found at Orizaba, Mexico, in the years 1903, 1904, and 1905, *
there appears a low place variation, with no modal change of the parts or of
the entire species. (See table 37.) In L,. dilecta, found at Puebla in the
years 1903, 1904, and 1905, there is no place variation in the color of the
pronotum, nor in size, but considerable is found in the elytra, with slight
modal fluctuations. (See table 38.)
Enough data has been presented to show conclusively the existence of place
variation in this genus of beetles, and also that some species are more variable
in this respect than others. It is perfectly obvious, therefore, that in any
'The records for 1906 at Orizaba, Mexico, show no change in the place variation of
/,. undecimlineata.
102 VARIATION IN LEPTINOTARSA.
study of variation, whether individual or geographical, or for whatever pur-
pose, but especially in evolution studies, one may be easily led into serious
errors by the gathering of material here and there, and at one time only.
We must first discover whether this phenomenon exists in the form chosen
for study, by observations covering several generations, and if it is found
its limits must be determined as near as possible. This variation may be
expressed by any of the variation constants, or by oscillations of the empirical
mode expressed in deviations in classes from the empirical race mode, i. e.,
the empirical mode of the species.
The causes of place variation in any species are the fluctuations from time
to time of the individual factors in its environmental complex. These are
moisture, temperature, and food commonly, and other factors less frequently.
It is possible in many cases to determine rather exactly the cause of a given
place variation, but the determination is not of any great value. Thus from
the data given it is possible to show that excessive variations were found
during this or that deviation from the normal.
As far as I am able to determine from observations and experiment, place
variations result in no permanent modifications, nor do the changes seem to
be inherited. They are produced in these beetles during the larval and pupal
periods, are somatogenic, and purely fluctuating. Place variation results fre-
quently, as can be seen in the data given, in the production of prophetic and
historic skewness ; indeed, in the material which I have studied I believe that
most of the cases of skewness are due to this cause and do not have the
meaning that might be attributed to them.
This place variation must necessarily be a troublesome factor in the study
by biometric methods of evolution, geographical variation, or selection. In
my own work it has been the rock upon which many cherished schemes have
been wrecked, and I suspect it has not yet completed its destructive work. If
one would study any of the broader problems of evolution by biometric meth-
ods he must first of all determine whether in the material chosen for study
this phenomenon exists, and if it is found, too great care or too long a time
can not be spent in the elimination of this factor. At present I know of but
one method of doing this — that is, by collecting data and material over a suffi-
ciently long period to determine the range of this form of variability. Un-
happily this demands time, patience, and often funds which the investigator
will not or can not afford. The failure to take into consideration this place
variation vitiates the validity of a large part of the biometric work that has
been done, and there is no reason to think that it will be otherwise in the
future. The consideration of this phenomenon will in a very large percentage
of cases make biometric work and the statistical study of variation and evolu-
tion efforts of much longer duration than hitherto, but the demonstration of
the existence of place variation in no wise vitiates the results which may be
obtained by biometric methods when this phenomenon is given due consider-
PLACE VARIATION. IO3
ation. We must use care and discretion, for biometric expressions are onlv
a little more precise, only a little less liable to error, than qualitative expres-
sions of biological facts.
Shull (1904) well remarks: "To some these results may seem to preclude
the possibility of deriving- anything of further value from quantitative studies
of variation, while to others many new problems of great interest and impor-
tance will be suggested."
As yet we know very little of the laws which govern this phenomenon, or,
indeed, of the part which it may play in evolution. In my own studies, data
have been gathered which point to the importance of this phenomenon, and
may indicate some of the laws by which it is governed.
The data in the foregoing tables show conclusively that place variation
affects every part of the animal, and that not all species are alike in their
variability. For example, dcccmlincata and inultitcriiiata are highly variable
in this respect, while signaticollis, iindecimlincata, and dilecta are among the
least variable of all the species. If we compare the place variability of these
species with the individual variation of the entire population it is at once
apparent that the species with a high individual variability also present a high
place variability. Moreover, the highly variable species deccmlineata and
multitceniata are also widely distributed, while those of low variability are
more restricted in their habitat, as signaticollis and dilecta. Hence it follows
that wide-ranging species, which are more variable than those of restricted
range, are also more variable in one place than are those of narrower distri-
bution. This wide distribution, high individual, and high place variation all
indicate in the species a high degree of sensibility to stimuli, and an adapta-
bility and lack of conservative tendencies in the population.
This difference between species we may examine to advantage in the speci-
mens of multitceniata obtained at Guadalupe, Federal District, Mexico, and
those of signaticollis from Cuernavaca, Mexico. One lies on the east of the
Sierra de las Cruces, the other on the west. One lives in a habitat where the
yearly rainfall is from 16 to 20 inches, the other where it is from 50 to 60
inches. In the first generation of multitceniata in 1903 the polygon of place
condition was fairly normal, while that of the second generation was strongly
skewed prophetically. During the growth of the first generation conditions
were average ; during that of the second the rainfall was far in excess of the
average. We may attribute this place variation to the increased rainfall,
although it is possible that it was only the indirect cause. During this same
season signaticollis, although subjected to the same variation in precipitation,
showed no increased place variation. In the following year (1904), again,
under average conditions, the polygon of the first generation in multitceniata
showed a historic skewness and that of the second a normal condition, while
signaticollis remained invariable. Of the two species one is immensely more
susceptible to changes in environment, undergoing modification very readily.
8— T
104 VARIATION IN LEPTINOTARSA.
In the section on variation I have shown that these two specie?, decemline-
ata and multitarniata, are the parents of many extreme variates (mutants).
In a species which is wide-ranging and highly variable, both individually and
as to place, which is highly susceptible to environmental change and shows
strong place variation, is there any relation of this phenomenon to the pro-
duction of sports or mutants? and, if so, what? As far as my evidence goes,
it is certain that in the second generation of 1903, at Guadalupe, the per
cent of the form melanothorax was 2.4, or more than ten times the normal
under average conditions at that place. Now, this was not accidental, for
in the same year and in a corresponding generation the per cent was 3.1
at Puebla, which is more than thirty times the normal proportion of this
variation at that place. The condition in the environment produced a wide
place variation at the time and a larger percentage than normal of extreme
variations ; whereas in succeeding generations the place variation was oblit-
erated and the normal per cent of melanothorax appeared. These observa-
tions would point to the conclusion that the change in the population for a
given generation due to place variation gave opportunity in the deviation of
the entire population from the normal, for a larger percentage to appear as
extreme variations. From observations in nature it is difficult to tie results
to special causes, to be sure that what we see happen is the result of this or
that cause. Likewise in this case it looks as though there were an important
and fundamental relation between this place variation and the production of
numerous sports or extreme variations in multitceniata. In 1904 I observed
that at Tierra Blanca the unusual condition of excessive precipitation pro-
duced angiistovittata from imdccimlineata in abnormal proportion, and the
same has been found in deccrnlincata. De Vries found that in some years his
Oenothera lamarckiana mutants were more abundant than in others. Is it
not possible that had he investigated he would have found here also this same
relation ? If this relation between place variation and the abundant produc-
tion of extreme variations in plastic species be generally true, and it certainly
is for this genus of beetles, are we not a step nearer to understanding better
the nature of mutants and the manner in which they are produced? Is it not
possible also that here too we have a clue as to how we shall proceed in order
to produce experimentally these variations?
Place variation must, I believe, be regarded as a phenomenon of impor-
tance, not only in its possibilities for producing false appearances, but also
because of the relation which it seems to bear to the appearance of extreme
variations or mutants. Whether or not it is of any moment in evolution by
slowchangeswe are not yet in a position to determine, but our evidence would
seem to indicate that it is not, as its flucuations are about equal on either side
of the average. Long-continued observation of this phenomenon would be
highly important, as it would perhaps go a long way towards settling some
disputed points regarding the method of evolution in animals.
GEOGRAPHICAL VARIATION.
I05
The following laws have been found to hold concerning place variation, as
far as the present investigation goes :
(1) Place variation is a universal phenomenon, but it varies in its extent,
being greatest in species of wide distribution and high individual variability
and least in species of restricted habitat and low variability.
(2) Place variation is productive of prophetic and historic skewness in
the polygone of distribution. (Unless this factor be accounted for in skew
polygons we may safely attribute skewness to this cause.)
(3) The extent of place variation shown by a species is a good index of
its susceptibility to external stimuli.
(4) In place variation, whenever there occurs an extreme oscillation of the
population there is an accompanying production of an unusually large per-
centage of extreme variations or mutants.
GEOGRAPHICAL VARIATION.
Variation in general color, or a tendency toward melanism or albinism, is
a common and long-recognized phenomenon, and one exhibited by almost all
wide-ranging genera. This variation has been attributed to changes in
environmental factors, such as temperature and moisture, to latitude or longi-
tude, to altitude, and more rarely to topography and other natural conditions.
In Leptinotarsa abundant variation in the general hue is found from place to
place, which may be attributed to moisture, temperature, food, sunshine, soil,
and other natural causes. There at least exists a striking correspondence
between the variation of these beetles and the changes in their environment.
In the study of the geographical variation of this genus the statements of
the conditions found in different places have in every case been based upon a
number of generations taken from one locality. In this way the error which
might otherwise be produced by place variation is, I believe, to a considerable
extent eliminated.
The data of geographical variation as it has been worked out is most easily
presented in tables, in which 20 equals total melanism and o total albinism.
Tabi.e 39 —Geographical variation in the general color of Leptinotarsa.
Cass..
Per cent of frequencies.
No.
of
gen-
era-
tions.
I
2
3
4
5
6
7
8
9
10
II
12
13
14
15
16
17
18
19
2m
/.. undeeimlineata :
4
12
16
2
36
10
24
19
2
4
1
I
8
11
13
8
5
6
10
2
2
2
7
1
4
5
I
5
4
2
2
Costa Rica
21
33
22
17
25
ill
5
>7
22
is
24
6
39
.i*
53
50
3°
25
36
53
50
-,11
1
13
12
21
12
'5
66
10
14
12
1.1
1
3
7
4
4
6
5
3
■
Oaxaca
Tierra Blanca, Vera Cruz...
Orizaba, Vera Cruz
L. dt versa :
I
L. signalicollis :
1
3
5
4
4
3
8
8
8
9
1
2
2
3
I
106 VARIATION IN LEPTINOTARSA.
Tablk 39. — Geographical variation in the general color of Lebtinotarsa — Continued.
Class....
Per cent of frequencies.
No.
of
gen-
era-
tions.
1
2
3
4
5
0
7
8
1
1
5
7
9
3
5
i'
11
1
6
5
10
4
14
■in
as
18
20
5
3
1
4
2
7
11
12
30
20
24
14
■j.-,
ii
15
5
9
11
10
18
12
;■.>
1 ;
'7
30
2;
21
as
is
2L
17
23
I".
<3
21
11
10
12
20
'5
■7
22
lo
!.-.
39
10
20
'4
12
9
8
4
13
9
8
11
19
21
20
21
6
15
1.
5
1
9
4
4
9
10
11
9
4
I
16
5
2
I
'7
2
18
2
19
1
20
L. multittxniata ■
6
5
4
3
5
4
2
4
1
4
2
5
4
5
1
2
4
5
4
7
3
2
5
4
6
16
6
4
7
3
8
2
3
6
4
5
2
2
6
5
3
4
4
3
7
2
2
4
2
2
3
2
3
2
4
5
2
2
3
4
4
2
b
I
1
4
2
2
1
2
1
7
4
2
1
3
L. oblongata ;
1
1
1
L. rubicunda ;
1
2
100
L. intermedia :
1
2
is
17
19
1
2
2
6
21
11
Ii
6
9
1
3
24
5
6
21
19
18
6
2
11
7
5
13
15
5
6
6
18
24
23
21
2
6
2
20
I
-.s
52
14
12
3
6
;ss
2
2
8
31
is
10
5
20
11
9
IQ
::i
11
14
is
i*
a,
8
5
10
1:
is
19
21
8
23
1
9
3
L. decemltneata ;
1
3
4
10
1
19
\»
5
1
3
20
22
15
14
is
31
17
is
21
14
30
.11
22
17
20
3
14
31
30
20
1
1
11
14
i7
27
21
IS
30
20
9
39
19
iS
13
11
9
1
30
19
20
15
9
3
20
34
12
14
15
11
2
18
>5
15
4
5
4
2
2
11
14
4
10
II
2
1
10
8
2
2
1
1
1
6
3
2
3
7
1
4
1
2
4
2
4
9
2
2
1
10
ii
11
il
2
2
1
2
1
1
3
1
Ontario and western New
7
5
6
2
I
3
1
1
I
1
2
3
20
1
6
2
14
3
20
15
14
2
M
7
7
2
10
5
6
1
5
1
1
1
1
1
2
60
3
5
*9
24
<;■;
4
51
5
85
so
83
71
40
11
36
en
70
28
71
I
2
ss
83
5
1
8
15
15
23
00
r,s
10
1
2
3
2
5
4
2
i
r
22
L-,
Ill
25
9
13
2
1
82
1
1
20
2
1
3
3
IS
64
21
60
11
12
10
11
22
28
32
31
10
10
9
1
L. juncta :
1
3
IS
17
3
I
I
4
2
1
1
4
20
18
22
1
1
4
2
4
13
14
19
2
1
2
1
2
9
8
9
1
L. dejecta :
Middle and western Texas
:■
1
2
L. dtlecta :
1
L. lineolata ;
6
2
2
II
10
6
L. haldemani :
L. "violescens :
'.Ml
85
L, dahlhomi ;
10
16
5
L. hbatnx ;
3
4
GEOGRAPHICAL VARIATION. 107
Table 40. — Geographical variation in the pronotum of Leplinotarsa.
Class...
Per cent of f reqneucies.
No.
of
gen-
era-
tions.
1
2
3
I
5
0
:
1
s
3
9
11
1
10
17
2
1 [
3d
S
2
4
I
12
20
21
3
5
2
2
1
13
9
10
10
9
1
5
9
M
3
20
34
22
i<
26
III
5
17
23
17
2.
12
8
7
5
12
10
S
11
18
20
21
21
5
15
16
17
18
19
20
/.. undecimlineata :
2
2
2
7
1
4
5
r
5
4
2
2
6
5
4
3
5
4
2
4
1
4
2
5
4
5
I
2
4
5
4
7
3
2
5
4
6
16
6
4
7
3
8
2
3
6
4
5
7
2
2
6
5
3
4
4
3
7
2
2
4
2
2
3
2
3
2
4
8
II)
fJ
sa
15
29
2 j
-.c,
-.1
i-.i
-.0
7
7
6
1 1
i'
2u
17
17
JO
9
' 1
1 1
14
6
2
1
1
1
2
5
5
5
4
1
L. diversa :
I.. sigiiaticoHis :
2
2
I
4
1
[3
:s<>
21
20
15
;-
:i
19
5
'J
it
12
18
5
4
3
2
is
IS
15
3a
18
21
■>.<.>
21
20
21
20
15
9
;
11
13
20
'5
9
1 1
1
14
17
21
11
35
if,
f'.l
21
1
2
2
L. mullitcpniata :
1
2
2
2
3
2
5
14
5
1.;
IT,
39
5
is
2'''
I
I
2
1
3
7
2
2
1
1
1
i"
3
5
S
6
2
2
7
4
1
8
3
L. oblongata :
10
4
1
1
L. ritbtcunda :
3
1
99
L. intermedia :
1
1
45
16
19
2
1
2
6
12
IO
5
10
1
3
24
9
6
22
19
17
5
3
10
8
4
id
15
6
X
9
15
2C
24
20
2
6
4
2>
I
5S
.1
I-i
12
4
8
3 s
I
I
9
31
30
10
4
20
1 1
o
18
:ji
9
5
18
■-is
27
2".
5
10
1-
17
20
I
1
4
20
20
16
14
ir,
30
21
at
20
16
IS
30
23
17
20
2
H
11
30
[9
35
9
20
1
II
2
I
1
IS
III
4
f. decemlineata :
1
4
5
9
1
1
I
ID
14
27
24
19
::o
21
10
31
20
14
11
10
1
30
■9
19
16
1
10
5
21
3 1
16
14
17
12
2
18
18
15
3
4
5
1
20
1?
2
3
2
II
I -I
I
10
9
1
1
9
lb
4
2
3
I
9
3
1
5
7
1
I
3
3
3
3
4
4
9
1
1
4
Ontario and western New
6
8
2
I
1
4
6
I
2
13
15
13
2 1
1
I
3
19
1
2
7
15
15
S
8
2
9
6
4
I
4
I
1
4
9
20
14
24
61
.7
11
2
2
55
4
t
21
60
1
50
4
S5
sr»
NO
si
no
u
H
37
/.. juncta :
4
J
Z(
1
20
18
2
2
1
2
1
■1
6
19
*;,-»
GO
10
12
II
I'
2.
28
31
3t
3
»9
17
21
2
5
1
1
14
16
IS
2
1
11
K
L. defect a :
Middle and western Texas
3
1
2
3
3
2
■i
4
1
I
2
6
L. dilecta :
I
I
IC
II
3
L. lineolata :
3
i
io8
VARIATION IN LlCPTINOTARSA.
Table 41.— Geographical variation in the epicranium of L. multitaniata, oblongata,
intermedia, decemlineaia, rubicttnda, and mclanothorax.
Class...
Per cent of frequencies
No.
of
gen-
era-
tions.
1
2
I
1
3
4
5
6
7
8
9
10
L. multitaniata:
7
5
3
2
7
8
8
7
8
9
13
14
7
11
14
7
68
60
73
80
1
IO
17
24
27
15
18
27
12
11
14
10
9
5
M
21
34
35
18
25
28
23
2
3
1
1
15
20
30
12
14
23
30
12
33
1
1
31
8
6
1
34
8
2
15
2
1
3
1
6
5
4
3
5
4
2
4
I
4
2
5
4
5
1
2
4
5
4
7
3
2
5
4
6
16
6
4
7
3
8
2
3
6
4
5
7
2
2
I
I
1
5
5
11
4
It
'7
12
7
6
L. oblongata .■
I
I
2
I
3
4
3
2
L. rubicunda ;
73
1
6
I
98
L. intermedia :
5
15
24
58
12
60
21
14
1
58
54
15
6
21
48
51
5
2
3
13
8
51
58
60
61
48
47
63
60
27
40
42
43
50
■7
18
2
2
4
19
23
14
7
17
59
58
24
19
20
14
19
19
14
19
5
21
26
22
20
4
3
I
I
L. decemlineaia ;
7
11
18
69
14
2
1
1
S
9
21
14
61
18
20
S
9
5
4
8
7
8
8
z
14
10
12
12
I
2
45
18
13
8
9
1
4
1
1
12
47
S
I
5
1
I
10
I
I
2
1
13
9
13
18
20
21
11
d
13
11
10
14
2
1
1
2
3
4
1
2
•A
2
1
I
Ontario and western New
York
1
I
2
3
3
1
I
4
8
7
9
3
I
2
2
I
In the data presented in tables 39 to 41 certain features of geograph-
ical variation are apparent. Some species are more variable than others, and
of these more variable species those of wide range are generally most vari-
able and those of limited range least variable. This, however, is not always
true. Thus, undecimlineata appears in all of the tables as a wide-ranging spe-
cies, found from Mexico to Panama, yet it is relatively invariable ; whereas
dahlbomi is also a wide-ranging species found from Guatemala to Texas
and is highly invariable, although it lives in many different environments.
On the other hand, multitceniata is highly variable and decemlineata only a
little less so, while the species of the dilecta and haldemani groups present
only a moderate amount of variation.
GEOGRAPHICAL VARIATION. ICXJ
Fable 42. — Geographical variation in the ventral surface of Leptinotarsa.
Per cent of frequencies.
No.
of
gen-
era-
ions.
Class.... 1
2
3
1
5
6
7
8
9
1
10
1
3
3
2
11
1
8
7
8
3
1
1
1
10
IN
It
".I
I
1 2
5
22
21
20
7
3
3
7
II
1 J
23
22
i
13
24
li
13
m
21
30
26
22
23
22
7
3
14
to
17
is
17
Is
Is
-.11
52
5
h
1
[8
5
6
7
II
17
IS
':
2
7
it,
4
2
I
I
3
1
2
I
I
I
17
3
1
18 19
1
L. multitccniata .■
1
1
6
5
4
3
5
4
2
4
1
4
2
5
4
5
1
2
4
5
7
3
2
2
5
4
6
16
6
4
7
3
8
2
3
6
4
5
7
2
2
6
5
3
4
4
3
7
2
2
1
1
'
I
1
L. oblongata;
1
I
2
2
7
4
5
6
'9
19
1
...
1
L,. rubicnnda ;
12
60
20
4
1
2
117
L. intermedia
1
S
I
I
2;
14
3
2
9
b
7
'7
16
8
11
10
8
21
5
3
9
10
9
4
60
-.1
15
12
S
4
6
5
22
24
ID
13
iS
16
7
6
1
I
1
9
2
16
1 ■
L. decemlineata ;
5 6
3 in
12
to
2
2
I
S3
32
10
4
2
24
1.-.
II
3
7
10
20
B
7
18
18
19
II)
IJ
18
16
16
18
20
2
10
11
16
16
j
2
14
12
13
II
43
20
19
43
4.-,
J 5
4 0
8
4
2
30
28
2 2
20
21
5
4
20
20
21
25
3
3
I
.!•.
II
8
5
4
3
4
4
5
4
5
1
1
7
8
1
1
2
1
I I
'
3
2
2
1
3
4
2
I
I
I
5
S
lr.
3
5
4
9
8
3
2
1
2
12
51
19
1
1
2
1
1
1
1
1
Ontario and western New
1
1
1
1
4
17
10
2
17
1 -i
4!
II)
1 1
18
'9
20
21
5
6
6
4
4
2
2
1
1
2
3
6
4
22
15
16
60
r.s
11
I
1
-,11
1
L. juncta ;
I
5
1 5
1 2
3 19
6 20
196O
60 3
IS 52
60 6
18
18
2
1
]
9
3
1
3
4
25
17
21
L. defecta ;
Middle and western Texas.
Indian Territory
Southeastern Colorado
I
4 8
1 9
1
2
Table 43. — Geographical variation in the color of the elytra of L. dilecta, calceata, and
lineolata.
Class...
Per cent of frequencies.
No.
of
gen-
era-
tions.
6
7
8
9
10
L. dilecta .
2
3
1
2
3
81
9
Io
4
7
2
80
T5
9
14
11
4
2
2
3
2
3
2
4
7 80
1
8
I
2
86
85
90
10
8
10
L. lineolata :
I
1
IIO VARIATION IN LEPTINOT ARSA.
In decemlineata the range is from an albinic condition found in the Rio
Grande Valley and in the southern part of Arizona to the highly melanic con-
dition found on the Atlantic coast. From Colorado eastward into Illinois and
Ohio and on farther to the Atlantic coast there is apparent an increasing ten-
dency toward melanism, which culminates in New England and the maritime
provinces of Canada. Along the Great Lakes-St. Lawrence Valley this same
tendency is found, whereas toward the South an albinic tendency prevails
which is very distinct to the southwest and less so to the southeast. In this
respect the facts presented are in no wise different from those that have long
been known, that all animals in the United States are in general more melanic
to the north and east, less so to the south, southeast, and middle west, and
albinic in the extreme south and southwest. As long ago as 1877 Allen recog-
nized this tendency in birds and mammals, and Enteman finds the same to
be true in Polistes. L. decemlineata, however, is a form that has in recent
years invaded the area in which it is now found, and since becoming a mem-
ber of the fauna of a wider region, has responded to the conditions of the
new environment, and in adapting itself to new habitats has varied in the
same general directions as do the other animals with which it lives. It
is, therefore, to the environment and not to the animal itself that we must
look for the stimuli which are the cause of these modifications. The factor
assumed by all to be the most powerful in this general color variation is
moisture, and probably temperature is of equal importance. I have gathered
some interesting data on this subject concerning decemlineata.
Live material was gathered from various parts of the United States and
Canada and was reared at Chicago under various conditions. The specimens
which will be considered particularly were from Massachusetts, Georgia, Ala-
bama, Kansas, northwestern Texas, Dakota, and Manitoba. Compared with
the material collected at Chicago the beetles from these various localities pre-
sented diverse appearances and conditions of coloration. At Chicago, during
the years 1902 to 1905, all were reared under uniform conditions, the result
in all cases being a complete and rapid change to the prevailing condition of
coloration at Chicago. In all but about 1 or 2 per cent of the specimens this
change was produced in one generation, and at most two were required ; nor
was there any tendency toward reversion in succeeding generations when
reared under the same conditions.
At the same time beetles from Chicago were in experiment subjected to
the same conditions of temperature and moisture which obtain in different
sections of America, with the result that in one generation modification to the
conditions characteristic of the different regions was produced, and this per-
sisted as long as the conditions remained unchanged. In this way I have
experimentally modified the general color of these beetles in one direction or
another at will, producing by experiment the conditions found in nature in
GEOGRAPHICAL VARIATION. Ill
diverse habitats. These modifications, however, are no more permanent than
those found in nature. Indeed, the conclusion I have reached is that the geo-
graphical variation is due solely to place variation, which in a particular local-
ity is rather constantly in one direction. As far as this species is concerned,
although marked geographical variations are found, the evidence is conclu-
sive that they are not permanent, and that although the species varies in dif-
ferent localities it has not undergone any constitutional change. In Colorado,
New Mexico, and western Texas decemlineata has lived for some hundreds
of generations. At Chicago it has been a resident of constant conditions from
1865 to 1905, a period of 40 years or 80 generations. At this latter place it
varies considerably in general color from the condition in Colorado and
western Kansas, and yet a period of 80 generations under uniform conditions
has not been sufficient to impress upon the species any degree of change.
Likewise the long period of time that it has lived in the Southwest has not
changed the constitution of the species. As far as the evidence goes, and
I believe that in this case it is remarkably complete, since so much is known
of the history of the species, it seems certain that this form has not been
altered by its existence in any one of a dozen different habitats, and that it has
retained the same constitutional structure and character. Place variation only
is found, and this is in a constant direction as long as environmental condi-
tions remain constant and change only as they change.
The geographical variations in general color shown by L. multitccniata are
of the same nature as those found in decemlineata, and are altered or pro-
duced at will in experiment. As far as I can determine they represent, not a
permanent condition in the species, but merely a place variation which is
constant as long as the conditions of existence are constant and varies as fast
and as frequently as its environment varies. An important but seemingly
unanswerable question in this connection is, How long must a species be sub-
jected to given conditions before it will be permanently changed in a given
direction? How long must decemlineata live at Chicago in order that the
general color conditions characteristic of Chicago may be so firmly impressed
upon its constitution that when it is transplanted into slightly different environ-
mental conditions it will retain its Chicago character, at least for some consid-
erable time? The evidence from this investigation would indicate that varia-
tions produced by different environmental complexes produce only tempo-
rary variation and are not incorporated in the constitution of the species, at
least not in some hundreds of generations. L. multitccniata has certainly occu-
pied its present habitat for a long period of time, at least during thousands
of generations, and yet in all this time the place variation tendency of differ-
ent localities has not become incorporated in the constitution of the species,
and it is tc-day as variable in this respect as it probably was at the beginning.
112 VARIATION IN LEPTTNOTARSA.
Strongly contrasted with the species just discussed are forms like L. unde-
cimlineata and dahlbomi. Undccimlineata, while variable, does not show
even in its wide range of habitat any great geographical variability, nor has
it been possible in experiment to produce any extreme modifications. L. dahl-
bomi, of all species in the genus, has been found in the most varied habitats- —
in the hot, humid valleys of Guatemala and southern Mexico, in the dry, cold
plateau, in the dry, hot Rio Grande Basin, in Yucatan, and in other diverse
places, yet in all it is the same conservative species, and as far as known is
almost free from variation of any kind.
If the data of geographical variation were to be presented for the different
elements of the color pattern, for size, variation in punctation, and for gland-
ular openings, it would but be a repetition of what has just been said concern-
ing general color variation in these beetles, and nothing would be added to
the value of the above discussion.
It appears that geographical and place variation are closely related; that
place variation is in reality the cause of geographical variation. Of most
importance, however, is the fact that, as far as it has been possible to deter-
mine, geographical variations are transient and form no permanent part of the
species. They are, as it were, a suit of clothing used as the weather dictates,
and produce no more permanent constitutional modification than does the act
of changing one's dress.
Some authors, as, for example, Allen, have attempted to establish a rela-
tionship between latitude and longitude and variation. As far as this material
is concerned, any such relationship is purely incidental and of no importance.
Correspondence between geographical variation and any artificial scheme for
dividing the earth's surface by arbitrary lines would, on the employing of
enough imagination, show a relationship. The only real relations of geo-
graphical variation that I can discover are those to natural features of the
earth's surface, to climate, topography, and other natural phenomena.
We have already seen the relation which exists between place variation and
geographical variation, and I have shown that place variation is due to the
varying conditions of different environmental complexes. In the tables given
are found examples of the most exact correspondence of variation to differ-
ent portions of the habitat. These relations will come out better and in their
true light in the next section of this paper.
EVIDENCE CONCERNING EVOLUTION. II3
DISTRIBUTION AND VARIATION AND THE EVIDENCE THEY
AFFORD CONCERNING EVOLUTION.
In the first chapter and in the first portion of this one has been presented
the data of distribution and variation, these being the two sources from which
we are prone to draw extensively for evidences of evolution. When taken
alone, as has been seen, neither is able to yield much that is certain concern-
ing evolution. Considered together, they may yield important information
concerning evolution in organisms. Distribution and variation often, nay,
almost always, are able to give rather exact ideas as to what the evolution in
a given group has been. In this section we shall examine our data first, for
evidence concerning the evolution of the genus Leptinotarsa, and second, for
the processes which brought about that evolution.
We can most easily and rapidly arrive at some conclusion as to the evi-
dence which the study of distribution and variation affords concerning the
evolution and phylogeny of this genus by examining a specific case with care.
Let us first consider the evidence in the lineata group.
The distribution of this group of 12 species of beetles has been described
in the first chapter and figured on plate I. Considered from the standpoint of
their distribution, the whole group may be separated into three sets of species,
each set characteristic of a certain area.
It has already been shown how all available evidence points to the Guate-
mala-Chiapas plateau as the center of origin of this group of beetles. From
this point uudccimliiicata can be traced northward to the escarpment of Mex-
ico and southward to Panama. Its variations are greater in the extremes of
its range than in its central portion, but these variations are, as has been
shown in this paper, fluctuating and place variations, as far as can be deter-
mined at present. On the north of the habitat of undecimlineata two closely
related forms occur, signaticollis and angnstovittata, the latter being a varia-
tion from undecimlineata. We have seen how these three forms, in their posi-
tions on the slopes of the Mexican highlands, are, first, undecimlineata at the
lowest points ; second, angnstovittata midway, and third, signaticollis at the
lower edge of the escarpment. Among these three forms there are many
close resemblances ; the chief differences are those in the elytral ornamen-
tation, the color pattern, and the punctation. These facts argue strongly
that these three species represent one line of descent. In this case we have
the rather unusual observation of one species arising directly from another,
which lends most convincing support to our argument. But to consider
only the facts from distribution and variation. The distribution of the three
species like windrows along the slope of the Mexican highlands, and the fact
that they show in their variations a series of modifications which overlap and
are in one direction, offer a condition not easy to explain otherwise than upon
the assumption that the species and their variations as we now find them rep-
ii4
VARIATION IN IvEPTINOTARSA.
resent a line of development in which the three species are steps in a certain
direction of evolution. It would be difficult to invent a hypothesis to account
for this case other than the one advanced above that would not contain obvious
absurdities.
The species multitceniata, oblongata, intermedia, melanthorax, rubicunda,
and decemlineata, which are very closely allied forms, are all, excepting
d ec emlin eata, confined entirely in their distribution to the Mexican table-lands.
This is suggestive. As I have shown in the first chapter, we may regard mid-
titceniata as the central species in this group, variable, plastic, and able to live
in diverse habitats. It has spread over a wide area, and as it has been dissem-
inated new species have been produced. In all their variations this series of
species shows the same kind and direction of modification ; they differ, how-
ever, in the extent of their variation and in the portion of the possible scale
thereof which they exhibit. Thus, multitceniata, intermedia, and decem-
OBLONQATA
RUBICUNDA
DECEMLINEATA
t
INTERMEDIA
MULTIT/ENIATA
MELANOTHORAX
SIQNATICOLLIS
ANQUSTOVITTATA
DIVERSA
HYPOTHETICAL ANCESTOR
Text-figure 4.— Scheme of phylogenetic development of the hneata group,
based upon the data of distribution and variation.
lineata, form a series from north to south over the continent, which show
not only geographically arranged steps, but steps also in the variations
presented; from multitceniata, with variation entirely in the melanic half of
the possible series, through intermedia to decemlineata, with its variations in
the middle and albinic end of the possible variation series. Thus these three
species, and to the south, in like manner, oblongata, show differentiation in one
direction, both geographically and in their fluctuating variations. Melano-
thorax and rubicunda, which live in the same habitat with multitceniata, rep-
resent more local species than the rest, while juncta and defecta show one and
the same sort of a series in their geographical distribution and their variations.
Is it possible logically to derive any other conclusion from the above data
than that the conditions of distribution and variation as found indicate the
direction and order of evolution and the relationship of the species? Dis-
carding impossible hypotheses, we are forced to one of two alternatives —
EVIDENCE CONCERNING EVOLUTION.
"5
either that the condition is due to chance, and does not really represent the
orderly development of species which it would seem to, or that the data of
variation and distribution give the true order of evolution. It must be ad-
mitted that the data do permit of the construction of a plausible scheme of
the phylogenetic relationship of these species, such as the one shown in text-
figure 4.
A scheme like that given in text-figure 4 is highly plausible, and is apt to so
strike one's fancy that other possibilities are ignored, as, for example, the one
that the relations suggested by the data employed are due to chance, and have
no real meaning.
DISTINQUENDA
\
PUDICA
\
OBLITERATA
TVPOQRAPHICA
LINEOLATA
CALCEATA CACICA
^ /
DILECTA NITIDICOLLIS
NOVEMLINEATA
FLAVITARSIS
HYPOTHETICAL ANCESTOR
Text-figure 5.— Scheme of the phylogenetic development of the dilecla group,
based upon the data of its distribution and variation.
The dilecta group consists of a series of species which are widely dissem-
inated and variable and nearly as numerous as those in the lineata group. On
plate 12 is represented their distribution over northern Central America and
Mexico. Here also are found series of species which in their geographical
distribution and their variations show conditions similar to those in the lineata
group. The species flavit arsis from Guatemala, nitidicollis from Chiapas,
and cacica from the Atlantic slope represent, in as far as their known distribu-
tion and variations can indicate, a line of species differentiation from highland
to lowland and in a given direction. Another series consists of the species
dilecta from the Oaxaca-Guerrero highlands and northward over the South
Mesa of Mexico, with its lowland modification, calceata, and its close allies to
the north, lineolata and typographica. Still another series is that made up of
the species novemlineata, oblitcrata, distinguenda, and pudica. It has not
been possible to obtain as full data of the distribution and variation in the
dilecta group as in the lineata group, but all known facts point to species
formation in certain directions over the continent— one line of differentiation,
that found in the series Havit arsis, nitidicollis, and cacica extending to the
north and eastward ; another, represented by dilecta and its allies, spreading
northward, and so on. The scheme of the phylogeny of this group, shown
in text-figure 5, is based upon its distribution and variation.
u6
VARIATION IN LEPTINOTARSA.
This scheme has exactly the same chance of being true or false, and for the
same reasons, as has the one constructed for the lineata group. In a like
manner and upon similar data we may build up schemes of phylogenetic de-
velopment for the other groups of Leptinotarsa, or for the entire genus. In
text-figure 6 is represented such a scheme.
In text-figure 6 the phylogeny of the genus Leptinotarsa, as revealed by the
study of its variation and distribution, shows uniformly a species formation
EVIDENCE CONCERNING EVOLUTION. WJ
along certain definite lines and in correlation with the natural features of its
habitat. The question is whether the many series of species found represent
chance arrangements, or whether they indicate species differentiation along
definite lines, in correlation with the climatic and topographic areas of their
habitat. If in the genus there were only one or two such cases, we might
perhaps conclude that they were the result of chance; but when the entire
genus shows this phenomenon it is highly improbable that the conditions
could be due to chance modifications. On the other hand, it is almost certain
that the interpretation of the relationships and the direction of the evolution
of species indicated in text-figure 6 is, in the main, correct. In this genus, at
least, the study of distribution and variation gives clear ideas as to what the
course of evolution has been. We are fortunate in that there are few gaps
due to apparent extinctions and in the unusually large amount of exact and
reliable data available.
Even though the truth of the above conclusion is admitted to be incontest-
able, no evidence is produced as to the methods by which the evolution of the
genus has been brought about. As far as the above conclusion regarding
phylogeny is concerned, it could be explained by any one of the current
hypotheses concerning evolution, at least as far as any facts yet discovered
are concerned. Slow, accumulated variations and natural selection, directive
evolution (Eimer's orthogenesis), neo-Lamarckism, and the mutation theory
would each be able in the opinion of its supporters to explain the condi-
tions found.
Can we, in the study of evolution as found in this genus, derive from the
data of distribution and variation any evidence as to the method of evolution?
This is often attempted, but the conclusions are badly strained and distorted,
and not infrequently are entirely unfounded. Nowhere in the data presented
is there anything to indicate that slow variations and natural selection has
been the method of evolution. We have already considered this case in part
in the discussion of place and geographical variations, where it has been shown
that as far as can be determined slow variations in a given direction have
not been incorporated into the species, even after many hundreds of genera-
tions. The evidence derived from the place and geographical variations of
L. multitaiiiata and dccemlineata would be difficult to explain away by this
theory. Its advocates can, and probably will, take refuge in the argument,
frequently employed, that we have not evidence covering a sufficiently long
period of time ; but they have not yet defined what is meant by "a sufficiently
long period of time." We have in dccemlineata one of the most satisfactory
cases known of those which are calculated to give evidence on this point, with
complete data and actual specimens for many generations, together with mod-
ifications of the required kind. It is shown conclusively, howevei, that the
changes have not as yet become a part of the species. It may be suggested
that perchance we may be dealing with one of those rare and special cases
Il8 VARIATION IN LEPTINOTARSA.
which are difficult to explain on any basis ; but evidence does not appear any-
where in the data of variation and distribution in the whole genus in support
of the theory of evolution by slow variations and natural selection. Are these
series of species, with their differentiations, in direct and orderly succession
from their center of origin over the country, and the continuity of their varia-
tions, to be taken as certain evidence in support of this view ? In any one of
the many examples of this kind which could be mentioned is there any more
probability of their having arisen by slow variations and natural selection
than by orthogenetic mutation ?
In regard to the other hypotheses of the method of evolution, do they
secure any support from the data under consideration ? In the entire genus
there is found present to a marked degree evolution and fluctuating variation
in narrow limits in definite directions. Does it follow that orthogenesis (in
Eimer's sense) has been the method of evolution pursued by this genus of
beetles ? Excepting for the existence of definite directions of evolution there
is no evidence in support of this view. The determinate character of the
variations and species differentiation can be just as well explained on the basis
of slow variation and natural selection by the elimination of all variations and
species but those within narrow limits as upon Eimer's theory of orthogenesis.
It would be equally easy for the neo-Lamarckians to devise a plausible argu-
ment as to the origin of the genus by the method of evolution demanded by
their hypothesis, and likewise for the advocates of species formation by muta-
tions and segregation in fittest environment to make out of the data a complete
and satisfactory case in favor of their theory. We do know, however, that
certain extreme variations or mutants occur, as, for example, mclanothorax
and rubicunda, and while these may argue in favor of species origin by
mutation, in the remainder of the species no more such examples are known ;
and, as far as any evidence yet discovered is concerned, these have not pro-
duced permanent species. Moreover, the origin of the species in the genus
can be explained by any one of the hypotheses we may choose, or rubicunda
and mclanothorax may be passed by as exceptional cases and of little sig-
nificance.
Viewing the data and evidence from a strictly impartial standpoint, as one
having no more interest in one hypothesis than in any other, as far as I can
discern there is no certain basis for asserting what method or methods of evo-
lution have been followed in this genus of beetles. Some species, multita-
niata and dccanlineata, afford evidence strongly opposed to evolution by slow
variation, while others, melanothorax and rubicunda, support the theory of
evolution by mutations; but, although this evidence is suggestive, it is well
to keep in mind the fact that there are other possible methods of evolu-
tion. Although frequently far-reaching conclusions are based upon evidence
derived from these sources, such a proceeding is not justified. The neo-
Lamarckian is prone to draw from variations distributed in time and space
EVIDENCE CONCERNING EVOLUTION. 1 19
conclusions warranted from his point of view, but not from that of others.
So, too, the advocate of orthogenesis, following Eimer, selects stages in varia-
tion here and there, and isolated facts of distribution, and draws sweeping
conclusions therefrom. The supporters of selection can see no possibility of
other methods of evolution, and the advocates of mutation are wildly enthu-
siastic over their newly discovered treasure.
The only conclusion regarding the method of evolution that, from an im-
partial standpoint, I am able to draw from this investigation of variation and
distribution is that we can get no undoubted evidence to warrant a decision in
favor of any one hypothesis more than another. We can interpret the condi-
tions found by any of the current hypotheses; but explaining a condition by
an hypothesis is not the same as that the conditions found are evidence in
support of an hypothesis, although it is often so used. Although it is conclu-
sive that no direct evidence of value or reliability concerning the method of
evolution is to be derived from distribution and variation, three suggestive
facts come out in the clearest possible manner : First, that species differentia-
tion and variability have followed in the same lines or directions of evolution
as is shown in distribution and variation; second, that three forms, melano-
thorax, angustovittata, and rubicunda, arise as extreme variations and are
stable forms ; third, that there is strong evidence, in the case of multitaniata
and decemlineata, against the idea of the slow modification of a species by the
incorporation into its constitution of fluctuating place or geographical varia-
tions. These facts have been investigated experimentally and otherwise and
the results will be presented later.
The study of distribution and variation, although not affording conclusive
evidence concerning the method of evolution, iJom give rather complete and
reliable data upon the probable phylogenetic development of the genus, and
also information concerning the limits, directions, and lazvs of variation, dis-
tribution, and migrations. This information is important, nay, indispensable,
in the study of evolution, in that it shows us what the natural history of the
material under investigation has been. This knowledge of the natural history
of the genus I regard as of fundamental importance; and it can not, I believe,
be neglected xmthout seriously impairing the value of researches in experi-
mental evolution.
The conclusions to be drawn from this study of distribution and variation in
the genus Leptinotarsa are that species differentiation's far as it is possible to
determine, has been definite and not promiscuous, and that all variations are
definite and not promiscuous, and also that both species differentiation and
all variations are in the same directions and are in a most remarkable manner
correlated with natural features of the general habitat. The data of distribu-
tion and variation can give no real basis for determining the method of evolu-
tion, nor does it enable us to arrive at the causes thereof. We can arrange
the species of the genus in series, and opposite them place climatic and other
9— T
120 VARIATION IN LEPTINOTARSA.
environmental factors, in plausible juxtaposition, and base far-reaching con-
clusions thereon ; but in any such process there are far too many assumptions
that this possible factor and that observed result are cause and effect. The
fact is lost sight of that the causes are numerous, and their action only
remotely comprehended. Too often we neglect to consider tendencies inher-
ent in the species, or to learn its physiology, its peculiarities of development,
its habits and instincts, and much other information which is apparently of
slight consequence, but which is of great importance in the economy of the
species. It is necessary to know our material intimately and to experiment
widely before we attempt to correlate observed variations with environmental
factors or to determine the causes and methods of evolution.
CHAPTER III.
COLORATION IN LEPTINOTARSA.
In this chapter, which deals with the colors and color patterns in the genus
Leptinotarsa, I have attempted to follow the development of the color phe-
nomena, both ontogeneticallv and phylogenetically, and to determine the laws
by which they are controlled and the environmental factors by which they are
modified.
COLOR PHENOMENA IN INSECTS.
COLOR CLASSES.
The colors of insects are divided into three chief categories, dependent upon
the causes by which they are produced. These classes of color we designate
as chemical or pigmental, physical or structural, and chemico-physical or
combination.
CHEMICAL OR PIGMENTAL COLORS.
These colors owe their existence to the presence of a substance of a definite
chemical composition, which has the property of absorbing some wave-lengths
of light and of reflecting others. These compounds may be the product of
the metabolism of the animal, or derived from the food (Poulton, 1893), or
they may be accidental inclusions. They may also be colored substances or
coloring substances. The colors due to this cause are black, brown, orange,
yellow, drab, many reds, rarely blue, green, and white. The pigments which
produce them are soluble in various reagents.
It has been shown that the color-producing substances may be divided into
three main classes, which differ in their location in the body and in chemical
composition. It was demonstrated by Hagen that these color-producing sub-
stances are divisible into two classes, the dermal and the hypodermal, the der-
mal being located in the cuticula and the hypodermal in the cells of the hypo-
dermis. Poulton and others have shown the existence of color-producing
substances in the fat body and haemolymph. These I have called subhypo-
dermal colors. The dermal colors and the substances which produce them
are located in the outer portion of the cuticula, and are diffuse pigments, and
not present in the form of granules, as maintained by Hagen and Enteman.
The hypodermal colors have been shown to be usually in the form of granules
located in the hypodermal cells, or, more rarely, derived pigments.
The solubility of these pigments has been studied by Coste, Urech, Mayer,
Tower, and Enteman. From these researches it appears that the color-
producing compounds are all soluble, but not with the same degree of ease.
122
COLORATION IN L^PTINOTARSA.
Some of these pigments are very unstable and break down on exposure to
light or at death, while others are of great permanence and are soluble only
after the most vigorous treatment with strong mineral acids. The colors,
their location in the body, their solubility, and degree of permanence are sum-
marized in the following table:
Chemical
pigmental
colors.
(a) Cuticula or
dermal col-
ors.
(b) Hypoderrual
colors.
(c) Subhypoder-
mat colors.
Black, dark
brown, brown,
straw yellow.
Located in the
primary c u t i -
cula.
Permanent, insoluble, or solu-
ble only on long standing in
water, alcohols, ethers, oils,
weak acids, or alkalis. Solu-
ble on standing or with heat
in strong mineral acids and
alkalis, with dissolution of
the cuticula.
f Chrome yel- ") rn__,„H in hv f Permanent, insoluble, or solu
i i^.„ r *> A i i*ocaieu in ny- , ,. „;ti, ^:a;»..u„ ;„ „,„*„
(i)-f
low, red
ve rmilion
scarlet
blue.
pod e r in a 1
cells as gran-
ules.
(2)
Green,
yellow,
white.
Located in or
between the
h y podermal
cells as a dif-
f u s e pig- 1
meat.
Derived
pigments.
r r f * n ( Located in body )
yellow! \ ? a v i t y in fat I D e r i v e d
white. I
body and
haemolymph.
pigments
ble with difficulty in water,
alcohol, weak acids, or alkalis.
Soluble in ether and other
fatty solvents.
' Not permanent;
fade at death or on
exposure to light.
Soluble in water,
alcohols, etc.
Chlorophyll and
xanthrophyll de-
rivatives.
Not permanent. Lar-
val colors. Soluble
in water, alcohol,
etc.
The chemical composition of the compounds which produce color has been
studied by Krukenberg by spectrum analysis, and by this and other methods
by Hopkins, Griffiths, Urech, Poulton, Mayer, Tower, Dewitz, Enteman, and
others. The cuticula pigments are diffuse pigments elaborated from the pri-
mary cuticula through the action of an enzyme which I have isolated and
called chitase. The product of this enzyme action is an azo compound, oxy-,
di-, or amido-azo, and shows all the characteristic reactions of this group
of organic compounds. It is maintained by Enteman that the pigment is
elaborated in the hypodermal cells, and passes out through the pore canals of
the cuticula to the primary cuticula, where it finally remains ; while others,
working on a wider range of material, find that it arises as a diffuse pigment
through the action of an enzyme which has been elaborated as the result of
chromatolysis of the nuclei of hypodermal cells upon the primary cuticula
when in a soft condition — the action of the enzyme being akin to coagulation
or gelatinization and the products being hard, stiff cuticula and the azo
compounds.
The pigments which produce the hypodermal colors have been studied by
Krukenberg by spectrum, by Zopf and Tower in the hypodermal cells proper,
and by Hopkins, Griffiths, Urech, Mayer, Von Linden, and others in the scales
of Lepidoptera. In the hypodermal cells proper Zopf and Tower find that
the colors in Coleoptera, Hemiptera, Lepidoptera, and Hymenoptera are due
to granules of pigment of a fatty nature. Some of these were shown to be
lipochromes by Zopf in Coccinellidse, and by Tower in other families of Cole-
COLORS OF INSECTS. I 23
optera and in Hemiptera, Lepidoptera, and Hymenoptera. These pigments
are elaborated in the cells in which they are found, and have no connection
with the red pigments of the Malpighian tubules at pupation. In the scales of
Lepidoptera, Hopkins has demonstrated in the Pieridae the existence of com-
pounds of uric acid which produce white and yellow pigments. Griffiths fur-
ther isolates and studies a green pigment which is allied to uric acid, or is a
uric-acid derivative called lepidopteric acid (CnH^AzsNgOjo?). Urech,
Mayer, and others have made various experiments and tests with the pig-
ments in the scales of Lepidoptera, but without any definite results as far as
their chemical nature is concerned. Mayer has shown that the pigments in
the scales of some butterflies and moths are elaborated from the haemolymph,
possibly through the action of ferments. He has performed some interesting
and instructive experiments with the haemolymph of various moths, but comes
to no satisfactory conclusion concerning the nature of the pigments.
The subhypodermal pigments have been most completely investigated by
Poulton and his pupils in Lepidoptera, and by myself in Coleoptera. It
appears from these researches that these pigments are largely characteristic of
phytophagous larvae, or of carnivorous larvae which feed upon certain phy-
tophagous species. Poulton has shown clearly that these pigments are derived
from the food or, largely, from pigments contained in the food.
PHYSICAL OR STRUCTURAL COLORS.
These colors are produced by some of the various physical principles
whereby light impinging upon a body is modified by reflection, refraction,
defraction, or dispersion. These results are brought about by polished sur-
faces, lamellae, pits, striae, scales, or other surface modifications. As far as is
known there is only one physical color in insects, namely, white, which is
produced either by air in scales, by the fiat faces of crystals, or by fine gran-
ules in the fat body, all of which give total reflection.
Although colors are produced by physical causes in insects with great fre-
quency, they are, as shown by Mayer in Lepidoptera and by myself in other
orders, usually combined with chemical or pigmental colors, and it is to these
that insects owe their varied and brilliant hues.
CHEMICO-PHYSICAL OR COMBINATION COLORS.
These colors were first clearly recognized and defined by Mayer (1897) in
Lepidoptera. To this class belong all metallic, iridescent, and pearly colors,
as well as blue, green, violet, and many reds. The combination of a structural
and pigmental color in the same scale was early noticed by Urech, who showed
that in some Vanessas the scales had pigments which produced color and
interference colors due to striae, and he gave a similar case in certain Lycae-
nidae. These colors I have studied extensively, and I find them to be the
124 COLORATION IN LEPTINOTARSA.
most widely distributed of all insect colors. The results of their examination
have been summarized in the form of a table, which is here reproduced :
f
Chemico-
physical
(Reflection, ) Colored surface hrn3nV 1 Caused by thin, polished lamella:,
■. < pigmental •< with polished \ vellnws f or b^ a poshed surface over a
' (. colors. I appearance. reds layer of pigment.
(41 /Retraction, pig- > M,t_111. --inrs 1 Caused by refractive lamellee over a
(0M mental colors. ]• Metallic colors. | layer of absorptive pigment.
(nffrartinn nio- l (Caused by surface structures, pits,
colors. I W mertilcolore Iridescent colors. 1 ridges, stria:, or refractive lamella;
(_ over an absorptive pigment layer.
J Various iridescent, metallic, opales- "1 ~ c„„j i „, » „,„„ *
cent colors in which colors of classes [ CoT" ^rt 't r! y r If r
(a), (*), and (c) arc combined in one J'Xrr?X .'„? w f
L I color effect. J other orders of insects-
DEVELOPMENT OF COLORATION.
The ontogenetic development of color and color patterns in insects has been
studied most in Lepidoptera and especially in the wings. The earliest studies
of note were those of Schaffer, von Bemmelen, and Urech. These were fol-
lowed by the more elaborate researches of Hasse, Mayer, and von Linden. In
the other orders of insects studies have been made by Kunckel de Herculais
in Orthoptera, Tower in various orders, but especially in Coleoptera, and
Enteman in the hymenopterous genus Polistes.
The studies upon the origin of color in the wings of Lepidoptera have
shown that there is a regular order of development, not only of colors, but
also of the areas in which spots and stripes appear. The wings, which are
colorless at first, become opaque, yellowish, or light drab, beginning first near
the base of the wing and spreading distalward. Soon spots, stripes, and adult
markings appear, at first proximad, then more and more distad, then between
the nervules, and, last of all, upon the nervules. First the purely pigmental
or chemical colors develop, and these are followed by the chemico-physical.
Upon the body in various orders of insects I have found, as has also Ente-
man, that the color develops in segmentally arranged spots located upon im-
portant sclerites, usually over muscle insertions, and that from these centers
it spreads out, giving a surface either uniformly colored or spotted, if the
primary centers remain unfused. Development proceeds in general from
before backward, and from proximal to distal. The order of succession in the
colors has been found by these authors to be much the same as that in the
wings of Lepidoptera ; the lighter and simpler colors appear first, followed by
those which are darker and more complex, and these by the appearance and
development of the chemico-physical colors. The ontogeny of larval and
pupal coloration has not been studied.
PURITY OF INSECT COLORS.
The purity of insect colors has been investigated by Mayer (1897), who
has found that there are few or perhaps no pure colors in butterflies. All the
examinations made by this author have shown that a given color is in reality
COLORS OF INSECTS. I 25
composed of several colors, and in all he found a surprisingly large percentage
of black. The colors of other orders of insects have not been studied as
regards their purity.
EVOLUTION OF THE COLOR PATTERNS OF INSECTS.
Studies in the evolution of the color patterns of insects have been made
almost entirely upon Lepidoptera ; in fact, outside of this order only two
papers, excepting short notes scattered here and there, have been published.
LAWS OF COLORATION.
The first author to deal with the laws of coloration in insects was Higgins,
who maintained that the simplest and most fundamental color pattern is a
uniform tint, and that later the pigment developing along the nervules forms
darker markings which follow the lines of the nervules. Spots and ocelli he
believed to arise by the migration of crescent-shaped areas from the nervules
into the interspaces. The ontogenetic studies of von Bemmelen, Hasse,
Mayer, and von Linden produce no data in support of this view. On the
contrary, Mayer and von Linden show that the scales over the nervules and
along their courses remain uncolored until after other parts are pigmented.
It was shown by Darwin that ocelli can be traced through a continuous series
backward into dark spots with lighter centers. Scudder and Bateson, in
Lepidoptera, showed that ocelli are always located in interspaces and rarely
upon the nervules. Mayer advanced the following laws as those which gov-
ern coloration in Lepidoptera :
(a) Any spot found upon the wing of a butterfly or moth tends to be bilaterally
symmetrical, both as regards form and color ; and the axis of symmetry is a line passing
through the center of the interspace in which the spot is found, parallel to the longi-
tudinal nervules. (b) Spots tend to appear, not in one interspace only, but in homolo-
gous places in a row of adjacent interspaces, (c) Bands of color are often made by
the fusion of a row of adjacent spots, and, conversely, chains of spots are often formed
by the breaking up of bands, (d) When in process of disappearance, bands of color
usually shrink away at one end. (c) The ends of a series of spots are more variable
than the middle. This is only a special case of Bateson's law. (/) The position of
spots situated near the outer edges of the wing is largely controlled by the wing folds
or creases.
The same author (Mayer) later distinguishes three kinds of markings on
the wings of Lepidoptera — spots, bands, and combination markings— the latter
being a combination of spots and bands and more numerous than the first
two classes. He also gives the following laws concerning their behavior :
Bands, rows of spots, and combination markings are usually unbroken, of uniform
color, and isolated one from another; and only rarely do we find them broken, parti-
colored, or fused one with another. All departures are much more apt to affect the
ends than the middle of a marking. Rows of spots and combination markings are
126 COLORATION IN LEPTINOTARSA.
more apt to show these variations than are bands. Each genus or family follows these
general laws of variations, but each displays modifications peculiar to itself; these
modifications being sometimes of so marked a character as to constitute an exception
to the general rule. Such wide departures from the rule are exhibited by broken rows
of spots on the fore wing of Hesperida?. Also by double spots in rows of otherwise
single spots on the hind wing of Hcsperidae. In both of these cases the variation is
commoner in the middle than at the ends of the row. Like kinds of markings are more
apt to fuse than are unlike. For example : Rows of spots often fuse with other rows
of spots, or bands with other bands ; but fusions between rows of spots and bands are
rare. Fusions are commonly mere contacts, which are terminal for at least one of the
fused rows. Each pair of fused markings is more apt to come in contact at their ends
than at their middle. A definite relation obtains between the number of markings on
the fore and hind wings. Also the markings on either wing tend to appear in definite
order of frequency which is peculiar for each genus or family, but which, when plotted,
gives a fairly regular frequency polygon, with one or two maximum points. The above
statement applies to the number of spots in a row, and the length of bands, as well as
to the number of rows of markings on the wings. These relations are too definite and
orderly to be due to accident, and appear to be the expression of a natural law.
Among the other orders of insects I have shown that the color patterns on
the body of Coleoptera, Orthoptera, Hemiptera, and Hymenoptera are com-
posed of segmentally arranged spots, originating upon important sclerites or
over muscle attachments. I consider the color pattern to be a segmental one,
showing in successive segments a repetition of the same spots and color areas.
Enteman has come to a similar conclusion concerning coloration in Polistes.
PHYLOGENY OF COLORATION.
The first thorough and consistent attempt to study and trace the phylogeny
of insect coloration was made by Darwin, who developed around this subject
to a considerable extent his theories of sexual selection, protective coloration,
warning color, etc. Natural selection and segregation were for this author
sufficient to account for the phenomena of insect coloration. The researches
of Darwin, Wallace, Bates, Weismann, Poulton, Hasse, and others upon the
coloration of insects, especially that in Lepidoptera, have developed a school
which stoutly maintains that the brilliant colors of many insects, especially
those in males, have been developed by a process of sexual selection. They
likewise hold that mimicry, protective resemblance, etc., are due to selective
influences alone, and Weismann has invented some ingenious theories to
account for the rise of these phenomena by means of selection, natural, sexual,
or otherwise.
Strongly opposed to the views of Darwin, Wallace, Weismann, and their
followers is the theory promulgated by Eimer, who, in 1889, conducted an
extensive research upon coloration in the Papilionidae, and in 1898 published a
still more extensive work along the same line. Eimer and his followers hold
that the color patterns of insects are only slightly influenced by natural selec-
COLORS OF INSECTS. 12J
tion, but are due to the influence of environmental conditions, food, moisture
and temperature, which have caused variation along a few lines and in no
others ; and it is in these lines that species are developed by the segregation of
units at points upon these lines of modification. This school holds the primi-
tive color pattern to be that of longitudinal stripes which break up into spots
and then fuse to form, first, transverse bands, and later a uniform color. New
markings appear upon the body from behind forward and from above down-
ward, or conversely, and old ones disappear in the same order and succes-
sion. Most of the data used in support of these views is drawn from the
study of adult coloration in Lepidoptera, wherein most of the bands and rows
of spots are transverse. But Eimer and his pupils boldly maintain that these
are longitudinal because parallel with the long axis of the body, and they
base their arguments for the validity of their doctrine thereon. With the
exception of von Linden, all of Eimer's followers base their conclusions upon
the study of adult color patterns, arranging these in series convenient for
their purposes. Yon Linden has studied the ontogeny of coloration in a
number of Lepidoptera, and claims that her results are in strong accord with
Eimer's views. However, an impartial examination of her observations and
figures fails to reveal any real confirmation of the theories so vigorously de-
fended. Likewise, the researches of Schaffer, von Bemmelen, Urech, Hasse,
and Mayer fail utterly to give support to the views of Eimer and his pupils.
The researches of Piepers upon Pieridas and the larvae of Sphingidae and
my own studies upon several orders of insects show that there are definite
lines of color-pattern development (phylogenetically), which are inde-
pendent of environmental conditions and inherent in the animals them-
selves. Thus, in the Pieridse, Piepers shows that red, orange, yellow, and
white often follow in phylogenetic sequence, and black increases in amount
as this progressive change goes on. Mayer holds that the force of Piepers's
arguments is weakened by the fact that the colors of Pieridse are produced
by uric-acid derivatives, while those of other Lepidoptera are not.
In my paper of 1903 I have shown that in several orders of insects there
are stages in development and many color-pattern features which are com-
mon to a wide range of insects, and that these are dependent upon physiolog-
ical activities in a very definite way and are in no wise related to environ-
ment. Thus, in the pronotum of various Coleoptera, it is shown that even in
distantly related families or in species living in diverse habitats there is a long
series of developmental stages in common ; and that it is only at the very end
of the series that specific and generic differences come in. In this instance
inherent structural and physiological characteristics are the controlling fac-
tors. Additional studies will undoubtedly verify and extend our knowledge
of the range of these phenomena.
128 COLORATION IN LEPTINOTARSA.
EFFECT OF ENVIRONMENT UPON COLORATION.
Many well-known facts point to the immediate action of temperature, mois-
ture, food, light, etc., upon the coloration of insects. That changes in tem-
perature can and do effect modifications in coloration and dimorphism is
abundantly proven by the work of Dorfmeister, Edwards, Weismann, Merri-
field, Urech, Fischer, and Standfuss, while the last two authors have been
able to produce an inheritance of the modification in the second genera-
tion. Seasonal variations as the result of the varying conditions of wet and
dry seasons are of common occurrence in the tropics. As a rule, however,
the modifications produced in experiments along this line have not been
inherited in subsequent generations ; and this has given Weismann, Poulton,
and others a basis for strong arguments against the possibility of the validity
of the Lamarckian doctrine of evolution in insect coloration through the
inheritance of acquired characters.
PROTECTIVE RESEMBLANCE, WARNING COLORATION, MIMICRY.
Wallace's theory of warning coloration and Darwin's theory of protective
resemblance, with its special phase of mimicry elaborated by Bates and
Muller, have secured strong support from the observations in nature of
Bates, Midler, Weismann, Tremen, and others, and from the experimental
researches of Poulton and Finn. These phenomena can at present be best
explained by the hypothesis of selection. Attempts at explanation by other
theories always show lamentable weakness or peculiarities of logic not easily
explained. In fact, natural selection as a cause of evolution in insect color-
ation receives its strongest support from the data and observations cen-
tered about these phenomena. Arguments against the theory of protective
resemblance, warning coloration, and mimicry are few and of doubtful valid-
ity ; and the facts used are cases of convergent evolution of questionable
application. Altogether, no other phenomena, I believe, argue as strongly and
successfully for the truth of the principle of natural selection as these mar-
velous phenomena of insect coloration.
THE PROBLEMS OF INSECT COLORATION TO-DAY.
It was truly and well said by Darwin that "the coloring of insects is a
complex and obscure subject." Before we can proceed much farther in the
subject we must understand better the nature, source, and development of
color and color-producing substances and their relation to the physiology of
the animals in question. When we have this knowledge at hand then shall we
be prepared to investigate the action of environment upon coloration, and to
understand more fully whether evolution has been along definite lines —
according to an orthogenetic principle, or fortuitous, guided by natural
selection.
PIGMENTAL COLORS. 120.
At present the explanation of color phenomena is most difficult. Warning
coloration, protective resemblance, and mimicry are, I believe, adequately and
only explainable on the basis of natural selection. With other phenomena it is
not so. The myriad hues and color combinations and patterns of insects and
the brilliant colors of the males of many forms can not, I believe, be explained
by natural selection, inheritance of acquired characters, orthogenesis, or
mutation ; at least no known explanation seems to be satisfactory to any
considerable body of investigators. Nowhere else in the animal kingdom has
evolution, as it were, so run riot as among insect colors ; and yet its study has
made comparatively little progress in recent years. The evolution of insect
coloration in general seems at present to be practically inexplicable. A por-
tion here and there is elucidated by one or another of the theories advanced ;
a few facts, a few laws, are known ; but, on the whole, our knowledge of these
intricate and interesting phenomena is truly meager.
THE COLORS OF THE GENUS LEPTINOTARSA.
In the beetles which form the basis of this research, all of the three cate-
gories of color are found, either in the imagines or in the larvae. These will
be considered under their natural divisions.
CHEMICAL OR PIGMENTAL COLORS.
It is to the colors of this category that the genus owes its characteristic
coloration ; in fact, physical or chemico-physical colors enter but little into its
color phenomena. In general the colors of the adults are cuticula and hypo-
dermal, and those of the larvae cuticula, hypodermal, and sub-hypodermal.
The cuticula colors, black, brown, and yellow, are at once recognizable by
the fact that they are insoluble in water, alcohol, oils, ether, weak acids, or
alkalis. In L. undecimlineata and decemiineata the black spots and stripes
are not pure black, but a deep brown that owes its black appearance to the
density of the deposit of pigment. If an elytron or a pronotum of one of these
beetles be boiled in alcohol, water, or one of the essential oils at one atmos-
phere of pressure, even for several hours, no perceptible change in color is
noted. Long standing in water or alcohol, or boiling in these reagents under
two or three atmospheres pressure, will in a few hours result in complete
extraction of the pigment. The same treatment with dilute HN03, HCL,
H,804, NH4OH, or KOH will produce solution after a sufficient period of
boiling — two or three hours. These cuticula colors are absolutely permanent
in these beetles, five years of constant exposure to light not effecting any
perceptible change in them.
The black cuticula colors are uniform throughout the genus, i. e., they are
due to dense deposits of a deep-brown pigment having the solubility and per-
manency given above. From the black we can pass through a graded series
to the browns, such as those of the brown areas in /,. setterstedti, or the dark-
brown spots and stripes of dilecta, belti, and others. The colors of these spe-
I30 COLORATION IN LEPTINOTARSA.
cies are produced by cuticula pigments, are permanent, and are with difficulty
soluble in the same reagents as the darker colors.
Between the browns and yellows transitions in the adult or in the larva? are
few, but in the ontogeny of the animal they are common, as in the develop-
ment of the black or brown areas of the species mentioned. The yellow,
as in L. decemlineata and its allies, is only slightly developed and is of no
great permanence, fading on exposure to light, with age, or on treatment with
reagents. It is in these beetles the least permanent of the cuticula colors, and
is in fact of relatively little importance in the production of color patterns.1
The dark markings — that is, the black or brown spots — of the larvse and
pupae of this genus are also cuticula pigments like those discussed above.
These are the pigments described by Poulton in lepidopterous larvae as
"proper to the species."
When studied in sections of 3^ P- to 6§ !'■ thickness, these pigments are
always seen to be diffuse and not in the form of granules, as claimed by
Hagen, and recently by Enteman in Polistes. They are limited entirely to
the primary cuticula, and the transitions from pigmented to unpigmented
areas is rapid but continuous.
Always closely associated with the cuticula pigments in the production of
color patterns in these beetles are the hypodermal pigments, which, as far as
I know, are always yellows and reds in this genus. These two colors exist in
the form of small granules in the hypodermal cells, and, although more or
less permanent, they all fade at death or on exposure to light, and are easily
soluble in water, alcohol, and other organic solvents. The chrome yellow so
often seen in decemlineata when freshly emerged is due to these pigments;
but undecimlineata never possesses them and is, therefore, of a white color
with black markings. In decemlineata, however, the yellow color fades
rapidly on exposure to light or with age, so that old specimens have the light
areas between the black stripes nearly white. The white is opaque, however,
and not translucent, as in undecimlineata. When freshly emerged from the
pupa decemlineata has a distinct reddish color, which gives away to chrome
yellow before the animal leaves the pupal cell, and later in life becomes white.
In rubicunda, however, the red color is retained and grows more intense. It
is not lost on exposure to light or with age, but fades to a dull yellowish white
at death. In rubiginosa in life the color is deep red, but at death changes to
a dull brick-red or yellow.
These same yellows and reds are also found in the larvae of the species
described, and, as far as surface observations can determine, are continued
over from larva to pupa and imago. They are difficult to investigate in
sections because the ordinary reagents of microscopic technic dissolve them
completely long before the tissue is ready for study. By freezing, good
1 These cuticula yellows must not be confused with the yellow colors often produced
in the cuticula at death. These latter colors are due to post-mortem changes and to
drying of the integument.
COMBINATION COLORS. I3I
sections have been obtained which show the location and character of these
pigments.
The relation between the cnticula and hypodermal colors in the production
of color patterns in these beetles is this : The hvpodermal color plays the part
of a ground upon which the darker browns and black stand out as spots and
stripes ; or, if the cnticula colors are extensively developed, the hypodermal
colors show through holes and rifts in the cuticula color as lighter spots.
Only rarely does one class prevail to the exclusion of the other. In rubigi-
nosa the prevailing color is hypodermal, red, the dermal being confined to the
appendages, whereas in modcsta no hypodermal color is visible, and in punc-
ticollis only a few small, yellowish spots are found.
The subhypodermal pigments play no part in the adult coloration of these
beetles ; for, although they are present in the larva, pupa, and imago, it is only
in the larvae of a few species that they are visible. When found the color is
always a yellow or a reddish yellow, due to a modified, derived pigment, and
it helps to intensify the yellow or red of the hypodertnis. Sometimes, as in
decemlineata, a change of food will change the nature of this derived pig-
ment, and thereby the ground color of the larva. In this form larval color
ranges from deep red to a pale yellowish red, due solely to variation in the
subhypodermal pigment contained in the larval haemolymph. In a few spe-
cies, undecimlineata and diversa, the haemolymph and fat body are unicolored,
so that the larvae have a white or a translucent appearance, depending upon
their stage of development. These larvae are translucent in early stages, but
later on the development of the fat body gives them first a chalky white and
then a pale creamy white ground color.
PHYSICAL OR STRUCTURAL COLORS.
Pure physical colors are absent in this genus, excepting white, which is
produced by the fat body in larvae {undecimlineata) or by reflection from
crystals or other surfaces in the integument. This is of no importance in the
coloration of the genus.
CHEMICO-PHYSICAL OR COMBINATION COLORS.
In the larvae and pupae of Leptinotarsa these colors are wanting, but in the
imago they, in connection with the cuticula colors, are productive of interest-
ing changes. These changes consist in the forming of metallic colors, green,
blue, violet, and combinations of these with black, green, greenish black,
blue-black, etc., colors very common in this genus. These colors are all due
to thin lamellae over dark absorptive pigment. Scales and striae are wanting,
and pits are not developed sufficiently to produce physical color changes.
The effect of these chemico-physical colors in this genus is to transform to a
metallic color areas that would otherwise be a dull black, and thus in a sim-
ple but effective manner to produce differences, often of specific value.
I32 COLORATION IN LEPTINOTARSA.
THE DEVELOPMENT OF COLORS IN LEPTINOTAESA.
CHEMICAL OR PIGMENTAL COLORS.
The most important of the chemical colors are those located in the cuticula
and in the hypodermis. In the ontogeny of the animal, however, the cuticula
colors are lost at each ecdysis, and are replaced by new developments thereof,
whereas the hypodermal colors persist from larva to adult. In this account
we shall first consider the development of the cuticula colors and the struc-
tures that carry them.
CUTICULA COLORS.
A section through the body wall of these beetles in any stage shows the
following structures : On the outside is the primary cuticula, a layer of hard,
homogeneous, unstratified material, everywhere of uniform thickness, not
penetrated by pore canals, and carrying deep brown or yellowish diffuse cuti-
cula pigments. Beneath this is a layer of varying thickness, composed of
alternating layers of material having different refractive indices, and pene-
trated everywhere by minute pore canals, which reach outward to the edge of
the primary cuticula. This layer is devoid of pigment, except in rare cases,
and is always unpigmented in Leptinotarsa. The inner surface of this cuti-
cula, the secondary cuticula, is uneven. Into the uneven places the hypoder-
mal cells fit exactly, and from them send out long, delicate, protoplasmic pro-
cesses, which fill the pore canals in life. The cells of the hypodermis vary in
size and shape according to their position in the body and the ontogenetic
stage of development in the individual. They are, however, always columnar,
and are backed up on the inner surface by a delicate layer of mesodermal
cells, endothelium, which lines the body cavity. In these hypodermal cells
are found the granules of the hypodermal pigment and beneath them the dif-
fuse subhypodermal color.
During the ontogeny of these beetles the integument undergoes a series of
cycles of development whereby the cuticula colors are lost and redeveloped
several times. These cycles of development are associated with ecdysis, and
are important for the understanding of color ontogeny, especially that in the
larvae.
In the embryo, at about the end of the embryonic period, the cells of the
hypodermis (ectodermal cells) secrete rapidly an even, homogeneous layer of
cuticula over the entire surface. This layer is without visible structure,
stains intensely in many stains, and is soft and pliable. No colors are seen at
first ; but in the few hours immediately preceding hatching color develops, at
first as faint yellowish or opaque areas. These later change to a light yellow-
brown, and remain so until after hatching, when they rapidly become darker,
and finally black. These darker areas constitute the colors "proper to the
CUTICULA COLORS. I33
species," and are developed in a short time — one or two hours after hatching.
Almost as soon as the embryo emerges from the egg the primary cuticula
begins to harden rapidly ; and about the same time, and especially after the
larva begins to feed, the secondary cuticula begins to be deposited, and its
deposition continues until near the end of the first larval instar. The deposi-
tion of this second layer, however, does not completely separate the primary
cuticula from the hypodermal cells, a connection being retained by means of
the numerous protoplasmic processes occupying the pore canals of the second-
ary cuticula. Near the end of the larval instar the protoplasmic processes are
withdrawn from the pore canals, the hypodermal cells lose their attachment
to the lower side of the secondary cuticula, and the hypodermis and cuticula
separate. The ordinary process of ecdysis now removes the old cuticula ; but
before this is accomplished a new, fresh, colorless layer of primary cuticula
has developed beneath the old.
With each succeeding ecdysis all the cuticula colors of the entire body are
lost and are redeveloped in the few hours immediately following. Although
they usually reappear in the same areas where they existed in the previous
stage, this process allows of variation in the coloration from instar to instar.
The development of the cuticula color is always the same, no matter whether
in a freshly hatched larva or in the transforming imago. Likewise the
stages are the same — a cloudy appearance in the primary cuticula, and then a
faint yellow followed by darker yellow, brown, and black. Cessation of devel-
opment through any cause at any point in this series gives for the perma-
nent coloration that color of the stage when development ceased. At each
renewal of the colors the full development is attained in a short period of time
immediately following the casting of the old cuticula covering.
Without exception in this genus, and, as I have shown, in insects in general,
the successive development of cuticula color in the ontogeny of insects fol-
lows the same lines laid down in the general account of the process and its
repetition. We shall now consider in detail the development of cuticula
colors in this genus.
THE CUTICULA, ITS STRUCTURE AND COMPOSITION.
It has been shown by Vossler and Tower that the cuticula portion of the
body wall of insects is composed of two distinct kinds of substances — an
outer, homogeneous, primary cuticula, and an inner, stratified, secondary cutic-
ula. The inner layer differs from the outer in that it gives a cellulose reaction
and is akin to carbohydrates. The outer layer of the cuticula is, I believe, a
derivative of some gelatinate and the secondary cuticula a carbohydrate.
Pupae of decemlincata, two to five days old, taken before pigmentation or
deposition of the secondary cuticula had begun, were crushed in a mortar,
washed in distilled water, to remove all loose material, and then macerated
with 1 per cent KOH for twenty-four hours at 18° to 22° C, with frequent
134 COLORATION IN LEPTINOTARSA.
agitation. By this method all tissues were removed and pure, or nearly pure,
quantities of the primary cuticulawere obtained and subjected to various tests.
Ledderhose showed that "chitin," in cold, concentrated solutions of HCL
or H2SO «
"OCM
0 s
*• aZ
fee «
posterior prouotal band
anterior pronotal band
anterior and posterior
bands of tergal spots
upon abdominal
segments
posterior lateral epicranial^
anterior lateral epicranial..
antenna ry
gular,
,'protboracie spiracolat
j-4- i-iii. — — ' uner tergals
|«i ■" —middle tergals
Uj- ——.outer tergals
_ — Ruiracuta spots
— ,_baso-pleural
——outer steraals
— "■ middle Hterrmts
Bpiraeulaf, l to 7 baBO-plcurnl (abd)
"metathoracic wing spot
mesothoracic wing spot
baso-plcural (thoracic)
Text-figure 7.— Diagrammatic representation of the centers of color development
found in the young larvse of Leptinotarsas. Drawn from a young larva of L. unde-
cimtineatataa seen in side view.
directly derived from modified or combined, segmentally placed centers. In
Lepiinotarsa this idea has been put to a most exact test with the result, as will
appear in the following pages, that it appears to be in all respects valid.
In the young larvae of very many Chrysomelidse, Coccinellidae, Cassidae, and
other forms is found a system of color centers which is highly suggestive and,
as we shall see later, ideal in its simplicity. In late embryos or newly hatched
144
COLORATION IN LGrTINOTARSA.
larvae of many Leptinotarsas (undecimlincata, decemlineata, signaticollis, etc.)
we find color areas or centers which are represented diagrammatically in text-
figure 7 as seen from the side, and in text-figure 8 as seen from above. An
abdominal segment (text-fig. 9), which represents the simplest condition,
shows the skeletal ring composed of paired tergae, pleurae, and sternae. Upon
each of these are color centers which are placed in a constant morphological
position and have definite relations to the structure of the animal. Upon the
tergum are found on either side of the median line three pairs of color centers
divided by a crease or furrow into two series. These are designated the outer,
middle, and inner tergal centers, and they exist as an anterior and posterior
series (text-figs. 7 and 8). Each of these centers may consist of one, two,
anterior median pronotal spot
posterior median
pronota) spot
Inner tergale
anterior lateral epicranial
** ^-posterior lateral epicranial
__aoterior pronotal band
— . — — — — — — -posterior pronotal band
_ _ _ — ——-outer tergal
— — . middle tergal
wing spots
&— apiracula
eplracula
— ——anterior tergal band
"*"■•■-— posterior tergal band
Text-figtjrb 8.— Diagrammatic representation of the color centers shown in text-figure 7
when seen from the dorsal side.
or three small, closely placed spots, and in many species they are the locations
of spines. This type of coloration is found not only upon all the tergae of the
abdominal segments, but also upon the last two thoracic tergae. Upon the
anterior thoracic tergum or pronotum, however, there appear in the larvae
the anterior and posterior pronotal bands (text-figs. 7 and 9), which are homo-
dynamous with the anterior and posterior tergal system of centers of the pos-
terior segments. This same type, further obscured, is seen in the epicranium,
where there exist the anterior and posterior lateral epicranial areas, each
homodynamous with the abdominal tergal areas. The anterior and posterior
median spots (text-fig. 8), which often appear upon the pronotum, are the
only ones not found in this series of abdominal tergal centers. On the side
ONTOGENY OF LARVAL COLOR PATTERNS.
MS
pieces or pleurae, on all abdominal segments and on the prothoracic segments,
are found the spiracular spots which are represented on the wing-bearing
segments by the wing spots. Beneath these, on the abdominal segments, at
the base of the pleurae, are one or, rarely, two basal tergal spots (text-figs.
7 and 8).
On the sternae also a series of color centers is found corresponding to
those of the tergae in that there are outer, middle, and inner areas, but these
are rarely present as anterior and posterior systems, although frequently in
variations there are found strong indications of this anterior and posterior
division. These areas continue forward in a homodynamic series over the
thoracic segments.
This system of color centers is not confined to the Coleoptera nor to a few
families thereof, but is, I find, of most general occurrence among the Tra-
cheates,and,as I shall show in another paper, is the fundamental pattern upon
which all insect coloration is founded. With this system of color centers it
Tergum
Pie
inner tergal
middle tergal
at body
---outer tergal
.-pSpiracula
_ _baso-pleural
■* -.alimentary canal
ventral-nerve cha
inner sternal!
outer sternal
middle sternal
Text-figure 9. — Diagrammatic representation of a segment, showing the position of the
different spots upon the sclerites which compose it.
is a simple process to develop longitudinal stripes or transverse bars, or both,
or combinations thereof. It would also be easy for the followers of Eimer to
see in this a system of longitudinal stripes, and for others to see a system of
transverse bars, as the fundamental color pattern. In truth, however, this
system of color centers is neither of the latter ; it is a metameric repetition of
homodynamous areas, or modifications thereof, on every individual segment;
and any other interpretation is warped in favor of some particular theory.
It is true that the various elements may be so modified as to give on super-
ficial examination the general idea of stripes or bands, and as such these com-
binations do function in color patterns ; but in ontogeny and phytogeny, in
experiment and in heredity, each metameric portion of any such color mark-
ing behaves as a metameric structure, and in no other way.
In the larvae of L. undecimlincata at this time the color scheme is to a large
extent the same as that shown in text-figures 7 and 8, excepting that the ante-
rior and posterior pronotal stripes and all the areas of the epicranium are com-
bined into one uniform black color. Upon the abdominal and last two thoracic
I46 COLORATION IN LEPTINOTARSA.
segments, however, the elementary color pattern is retained in full, as may be
seen by comparing fig. 1 on plate 17 with text-figures 7 and 8. Each color
center bears one or more single or slightly branched spines on the tergal and
sternal sclerites of each segment. In the pleural region these spines are
small, and are limited to the baso-pleural centers. As the small larvae push
about through the abundant trichomes on the leaves of their food plants, a
large accumulation of these become lodged among and cemented to the
spines by the secretion of the dermal glands, until the larva presents the color
and appearance of a ball of dislodged trichomes. At the first ecdysis the
color pattern of the larva changes strikingly, as is shown in fig. 2 on plate
17. In the second stage the epicranium and pronotum remain black, while
all the color centers are lost, excepting the abdominal and prothoracic spirac-
ular spots, the wing spots, and the baso-pleural thoracic spots. The integu-
ment in this second stage is smooth and entirely devoid of spines on the tergal
and sternal elements of the segments. Owing to the sticky secretion of the
dermal glands a deposit of trichomes gathers on the larva in this stage,
though, owing to the absence of spines, this deposit is thinner than in the pre-
vious stage. The true body color now is a pale, transparent greenish-white, as
is shown in the figure. Between this stage and the final one shown in fig. 3
on plate 17, there are no modifications in the color markings, although the
general body color is changed to an opaque, pale yellowish-white, which is
due solely to the development of the fat body. In the final stage the increase
in the body surface, which is not accompanied by any great increase in the
number of dermal glands, results in there being on the body only a very
slight deposit of trichomes, if any at all, so that the larvae are freely exposed
on the leaves of their food plant. The final body color varies from a pure
white to pale lemon yellow, the average condition being represented by fig. 3
on plate 17.
The nearest relative of L- undecimlineata is L. diversa, and it is of interest
to compare the ontogeny of the larvae of the two species. In the first instar
diversa has the appearance and all the characteristics of a larva of undecim-
lineata at the same stage of development. In fact, the larva? of the two spe-
cies are then alike, as may be seen by comparing figs. 1 and 4 on plate 17.
During the first instar L. diversa has likewise the habit of covering itself with
the trichomes of its food plant, so that the two can not be distinguished from
each other either in coloration or in habits. The second instar is, as in unde-
cimlineata, very unlike the first, there being a striking loss of color centers
and of spines, while in color and habits the two species are at this stage almost
duplicates of each other, as may be observed in figs. 2 and 5. In the final
stage, however, as is shown in figs. 3 and 6, there appear striking differences
between the two species in color and color pattern. L. diversa now develops
a series of black tergal markings that are characteristic and constant, and
which clearly differentiate the two species of larva?. The black tergal mark-
Explanation of Plate 17.
This plate illustrates the changes in the coloration exhibited by the larvse of several
species of Leptinotarsa during their ontogeny.
Figs. 1, 2 and 3, L. undecimlineaia Stal. From Tierra Blauca, Vera Cruz.
Figs. 4, 5, and 6, L. diversa (n. sp.). From Guadalajara, Jalisco, Mexico.
Figs. 7, 8, and 9, L. signaiicollis Stal. From Cuernavaca, Morelos, Mexico.
Figs, io, 11, and 12, L. multitceniata Stal. From Guadalupe, Federal District,
Mexico.
Figs. 13, 14, and 15, L. oblongata (n. sp.). From Cuernavaca, Morelos, Mexico.
Figs. 16, 17, and 18, L. rubicunda (n. sp.). From Toluca, Mexico, Mex.
Figs. 19, 20, and 21, L. dccemlineata Say. From Chicago, Illinois.
Figs. 22, 23, and 24, L.junda Guer. From Experiment, Georgia.
Figs. 25, 26, and 27, L. haldemani Stal. From Cuernavaca, Morelos, Mexico.
Figs. 28, 29, and 30. L. dilecia Stal. From Cuernavaca, Morelos, Mexico.
Figs. 31, 32, and 33, L. violescens Suffr. From Orizaba, Vera Cruz.
Figs. 34, 35, and 36, L. rubiginosa Rogers. From Santa F6, Federal District,
Mexico.
W. L. TOWER
**• 'l-'^ **• ^
•
.. •-.% &L* <$£&* ...::>r,
s
5
14- 20.
•■••n
:^"'
W LT del
STAGES IN DEVELOPMENT OF THE COLOR PATTERN OF LARVAE OF
LEPTINOTARSA DURING OMTOGENY.
ONTOGENY OF LARVAL COLOR PATTERNS. I47
ings develop around the centers of color described as anterior and posterior
bands, which later fuse, first in the middle portion of the body and still later
toward the ends, but which are always distinguishable laterally. These
markings are confined to the abdominal segments. The general body color
also changes in this stage from the pale, transparent lemon yellow of unde-
cimlineata to a chalky or ocher yellow. The color of the head, and some-
times that of the pronotum, shows a change to brown, as do also the legs.
Closely allied to the two preceding species is signaticollis, whose larval col-
oration has ontogenetic stages very much like those of diversa. The first
stage in signaticollis (pi. 17, fig. 1) is the same as the corresponding ones
in diversa and undecimlineata; the spines and viscous secretions enable the
trichomes of its food plant to become entangled and cemented to the body in
signaticollis in the same manner as in the two species just described, so that
in life on their native food plants the three species are indistinguishable. So,
too, the second stage of signaticollis is like the second stage of the other two
species, in that all excepting the spiracular, wing, and basal-pleural thoracic
spots are lost ; and there is the same habit of decorating the body surface with
the trichomes of its food plant. In the third and last stage, however, occurs
a striking change in coloration and habits. The color centers now redevelop
color on all the abdominal segments excepting the first three or four baso-
pleural spots and the anterior and posterior tergal areas (pi. 17, fig. 9). The
tergal bands arise on the three pairs of tergal centers and fuse first distalward,
and then, beginning mediad, fuse caudad to some extent ; but a large part of
both tergal bands remain free, as in the figure given. Sometimes the two do
not fuse or coalesce at all, but are clearly distinct. The general body color
likewise changes from a pale, transparent yellow to an opaque ocher or chrome
yellow. This change is due first to the appearance of lipochrome pigments in
the hypodermis, and, second, to the development of the opaque yellow fat
body. There is a striking parallel in the ontogeny of the coloration of diversa
and signaticollis.
In L. mnltitaniata and its allies the color development is also suggestive.
The first larval stage (pi. 17, fig. 10) has a color pattern exactly like those
of undecimlineata, diversa, and signaticollis, the four species being in their
coloration so nearly alike in the first larval stage as to be indistinguishable.
L. mnltitccniata, however, does not have the habit of covering its body with
trichomes, and it soon develops a deposit of bright yellow lipochrome hypo-
dermal pigments, which give it a bright yellow color of about the same shade
as the large, hard yellow spines of its food plant, Solanum rostratum. In the
second stage of development a change of coloration occurs, but one not as
pronounced as those in the three species just described. In mnltitaniata all
the spiracular, wing, and basal-pleural spots are retained in this stage, as are
also the inner and middle pairs of tergal centers on the last two thoracic and
the first three or four abdominal segments (pi. 17, fig. 11). The body color
148 COLORATION IN LEPl'INOTARSA.
is also bright opaque yellow, quite different from the translucent pale yellow
of the previously described species. Upon the pronotum a change is also
found in the reduction of the anterior pronotal band. In the last stage, which
is shown in fig. 12 on plate 17, almost no further change of coloration is
found.
Closely allied to multitccniata are oblongata, decemlineata, rubiciinda, and
juncta. L. oblongata, which in the adult condition is almost undistinguish-
able from multitccniata, is in the larval stages quite different, as may be seen
from the representation of the first larval stage in fig. 13 on plate 17. In
this form the only color centers that are represented are the spiracular, wing,
and baso-pleural thoracic spots. All other spots, as well as all spines, are
wanting in the earlier stages of this species. Like multitccniata, it is at first
almost white, but soon changes to corn yellow. In the second and third
stages no modifications of color occur, excepting the loss of the anterior
pronotal stripe in the last stage, although the general body color gradually
becomes more opaque and yellow (pi. 17, figs. 14 and 15). This species,
oblongata, is quite remarkable for its early complete loss of all but a few color
centers ; but that the animal is not without the fundamentals of these is shown
by their brief appearance in the late embryo, and by their reappearance as
color centers in the pupa and imago.
L. intermedia, rubicunda, and decemlineata present stages of development
almost exactly like those of multitccniata. All have in the first larval stage a
complete system of color centers and a full armament of spines, as are shown
in rubicunda in fig. 16, and in decemlineata in fig. 19 on plate 17. The gen-
eral body color of these forms differs from that of multitccniata in that rubi-
cunda is bright red, intermedia a dull yellow, and decemlineata a dull red. In
the second stage each of these four retains the spiracula, wing, and all baso-
pleural spots, and the inner and middle tergal pairs of color centers on the last
two thoracic and the first three or four abdominal segments. The outer part
of the anterior band on the pronotum is lost in rubicunda in the last instar,
and in intermedia and decemlineata in the second instar.
The species juncta and defecta also exhibit interesting stages of color
ontogeny. In juncta, in the first larval stage, only the spiracular and
baso-pleural spots are present, and there are no spines (pi. 17, fig. 22).
In the second and last stages (figs. 23 and 24) the baso-pleural spots have
been obliterated, so that only the posterior pronotal band remains, while the
baso-pleural spots of the thoracic segments are small or wanting. Often all
the color areas are light brown in color.
These species of Leptinotarsa all belong to the lincata group, which con-
tains the most generalized members of the genus. In the coloration of the
larvae of these beetles I find that in undecimlineata, signaticollis, angustovit-
tata, diversa, multitcrniata, intermedia, decemlineata, rubicunda, and melano-
ONTOGENY OF LARVAL COLOR PATTERNS. I49
thorax, the first larval stage has a color pattern of a generalized character,
with all the color centers described in the beginning of this section clearly rep-
resented. The second stage, however, shows in all a marked change in color
pattern in the reduction of many centers. The centers first to go are the pos-
terior tergal and sternal, and the last are the thoracic baso-pleural, while the
spiracula and wing centers are never lost, although the posterior members
may be reduced. On the pronotum it is the anterior band that is reduced.
In the species oblongata, juncta, and defecta the first larval stage possesses a
color pattern like the second stage of the other species of this group ; and if
the color centers not found in the larvae appear at all it is only in the embryo
as very transient characters, or in the imago. Thus it appears that these cen-
ters are not lost when the stages are missing, but are only quiescent.
In other groups of this genus the larvae likewise show stages or series of
stages which are of interest. These, however, do not present the generalized
conditions found in the lincata group, although the coloration of the larvae of
the entire genus starts with a condition comparable to the earlier stages in that
group. In haldemani and violescens the first larval stage shows the spirac-
ular, wing, baso-pleural, and sternal centers of coloration on the anterior four
or five abdominal segments. The general body color is red, due to red lipo-
chromes in the hypodermal cells (pi. 17, figs. 25 and 31). In the second
stage of both species all of the color centers are lost, excepting the spiracular
and wing and traces of the baso-pleural thoracic spots. The wing spots are
reduced to crescentic markings and the pronotal bands are much reduced,
as is shown in figs. 26 and 32, plate 17. On the head the reduction of color
is carried to the extent that all dark markings are lost, excepting the
retinal pigment and that at the tips of the mandibles. There is also a marked
change in the body color from opaque reddish to a nearly transparent pale
yellow-white. In the last stage the reduction of the dark color is somewhat
more marked, but the chief change is in the body color, which becomes pale
opaque yellow in violescens and corn yellow in haldemani (pi. 17, figs. 27
and 33).
In L. dilecta is found a series of color-pattern stages which are in the main
comparable to those just described for haldemani and violescens. The larva
is at first red, with spiracular, wing, baso-pleural, and sternal centers of col-
oration represented, and with spines on the ventral surface (pi. 17, fig. 28).
In the second stage all excepting the spiracula and wing centers are lost, the
pronotal stripes are reduced, and color disappears from the head. The body is
a general transparent pale yellow-white. In the last stage the color becomes
grayish yellow, the anterior tergals unite to form bands, the wing spots
are mere crescentic lines, and the larva shows the condition rather unusual for
this genus, which is shown in fig. 30. In forms like rubiginosa larval devel-
opment starts with the centers much reduced, and finally reaches the condition
shown on plate 17, fig. 36.
I50 COLORATION IN LEPTINOTARSA.
In all of the species in which the ontogeny of larval coloration has been
described, and in all in which it has been studied, there appear sharply marked
stages, some generalized, others highly specialized. We find, also, that cer-
tain color centers are exceedingly persistent, and others transient; that, in
general, the tergal and sternal centers are the less persistent, and that the pos-
terior members are more variable in their degree of permanency than the ante-
rior. The sharp demarkation of stages is, of course, due to the phenomenon
of ecdysis, which periodically removes one entire set of coloration characters
and affords an opportunity for them to be replaced by a similar set or a differ-
ent one, as the case may be. In the phenomenon of ecdysis there are at work
processes of high importance. Thus, in the first instar in L. divcrsa, there
are groups of metamerically arranged cells which develop the necessary
enzyme for the production of the color centers, and, accompanying them, are
cells which develop spines; but at the first ecdysis these spines are all
removed, which means that the setigerous cells hold in abeyance their power
of forming spines. Likewise, the cells which develop the enzyme necessary for
the formation of the many centers of coloration in the first stage do not exer-
cise this function in the reconstruction of the color pattern during the first
ecdysis, and only the spiracula, wing, and basal-pleural thoracic centers retain
this function through the first ecdysis. At the second ecdysis, when the sec-
ond color pattern is removed and the third is developed, color-enzyme forming
cells, inactive in the formation of the second pattern, become active and pro-
duce the dark tergal markings of the last larval stage. These tergal markings
develop first about the primary centers, but they are soon joined by deposits of
pigment, and thus form the anterior and posterior bands.
The pigment and its formation is the same in both sets of colors ; but why
does color appear in some centers in the first stage and disappear entirely in
the second, and then reappear in the third stage in increased area ? Here are
fundamental processes which will be considered at some length later on.
As has been stated, all traces of the cuticula colors are removed at ecdysis,
and only the general body color is left to the larva. On its emergence from
the old larval skin the larva is yellow, red, or some other color, depending
upon the color of the body, and is without a trace of dark markings, excepting
the pigment of the ocelli on the head. The first dark colors to develop are the
spiracula and wing spots, which are a homodynamous series of color centers,
and the color in these areas ma}' even begin to form before the larval skin is
cast off. The color changes have been described elsewhere and need not be
restated. Color next appears in the appendages of the head — first in the
mandibles, next in the palpae and other mouth parts, then in the legs, and
finally in the baso-pleural centers. In figs. 1 to 4, on plate 18, is figured
the development of the second color pattern of signaticollis, and in figs. 5 to 8,
on the same plate, that of the third color pattern for the same species.
ONTOGENY OF LARVAL COLOR PATTERNS. 151
At ecdysis in this genus the color pattern or the dark elements thereof
develop around the centers of color and from there spread out to form larger
colored areas. In sequence the development is as follows : ( i ) Spiracula and
wing, (2) baso-tergal, (3) inner tergal, (4) middle tergal and inner sternal,
and (5) outer tergal and middle and outer sternal. The anterior members
appear earlier and are at all times more completely developed than are the
posterior members of the same homodynamous series. Exceptions to the
above are found only in some few species, in the larvae of which the sternal
centers develop before the tergal. These exceptions are of such small mo-
ment in larval coloration that they may be passed over as of no significance.
It is of interest to note that the order of the appearance of color centers at
ecdysis and of their development in the embryo are identical. In fact,
ecdysis, at regularly recurring periods reduces the color pattern of a larva to
an embryonic condition from which may arise the same color pattern or a new
one; but the new pattern is always developed around the same old centers.
In the ontogeny of the coloration of the larvae each species studied passes
through a series of stages, beginning with one more or less simple or gener-
alized and ending with another of a more special character. Some species, as,
for example, undecimlineata and signaticollis, start with a very generalized
color pattern and pass rapidly through stages which end in a fairly specialized
condition. Other species, oblongata and dilecta, do not begin their larval
coloration with as generalized a pattern as does undecimlineata and others,
but with a pattern which corresponds very exactly to the second stage in the
development of the color in undecimlineata. In these instances have we an
example of Hyatt's principle of acceleration? And have we here also the
existence of localized stages in development corresponding to phyletic stages ?
These are questions whose discussion can best be postponed until after we
have examined the data of the ontogeny of imaginal coloration.
The ecdysis at pupation removes all traces of larval cuticula colors, so that
such colors as exist during the pupal stage are developed therein. The pupal
stage, which is passed in the ground, is in most species characterized by
an absence of color, excepting the spots upon the abdominal tergae and
sternae, which in some species, as, for example, in signaticollis and oblongata,
are modified into black bands. These areas of markings, which can be of no
possible use to the animal, as it is completely hidden in the earth, represent
simply the revival of the color-producing power of the cells in the tergal
and sternal centers. All this pupal coloration, however, has no relation to
that of the imago, as it is cast away at emergence. The imaginal color
develops either under the pupal skin before emergence or in the hours imme-
diately following the final transformation.
152 COLORATION IN LEPTINOTARSA.
IN THE IMAGINES.
In all the species of Leptinotarsa which I have studied, the development of
the imaginal color pattern begins in the latter part of the pupal stage and is
completed after emergence while the young imago is still in the pupal cell.
The full intensity of the adult coloration is not attained until after the beetle
has emerged from the ground and has been feeding for several days, that is,
just before it begins to breed. Soon after reproduction has commenced, how-
ever, the colors begin to fade.
The Ontogeny of Color on the Epicranium and Pronotum.
The color patterns on the epicranium and pronotum begin to develop on the
average about three days before the final transformation ; but in some species,
and in individual instances, the time may be longer or shorter. Thus, in
decemlineata, color may appear on the epicranium four or five days before the
insect emerges, while in oblongata, dilecta, and other species there may be no
trace of color upon these parts until ten or twelve hours before the final trans-
formation.
The color development upon the epicranium of undecimlineata is typical
of those species which have a pattern composed of spots of dark cuticula color
upon a lighter background. On plate 18, figs. 9 to 14, are represented six
stages in this development. In the first stage (fig. 9) two lightly traced areas
upon the epicranium are observed as two faint yellow-brown spots. These
mark the beginning of the two centers (g and g') which in later stages fuse
to form a median heart-shaped spot. On the pronotum no color is yet visible.
A stage closely following this is represented in fig. 10, with the color on the
epicranium deeper and with two new faint yellow-brown areas on the prono-
tum in the position later occupied by the spots a and a'. Simultaneously two
color areas have appeared in the outer posterior portion of the epicranium,
which are the beginnings of the spots h and h'. In each of the spots a and a'
on the pronotum there are traces of anterior and posterior centers which are
usually visible for a brief period. In the third stage (fig. 11) the spots g and
g' have grown considerably larger and have fused across the median line pos-
teriorly, while h and W have also grown and extended medianward. On the
pronotum a and a' have increased in depth of color and have spread peripher-
ally, while anteriorly new areas, the beginnings of b and b' and c and c',
have appeared, and posteriorly e and e' are recognizable. In the following
stage (fig. 12), g and g' have grown larger and have also extended poste-
riorly until they almost unite with h and h' in the median posterior portion of
the epicranium. This union is completed in the next stage (fig. 13). Upon
the pronotum d and d' and / and /' now appear, and the a, b, c, and e spots
have grown in size and depth of color. In the remaining stages represented
(figs. 13 and 14) the color pattern is finished; g and g' completely fuse with
ONTOGENY OF ADULT COLOR PATTERNS. I53
each other and also with h and h' (fig. 14), so that the epicranium has a very
considerable portion of its surface black. On the pronotum a and a' fuse
posteriorly with each other to form the usual V-shaped marking, and ante-
riorly with b and V , while on either side e and / fuse with each other (fig. 14).
The changes in the color of the cuticula pigments from a faint yellowish-
brown through brown, deep brown, and opaque brown which appears black,
are shown on the plate and are the ontological series of changes usual with
colors of this class.
During the development of the cuticula color of the epicranium and prono-
tum the hypodermal color changes little, and the insect emerges as an imago
with the pale yellow-white tint shown in figs. 9 to 12. Neither does the
hypodermal color change to any extent during the life of the adult, excepting
in the latter part, when it becomes white.
For the purpose of comparison, the development of the color pattern of
the pronotum and epicranium of multitccniata is shown on plate 18 in figs.
15 to 20. In this series the figures show, as far as the cuticula colors are con-
cerned, the same order of appearance of spots and increase in size and the
same directions of fusion between the elementary centers as those found in
undecimlineata. We note a change in the hypodermal color, however, in that
there has been a slight development of a yellowish tinge ; and after the insect
leaves the ground and begins to feed this color appears rapidly, soon attain-
ing the intensity characteristic of the sexually mature beetle (fig. 20). In
this feature multitaniata shows a higher or more advanced degree of devel-
opment than that found in undecimlineata. In other respects the type of
ontogenetic development found in both species is, as far as the cuticula colors
are concerned, the same as that found in all of the lineata group. In the
development of the color on the epicranium and pronotum of melanothorax
(pi. 18, figs. 21 to 26), there are some features of interest. Comparison of
figs. 21 and 22 of melanothorax with figs. 9, 10, and 11 of undecimlineata
shows little difference between the two species, excepting that melanothorax
has crowded as much into two stages as undecimlineata has in three ; or rather,
melanothorax has pushed back to an earlier period in ontogeny the appear-
ance of certain spots. In figs. 23 and 24 of melanothorax a difference begins
to appear in that on the anterior border of the pronotum the anterior end of a,
together with b and c, become involved in a general band-like fusion, and the
same takes place on the posterior end with a and d and c ; thus are formed the
anterior and posterior bands, which spread rapidly (fig. 24) and fuse until
they finally cover the entire pronotum (fig. 25) excepting a narrow margin,
which is also soon obliterated (fig. 26).
The cuticula colors in these three species, especially those upon the prono-
tum, have the same color centers as those which appear in the ontogeny of the
larva! colors, and they follow the same course of development. There are dis-
154 COLORATION IN LEPTINOTARSA.
tinguishable three anterior and three posterior centers, corresponding to the
inner, middle, and outer pairs of tergal centers found in the larvae. The inner
and outer pairs of centers develop before the middle pair; and lateralward
fusions form anterior and posterior bands such as are found in the larvae
and in the ontogeny of mclanothorax. These fusions are of exactly the same
type as those shown in the abdominal segments of larvae of signaticollis.
Suggestive also along this same line are stages in the ontogeny of species
like violesccus and dilecta, which have a unicolorous pronotum. Stages in the
color ontogeny of violesccus are represented on plate 18, figs. 27 to 32, and
of dilecta in figs. 33 to 38. Figs. 27 to 29 show features of color levelopment
in violesccus comparable with stages found in mclanothorax and to be inter-
preted in the same manner.
Those species of Leptinotarsa which have unicolorous epicrania and pro-
nota and some development of chemico-physical colors also present interest-
ing stages of growth. Illustrations of the development of the color pattern ot
two such species, dilecta and violescens, have already been given and the
earlier stages have been discussed. In the later stage the relation of the cutic-
ula color to the appearance of the metallic luster of the adult is seen. In the
stages shown on plate 18 it is evident that the cuticula color must precede
the physical portion of the chemico-physical coloration and must reach a con-
siderable degree of intensity before the latter can become visible ; for as the
cuticula color gains in intensity and absorptive power the chemico-physical
color becomes more and more apparent.
The Ontogeny of Color on the Wings,
on the elytra.
In Leptinotarsa no color develops on the elytra until after the animal has
left the pupal stage and the elytra are fully expanded and possess a relatively
firm texture. After their emergence from the pupa and the development of
their adult form, these beetles, like most of the Coleoptera, spend many hours
and often several days in the absolutely dark pupal cell ; and it is in this period
of adolescence that the wing colors are developed.
The ontogeny of coloration in the elytra of L. undecimlineata is shown on
plate 19, figs. 1 to 6. In fig. 1 the elytron is represented as it appears several
hours after emergence, when it is of a uniform pale yellow-white with delicate
traces of dark color, and when the trachea are well filled with air and stand
out conspicuously against the darker opaque body-color beneath. The first
well-developed traces of markings are shown in fig. 2, where faint yellow-
brown areas of color appear anteriorly in the costal and subcostal spaces and
extend distalwards over perhaps one-third the length of the elytron. This
color extends posteriorly, deepens rapidly, and also appears in the other inter-
spaces of the wing (fig. 2) ; and this process is continued until each of the
Explanation of Plate 18.
This plate illustrates the development of coloration in the larvae at
ecdysis, and upon the head and epicranium of the imago in the pupal stage.
Figs, i, 2, 3, and 4, L. signaiicollis Stal, illustrate the change
of the larvae at the second ecdysis. From Tierra Blanca,
Vera Cruz, Mexico.
Figs. 5, 6, 7, and 8, L. signaiicollis Stal, illustrate development of
coloration at the second ecdysis. From Cuernavaca, Morelos,
Mexico.
Development of coloration upon the epicranium and pronotum of the
imagines:
Figs. 9 to 14, L. undecimlineata.
Figs. 15 to 20, L. multitaniata.
Figs. 21 to 26, L. melanolhorax.
Figs. 27 to 32, L. violescens.
Figs. 33 to 38, L. dilecia (green form).
W. L. TOWER
PLATE 18
'-
f\*
r\v
m *^# <^^
„ .., - - .
• • • • $&J «fc
..«/;. .^l* vVu
.*». .-'*
cubitus- ',-•*'
anal- " a
transverse vein- _ f»-
costal"^^^
subcostal" *
discal cell"
i
anal border.
costal border.
transverse vein - ~rz
discal cell
anal
cubitus
medius
rauious
Text-figurk io. — Diagrammatic figures of an elytron (A), a coleopterous hind
wing (B), and a butterfly /ore wing (C), to show homology of parts and the
obliteration of the margin in elytra.
The wings, which are flattened, bag-like evaginations of the body wall,
variously modified and ornamented, carry the most complicated, and, for cer-
tain lines of inquiry, the most significant parts of the color pattern of the
entire insect. The coloration of the wings has justly received the lion's share
of attention from those studying insect coloration. This is especially true of
the Lepidoptera.
l62 COLORATION IN LEPTINOTARSA.
The normal position of the wing is that in which the axis forms a high
angle with the longitudinal axis of the body and is not parallel with it, as some
writers have claimed. Each wing is composed of three clearly marked areas :
First, the anterior, which is supported by the costal and subcostal veins and
which we shall call the anterior system; second, the middle area, devoid of
veins, or the discal cell lying between the subcostal and ramous veins ; and
third, the area supported by the ramous, cubitus, medius, and anal veins, which
we shall call the posterior system. Each of these areas is clearly seen in the
wings of all insects, and even in the specialized elytra of Coleoptera (see text-
fig. 10). In text-figure 10 c are represented these areas in the wing of a but.
terfly, and in text-figure 10 B in the hind wing of a Lcptinotarsa. Although
there are great differences between these, the homologous areas are clearly
seen. In text-figure 10 a the elytron, although specialized, shows the anterior
system condensed and crowded into the costal border, the tip of which is
further bent caudalward until it meets the anterior vein of the posterior
system, the ramous. At the junction of these the apex of the elytron is
formed, and between them is inclosed the discal or central space. Posteriorly
the anal and cubital veins are crowded together. In the elytron the costal
border is not diminished, but the margin (text-fig. 10 a) is obliterated, and
the costal border and the outer extremity of the posterior system meet, while
the anal area is condensed. The least modified area of all is the central or
discal area. This is often divided, almost always in many orders, by a trans-
verse vein from the subcostal to the ramous (text-figs. 10 a and 10 b), and in
the elytra of beetles this position is marked in many genera and families by a
cross vein of varying size and distinction. Although highly specialized, the
elytron is clearly homologous to the wings of other insects.
In the ontogeny of elytral coloration I have shown that color appears ear-
liest at the proximal portion of the wing in the interspaces between the veins,
and first of all in the central space, and that from these beginnings it spreads
distalward in species with color stripes between the veins, and across the veins
in unicolorous species. Of importance is the fact that in Coleoptera, as in
Lepidoptera, as has been shown by numerous writers, color should first appear
in the interspaces between the veins at the base of the wing, in the cen-
tral cell, and the area immediately anterior and posterior thereto. There is
therefore between the wings of these orders a complete homology, not only of
the larger parts, but also of the initial stages of color development. In the
further development of the color pattern of the wings of these beetles it
appears from the observations given that color, whether it be spot or stripe,
develops first between the veins and later over them. In this they are in per-
fect accord with the Lepidoptera. In the development of the transverse bands
of spots, which represent a system not found in Lepidoptera, color is found
over the veins. These transverse bands of spots consist of the basal band at
the proximal end of the wing, the distal band at the apex of the wing, which
LOCALIZED STAGES IN DEVELOPMENT. 163
is the condensed marginal band of other insects, and the outer, middle, and
inner transverse bands, of which the middle lies over the transverse vein
across the distal cell.
The wing in development starts from a minute invagination of cells in the
region of the wing spot, which is an area, as shown by Verson, myself, and
others, homologous to the spiracular center of other segments. This invag-
ination grows inward, and the dorsal portion thickens and grows downward
into a tongue-like process either in a pocket of the hypodermis or between the
hypodermis and cuticula. These thickenings are found very early, as are also
the cell arrangements of the anterior and posterior systems and of the central
space. Likewise, before pupation, the interspaces between the veins are fully
formed. Later on, in these interspaces in the pupal wings, the enzyme-form-
ing cells are found, but of their earlier history we are as yet quite ignorant.
The manner of the growth of the wing and the arrangement of the glands,
etc., in parallel bands strongly suggest that there exist in the wing fundament
of all insects, mother cells, which, by division, give rise to the rows of pigment-
enzyme forming cells in the interspaces, and that from these rows they spread
in all directions in species of a unicolorous condition, or in special regions
only in banded species. It would be difficult to account for these conditions
of color development upon any other basis ; and when to this is added all the
data of variation, of phylogeny, and of experiment, no other hypothesis seems
tenable. It is known that the whole process of wing development is similar
in all insects, even, as I have shown, in forms which are as diverse as those
with incomplete and complete metamorphosis ; and inasmuch as the initial
stages in color development are identical in all, it is highly probable that the
above hypothesis is correct. With this hypothesis and the data of ontogeny
which so strongly support it, and with our knowledge of the existence of
color centers upon the body from which the color patterns develop, we have
a basis for the working out and logical interpretation of the phenomena of
localized stages in growth and of the laws of the evolution of insect colora-
tion and the various phenomena connected therewith.
LOCALIZED STAGES IN ONTOGENY OF COLORATION.
The existence in the young of ontogenetic stages or conditions simulating
adult characters of related species or genera, ancient and modern, has been
abundantlv described and studied in plants, and especially in fossils of both
plants and animals. The idea that embryonic, larval, and juvenile conditions
are the atavistic reappearances of ancestral adult stages, an old idea revived
and modified by L. Agassiz and later crystallized by Haeckel into the dictum
that "Ontogeny repeats phylogeny," has been greatly misused, especially by
paleontologists, although it is in many cases a valuable aid in phylogenetic
study. Jackson, Cushman, and others have described in plants and animals
164 COLORATION IN LEPTINOTARSA.
having periods of interrupted growth, stages and characters found at the
resumption of growth which they believe to be the atavistic reappearances of
ancestral adult characters. More recently, Shull, investigating these stages
in Slum cicutcrfolium, comes to the conclusion that these stages represent, not
ancestral adult characters, but simpler conditions of structure which are
explainable upon physiological instead of phylogenetic grounds. In insects
these stages have not, as far as I am aware, been investigated.
In the ontogeny of both the larval and imaginal coloration in Leptinotarsa
there are clearly distinguishable the following stages: (a) The preparatory
stage, the time when the color is the uniform yellow of the pupa, and the
period of active zymogen formation; (b) the initial stage, during which the
color appears over the centers of coloration and the color pattern is marked
out; and (c) color intensification stage, when the colors develop to their full
intensity. These stages are clearly defined and recognized in all insects
whose color development has been studied. Do these three stages, which are
common to all insects in their development, represent three great steps in the
evolution of the coloration of the class ? Are we to believe that the prepara-
tory stage is the atavistic repetition of the adult coloration of the hypothetical
ancestor of the tracheate phylum? that the initial or second stage represents
the first developments of coloration, and the third the condensed recapitula-
tion of more modern specialization and differentiation in insect colors ? Such
a view is perfectly in accord with those of Cope, Hyatt, and others. The
interpretation of these stages, widespread though they are in insects, does not
rest with studies in phylogeny ; it is a question solely of the investigation of
adaptation to ontogenetic ends ; that is, they are physiological and develop-
mental, and not phylogenetic phenomena.
Almost all tracheates are subject to periodic changes of the chitinous por-
tion of their body wall, and at these ecdyses the color pattern, as far as the
markings arising from the dark cuticula pigment are concerned, is removed
and must be replaced by a new development thereof. The removal of the
chitinous layer reduces the animal to a condition which, as far as the integu-
ment is concerned, is practically embryonic. This is a fact which any one
who wishes may verify. From this condition the animal must retrieve itself
as soon as possible by the development of a new cuticula carrying new color
markings. At each recurring ecdysis this process is repeated, and the three
main stages are passed through. We admit that ecdysis is necessary to
growth in these forms, and that growth is in a sense periodically interrupted
by ecdvsis. It is probably true that in the evolution of insect color there were
three stages corresponding roughly with the stages in development already
described ; but that these should be repeated so exactly at each ecdysis is a
phenomenon the reasons for which it is hard to imagine. On the other hand,
its interpretation on the basis of an ontogenetic adaptation alone would involve
LOCALIZED STAGES IN DEVELOPMENT. 165
no difficulties in reasoning. Ecdysis removes all dark color from the body,
and this is redeveloped, and in development passes through stages which may
in a broad way simulate stages which we suppose may have existed in the
general evolution of color. The fact that in diverse orders of insects the same
conditions are found is not necessarily evidence for its phylogenetic interpre-
tation, but rather for the opposite, because the process of ecdysis is the same
in all insects, and like destructive and constructive stages and processes are
common to all. It is also true that the supposed phylogenetic stages are
products of our imagination, creatures brought into being for the purpose of
making observed conditions in ontogeny agree with cherished hypotheses.
Direct evidence upon this question is of course difficult to obtain, and it must,
at least for the present, remain one of opinion as far as its larger phases are
concerned. The smaller and more special phases, however, may perhaps be
more easily interpreted, and to these we shall next direct attention.
On plate 17 are shown the succession of color patterns found in the larvae
of twelve species of Leptinotarsa. These figures represent well the different
types of species and color patterns found. In them we have clearly marked
"localized stages in development," and the question to be considered is
whether they are to be interpreted on the basis of atavism or on that of
physiology and development.
On superficial examination it is strikingly apparent that in the lineata
group, for example, the larvae all have a similar generalized color pattern in
the first larval stage, that this is succeeded by one quite different in the second
stage, and in some species by still another in the third instar. When we trace
back the phytogeny of the genus we find that it goes directly to Zygogramma,
where many of the adult larvae have color patterns like those shown in the
second stage of so many Leptinotarsas. Back one step more in the ancestry
is Calligrapha, in which many of the adult larvae have spines and color pat-
terns like the young of Leptinotarsa. What further proof of biogenesis could
we desire? In the first instar of Leptinotarsa we have a Calligrapha condi-
tion, in the second a Zygogramma condition, and finally the pattern charac-
teristic of Leptinotarsa is attained. If we accept in all sincerity Haeckel's
dictum and apply it in the rigid manner of Cope, Hyatt, Jackson, and others,
the superficial evidence appears complete.
In the face of evidence apparently so indubitable, I admit that the chances
that any other explanation will meet with favor are poor indeed ; yet I am
thoroughly convinced that these stages figured in the species of Leptinotarsa
on plate 17 are not to be explained upon the basis of atavism, but upon that of
the morphological and physiological constitution alone, and that the connec-
tion with ancestral or racial conditions is only a general one. Evidence in
support of this side of the question is not entirely wanting.
IDD COLORATION IN LEPTINOTARSA.
It has been shown that color in these beetles is limited in its appearance to
definite centers, and that all patterns must, as far as is known, originate from
continuations of these centers, so that color patterns are in a degree limited.
This character is one of such wide distribution in insects that I consider it to
be a character of the phyletic race — one that is so firmly fixed in the constitu-
tion of the tracheate protoplasm that it can not be eradicated. It is one of
those phyletic characters whose origins are unexplained excepting by theory.
In the Coleoptera, and especially in the Chrysomelidas, the existence of these
centers in development is beyond doubt. Their presence gives to each species
a like background out of which each species creates its own distinctive color
pattern. That is, all start in ontogeny endowed with an identical arrange-
ment of color-producing centers, which become variously modified, sup-
pressed, or accentuated in different stages and in different species.
Strong evidence against the idea of specific ancestral influence is found in
the three species iindecinilincata, multitccniata, and oblongata, L. undecim-
lineata is undoubtedly the ancestor of multitccniata, and oblongata arose
from the latter. On plate 17, figs. 1 to 3, are given the three stages of
uudeciniliueata, beginning with a condition which is generalized and ending
with that shown in fig. 3. L. multitccniata (figs. 10 to 12) never goes as far
in the modification of its color pattern as its ancestor undecimlincata, while
oblongata starts in life with a condition far beyond its parent in its specializa-
tion. Here is a condition difficult to explain upon the basis of a literal inter-
pretation of the biogenetic law, of a species starting its ontogeny far in
advance of the adult parental condition. This surely is Hyatt's idea of accel-
eration in development carried on with great speed. Now, I find that the
cells which form the centers of coloration are developed in the embryo of
oblongata, but that they remain quiescent until the pupal period, when they
become active and produce color areas in the proper place, as is shown on
plate 20. I am not able to understand how this condition found in oblongata
can be the result of ancestral influence. Surely there are no indications of the
process in the parent, and in the grandparent species the same modifications of
color have taken place in the larva in the second instar (fig. 2). The extreme
advocate of the law of biogenesis will in this instance claim that through
acceleration in development the earlier stages have been dropped off in the
embryo, so that the larva of oblongata starts in life with the mature color pat-
tern of its grandparent species. This would all be very well were it not for
the presence of the color-forming cells in the late embryo, and the fact, above
mentioned, that they remain inactive throughout the larval life and become
active and produce color areas in the pupa. This can not be successfully
explained by the theory of temporary retardation of development, because we
have the phenomenon slightly varied in many species. Nor do I see how how
the idea of the recapitulation of the immediate ancestral stages in growth can
LOCALIZED STAGES IN DEVELOPMENT. 167
throw any light upon the subject. These localized stages in the development
of the color pattern of Lcptinotarsa are of value phylo genetically only as
indicating a similarity of constitution and organization between species.
That each stage represents the repeated color pattern of some ancestor is
a view not in accordance with the conditions shown to exist in this genus of
beetles.
I have interpreted these localized stages on the basis of similarity of orig-
inal constitution and modification in development. The successive stages in
Lcptinotarsa are not, however, simpler conditions, but more specialized. They
are not comparable to the condition found by Shull in Sium, where the local-
ized stages are simpler states ; although in both cases the explanation is phys-
iological and developmental and not phylogenetic.
/;; Lcptinotarsa each species, as far as we can discover, starts in larval
development endowed with a definite system of color-enzyme producing cells;
that is, all start alike from a racial condition. Given this racial endowment,
each species from the start modifies, holds in check, or increases the activities
of the centers in its own manner, without any dependence upon the actions of
its immediate parental or grandparental species. In the evolution of the color
pattern, in the rise of new species, each species inherits this general racial
system of coloration entire in its germ plasm, and the fundaments thereof
appear in development. In the production of the nezv race the capacity to
modify this general color scheme is inherent in the germ plasm, and in hered-
ity is transmitted to the offspring of the same kind as in the parent; but in the
new race it is changed, the modified capacity producing new developments of
the color centers which we recognize as the differentials of the nezv races
or species.
I am convinced that in these beetles at least the apparent localized stages in
ontogeny are nothing more than the physiological expression of similarities
in constitution and in developmental tendencies.
Likewise the stages in development found in the imaginal color pattern are
best and most logically explained in this manner. Ancestry exerts its influ-
ence only by providing a general racial or generic background out of which
each species takes characters and variously modifies them as its specific con-
stitution demands, to create its color pattern in the different stages of its
ontogeny. Each species, however, is absolutely dependent upon this ancestral
background for its characters, and can not, under any conceivable conditions,
develop traits or characters which are not based upon some trait or character
existing- in the ancestral background.
i68
COLORATION IN LF.PTINOTARSA.
THE EXPERIMENTAL MODIFICATION OF COLOR AND COLOR
PATTERNS.
Animal colors are to a considerable extent direct!}* modified by the action of
external agents, such as temperature, moisture, food, altitude, and light. In
the genus Leptinotarsa variations in color corresponding to different localities
or seasonal changes are common, and these have been considered under the
topics of place and geographical variation. The methods of biometry are
not, however, adequate for the analysis or interpretation of these color
changes ; as from this source we gain knowledge only of the existence of the
phenomena, of their extent, and of the laws governing the observed changes.
We shall now inquire into the causes of color changes — their nature and
extent.
MODIFICATION OF COLOR IN L. DF.CF.MLINEATA.
My experiments in the modification of both larval and adult color in decem-
lincata cover a period of ten years, and comprise work with temperature,
moisture, food, light, soil, composition of the atmosphere, and atmospheric
pressure.
TEMPERATURE EXPERIMENTS.
Temperature experiments were begun in 1894, and have been continued
each year since then. The results are given in condensed form, the data of
many experiments of the same kind being combined into one comprehensive
statement.
Experiment i. — To determine the effect of a slightly increased temperature upon the
color and color pattern of L. decemlineata.
Conditions. — Temperature on the average 5.780 C. above that in nature.
Other conditions normal.
Apparatus. — Glass cages and glass breeding tanks, warmed by the sun and
protected from radiation at night.
In these experiments, which were conducted during the years 1894 to 1901,
5,500 larvae of this beetle taken at random from potato fields were used.
They were placed in the conditions of the experiment in late larval life, and
remained there until the full imaginal colors had been developed. The tem-
perature records for the series are given in the following table :
Table 44. — Temperature conditions.
Conditions —
7 a. m.
1 p. m.
8 p. m.
Maxi-
mum.
Mini-
mum.
Aver-
age.
Deviation
from
normal.
In nature
In experiment.
°C.
19.2
23-9
° C.
30.62
33-5
0 C.
I7.389
27.17
°C.
33-5
38
"C.
13.0
19
22.4°3
28.188
°C.
0
+5.78
MODIFICATION OF COLOR.
169
In these experiments 10.8 per cent died in the larval stage, 46.1 per cent in
the pupal, and 43.1 per cent completed their transformation. Of the latter
there was about an equal division as regards sex, there being 51.5 per cent
males and 48.5 per cent females. The size and form were normal.
In the shape, position, and fusions of the spots the color pattern was un-
modified ; but the size of the spots or the amount of dark color was increased
so that as regards general color there was a slight shifting of the modal con-
dition. This is shown in the following table, which gives the conditions of
the parent generations and those of the material used in experimentation and
in the control :
Table 45- — General color of beetles used.
Class
5
6
7
8
9
10
11
12
13
14
15
P.ct.
I
16
17
P.ct.
I
Parents ....
Control
Experiment .
P.ct.
I
P.ct.
2
2
P.ct.
5
7
1
P.ct.
20
10
I
P.ct.
42
23
5
P.ct.
18
30
II
P.ct.
9
16
18
P.ct.
2
6
46
P.ct.
I
5
10
P.ct.
I
5
P.ct.
I
Empirical mode of parents 9
Empirical mode of control 10
Empirical mode in experiment . . 12
Modal deviation of parents o
Modal deviation of control -p-1
Modal deviation in experiment. . +3
In all of these experiments there was a slight histonic skewness of from
one to four classes in each separate lot of individuals experimented upon, the
average being three classes. The skewness is histonic in this case toward a
more melanic condition, such as characterizes the ancestor of this beetle,
L. mnltitcvniata.
Experiment 2. — To determine the effect of a considerable increase in temperature upon
the color and color pattern of L. decemlineata.
Conditions. — Temperature on an average 9.7670 C. above that in nature.
Other conditions normal.
Apparatus. — Glass cages and special breeding tanks warmed by the sun or
by artificial heat, and protected from radiation at night.
In this experiment, in which are combined the data and results of many
experiments extending from 1893 to 1904, 7,800 larvae taken at random from
potato fields were used. These were all placed in the conditions of the ex-
periment in the last larval instar before the beginning of the prepupal period.
Food, moisture, light, soil, and other conditions were kept normal, the tem-
perature only being varied. The temperature records are as follows :
Table 46. — Temperature conditions.
Conditions—
7 a. m.
1 p. ra.
S p. ni.
Maxi-
mum.
Mini-
mum.
Aver-
age.
Deviation
from
normal.
In nature
In experiment. .
°C.
19.2
25
° C.
30.62
38.5
°C.
17.389
33
°C.
33-5
41.4
°C.
13
22
°C.
22.403
32-17
0 C.
O
+ 9- 767
170
COLORATION IN LEPTINOTARSA.
In these experiments 55 per cent died in the larval stage, 45 per cent in the
pupal, and 5 per cent completed their transformation. These were normal in
shape, but 23 per cent below the average in size. The color and color pat-
terns were changed as follows : The hypodermal color became pale yellowish-
white, and the areas of cuticula pigments were considerably altered. All of
the elytral stripes were shortened and narrowed, and showed no fusions ; the
spots on the pronotum and epicranium were also without fusions and small,
/ and f being continually absent, and d and b being represented only by traces
of color. On the ventral surface nearly all the color was lost, traces only of
the outer sternals remaining, and rarely of the inner sternals. The general
appearance of the imagines and their seriations as regards general color are
shown in the following table :
Table 47. — General color of beetles used.
Class
1
2
3
4
5
6
7
8
9
10
11
12
13
14
P.ct.
P.ct.
P.ct.
P.ct.
P.ct.
P.ct.
P.ct.
P.ct.
P.ct.
P.ct.
P.ct.
P.ct.
P.ct.
P.ct.
I
2
a
20
42
18
9
2
1
Control
2
7
IO
23
30
16
6
s
I
Experiment .
2
4
6
34
22
II
9
5
4
2
I
Empirical mode of parents 9
Empirical mode of control 10
Empirical mode in experiment . . 4
Modal deviation of parents o
Modal deviation of control .... +1
Modal deviation in experiment. . — 5
In each of the separate experiments comprising this second set the resulting
individuals showed a strong prophetic skewness toward an albinic condition.
This was due to the general reduction in size of all areas of cuticula pigment,
to the loss of many, and to the change in the hypodermal color already noted.
Experiment 3. — To determine the effect of a high average deviation of temperature
upon the color and color pattern of L. decemlincata.
Conditions. — Temperature on the average 12.0970 C. above that in nature.
Other conditions normal.
Apparatus. — The same as in experiment 1, artificial heat, however, being
used to some extent.
Ten thousand one hundred larvas of L. decemlincata, taken at random from
potato fields, were subjected to the conditions of this experiment, which
extended over the years 1894 to 1904. All the larvae were placed at the
beginning of the last instar, a temperature condition similar to that found in
nature, and the temperature was raised in four or five days to the desired
degree. Food, moisture, light, soil, etc., were kept normal. The tempera-
ture records are given in the following table :
MODIFICATION OF COLOR.
Table 48. — Temperature conditions.
171
Conditions —
7 a. m.
4 P- to.
S p. m.
Maxi-
mum.
Mini-
mum.
Aver-
age.
Deviation
from
normal.
In nature
In experiment.
°C.
19.2
2S
°C.
30.62
45-5
°C.
I7-389
3°
33-5
47
13
25
°C.
22.403
34-5
°C.
O
+ 12.097
In these experiments 95 per cent died in the last larval stage, especially in
the prepupal period, 4 per cent in the pupal period, and 1 per cent appeared
as imagines. These latter were equally divided as regards sex, and were
39 per cent below the average in size. The hypodermal color was pale
yellowish-white, without a trace of lipochrome pigments, and the cuticula
color was much reduced. On the epicranium all spots were completely
absent ; while on the pronotum b, d, f, and e were absent, c was very small,
and a was divided into two spots, the anterior and posterior. On the elytra
the costal, anal, and cubital stripes were either wanting or very rudimentary,
while the others were reduced in size and were light brown in color. Upon
the ventral surface the legs had no dark color and all spots were reduced,
excepting the outer sternals, which were represented by the merest traces
upon the thoracic and first two abdominal segments. The general color of
these beetles is shown in the following table :
T
ABLE
49."
-General
color of beetles used.
Class
z
P.ct.
2
P.ct.
3
P.ct.
4
P.ct
5
Pet.
6
P.ct
7
P.ct.
8
P.ct.
9
P.ct.
10
P.ct.
11
P.ct.
12
P.ct.
>3
P.ct.
14
P.ct.
Parents
I
2
S
20
42
iS
9
2
I
Control
2
7
IO
23
30
16
6
5
I
Experiment. .
3
14
38
23
15
4
3
Empirical mode of parents 9
Empirical mode of control 10
Empirical mode in experiment . . — 3
Modal deviation of parents o
Modal deviation of control' +1
Modal deviation in experiment . — 6
In experiment 3 the limit is almost reached of the average deviation of tem-
perature to which it is possible to subject these insects without their being
previously adapted to a higher temperature, an average increase of i° C.
above the condition of experiment 3 resulting in a mortality of 100 per cent.
Experiment 4.— To determine the effect of a slightly lowered average temperature upon
the color and color pattern of L. undecimlineata.
Conditions. — Temperature on an average 6.50 C. below that in nature.
Other conditions normal.
Apparatus. — Special breeding tanks with inner and outer glass walls, be-
tween which there were 2l/i to 3 inches of constantly changing cooled water.
Ice boxes were also used.
172
COLORATION IN LEPTINOTARSA.
In this experiment, which extended over the years 1895 to 1903, 6,000
larvae were used, taken at random in the last instar. In the earlier experi-
ments, those covering the years 1895 to 1898, an ice box was used ; and in the
later ones the special tanks. Food, moisture, light, and other conditions were
kept normal. The temperature records are as follows :
Table SO. — Temperature conditions.
Conditions —
7 a. m.
1 p.m.
8 p. m.
Maxi-
mum.
Mini-
mum.
Aver-
age.
Deviation
from
normal.
In nature
In experiment.
°C.
19.2
14.306
°C.
30.62
1S.409
°C.
I7-339
15
°C.
33-5
23-5
°C.
13
12
° C.
22.403
I5-903
°C.
0
-6.5
In these experiments 4 per cent died in the larval stage, 6 per cent in the
pupal, and 90 per cent appeared as imagines. In size these were normal, or
slightly larger than normal. The average duration of pupal life was
increased one and a half days. The color and color pattern were modified
as follows : The hypodermal color became a more intense yellow, often with
a reddish tinge; the cuticula color became a dense black, and all the areas
thereof were slightly increased in size, which gave the specimens a more
melanic appearance, as is shown in the following table of their seriations :
Table Si. — General color of beetles used.
Class
5
6
7
8
9
10
11
12
13
14
15
16
Parents
Control
Experiment .
P.ct.
1
P.ct.
2
2
P.ct.
5
7
1
P.ct.
20
IO
2
P.ct.
42
23
3
F.ct.
18
30
7
P.ct.
9
16
12
P.ct.
2
6
22
P.ct.
I
5
39
P.ct.
I
10
P.ct.
3
P.ct.
I
Empirical mode of parents 9
Empirical mode of control 10
Empirical mode in experiment.. . 13
Modal deviation of parents o
Modal deviation of control +1
Modal deviation in experiment. . . +3
A striking parallel exists between the seriations of this experiment and
those of experiment 1 in that the conditions of both act as a stimulus toward
an increase in the amount of pigmentation, and hence toward a more general
melanic appearance.
Experiment 5. — To determine the effect of a considerably decreased average temper-
ature upon the color and color pattern of L. decemlineata.
Conditions. — Temperature on the average 9.5° C. below that in nature,
with other conditions normal.
Apparatus. — The same as in experiment 4.
MODIFICATION OF COLOR.
173
In this experiment, which extended over the years 1895 to 1903, 6,300
larvae of L. decemlineata were used. All were taken at random in the last
instar. Food, moisture, light, soil, and other conditions were kept normal.
The temperature records are as follows :
Table 52. — Temperature conditions.
Conditions —
7 a. m.
1 p. m. S p. m.
Maxi-
mum.
»C
33-5
18.5
Mini-
mum.
Aver-
age.
Deviation
from
normal.
In experiment. .
°C.
19.2
9
°C.
30.62
17.109
°C.
I7.3S9
10.600
°C.
13
5-4
°C.
22.403
12.903
O
-9-5
In these experiments 64.7 per cent died in the larval stage, 25.3 per cent in
the pupal, and 10 per cent completed their transformation. Of these 50 per
cent were males and 50 per cent females. The pupal period was on an aver-
age 32 days, or more than three times the duration of this stage in nature;
while the imagines were on an average 6.2 per cent below the normal in size.
The hypodermal color had become lighter, being yellow. The cuticula pig-
ments of the dorsal side, however, were only slightly reduced, and no spots
were entirely absent; but fusions between the spots were not common. On
the ventral surface all spots were reduced in size, but none were absent. The
general appearance was slightly albinic, as may be seen in the following
seriations :
Table S3. — General color of beetles used.
Class
4
5
Pel.
6
P.ct.
7
8
/' r/.
9
Pet.
10
11
P.et.
12
P.ct.
13
P.ct.
14
P.ct.
P.ct.
P.ct-
P.et.
Parents.. ..
I
2
S
20
42
IS
9
2
I
Control
2
7
IO
23
30
16
6
5
I
Experiment.
I
2
7
9
36
20
15
8
1
I
Empirical mode of parents 9
Empirical mode of control 10
Empirical mode in experiment.. . S
Modal deviation of parents o
Modal deviation of control +1
Modal deviation in experiment . — 1
Experiment 6. — To determine the effects of a large decrease in the average temperature
upon the color and color pattern of L. decemlineata.
Conditions. — Temperature on the average 13.40 C. below that in nature,
with other conditions normal.
Apparatus. — The same as in experiments 4 and 5.
In this experiment, which extended over the years 1897 to 1903, 3,000 larvae
taken at random in the last instar were used. The temperature records are
as follows :
174
COLORATION IN LEPTINOTARSA.
Table 54. — Temperature conditions.
Conditions —
7 a. m.
1 p. m.
8 p. m.
Maxi-
mum.
Mini-
mum.
Aver-
age.
Deviation
from
normal.
In nature
In experiment. .
0 £-
19.2
7
° C.
30.62
12.009
°C.
17.339
8
0 C.
33-5
15-5
0 C.
13
6
° C.
22.403
9.003
° C.
0
-13-4
In these experiments 88 per cent died in the larval and prepupal stages,
7 per cent in the pupal, and 5 per cent appeared as imagines, of which there
were 70 females and 80 males. The pupal period averaged 43 days, or over
four times that in nature. The hypodermal color was a light yellowish-white,
and many of the spots of the cuticula were absent or small. On the epi-
cranium g and g' were separate, small, and light brown in color, while h and
hi were often absent. On the pronotum there were no fusions, and b, d, and f
were always very small and frequently absent. The elytral stripes were
shortened and were light in color, while upon the ventral surface the middle
sternals were largely wanting, and the legs were almost devoid of dark color.
The general appearance was decidedly albinic, as is shown by the following
seriations :
Table 55. — General color of beetles used.
Class
4
5 6
1
7
8
9
10
11
12
13
14
Parents
Control
Experiment. .
P.ct.
2
P.ct.
I
10
P.ct.
2
2
24
Pet.
5
7
22
P.ct.
20
IO
20
P.ct.
42
23
12
P.ct.
18
30
S
Pet.
9
16
2
P.ct.
2
6
P.ct.
I
5
P.ct.
I
Empirical mode of parents 9
Empirical mode of control 10
Empirical mode in experiment. . . 6
Modal deviation of parents o
Modal deviation of control +1
Modal deviation in experiment. . — 3
Experiment 7. — To determine the effect of an extremely low temperature deviation
upon the color and color pattern of L. decemlineata.
Conditions. — Temperature on the average 23. 50 C. below that in nature.
Relative humidity increased in some experiments 7 to 10 per cent ; light
decreased to one-half its normal intensity.
Apparatus.- — Glass tanks with jacket of ice and salt between. Use was
also made of an ice box.
In this experiment, which was conducted during the years 1900 to 1904,
2,000 larvse were used. These were taken at random from potato patches in
the last instar and were placed in glass vessels. Food, moisture, and soil were
the same as in nature, but the light was greatly diminished or entirely want-
MODIFICATION OF COLOR.
175
ing. Light, however, is a factor of no importance, as the colors are always
developed in the dark. The temperature records are as follows :
Table 56. — Temperature conditions.
Conditions —
7 a. m.
i p. m.
8 p. in.
Maxi-
mum.
Mini-
mum.
Aver-
age.
Deviation
from
normal.
In nature. . . .
In experiment. .
"C.
19.2
- 3-3°
30.62
o-5
°C.
17-389
- -5
°C.
+3
2
°C.
-13
- 4-5
°C.
22.403
— I.I
"C.
0
-23-5
In this experiment 92 per cent died in the larval and prepupal stages, 6 per
cent in the pupal, and 2 per cent, 20 males and 12 females, completed their
metamorphosis. The pupal period averaged 64 days, or more than six times
its duration in nature. The imagines were all strongly modified toward an
albinic appearance, having a pale yellow-white hypodermal color. The dark
color of the epicranium was wanting, as were also the spots b, d, e, and / on
the pronotum ; c and a were much reduced, and a was divided. On the elytra
the costal and anal stripes were wanting, and the others were reduced in size
and were lighter in color. Upon the ventral surface the legs were devoid of
color and only the outer sternal spots were found, and these only upon the
thoracic and the three or four anterior abdominal segments. The extreme
albinism of the beetles is apparent from the following seriations :
Table 57. — General color of beetles used.
Class
2
3
4
5
6
7
8
9
10
II
12
13
P.ct.
P.ct.
P.ct
P.ct.
Pel.
P.ct.
Pel.
P.ct.
P.ct.
P.ct.
P.ct.
P.ct.
Parents
I
2
5
20
42
18
O
2
I
Control
I
2
6
22
28
2,S
9
s
2
Experiment. .
1
13 *3
29
13
1
Empirical mode of parents 9
Empirical mode of control. . 9
Empirical mode in experiment. ... 4
Modal deviation of parents o
Modal deviation of control o
Modal deviation in experiment .. — 5
Between this experiment and experiment 3 there was a close similarity in
the mortality and modifications produced, and also in the fact that in both
any further decrease of average temperature was fatal to 100 per cent of the
beetles, so that as far as the development of color was concerned, experiments
3 and 7 attained to the highest limits at which color and color patterns could
be produced in these beetles without their being previously acclimatized to
higher or lower temperatures.
In experiments 1 to 7, 40,700 larva; of decemlineata from Massachusetts,
Long Island, Ohio, and Illinois were used, and from these there emerged
13— T
1 76
COLORATION IN LEPTINOTARSA.
8,861 imagines. These beetles were subjected to conditions of temperature
such as might be produced in nature by the coming on of a cold or hot spell
when the larva? were about to pupate, and the continuance thereof during the
pupal period. Such cold or hot spells do occur frequently in nature, and are
productive of modifications in the coloration of animals and plants. These
experiments show what would happen in the way of modified color or color
patterns if by accident or intentionally a lot of nearly grown larvae were to be
transported from, let us say, the average conditions of the northeastern United
0
eviations i
i temperat
ure
- from
the normal, -
used
n experiments
•
2 3 4 6 6 7 8 9 10 1 1 1 2 13 14 15 16 17 13 19 20 21 22 23 24
SO
19
1."
1?
16
15
14
18
12
11
10
s
8
7
e
4
-\
s
S
^1 s
\
sy
>
/.
\
V
H
\
"s
s.
>.
.<>
-
4
3
\z
' —
'
a
1
\
V
I
.Experiments with temperatures above the normal.
..Experiments with temperatures below the normal.
Text-figure ii.— Diagrammatic representation of the results obtained in experiments i to 7, showing
in both high and low temperature experiments the production of metallic conditions by the first devia -
tion followed by the production of albinic tendencies as the temperature is further increased or
decreased. The heavy line across the middle of the figure on class 9 is the modal class of the
parents above and below which the tendency is metallic and albinic, respectively. The deviations in
temperature are given in degrees centigrade.
States to Arizona or New Mexico (experiments i and 2), or to the Yuma
Desert (experiment 3), to Nova Scotia or Newfoundland (experiment 4), or
to Labrador (experiments 5, 6, and 7).
The general results derived from these experiments have been put into the
form of a table in text-figure II. In this figure the high-temperature experi-
ments are represented by the solid lines and the low by the broken lines. In
both sets of experiments the result produced by either a higher or a lower
temperature is the development of a greater amount of pigmentation and a
MODIFICATION OF COLOR. 177
consequcuL melanic tendency in variations. This stimulus in both directions
to increased pigmentation reaches a maximum between 50 and 70 C. devia-
tion from the normal. Beyond these, as the temperature further deviates,
there is a rapid fall in melanism, first to the normal, and then to a condition
below normal, until a marked albinic tendency is found; and this decrease in
pigmentation continues until the zero point is reached, beyond which no pig-
ment whatever is produced. The zero point is reached much sooner, how-
ever, in high-temperature experiments than in the low.
The increased pigmentation caused by a slight raising of the temperature
can be accounted for by the fact, which has been proven by experiment, that
pigmentation is due solely to the action of enzymes, and that the action of
these enzymes is accelerated by an increase in temperature up to a certain
point, and thereafter retarded. With low-temperature experiments, however,
the phenomenon of increased pigmentation is due to the prolongation of the
period during which the enzymes act. The lowered temperature inhibits the
developmental process, and consequently emergence from the pupa; but it
does not interfere to the same degree with the action of the color-forming
enzymes. Hence, even though the amount of enzymes produced in a low
temperature may be less than that produced in a temperature correspondingly
high, the resulting pigmentation may be approximately equal in both cases,
owing to the retardation of development in the lower temperature and the
longer continued action of the pigment-producing enzymes.
From these seven experiments, conditions of coloration were produced in
the imagines which simulate those found in nature in several geographical
areas. This becomes evident from a comparison of the distribution of the
variation in these experiments with that in the tables of place variation. Thus
the results produced by experiments 1 and 4 are like the melanic conditions
found in occasional generations in the northeastern United States ; or a like
parallel may be found by comparing the effects of experiments 2 and 5 with
the place variations of the south and west, or of experiments 3 and 6 with
those in Arizona and New Mexico. This production in experiment of condi-
tions found in nature in different geographical areas is conclusive proof of the
statement made in regard to place and geographical variations that they are
due solely to the natural conditions of existence.
Another point of interest in this connection is the fact that the variation of
the color pattern in experiment follows the laws found in individual variation,
and is in strict accordance with the observations and laws derived from the
study of color-pattern ontogeny.
These experiments are similar to those conducted by Dorfmeister, Weis-
mann, Edwards, Merrifield, Standfuss, Fischer, and others in that larvae were
suddenly transported just before pupation from normal conditions in nature
to the conditions of the experiment. The difference between my experiments
i78
COLORATION IN LEPTINO'fARSA.
and those of others is that a special effort has been made to change one factor
only in the environment. In this I have been successful. In most of the
experiments that have been recorded no attention was paid to relative humid-
ity. When pupae are inclosed in a dish or box in an atmosphere in summer
where the relative humidity is 60 to 70 per cent, and are then placed in an ice
box, the temperature is lowered, but the relative humidity is increased, and
thus two important factors are changed. Or if pupae are placed free in an
ice box the temperature is lowered, and the humidity is also increased almost
to saturation. On the other hand, in high-temperature experiments in which
an incubator is used the temperature is increased and the relative humidity
decreased to a point below that of the driest deserts. In such experiments
two or perhaps more factors are changed, so that the results are due to
changed complexes, and not to deviations in any single factor. In these
experiments a special effort has been made to change the one factor of tem-
perature alone, the relative humidity being kept between 65 and 70 per cent.
The general conclusions derived from these experiments accord well with
those of Fischer working upon Vanessidae, as also with those of a like nature
that have been made by others upon Lepidoptera. In the set of experiments
which follows the procedure has been quite different from that just described,
the animals having been reared throughout the entire generation under the
conditions of the experiment. In these experiments it was necessary that
much smaller numbers of insects be used, but the results are nevertheless
equally as conclusive as those of the preceding set.
Experiment 8. — To determine the effect of increased temperature upon the color and
color pattern of L. decemlineata when applied throughout life.
Conditions. — Temperature on the average 5.330 C. above that in nature,
with other conditions normal.
Apparatus. — Glass breeding tanks, as in experiments 1 and 2.
In these experiments, which were conducted in the years 1898 to 1904,
2,000 larvae were used. The eggs were taken from a state of nature within
eight or ten hours after being laid, and placed in the conditions of the experi-
ment. From the 3,200 eggs used 2,000 larvae were hatched. These were
kept under the best of conditions as regards food and cleanliness. The tem-
perature records in the experiments are as follows :
Table 58. — Temperature conditions.
Conditions —
S a. m.
1 p. m.
S p. m.
Maxi-
mum.
Mini-
mum.
Aver-
age.
Deviation
from
normal.
In nature
In experiment .
°C.
19
23
"C.
31
38
°C.
21
26
°C.
38
39-5
° C.
13
20
0 C.
23.66
29
0
+5-33
MODIFICATION OF COLOR.
i79
During these experiments 12 per cent died in the larval stage, 23 per cent
in the pupal, and 65 per cent, equally divided between the sexes, emerged.
These were 15 per cent below the normal in size, and their development was
accelerated about 3.5 per cent. The hypodermal color was normal or slightly
darker than normal, but the dark markings were increased in size so much
that the imagines presented a distinct melanic appearance. On the ventral
surface especially the increased pigmentation was marked, the middle and
inner sternal spots being found in 80 per cent of the variates. The seriation
of these beetles into classes is as follows :
Table SO. — General color of beetles used.
4
5
6
7
s
9
10
ir
12
13
M
P.ct.
P.ct.
P.ct.
Pet.
Pet.
P.ct.
Pel.
Pet.
P.ct.
P.ct.
Pet.
Parents
I
2
S
20
42
18
q
2
I
Control..
1
I
7
8
17
39
iq
6
I
I
1
4
6
12
20
48
7
2
Empirical mode of parents 9
Empirical mode of control 9
Empirical mode in experiment... 12
Modal deviation of parents ....
Modal deviation of conti ol
Modal deviation in experiment
o
o
+3
The conditions of this experiment are nearly the same as those of experi-
ment 1, the resulting modifications of the elements of the color pattern, of
the colors themselves, and of the whole color effect being almost identical,
as may be seen by a comparison of tables 45 and 59. In this experiment,
however, the mortality was decreased, there being a total difference of 22 per
cent in favor of the latter.
Experiment 9. — To determine effects of an increased average deviation of temperature
upon the color and color pattern of L. decemlineata when applied throughout life.
Conditions. — Temperature on the average 9.5 ° C. above that in nature, with
other conditions normal.
Apparatus. — The same as in experiment 8.
In these experiments, which were conducted in the years 1898 to 1904,
1,600 larvae were used, which hatched from 2,100 eggs placed in the experi-
ment rooms. The temperature conditions during the experiment were as
follows :
Table 60.— Temperature conditions.
Conditions —
8 a. m.
1 p. m.
8 p. m.
Maxi-
mum.
Mini-
mum.
Aver-
age.
Deviation
from
normal.
In nature
In experiment. .
"C.
19
28
0 C.
31
41 5
° C.
21
30
° C.
33
43
° C.
13
20
°C.
23.66
33- 16
° C.
O
+ 95
i8o
COLORATION IN LEPTINOTARSA.
During this experiment there was a mortality of 55 per cent in the larvae
and of 30 per cent in the pupa:, 15 per cent emerging as imagines. These
were 35 per cent below the normal size and equally divided between the sexes.
The hypodermal color was pale yellow and the dark markings were much
altered. Upon the epicranium and pronotum the spots were small, and there
were no fusions. The epicranial spots h and ft' were wanting, as were also the
pronotal spots / and /' and often d and d', while b and V were much reduced
and a and a' were nearly divided into two spots each. In the larvae the
basal-pleural spots were lost in part of the larvae in the second instar, and the
tergal spots all disappeared. The cuticula pigments were also lighter, being
a dark brown. The general color tendency was albinic. as is shown in the
following seriations :
Table 61. — General color of beetles used.
Class
4
5
6
7
8
9
10
II
12
13
P.ct.
I
I
Parents
Control
Experiment. .
P.ct.
1
2
P.ct.
I
I
12
P.ct.
2
7
42
P.ct.
5
8
27
P.ct.
20
17
II
P.ct.
42
39
3
P.ct.
18
19
I
P.ct.
9
6
5
P.ct.
2
I
Empirical mode of parents 9
Empirical mode of control 9
Empirical mode in experiment ... 6
Modal deviation of parents o
Modal deviation of control o
Modal deviation in experiment . — 3
Experiment 10. — To determine the effect of a large average increase in temperature
applied throughout life upon the color and color pattern of L. decemlineata.
Conditions. — Temperature on the average 14.50 C. above that in nature,
with other factors normal.
Apparatus. — The same as in experiments 7 and 8.
In this experiment 2,000 larvae hatched from 3,850 eggs were used. These
eggs were taken at random from potato plants soon after they were laid, and
were placed at once in the conditions of the experiment. The temperature
records are as follows :
Table 62. — Temperature conditions.
Conditions-
S a. m.
1 p. ra.
8 p. ra.
Maxi-
mum.
Mini-
mum.
Aver-
age.
Deviation
from
normal.
Ill nature
In experiment.
°C.
19
33
° C.
31
46.48
°C.
21
35
°C.
33
48
°C.
13
2S
°C.
23.66
38.16
°c.
0
H4.5
MODIFICATION OF COLOR.
181
In these experiments 62 per cent died in the larval stage, 34 per cent in the
pupal, and 6 per cent completed their transformations. In the larvae at hatch-
ing- the dorsal color marks were small, and many were absent; and in the sec-
ond instar all spots had disappeared, excepting the spiracular and wing, and
the pronotum and epicranium were light brown in color. The imagines were
39 per cent below the normal in size and were equally divided between the
sexes. The hypodermal color was white, and there were no dark spots upon
the epicranium, nor any upon the pronotum excepting a and c. The elytral
stripes were shortened, and the anal, cubital, and costal were wanting. On
the ventral surface all spots were obliterated, excepting the anterior sternals
upon the thoracic and abdominal segments, and those upon the tarsi of the
legs. The general appearance was strongly albinic, as is shown by the
following seriation :
Table 63. — General color of beetles used.
Class
i
2
3
4
5
6
7
8
9
10
11
P.ct.
9
1
12
13
P.ct.
I
6
Parents
Control
Experiment* . .
P.ct.
I
P.ct.
7
P.ct.
46
P.ct.
I
33
P.ct.
I
I
22
P.ct.
2
7
8
P.ct.
5
8
2
P.ct.
20
17
I
P.ct.
42
39
P.ct.
18
19
P.ct.
2
I
* Number of individuals.
Empirical mode of parents 9
Empirical mode of control 9
Empirical mode in experiment ... 3
Modal deviation of parents o
Modal deviation of control o
Modal deviation in experiment. . — 6
Experiment ii. — To determine the effect of a small average decrease in temperature
upon the color and color pattern of L. dccemlincata when applied throughout life.
Conditions. — Temperature on the average 6.50 C. below that in nature,
with other conditions normal.
Apparatus. — The same as in experiment 8.
In this experiment 3,300 freshly laid eggs taken at random were placed in
the conditions of the experiment as soon as they were obtained, and from
these 2,700 larvae were hatched. The experiments were conducted during
the years 1898 to 1904. The temperature records are as follows :
Table 64. — Temperature conditions.
Conditions—
8 a. in. i 1 p. 111.
8 p. m.
Maxi-
mum.
Mini-
mum.
Aver-
age.
Deviation
from
normal.
In nature . .
In experiment .
0 C. ° C.
19 31
16 19.48
°C.
21
16
°C.
33
20.5
° C.
13
12
° C.
23.66
17.16
°C.
0
-6.5
182
COLORATION IN LEPTINOTARSA.
In these experiments no color modifications were noticed in the larvae, but
the length of the larval life was prolonged 10 to 12 per cent. During the ex-
periment 21 per cent died in the larval stage, 15 per cent in the pupal, and 54
per cent emerged, of which 47 per cent were females and 53 per cent males.
The color modifications were almost exactly the same as those described for
experiment 5. The following seriations have been made to illustrate the gen-
eral color tendencies :
Table 6s. — General color of beetles used.
Class
4
5
6
7
8
9
10
Pet.
11
P.ct.
12
P.ct.
13
P.ct.
P.cl.
15
P.ct.
16
P.ct.
P.ct.
P.cl.
P.ct.
P.ct.
P.ct.
Pet.
Parents
I
2
S
20
42
18
9
2
I
Control
1
I
7
S
17
39
IQ
6
I
I
1
4
8
IO
23
42
9
3
Empirical mode of parents 9
Empirical mode of control 9
Empirical mode in experiment. . . 14
Modal deviation of parents o
Modal deviation of control o
Modal deviation in experiment. . +5
Experiment 12. — To determine the effect of a considerable average reduction in temper-
ature upon the color and color pattern of L. decemlineata when applied throughout life.
Conditions. — Temperature on the average 120 C. below that in nature, with
other conditions normal.
Apparatus. — The same as in experiment 8.
In these experiments, which were conducted during the years 1900 to 1904,
2,900 eggs were placed in the condition of the experiment, and from these
1,500 larvae were hatched. The temperature records are as follows:
Table 66. — Temperature conditions.
Conditions —
8 a. m.
1 p. m.
8 p. m.
Maxi-
mum.
Mini-
mum.
Aver-
age.
Deviation
from
normal.
In experiment. .
°C.
19
10
°C.
31
14.98
°C.
21
IO
°C.
33
15-5
°C.
13
8
°C.
2^.66
ii.66
°C.
0
— 12
In these experiments the hypodermal color of the larvae was duller than
normal, all the tergal and posterior basal pleural spots were dropped out, and
the legs, head, and pronotum were dark brown in color. During the experi-
ment 60 per cent died in the larval stage, 22 per cent in the pupal, and 12 per
cent appeared as imagines. These showed the same color modifications as
those described for experiment 5. Their seriations into color classes are as
shown in table 67.
MODIFICATION OF COLOR.
Table 67. — General color of beetles used.
183
Class
4
5
6
7
8
9
10
11
12
>3
Parents
Control
Experiment*.
P.ct.
I
P.ct.
I
I
P.ct.
2
7
5
P.ct.
5
8
50
P.ct.
20
17
20
P.ct
42
39
20
Pel.
18
19
5
P.ct
9
6
Pet.
2
I
I
I
* Number of iudividuals.
Empirical mode of parents 9
Empirical mode of control 9
Empirical mode in experiment ... 7
Modal deviation of parents
Modal deviation of control
Modal deviation in experiment
Experiment 13. — To determine the effect upon the color and color pattern of L. decem-
lineata of a large decrease in average temperature when applied throughout life.
Conditions. — Temperature on the average 20 ° C. below that in nature, with
Other conditions abnormal, as described in experiment 7.
Apparatus. — The same as in experiment 7.
In these experiments 4,250 freshly laid eggs collected at random were used ;
of these 1,150 hatched, but only 1,000 lived to begin feeding. The larvae
were modified in their coloration, as in experiment 12, and the larval period
was three times its normal length. The following temperatures were used :
Table 68. — Temperature conditions.
Conditions—
8 a. m.
1 p. m.
8 p. m.
Maxi-
mum.
Mini-
mum.
Aver-
age.
Deviation
from
normal.
In nature
In experiment. .
0 C.
19
2
° C.
31
5.98
°C.
21
3
°C.
33
6.5
° C.
13
O
°C.
23.66
3 66
O
— 20
In these experiments 79 per cent died in the larval stage, 1 1 per cent in the
pupal, and 10 per cent completed their transformations. The pupal stage
was prolonged to five times its normal length. The variations found were
exactly the same as those described for experiment 7. The sedations of the
imagines are as follows :
Table 69. — General color of beetles used.
Class. ...
2
3
4
5
6
7
8
9
10
II
12
13
14
Control
Experiment. .
P.ct.
1
P.ct.
10
P.ct.
I
43
P.ct.
1
1
21
P.ct.
2
7
15
P.ct.
5
S
8
P.ct.
20
17
2
P.ct.
42
39
P.ct.
18
19
P.ct.
9
6
P.ct.
2
I
P.ct.
I
I
P.ct.
Empirical mode of parents 9
Empirical mode of control 9
Empirical mode in experiment. ... 4
Modal deviation of parents o
Modal deviation of control o
Modal deviation in experiment . . — 5
1 84
COLORATION IN LEPTTNOTARSA.
In this set of experiments (8 to 13), in which the beetles were subjected to
the conditions of the experiment throughout life, beginning a few hours after
fertilization, results were obtained in imaginal color modification that are
identical with those in experiments 1 to 7, where the beetles were used during
the late larval and pupal stages only. These results I have represented in
text-figure 12. Hence the conclusion to be drawn from these experiments is
that as far as adult coloration is concerned it makes no difference whether the
stimulus of increased or decreased temperature be applied for the entire life
or for the pupal stage only. That is, there does not appear to be any accu-
mulated effect of a temperature stimulus.
tDe/
ations
in
:en
peratu
re,
- from
norma
[, -
used in experiments ■
'
>
8 7 6
J 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 fi5°
B
5
11
8
*
M
V
<*
>
\
\
\
\
,12
6
9
^,
^~
■.^
"
-^
13
\
.<]
\
\
V
V
\
.Experiments with temperatures above the normal
——Experiments with temperatures below the normal
Text-figure 12.— To illustrate graphically the results obtained in experiments S to 13.
It has been shown that the influence of high and low temperature upon the
larval color pattern is much the same as that upon the adult. Therefore it
would seem that the effect upon the color-enzyme forming cells at each period
of activity is immediate and persists only as long as the stimulus is applied,
and that there is no accumulated influence stored up through the larval period.
This has been conclusively shown by a set of experiments made to test this
point. In these experiments larvae reared up to the prepupal period under
the conditions of experiments I to 7 were at pupation placed in normal sur-
roundings. Now, if the effects of stimuli applied to the larvae were cumula-
tive, the fact ought to be apparent in the adults. Two experiments taken
from this set will serve to illustrate the results obtained from this entire series
of seven experiments.
MODIFICATION OP COLOR.
I85
Experiment 14. — To determine the effect upon adult coloration of L. decemlineata of a
considerable increase in average temperature applied throughout the larval stage only.
Conditions. — Temperature on the average 14.5° C. above that in nature,
with other conditions normal.
Apparatus. — The same as in experiment 10.
In this experiment 2,000 larvae were used, hatched from 3,300 eggs which
were taken at random soon after being laid, and placed in the conditions
of the experiment. The larvae were kept there until time for pupation — that
is, until the beginning of the prepupal period — when they were returned to
normal surroundings. The temperature records are as follows:
Table ;
"O. — Temperature conditions.
Conditions-
8 a. m.
1 p. m.
8 p. ill.
Maxi-
mum.
Mini-
mum.
Aver-
age.
Deviation
from
normal.
Ill experiment. .
°C.
19
33
°C.
31
46.48
21
35
°C.
33
48
"C.
13
28
°C.
23.66
38.16
O
+ 14.5
In this experiment, which is an exact duplicate of experiment 10 as far as
the larvae are concerned, the mortality was approximately the same in the
larval stage, but lower in the pupal — that is, 60 per cent died in the larval
stage, 10 per cent in the pupal, and 30 per cent appeared as imagines. The
imagines were normal, as is apparent in the following table :
Table 71. — General color of beetles used.
Class
4
5
6
7
8
9
10
11
12
13
Parents
Control
Experiment. . . .
P.ct.
1
1
P.ct.
I
I
2
P.ct.
2
7
3
P.ct.
5
8
7
P.ct.
20
17
15
P.ct.
42
39
45
P.ct.
18
19
18
P.ct.
9
6
6
P.ct.
2
I
2
P.ct.
I
I
I
Empirical mode of parents 9
Empirical mode of control 9
Empirical mode in experiment ... 9
Modal deviation of parents o
Modal deviation of control ....... o
Modal deviation in experiment. . . o
Experiment 15. — To determine the effect upon adult coloration of L. decemlineata of a con-
siderable average decrease in temperature when applied throughout the larval stage only.
Conditions. — Temperature on the average 12° C. below that in nature,
with conditions as in experiment 10.
Apparatus. — The same as in experiment 10.
In this experiment, from 3,100 eggs were obtained 1,500 larvae, which were
kept in the conditions of the experiment until the beginning of the prepupal
period, and were then returned to a normal environment.
1 86
COLORATION IN LEPTINOTARSA.
The temperature records are as follows :
Table 72. — Temperature conditions.
Conditions —
8 a.m.
r p. ill.
8 p. in.
Maxi-
mum.
Mini-
mum.
Aver-
age.
Deviation
lrom
normal.
In experiment.
°C.
10
°C.
14.98
°C.
21
IO
°C.
33
15-5
°C.
13
8
°C.
23.66
11.66
°C.
0
— 12
In this experiment the larvae showed the same color changes as those in
No. 10. The mortality was 60 per cent in the larval stage, 12 per cent in the
pupal, and 38 per cent emerged as imagines. The seriations of these are
shown in the following table:
Table 73. — General color of beetles used.
Class
5
6
7
8
9
10
11
12
13
Perct.
Per ct.
Per ct.
Perct.
Per ct.
Per ct.
Per ct.
Per ct.
Perct.
Parents
I
2
5
20
42
18
9
2
I
Control
2
7
8
17
39
19
b
I
I
Experiment. .
I
5
7
18
40
20
7
2
Empirical mode of parents 9
Empirical mode of control 9
Empirical mode in experiment. .. . 9
Modal deviation of parents o
Modal deviation of control o
Modal deviation in experiment. . . o
In these experiments the adult coloration was not affected by temperature
stimuli applied throughout the larval stage alone. This fact is made clearer in
text-figure 13, where the curve of the entire series of these experiments is rep-
resented. Not the slightest indication is here apparent of any influence upon
imaginal coloration of the conditions of temperature to which the larvae have
been subjected.
The modifications of coloration produced by external stimuli during the
ontogeny of L. decemlineata are purely somatic variations, there being no
accumulations of the effects of stimuli applied at one stage upon the coloration
of a later one. The influence of temperature upon coloration is felt during
the development of coloration of a particular instar only, and consists in the
acceleration or retardation of the action of the color-producing enzymes.
Neither is there any accumulation of the results of repeated temperature stim-
uli nor any increase of response even when repeated for several generations.
This is apparent in text-figure 14, where are given the results derived from a
series of experiments made in the years 1901 to 1904, in which these beetles
were subjected to the conditions of increased or decreased temperature during
MODIFICATION OF COLOR.
I87
^Deviations in temperature, - from the normal, - used in experiments. ^
!
-
■
10 1
■ 1
1
1
S 16 1
18 19 20 21 22 23 24
■:
c
.'
.
Ill)
1 I
Ujj.
1
' '
1 1 1
1 II
it.
t ■ 1
*
14
^
_
+ t~i
c
!- -
— r
- f
-t-
-;
^
- —
4-
F
4- -
0
0
=FF==F
.8
•
-
1
I
(K
i
4
-
2
Text-figure 13.— To illustrate the results obtained in a series of experiments of which two (14 and 15
are described in the text, the results obtained in the others being shown by the figure.
1901
1902
1903
1904
Gen. I.
Gen. M..
Gen. {TJ.
Gen.JV.
Gen.T.
Gen. VI.
Gen
. VM.
Gen. VIII.
x Deviations in temperature, - from normal,- used in experiment ,
5 10 15 5
5 10 15 5 10 15 5 10 15
Text-figure 14. —To illustrate the results that were obtained in a series of experiments as described
in the text when continued for a number of generations. Kach of the different lines represents a
series of experiments and the arrow heads the direction takeu by different parts of each set when
the stimulus was removed or applied. In all the sets of experiments represented lines going toward
the modal line of the parents indicate the rapid return of the beetles to the modal couditiou when
the stimulus was removed.
1 88
COLORATION IN LEPTINOTARSA.
their color ontogeny in successive lineal generations. Moreover, if the stim-
ulus be removed the coloration at once shows its ephemeral character by going
back to the normal condition, and does not in succeeding generations exhibit
any tendency to repeat the modification. Some of my material, which was
experimented upon along this line for from twelve to fifteen lineal genera-
tions, still showed no trace whatever of permanent modification even with
intense selection.
I conclude, therefore, that deviations in temperature, when applied to L.
decemlineata, act simply as stimuli toward the acceleration or retardation of
pigmentation, and that the responses attain their maxima at once, are never
increased (unless the stimulus be changed), and disappear as soon as the stim-
uli are removed. The results in color modification are purely transient, so-
matic variations, which, however, may produce a strong prophetic or historic
skewness in the variation polygon, and may correspond very closely, as I have
found in many of my experiments, to those variations found to exist in nature
as place and geographical variations.
Nowhere in these experiments does evidence appear which would indicate
that temperature has any specific effect upon the coloration of these beetles.
It in all cases acts simply as a stimulus toward the production of results which
are exactly the same as those which I have brought about by other agencies.
Moisture Experiments.
The effects of moisture upon the production of color modification has been
determined by subjecting L. decemlineata to varying percentages of relative
humidity. Numerous experiments have been tried which I shall combine into
a few large experiments, as in those with temperature.
Experiment 16. — To determine the effect of a moderate increase in relative humidity
upon the color and color pattern of L. decemlineata.
Conditions. — Relative humidity 10 per cent above the average in nature,
with other conditions normal.
Apparatus. — Glass tanks arranged for the absolute control of moisture.
In these experiments, which were conducted during the years 1898 to 1904,
4,500 eggs collected at random were used, from which 4,000 larvae hatched.
These were kept in the following conditions throughout life :
Table 74.— H
tmidity
conditions.
Humidity—
7 a. in.
10 a. in.
r p. m.
3 p. m.
8 p. m.
Maxi-
mum.
Mini-
mum.
Aver-
age.
Deviation
from
normal.
Iu nature
In experiment .
Perct.
*IOO
*IOO
Perct.
65
85
Per ct.
•SO
60
Per ct.
55
75
Per ct.
*IOO
*I0O
Per ct.
IOO
IOO
Per ct.
43
55
Per ct.
74
S4
Per ct.
0
+10
♦Dew.
MOnil'ICATION Ol* COLOR.
189
In these experiments 6.3 per cent died in the larval stage, 8.7 per cent in the
pupal, and 85 per cent emerged. The coloration of the larvae was a deep wine
red, and all dark spots were large and of a polished black. On the tergae the
inner tergal spots were quite uniformly present. The imagines were either
normal in size or slightly large. The hypodermal color was a strong ocher
yellow, tinged frequently with red. The dark markings, which were large
and frequently fused, were of a dense black color. Their general appearance
was melanic. as is shown by their seriations in the following table:
Table 75. — General color of beetles used.
Class
5
6
P.ct.
2
7
7
P.ct.
5
8
1
8
P.ct.
20
17
I
9
10
P.ct.
iS
19
6
11
P.ct.
9
8
18
12
P.ct
2
I
40
13
P.ct.
I
I
19
14
P.et.
8
15
P.ct.
I
16
P.cl.
1
Parents
Experiment . .
P.cl.
I
2
P.ct.
42
37
5
Empirical mode of parents 9
Empirical mode of control 9
Empirical mode in experiment. ... 12
Modal deviation of parents o
Modal d -viatiou of control o
Modal deviation in experiment . . +3
Experiment 17. — To determine the effect of a large increase in relative humidity upon
the coloration of L. decemlincata.
Conditions. — Relative humidity on the average 21 per cent above that in
nature, with other conditions normal.
Apparatus. — The same as in experiment 16.
In these experiments, which were conducted during the years 1898 to 1904,
3,500 eggs were used, from which were hatched 3,100 larvae. The conditions
of experimentation were as follows :
Table 76. — Humidity conditions.
Humidity—
7 a.m.
10 a. 111.
Per ct.
85
1 p. m.
3P.rn.
Sp. m.
Maxi-
mum.
Mini-
mum.
Aver-
age.
Deviation
from
normal.
In nature
In experiment
Per ct.
*IOO
*IOO
Per ct.
50
85
Per ct.
100
90
Per ct.
*IOO
*IOO
Per ct.
IOO
IOO
Per ct.
43
82
Per ct.
74
95
Per ct.
0
+ 21
In these experiments 12.4 per cent died in the larval stage, 17.6 per cent in
the pupal, and 70 per cent emerged. The larval color pattern was changed in
that the hypodermal color became paler, and had a translucent appearance and
the dark spots were not as large nor were the dorsal areas as well developed.
The imaginal color was also modified, the hypodermal color being pale yellow,
with the dark areas smaller than is normal and with few fusions. In size the
beetles were normal. The general color tendency was albinic, as is shown in
table 77.
190
COLORATION IN LEPTINOTARSA.
Table 77- — General color of beetles used.
Class
3
4
P.ct.
5
P.ct.
6
P.ct.
7
8
P.ct.
9
P.ct.
10
11
12
13
P.ct.
P.ct.
Parents
I
2
5
20
42
18
°
2
I
Control
2
7
8
17
37
10
8
1
I
Experiment .
2
3
7
17
38
21
9
2
1
Empirical mode of parents 9 I Modal deviation of parents o
Empirical mode of control 9 ] Modal deviation of control o
Empirical mode in experiment 7 | Modal deviation in experiment . — 2
Experiment 18. — To determine the effect of a saturated atmosphere upon the coloration
of L. deccmlineata.
Conditions. — Relative humidity 26 per cent above that in nature, with other
conditions normal.
Apparatus. — The same as in experiment 16 .
In this experiment, which was conducted during the years 1899 to 1904,
2,000 larvae, hatched from 4,450 eggs, were subjected to the following condi-
tions :
Table 78. — Humidity conditions.
Humidity —
7 a. m.
10 a. m.
1 p. m.
3 p. 111.
8 p. m.
Maxi-
mum.
Mini-
mum.
Aver-
age.
Deviation
from
normal.
In nature
In experiment.
Per ct.
*IOO
*IOO
Per ct.
65
*IOO
Per ct.
50
IOO
Per ct
55
*IOO
Per ct.
*IOO
*IOO
Per ct.
IOO
IOO
Per ct.
43
IOO
Perct.
74
IOO
Per ct.
0
+26
In the experiments of this series 52 per cent died in the larval stage, 38 per
cent in the pupal, and 10 per cent emerged. In the larvae the hypodermal
color was a light, translucent, yellowish red, and the dark areas were small
and brown in color. All dorsal markings were wanting, as were also the pos-
terior four or five basal pleural centers. In the imagines the hypodermal
color was pale yellowish-white and the dark spots were much reduced in size.
On the epicranium h and h' were wanting and g and g' were represented by
mere traces of color. On the pronotum / and /' were absent entirely ; d and
d', b and b' were absent in 40 per cent of the variates, and e and e' in 12 per
cent. The spots a and a' were in 8 per cent of the variates divided into
anterior and posterior parts. On the elytra all stripes were reduced and the
anal was absent. On the ventral surface all spots were represented by mere
traces, and the entire series of middle sternals was absent. The legs, with
the exception of the tarsse, were devoid of color and the general appearance
was decidedly albinic, as is shown in the following seriations :
MODIFICATION OF COLOR.
Table 79. — General color of beetles used.
191
2
3
4
5
6
7
8
9
10
11
12
'3
Parents
Control
Experiment.
P.cl.
2
P.ct.
9
P.ct.
39
P.ct.
I
2
2S
P.ct.
2
7
12
P.ct.
5
8
7
P.CJ.
20
17
2
P.cl.
42
37
1
P.ct.
IS
19
P.ct.
9
S
P.cl.
2
1
P.cl.
I
I
Empirical mode of parents 9
Empirical mode of control 9
Empirical mode in experiment 5
Modal deviation of parents o
Modal deviation of control o
Modal deviation in experiment. . —5
Experiment 19. — To determine the effect of a moderately diminished average relative
humidity upon the color and color pattern of L- decemlineata.
Conditions. — Relative humidity on the average 8 per cent below that in
nature, with other conditions normal.
Apparatus. — The same as in experiment 16.
In this experiment, which was conducted during the years 1898 to 1904,
2,500 larvae, hatched from 3,160 eggs, were used. The conditions of experi-
mentation were as follows :
Table 80. — Humidity conditions.
Humidity—
7 a. m.
10 a.m.
1 p. m.
3 p.m.
8 p. m.
Maxi-
mum.
Mini-
mum.
Aver-
age.
Deviation
from
normal.
In experiment .
Per ct.
IOO
100
Per ct.
65
60
Per ct.
5°
40
Per ct.
55
40
Per ct.
IOO
90
Per ct.
IOO
IOO
Per ct.
43
33
Per ct.
74
66
Per ct.
O
-8
In the experiments of this series 17 per cent died in the larval stage, 18 per
cent in the pupal, and 65 per cent emerged. The larvae were modified
slightly, the hypodermal color becoming dull red and the dorsal abdominal
spots being largely reduced. The imagines were 6.5 per cent below the size
of the parents and of those under control. The hypodermal color was chrome
yellow, with frequent fusions among the darker spots on both surfaces. The
range of variation was large, but the general appearance was melanic, as may
be seen in the following table :
Table 81. — General color of beetles used.
Class
5
6 7
8
9
10
II
12
13
14
15
16
17
18
P.ct.
I
Parents
Control
Experiment. . . .
P.ct.
I
2
P.ct. Put.
2 5
7 8
1 2
P.cl.
20
17
2
P.ct.
42
37
3
P.ct.
18
19
15
P.ct.
9
8
31
P.ct.
2
I
18
P.ct.
I
I
15
P.ct.
6
P.ct.
3
P.cl.
2
P.ct.
I
Empirical mode of parents 9
Empirical mode of control 9
Empirical mode in experiment. . . 11
14— T
Modal deviation of parents o
Modal deviation of control o
Modal deviation in experiment. . +2
192
COLORATION IN I.EKflNOTARSA.
Experiment 20.— To determine the effect of a considerable decrease in relative humid-
ity upon the color and color pattern of L. decemlineata.
Conditions. — Relative humidity on the average 14 per cent below that in
nature, with other conditions normal.
Apparatus. — The same as in experiment 16.
In this series of experiments, conducted during the years 1899 to 1904,
3,860 eggs were used which were taken at random and subjected soon after
being laid to the conditions of the experiment. From these 2,000 larvae were
hatched.
The conditions of the experiment were as follows :
Table 82. — Humidity conditions.
Humidity—
7 a. in.
10 a.m.
1 p. m.
3 p.m.
8 p. m.
Maxi-
mum.
Mini-
mum.
Aver-
age.
Deviation
from
normal.
In experiment.
Per ct.
100
95
Per ct.
65
55
Per ct.
5°
40
Per ct.
55
40
Per ct.
100
70
Per ct.
100
100
Per ct.
43
30
Per ct.
74
60
Per ct.
O
-14
In this series of experiments 38 per cent died in the larval stage, 42 per cent
in the pupal, and 20 per cent emerged. The larvae were all small in size and
light in color, and the basal pleural spots were wanting. The imagines were
25 per cent below the average size of the parent, and were strongly albinic.
The hypodermal color was light yellow. On the epicranium the spots were
entirely absent in 70 per cent of the variates, and in the remainder rudiments
of g and g' only were found. On the pronotum the spots f and f and d and d'
were absent and b and b' and c and e' were reduced to mere traces. On the
elytra all the stripes were reduced posteriorly and the anal and costal were
absent. The ventral surface was almost unicolorous, traces of the outer ster-
nal spots only being found in the thoracic and anterior abdominal region. The
entire lot was strongly albinic, as may be seen in the following seriations :
Table 83. — General color of beetles used.
Class
3
4
5
6
7
8
9
10
11
12
13
P.ct.
P.ct.
P.ct.
P.ct.
P.ct.
P.ct.
P.ct.
P.ct.
P.ct.
P.ct.
P.ct.
Parents
I
2
5
20
42
18
9
2
I
Control
2
7
S
17
37
19
S
I
I
Experiment. .
2
12
40
23
10
7
3
2
1
Empirical mode of parents. 9
Empirical mode of control 9
Empirical mode in experiment. ... 5
Modal deviation of parents
Modal deviation of control
Modal deviation in experiment.
o
o
-4
MODIFICATION OK COLOR.
193
Experiment 21.— To determine the effect of a large average decrease in the relative
humidity upon the coloration of L. dccemlineata.
Conditions. — Relative humidity on the average 24 per cent below that in
nature, with other conditions normal.
Apparatus. — The same as in experiment 16.
In this series of experiments, conducted during the years 1899 to 1904,
there were used 3,000 larvae, hatched from 5,300 eggs collected at random and
subjected to the conditions of experiment soon after being laid. The condi-
tions were as follows :
rABLE 84. — Humidity conditions.
Humidity-
7 a. ra.
10 a. ra.
1 p. m.
3 p.m.
8 p.m.
Maxi-
mum.
Mini-
mum.
Aver-
age.
Deviation
from
normal.
la nature. . . .
In experiment.
Per ct.
IOO
75
Per ct.
65
40
Per ct.
5°
30
Per ct.
55
40
Per ct.
IOO
65
Per ct.
IOO
83
Per ct.
43
25
Per ct.
74
50
Per ct.
0
-24
In this series of experiments 52 per cent died in the larval stage, 40 per
cent in the pupal, and 8 per cent emerged. All larvae and imagines were
developed in the same directions as those in experiment 20, although to a
greater extent. The size of the imagines was 38 per cent below that of the
parents and the general color appearance was strongly albinic, as is shown in
the following table :
Table 85. — General color of beetles used.
Class
1
2
P.ct.
3
4
P.ct.
5
P.ct.
6
P.ct.
7
P.ct.
8
P.rt.
9
P.ct
10
P.ct.
11
Prt
12
Prt
13
Prt
P.ct.
P.ct.
Parents
I
2
5
20
42
t8
9
8
2
1
Control
2
7
8
17
37
IQ
I
1
Experiment. .
3
6
43
30
12
5
1
Empirical mode of parents 9
Empirical mode of control 9
Empirical mode in experiment 3
Modal deviation of parents o
Modal deviation of control o
Modal deviation in experiment . — 6
Experiment 22. — To determine the effect of an extremely low relative humidity upon
the color and color pattern of L. decemlineata.
Conditions.— Relative humidity on the average 40 per cent below that in
nature, with other conditions normal.
Apparatus. — The same as in experiment 16.
In these experiments, conducted during the years 1899 to 1903, 6,500 larvae,
hatched from 14,300 eggs taken at random, were used. The conditions of
experimentation were as follows :
TQ4
COLORATION IN LEPTINOTARSA.
Table 86. — Humidity conditions.
Humidity —
7 a. m.
to a. m.
i p. m.
3 P l«
S p. m
Maxi-
mum.
Mini-
mum.
Aver-
age.
Deviation
from
normal.
In nature ...
In experiment.
Pen
IOO
45
Per ct.
65
35
Per ct.
50
20
Per ct.
55
30
Per ct.
IOO
40
Per ct.
IOO
55
Per ct.
10
Per ct.
74
34
Per ci.
O
-40
In these experiments 93 per cent died in the larval stage, 6.6 per cent in the
pupal, and 0.4 per cent emerged. All stages were highly albinic, and the
imagines extremely so, as is shown in the following table :
Table 87. — General color of beetles used.
Class
1
P.ct.
5
2
P.ct.
17
3
P.ct.
■■
3
4
P.ct.
I
5
P.ct.
I
2
6
P.ct.
2
7
7
P.ct.
5
8
S
P.ct.
20
17
9
P.ct.
42
37
to
P.ct.
18
19
II
P.ct.
9
8
12
P.ct.
2
I
13
P.ct.
I
I
Parents
Control
Experiment*.
♦Individuals = 26.
Empirical mode of parents 9
Empirical mode of control 9
Empirical mode in experiment 2
Modal deviation of parents ....
Modal deviation of control
Modal deviation in experiment. .
In this series of experiments the hypodermal color was a pure, translucent
white without any development of lipochrome pigments ; all the dark mark-
ings were lighter in color and many were entirely wanting
The general results derived from these humidity experiments have been
put into text-figure 15. It is apparent from a study of these records that
the results of experiments with deviations of humidity are almost exactly the
same as those which were obtained from experiments with deviations of tem-
perature. Such deviations from the normal, either toward an increase or a
decrease, produce up to a maximum increased pigmentation and a consequent
melanic tendency, but beyond this the effect is reversed, pigmentation is re-
tarded, and the tendency toward albinism becomes more and more pronounced
as the deviation from the normal becomes greater. These results are well
shown in text-figures n to 15. Neither temperature nor humidity seems to
have any specific influence upon the coloration of this beetle ; but each acts as
a stimulus toward the production of an increase in activity under slight devia-
tions, and toward a decrease under larger deviations ; and the curve of the
response is similar to that obtained from very many physiological experiments.
MODIFICATION 01' COLOR.
195
Experiments with Temperature and Moisture Both Variable.
In nature the deviation of one environmental factor alone is of rare occur-
rence, for when one factor is changed, as, for example, temperature, a corre-
lated modification also occurs in others. Thus, in nature, if the temperature
in a given environment is lowered, the percentage of relative humidity is
increased, evaporation is decreased, the sensible temperature is altered, winds
1
V
De
viations in percen
Bg
.5
of
Re!. Hum.
- from norn
•I,
used in experiments
2 3 4 5 6 7 8 9 10 11 12 13 14 15 18 17 16 19 20 21 22 23 24 25 26 38 39 40 4
S
1 1
IS
1 ■
E
c
.-,
2
E
c
n
1 =
i'
13
12
11
i:
ll6
^
j
/.
X
/
'<*
I
T
1
-J
SB
1
T
■ -
to
1
1
1
'
9
1
■
1
1
s
-■
1
1
_,
1
1
1
o
3
JO
Q.
suoi}Bjaua§ }uajBd jo ssep |epoj/\|
Text-figure iS.— To illustrate the results obtained in the different parts of experiment as
described in the text.
204 COLORATION IN LEPTINOTARSA.
to experiment ; and likewise lots were taken from the control and placed in
experiment, and subsequently returned to control. In this way a complete
check was kept on the experiments. In text-figure 17 are represented the
generations experimented upon and the proceedings followed with each.
Experiment 27 was divided into two parts— the experiment proper (27a)
and the control (27Z?). In 27a the beetles were subjected to the conditions of
experiment during ten lineal generations, with results shown in text-figure
18. A maximum deviation in coloration was produced at once toward a me-
lanic state (class 15), from which there was no deviation either above or
below in the succeeding generation. In the third generation of 27a the prog-
eny were divided into two lots of equal size, one of which was kept in the con-
ditions of experimentation, and the other returned to normal conditions. This
second lot, known as 27a1, after being bred during four generations in nor-
mal surroundings, was further separated into two portions, one of which was
still kept in normal conditions as 27a13, while the other was returned to the
conditions of experimentation as 270lb. When the beetles in 27a1 were
returned to normal surroundings theyat once resumed their natural characters
(seetext-fig. 18), and did not deviate therefrom during the four generations of
27a and the three of 27a13, or seven in all. However, the effect upon 27alb
of being returned to the conditions of experiment was an immediate return to
the maximum melanic tendency before observed. From the sixth generation
in 27a another lot of beetles, 2702, were taken and reared in normal conditions,
with the result that they also immediately reverted to the parental condition,
and the same was true of 27a3 in the ninth generation. In experiment 27ft
there appears a slight oscillating variability ( text-fig. 18), which, however, is of
no consequence. In the second generation 27b was likewise separated into two
lots of equal size, one of which, 27b, was retained as a control, while the other,
27b1, was placed in the conditions of experimentation for four generations,
and later in the seventh generation returned to control with 27b. The effect
upon 27b1 was an immediate production of the maximum melanic condition
(class 15), which was retained throughout the four generations of experi-
mentation, and lost only when 27b1 was returned to control.
In experiment 27 there was no artificial selection, all imagines being allowed
freedom to mate and breed as in nature ; hence the only selective influences
present were those exercised in the mating of the beetles and by the condi-
tions of the experiment, which eliminated a small percentage.
From the data of this experiment the following conclusions are drawn :
(1) A deviation in an environmental complex at once causes the polygon
of somatic variation and the modal class to shift as far from the normal as it
can go under the given condition, and keeps them there until there is a return
to the normal environmental complex, when the somatic variations also at
once return to their normal state.
MODIFICATION OF COLOR. 205
(2) The color variations employed in experiment, which are purely somatic,
are the direct result of a response to changed environmental conditions, in
terms of increased or decreased activity in pigmentation. They may change
as rapidly, as frequently, and in as many different directions as the conditions
producing them change, and they have no influence whatsoever upon the col-
oration of succeeding generations.
In this whole line of experiments, conducted with the purpose of deter-
mining the effects of temperature and moisture upon the coloration of decem-
lineata, many thousands of individuals have been subjected to experiment for
single generations, either throughout ontogeny or during the pupal period,
and others for successive lineal generations. I find in all, even where the most
extreme stimuli have been used and great modifications of coloration have
been produced, that the determinate character of the variations, their order of
appearance, and their degree of stability follow strictly the laws of variation
already stated.
Temperature and moisture, the cardinal factors of environment, do not,
however, as is often stated, have any specific effect upon the coloration of
decemlineata, but each acts simply as a stimulus to accelerate or retard the
physiological processes involved in the production of coloration, and whether
alone or in combination produce identical results. Of the two factors when
in combination moisture is by far the more important and the determining
agent in the modification of color, excepting when the other is more highly
abnormal, a condition not often found in nature.
The modifications produced in experiment are purely somatic and follow
exactly the laws of fluctuating variation. In many experiments, however,
the modifications effected by experiment were exactly like those found in dif-
ferent geographical areas, as, for example, when in beetles from Chicago
there were produced conditions characteristic of those from Arizona. This
fact is at once important and suggestive on account of its bearing upon the
interpretation of the phenomena described in the section on "Place and Geo-
graphical Variation." Finally, in experiment 27, it is clearly demonstrated
that the somatic variations in color are not inherited, but that they are fluc-
tuating, transient, and due solely to environmental stimuli which accelerate or
retard color development. It is from experiments like No. 27 and those pre-
ceding it that the proper explanation is found of the place and geographical
variations of this plastic species.
Experiments with Different Foods.
The influence of specific foods upon animal coloration has been investigated
to some extent in birds and insects. Poulton has shown in Lepidoptera and I
in Coleoptera the part that derived plant pigments play in the coloration of
larvae, where they are usually found as subhypodermal colors, which are not,
206 COLORATION IN LKPTINOTARSA.
as far as is known, inherited by succeeding generations, but may serve, how-
ever, as valuable protection during the life of the individual.
With decemlineata, experiments have been tried with the following food
plants: Solanum tuberosum, rostratum, eleagnifolium, carolinense, dulca-
mara, nigrum, and mcioitgcua, Lycospersium esculentum, and Physalcs, three
species. Larvae fed upon these different plants showed some modifications
of the subhypodermal color, but none of importance. Thus, Solatium nigrum
produced a redder color in the larv.-e and 5". carolinense a light dirty wine-
red, but I was not able to obtain the results attributed by Riley to the influence
(if this food, and I conclude that they were almost certainly due to tempera-
ture and moisture, and not to food. From S. rostratum a yellowish color was
obtained in the larvae, but the woody Solanums produced no changes.
With etiolated food white and transparent larvae were obtained exactly as
in Lepidoptera, but such modifications are manifestly of no importance in spe-
cies evolution. Also, when the food supply is scanty or the nutrition of the
larvae is interfered with, changes in coloration result, but these are due to mal-
nutrition and starvation, and not to the direct influence of any specific food.
Experiments with Soils.
The direct influence of soils upon animals is exerted in three general ways :
First, by supplying the inorganic salts needed in the formation of protecting
or supporting tissues ; second, by a continued production of proper conditions
of moisture for hibernation or development in the ground ; and third, by the
control and modification of temperature.
The first of these influences may be discarded, as in Leptinotarsa no salts
are used in the production of protective or supporting tissues. The second
and third, however, are of great importance, not because of the chemical com-
position of the soil, but because of its control of humidity and temperature,
and, through them, of coloration ; for all species of this genus pupate in the
ground, and are, therefore, subjected to the alterations in temperature and
humidity which soils induce.
The modifying effects of soil are especially important in ecological studies,
and these will be considered in a subsequent paper. These influences are ex-
erted largely through the contained water — that is, a dry soil becomes intensely
hot by day and cools to a low degree at night by radiation, while a soil con-
taining much water warms slowly by day and cools slowly at night, owing to
the great amount of heat which water is capable of absorbing, and the reluc-
tance with which it is given up. The water content of soil is controlled, not
by abundant rainfall, nor by telluric water, but almost wholly by adhesion and
capillarity in the soil — that is, physical conditions alone, such as permeability,
capillarity, and the power to absorb and to retain water are the factors which
influence the moisture content of soil, and hence the coloration of these
MODIFICATION OF COLOR.
207
beetles. In all soils the pores which do not contain water are filled with air in
which the percentage of relative humidity is controlled by the amount of water
in neighboring pores. Likewise the cells in which these beetles pupate are
filled with air, the relative humidity of which is controlled by the water in the
pores of the surrounding earth.
With Lcptinotarsa I have used the following soils in experiment: (1) Gla-
cial clay, (2) glacial till, (3) fine-grained soil, rich in humus, (4) fine sandy
soil, poor in humus, and (5) coarse sandy soil. The absolute water capacity
of these (water remaining after the excess had been drained off) and the rel-
ative humidity of the cavities therein were as follows :
Table 95. — Soils used.
Soil.
Absolute
water
capacity.
Relative
humidity
in small
cavities.
Air
capacity.
Source of soil.
Glacial clay
Per cent.
40.3
40. s
32.6
150
10.3
Per cent.
92
85
74
43
33
Per cent.
5°-7
59-2
64.4
85.0
89.7
Brick clays, Bridgewater, Mas-
sachusetts.
Moraines of eastern Massa-
chusetts.
Market-garden soil in Chicago.
Edge of sand dunes near Chi-
cago.
Sand dunes near Chicago.
Glacial till
Fine soil rich in humus.
Fine soil poor in humus.
Larvae of L. dcceinlineata were allowed to pupate in these soils, which were
frequently flooded and drained, so that the absolute water content and relative
humidity were kept constant. The resulting modifications expressed in the
sedation of their variations in the imagines are as follows :
Table 96 —
■Coloration
of b
eetles, with
various
soils used.
Class ...
3
4
5
6
7
8
9
10
11
12
13
14
15
16
-a
v 2
«3
Pel.
P.ct.
Pel.
P.ct.
P.ct.
P.ct.
Pet
P.ct
P.ct.
P.ct.
P.ct.
P.ct.
P.ct.
P.ct.
P.ct.
I
2
S
20
42
18
9
2
I
72
I
3
7
II
21
47
8
I
I
92
Glacial till
I
7
8
18
39
iq
7
I
8S
Fine soil rich in humus
I
,S
IO
2,S
39
17
2
1
74
Fine soil poor in humus
I
4
45
23
22
3
2
43
2
9
44
20
14
9
I
I
33
In this table the strong influence of the soil in modifying moisture condi-
tions and producing color changes is apparent. The polygons of variation
shift either toward albinism or melanism as the moisture content of the dif-
15— T
208 COLORATION IN LEPTINOTARSA.
ferent soils is high or low, and the deviations are the same as those found in
the experiments with moisture.
I conclude, therefore, that clay soils (glacial clays, glacial tills, etc.) are
productive of melanic tendencies in color variation ; that from fine-grained
soils rich in humus we have average or slightly melanic conditions ; while
sandy soils poor in humus and coarse sands produce albinic tendencies.
These results correspond well with conditions found in nature. Over the
northeastern part of the United States and Lower Canada, in the region of
glacial soils which contain a high percentage of clay, the general tendency of
coloration in decemlineata is melanic. In the prairie country and in the re-
gion of residual soils, in the middle latitudes of the United States, average or
slightly melanic conditions are found, while in the west and southwest, in the
sandy soils poor in humus, the tendency is albinic. Soil acting upon color-
ation through the contained moisture may, therefore, be an important factor
in geographical and place variation, as also in distribution and ecology.
Experiments with Light.
It is frequently stated that light has an important influence upon coloration,
and there is frequently assumed to exist a sort of bio-photography of the
environment upon the animal ; but with the exception of the results of the
experiments of Poulton and his pupils upon Lepidoptera there is not to my
knowledge any evidence of value on this subject.
For several years (1897 to 1904) I have experimented with decemlineata
to determine the action of light upon coloration. All separate colors of light
or different intensities thereof — parallel rays, cross-rays, and absolute dark-
ness— have been used, but with none have I been able to obtain any variations
in color in decemlineata. I therefore conclude that in this beetle at least
light — that is, its action in different wave-lengths and intensities — certainly
has no effect upon color-pattern variations. The beetle responds to light,
but in other ways, and not by changes in coloration.
Experiments with Atmospheric Pressure.
Experiments with altered atmospheric pressure, wherein alpine conditions
were simulated, have been tried, but not with any success. While it is easy
to produce rarefication of air and the conditions of temperature and moist-
ure found in alpine regions, it is not possible to produce the same kind of
atmosphere, nor the intense sunlight. These experiments were continued for
three years, but although as far as atmospheric pressure was concerned it
was possible to obtain conditions equal to an altitude of 22,000 to 23,000 feet,
the beetles showed no resulting modifications of coloration. When, however,
opportunity came to transport them to an altitude of 9,000 feet, a marked
change was noticeable, which indicates that something was lacking in the
environmental elements in the experiment made in the laboratory.
MODIFICATION OF COLOR. 209
Experiments with increased atmospheric pressure have also been made, but
thus far without any decisive results. I conclude, therefore, that as far as
variation in atmospheric pressure is concerned, it is a matter of no moment in
the color variations of dccemlineata. However, altitude in nature is a factor
of importance, but whether the modifications produced are the effects of
atmospheric pressure or of other factors that are active in alpine regions has
not yet been determined.
From these experiments upon this beetle, extending over a period of ten
years, I have reached the following general conclusions :
(i) Somatic color variations are due to the action of environmental stim-
uli which accelerate or retard the physiological processes involved in color-
pattern formation, and thus cause the color pattern and its elements to vary
in definite directions toward an increase or decrease of pigmentation.
(2) The environmental factors do not have any specific effect, but each acts
as a stimulus, which, when applied to the animals, produces modifications in
coloration by accelerating or retarding the physiological processes which are
productive of coloration ; hence many different stimuli may produce identical
results.
(3) Temperature and moisture are the two prime factors in the production
of color changes, and of these moisture is the more important.
(4) Soil acts as an important factor indirectly through moisture.
MODIFICATION OF COLOR IN L. SIGNATICOLLIS.
It has already been shown that signaticollis is a stable species, of low varia-
bility and of very conservative constitution ; and I have therefore selected it
for experimental work as the type of one extreme in certain species of the
genus, while dccemlineata represents the other extreme. My experiments
with signaticollis do not extend over as long a period as those with decemline-
ata. but the results are quite as conclusive.
Temperature Experiments.
Experiment 28. — To determine the effects of a slight average increase in temperature
upon the coloration of L. signaticollis.
Conditions. — Temperature on the average 6° C. above that in control, with
other conditions normal.
Apparatus. — Glass tank, as in experiment 23.
In this experiment 800 larvje from three generations were used. The
material was brought from Cuernavaca, Morelos, Mexico, and subjected to
experimentation in the greenhouse at the University of Chicago. The tem-
perature conditions were as shown in table 97.
2IO
COLORATION IN LEPTINOTARSA.
Table 97. — Temperature conditions.
Conditions—
7 a. m.
10 a. m.
1 p. m.
3 P-ni.
S p. m.
Maxi-
mum.
Mini-
mum.
Aver-
age.
Deviation
from
normal.
In control
In experiment .
20
23
°C.
24
28
°C.
32
40
°C.
25
32
°C.
21
29
°C.
35
43
0 C.
18
20
°C.
24.4
3°-4
"C.
+ 3
+ 6
In the experiment 9 per cent died in the larval stage, 21 per cent in the
pupal, and 70 per cent emerged. In both larvae and imagines the coloration
was unmodified, as is shown for the adults in the following table :
Table 98. — Coloration of adults.
Class
10
11
12
13
14
15
16
P.ct.
P.ct.
P.ct.
P.ct.
P.ct.
P.ct.
P.ct.
Parents
3
4
7
16
50
20
Control
1
S
8
18
46
22
Experiment..
1
2
5
7
16
45
24
Empirical mode of parents 15
Empirical mode of control. 15
Empirical mode in experiment .. . 15
Modal deviation of parents o
Modal deviation of control o
Modal deviation in experiment. . . o
Experiment 29. — To determine the effects of a considerable average increase in tem-
perature upon the coloration of L. signaticollis.
Conditions. — Temperature on the average 12° C. above that in the control,
with other conditions normal.
Apparatus. — The same as in experiment 23.
In this experiment the material was the same as that used in experiment 28.
Nine hundred larvae from three generations were subjected to experimenta-
tion under the following conditions :
Table 99. — Temperature conditions.
Conditions—
7 a. m.
10 a. m
1 p. m.
3 p.m.
S p. m.
Maxi-
mum.
Mini-
mum.
Aver-
age.
Deviation
from
normal.
In control
In experiment. .
°C.
20
3°
°C.
24
32
° C.
32
44
°C.
25
39
21
37
°C.
35
46
° C.
18
25
°C.
24.4
36-4
°C.
+ 3
+ 12
In this experiment 34 per cent died in the larval stage, 56 per cent in the
pupal, and 10 per cent emerged. In both larvae and imagines a small increased
variability was found, but there was no modal change, as may be seen in the
table 100.
MODIFICATION OF COLOR.
Table ioo. — General color of beetles used.
211
Class
6
II
12
13
14
15
16
P.ct.
20
22
20
17
P.ct.
2
Parents
Control
Experiment. .
P.ct.
I
P.ct.
I
P.ct.
3
i
2
P.ct.
4
5
4
P.ct.
7
8
9
P.c(.
16
18
14
P.ct.
50
46
47
In experiments where an average increase in temperature of more than
12° C. was used the mortality was too high for the observations to be of any
value. Experiments conducted with lowered temperatures gave results like
those in increased temperatures. In fact, in all the experiments with signati-
collis the same result was obtained — that there was no modal shifting of the
species as the result of increased or decreased temperature. This is in close
accord with my observations of this species in nature, where it is apparently
extremely stable and conservative, and varies as little as in experiment.
Moisture Experiments.
L. signaticollis was also experimented upon with deviations in relative hu-
midity similar to those used with decemlineata, but without any resulting
color modification.
In signaticollis is found a species which differs greatly from decemlineata in
that it does not respond to stimuli by altered coloration. Pure somatic varia-
tion in signaticollis is therefore very limited and of a low degree, a fact that
has already been arrived at in the study of variations. We are not to conclude
from this that signaticollis does not vary, because it gives rise to most inter-
esting modifications. These, however, are germinal, and not somatic. But,
as far as purely somatic modifications of color are concerned, signaticollis is
highly conservative, and a striking contrast to decemlineata.
MODIFICATIONS OF COLOR IN OTHER SPF.CIES OF LEPTINOTARSA.
Experiments with the environmental factors used with L. decemlineata and
signaticollis have to some extent been tried upon undecimlineata, dilecta, vio-
lescens, oblongata, juncta, rubicunda, and multitceniata. From these experi-
ments, the data of which it seems unnecessary to give, I find that species like
multitcrniata and undecimlineata, which are variable in nature, are likewise
variable in experiment ; while those which are stable in nature are also stable
in experiment. However, this rule applies only to the variations produced in
the organism during ontogeny — that is, after fertilization — by the action of
such external stimuli as those which are used in the production of the varia-
tions which we usually call fluctuating, and which are largely the cause of the
high place and geographical variability in adults.
212 COLORATION IN LEPTINOTARSA.
With these several species the responses in color changes were always defi-
nite and in the same directions as those already described for decemlineata.
Likewise, they were not the effects of specific agents, but of stimuli which
accelerated or retarded the processes of pigmentation. All responses reached
the maximum at once, showed no accumulated effects of repeated stimulation,
disappeared as soon as the stimuli were removed, and did not reappear until
they were applied again.
GENERAL RESULTS DERIVED FROM EXPERIMENTAL MODIFICATION OF COLOR.
In all investigations dealing with variation we must, whenever possible,
distinguish clearly between variations which arise in the germ plasm and
those which arise in the soma after fertilization, or during ontogeny. It is
therefore important in experimental work that the conditions of experiment
be applied at such a time that either the germ cells or the soma alone may be
affected, this being the only method by which we shall be able to arrive at a
solution of the vexed question of the part taken by somatic variations in
evolution.
In my experiments with Leptinotarsa I have been fortunate in that this
much-desired result has been easily accomplished on account of the fact that
the germ cells do not begin growth until after the time in ontogeny when the
color patterns can be influenced by external stimuli. Although the germ
cells are present during all the time of experiment, they are in the division
periods, in which condition, as far as I can discover, they do not seem to be
influenced by stimuli. It is during the growth and maturation periods that
they are most sensitive, and these in Leptinotarsa do not begin until after
coloration is completed. It was possible, therefore, during ontogeny, and
especially during the development of the adult coloration, to apply strong
stimuli to which the germ cells were not at that time sensitive, and obtain a
transient somatic variation, and later on to apply the same stimuli during the
period of the growth of the germ cells and obtain modifications which -were
similar, but which were permanent in heredity and in subsequent generations.
Although later on I shall devote an entire chapter to a consideration of this
subject, a statement of the above results seems necessary at this point.
The correctness of my methods of experimentation and the observations
upon which it is based is proven by the fact that, among the thousands of
somatic variations zvhich I have produced in my experiments on color modifi-
cations, not one has ever given the least indication of permanency in succeed-
ing generations, or in crossing; nor can they be preserved by the most intense
selection after the removal of the stimulus zvhich produced them. This possi-
bility in Leptinotarsa of so surely isolating somatic and germinal variations
and their production has enabled me to conduct experiments and obtain
results which would not be possible with the material used by other workers
along this line.
MODIFICATION OF COLOR. 213
Among the long list of workers who have experimented upon Lepidoptera
with deviations of temperature and moisture, Weismann, Standfuss, and
Fischer stand preeminent as the most successful. Although each of these
workers obtained modifications which were produced in experiment and
inherited in succeeding generations, the fact that in the Lepidoptera used the
germ cells were in the more sensitive stages during the period of color for-
mation when the experiments zvere performed, leads one to conclude that the
apparent inheritance of somatic modifications zvas due to the direct result of
stimuli applied to the germ cells, and not to the inheritance of somatic modi-
fications. Superficially somatic and germinal variations are often indis-
tinguishable, but their different nature is shown in heredity. It is a fact, the
importance of which can not be overestimated, that out of the thousands of
experimentally produced color variations (somatic variations) recorded in the
literature of the last fifty years, as also in the experiments described above,
not one case of the unquestionable inheritance of such modifications has been
described. Tltis fact is a most difficult one for the neo-Lamarckian to explain
away, because in it lies the direct experimental refutation of their cardinal
principle.
In a large portion of the published work along the line of the experimental
modification of color the experiments described have been of too short dura-
tion to be productive of any generally reliable results. Many workers have
been contented to produce "aberrations," which have been interpreted as
atavistic repetitions of ancestral stages (Dixey, Merrifield), without any atten-
tion being given to the question of the natural variability and its direction
in the species experimented upon. There has been also a marked tendency to
ascribe specific effects to both temperature and moisture ; whereas, as both
Fischer and I have shown, like results are produced by diverse stimuli. In
this connection it must be kept in mind that the experiments of Weismann,
Edwards, and others with dimorphic species of butterflies are not to be com-
pared with the experiments herein described ; because in their experiments it
is the production of changes in the alternation of dimorphic generations
that is brought about by experiment, and this is a» phenomenon in no way
associated with the experimental modification of coloration in non-dimorphic
species.
From the above-described experiments with Leptinotarsa I have drawn the
following conclusions, which are, I believe, generallv true for all insects. At
least, all published data, on careful analysis, falls directly in line with these
conclusions, although often it is not so interpreted by its authors.
(1) The different factors of the environmental complex do not have any
specific influence upon coloration, but all act alike as stimuli, either alone or
214 COLORATION IN LEPTINOTARSA.
in combinations, to accelerate or retard color development, and thus to modify
coloration in the following ways :
(a) Toward melanic or albinic conditions, which are the most general and
important in coloration.
(b) Toward suppression or accentuation of particular color areas or
groups thereof.
(c) Toward changes in the colors themselves.
(2) The factors most potent in the modification of coloration are temper-
ature and moisture ; soil and altitude act indirectly through moisture and tem-
perature, while the influence of food, light, and other factors is very slight.
(3) Any factor acting as a stimulus produces at once the maximum re-
sponse which the deviation in the factor is capable of producing, and this
maximum response remains constant as long as the stimulus is in force, but
varies as the stimulus varies, and is lost when the stimulus is removed.
(4) Any factor which deviates either above or below the normal has the
effect up to a certain point of producing increased pigmentation, and beyond
that point of retarding it.
(5) Variations produced by the action of environmental factors during
ontogeny always follow the laws of fluctuating variation. New combina-
tions of color characters never appear as the result of stimuli applied during
ontogeny, and the modifications found are all in the line of accentuation or
reduction of the color characters natural to the species.
(6) The variations produced in experiment resemble in their polygons
of distribution and in their modal classes conditions found in nature in places
or in seasons in which the conditions of existence are similar to those of the
experiment ; and a variation found in nature is to be interpreted as the result
of a deviation of some factor of the environment acting as a stimulus to pro-
duce a modification of coloration.
(7) The variations produced by somatic stimuli are never inherited, no
matter how long the stimuli be applied. They are therefore of no importance
in evolution. They are of importance, however, in a consideration of the
phenomena of place and geographical variation.
(8) Species of high variability in nature are also highly variable in experi-
ment, and conversely, those which are constant in nature are the same in ex-
periment ; hence the observed variability of a species is a good index of the
presence or absence of somatic plasticity, but is not necessarily an indication
of its ability to produce germinal variations and become a factor in evolution.
To these conclusions the objection might be raised that in nature the fac-
tors of environment act as stimuli not only after the eggs are laid — that is,
during ontogeny — but also before; that they would have the same influence
upon the germ plasm as upon the soma, and should, therefore, produce in the
germ plasm variations of permanency. In all animals, however, the soma
acts as a protecting or insulating layer which, during the growth and fertiliza-
ADAPTATION IN COLORATION. 21 5
tion of the germ cells, shields them from external stimuli by its capacity to
absorb or compensate therefor. Of more importance, however, is the fact
that animals, or Leptinotarsa, at any rate, reproduce or mature the germ
cells only under favorable conditions, and then within narrow limits. Of
still greater moment is the fact that the germ cells are not equally sensi-
tive to stimuli in all their stages of development. Thus, in Leptinotarsa,
in both tropical and temperate latitudes, the germ cells do not develop nor
reproduction take place until the conditions of temperature and moisture are
favorable. For example, I have found specimens of signaticollis which
were active as early as April 15, whose germ cells did not begin to grow or
that did not begin to reproduce until about June 1, when the arrival of the
rainy season insured the necessary amount of moisture. In the dry plateau
of northern Mexico and the southwestern part of the United States Leptino-
tarsa may emerge from aestivation weeks before the rains come, but they
do not breed until the rains begin. Likewise, in the northern United States
and Canada, decemlincata may emerge from the ground in April, but the
germ cells do not begin to grow until the coming of the warm, moist days in
May, or possibly June. In experiment it has been my most difficult problem
to induce these beetles to breed under changed conditions. Germ-cell growth
and reproduction for a given species takes place, therefore, only under condi-
tions within the narrow limits to which it is adapted, but after the eggs are
fertilized and laid, hot, cold, wet, or dry spells act as stimuli upon the soma
and produce results like those which we have already described under "Place
and geographical variation."
Permanent, heritable color modifications of Leptinotarsa have been found
in nature, and are indistinguishable from somatic variations excepting in their
capacity for being transmitted to subsequent generations. They, however,
have no relation to the variations herein described.
ADAPTATION IN COLORATION.
WARNING COLORATION.
Almost all the species of Leptinotarsa are conspicuously colored white and
black, yellow and black, red and black, metallic green, or blue ; and all, both
larva; and adults, live freely exposed upon their food plants. The combina-
tions of yellow, white, or red with blacks are regarded as warning colors,
especially by the supporters of the theory of natural selection, and serve to give
notice of the supposed inedibility of the possessor of the colors. This theory
of warning colors has been developed mostly by the English naturalists, Dar-
win, Wallace, and Bates, and by Miiller, Hasse, and others from observations,
and by Poulton from experimentation, all of whom consider warning color-
ation to be due to natural selection. I have had an opportunity to examine
this phenomenon quite extensively in Leptinotarsa, and to obtain data from
nature and in experiment that goes to support the idea of utility in the above
color combinations.
2l6 COLORATION IN LEPTINOTARSA.
The adults of almost all Lcptinotarsa have strongly contrasting colors, and
are visible from a distance of many feet, often yards. For example, L. un-
decimlineata, with its white hypodermal color and deep, glassy greenish-black
markings, is an object which may be easily seen from a distance of from 30 to
40 feet, and often, when viewed from an elevation, for greater distances. It
stands out clearly and sharply against the green of its food plant, where it
lives boldly exposed upon the upper surface of the topmost leaves in a situa-
tion where it is at once recognizable and is most easily seen and attacked. The
photograph (plate 21, fig. 1) shows the sharp contrast which this species
makes with its food plant, as also the total absence of even the most super-
ficial resemblance to any object in its environment which might afford it pro-
tection from its enemies.
I have frequently taken a position in which I could observe a colony of un-
decimlincata in nature, and I have never seen one of them attacked at any
time of day by any of the insectivorous vertebrates. I have repeatedly seen
birds — tanagers, warblers, fly-catchers, etc. — visit the same branch upon
which the beetles were resting, and take other insects (Viabrotica, etc.), but
pass undecimlincata by. The same observations were made with reference
to several genera of Lacertilia. Additional evidence that they are not eaten
has been obtained by an examination of the stomach contents of insectivorous
birds, both diurnal and nocturnal, but no trace of these beetles was found
therein. Yet I have found, although rarely, mutilated specimens with the
elytra broken and the body crushed and torn, as if it had been picked up by a
bird or lizard and partly chewed, and then thrown away. These mutilations,
however, might have been due to injury from other causes, such as being
stepped upon by cattle or other large mammals, and can not be considered as
good circumstantial evidence of their being taken as prey by insectivorous
vertebrates and then rejected on account of some objectionable quality.
Various insectivorous lizards and toads which have been confined in cages
with these beetles have rejected them as food. If, however, no other food is
provided and the starvation point is reached undecimlincata is eaten, and
without any apparent harm. This fact that under extreme shortage of food
lizards and toads eat these beetles is of absolutely no significance, because all
animals — even man — will eat anything under the conditions of starvation;
hence feeding experiments in which the supposed objectionable food only is
supplied have no weight zvhatsoever against the idea of warning or protective
coloration. When a variety of food was supplied the toads and lizards passed
undecimlincata by unmolested, although other forms only a few centimeters
away were seized and eaten. I have frequently in nature seen lizards exam-
ine the beetles carefully, and then pass them by to take other food ; and I do
not think that there can be the least doubt that the conspicuous black and
white coloration of undecimlineata secures to it a high degree of immunity
from its enemies.
ADAPTATION IN COLORATION. 2\J
Other species of the genus also possess the same combination of characters.
Thus, lacerata, the largest species in the genus, is a conspicuous object on
account of its black and white color. Like undecimlineata, it lives exposed
upon the upper side of the leaves of the topmost branches of its food plant,
where it is most liable to attack. On the green leaf in the bright tropical sun-
shine its black and white pattern is conspicuous, even at a distance of 50 to 75
feet, and could not fail to be observed by the keen-eyed insectivorous birds
and lizards ; yet, in spite of its coloration and position, I have never seen it
attacked, nor have I ever found mutilated specimens.
The adults of L. rtibiginosa are of a uniform bright red color, and as they
are large the red color it makes upon the green leaves of its food plant may
be seen with ease from a considerable distance. Other species of the genus
are also conspicuously colored — violescens and libatrix, with their brilliant
metallic blue-green colors, and dilccta and its allies, and many others — and
they all rest in exposed positions.
The species multitccniata, oblongata, melanothorax, rubicanda, and decem-
lineata have a conspicuous color pattern of yellow and black or red and black,
and are quite as well protected from their enemies as are those with white and
black. During the years of the dissemination of deccmlineata numerous
experiments were made to discover whether domestic fowls — ducks, geese, and
turkeys — would eat the intruders. The universal observation was that decem-
lineata was eaten sparingly, if at all, and then mostly by young birds. One
ingenuous experimenter in Massachusetts inclosed fowls in a yard without
other food than these beetles, and announced with great satisfaction the fact
that after some days the beetles "were eaten with great relish." It would be
interesting to know whether the same might not have been recorded of the
experimenter himself under similar conditions. Various birds, orioles, gros-
beaks, cuckoos, etc., have been observed by others to attack this beetle, and the
"bobwhite" eats potato bugs freely (vide Yearbook U. S. Dept. Agr., 1903) ;
also, they have been found in the stomachs of a considerable number of species,
but there is little, if any, evidence to indicate whether these birds were young
or whether they were experienced. I have also frequently seen deccmlineata
in nature picked up by orioles, blackbirds, catbirds, robins, grosbeaks, and
other birds, and then dropped without being crushed or eaten.
L. decemlineata has on the elytra and around the edge of the thorax closely
set rows of compound glands which secrete an oily, bad-tasting and bad-
smelling fluid. If one of these beetles be picked up between the thumb and
forefinger gently, and the elytra examined under a hand lens, no secretion will
be found, but if ever so gentle a pressure be exerted upon the beetle, and the
elytron watched under a hand lens, minute drops of the yellow, oily liquid will
be seen to be suddenly extruded from each of the numerous glands. When a
bird picks up one of these beetles and exerts pressure upon it with either beak
or tongue, the result is an immediate extrusion of the repugnatorial fluid, in
2l8 COLORATION IN LEPTINOTARSA.
small amounts if the pressure is slight, and in greater amounts as the pressure
is increased. This liquid is volatile, and to the human senses has an unpleas-
ant odor and a rather unpleasant taste, and I see no reason why it might not
be equally disagreeable to insectivorous animals. At any rate, the glands as a
protective device have a high degree of efficiency, and although some of the
insects are eaten, the great majority are well protected thereby. I have ob-
served that young fowls, when turned loose in a potato field where the beetles
are common, attack them at first eagerly, but learn in three or four days that,
although they are easily found and tempting, they are not choice morsels as
food. Hence, after a few days, they are recognized by their striking yellow
and black surface and are avoided thenceforth, and to all intents and pur-
poses they are immune from that particular generation of fowls.
I have also tried the experiment of mixing decemlineata with edible insects
in about equal quantities and feeding them to fowls, with the result that the
experienced fowls unerringly selected the edible and rejected the distasteful
insects, whereas the inexperienced fowls did not so certainly select the agree-
able from the nauseous. The different elements in the mixture were at first
taken in about the proportion of edible 65 per cent and inedible 35 per cent,
which would indicate that even although the new food had not been pre-
viously encountered and its distastefulness experienced, the mere fact that its
coloration was novel and different from the usual run of food caused it to be
investigated with some care before being consumed in quantities. At any
rate, the fowls soon learned to recognize and avoid this insect on account of
its conspicuous coloration. This fact was proven by an experiment in which
beetles whose yellow areas had been painted with blackened shellac were fed
to experienced fowls. The insects were eagerly seized at first, but as soon
as their distasteful qualities were discerned they were rejected.
It can not be doubted that decemlineata, on account of its striking color-
ation and its accompanying inedible qualities, enjoys a high degree of im-
munity from its vertebrate enemies, and not because there is any instinctive
avoidance of the beetle by fowls and birds, but because each succeeding gen-
eration very soon becomes acquainted with its distasteful qualities and learns
to associate it with the sharply contrasted colors of its color pattern. In this
process of educating its public enemies decemlineata yearly sacrifices a small
percentage of its numbers, probably less than half of 1 per cent, but this sacri-
fice gives to the remaining 99 per cent far greater immunity from enemies
than could otherwise be obtained.
L,. multitceniata and oblongata, which possess this same coloration and the
glands, are also, as far as I can discover, practically immune from the attacks
of insectivorous enemies. In fact, every species in the genus Lcptinotarsa
possesses these glands to a greater or less degree, and almost all are marked
with bright colors in most striking contrast to one another and to their sur-
roundings, which render them as conspicuous as possible. Moreover, as far
PLATE 21.
Photographs showing conspicuous nature of coloration of larvae and imagine: ol I i ; tinotarsii
Fig-, i. L. nndecimlineaiti; the striking black and white coloration of this beetle makes it .1 conspicuous
object upon the leaves of its food plant.
L. signaticotlis; larvae in the last larval stage, living exposed on leaves of the food plant ami
rendered conspicuous by their bright yellow and '
ADAPTATION IN COLORATION. 2IO,
as I have been able to discover, the genus as a whole is almost entirely free
from the attacks of insectivorous animals upon the adults. I can, therefore,
come to no other conclusion than that the conspicuousness of the coloration
of the adult beetles serves to advertise for the mutual benefit of themselves
and their enemies their inedible character.
The larva? of Lcptinotarsa also possess striking colors and display them in
a most conspicuous fashion upon their food plants. In rubiginosa, rubicunda,
and decemlineata the larvae are bright red; in diversa and signaticollis, when
mature, they are yellow and black; in lincolata they are banded black and
white, and in mitltitcritiata and oblongata, when mature, they are a brilliant
yellow. All larvae also possess glands which secrete pungent and ill-tasting
fluids, and extrude them on the least provocation. To vertebrate enemies
these larvae are quite immune, as I have repeatedly proven with decemlineata
by feeding it to fowls, and with multitccniata and oblongata by feeding them
to toads and lizards.
These colors do not, however, afford protection from insect enemies, as I
have abundant evidence that the brilliantly colored larva; of multitccniata, sig-
naticollis, diversa, undccimlincata, decemlineata, and lacerata are attacked by
predacious Hemiptera for the purpose of sucking out the body fluids. How-
ever, the attacks of these enemies are of little moment, and do not seem to
appreciably reduce the numbers of any species of Leptinotarsa. The Hemip-
teron attacks the beetle by making a small puncture and sucking out the
haemolymph, and as it does not come in contact with the glandular secretion
it is not repelled thereby. Although a considerable number of larvae may be
destroyed by these insects, it is the vertebrate enemies that are most effective,
and against which protection is really needed.
In almost all Leptinotarsas it is the mature larvae only that are brilliantly
colored. The earlier stages, which are often devoid of bright colors, are pro-
tected by their resemblance to their food plant or to other objects in their
environment. Thus, in signaticollis and tindecimlincata, the larvae of the first
two instars, covered with trichomes gathered from their food plant, crowd
closely together so that they resemble withered leaves or dust; while in the
last stage they live freely exposed, but are protected by their brilliant col-
oration. These conditions will be considered later on under the head of
"Protective resemblances in coloration."
Although I find in Lcptinotarsa, in both adults and larvae, the phenomenon
of warning coloration present and producing the immunity from vertebrate
enemies which the originator of the theory attributed to it, I must acknowl-
edge that to find and recognize the phenomenon and its utility in the economy
of the species is one thing, but to discover how the various combinations of
strongly contrasted colors came about, or how the final adaptation of warning
colors to environment was accomplished, is another and entirely different
question. Likewise, to explain the observed phenomena and their undoubted
220 COLORATION IN LEPTINOTARSA.
utility by one hypothesis is one thing, and very easy, but to determine the
cause of the phenomena is quite another, and not easy.
The most approved current hypothesis employed in the explanation of the
phenomenon of warning coloration is that of natural selection. All are
familiar with the use made of this hypothesis by Darwin, Wallace, and later
writers in their treatment of the subject of protective coloration, and the
plausible explanations that have been made upon the basis of this hypothesis
of natural selection. While I think that no one can doubt the utility of the
end result — that is, the fully developed warning coloration— it is indisputable
that we are as yet ignorant of the earlier stages in the development of warn-
ing coloration, excepting as we create them in our imagination. No one, to
my knowledge, has as yet recorded from actual observation the beginning of
the development of warning coloration either in nature or in experiment, and
the supposed utility of the initial stages still remain an unproven assumption.
Altogether the explanation of warning coloration by natural selection, while
it may be a true one, is at present based upon assumptions the correctness of
which is open to doubt. Those writers who, in highly involved language and
through circuitous logic, dogmatically assert that these phenomena "can only"
have come about by natural selection, are themselves indulging in the same
dogmatism as that which once so generally characterized a kindred branch of
human knowledge, and which the same scientists have vigorously denounced
and held up to ridicule. In Leptinotarsa we shall for the present recognize
the existence and utility of the phenomenon of warning coloration without
attempting at present to account for its production.
PROTECTIVE RESEMBLANCE.
The widespread distribution of this phenomenon among all animals, and
especially among insects, and the interest which these wonderful adaptations
have aroused, have made them the basis of some of the wildest and most
absurd speculations to be found in modern biological literature.
In Leptinotarsa color adaptations which arise as protective resemblances
are not numerous. The eggs, many of the larvae, and most adults are con-
spicuously colored. Some few larva?, however, do show protective coloration
in the early stages. Thus, in multitcemata, oblongata, rubicunda, and melano-
thorax the young larvae are all light yellow, exactly the color of the long yel-
low spines on their food plants, among which they rest. In these larvae the
only protection comes from the general correspondence in color, and is, I
believe, purely a matter of accident, and not an adaptation. At any rate, the
larvae quite as frequently lie on the green part of the plant as upon the yellow
spines, and it would take more imagination than I possess to make out of this
a case of protective resemblance.
Some larvae, undecimlineata, diverse, signaticollis, dilccta, violescens, liba-
trix, and other species, are colorless, or nearly so, in certain stages. This
ADAPTATION IN COLORATION. 221
semi-transparent condition is usually found in the second instar, when the
larva is full of clear hzemolymph and the fat body has not yet developed.
They are then translucent, and resemble closely the color of their food plant.
This combination of conditions, which gives a considerable degree of conceal-
ment to the larvae, is to be interpreted as accidental, and not as a real adapta-
tion for the purpose of concealment and protection.
The adults of Leptinotarsa, as far as I know, present no protective resem-
blances in color adaptation, but all show the warning coloration already
described. Mimicry also is entirely wanting, and while it would be easy
to match up Leptinotarsa with species in Doryphora,Labidomera,Calligrapha,
Stilodcs, etc., and create cases of mimicry, there would be no foundation
in fact for them. If I had never seen these genera in nature, and did not
know that they and the species having similar color patterns are not found
together — that is, if I knew them from museum specimens alone — I might be
pardoned for creating with them more new cases of mimicry; but in these
beetles in nature there is not the slightest trace of mimetic phenomena, and
the creation of any cases of mimicry is, I believe, absolutely unpardonable
unless the cases have been thoroughly studied in nature. The literature is
already overloaded with purely imaginary examples of this phase of protective
coloration.
The adaptations in coloration in Leptinotarsa are largely warning colora-
tion with a few examples of protective resemblance, and no really good evi-
dence has been found which would tend to show how these came into exist-
ence. It is true that oblongata, rnbicunda, and melanothorax all have warn-
ing colors, and that they arise by sudden transformation from multitceniata;
but multitceniata is itself provided with warning colors, and as these species
arise from a species in which warning colors were already developed, we shall
have to go farther back in their genealogy for the origin of the phenomenon.
These species afford evidence only of the transmissibility of warning colors
from species to species, and not of its first production. The fact that warning
coloration is handed on in evolution from species to species is, I believe, of
considerable importance. In these three species, at least, this character is
handed down from the parent species to the descendant species in its full
intensity.
Inasmuch as this phenomenon is general throughout the genus, and is, as I
have just shown, handed on in evolution from one species to another, I con-
clude that the wide distribution of striking colors which serve as warning
colors has not been developed within the genus Leptinotarsa at all, but that it
is inherited by this genus from its ancestors. Hence it is not to be expected
that evidence of the origin of the phenomenon would be found in Leptino-
tarsa; and in my experiments with these beetles none has appeared.
222
COLORATION IN LEPTINOTARSA.
THE EVOLUTION OF COLORATION.
One of the most conspicuous traits of the genus Leptinotarsa is the devel-
opment of its specific characters along definite lines, and not in all directions.
Upon the elytra of Leptinotarsa the color pattern is composed of stripes,
spots, or bands, or combinations thereof. The stripes are color marks parallel
to and between the veins, the bands cross the veins at right angles, while the
spots usually lie between the veins, or sometimes over them. These elytral
color patterns are represented in table 101, where the number of species
characterized by the different patterns is also shown.
Table ioi. — Different types of color pattern in the elytra of Leptinotarsa, and
the number of species which exhibit them.
Types of color pattern.
Number
of
species.
Types of color pattern.
Number
of
species.
Unicolorous, light
Unicolorous, dark, with
or without metallic
colors
Striped
2
7
16
Striped and spotted
Striped and banded
4
5
i
o
8
Banded and spotted
Spotted
It is apparent in this table that among the 43 species in the genus, 16 have
an elytral color pattern which consists solely of stripes, and 4 have a pattern of
both stripes and spots, in which the stripes are dominant. Therefore, the pre-
vailing element in the coloration is the stripe, 20 species in the genus having
striped elytral patterns and the remainder unicolorous, spotted or banded. From
the table it is also evident that there are four directions of elytral color-pattern
evolution in the genus — first, toward a unicolorous condition (9 species) ; sec-
ond, toward a striped condition (20 species) ; spotted (8 species), and banded
(6 species). If we compare the table with the groups of the genus outlined
in the first paper we find that the flavopustulata group is made up of spotted
or banded-striped species ; the haldemani group of striped and dark unicolor-
ous species ; the lacerata group of spotted, banded, banded-striped, and dark
unicolorous species; the rubiginosa group of light unicolorous species; the
lincata group of striped or light unicolorous species, and the dilccta group of
striped, striped-spotted, or spotted species. The most casual inspection of
plates 22 and 23, where the color patterns of the genus are represented,
will at once reveal the fact that the four types of color pattern are really due
to the combinations possible between the two systems described in the section
on color ontogeny, namely, stripes or bands, the spots resulting simply from
the breaking up of the two other elements.
In the second chapter it is shown that the variation of the color pattern
follows along definite lines, and later, in the discussion of color ontogeny,
that color first develops between the veins, and crosses them at five definite
points; and when it fails to develop it is lost first at certain regions or
EVOLUTION OF COLORATION. 223
interbands. (See L. dilecta, plate 19.) Likewise on the figures on plates
22 and 23 it is apparent that species differentiation has followed along the
same lines. For example, the banded and spotted color pattern represented
on plate 22 shows plainly the combination of the bands with the spots.
The immediate ancestor of these beetles is the genus Zygogramma, and
the next removed is Calligrapha. The color pattern on the elytron of C.
intermedia, which is represented on plate 22, fig. 1, is one from which all of
the elvtral designs in Calligrapha are easily derived by slight changes of the
elements thereof. Thus, a slight change consisting in the fusion of the ele-
ments in the central space, and the further development of the proximal, mid-
dle, and distal band, or of spots representing them in the ramous interspace,
will give the conditions found in Zygogramma clathrata, a highly variable
species, which is represented in two of its "variates" in figs. 2a and 2b on
plate 22. Likewise, by a continued development of the color centers of the
central space and of the transverse bands, the pattern found in another
variable species, Z. ornata, is easily produced. This species also shows two
divergent varieties — figs. 3a and 3b, on plate 22.
From the variety Z. ornata, represented in fig. 36, the elvtral patterns of
L. hogei (fig. 4) and dahlbomi (fig. 5) might easily be derived. Also from
the variety 36, by a continued increase of pigmentation, the forms L. hal-
dcmani (fig. 6) and its allies might easily be produced. Or the series L.
heydeni (fig. 7), modesta (fig. 16), and puncticollis (fig. 15), might have orig-
inated from the variety 30 of Z. ornata. Also the species L. lacerata (fig. 8)
and chalcospila (fig. 14) might be derived from the same condition by a
reduction of the spots and a development of the bands (lacerata) , and then by
the further reduction of the bands the spotted pattern of chalcospila might be
produced. Also from the pattern found in Z. clathrata, variation b, the pat-
tern of the elytron of L. stali (fig. 9) could be produced by the reduction of
the stripe of the central space in the region of the proximal interband ; and
likewise, from the variety 2a of Z. clathrata, the color pattern of L. Havo-
pustulata (fig. 10), belti (fig. 11), dohmi (fig. 12), and evanescens (fig. 13).
On plate 23 are represented color patterns of another type, which, how-
ever, can all be derived in much the same way. If we start with a pattern
such as is represented in fig. 1 of Z. dulcis, which is almost the exact counter-
part of C. intermedia, geographica, etc., we can pass from a Calligrapha type
of pattern through the species of Z. malvcs, stali. aneovittata, and novem-
virgata to the striped Leptinotarsa elvtral pattern, simply by the fusion of
spots in the interspaces to form stripes, a process well represented in figs.
I to 5. Z. novem-virgata is an extremely variable form which shows numer-
ous elementary species, and from this group of elementary species the condi-
tions of color pattern represented on plate 23 could easily have been derived.
It is perfectly evident that as far as the elvtral color patterns are concerned
the species of the genus Leptinotarsa can be arranged in a perfectly consistent
16 — T
224 COLORATION IN LEPTINOTARSA.
and complete orthogenetic series. These elytral color patterns are the most
important of the specific differentials, are perfectly stable characters in each
species, and are as reliable as indicators of relationship and evolution as any
other characters we may examine. Let us ignore for the moment the ques-
tion of the relationship of the species and the order and evidence of their evo-
lution, and concentrate our attention upon the very plain case of orthogenesis
presented in the effort to discover why this orthogenesis exists.
In the construction of plates 23 and 24 the species were not selected to
illustrate orthogenesis, but every species of the genus is represented save
two — L. rubiginosa, a unicolorous bright red form, and L. zetterstcdti, a
brown-banded species. Both these excluded species stand wide apart from the
rest of the genus, and are more nearly related to species of Zy go gramma.
They are in reality the end results of lines of modification in that genus, and
show the same phenomena of orthogenesis. The fact that all the species of
the genus show this phenomenon is of itself adequate proof without further
argument of the formation of species in definite directions. (I use the word
orthogenesis in the sense in which it was first used, and not in that employed
by Eimer.) However, upon whatever basis we are content to explain this
condition, the fact remains that the species as we now recognize them have
elytral color patterns of such a character that their arrangement in an ortho-
genetic system is accomplished with ease. This fact might be explained upon
the basis of Eimer's orthogenesis, by an internal perfecting tendency irresist-
ibly forcing the species in certain directions, by natural selection through an
elimination of all variations excepting those along the lines which were of
utility to the species, by the neo-Lamarckian factors, or by mutation. And
yet, with each or all of these hypotheses, we are far from any real elucidation
of the question as to why the establishment of species along certain lines has
taken place. A possible explanation of the problem is to be found in the
phenomena of development, and it is quite possible that the final interpreta-
tion will rest upon the laws of ontogenetic development and physiology alone.
In the ontogeny of elytral coloration and in the coloration of adults it has
been shown that color develops in stripes, bands, or spots, and that the evolu-
tion of patterns has been orthogenetic, by means of various combinations of
these bands, stripes, or spots. If now we can arrive at some explanation of
these fundamental color elements, we shall be better able to understand the
whole problem. In order to do this we shall first have to consider some of
the phenomena in the development of the wings.
The wings of these beetles and of all insects arise from thickenings upon
the sides of the last two thoracic segments, followed by invagination, and
later by evagination, to form the wing fundament. In Leptinotarsa they
always arise beneath the wing spots, which, as I have shown, are homo-
dynamous with the spiracular spots. The wing fundament consists at first of
a minute area in which the cells are more closelv crowded together than in the
Explanation of Plate 22.
This plate illustrates the orthogenesis in the color patterns of the elytra in the genus
Leptinotarsa.
Fig. 1, Calligrapha intermedia.
Fig. 2, Zygogramma clathrata (two
varieties).
Fig. 3, Zygogramma ornata (two varieties).
Fig. 4, Leptinotarsa hogei.
Fig. s, Leptinotarsa dahlbomi.
Fig. 6, Leptinotarsa haldemani.
Fig. 7, Leptinotarsa heydeni.
Fig. 8, Leptinotarsa lacerata.
Fig. 9, Leptinotarsa stali.
Fig. 10, Leptinotarsa fiavopustulata.
Fig. 11, Leptinotarsa belti.
Fig. 12, Leptinotarsa dohmi.
Fig. 13, Leptinotarsa evanescens.
Fig. 14, Leptinotarsa ehalcospila.
Fig. 15, Leptinotarsa puncticollis.
Fig. 16, Leptinotarsa modesta.
Fig. 17, Leptinotarsa litigiosa.
Fig. 18, Leptinotarsa tlascalana.
Fig. 19, Leptinotarsa Ubatrix.
Fig. 20, Leptinotarsa cyaneseens.
Fig. 21, Leptinotarsa violeseens.
PLATE 22
PS
I 4
>2 ■ # ;
I I!
9$
I IF
Tn illustrate orthogenesis in the color patterns "l the elytra of the genus Leplinotarsa
EVOLUTION OF COLORATION. 225
neighboring hypodermis, and which has a single distinct trachea distributed
to it from the main tracheal trunk. In the early embryo the homodynamous
thickenings occur upon all of the segments, in the position of the spiracular
spots, and from these develop the tracheal invaginations and the pigment-
enzyme cells of the spiracula spots. The entire set of homodynamous spirac-
ula and wing spots receive from the same source in the embryo some cells
which are later specialized for the purpose of producing pigment enzymes,
others for the development of the tracheae, and still others for the general
hypodermis. The existence of these specialized cells has been demonstrated
cytologically in the embryo, and experimentally in the larva.
In the development of the wing the fundament takes out of the tissue be-
neath the wing spot ordinary hypodermal and color-producing cells. As the
wing fundament forms a connection is retained at the point where the trachea
touched the fundament, and it becomes the apex or margin of the wing. The
wing grows largely from the apex, the base remaining relatively fixed. As the
apex grows outward the various areas of the wing are laid down, and into the
spaces formed the permanent tracheae enter later. As the wing grows outward
from its base the outer end (apex or margin) forms behind it the structure of
the adult wing, and from the color mother cells there are left in rows, bands,
or spots the cells which are to produce the color pattern. In this growth of the
wing the base is always ahead in development, and is completed first, the tip
last. The essential fact in wing development, as far as coloration is concerned,
is that the wing is derived from one of the systems of the homodynamous color
areas on the body wall of the insect, and that from this area it receives both
hypodermal and pigment-forming cells, the latter of which give rise to the
cells which form the adult coloration. This arrangement, which is character-
istic of all Coleoptera at least, and I believe of insects in general, is the basis
upon which we must explain color patterns and color evolution. It can be
traced back to the homodynamous centers, and from there to the highly prob-
able but not yet demonstrated mother cells of the body centers of coloration ;
then still farther back to the original blastomere, then to the formative sub-
stance of the ovum, and finally, in our imagination, to the more remote and
less tangible elements of organic structure and activity. From this we may-
gain some little appreciation of how fundamental and how early provided for
in development is the coloration of an animal. Here in the general phyletic
constitution of insects we see a fundamental system provided out of which is
to be shaped a great variety of color patterns by many forces acting in diverse
directions.
In the phyletic development of color patterns of genera and species this
fundamental plan of coloration is variously modified, and this furnishes a
basis for the determination of color pattern homologies in insects. We are
not able to explain why in Lcptinotarsa the mother color cells leave behind
rows of cells to form stripes in some species and bands or spots in others. We
226 COLORATION IN LEPTINOTARSA.
say it is due to the morphological and physiological constitution of the species,
to ids and determinants, or to internal perfecting forces, all of which explana-
tions are attempts to appear wise when we are ignorant.
When the facts as to the source of the color-producing cells of the wings
and their general behavior in ontogeny are joined to those of the variation
and ontogeny of coloration — the production of stripes, bands, or spots, or uni-
colorous conditions — they enable us to gain a clearer comprehension of color-
ation than is otherwise possible and to understand how orthogenesis is pro-
duced in variation and evolution, yet we are still ignorant as to the causes.
Let us examine our material in the light of the above conclusions, and we
shall be better able to explain the conditions found in elytral coloration. In
calceata, dilecta, lineolata, and typographical (plate 23, figs. 13 to 16) we
see species formation along a definite line and a distribution from southern
Mexico to Texas. In this series the evolution, variation, and distribution all
show indisputably the existence of orthogenesis ; for it is highly improbable
that the conditions could be due to chance. Finally, when we take into
consideration the fact that the entire genus shows these conditions of dis-
tribution, variation, and coloration there exists not the slightest doubt of
orthogenesis in the genus Lcptinotarsa.
The existence of a fundamental color pattern which each species or genus
variously modifies in its own fashion has been proven, but why each species
should do thus or so with this general plan of coloration is exactly the same
problem as why species do thus and so with the fundamental plan found in
other organs. This is the problem of the origin of species, and the explana-
tion of the existence of the general plan of each system of organs belongs to
the still more remote problem of the origin of phyla.
Although we can not at present explain the observed conditions otherwise
than through speculation, we have gained evidence which is of value in the
interpretation of the phenomena of distribution, variation, and coloration.
We might examine other characters in the genus, such as the epicranium,
pronotum, legs, etc., but we should consume space and time unnecessarily and
arrive at the same conclusion.
It must be admitted that there exists in Leptinotarsa a race tendency, no
matter how it may be accounted for, which from the general system of color
centers existing in the genus as a phyletic character produces coloration of a
certain type, spots, bands, stripes, or combinations thereof, which type is
variously modified by each species. The study of coloration deals largely
with these racial tendencies, and leads to the following general conclusions :
( 1 ) Centers of color origin. — In ontogeny and in evolution (species forma-
tion) color appears first in centers which upon the body are metamerically
repeated spots and upon the wings are either spots, stripes, or bands. In devel-
opment the anterior or proximal centers develop first because they are first
Explanation of Plate 23.
This plate illustrates orthogenesis in the color patterns of the elytra of the genus
Leptinotarsa.
Fig. 1, Zygogramma dulcis.
Fig. 2, Zygogramma malvce.
Fig. 3, Zygogramma stali.
Fig. 4, Zygogramma anovittata.
Fig. S, Zygogramma novem-virgata.
Fig. 6, Leptinotarsa fiavitarsis.
Fig. 7, Leptinotarsa nitidicollis.
Fig. 8, Leptinotarsa cacica.
Fig. 9, Leptinotarsa novemlineata.
Fig. 10, Leptinotarsa obliterata.
Fig. 11, Leptinotarsa pudica.
Fig. 12, Leptinotarsa distinguenda.
Fig. 13, Leptinotarsa calceata.
Fig. 14, Leptinotarsa dilecta.
Fig. 15, Leptinotarsa lineolata.
Fig. 16, Leptinotarsa typographica.
Fig. 17, Zygogramma novem-virgata.
Fig. 18, Zygogramma novem-virgata.
Fig. 19, Leptinotarsa juncta.
Fig. 20, Leptinotarsa defect a.
Fig. 21, Leptinotarsa dejecta.
Fig. 22, Leptinotarsa undecimlineata.
Fig. 23, Leptinotarsa angustovittata.
Fig. 24, Leptinotarsa signaticollis.
Fig. 25, Leptinotarsa diversa.
Fig. 26, Leptinotarsa decemlineata.
Fig. 27, Leptinotarsa intermedia.
Fig. 28, Leptinotarsa melanothorax.
Fig. 29, Leptinotarsa rubicunda.
Fig. 30, Leptinotarsa multitaniala.
Fig. 31, Leptinotarsa oblongata.
iii;
I
r . i
I
i '
■y
I
fig
rt1
.IS
-
«> >\>
)
» #
n
1 1
* V
! 2
i
■
l ■ in the color patterns of th< elytra ol Lhe genus Lep tar; t.
EVOLUTION OF COLORATION. 227
laid clown in ontogeny, and hence mature first. This phenomenon of priority
in development is not at all of a phylogenetic nature; it allows, I believe, of
no other interpretation than that it is simply a phase of the general law that
the anterior and proximal portions of an animal develop ahead of and faster
than the posterior and distal. In variation the posterior distal and dorsal are
most variable, as they are the last developed, and are hence the longest sub-
jected to the causes of somatic variation when in an undifferentiated state.
(2) Color patterns. — The color patterns of the body are built up of color
areas derived from the phyletic system, of segmental centers of coloration,
which appear in the embryo. Of these centers the species uses those which
its peculiar constitution demands and suppresses all others. Upon the wings
color appears in centers, but each species in the ontogeny of the wing develops
pigment cells in the centers where the constitution of the species demands
them. In general these are developed in stripes, bands, or spots, located at
very definite and invariable positions upon the wing. Stripes are always
found between the veins and bands at right angles to them (plate 24, fig. 2),
and at the following places: Base, apex (margin), middle, above the trans-
verse vein, one half-way between the basal and middle bands, and another
half-way between the middle and marginal. Spots occur most frequently at
the point of intersection of bands and stripes.
(3) Color characters. — In development and evolution the units of color-
ation upon the body are the segmentally placed centers, while upon the wings
they vary in different races, but are in general whole bands or stripes, or por-
tions thereof.
(4) Color-pattern evolution and color variations, both somatic and germi-
nal, coincide, and are orthogenetic upon the body because of the existence of
the system of phyletic color centers, which follow in variation and evolution
the laws governing metamerically repeated structures, and on the wings
because of the phyletic system of stripes and bands which follow their own
laws.
(5) Color-pattern evolution tendencies exist which produce patterns of
definite type due to some force characteristic of the germ plasm. This force
takes certain color centers from the phyletic system in the embryo, and by
various modifications thereof produces diversity in coloration. In Leptino-
tarsa there exists a general generic tendency which controls the entire genus,
a group tendency which governs groups of species, and a race tendency which
is limited to a few species. These tendencies are not to be confounded with
the "internal perfecting principle" of Nageli and others, but they are entirely
a matter of germ-plasm constitution and of heredity, and any change of the
latter will modify the tendency. The difference betwen these two ideas is
this : the "internal perfecting principle" involves the conception of an ideal
for the race toward which it is driven by the directive force of some more
remote agency ; while a "tendency," as I use the term, is an expression for a
228 COLORATIOX IN LEPTINOTARSA.
more or less stable condition which exists in the germ plasm for longer or
shorter periods of time, through whose influence certain characters are taken
from the general phyletic backgound and specialized in various ways to build
up the color pattern of the race or species.
(6) The color phenomena of insects show, first, a general phyletic color
pattern which consists of the metamerically repeated homodynamous centers
of coloration and the wing stripes and bands ; and, second, ordinal, generic,
and racial types of color pattern, which consist of modifications and specializa-
tions within groups, superimposed upon the phyletic type. The phyletic type
of coloration belongs with other characters of the phylum, and its explanation
can be undertaken only when we know better the origin of phyla. The ordi-
nal, generic, and racial types are those with which we are concerned in present
day studies of evolution, and they are open to investigation by experimental
and other methods of research.
In this discussion of coloration data and observations from several lines of
research have been brought together and used in building up the final con-
clusions and conceptions given in the last section of this paper. Coloration
is here shown to be a fundamental character of animals, and not an accident.
Moreover, it is found to be governed by laws as strict as those which control
all morphological and physiological characters of animals in general. It is
therefore a trustworthy character for use in experimental evolution, being as
capable as is any other character of giving accurate data as to the laws of
evolution. Finally, the conclusions reached herein may be used for the pur-
pose of more fully verifying and comprehending the facts and tentative con-
clusions set forth in the discussions of variation and distribution.
Although in this study of coloration important and fundamental conclu-
sions are reached as to the laws governing its evolution, we in nowise derive
any valid basis for speculation as to the causes of evolution beyond that
already brought forward in previous chapters. The experiments in somatic
color modification do remove, to a large extent, if not absolutely, from the
range of possibility as causes of color evolution, the neo-Lamarckian factors —
the inheritance of somatic variations acquired through the action of external
stimuli upon the soma during ontogeny. Hence it is to the germ plasm in
the period before cleavage and differentiation begins that we must turn our
attention in the study of evolution.
If somatic variation is eliminated as a cause of color evolution, the only
other possible explanation of the phenomenon that presents itself is that varia-
tions of permanence and utility in evolution arise in the germ plasm; but
whether these variations are due to selection, mutation, or to the direct influ-
ence of environment, is a matter for observation and experiment, and not for
idle speculation.
Explanation of Plate 24.
This plate illustrates the laws of coloration in the genus Leptinotarsa.
The figures are semi-diagrammatic representations of the condition found in different
species of the genus. The lettering used on this plate is the same as employed in text-
figures 1, 2, 3, 7, 8, 9, and 10.
Figs. 1 to 8, elytra of several species to illustrate the laws of coloration of the elytra.
Figs. 9 to 16, pronota of several species to illustrate the laws of coloration.
Figs. 17 to 22, to illustrate the laws of coloration upon the body segments.
PLATE 24.
tin lit
pra
i bn /"•■■
'•■. :; ■?■
A
,.-•
■•;
:/:•.
#-••••
17 I I
J
f «»
To illustrate the laws of coloration in 1 1 it- genus Leptinotai i
CHAPTER IV.
HABITS AND INSTINCTS IN LEPTINOTARSA.
In the preceding chapters on geographical distribution, variation, and color-
ation in the genus Leptinotarsa morphological characters alone were consid-
ered, and the ways in which they show evidences of evolution. In this chap-
ter we shall have to deal with characters and processes that are physiological,
instinctive, and selective. These we shall discuss under three general heads :
(i) Habits and instincts connected with reproduction; (2) habits and
instincts connected with hibernation and aestivation ; (3) habits and instincts
connected with self-preservation.
HABITS AND INSTINCTS CONNECTED WITH REPRODUCTION.
A considerable portion of the habits and instincts found in the genus Lep-
tinotarsa are centered about the reproduction of these beetles, and contribute
directly or indirectly to the success or failure of this process. Some of these,
such as the mating habits, are concerned directly in the reproduction of the
species, while others, such as preferential or selective mating, while they are
not so vital to the perpetuation of the species, are of paramount importance
in the molding of the character of the subsequent generation. Of signifi-
cance also is the adaptation exhibited by these beetles in the habits and
instincts which center around the process of reproduction ; and inasmuch as
these habits and instincts have at various times shown their sensitiveness to
the processes of evolution by being modified permanently in experiment in
the same way as have morphological characters, an attempt will now be made
to gain as clear an understanding as is possible of these habits and instincts
as they occur naturally in the genus, and of their importance to the species
displaying them ; for by this means may be gained a basis in fact for the dis-
cussion of the modified conditions that have been produced in experiment.
Finally, as much as possible must be learned concerning the adaptation of
these characters to the needs and surroundings of the individual species.
MATING HABITS.
As far as I have been able to discover, the mating habits of these beetles are
much the same in all the species in the genus, so that a description of them in
two or three species will suffice for all.
229
230 HABITS AND INSTINCTS IN LEPTINOTARSA.
In all the species of this genus the beetles on emergence from hibernation
or aestivation, or when freshly hatched from the pupa, are sexually immature
and do not seem to possess any attraction for the opposite sex. In both tem-
perate and tropical regions, when the different species of Lcptinotarsa emerge
in the springtime from the dormant condition in which they have passed the
winter or the dry season, as the case may be, they are emaciated, weighing
less than half as much as they did when they disappeared in the autumn
before. Moreover, the germ cells are not developed, and the scent glands con-
nected with the reproductive organs are inactive and devoid of secretions. At
this time the sexes mingle as so many individuals, without the least trace of
interest in each other. They fly or move about their food plants or in the
observation cages in search of food and water, and when they find them they
set upon them with a most ravenous appetite, so that large quantities are
devoured in a short time. But after feeding for several days, or perhaps
longer, the beetles have become well nourished, the germ-cells have matured,
the reproductive organs are ready for use, and the characteristic habits of
the mating period begin to appear. The changes which take place in this
interval between emergence from hibernation or aestivation and the beginning
of reproduction will be considered in the second division of this paper.
As soon as the beetles have attained sexual maturity they become exceed-
ingly restless and active. They move about in search of the opposite sex, and
if it is not to be found this restlessness results often, especially in the males,
in movements so violent and prolonged that the animals become completely
exhausted, and death not infrequently follows. If during the neutral period
the sexes be separated and allowed to attain maturity in separate observation
cages, they will, after the period of neutrality, show a great and increasing
activity. This is manifested in both sexes by a cessation of feeding and an
incessant, aimless wandering about at a rate of movement so excessive that
complete exhaustion and death may follow in a day or two. When thus sepa-
rated the females are less violent in their activities than are the males, and
both sexes show large individual differences in their manifestations of
uneasiness at the absence of the opposite sex.
The experiment was tried, with rather interesting results, of keeping the
sexes separated entirely throughout the breeding period. When decemlincata
was thus experimented upon some of both sexes died in a few days, but
others, especially among the females, lived for two or even three months when
kept in proper conditions of temperature and moisture and given food of
good quality. All experiments of this kind are summarized in the following
table, in which is shown the duration of life of the two sexes when separated
and not allowed to breed .
MATING HABITS. 23I
Table 102. — Duration of life when the tn-o sexes are separated and not allowed to breed.
Duration
of life.
Males.
Females.
Duration
of life.
Males.
Females.
Days.
Per ct.
Per ct.
Days.
Per ct.
Per ct.
5
18
6
55
I
I
10
33
20
60
I
15
36
24
65
I
20
4
15
70
25
2
10
75
I
30
1
5
80
I
35
1
3
35
3
40
1
5
90
2
45
1
1
95
1
50
1
100
1
From this table it is evident that the lack of an opportunity for reproduc-
tion caused the untimely death of at least half of the beetles experimented
upon, but that there was in the population a certain small proportion, espe-
cially of the females, that were able to survive for a long period of time.
When the beetles are allowed to reproduce normally the death rate is as
follows :
Table 103. — Duration of life when beetles are allowed to reproduce normally.
Duration
of life.
Males.
Females.
Duration
of life.
Males.
Females.
Days.
Per ct.
Per ct.
Days.
Per ct.
Per ct.
5
2
I
45
I
2
lo
8
3
5°
I
2
15
22
6
55
I
20
35
2S
60
I
25
21
20
65
I
3°
6
17
7°
3
35
3
IO
75
1
40
1
4
A comparison of the two tables shows at once that the prevention of nor-
mal reproduction produces a decided increase in early mortality, but that in
both of the series there are individuals, both male and female, that live to a
much greater age than their fellows. This ability of some of the beetles to
live for a considerable time has been put to good use in experiment in the
preservation of variations which for one reason or another were not able to
breed with their own generation, but could be kept alive until reproduction
was effected with the generation following. It is quite possible that the same
process might occur in nature, where in a restricted locality all the progeny
of a given generation were, for some reason, of one sex, and were therefore
unable to breed until beetles from another colony should reach the first one in
the following generation. The normal length of time between generations in
232 HABITS AND INSTINCTS IN LEPTINOTARSA.
this genus is thirty-five days, and yet I have often bred these long-lived adults
with the offspring of their own generation, and on several occasions with
those of the following generation.
With other species of the genus, as, for example, multitamiata, oblongata,
iindecimlineata, dilccta, and signaticollis, the same results have been obtained,
excepting that there is a variation in the length of time that each species will
live.
It is apparent that there is a high degree of difference between the species
of the genus in their ability to resist the abnormal condition of not being able
to reproduce at the proper time, and that this resistance is greatest in those
species which are from habitats where there is a great chance of unfavorable
conditions, as, for example, from the plateau of Mexico, where local condi-
tions may make necessary the prolongation of the reproduction period. In
the species decemlineata, inultitivniata, and oblongata this ability is developed
to a marked degree, and these species live in habitats where the conditions of
existence are much more variable than are those of the species dilecta, unde-
cimlineata, or violescens, which do not show as much resistance along this
line. This ability to hold in abeyance the reproductive capacity for longer or
shorter periods of time is characteristic of the entire genus ; it is quite possible
that it is an adaptation which has been preserved by selection, but I do not
believe it could have been caused thereby. Of course, if one is credulous
enough to accept Weismann's doctrine of germinal selection he might con-
sider that the whole phenomenon has been originated and controlled by select-
ive processes. If it is an adaptation, it is one that is not yet fully formed, in
that some of the individuals of the species do not show the slightest ability to
resist the prevention of their reproductive activities. However, such a char-
acter does offer an opportunity to test the possibility of creating experiment-
ally by selection an adaptation to changed conditions of existence ; because if
in experiment one could create a race in which all the individuals possessed
this capacity, it could not be considered in any other light than as an adapta-
tion developed in the species to enable it to meet unfavorable conditions in its
environment at the time of its normal reproduction.
The importance of a habit like this to the species living upon the Mexican
Plateau, for example, is very great. In the North Mesa it often happens that
when a slight rain occurs in the spring these beetles emerge from aestivation,
and then there is no more rain for weeks. Inasmuch as very circumscribed
conditions are necessary for successful reproduction, it results that unless the
species is able to meet these suddenly introduced unfavorable conditions it is
exterminated. It is here that the habit which we have been discussing would
be of value for the purpose of preventing a wholesale destruction of the spe-
cies. Furthermore, only those individuals which possessed this trait would
be able to survive and reproduce, and thus the character would be preserved
MATING HABITS. 233
by selection. However, before natural selection could act, this character
must already have been in existence and developed to the extent that it had a
selectional value to the species. How such a character could first arise and
attain a development sufficient to give it selectional value is quite another
problem. Here we simply admit the existence of the character and its use-
fulness in the economy of the species when it is developed to such an extent
as to be of selectional value; its origin and its development up to this point
will be considered later.
Under normal conditions of existence these beetles, after they have passed
the period of feeding and maturing the reproductive products, begin a most
ardent courtship, which in a relatively short time results in mating and fertili-
zation. In general the males mature first, and begin actively at once to
search for females that are in condition for reproductive purposes, or that are
willing to receive them in coitus ; for often the female receives the male before
her eggs are ready for fertilization.
The courting habits of these beetles are of the simplest kind. The male
normally approaches the female quickly, with his antennae vibrating rapidly
in a rotary manner in front of his head, and after a few passes of the antennas
over the body of the female he climbs quickly upon her back, and there
attempts to secure a firm hold with his legs. If the female is not averse to
mating she stands quietly, but if, because of immaturity or for some other
reason, she does not care for the attention of this male or of any male at this
time, she will run away or take flight at his approach. Even when the male
succeeds in gaining a firm foothold upon the back of the female, he can not
copulate unless she opens the plates which surround the genital passage.
Often when a female is busy feeding a male will approach her quickly and
gain a perch upon her back before she is able to escape, and then with ex-
truded penis will attempt to force an entrance between the plates which guard
the entrance to the vagina, but I have never seen one accomplish this.
Usually after a few vigorous attempts to copulate the male retracts the penis
and rides around on the back of the female, often for an hour or more, before
he becomes weary of this occupation and seeks for a more responsive female.
When the female is in a receptive condition and the male has secured a
firm footing upon her back, he settles backwards upon her body, the two
plates which guard the entrance to the vagina are opened, and the penis of
the male is allowed to enter. The beginning of copulation is accompanied
in both animals by a few convulsive movements which consist of spasmodic
expansions and contractions of the abdominal segments in both, and of the
raising and lowering of the body of the male. These convulsive movements,
however, last for only a short time, when both animals become quiet, and the
female frequently resumes her feeding. At intervals of five, ten, or fifteen
minutes thereafter, however, the male raises and lowers his body rapidly
17— t
234 HABITS AND INSTINCTS IN LEPTINOTARSA.
three or four times, and then becomes quiet again. Copulation lasts from
two or three minutes up to ten or twelve hours, the first coitus, however,
being usually far longer than those which occur subsequently. When the
act is completed the two sexes separate, and within a short time each may be
found in copulation with some other individual.
As far as I have been able to discover, there are no complicated habits asso-
ciated with the mating of these beetles, it being effected with the least trouble
or ceremony when they are ready for copulation. Rarely, in some species, as
in dccemlineata, oblongata, dilecta, and others, the two animals contemplating
mating will first feel each other over with their antennae, while others will
walk around as if inspecting each other. In oblongata the male will some-
times open his elytra to expose the bright red color of his hind wings, but as
he is quite as apt to do this when the female is not present it is probable that
this action has no particular significance. After these more complicated
courting movements copulation is no more apt to follow than it is if the male
approaches the female and attempts to begin copulation without any previous
ceremony. When compared with the accounts that have been published of
the courting habits of various other insects those of Leptinotarsa seem simple
indeed. I have often watched for these highly complicated and romantic
courtships among insects, but I have not yet been able to see them ; and I
should be very much interested to know just how much of the ceremonies
described was furnished by the insects and how much by the imagination.
The directness of the mating of the sexes in Leptinotarsa and its lack of
ceremony leaves very little room for the play of sexual selection in these
forms, for, although there is no dearth of bright colors, they do not seem to
play any part in the bringing together of the sexes. Indeed, the only action
that can be interpreted at all as in the nature of ceremony is the opening of
the elytra of the male to display the bright color of the hind wings ; but as
this is done as frequently in the absence of the other sex as in its presence,
and also by both sexes, no one but the extreme advocate of sexual selection
could see in this act one of utility as an attraction to the opposite sex.
The attracting of the two sexes is largely accomplished by the odors
secreted by the glands of the body, and especially by the glands of the genital
organs. I have often tried the experiment with dccemlineata, oblongata,
dilecta, and niidccimlineata of removing the antennae and palps of the males
and then turning them loose in an observation cage with a number of females
in condition for copulation. As far as observed, the males were unable to dis-
tinguish the sex of their companions, and if they attempted copulation at all
they were as apt to choose males as females. However, if by accident a male
minus his olfactory organs came upon a female and succeeded in mounting
her back, he was just as able to copulate as any other male, but he had to
depend upon chance to bring him to the female. Under normal conditions a
MATING HABITS. 235
male frequently mounts the back of another male, but I have never seen one
attempt to copulate. When mutilated, however, they try most vigorously
and persistently for a long time to effect copulation with other males, thus
showing their inability to recognize the difference between the sexes without
the help of their olfactory organs.
Similar results were obtained by the experiment of painting the antennae
and palppe with paraffin, balsam, shellac, and other substances which prevent
the use of the olfactory organs, but do not give the shock occasioned by the
removal of the antennas and palpas. Beetles thus painted were quite as unable
to distinguish the sex of their companions as were those without these organs ;
but when this coating of paraffin was removed they could distinguish the other
sex as readily as those beetles that had not been experimented with.
I also tried the experiment of painting the bodies of the animals black, red,
white, green, and other colors, using largely individuals of the species decem-
lineata, multitaniata, dilecta, undecimlineata, violescens, and lacerata. As far
as I have been able to observe, the giving to either male or female a color not
natural to it had no influence whatsoever upon the mating in any species,
neither when the color was applied to one sex alone nor when it was applied
to both.
Experiments were tried in which the abdomen of the female was covered
with shellac, paraffin, or other substances which would prevent the odor of the
sexual glands from escaping; the result was that the males passed by the
females so treated without recognizing them as females at all. Experiments
with various mutilations were tried, but in all the same result was obtained,
that if the olfactory organs and the scent glands remained intact the sexes
unerringly recognized each other ; but any interference with either of these
organs at once prevented the recognition of the opposite sex. In these
beetles, as in Lepidoptera (Mayer, 1900), sex recognition is accomplished
solely through the olfactory organs and scent glands, the colors playing no
part whatsoever in the process. Whether these glands vary or not it has not
been possible for me to demonstrate, although there is no reason why they
should not show variation in the same manner as do other characters of the
body. If there is a variation in the scent glands either in number, in the
strength of their secretions, or in the ability of the male to perceive the odor,
it is not great enough to produce any observable selection in the mating of
those beetles which have come under my observation.
In all the species of the genus Leptinotarsa copulation occurs many times
in the reproductive period, and with different individuals, so that often by the
time the first eggs are laid the receptaculum seminis contains spermatozoa of
several different males, and hence the offspring of one female may have
several different males for the other parent. This results in a great com-
mingling of the species in reproduction, and a consequent tendency to keep
236 HABITS AND INSTINCTS IN LEPTINOTARSA.
the species in a mediocre condition by decreasing greatly the probability of
extreme individuals arising in any great numbers through their exclusive
breeding.
Although copulation is accomplished with ease and without ceremony in
these beetles, it is not effected unless the conditions of existence are proper
for the act, or nearly so. The reason for its failure is largely, if not entirely,
the non-development of the germ-cell, any change in the normal rhythm of
reproduction being usually associated with changes in the environment, such
as excessive heat, cold, or moisture.
The laying of the eggs does not present any features of special interest,
excepting that if the conditions of existence are too extreme the eggs, although
they may be fully formed and fertilized, are not laid, but are retained in the
passages of the female reproductive organs until they are resorbed, or, as
more frequently happens, until the female dies. This failure to deposit the
eggs seems to be caused solely either by the drying of the secretions or by the
inability of the female to produce them in the accessory glands.
In the normal laying of the eggs in deccmlineata the female selects a proper
spot, and, raising her body high upon the legs, protrudes the posterior seg-
ments, opens slightly the genital plates, and, touching the tip of the abdomen
to the leaf, allows a drop of a yellow oily fluid to escape ; then as the abdomen
is slightly raised the egg protrudes and is fixed by its posterior pole in the
drop of fluid already deposited. The fluid now hardens rapidly and cements
the egg in place. One egg having been deposited, the female moves along a
slight distance, and there places another by the side of the first, and so on until
there is a row of from five to ten eggs in a nearly straight line on the leaf. A
second, third, fourth, and often as many as ten rows are thus laid, each of
which is more or less closely placed to the previously laid rows, and forms
therewith a compact bunch. I have observed the egg-laying in some 20 spe-
cies of this genus, and in all the procedure is the same. All the eggs that are
ripe at a given time may be deposited in one spot, or the beetle may lay a few
and then move on to another place to deposit a few more, and thus the 30 to
75 eggs which mature at one time in these beetles may be laid on a half dozen
different plants.
In all of these beetles the eggs are developed in batches, each of which is
laid before another is developed. The average number of eggs laid by the
females in this genus is 375, although there is considerable variation both in
the number of eggs laid by the different species and in the number of batches.
The data on hand along this line are brought together in table 104, on
the following page.
MATING HABITS.
237
Table 104. — Number of eggs and number of batches of eggs laid by different species
of the genus Leptinotarsa.
Species.
Average
number of
eggs.
Average
number of
batches.
Average
Variation in egg-laying.
egg-laying
period
in days.
Maximum
number of
eggs.
Minimum
number of
eggs.
Maximum
number of
batches.
Minimum
number of
batches.
L.decemlineata. .
multitieniata. .
oblongata
rubicunda
undecinilineata
signaticollis . .
dahlbomi
violescens ....
zetterstedti . . .
rubiginosa. . . .
haldeniaiii. . . .
lineolata. . .
45°
425
300
300
400
425
200
175
175
250
270
200
200
175
200
12
IO
IO
IO
12
6
12
14
8
7
7
9
11
6
7
3°
3°
25
3°
25
35
15
20
14
20
20
20
20
23
22
600
783
44S
372
621
643
310
242
2IO
290
381
242
260
234
300
190
211
114
185
154
396
107
31
IO
200
200
176
I40
152
I09
18
22
14
14
20
12
15
20
IO
12
14
13
17
4
5
4
3
7
6
4
5
10
10
6
6
5
6
7
8
8
Concerning the breeding habits of these beetles three facts have been
learned which are of importance in succeeding studies. First, the habits of
these beetles in normal reproduction are promiscuous; hencetheoffspringfrom
any one female may have several different male parents, and from material
taken in nature in copulation or otherwise we can not be sure of results in
breeding experiments until we have reared such material in experiment for at
least one or two generations, and are certain of the purity of the race under
experimentation ; second, the eggs are laid and developed in batches, which in
the genital passages of the female do not mingle with successive batches —
that is, each is distinct from that which precedes and from that which fol-
lows— a fact that has been made great use of in the experiments in the pro-
duction of modifications ; third, the conditions of existence may prevent the
deposition of the eggs either through the inhibition of copulation or of the
development of the ova, or through the mechanical prevention of egg-laying
by the inability of the female to secrete in the accessory glands the fluid which
enables the eggs to pass easily through the lower part of the genital passages,
and which serves also to cement the eggs to the leaves. These three facts are
of paramount importance in the carrying on of continued and successful
experiments in pedigree breeding and in the experimental production of new
races and modifications.
It has been clearly shown that there is not the slightest trace in the breeding
of these beetles of any selective influence due to color or form, that elaborate
mating habits are wanting, and that the sole means of sexual recognition is
through the olfactory sense and the odoriferous secretions of the accessory
238 HABITS AND INSTINCTS IN LEPTINOTARSA.
glands of the reproductive organs, which are active on the maturing of the
genital products. It appears, further, that as far as can be determined there
is no selective influence exerted upon mating by the variations which probably
exist in the strength of the glandular secretions and in the ability of the dif-
ferent males to detect the scent. If such selective influence be present, it is
so slight that it is not to be detected by any method known to me.
ASSORTIVE MATING.
Assortive mating, which implies the tendency of certain conditions to be
brought together in reproduction with greater frequency than other condi-
tions, is believed by some of the modern selectionists to be a constant and a
rather important factor in the mating of animals. This idea is a further and
perhaps more logical extension of the sexual-selection principle which has
been used to account for many ornaments found in animals. In the genus
Leptinotarsa assortive mating is found in the habit that extreme individuals
have of showing a strong tendency in mating to select partners of a more
mediocre condition than themselves, whereas the more ordinary individuals
usually select partners that are their exact counterparts ; and this is true of
each pairing, whether it be a first or a subsequent one. There seems to be
inherent in the species of this genus the habit of rejecting in reproduction
those individuals which possess extreme or unusual characters. I have
investigated this phenomenon in decemlincata, oblongata, undccimiiv.cata, and
dilccta, all of which species have furnished evidence of the existence of this
selective influence.
The most satisfactory material in the genus for the investigation of this
problem is the species oblongata, in which the assortive mating as regards
size, general albinic or melanic tendencies, and the dimorphic red and yellow
conditions in both sexes affords an excellent opportunity to test the action and
existence of this phenomenon.
In respect to size there is a most distinct assortive mating in this beetle, a
fact that has been determined both statistically and by a close study of the
beetles in nature. In these observations it was apparent that the extreme
individuals were not, as a rule, able to obtain partners in the reproductive act.
Thus the largest and smallest of the individuals of both sexes were far more
apt to be found unmated in nature than were the more mediocre individuals.
This is much more clearly shown in the tables of the seriations of these beetles
in respect to size. In these seriations the classification has been made upon
the basis of ten classes, in which 1 represents the smallest condition known
and 10 the largest. The unit used in the measurement of these beetles was
the perpendicular distance between parallel lines passing through the apex of
the elytron and the median posterior side of the pronotum when the beetle
was at rest. The entire length can not be used because of the great variation
ASSORTIVE MATING.
239
in the position of the head and pronotum, which would increase or decrease
the bod}' length in individual cases without having any significance, and
would, therefore, introduce a large error into the result. The unit chosen is
not subject to any such error, and is a reliable index of size.
Table 105. — Assortive mating found in L. oblongata, size being taken as the selective
character.
Sex.
Normal range of variation— Class.
1
2
3
4
5
6
7
8
9
10
Males
Per ct.
2
Per ct.
I
3
Per ct.
3
4
Per ct.
6
Per ct.
15
5
Per ct.
40
12
Per ct.
17
38
Per ct.
10
19
Per ct.
6
IO
Per ct.
1-5
3
4
Per cent oj males of each class which mated with individuals in the various classes into
which the females were grouped as regards size.
Class of males.
Class of females.
I.
2.
3-
4-
5-
6.
7-
8.
9-
10.
Per ct.
90
6
4
Per ct.
10
70
13
7
Per ct.
2
6
71
12
Per ct.
13
74
12
1
Per ct.
I
IO
76
II
2
Per ct.
5
IO
70
13
2
Per ct.
Per ct.
I
2
85
IO
2
Per ct.
Per ct.
3
6
90
In the above table there is shown to be a striking correlation between the
size of the sexes in each pair, matings between males and females of very un-
like size not occurring at all in the series. Moreover, the number of unmated
individuals in any population is far greater among the extremes than among
the more mediocre individuals of the population.
It is highly significant that extremely large or small beetles are unable to
find partners among the more moderate individuals, and are thus effectually
prevented from becoming parents to the following generation, and thereby
perpetuating their more or less extreme conditions, should they happen to
possess heritable qualities. This inability of extreme individuals to find part-
ners does not seem to be due to any aversion to them exhibited by the more
mediocre part of the population ; it may rather be attributed entirely to the
fact that the exremely large or small beetles are not able to effect copulation
with the others. In mating not the slighest appearance is found of a desire
24O HABITS AND INSTINCTS IN LEPTINOTARSA.
to avoid the extremes, a large male being on the average just as apt to
attempt to mate with a small female as with one more nearly his own size,
allowing, of course, for the greater opportunity for finding a female more
nearly suited to his size among the females of the general mediocre popula-
tion. At any rate, the attempt of a large male to mate with a small female
invariably results in failure, owing to his inability to force an entrance into
the small genital passage of the female. On the other hand, large females
and small males do not seem to be able to mate with any degree of success on
account of the lack of adaptation in the size of the reproductive organs.
I have attempted to obtain statistical data upon this point, but the difficulty
of dissecting out the organs and the distortion that inevitably results are so
great that the error introduced is probably greater than the range of varia-
tion. At any rate, it is evident from the table that the extremes only very
rarely mate, and that when they do it is most frequently with mediocre indi-
viduals, as might be expected. Therefore, there is to a large extent in the
mating of these beetles an elimination of all but the mediocre individuals from
the parenthood of the following generation. Hence it follows that no matter
whether the characters of the extreme individuals be heritable or not, there is
little or no chance for transmission of them to the following generation, and
thus an effective although unconscious selective mating as regards size is
brought about in this species. The selective factor is, as has been shown by
•Crampton in Lepidoptera, a lack of adaptation in the size of the reproductive
organs of the sexes ; the end result is the preservation in the population of an
extremely mediocre condition as regards size, with a very few only in the
more extreme classes.
Assortive mating between beetles of the same general color value — that is,
between beetles which belong in the same class of color value as regards
albinism or melanism — has also been studied in oblongata, but with quite dif-
ferent results. As regards color alone there does not seem to be the slightest
trace of any selective influence in the mating of these beetles. Among 500
copulating pairs examined, it was found that a highly albinic male was on the
average just as apt to mate with a melanic female as with one nearer its own
color. Of course, if the albinism or melanism was associated with abnormal-
ities in size, there was a rejection of the individual, but upon the basis of size
and not upon that of color. From the general habits of this beetle in mating
and its entire lack of appreciation of color, we should not expect to find any
assortive mating upon this basis. Nor has there been discovered the least
trace of any selective mating or segregation on this basis in the dimorphic
males and females of this species; that is, males and females of like color
have not been found in copulation with any greater frequency than have those
of different colors. The data from 500 copulating pairs, taken from three
different generations, have not shown the slightest trace of preferential
mating in regard to these characters.
ASSORTIVE MATING. 24I
I have examined this phenomenon in decemlincata, undecimlineata, and
dilecta, and have obtained results exactly like those just described for ob-
longata. Among 300 copulating pairs of decemlincata from Cold Spring
Harbor, Long Island, New York, ioo pairs each from Woods Holl, Massa-
chusetts, and Newport, Rhode Island, 200 from Yellow Springs, Ohio, and
300 from Chicago, Illinois, or 1,000 pairs in all, I found that extreme individ-
uals were to a large extent absent, and I conclude that it was because of their
being unable to copulate properly with the opposite sex. However, when
color wras taken as a basis of selection, there was not the slightest trace of
preferential mating.
It appears, therefore, that selective mating exists in these beetles, but that
it is not found in any character excepting size, and that even here it is due
solely to the inability of abmodal individuals to properly perform the sexual
act. We do not ordinarily realize how narrow are the limits within which
successful copulation can take place in insects, or how slight a variation is
sufficient to prevent the performance of the sexual act with such completeness
as to insure the leaving of progeny. The wrork of Crampton has shown this
most clearly in Lepidoptera, where, fortunately, the genital organs can be
measured without the introduction of an enormous error. Although in Lep-
tinotarsa successful measurement is impracticable, it is certain that a like
slight variation in the size of the genitalia or in their position is enough to
eliminate the possessor thereof from participation in the perpetuation of the
species. Hence there results a most effective selective mating in which only
the modal individuals, or those in near-by classes, are able to leave any large
number of progeny, and whether the variations are heritable or not, the
extreme variations have little or no chance to be transmitted to succeeding
generations. It is noticeable in all of the seriations which have been made of
different characters in this genus that there is a constant high percentage of
the individuals in or near the modal class, nearly 75 per cent in the polygon
of variation being usually found in the modal class or in the classes imme-
diately above and below it ; and this is due largely, if not entirely, to the
selective mating which eliminates in reproduction the perpetuation of the
more extreme conditions.
In the preferential or assortive mating in these beetles we must therefore
recognize a selective process which acts to keep the species as close as possible
to the mode or standard by the elimination of extremes. Hence it is a con-
servative factor, and not one that would tend toward the production of modifi-
cations. If, however, from any cause whatsoever there should occur a rapid or
slow shifting of the mode of the species, preferential mating would still keep
it closely aggregated about the new mode, and thus bring about a modification
in the entire polygon of distribution. In this way selective mating has un-
doubtedly been and is now a valuable aid in the formation of races and spe-
242 HABITS AND INSTINCTS IN LfiPTINOTARSA.
cies. It is just the kind of a process which would be best adapted for the
splitting up of a polymorphic species through the limitation of reproduction to
the more mediocre of the individuals in each group, and the tendency to
eliminate intermediates, and also perhaps to prevent the crossing of closely
related forms. Assortive mating is no doubt equivalent to some one of
Gulick's (1905) divisions of segregation, but I can not see the need of
the extremely finely divided divisions of segregation, and prefer to use for
the present the simplest terminology. Although assortive mating (or seg-
regation or isolation in reproduction) is undoubtedly a valuable aid in the for-
mation of species, it must nevertheless find characters already developed to the
extent of having a selective value before it can begin to exert any influence.
It can not bring into existence anything; it can only preserve that which is
already in existence. Moreover, in the genus Leptinotarsa at least, a great
many characters, as, for example, color and ornamentation, are not influenced
at all by this form of selection, so that if, perchance, these non-selective char-
acters should exist in combination with characters which have a selective
value, they would be carried along with the selective characters in the segre-
gations which assortive mating brings about. In this way non-utilitarian
characters might be preserved by selection through being associated with
others which are perhaps less conspicuous, but which have selective value.
However, in order to explain all nature and evolution through selection, one
must not assert that all apparently non-selective characters are correlated with
selective characters which have not as yet been discovered, and perhaps never
can be. Assortive mating has been invaluable in the production of new
characters in experiment and in the perpetuation of the new races which have
arisen in my experiments, and I can see no valid reason why it should not be
an equally important factor in nature, although this is not as capable of
demonstration or of experimental proof. It is possible, however, as I shall
show later, that there is the best of circumstantial evidence that this selective
process has been a powerful factor in the evolution of this genus of beetles ;
and if it is a potent factor here, why should it not also be one in other cases?
Although in these beetles, as far as we can determine, color and ornamenta-
tion seem to play no part in selective mating, there is no reason apparent why
these factors may not be of selective value in other animals. It is well to re-
member that we are viewing these beetles as outsiders, and are not, therefore,
able to appreciate the minute individual differences which they exhibit and
which may play important roles in the mating of the species. It is quite pos-
sible that there are in animals characters of a selective value which we are
unable to recognize, but which the animals themselves appreciate. Even in
our own species, the individuals when in a crowd appear so much alike that
we are unable to recognize persons whom we know from strangers. In
foreign races of our own species we recognize individual differences and
NUMBER OF GENERATIONS. 243
attractive qualities with difficulty, or not at all ; and the same difficulty must
be experienced in our observation and appreciation of individual differences
and attractive qualities in animals, but to an immensely greater extent. This
limitation of our appreciation of the qualities of animals should not be used as
a screen behind which to take refuge with some cherished hypothesis, there to
maintain its universal validity ; for although these processes and conditions
may exist, their existence is not open to proof or disproof.
NUMBER OF GENERATIONS.
The number of generations in Lcptinotarsa each year, in both temperate
and tropical latitudes, is a remarkably constant character, and might well be
used as a generic differential. As far as I know, the number in all of the spe-
cies is limited to two. Thus, there are two generations throughout the range
of decemlineata, although Lugger has recorded three in Minnesota, and others
have supposed that there may be three in the southern United States. I have
not, however, been able to get decemlineata to breed more than twice in a
season without a period of hibernation or aestivation. In the spring decem-
lineata emerges from the ground, and after a period of feeding, during which
the germ-cells are also maturing, it breeds and lays the eggs for the first gen-
eration. These are usually all deposited at about the same time, but there are
always for a month or more some individuals that are laying eggs, and of
course the larvae and imagines resulting from these eggs which are last laid
are much later in maturing than are the majority of the population. The
first brood, on emergence, feeds for a few days, and then deposits the eggs for
the second generation. The majority of these eggs hatch late in the summer,
and after the animals feed and fly around for a month or more they burrow
into the ground, and there hibernate until the following spring. The second
generation does not develop the germ-cells nor show any reproductive activity
until after it has passed through a period of hibernation or aestivation. Beetles
are found breeding even late in the autumn, but these are the belated indi-
viduals of either the first or second generation. As far as I can discover, the
life cycle in this species is that given above. In the tropics, where one might
think that the constant high temperature would keep the beetles breeding all
the year through, there are likewise only two generations a year in all the
species that I have studied, it being necessary in all of them that the second
generation have a period of rest before resuming its reproductive activity.
Normally this rest is passed in the ground in aestivation, but by stimulation
the beetles may be kept active, although they will not breed until a period of
at least two or three months has elapsed. The species in the genus are there-
fore double-brooded, the second brood undergoing hibernation or aestivation
before reproductive activity is resumed. Owing to the long time in each gen-
eration during which the beetles may breed, there are found in nature at the
244
HABITS AND INSTINCTS IN LEPTINOTARSA.
same time throughout the whole season eggs, larvae, and adults, so that it
seems on first sight that there are a large number of generations in every year
in the tropics. In all the species of the genus the only difference between the
two generations is that the second does not develop the germ-cells until after
a period of rest, while the other develops them at once, or soon after
emergence.
The length of the season during which any given species is active varies
greatly in the different species, and also in the different parts of the continent
of North America, the season being longest in the moist regions of Vera
Cruz, Chiapas, and southward, and shortest in the arid semidesert and desert
regions of northern Mexico and the southwestern United States. The
lengthening of the period of reproduction is the cause of the long time during
which these beetles are found active in the more humid districts, and, vice
versa, the reduction of the length of the reproductive period is responsible for
the lessened period of activity in the more dry regions. Nowhere, however, is
the duration of the embryonic period, of the larval life, or of the pupal stage
subject to any great amount of variation. The time of the appearance of the
different generations, their duration, and various other data have been con-
densed into the following table :
Table 106. — Compilation of data concerning generations in the genus Leptinotarsa.
Species.
No. of
gener-
ations.
Date of
emer-
gence.
Date of
breeding.
First gen-
eration.
.Second
gener-
ation.
Average
length of
repro-
ductive
period.
Length of
season.
h. undecimlineata .
signaticollis ....
angustovittata. .
multitteniata . . .
oblongata
rubicunda
intermedia
decemlineata . . .
2
2
2
2
2
2
2
2
2
2(?)
2
2
2
2
2
2
2
2
2
2
May 15
May 1
May 20
May 15
June 7
June 1
June 1
June 10
May 15
May 15
May 15
June 1
June 1
May 10
May 30
May 30
June 1
June 1
May 25
May 30
June 10
June 1-10
June 15
June 1-10
June 15
June 10
June 10
June 15
June 1
June 1
May 25
June 10
June 5
[tine 1
June 10
June 12
June 14
June 15
June 1
June 5
July I
July 10
July 15
July 10
July 15
July 15
July 15
July IS
July 1
July 1
July 1
July 1
July 10
July 10
July 15
July 20
July 20
July 15
July 15
July 15
Sept. 15
Sept. 1
Sept. 1
Sept. 1
Sept. 5
Aug. 25
Aug. 25
Sept. 10
Sept. 1
Aug. 20
Aug. 20
Aug. 15
Aug. 25
Sept. 1
Sept. 15
Sept. 1
Sept. 1
Sept. 10
Sept. 15
Sept. 1
Days.
20
3°
18
21
35
20
14
30
35
20
20
20
23
25
15
22
25
20
14
19
Days.
240
I50
I50
200
I50
I50
120
loo
125
I50
I30
I20
120
I20
IOO
IOO
120
I50
150
I50
defecta
libatrix
lacerata
modesta
AESTIVATION AND HIBERNATION. 245
HABITS AND INSTINCTS CONNECTED WITH HIBERNATION AND
AESTIVATION.
As has been shown in the preceding section, all Leptinotarsas either hiber-
nate or activate in each alternating generation. This phenomenon is one of
importance in the life cycle of each species, as it occurs in that portion of the
year when conditions of existence are least favorable. As far as is known, all
of the species in the genus are in this dormant state for at least three months
of the year, and the larger portion of them for five or more months. Hiber-
nation and aestivation are fundamentally one and the same process, the term
hibernation being applied to the dormant period produced by lowered temper-
atures, and aestivation to that which follows the coming on of the dry season
in the warmer parts of the earth. There is, I believe, no reason for the exist-
ence of two terms to describe this phenomenon, as both apply to the same
physiological process, the only difference being that found in the factors which
initiate them. In this paper I shall use them as the synonymous terms, which
they really are, and we shall see later that in these beetles the two processes
are the same.
Throughout the whole of the United States, after the second brood of
decemlineata has emerged in late September or early October, it devours a
large amount of food in the first few days of its adult life, and stores up in the
fat body a reserve supply to be used in the ensuing period of hibernation and
in starting the germ cell onward in development in the spring. This storing
up of food accomplished, the beetles stop feeding, and the preparations for
hibernation are begun. This is initiated by the emptying of the alimentary
canal of all food and the elimination by the malpighian tubules of the excess-
ive quantity of waste products. These are often red, as they are at pupation.
This clearing out of the alimentary canal and the elimination from the body of
as much of the waste of metabolism as is possible requires from three to as
many as ten days for its accomplishment, it being effected rapidly in warm,
clear weather, and retarded in cold or cloudy weather. The beetles now no
longer pass the nights upon the plants, but late in the afternoon, as the sun's
rays become more and more slanting, they are found burrowing into the
ground, often to a depth of 3 or 4 inches, where they remain until the next
morning, crawling out again when the sun's rays have warmed the earth ; and
if the day be cool the beetles do not come out in any numbers, and may even
remain in the ground for several days if the temperature is constantly rather
low.
After the contents of the alimentary canal have been emptied out and the
excretion of the excessive amount of waste products from the malpighian
tubules has commenced, the body weight begins to fall rapidly and continues
to do so until on the average there has been a loss of about 30 per cent of the
gross weight of the beetle by the beginning of preparation for hibernation.
This reduction in weight is occasioned by the emptying of the contents of the
246 HABITS AND INSTINCTS IN LEPTINOTARSA.
alimentary canal, which, on the average, makes 3 per cent, and by the removal
of the watery material from the excretory organs and the evaporation from
the body surface, which make the remaining 27 per cent. The average dry
weight of the beetles does not decrease at all, however, but is rather increased
by the excessive feeding that has gone before ; hence in its net dry weight the
beetle is heavier when it goes into hibernation than it was when it emerged
from the pupa, although its gross weight is somewhat less.
It is evident that preparation for hibernation consists largely in a reduction
of the watery contents of the body and in an elimination of all food and
other substances from the alimentary canal. In other words, preparation for
hibernation consists in a physiological change in the constitution of the body
for the time being and a consequent lowering of the freezing-point of its tis-
sues in exactly the same way that the spores of many plants and the over-
wintering eggs of Rotifers prepare for the coming on of unfavorable conditions
in their environment. Through the loss of water the protoplasm becomes
denser, and takes on characters that are at once recognized in cytological
preparations. The cells of the body also become shrunken and flattened,
the nuclei take on a most extreme vegetative appearance, and in some
cases it has not been possible to demonstrate the existence of chromatin in
cells where six or eight months later abundant and active chromatic condi-
tions will be found. In all of the cells the protoplasm takes on a peculiar col-
loidal or else a fine granular appearance and retains it throughout the whole
period of hibernation. In the hypodermal cell the lipochrome colors fade and
the beetles become light straw yellow, or often white.
All parts of the body participate in the preparation for hibernation, in which
the essential process is the elimination of water until the maximum and min-
imum at which the protoplasm of the animal can still survive is increased in
both directions to such an extent that lowered or increased temperature will
not be likely to produce fatal results. After the beetles have made prepara-
tions for hibernation they do not long delay the final burrowing into the
ground and the beginning of the long period of dormancy. L. decemlineata
hibernates at an average depth of from 18 inches to 2 feet, and I believe that
if the winter is very cold it may in some cases work down even farther.
In decemlineata I have not by any process been able to prevent this prepara-
tion for hibernation in beetles which should normally hibernate. By high
temperatures and moisture it is possible to greatly inhibit the process, often
to the extent that it is not completed, and as a result many of the beetles
die. Even if they do not burrow into the ground, the hibernation prepa-
ration is completed at the proper time in the life cycle of the species, and,
although the beetles are active and out of the ground, they do not feed for
weeks at a time, but rest upon the sides of the observation cages or on the
food plants. In this way they may pass three or four months, when they
again begin to feed, and in a few days are ready for reproduction.
ESTIVATION AND HIBERNATION. 247
Upon emergence from hibernation changes are found which are the reverse
of those which take place in the preparation for hibernation. Immediately
upon emergence a beetle voids the waste products that have accumulated in
the malpighian tubules, and, if food is at hand, begins to devour it most rav-
enously, so that it increases in weight with marvelous rapidity, until in three
or four days from the time of emergence it weighs more than it did at the
beginning of the preparation for hibernation. The beetles on emergence from
hibernation weigh less in both net dry and gross weight than they did at the
beginning of hibernation, which shows that there has been a low grade of
metabolism going on throughout the period of hibernation.
We see that there is a rapid gain in the water content of the body as well as
a net gain in the solid. In this period important cytological and histological
changes are also going on in the soma cells of the body. The protoplasm be-
comes more watery and vacuoles appear, while cells become larger and more
turgid and the chromatic elements in the nuclei increase greatly in size and
stain deeply, thus presenting all the signs of intense activity. In general the
body on emergence passes through the same changes as those which are con-
nected with the preparation for hibernation, but in the reverse order.
This process is a deep-seated one in decemlineata, and one that can not by
any known method be postponed or prevented. The process can be modified
in various ways, but it can not be eliminated from the life cycle of the species,
and, although through excessive and highly abnormal stimuli I have succeeded
in causing a semi-active hibernation, I have not been able by selection and
continued stimulation to modify this habit in the slightest.
Hibernation has, in decemlineata at least, a decided selective influence which
acts to produce much the same result as does assortive mating. That is, a
very small number, if any, of the more extreme individuals of a population are
able to survive the hibernation period, and the most extreme of the variations
found in any generation in the autumn are almost certain not to emerge in the
following spring. In nature the second generation frequently contains a con-
siderable percentage of rather extreme individuals, so that the polygon of
variation is strongly skewed in one or the other direction ; but in the spring
the beetles which emerge from hibernation are not as strongly skewed, if at
all, which shows the leveling action of hibernation through the elimination of
extremes in variation. Just why these extreme variations should not be able
to survive the hibernation period is at present unexplained. The selective
influence of hibernation is well shown in the table on page 248.
The selection produced through the elimination of the extremes during
hibernation is entirely in the direction of conservatism, tending to preserve
those individuals only which conform most exactly to the modal standard of
the species at a particular time and place. Hence, in the preservation of this
species and in the evolution of species in general, this selective influence would
act in precisely the same manner as does assortive mating. Likewise a large
248
HABITS AND INSTINCTS IN LfiPTINOTARSA.
number of the attributes of an animal, as, for example, all the color characters,
would fail to come directly under the influence of this kind of selection
because of their non-selective value, and their elimination or preservation
would therefore depend upon whether or not they were associated with some
character which did have a selective value to the species.
Table 107. — Selective action exerted upon L- decemlineata by the hibernation period.
[Hibernation in the winter of 1903 to 1904 at Chicago, Illinois. The table is based upon the
species as a whole, and has been obtained by seriating the individuals into classes accord-
ing to their likeness to each other.]
Period.
Class.
2
3
4
5
6
7
S
9
10
11
12
13
Before hibernation.
After hibernation
o-5
I
i-5
2
3
4
I
19
25
42
54
21
19
3
1
2
1
The species of this genus that live in the tropical or semi-tropical regions
also pass a portion of each year in a dormant state, which has been termed
aestivation. The cause of this hibernation in tropical species is the coming on
of the dry season, which produces the same physiological results as does cold
in the northern latitudes. I have studied the conditions of aestivation in a
number of species in this genus, two of which will serve to illustrate the phe-
nomenon as it is found in the tropical members of the genus. The conditions
found in undecimlineata are representative of species which live in the more
moist regions, and those of inultitccniata of species which live on the high and
dry plateau, where the moisture conditions tend strongly toward aridity. The
proper apparatus has not been available for the accurate weighing of these
two species in nature, but this in no way vitiates the general results that have
been derived from the study of aestivation in these beetles.
Lcptinotarsa undecimlineata, like decemlineata and all the other species in
the genus, hibernates in the second generation of each year, no matter whether
the proper time is reached in the rainy season, at its close, or thereafter.
When the beetles are ready to hibernate they will do so, in spite of the condi-
tions immediately surrounding them. The second brood in this species is
found in general in the months of September and October, hibernation begin-
ning in October or November with the cessation of the rains. After emerg-
ing from the pupa the second generation of this beetle feeds often for as long
a time as three weeks or a month, but not as ravenously as does decemlineata
in the north. During this time the weight, both gross and dry, increases
steadily until the animal gradually ceases feeding. The intestine is then emp-
tied of all food and the malpighian tubules secrete an excessive amount of
waste, which is often tinged reddish in color. Then there is the reduction in
AESTIVATION AND HIBERNATION. 249
the gross weight due to the loss of water. There is, however, no loss in the
dry weight during this preparation for aestivation. Yet even when fully pre-
pared for the advent of the dry season, the beetle does not at once enter upon
the dormant state, but may live in a rather inactive condition upon its food
plant for days or even weeks before it burrows into the ground to pass the
dry season. When, however, the dry winds come at the cessation of the
rains, the beetles usually seek retirement in the ground, but they do not now
burrow down more than 6 or 8 inches, and often may be found within only an
inch or so of the surface. They remain in aestivation until the rains come at
the beginning of the following season, when they are awakened to renewed
activity and another series of generations. On emergence from aestivation
they go through the changes found in the preparation therefor, but in reverse
order.
The preparation for aestivation in uiidccimlincata results in a reduction of
the water content of the body, and hence in an increase in its capacity to meet
the high temperatures which often occur, especially at the end of the dry
season, when the lack of moisture and the intense sunshine in a cloudless sky
heat the soil up to a high degree. Soil temperatures taken on the savannas
of Vera Cruz in April, 1904, in places where these beetles were activating,
were frequently as high as 6o° and 65 ° C. It is for the purpose of success-
fully passing through these high temperatures at the end of the dry season
that the physiological changes involved in aestivation are due. Moreover, in
the same regions the temperature frequently falls rather low at night during
the dry season, although there are never frosts, 30 C. being the lowest re-
corded temperature.
With multitcrniata, which lives on the cold, dry plateau, the preparation for
aestivation is the same as that described for the two preceding species. This
form has to encounter in the dry season not only high temperatures by day,
but also freezing conditions at night, and great desiccation at all times for
seven or eight months of the year. Although the environment differs greatly
in the three species, the manner of preparing for the season of the year when
unfavorable conditions of existence are most apt to obtain is the same in all.
The process is simply that of storing up a reserve food supply in the fat-body
and removing all matters which might ferment, decay, or become poisonous
to the animal, and the reduction of the water content to a lower percentage.
It is this last which gives to the organism the ability to resist the extremes of
temperature and moisture which it surely will encounter in the course of the
ensuing winter or dry season. The concentration of the protoplasm very
greatly widens its range of existence for the time being. This concentration
of the protoplasm in hibernation and aestivation has been studied much in
plant spores and in the lower animals, and the general conclusion has been
reached that increased capacity to resist high or low temperatures or extreme
desiccation is always due to the loss of water in the protoplasm ; and like-
18— T
25O HABITS AND INSTINCTS IN LEPTINOTARSA.
wise, in the higher forms, as in insects, increased capacity to resist extremes is
due to the lowering of the water content in the protoplasm.
In the more arid portions of the American continent which are inhabited by
these beetles, as, for example, on the Pacific coast of Mexico or the northern
portion of the same country, the dry season is frequently prolonged for weeks
or even, in extreme cases, for a year or more. In many such places these
beetles live and are not exterminated by this excessive variation in their
environment, or, more correctly, there is no evidence that they are, and it is
of interest to know how they are able to withstand this extreme condition. I
have not had an opportunity to study this phenomenon in nature, but I have
reproduced in experiment such conditions with success and with suggestive
and conclusive results.
In September, 1902, I placed a lot of 300 decemlineata in a cage which was
prepared for maintaining constantly a high temperature and a low percentage
of humidity. These beetles, which were of the second generation of 1902,
had been reared under normal conditions and would have hibernated in a
few days, and yet hibernation was postponed for weeks, although by about
December 1 all that were alive had entered the ground. The beetles in this
experiment were kept constantly dry and warm, although the diurnal range
in temperature was considerable, and a small amount of moisture was added
from time to time to compensate for that which would be supplied in nature
by dew and by the soil lower down. The beetles continued their hibernation
through the remainder of the year 1902, all the year 1903, and the winter
of 1903-1904, and only finally emerged when, in May, 1904, I gradually
increased the amount of moisture in the tank until the conditions found in
nature at that time were reproduced. Near the end of the month 76 beetles
emerged and began feeding, and during the summer of 1904 two normal gen-
erations were produced, the second of which hibernated in due course of time.
As far as I was able to see this prolonged aestivation did the beetles no harm
whatsoever ; and it is quite possible that if I had been able to restrain my
curiosity as to whether they were alive or not, they might still be contentedly
hibernating in their tank, to emerge at some future time. This experiment
shows how these beetles may be enabled to pass successfully over unfavorable
conditions of long duration, and I have not the slightest doubt that this is the
way in which the beetles of Chihuahua, San Luis Potosi, and other States in
northern Mexico manage to survive where there occur from time to time
seasons or a succession of seasons in which there is not enough rainfall to
start vegetation and animal life into activity. At any rate, the beetles are
found in these regions now, and if they have been exterminated in the past,
they have been able to reoccupy the region by immigration from the sur-
rounding territory with a swiftness which is astonishing and improbable,
owing to the slowness of the dispersal of these beetles in arid regions. More
crobable, especially in the light of this experiment, is the hypothesis that they
ESTIVATION AND HIBERNATION. 25I
are able to extend hibernation much longer than is normal, and thus bridge
over periods when impossible conditions of existence prevail.
Closely associated with the hibernation of the second generation in these
beetles is the quiescent or resting period in the cycle of the germ plasm. That
is, the germ cells do not develop at all in the autumn, but remain as oocytes
or spermatocytes, which are relatively few in number, until the following
spring after emergence from hibernation, when they develop rapidly. This
period of inactivity in the reproductive elements has been regarded as due to
some inherent necessity for rest in the germ plasm. However, it does not
seem to me to be of this nature ; it is rather in the nature of a very deep-seated
adaptation, like aestivation, which has been developed and retained not only
in this genus, but also in the whole insect phylum, for the purpose of -enabling
them to pass successfully over the season of the year when unfavorable con-
ditions of existence are most apt to occur. In my experiments a race arose
suddenly in which there were five generations, and then a period of rest, and
then five more, and so on. In all insects, however, there is normally this
period of cessation of reproductive activity, which comes at some time in the
life cycle of each species, and corresponds in all exactly with the time of the
year when unfavorable conditions of existence are most apt to occur. In some
species it is passed over in the egg, in others in the larva, the pupa, or the adult,
and there seems to be no general law by which to determine in which of the
stages of existence the dormant period shall be passed. Inasmuch as closely
related forms pass this period in different stages, and after different numbers
of generations, there seems to be no alternative to the conclusion that we
must regard this phenomenon as an important and fundamental adaptation
which has been developed and preserved in the phylum for the reasons already
stated. The points of chief interest in this research, however, are the adapta-
tion which appears in alternative generations purely as a matter of hereditary
influence and the fact that changes in the environment can not postpone the
habit, although they may modify it to a considerable extent in its unimportant
features. The deep-seated character of the physiological process involved in
aestivation and its appearance in every second generation as an inherited habit
or instinct makes it one of the most interesting of the physiological characters
found in these beetles, or, indeed, in all insects. It has therefore been used in
this research as a character well adapted for experimental investigation, and
the attempts to obtain permanent modifications of it have been remarkably
successful. Because of its clear-cut nature it has been of the greatest service
in the investigation of the inheritance of habits and instincts in these beetles.
In nature, besides the utility of this perod of rest as a means of carrving
the beetles over the dry season or the winter, it also acts as a strong selective
factor. This has been shown statistically for decemlineata, and could be for
other species were it worth while. The selection, however, is all in the direc-
tion of conservatism — that is, only those individuals that conform most
252 HABITS AND INSTINCTS IN LKPTINOTARSA.
exactly with the standard of the race are able to survive in any numbers.
That is, here natural selection eliminates, and not perpetuates, the extremes
of variation. It is a self-evident fact that the major portion of any species
numerically strong must be well adapted to its conditions of existence, and
these beetles are no exception to the rule. In these beetles, however, we find
that the selective action of assortive mating and hibernation, both of which
are strong and important selective agents, tend in the direction of segregation
and an elimination of the individuals farthest removed from the standard of
the race. This conservative or segregational selection has been an important
factor in the evolution of the genus Lcptinotarsa, and also in evolution in
general.
HABITS AND INSTINCTS CONNECTED WITH SELF-PRESERVATION.
The devices exhibited by these beetles in one way or another, which enable
them to obtain a measure of immunity from the attacks of their enemies, have
already been examined in a preceding chapter. These have been found to be
largely bright colors which, accompanied by the exposed situation in which
the beetles live, make them conspicuous objects upon their food plants, and
thus serve to advertise the inedibility of the possessors thereof. These warn-
ing colors do not entirely protect the beetles from their enemies, because we
find that each generation of enemies has to be educated, and in this education
a small percentage of the insects are killed. The influence of warning colora-
tion, however, is purely passive; yet there are in this genus habits that are
active agents of protection, and that are brought into play only on the ap-
proach of danger. It is these which will now be considered.
The most widely distributed protective habit shown by animals on the ap-
proach of danger is that of assuming a perfectly motionless position, with all
the organs and appendages as closely drawn together as is possible. Observa-
tions upon habits of this kind have frequently been made upon birds and mam-
mals, but less frequently upon lower forms. Many insects have this habit as
fully developed and display it in the same manner as do the higher classes of
animals. In the genus Lcptinotarsa it appears to a greater or less extent,
most frequently when the food plant is shaken or there is a sudden noise
or a shadow passes quickly across the beetles. Oblongata and signaticollis
respond to these stimuli, which are unusual in the habitat, by stopping quite
still wherever they are and drawing the legs and antenna; as close to the body
as is possible. In this perfectly motionless position they are far less easily
seen than they are when moving. The length of time during which this pose
continues depends upon the nature of the stimulus and its duration. If it is a
passing shadow which is not repeated, the beetle within a few seconds goes
on with its activities, whereas if the shadow passes back and forth over the
beetle repeatedly it remains in this pose for some minutes, and if the stimulus
PROTECTIVE HABITS. 253
is long continued seeks relief therefrom in flight, crawling rapidly to a darker
portion of the food plant, usually on the under side of a leaf, where it again
assumes a similar pose. Stimulation through noise, unusual odors, or the
shaking of the food plant produce exactly the same response.
During the breeding period the beetles are far less responsive to the above-
mentioned stimuli. Frequently they will take no notice of the odors, shaking
of the food plant, or noise which would at other times drive them to flight and
secretion in the darkest and most secluded spots. At such times the over-
powering influence of sexual instinct seems to render them oblivious to all
other stimuli. Throughout the entire genus there is little or no variation in
this habitat ; and it seems to be one that is inherited, for the beetles display it
as soon as they emerge from the ground. The protective value of the habit
seems to be relatively slight, for it does not serve to conceal the animals from
the enemies that are actively looking for them. Its real use seems rather to
be that of momentarily decreasing their conspicuousness so that they may be
passed unnoticed by any possible enemy. It also affords them an opportunity
to determine better the nature of the disturbing stimulus and the direction
from which it comes, and, possibly, also to decide what to do in order to
escape from the injurious effects which the cause behind the stimulus might
bring to them. At any rate, if the stimulus be long continued, the beetles
soon seek more efficient protection by other means, and these are either con-
cealment in some shaded and not easily visible portion of the food plant,
usually the under sides of leaves, or by more energetic measures, such as fall-
ing to the ground, and there remaining motionless as if dead until the danger
is past. The attempt to escape through hiding on the under sides of the
leaves shows the simple instinct of fear and the response by seeking safety in
flight. The imitation of death is, however, a more complicated instinct, which
is developed in some and not in other species, and also to different degrees in
the same species.
When an unsuccessful attempt is made to capture one of these beetles with
the fingers or forceps, it almost always relaxes its hold upon the food plant,
falls to the ground, and rolls away. It remains perfectly motionless where it
stops, with legs, antenna?, and all parts of the body closely folded together,
and whether it has landed on its back, head, or in a natural position, it does
not move until all danger has passed. Even in the most exposed situations
the habit is highly efficient as a protection, and in the usual conditions of the
habitat of these beetles, where the food plant is surrounded by grass and other
low, closely packed vegetation, it is a perfect protective device, as most dili-
gent and prolonged search is necessary if one wishes to discover them when
thus concealed.
The position assumed is to a considerable extent a specific characteristic.
Oblongata, multitccniata, and dccemlineata all likewise draw the appendages
close to the body, which is not the death position, either for these species or for
254 HABITS AND INSTINCTS IN LEPTINOTARSA.
the others. Although the position assumed is constant within a given species,
it varies in different species, as, for example, in rnbicunda, which falls with
the appendages held in a rigid position straight out from the body. In the
two species, multitccniata and rnbicunda, a point of great interest is presented.
In the production of the new species there is rapid change in this protective
habit as well as in the morphological characters. Hence it follows that modi-
fications in the habit may arise suddenly as well as by selection, and this opens
the question as to whether or not the habit itself may not have arisen in the
same way.
While being handled the beetles often assume this attitude for a short time,
but if the stimulation be continued they soon cease the attempt to escape
through concealment, and begin to struggle vigorously ; hence they do not
carry the habit to the extent that has often been described in many insects
that they allow themselves to be eaten or mutilated without manifesting a
sign of life. I have examined several reputed cases of this kind in the tropics,
but for some reason or other when under my observation they did not act as
they should according to the published accounts. At any rate, in the genus
Leptinotarsa the beetles do not carry the habit to the very evidently injurious
extent to which it has been carried by other forms that have been destroyed.
That the habit is really of use to the species in the genus which possesses it
there is not the slightest doubt. In nature I have seen undecimlineata, when
approached and examined by a gecko, relax its hold upon the food plant, fall
to the ground as if dead, and remain there until it considered that danger had
passed, when it would crawl back upon the plant. Of the utility of such an
act there can not be the slightest doubt. Of course it might often happen that
in falling the beetle would get into a worse predicament than the one from
which it sought to escape, but, as far as I can see, there is no possible means
of determining this point other than by having a trained observer constantly
watch the movements of a beetle throughout its adult life. As regards the
selective value of the habit, it seems to me to be small, owing partly to the
already existing high degree of protection afforded by warning colors, and
quite as much to the fact that an insectivorous animal is keen of eye and does
not often miss its mark. When, therefore, one of these beetles happens to be
the target, the aim is so true and the attack comes so suddenly that the beetle
has an extremely small chance of bringing this protective habit into play.
Rarely, as with the gecko, the beetle may have a chance to use the habit to
advantage, but such opportunities are few and far between, and hence can not
play any appreciable part in the preservation of the species.
After years of observation of these beetles I find it impossible to regard the
nature and purpose of this habit to be that usually attributed to the same reac-
tion in other animals. I have described the habit in detail, and have shown
that while it does have a certain protective value to the species, this value is
PROTECTIVE HABITS. 255
so slight that it is hardly worthy of consideration. After a long acquaintance
with these beetles, both in the tropics and in the temperate regions, it seems
to me that the habit is no more than the response produced by excessive fear.
There can not be the slightest doubt that fear is a psychic phenomenon which
is found in all animals and may be aroused by quite varied stimuli. In the
higher animals the continued application of a stimulus that excites fear pro-
duces a condition of muscle tension and inability to move, and frequently a
state of complete unconsciousness, while in the human race these states are
always accompanied by characteristic outward signs — pallor, trembling, pro-
trusion of the eyeballs, clenched fists, tense muscles, and other characters.
If a slight noise be made with a tuning fork these beetles pause and assume
the pose described; if the noise be continued they seek safety in flight; and
then, if it be greatly increased, they assume the so-called death attitude, from
which they emerge only after the stimulus has passed away and sufficient time
has elapsed for the beetles to recover from their fright. The attitude assumed
while in this state is caused simply by the nervous shock received by the body
of the beetle, and especially by the muscular system, the reaction differing in
no way from the well-known effects upon the muscular system of similar
states in the human subject. Therefore, until more conclusive evidence is
produced to show that the habit has the high utility so often attributed to
it, death-feigning must be regarded as the specific response to continued or
excessive fear-producing stimuli. In these beetles the habit is, I should judge,
quite as frequently productive of ill results as of beneficial ones. I am of the
same opinion as Morgan, that "the origin of these trophisms can not be ac-
counted for on the ground of their benefit to the individual or the race." Cer-
tain it is that in Lcptinotarsa there is not the slightest evidence that this habit
has any selective value, or that it is of any constant or marked utility to the
race. It may occasionally be of use in the preservation of an individual
beetle, but such rare and sporadic utility of the habit would not bring it under
the operation of natural selection, and I can not conceive of any rational way
in which it would have been produced by this agency. On superficial exam-
ination or on first sight this habit would appear to be one of great value to the
species, but a better understanding of the conditions of life in the species and
of the relation of enemies thereto shows clearly that the habit is really of little
utility in the life and death struggle of the species, and, therefore, has an ex-
tremely low selectional value, if any at all. For the purposes of this research,
however, it is a useful character in that it is inherited in its full intensity and
has a manner of outward manifestation that is more or less characteristic for
each species. Moreover, when a new species arises, the habit may also be
modified at the same time, which shows conclusively that not only are the
most superficial structures, such as color and ornamentation, able to make
rapid developments in their evolution, but that instincts and deep-seated
nervous reactions can also change as quickly and easily as do the supposedly
256 HABITS AND INSTINCTS IN LEPTINOTARSA.
less fundamental characters which are usually associated with this rapid
development of new forms.
The only other habit that can claim classification as a protective one has
already been discussed. It is the habit which the larvae of undecimliueata,
diversa, and signaticollis have of covering themselves with the trichomes of
their food plant. This, however, is an extremely passive habit, which is due
solely to the facts that the beetles feed upon species of Solarium that have a
rich development of trichomes on the under sides of the leaves, and that the
young larvae are supplied with spines in greater or less number and integu-
mentary glands which secrete a sticky fluid, so that the trichomes become
attached to the larva as it moves around upon the leaf. This habit is found
in the first two instars only, when the larvae are small and not active ; as they
become larger, however, the trichomes, even when they do adhere to the body,
are not apt to remain there, because of the more active movements. The effect
upon the young larvae is to produce a rather efficient disguise.
This habit or protective device is, however, accidental, because whenever
any of these beetles happen to feed upon a species of Solatium which does not
have the dense deposit of trichomes, they thrive just as well, both in nature
and experiment. Hence this apparently protective habit is simply the acci-
dental accumulation of trichomes upon the body which is found in these three
closely related species, because they all feed upon identical or closely related
food plants. As far as I can discover, the habit has neither utility to the spe-
cies nor has it injurious effects, and therefore it can not become a means of
selective action until the time shall arrive when it is either of advantage or
disadvantage to the species.
The three great groups of habits in these beetles have been examined not
with the purpose of an animal psychologist, but in order to learn what part
they play in the life of these beetles, which are being used as material for the
study of animal evolution. The reader must not, therefore, be greatly disap-
pointed if he fails to find herein the "final solution and explanation" of the
habits of these insects. In this paper an insight has been gained into the part
played by these various habits in the economy of the species, and to a certain
extent into the manner in which these habits may act in the evolution of the
genus. It has been clearly shown that some of the habits have a decidedly
selective influence, but in all the habits discovered this influence is exerted in
the direction of conservatism, and is of no use whatsoever in the preservation
of a newly-appearing variation, no matter how useful it may be to the species.
This conservative tendency, however, is a highly important factor in keeping
the species true to the modal type, and may thereby be a far greater force in
evolution than it would be if it concerned itself with every little variation pos-
sessing a slight degree of utility to the species. I am strongly of the opinion
that when we are really better acquainted with the process of natural selec-
PROTECTIVE HABITS. 257
tion we shall find that it has been the most important factor in evolution,
but that its importance as a conservator of species is due not to the tendency
usually attributed to it to preserve as many of the useful variations as pos-
sible, but to the fact that it holds the species, race, or variety to the modal
standard. So, too, it might with equal effectiveness be productive of the
development of varieties or of groups of species in highly variable forms
through this same strong tendency to segregate the species about one or more
conditions, and by the equally intense elimination of extreme or remote indi-
viduals. With any but the rabid selectionist it would go without comment
that the material must exist in such a state that selection can get a hold, or, in
other words, that the characters must have a selectional value either one way
or another before they can enter at all into the process. How these charac-
ters reach this state of development is a subject for experimentation and
research, and not for idle speculation. Finally, it has been seen that some
habits appear on first sight to be of great utility to the species, but are
not able to bear out this role when subjected to a broader and more careful
analysis.
CHAPTER V.
PRODUCTION IN EXPERIMENT OF RACES, NEW CHARACTERS, AND SPECIES
IN LEPTINOTARSA.
The species of Lcptinotarsa are clearly and distinctly separated from one
another, and each has its particular groups of characters. This isolation, or
lack of continuity, exists everywhere among all the species, when taken as a
whole, and also in their particular attributes ; and even though species possess
the same character in common, or the characters overlap, each in its totality
of characters is distinct from the sum total of characteristics discoverable in
its most immediate relative.
According to the Darwinian hypothesis this distinctness of species is sup-
posed to be the result of the extinction of intermediate forms ; however, few
traces of such a process have as yet been certainly demonstrated. In plants,
De Vries has described elemental species separated from one another by lesser
magnitudes than those separating species. The further extension of this idea
and its investigation in CEnothera has shown that in plants elementary species
arise by sudden development, so that the new is always separated from the old
by definite characteristics. This sudden permanent change of form is a phe-
nomenon more or less well known to plant and animal breeders, and has been
the means through which many of our domesticated animals and cultivated
plants are supposed to have arisen ; but it was until recently considered to be
a process peculiar to domesticated races, and the Darwinian idea of slow isola-
tion as the result of extinction was the generally accepted hypothetical process
in nature. The work of De Vries, however, shows conclusively that in plants
the rapid development of new forms takes place also in nature, and it is now
an established fact that it is not confined to domesticated races, but is common
to both wild and cultivated plants. Among animals, however, the state of our
information is far less satisfactory, and it is decidedly an open question as to
whether the origin of species in animals has been by slow modification and
extinction, rapid transmutation, or by "mutation."
The attack upon this problem in animals is far more difficult than in plants,
the obstacles to be overcome and the apparatus and expense demanded being
so excessive that but poor headway will be made as compared with that of
the plant evolutionist. Moreover, the problem is more complex in animals
than in plants, and the ramifications, whose solutions are necessary before
we can arrive at a really safe hypothesis, are numerous. The attack must
be made from two sources : First, from careful observation in nature of the
material under investigation, that the observer may become thoroughly
259
26o PRODUCTION OF RACES AND SPECIES IN LEPTINOTARSA.
acquainted with his material in all its natural phases ; and second, by long-
continued experimental investigation. In the study of evolution in Leptino-
tarsa I have carried on the two lines of investigation stated. The first line of
inquiry has been presented in the first chapters of this paper. In this chapter,
in which will be taken up the second line of attack, it will not be out of place
to review briefly and bring together the general results of the first inquiry,
since it is these conclusions which form the starting-point for the experi-
mental portion of this research.
The genus Leptinotarsa everywhere lives in situations where the earth is
covered with perennial grasses, with a greater or less growth of herbaceous
plants, some of which must be Solanacese, upon which the beetles feed. They
are typical grassland forms. This fact explains their general distribution
over the American continent — their presence in some parts and absence in
others. To the conditions of existence of a grassland habitat they are singu-
larly adapted. The general climatic conditions of grasslands are frequent
although often small precipitation during the growing season, which keeps
the superficial soil moist and the lower stratum of air humid through evap-
oration, followed by a dry season of greater or less duration. I have shown
how the beetles are limited by these very factors of soil and atmospheric
moisture in their general and local distribution, and how their life histories,
reproduction, aestivation, and physiology are adapted to these conditions.
But the grasslands are not all alike; some are more, others less humid;
some are warm, others cold ; hence a particular trend in species evolution in
the genus shows closely related species occupying adjoining but differing
environmental complexes; and we inquire whether these species have come
about by slow modification and extinction or by sudden transformation. We
have examined the variations, fluctuating, place, and geographical, and get
no satisfactory evidence as to the method of evolution, too much supposition
being required to connect variations with evolution. This study of distribu-
tion and variation gives no information concerning the method of evolution,
but it shows conclusively, first, that species formation has been definite, the
successive and related species having developed in contiguous habitats as the
genus evolved ; and, second, that variations are in the same direction as spe-
cies formation. Evolution has been definite and not promiscuous; it is
orthogenetic.
Throughout there appears a growing body of evidence tending to show
that selection does not create nor even preserve extremes of variation, but
that it acts as a conservative factor to eliminate extremes and to confine the
species close to the mode. Correspondingly, there is found in increasing
volume evidence that new races and characters develop rapidly, but always in
conformity with the general orthogenetic trend of evolution of the group or
genus.
PEDIGREE CULTURES. 26l
Two questions which can he solved only by experiment are before us: (i)
Do new characters, races, and species arise by rapid development only, or do
they arise by both rapid development and by slow variation and isolation
through extinction? (2) What is the cause of these transformations of char-
acters and species ? is it external or internal, or both ? and can they be pro-
duced in experiment ?
PEDIGREE BREEDING OF L. DECEMLINEATA.
EXPERIMENTS IN RACE PRODUCTION AND MODIFICATION BY
ARTIFICIAL SELECTION.
From 1896 to 1904 I carried on continuous and rather extensive experi-
ments in selection. In these experiments the beetles were kept under con-
stant conditions, especially during ontogeny, and in each generation indi-
viduals which showed extreme variations were isolated and became the
parents of following generations. This is a comparatively simple process,
and one which can without doubt be kept up for a long series of generations.
It is these experiments which will now be considered.
I began with the idea of showing experimentally that where there are
closely related species similar in most of their characters, but isolated by
gaps, evolution has been through slow progressive modal shifting and the
extinction of intermediate forms by selective processes. Thus, for example,
in the series multitaniata, intermedia, and decemlineata, multitceniata is most
melanic in general coloration, and decemlineata least. It is possible, by envi-
ronmental influence and selection, to change species one into the other?
Selection Experiments with Color Characters.
My experiments with color characters dealt both with special areas of color
and with general color tendencies, such as melanism, albinism, xanthism, and
rufism.
I attempted by selection to create a race of decemlineata in which the spots
a and a' upon the pronotum would always be fused posteriorly, as they are in
some species, and at the same time to increase their size. In this experiment
the most rigorous selection was practiced, and every opportunity given for
selection to produce the desired result, but to no purpose. In one part of this
experiment the selection was carried out in a most rigorous way for eleven
generations without any discoverable result (plate 25, A). In each genera-
tion the progeny of extreme melanic parents presented, as far as the character
in question was concerned, about the normal range of variation. The entire
negative result in this experiment may be interpreted as an example of the
total inefficiency of selection, or, what is more probable, as indicating that the
selected variations of the character were not capable of transmission, being
purely fluctuating somatic variations. In the chapter on coloration I have
262 PRODUCTION OF RACES AND SPECIES IN LEPTINOTARSA.
shown that color variations produced by environmental conditions during the
larval and pupal periods are not inherited at all.
In another part of this experiment a single pair of beetles gave off-
spring of two sorts — one with the extreme conditions of the parents, the other
with the conditions of the first set. The parents of this series were from
the same stock as that which gave entire negative results, and were indistin-
guishable from several other like pairs that were not able to transmit their
peculiarity to the next generation. There is, therefore, a difference in the
variations experimented with, one being inheritable, the other not. The rela-
tionship and behavior of these two series I have represented on plate 25. In
this plate, in which the curves of the selected parents and offspring are
brought together, it is clearly demonstrated in the series A that there is no
accumulation of the character through selection. The series B, however,
which arises from a single pair indistinguishable from the rest of the parents
of generation 5, shows two distinct groups of offspring. These were isolated
and reared in several succeeding generations. One of these groups (B') was
always strongly like the parents, and showed little or no tendency to revert to
the racial standard as did the individuals of B" . In the pair that were the
parents of series B, one or both had the selected variation of transmissible
quality, and passed it on to the following generation, but to part only. It
would appear, therefore, either that one parent had the character selected
transmissible and the other not, and that part of the offspring received the
transmitted character in inheritable form and part did not, or it is possible
that both parents had the character latent in some germ cells and not in others,
and that in reproduction there was a union of like germs. I believe that the
first proposition is the correct one ; but which parent possessed the character
in transmissible form I do not know in this case, and for this series of experi-
ments this is of little importance.
A glance at plate 25 shows that the series of cultures B' are very different
from A or B" in that in B' selection was able to keep offspring up to the
parental mode, and to raise the mode of the race. In the others, however,
i. e., B" and A, this was not possible, nor was I able to maintain the same,
nor approximately the same, modal standard for parents and offspring.
In B', however, I was not able to carry the selective development of the race
beyond the usual range of variations of the species ; that is, within the num-
ber of generations experimented upon in B' , and the same result was obtained
in six other identical series. The limits of variability of the species seem to
be limits beyond which artificial selective processes are unable to carry the
modification of the race. On the other hand, by selection we can rapidly
create races up to the natural limit of variation of the parent form, and main-
tain them there with ease by the same process, and I can see no reason why
artificial selection should not be able to carry the species beyond its natural
' Normal " range of variation.
Mode.
'Normal" range of variation.
Mode.
SELECTION EXPERIMENTS. 263
limits of variation, if selection is the powerful evolutionary factor its advo-
cates hold it to be. It is possible that my experiments were not conducted
over a sufficient number of generations to bring about greater changes.
From this series of experiments it is evident that in the character chosen
some variations are transmitted, others not; and, as far as my experience
goes, there is no middle ground. To all appearance these variations are
alike, and are distinguishable only by their behavior in pedigree breeding.
The results of the above experiment are paralleled by similar conditions
found in nature. Thus, from selected pairs of parents found in nature, only
about 4 per cent pass on their extreme characters to their progeny, showing
how small a percentage of the variations are transmissible. In nature, where
crossing is going on, such heritable variations are swamped as soon as they
arise — a fact that I have proven experimentally by allowing such variations
to breed freely with the general population. It is only when we practice
artificial selection or isolation that they are preserved. With such charac-
ters as the spots used in the above series of experiments, which are absolutely
neutral and could not by any conceivable process come within the domain
of selection, and are yet good, specific characters, it is certain that the
character as it exists in the different species could never have come about as
the result of selection and extinction. It is possible that it may be correlated
with some other characters of direct selective value, but I have not been able
to establish any such connection. As far as the experiments go, the variations
are either transmissible or not. and there appears to be no intermediate state
of partial or weak inheritance. Experiments were also tried with other
spots on the pronotum, c, d, and e, and with spots upon the epicranium and
upon the legs and ventral surface of the abdominal segments, with exactly the
same results.
When we combine variations and treat them statistically, we get results
that are false. Thus, on the basis of statistical method the variations of
B, A, A', B' are continuous and of the same kind, while in reality they are
different ; and by no method of biometry could we discover the two kinds.
These can be discovered only by the observation of individual cases and by
breeding. When we examine the pairs possessing heritable variations, it is
found that in the population of any generation they stand apart, separated
from the mode of the species, and that intermediate states are absent. It is
true that when many of these individuals are brought together, "lumped,"
they form a regular polygon of distribution. The reason for this is that
while each individual represents a distinct deviation from the mode of the
species, these deviations are not all of the same size — some are large, others
small, and the latter are far more numerous than the former. Moreover, we
do not find many of the larger variations in any given generation. Consid-
ered statistically, we should recognize these as continuous variations; exam-
19— T
264 PRODUCTION OF RACES AND SPECIES IN LEPTINOTARSA.
ined singly, they are unmistakably seen to be deviations of greater or less size
from the mode of the species.
I have shown that these spots behave as units in the ontogeny and evolu-
tion of coloration. In these selection experiments, also, they behave as units.
If we combine large numbers of cases they follow Quetelet's law in the
polygon of distribution ; but this in no wise presents the real fact, namely,
that each heritable variation is a divergence from the species mode in some
individual line of descent.1 The distribution, in the long run, according to
the law of error, of these diverging heritable variations is not without interest
or importance, indicating, as it does, that the causes productive of the trans-
missible variations also diverge from their standard, now more, now less, and
in different directions; but whether there is continuity or not can not be
determined until we know the causes. In any event, the distribution of the
causes of heritable variations according to the law of error results in the
long run in producing heritable variations in dccemlineata that are also dis-
tributed according to the law of error, so that the species is thereby kept upon
an even basis.
The behavior of general color characters, such as melanism, albinism,
xanthism, and rufism has been studied, especially with the object of creating
races by the selection of variations through successive generations. Out of
the many sets of experiments tried, two will serve to illustrate the general
results obtained. I shall describe experiments for the production of albinism
and rufism.
I attempted by selection to create an albinic race of dccemlineata in two
ways — first, by selecting for breeding the most extreme albinic variations
found in nature, and, second, by creating extreme albinic conditions in experi-
ment and breeding from them. For the first set of experiments the selec-
tion was made from numbers of copulating pairs found in nature. The
selected pairs were kept in separate cages, as were their progeny, the only
lumping of material being in the statistical treatment of it. The great major-
ity of such pairs and their offspring were not of any interest. Out of 311
pairs selected and mated in the years 1 896-1904, only 26, or 8J per cent of the
total number of pairs tried, were found capable of transmitting their partic-
ular variations. In many of these pairs it was certain that only one of the
beetles had the character in transmissible form, so that in 311 pairs, or 622
1 That is, between any given variation and the species mode the intermediate stages
are a greater or less distance back in the line of individual descent. Moreover, in the
lines of individual descent intermediate stages are always discernible, no matter how
suddenly the variation may seem to appear. In all of the cases that I have studied there
are no jumps or discontinuity of any kind in the heritable variations, but there always
exist transition stages, passed frequently in ontogeny, but nevertheless real and directly
observable, between the characters of parent and offspring. There is then no saltation
or "mutation," either large or small, in Lcptinotarsa, nor are there "immutable" unit
characters.
range of variation.
__/
Mode.
' Normal " range of variation.
A
50
w
Z 2
* 2
z s
W -=
wg
« »
n a
2 w
z w
:s
« z
2 °
z o
g
«
-r
•7"
(n
.
O
3
~
n
^
a
-
w
p
-
>
so
Z
M
>■
H
M
>
Z
-i
S
O
-1
■fl
s
w
'/
>
O
Z
o
X
M
>
S!
z
so
i
>
-■■
f/1
o
-
■/
."
o
z
V
IV
III
II
I
^J
I
J
\
\
i
J
c
J
^
^
i
!
>
\
s
f
1
I
1 1 — ^
J^
n
- ~T — r^ft.
VI
V
IV
III
II
I
L**k. ^
^
1
I— 4«s v
XJ
1
v
,
\
1
\
,/
' I ^
\
y
R
\
\
\
k /
N
^
VII
'Normal" range of variation.
PLATE 27
•■: i It
s
■z
>
- n -:
1 > to
= t
n > O
=■ - u
■u .- -:
a « >
« o 3
£ - -
« _• -
S s:' m
n > Z
a 2 .
m
n
z
z
£
^
-
i*i|i
2
u
1
yk
C
*
M
1 1 ,' '
ll 1.*
'
i I
|
}
Ml
D
/
nil
Mil
III
i 1
r
i
4
fc
3
l{
3
r ^
_i
\
~ i
>A"
: f
A
ft
'"'•V
%
^
^
-;
lis
s
1
T
i
l
i
sir
T
"Normal" rauge of variation.
A
«
«
«
SS
O
as
o
•a
■-J
K
«
«
XT.
C
f>
-J
«
«
n
«
o
z
o
>
a
z
D
K
pi
>
z
►
o
M
to
U
«
n
3
z
^L
gt ^f\ £E3T
Je=^ JAi 3\J /— H1~
1=1 ^^--:-r ;A £ — 3
e a^- •■- ~--T-.'~~t-z-~z.^/£ 3
£ t^\r J-r-
_a_ £^_ _/^r
civ ^""^^3 &^*
J Jt %^\ * i
m Js^la ^-^ ^^_
m^---?--j S-?S /3T
A ij £ " =-r F:K ^ - 5: :":_: - s 1
~" ""ft } ^ s fr 7
Jk / \ ^k_
Rf=\ \ ~~I — 3~
_,^=f=^ \gn|
^ | "' =s>. ^ :^
U *"£
m\
-Jl= yg=3
^=f=fs risrK
j — — ==v ^ a
~L t
Ai 7 a
_^gj j/T-jv
jj=T=1 Bi==\
^==j==^ ^ — = — w
It K
t ' £X
fi^^ 7=it=i
■■K 2 \
— C ^L
_se=v ~2— 5
/ 1 j\ ^aS
J==f : r**^ ^ = = 3^
It t
.si ^y.
— _^s ^-3^r ~
H^---fe-^ -_l
13 jMi 2^jl
I pS 2— — £T
>c! — — ^ ^" 3^l_
~2=^=^Ssi ■ — =^
JX ^=* -2-
_jj /===SlI ^
^^ — = = = V 1
^^===li==^J
^=^===^i jp===^^l
W. L. TOWER
PLATE 29
" Norma
F
'ang
5 of Vari;
h
ition
"
^
x-
*>
\
i
N
r
\
t-2
i
I
7^
/
^
VIII-
r
^5"
^
^^_
f
|
_C
i_ ■
c
$
,
\
»
^
T
,q
VII'
,
m
n
f
1
■
K
-,
r
l
R
J
^
\
>
s
VI ■
'•-
\
-
> ^ =
^
/
h_
CI
1
^
1
L
<
^7
r
V ■
i ^
v.
1
_4
"f
\
^,-
IV-
*- ^-" p
1
D
*
>
St
-
Ss
(.
iii-
y
J^
&
i "
-
- ■ -
z- =1
"%
^
A
y
/)
:
a
t
|
ii •
4
~
_-Jn
/
-S.
•
A
-\
"*,
B
/
7"
-1-
-A,
1
\
1
i -
A
V
t I
/
1
A
«*
1
ii
Classes 23 22 21 20 19 18 17 16 15 14 13 12 II 10 9 8 7 G 5 4 3 2 I a b. c d e. f g. h.
DIAGRAMMATIC REPRESENTATION OF THE RESULTS OBTAINED IN THE ATTEMPT TO
CREATE ALBINIC AND MELANIC RACES BY THE COMBINED INFLUENCE
OF SELECTIVE AND ENVIRONMENTAL STIMULI
SELECTION EXPERIMENTS. 265
individuals, the actual percentage of specimens showing heritable variations
was probably not far from 4 or 5 per cent.
In plate 26 I have shown the results obtained from three experiments, A, B,
and C, each starting from a single pair of beetles where extreme and non-
transmissible variations were the parents of each generation and were selected
most rigorously, all others being exterminated. In these experiments with
general color characters there was no influence exerted upon the progeny by
selection of the parents. In general color characters exactly the same results
were obtained as in the selection of unit color characters. In plate 27, how-
ever, are seen quite different results from those portrayed in plates 25 and 26.
Starting from a single pair of albinic beetles in which the selected character
was transmissible, and following the line of descent from generation to gener-
ation, the fact is graphically shown that the particular variation of the parent
was not only preserved, but carried close to the limit of normal variability
of the species, and that by selection the race was changed from one which
was variable to one which was relatively invariable — that is, selection resulted
in the production of a race of albinic beetles of low variability, which, no
doubt, it would have been easy to maintain for a long period of time. From
the third generation a selection was made for the parents of the fourth
of the most and least albinic individuals, A still being the albinic race, and B
the divergent race tending toward the opposite extreme. These two lots of
parents gave in the fourth generation two distinct polygons which overlapped
only in the slightest extent. By continued selection the polygons in the fifth
generation did not overlap, and in this generation further division was made
of B into B and C. These two lines were continued for several generations,
diverging from the A line, but not far nor rapidly. In the second generation
there arose two distinct groups separated by a wide gap, A and D, the latter
being the exact opposite of the A race. This D race was propagated, and by
selection produced the result shown in the polygons along the line of descent
D, giving in the last generation of the race a group of beetles of almost uni-
form condition. In all the lines of descent, A, B, C, and D, artificial selection
did just what it was found to do in the elements of coloration, namely, it
created a race of low variability about the standard chosen, which it main-
tained as long as selection was practiced ; but it did not carry the race beyond
the normal range of variation of the species. That is, artificial selection can,
as De Vries points out, produce races and maintain them, but its power to
develop these races beyond the natural range of variability is yet to be demon-
strated.
From the series of cultures represented in plate 27 it is shown that it is
easy by selection to create races from a species, which would, as long as the
artificial selection lasted, breed true to the ideal chosen. Such an experiment
was made and two races breeding true were produced. Their history is rep-
266 PRODUCTION OF RACES AND SPECIES IN LEPTINOTARSA.
resented in plate 28. From the parent generation two selected groups, one
melanic (B), the other albinic (A), were taken, and from these, two clearly
defined races without trace of intermediate condition were produced. During
each of eight consecutive generations slightly variable, light and dark races
were maintained. At the end of this time the material was divided and selec-
tion was stopped in one group and continued in the other, but the lots were
not allowed to interbreed. The removal of the selective factor at once
resulted in a regressive shifting of the mode of each unselected race and in
increased variability, and this change continued through the eleventh genera-
tion, when both unselected lots had moved back to the mode of the species.
These experiments with color characters show very clearly that artificial
selection is with transmissible variations a powerful factor and can greatly
accentuate any character and maintain it in an extreme condition, but that
there are limits beyond which I was not able to modify the characters by this
agency. The experiments also show that artificial selection works rapidly,
and not, as has been so often assumed, with extreme slowness. True, in
experiment I practiced a most rigorous selection, but not more rigorous than
that which the natural selectionists believe exists in nature. The general
feature of these experiments I shall postpone discussing until after the data
from other selection experiments have been considered.
The experimental production of general color variations and their preserva-
tion by selective breeding give many points of interest. In this I have con-
fined my attention almost entirely to extreme light and dark forms. To pro-
duce light forms I have used hot and dry conditions, and for dark forms,
warm and moist. The conditions productive of these variations have already
been discussed and recorded (Chapter III). The experiments herein recorded
differ from those already given in that the entire life of the beetles was passed
in the conditions of experiment, and not the larval and pupal stages alone, as
in the experiments upon coloration.
In plate 29 are brought together in diagrammatic form the data and general
history of cultures where both light and dark forms were produced and
further subjected to experiment. The black polygons represent the selected
groups of parents, the ruled polygons the offspring. The appearance of
"mutants" beyond the normal range of variability is indicated by the small
white polygons, pa = L. Pallida, with its range of variation ; de = defccto-
puiictata; m = L. melanicum. The series is a complex one, involving pro-
cesses other than artificial selection and introducing factors of interest which
are the key to further experimental study. At present we shall consider
only that portion directly concerned with selection or selective processes.
In the first, or parent, generation I selected 6 copulating pairs of beetles
from the hibernating population, and kept them and their progeny in natural
conditions. From the 6 pairs were obtained in the second generation 1,320
SELECTION EXPERIMENTS. 267
mature beetles, and from these, two groups of copulating pairs of 10 each
(A and B) were selected and reared in the third generation, but showed no
modifications as the result of selection. These hibernated, and selections
from each lot were reared in the fourth generation, but showed no modifica-
tion. I now felt sure that the material was pure, that is, normal, and carried
no tendencies to appear in divergent extreme variations. Accordingly, from
the two series selection was made of as nearly modal individuals as possible,
and the two selected lots were mixed and divided into two lots of 10 pairs
each, C and D. These were placed, as soon as possible after emerging, in
surroundings productive of dark and light conditions of coloration, and
allowed to breed, producing in the fifth generation two distinct lots of
descendants, one light, the other dark. These hibernated, and after emerging
in the following spring were allowed to breed, when it was found that out of
50 mated pairs 31, or 62 per cent, were able to transmit their particular varia-
tion in full strength, a huge increase over that found in selections from
nature. From each group 5 pairs were selected as the parents of the sixth
generation. These gave, as was expected, distinct lots of individuals more .
melanic and more albinic than their parents, and each also produced indi-
viduals differing in many respects from the parent stock, and beyond the
usual range of variability. In the five following generations the same thing
was repeated, as may be seen from plate 29; that is, from each group of
selected parents there came a general population less and less variable, and a
greater or less number of highly divergent forms beyond the normal range of
variability of the species. These latter we shall consider in another place.
In this series of cultures a normal parent stock has been subjected to
artificial selection aided by powerful environmental stimuli, both having the
production of the same end in view. The results, however, were a keen dis-
appointment ; the inability to produce in this experiment by selection and
powerful environmental influences a race much beyond the normal limits of
variability of the species might easily be taken to indicate the impotency of
selection. The ease with which the beetles moved back toward the species
mode when selection was no longer practiced and the conditions of existence
became modal, when joined to the data of place and geographical variations,
allows only of the conclusion that while differently colored races and modifica-
tions of this beetle occur in nature and are produced in experiment by artificial
selective processes and local environmental influences, such modifications are
limited by the natural limits of variation of the species and persists only as
long as the maintaining processes are present, and are utterly incapable of
existence under adverse or even natural conditions, reverting to the species
type. That is, artificial selections or local influences are able to modify, and
to a certain extent create races founded upon those variations which are
ordinarily killed off by natural selection (Chapter IV) ; but in the creation
268 PRODUCTION OF RACES AND SPECIES IN LEPTINOTARSA.
of such races we really have two forces— a species selective tendency and a
local (or artificial in experiment) — acting against one another, with the
result that selective divergence to a certain limit is attained, but beyond that
the racial divergence is slow or entirely stopped. When the local or artificial
selection is removed the species selective tendency causes a regression to the
type of the species. It may be objected that my experiments do not cover a
sufficiently long series of generations to have accomplished the result
intended, and this may be true ; but selection is a powerful formative factor
and works rapidly up to a certain limit, and this has been abundantly proven
by plant and animal breeding for fifty years. Why should it not also be
able to establish a race on permanent footing with the same rapidity? We
know from the rearing of domestic animals and plants that constant selec-
tion is necessary to maintain the race. In as far as these color characters
are concerned, by artificial selection we can easily produce and maintain a
race, but we can not establish it as an independent one ; we can create
isolated races from extreme variations, and by selection keep them isolated,
but we can not permanently establish them. On the whole, selection would
appear as a relatively impotent factor in the evolution of these color charac-
ters were it not for the fact that we usually try in our experiments to zvork
against the far more powerful natural selective tendencies of the parent
species. Two points we may note in passing, namely, the number of highly
divergent variations beyond the normal range of fluctuating variation pro-
duced in this series of experiments, and the increased percentage of indi-
viduals which show variations capable of being transmitted to the progeny.
Selection Experiments with Structural Characters.
It is often asserted that color characters are not reliable for the study of
variation and evolution. The assertion is, I think, due entirely to ignorance
of the real nature of color characters. I have made a series of selection
experiments with structural characters — general, such as size and shape;
and special, such as glands, punctation, spines, etc.
There exists in this species a very considerable variation in size between the
sexes, and also in the same sex, and this provides a good general structural
character for the experimental study of artificial selection and selective proc-
esses. Four main sets of experiments were tried, each aiming to create by
selection a race having certain characters. Four conditions were aimed at:
(a) To have both sexes large; (b) both sexes small; (c) females large and
males small; (d) and males large and females small. Only the first two sets
gave results of any value, the last two failing because of the difficulty of
breeding opposite extremes. Therefore I give here the data and conclusion
from the first two series. The general results of these experiments are given
in plate 30, the procedure and results being in general not unlike the experi-
ments with the color characters.
•Normal" range of variation.
PLATE 30
Mode.
SELECTION EXPERIMENTS. 269
I started in these experiments with 10 pairs of modal beetles that had hiber-
nated, and bred them, getting a second generation, normal in every respect,
and from these selected 10 modal pairs and obtained a third normal genera-
tion of 1,288 beetles. From this third generation, out of no pairs 2 were
found that had variations capable of transmission, giving two divergent
groups of beetles in the fourth generation. One of these consisted of beetles
larger on the average than the parents and the other beetles smaller than the
parents. From the fifth to the tenth generations I practiced constant rigorous
selection, and in the eighth there was a relaxation in the selection of part of
the large beetles, with the general result shown in plate 30. It is possible
to increase and decrease the size, but the mode of any one generation could not
be brought to the limits of the variability of the species. This result, with the
general structural character, size, differs in no way from that obtained with
color characters, nor from the common experience of breeders of domestic
animals and plants. The breeder of large horses has not yet by selection
been able to create dray horses as large as elephants, nor is it possible to
double the size of these beetles by the same process. The change produced
in size was, if selection was removed, as shown in plate 30, immediately oblit-
erated through regression to the standard of the race.
An interesting set of experiments with the scent glands was made. I have
shown how these glands act as protective structures, and how distasteful their
secretions are to insectivorous birds, reptiles, and amphibia. They are there-
fore highly useful to the beetles, serving as active protection, hence of real
utility, and they ought by selection to be capable of increase in number and
in quantity of secretion, a result that could not be otherwise than a decided
advantage to the race, since it is shown that the species most numerous,
widely distributed, and immune from attack are those with the largest number
of these glands.
All that the experiments with the glands show is that, given a variation
capable of being inherited, with it as a basis we can by selection create a race
that possesses as its modal condition very nearly the extreme variations in the
characters chosen for experiment ; but the race persists only as long as the
stimulus of selection is continued, and I do not see how this process could lead
to the establishment of independent races or species.
While the selection experiments with structural characters show that the
same results are obtained as with color characters, they are strong evidence
against the assumed ease with which useful characters can be augmented and
preserved by selection. They, however, show clearly that in these beetles
even a most useful protective device is not easily modified in a permanent
manner by artificial selective processes. In this series of cultures we are not
dealing with uncertain and unknown quantities. We- know the direct and
important use to which these glands are put by the beetles ; we know that in
the genus some beetles possess more, others less ; and that those species with
2/0 PRODUCTION OF RACES AND SPECIES IN LEPTINOTARSA.
the greatest numerical strength and extensive distribution are provided with
numerous protective glands. Moreover, we know from the experiments that
the parents were normal, and that the selected parents possessed the varia-
tions in transmissible form. In the culture I acted as the insectivorous
enemy, eliminating all but those with the most glands, which were allowed to
breed. In this experiment were combined the conditions ideal for selective
modification, but the results in no wise differed from those obtained from
experiments with characters of no utility to the species. The similarity
between the results obtained in characters of utility and characters of no
utility is most interesting, and admits of only one conclusion — that as far as
the experiments with these beetles go local or artificial selection is a weak
factor, able to produce little or no real change in the species, and unable to
create a permanent race by the preservation of extreme variations.
Selection Experiments with Physiological Characters.
During the years 1899-1904 various experimental pedigree cultures were
made in which physiological characters, such as habits and instinctive
responses, were chosen for experimentation. These characters were found to
be only in the slightest degree influenced by selection, and the changes found
were too small to be of any real value or afford a basis for conclusions. The
failure to get positive results was not due to the non-heritable qualities of the
characters, but to the failure to get hold of sufficiently extreme variations to
form a basis for selective experiments. All of the variations used were so
near the mode that they afforded no real basis for work, and I was not able
by selection to increase such variations as were found.
It has been shown in respect to these physiological characters that they
were exceedingly conservative. Thus, in the various habits associated with
reproduction or hibernation, only the modal individuals are able to go
through the habit with success, and a variation of any moment is sufficient to
exclude the possessor thereof from participation in the propagation of the
race. As a matter of fact, variations of these characters are rare, and the
species are separated by wide divergences, far larger than those existing in
the case of morphological characters.
When we bring together the results of the selection experiments upon
L. decemlineata several points of general interest are brought to light. First
and of fundamental importance is the fact that all of the heritable variations
are discontinuous in any given generation, and that continuity exists only in
individual lines of descent where discontinuity is impossible. The second
point of interest is the failure to carry, by selective processes, races beyond
the normal range of variation of the species. The third is the sudden devel-
opment of extreme variations differing in many points from the parents, and
lying beyond the range of fluctuating variations. These were found with
especial frequency in the series of experiments with general color tendencies.
APPEARANCE OF NEW CHARACTERS. 27I
In as far as we can judge from these experiments, while the selection of
the inheritable variations when continued for ten or twelve generations is able
to give well-defined races which persist as long as selection is practiced, it
is not able to establish these races so that they can exist alone. The results
from the experiments fall in line with the general experience of breeders of
domesticated plants and animals. In these beetles, as De Vries also points
out in plants, selection is able to produce races by isolation only (selection)
of extreme heritable individual variations.
I am aware that against this view the selectionists will at once say that far
more generations are needed, which I grant is possibly true ; but how many ?
I have shown that geographical races of this beetle living in places where
there have been the same environmental and local selective tendencies for at
least 400 or 500 generations have not yet become permanently modified in
form in such locations. It is true that a large part of geographical variation
is purely somatic, but not all ; hence, if local selective influences are so power-
ful, it would seem that a series of several hundred generations were ample
time for selection to begin to show its action. In experimental breeding
selection works rapidly, and in nature why should it not act with equal rapid-
ity— that is, where it is operative at all ? In any event, until there is evidence
to the contrary it will not be unfair to hold that pedigree cultures of ten to
fifteen generations do show fairly well how potent local or artificial selection
really is as a means of modification in these beetles. We must not confuse
local or artificial selection (isolation) with the specific selection characteristic
of the entire species. The former is productive of divergence, the latter of
conservation, and my experiments are in reality a contest between the two
with the conservative one victor in the end.
EXPERIMENTAL PEDIGREE BREEDING OF NEW CHARACTERS AND SPECIES.
There have arisen in my experiments, and I have found in nature, varia-
tions of L. decemlineata which differed from the parent stocks in one or
many characters and stood beyond the limits of fluctuating variations. These
sports — saltatory variations, discontinuous variations, or "mutations," what-
ever we may call them- — often bred true, and were able in many instances, but
not in all, to hand on their variations in full intensity. Whenever found
these interesting variations have been seized upon and made the basis of
instructive experimental cultures.
It has been shown that in nature three well-defined species of Leptinotarsa —
melanothorax, rubicunda, and angustovittata — arise through rapid transfor-
mation, and that rubicunda is possibly becoming established as a permanent
species. In treating of variations I have figured under the head of extreme
variations some of those observed to have arisen from decemlineata.
272 PRODUCTION OF RACES AND SPECIES IN LEPTINOTARSA.
The occurrence in nature of these variations from dcccmlineata is, as a
rule, rare. Owing to the abundance and concentration of dcccmlineata upon
potato fields, it is possible to examine huge series of these beetles. Some
idea of the infrequency of sports under normal conditions of existence can be
conveyed by the following observations made at various localities and for
different broods and years :
West Bridgczvater, Massachusetts, 1895. — First brood: Examined from one field
13,210 beetles, no sports; from a second field, 11,841, 1 sport, form melanicum. Second
brood : From same field as first brood, 21,400 beetles, I sport of form melanicum.
Cold Spring Harbor, Long Island, 1S99. — First brood : Examined from one field
14,600 beetles, 2 sports — forms pallida, melanicum. Second brood: From same field,
13,500 beetles, no sports.
Cabin John Bridge, Maryland, 1900. — Second brood: From one field 11,792 beetles,
82 sports — forms melanicum, minuta, pallida, tortuosa.
McKcesport, Pennsylvania, 1900. — Second brood: 9,460 beetles, no sports.
Yellow Springs, Ohio, 1001. — First brood: 16,002 beetles, no sports. Second brood:
14.200 beetles, 17 sports — pallida, minuta, immaculothorax.
Chicago, Illinois, 1903. — Hibernating beetles from collections on lake beach, 10,109
beetles, no sports. First generation: 12,019 beetles, 2 sports — form pallida. Second
generation : 17,008 beetles, no sports. 1903, from same localities, 16,200 beetles of the
hibernating generation, I sport — form minuta. First generation: 13.100 beetles, no
sports. Second generation : 12,400 beetles, no sports.
San Antonio, Texas, 1904. — Second brood: 1,100 beetles, 12 sports — tortuosa, albida,
minuta.
From these records in nature we see the great rarity of these forms, and
taking all the counts, out of 207,891 beetles examined there were 118 sports,
or in the ratio of 1,761 to 1. This high ratio, however, is due to the lot from
Cabin John Bridge, where there were an enormous number of sports, due to
most unusual conditions of environment. If we remove these the remaining
196,099 show 36 sports, or in the ratio of 5,447 to 1. Taking all the available
data gathered from 1894 to 1904, I find that on the average about 1 beetle
in 6,000 is of the class which is designated as sports, discontinuous varia-
tions, or mutants.
These figures demonstrate two interesting points : First, the average rarity
of these variations, and the necessity of handling large numbers of specimens
to discover them; second, the great production of these under extreme and
sudden changes of environment as at Cabin John Bridge. I have already
described the same phenomenon in the section on variation, wherein I showed
that place variation, when extreme, was accompanied by a great increase in
the number of these variations produced, as compared with the normal ratio
of production.
The finding of these extreme and permanent variations in nature at the
right time for experimentation, and the successful breeding of them is neces-
sarily a matter of chance and open to great liability to failure. In my
experience these sports are in nature mostly found singly, and in culture must
RARITY OF MUTANTS.
273
be crossed with the parent stock, because similar variations of the other sex
are usually wanting. Cultures of these become, then, largely experiments
in crossing these variations with the parent species, as they must and will be
in nature.
The variations of this kind observed to arise from decemlineata and the
number of specimens are: of melanicum, 31 ; tortuosa, 3; minuta, 2; immac-
ulothorax, 2; pallida, 63; rubrivittata, 1; defectopunctata, 1; albida, 1, and
obscurata. 4. Some of these I have been able to propagate with success;
others, like defectopunctata and obscurata. I have been unable to cross suc-
cessfully with the parents. It is, however, certain that these variations are
a normal product found in nature produced in different ratios — a ratio seem-
ingly dependent upon fluctuations in the environment. The point is that
these variations are natural and not the result of domestication or confine-
ment, and it can not be charged that in this material their appearance is due
to cultivation, because the foregoing data are drawn directly from nature,
and the material for the following pedigree cultures came from the same
source. My material is pure, never having been subjected to the supposed
vicious influences of cultivation or domestication.
It should also be noted here that the same kind of variations were found
to arise in the selection experiments. These were especially numerous in the
experiments where I attempted to produce extreme racial conditions through
changed surroundings, accompanied by selection, and they are also numerous
in the other selection experiments as the races diverge from the mean
and mode of the parent species ; but they are not numerous in any case,
excepting where changed environmental complexes have been employed in
conjunction with selection. In many respects these results resemble De
Yries's experiments with CEnothcra. excepting that here we have evidence
converging from many sources and pointing strongly to the conclusion that
the production of these rare variations is in some way connected with varia-
tions in the environmental complexes. This point we shall investigate in a
subsequent section of this chapter.
One of the earliest observed variations of this class from decemlineata is
the form pallida, which was used in some of my earlier cultures. (See plate
16, fig. 7.) It is not a variety of decemlineata formed by the taking away of
characters, but it is a step forward in the line of evolution represented by
L. multitceniata. intermedia, and decemlineata — that is, pallida is the next
logical step in the orthogenetic evolution of the series of species. This spe-
cies has been found in the last ten years many times and in divers parts of the
United States, and is not limited to any one region.
At various times I have been able to secure 2 or 3 and once 6 specimens of
pallida at the same time, and I have made experimental cultures of these with
interesting results. In August, 1901, I found 6 specimens of pallida at Clif-
ton, Ohio. Four of these were males and 2 females ; all were well nourished,
274 PRODUCTION OF RACES AND SPfiCIES IN LEPTINOTARSA.
robust, healthy specimens, and were allowed to hibernate through the winter
in nature. In the following spring two males and one female emerged from
hibernation, the others not being able to pass through this ordeal. With this
material as parents, two sets of experimental cultures were started, one a
normal cross between male and female pallida, the other between a male pal-
lida and a modal female dcccmlincata.
The pure cultures of pallida gave in the first generation 30 adult male and
21 female pallida without any trace of reversion to the ancestral species. The
polygon of distribution as worked out for this generation is normal, without
trace of skewness, and is separated widely from that of the parent species.
From this generation I allowed 10 male and 14 female pallida to breed
together in one cage, and obtained from the culture 201 male and 282 female
pallida, again without trace of tendency to revert to the condition of the
parent species. This third generation hibernated, and were unfortunately
almost entirely killed off by freezing, so that only 3 male and 4 female
pallida emerged from hibernation. I allowed these adults to breed freely
together as in nature, and obtained a fourth generation of 76 male and 81
female pallida, all true to type. From these 20 males and 20 females, repre-
senting the entire range of the variations found in pallida, were allowed to
breed freely, giving a fifth generation of 292 male and 306 female pallida.
The entire lot hibernated, and again a freeze so reduced the number that only
9 males and 7 females emerged in the spring. These were bred together
freely, giving a sixth generation of no males and 146 females true to type.
From these I selected 40 males and 45 females, covering the entire range of
variations found, as parents of the seventh generation, which gave every
promise of abundant and typical pallida, when the experiment was abruptly
terminated in July, 1904, by the accident which brought all of my experi-
ments to an end.
In this series of cultures with pallida there is a striking constancy of form.
No selection of parents for each generation was made — only a reduction
because my quarters could not accommodate all. Moreover, free interbreed-
ing was allowed in each generation as in nature. Under these conditions
pallida showed no tendency to revert to the parental species nor any trace of
skewness of the polygons of distribution.
In the first generation 10 male and 12 female pallida were placed in a
tank with 15 male and 15 female dcccmlincata, and all allowed to breed
freely. Crossings of male pallida X female decemlincata and male dccem-
lineata X female pallida, as well as normal crossings, were observed, and the
proportion of normal crosses to abnormal was 7 to 1. From this mix-
ture, allowed to breed freely as would be the case in nature, I got in the
second generation, L. dcccmlincata, 131 male and 114 female; L. pallida,
133 male and 162 female. These hibernated, and there emerged in the
following May, L. decemlincata, 10 male and 18 female; L. pallida, 9 male
PEDIGREE BREEDING. 275
and 10 female. I allowed these to breed freely as before, and obtained in the
fourth generation, L. decemlineata, So male and 106 female; L. pallida, 190
male and 210 female.
I then transferred the entire lot to nature in a situation where I could keep
control of them.
In the fifth generation I gathered all the adults and found : L. decemline-
ata, 211 male and 209 female; L. pallida, 509 male and 540 female.
It was evident that pallida was more than holding its own in this experi-
ment. All were subsequently returned to the plot where the experiment
was being conducted to hibernate. The following winter was a hard one,
and the plot where the beetles were was frequently subjected to freezing and
thawing — conditions most unfavorable to hibernation. In the spring there
emerged, L. decemlineata, 6 male and 10 female; L. pallida, 14 male and 15
female. These I allowed to breed together upon a plot of potatoes supplied
for the purpose, and counted in the sixth generation, L. decemlineata, 314
male and 301 female; L. pallida, 819 male and 761 female. It is clear that
pallida was becoming well adapted to the condition of its environment. I
felt that further experiment with this form unconfined in nature was neither
safe nor desirable, and exterminated the entire lot in the sixth generation.
It would have been of great interest to have allowed this experiment to go on
and to have given the insects freedom to spread over the country, to see how
fast and how far they would have gone, but the nature of the species, its huge
capacity for food, and the plants attacked unfortunately made such an experi-
ment entirely out of the question.
At the beginning of these experiments with pallida a single male pallida
was crossed with 4 females of decemlineata, giving the first hybrid generation
with the decemlineata character dominant. From these beetles I selected
lots of 10 males and 10 females, and got offspring as follows: (A) L. pallida,
21 males and 10 females = R; (B) L. decemlineata, 58 male and 51 female =
D + 2 Dr.
These lots w:ere now reared in succeeding generations.
Lot A. — These hibernated, and there emerged of pallida 3 males and 2
females, which were bred, and gave a fourth generation which consisted of
L. pallida, 21 male and 10 female, without traces of intermediate or decem-
lineata characters. These were bred for the fifth generation, giving of L.
pallida 41 male and 40 female. This generation hibernated, and thereafter
emerged and were bred in cultures, giving in the sixth generation of L. pal-
lida 82 male and 61 female, with no trace of decemlineata. This series
was in the seventh generation exterminated by an accident in late larval and
pupal stages. It is evident, however, that the material of lot A was either
pure pallida or that pallida was dominant to the extent of submerging and
keeping submerged any decemlineata characters.
276
PRODUCTION OF RACES AND SrECIES IN LEPTINOTARSA.
Lot B. — This lot, consisting of a mixture of hybrids and decemlineata from
the first hybrid generation (the D -\- 2 Dr), gave in the fourth lineal genera-
tion after hibernating, of L. decemlineata, 92 male and 91 female; L. pallida,
19 male and 31 female. These were separated and known as lot B 2 (de-
»i
/
Zen /
_1 /
2 '
UJ -
o<-o
UJc\j
Ocn
2?
<
a
_i
<
0-
G -i; ^
Si "C
> y
■«.
(*«
0
ex
<
/
6
H
/
Sj
u
H
2
f2
2
2
2
V
<
s
h^>
UJ _
0*0
UJ >o
UJ f^,
UJ —
hi
O
E
^ 5
Wo
UJ \-
«
6
0
oS
Q CO
Q CN
Oco
C
-
Z
0+
0
<2
0
a
a
E
1
Q 2
Q eg
Q in
0 ^
&
5
^
lit"
o*
---
_|
*o'
—
fc-j^5-
■d-5"
»
Zi
_ _
_i -
O
<^D^O
<<*o
< O
< 5
e, D. F.. Mex
II
rMULTlT/ENIATA
Eeared at Chicago, III.
rubTcunda
80 > JS£i2___
Hibernated.
*t.x 19
RUBICUNDA
MULTIT/ENIATA
'melanothorax
III .
•to,
61 9
1 $ crossed with ,
18 X, 219
'
'
MULTIT/ENIATA
1
MULTIT/ENIATA
HYBRIDS
RUBICUNDA
IV
50 , 48?
+8 rT, [Uj?
31 $, 3 9
killed
MELANOTHORAX HYBRIDS
killed
V
Transferred to Mexico:
6rJ, 59 16 J 89
DECEMLINEATA
18 , 22 9
17 r?, 20 9> C> 2 9;Emerse(1 ,rom *Btivation Jan. 2nd.
DECEMLINEATA
12c?, 149
DECEMLINEATA
3cT, t 79
DECEMLINEATA
15?, , 219
DECEMLINEATA
26 ?, I 329
DECEMLINEATA
40 ?, | 38 9
DECEMLINEATA
12?, I 119
DECEMLINEATA
10 (J, | 99
DECEMLINEATA
17c?, , 179
DECEMLINEATA
26. I 439
DECEMLINEATA
19?, I 219
DECEMLINEATA
48?, , 479 J
DECEMLINEATA
38,?, | 429
DECEMLINEATA
40?, J 359
DECEMLINEATA
18?, 32 9 J
Five successive geue-
rations of L. decern-
liueuta normal in ev-
ery respect excepting
'the number of gene-
rations iu each cycle.
Second cycle
Test-figure 26.
beetles of the lineata group, to which this beetle belongs, have more than two,
or rarely three, generations per year, and there are none known in the genus
that have over three. Clearly, then, this race with five generations in each
cycle is quite a new character in the genus, and was, as far as discovered,
290
PRODUCTION OF RACES AND SPECIES IN LEPTINOTARSA.
constant from the start, showing in the fourteen generations no tendency to
revert to the parental standard. As far as I can discover, this case can be
explained only as the direct response of the germ plasm to the extreme stimuli
used in the experiment. It seems impossible to account for the condition pro-
duced upon the basis of a latent character of this kind somewhere in the
ancestry of this genus. Moreover, all the beetles in this experiment had an
equal chance to be accelerated by the conditions of existence, so that this
factor could not by any stretch of the imagination be held to account for the
development of this change. As far as is known, the only changed factor in
this experiment was that used in the third generation upon the 7 males and 7
females for three-fifths of the reproduction period.1 From the germ cells
GENERATIONS,
I
II
III
IV
"V
VI
VII
VIII
IX
DECEMLINEATA
10 x 109
tr
DECEMLINEATA
48, 92$
+
DECEMLINEATA
98, 104?
10,? X 10O
\ —
DECEMLINEATA
80 J1, | 67?
(Parent stock.
After hibernation su
'Ji yo average fur : -
jected to temperature averaging n above normal, E. H.
reproductive cycle- % in normal condition.
I
DECEMLINEATA
74, 1 799
DECEMLINEATA
6'c?, 1 44$
DECEMLINEATA
48 £, J 79 9
DECEMLINEATA
88 I 190 9
UNDECIMLINEATA
Orizaba, V. C, Mexico, August, 1904,
74 $,
82$
Reared at Chicago and sestivated.
4 V Subjected to experiment.
UNDECIMLINEATA
38(?; 142?
UNDECIMLINEATA
60 , I 71 9
Continued at Chicago
as brood living free
upon plants in green-
house. Continued to
breed true without giv-
ing anymore augusto
vittata.
ANGUSTOVITTATA
eS, I 59
ANGUSTOVITTATA
| 23?
(^iativated lu February, takeu to Mexico in March.
ANGUSTOVITTATA
36,?, |349
ANGUSTOVITTATA
_92rT> 98 9 (N<™
activating. Jail., 1006.
Text-figure 30.
As far as was discovered in these cultures of L. undecimlineata the results
of the experiment in the stimulation of the growing germ cells were the same
as those obtained in the ohers experimented with, namely, permanent varia-
tions, which bred true to type.
CONCLUSION.
I have presented in these few pages some results obtained in a line of
investigation which is as yet in its infancy and which is being actively contin-
ued. Besides the experiments given, I have tried and am continuing other
experiments upon the same species, as well as upon other species of the genus
Leptinotarsa and species in closely related genera. Of this experimental work
I introduce here only my longest continued and clearest examples as part of
the argument for the method of evolution which I believe is followed in
insects. I shall present in some future report a full statement of these experi-
ments, and at that time shall be able to go more completely into the details
and theoretical aspects of the problem. Herein, however, I shall be content
to make a simple statement of the results, reserving theoretical consideration
PRODUCTION OF NEW SPECIES. 295
until such time as I have accumulated a sufficient body of observations to
warrant it.
A careful consideration of the various lines of experimentation recorded
and of the pedigree cultures and the data from observation in nature
irresistibly forces one to the conclusion that in these beetles the only
variations of permanence are germinal, and that the only possibility of
evolution is through germinal variations. Those germinal variations which
arise in nature are permanent, and the same variations, of the same degree
of permanence, are produced in experiment. The diverse kinds of evidence
produced in this and in preceding chapters all go to show that under varying
conditions of their surroundings these beetles vary, and that as they become
more and more extreme an increasing percentage of striking, permanent
variations is found; and as I have just shown, it is possible in experiment to
produce in this same way a variety of permanent modifications. From all
this evidence, however, there nowhere appears the least trace of a suggestion
of any specific action of the conditions of existence, but everywhere there
appears only the action of environment as a stimulus, while the response is
entirely determined by the organism. All of these variations of purely tem-
porary and of permanent kinds resolve themselves into responses of the
organism to the stimuli of its environment, but the nature of the response
is entirely determined within the organism. It is true that different inten-
sities of the same stimuli call forth different responses, but, as is shown in
the chapter on coloration, the response is entirely determined within the
organism, which is adjusted to different intensities of stimuli and reacts
according to its own method and on the basis of its own constitution, there
being no specific reaction called forth by a given stimulus.
I conclude, in the light of these experiments, that the production of herita-
ble variations, slight or extreme, represents in these beetles the response of
the germ plasm to stimuli. In my experiments these stimuli were external,
but there is no a priori reason why they might not also be internal. The
response, however, is absolutely determined within the organism.
Since the above statements were written, MacDougal (1906) has given a
preliminary account of his experiments with Raimannia, wherein salt solu-
tions were injected into the ovule just previous to fertilization, with the
result that variations, "potentially new species, were secured as a result of
chemical and osmotic action exerted on unfertilized ovules." These results,
he believes, "demonstrate conclusively that factors external to the proto-
plast may exert a profound influence upon its heredity characters, and call
out qualities not hitherto exhibited by the line of descent affected." These
results of McDougal's exactly confirm in plants the results that I have
obtained in these beetles, so that the point is now doubly certain that heritable
variations are produced as the direct response to external stimuli. In my
296 PRODUCTION OF RACES AND SPECIES IN LEPTINOTARSA.
own experiments I have used for three years past salt solutions as well as
electric stimuli, with the same results as are described in this paper for tem-
perature and moisture, and in all their action is simply as a stimulus, and
nothing more. The same is also probably true in MacDougal's plants. In
plants we shall undoubtedly find that the influence of salt solutions is far
greater than in animals, owing to the intimate relations of plants to the soil.
MacDougal admits that he has not been able to analyze the manner in
which this result is brought about in the germ plasm, and it seems to me a
waste of time and energy to speculate upon this point until we have some
direct investigation of this problem.
Of considerable importance is the strong evidence which points to the gen-
eral conclusion that these permanent variations arise during the growth
periods of the germ cells, and do not appear to arise before or after this
period. Just what this signifies is at present difficult to determine, further
experiments and cytological studies being necessary before we shall be able
to arrive at a more complete understanding of it.
CHAPTER VI.
THE PROBLEM OF THE ORIGIN OF SPECIES-DISCUSSION, SUMMARY, AND
CONCLUSION.
In the preceding chapters facts from observation and experiment have
been brought together and correlated within narrow fields, and these in turn
within broader fields, in the attempt to arrive at the more general axioms
concerning animal evolution as it is revealed through the study of Lcptiiw-
tarsa. This investigation began with the most general facts concerning these
beetles, their systematic position, distribution, and ecology, and from these
passed to more important and critical phases of their history. In this chap-
ter we shall attempt to bring together and correlate the accumulation of facts
and minor conclusions in the preceding chapters, that we may derive, if pos-
sible, more general axioms concerning the phenomena of evolution as discov-
ered in these beetles.
The slow evolution of the idea of evolution, from its crude beginnings to
its wonderful development within the past century, has not solved the riddle,
but only crystallized out the separate fundamental problems, which, although
interdependent, can for the present be best investigated as separate problems,
and the results combined into more general laws. The fundamental prob-
lems of animal evolution are those of —
(i) Heredity, the most important and most difficult cf solution.
(2) Variation, next to heredity the most widespread of organic phenomena.
(3) Method of evolution — that is, in what manner or by what methods
have species, genera, and larger systematic groups been brought into exist-
ence?
(4) Adaptation, or the adjustment of the animal to some one set of condi-
tions of existence.
In this paper we are largely concerned with the method of evolution — that
is, whether species come through variation and heredity by one or the other, or
all of the methods proposed during the last hundred years. In this chapter,
therefore, I shall consider the accumulated facts as they bear upon the prob-
lem of the method of the origin of species. The great factor, heredity, we
must accept as acting, as we know it does, without attempting to examine
into it at all. If we accept heredity as a process acting in nature, we shall
in no way invalidate any arguments or conclusions that we may enter into
concerning the origin of species. In thus broadly accepting heredity as at
work in evolution, I also include within this same acceptance that of the
297
298 PROBLEM OF THE ORIGIN OF SPECIES.
doctrine of the continuity of the germ plasm in its broad sense — i. c, the
genetic continuity of germinal material — as a true generalization, undemon-
strated, to be sure, but upheld by all of the best observational, experimental,
and theoretical work. The acceptance of the doctrine of the continuity of
the germ plasm, irrespective of how we conceive it to exist, is an absolutely
necessary accompaniment to the fact of the existence of heredity. With the
exception of Eimer and his followers, practically all biologists accept this
doctrine unreservedly in one form or another. Indeed, without it the most
ordinary facts of development and heredity become inexplicable, excepting
as miracles, and orderly evolution and development become inconceivable
processes.
THE METHOD OF EVOLUTION.
On the basis of our present knowledge the problem of the method of evolu-
tion presents two separate minor problems for consideration : ( 1 ) the origin,
or the cause and control of variations : (2) the preservation, or the action of
natural selection upon the new variations as they appear. These two
divisions of the main problem we shall here consider separately.
Variation and natural selection in their various aspects furnish the entire
solution to the problem of the method of evolution and of adaptation, and
could we arrive at a solution of the course and sequence of stages in varia-
tion and natural selection, we should have the basis for a complete solution
of the problem of the origin of species. Such a solution we can not hope to
arrive at, at least not for a long time to come, and probably ultimately the
best we can do is to push back to some impassable limit the description of the
sequence of stages, the primal stages remaining unknown.
The Nature, Cause, and Control of Variation.
The phenomenon of variation primarily owes its existence to the fact that
community of descent and heredity tend to produce the exact counterpart of
the parent organism ; the process of development, however, is not carried out
under absolutely constant or uniform conditions, but in a world wherein there
exist changing environmental states in endless perplexity. This results in
the turning aside of the line of development from the parental standard, per-
haps ever so little or only in one character ; but in this we have deviation or
variation.
While we can not hope to explain, excepting through philosophical specu-
lation, the origin of the phenomenon of variation, we do know the law or
sequence of events which variations follow. It has been abundantly dem-
onstrated in the physical and mathematical sciences that variations follow
precisely the law of error, and in organisms it is becoming more and more
certain that all variations follow most rigorously the same law. In organ-
NATURE AND CAUSE OF VARIATION. 299
isms, however, the conditions are immensely more complicated than in the
mathematical or physical sciences, and it will necessarily be a long time before
we can even begin in one single instance to follow the order of events in any
given variation in organisms.
In the explanation of the origin of variation in organisms the only assump-
tion we need make is that the original unit of organic matter was possessed
of the attributes which characterize organic matter to-day — motion, sensa-
tion, growth, and reproduction. This assumption can not meet with any
serious objection unless we change our ideas and definition of organic units.
Granted the existence of one single organic unit endowed as above, there is
no reason for introducing further complications by the explanation of phe-
nomena through undemonstrable hypotheses, because the fact of variation in
organic units can be explained solely through their existence in a natural
world surrounded by varying conditions of existence. The original organic
unit could not do otherwise than encounter changing conditions, or do other-
wise than react to the stimuli it encountered, even though it were so lowly
organized that all stimuli were perceived as one and reacted to by the sole
response of contraction. In this we have all the conditions requisite for
variation in full conformity with the law of trial and error.
The simplest existing organisms, greatly specialized, and the heredity
product of countless generations of similar simple organisms, react to stimuli
according to the method of trial and error. Thus Jennings says, regarding
the responses of the lowest organisms to stimuli :
In these creatures [Amcebae] the behavior is not, as a rule, on the tropism plan — a
set, forced method of reacting to each particular agent — but takes place in a much more
flexible, less directly machine-like way, by the method of trial and error. This method
involves many of the fundamental qualities which we find in the behavior of higher
animals, yet with the simplest possible basis in ways of action; a great portion of the
behavior consisting often of but one or two definite movements — movements that are
stereotyped when considered by themselves, but not stereotyped in their relation to the
environment.
It may be objected that Jennings is dealing entirely with sensory charac-
ters, reactions; yet wherein lies the real objection? Are not activities,
responses, behavior, indissolubly associated with structures ? Should we look
for one law of structure and another of activities? Undoubtedly structures
respond to stimuli. We know it to be a fact, and they follow in structural
variations the same law of trial and error. Were this not so, the observed
variations would not fall so truly into the probability of error distribution.
There does not exist, as far as I can discover, any real reason why organic
and inorganic variations are not the manifestation of the universal law of
error (trial and error) ; nor is there any necessity of introducing latencies or
predeterminations into the explanation of organic variations. Theoretically, at
300 PROBLEM OF THE ORIGIN OE SPECIES.
least, variation in organisms is to be explained from its start by the principle
of trial and error — trial due to stimuli, response according to the nature
of the organism. The response, however, is entirely conditioned within the
organism, and is manifestly a success or error. If now we are able to dis-
cover whether or not variations in complicated organisms are the responses
to stimuli, we shall be better able to judge whether organic variations are
explicable in conformity with the general law of variability of mathematical
and physical science, or as peculiar phenomena sufficient unto themselves
and needing undemonstrable hypotheses to explain their origin and workings.
In the second chapter there is brought together data concerning variations
in the genus Leptinotarsa, and the fact is demonstrated that variations in
these forms are distributed according to the law of the distribution of error,
as are the variations of all organisms thus far studied. In the third chapter,
where color characters are used as subjects, it is demonstrated that variation
is directly produced by stimuli — that from relatively invariable parents, stim-
uli produce variable offspring ; and again in the fifth chapter it is shown that
variations arise in direct response to stimuli. Similar facts exist in abundance
in the literature of the action of temperature and moisture on insects, of
pressure or friction, causing callouses and exostoces in Vertebrata, and of
changes of plant form through the stimuli of soil, light, and moisture. So an
endless recital of facts might be given to show that there is variation in
response to stimuli, and I have shown this to be true for these beetles. That
all variations are responses to stimuli will be admitted by some [neo-
Lamarckians], denied by others. That some variations may be responses to
stimuli, others not, will be the contention of the Weismannians.
In Leptinotarsa it is shown that there are variations which are inherit-
able, and others that are not ; and in Chapters III and V it is further shown
that the heritable variations as far as discovered arose in the germ plasm, the
non-inheritable during development in the soma. Further, the experiments of
Chapter V show that heritable variations arise as the response to stimuli
applied to the germ plasm. These responses, moreover, follow exactly the
law of error in their distribution about the normal mean. I maintain, there-
fore, that all organic variations are responses to stimuli, and are not due to
inherent tendencies or latencies, or the product of mystical elements.
The correctness of this conclusion regarding the nature of variations
will be immediately granted by many, but the supporters of Weismannism
will, I think, be prone to say that while the variations did result from the
stimuli used in the experiments, the variations were due to the existence of
ids, or determinants, carrying the variations in question, and were by the
stimuli employed brought forward, and so on. These elements, and the
Weismannian view that an "organ is undoubtedly predetermined in the germ
plasm," together with all its activities and variations, "can only" result in one
NATURE AND CAUSE OF VARIATION. 3OI
of two situations. Either the whole hypothesis of ids, determinants, and bio-
phores is a product of the imagination, or, if they do exist, further attempts
to solve the riddle of evolution will be futile.
In the experiments with Leptinotarsa new permanent variations arose as a
direct response to stimuli applied to the germ plasm (germ cells). In one
experiment, in the rise of a five-brooded race, there was a pure, perfectly
constant inheritable character arising as the response to stimuli applied to the
germ plasm. Eleven years of study of this and related genera have shown
that in none of the family, or relatives of the family, are there traces of -five-
brooded races or species. How, then, can we explain this on the basis of the
id-determinant-biophore hypothesis ? By supposing ids, determinants, or bio-
phores to have existed latent in the germ plasm from some extremely remote
ancestor who had a five-brooded condition, and then by germinal selection
lost it, a few stray ids of the five-brooded condition remaining latent for
countless generations, finally to appear in my experiments? If we accept
such an explanation science has no further excuse for its existence.
I believe that the rise of this five-brooded race is a direct experimental
demonstration of the falsity of the whole id-determinant-biophore hypothesis.
Possibly we could explain the origin of this race upon the supposition of
mutation, sudden jumps of ids or determinants from a two-brooded to a five-
brooded condition, but here we encounter the cherished hypothesis of germi-
nal selection, and obviously there was no time in this experiment for selection
to produce the results obtained. Moreover, the mutation hypothesis, with its
basic assumption of pangenes, each carrying "unit characters" absolutely iso-
lated from one another, affords only a superficial explanation. The most
logical interpretation is that this variation is a response of the germ plasm, in
changed chemical or physical constitution, due to stimuli which produce an
inheritable variation differing in no respect in its transmissibility and perma-
nency from the other variations produced.
\\ cismann admits that unicellular organisms may vary even in response to
external stimuli, which we know follow the general laws of variability, and I
can see no reason for a difference in kind of organic variability. Moreover,
my experiments with Leptinotarsa point clearly to the unity of variability
in its conformity to the general law of variability expressed in the method of
trial, with errors distributed according to the probability of error. Now,
while I believe that the law of variation is a general one, it does not and
never can explain anything. All that we can ever get out of this law of
variation, or any other, for that matter, is a more and more concise formula
whereby we may express our accumulated, systematized experiences con-
cerning variations. The best we can do is to describe and determine the rela-
tions and sequences of stages, and in the sense that an antecedent stage is the
cause of a following stage, we can determine "causes," but we shall never
302 PROBLEM OF THE ORIGIN OF SPECIES.
reach a point beyond which there will not be innumerable causative stages
which our perceptions can not discover. In this sense, then, the above law
of variation seeks only to express concisely and briefly the stages in their
general sequence as I perceive them.
Variation, then, in organisms resolves itself into trial and error. The trial
we can not conceive of as arising dc novo in the organism, but there must
have been an antecedent stage, which we call a stimulus, and it is shown
that when a stimulus is applied in experiment to organisms there is a trial
in that there follows a response (movement, modification) dependent upon
pre-existing stages in the constitution of the organism, and more remotely,
perhaps, upon the nature of the stimulus. Thus we see that even though we
can express the sequence of stages briefly by this law, and find it of use in
systematizing our perceptions of variation, in any single case, it simply states
what we may expect to happen when an organism meets the condition of
its existence.
This discussion brings us to the point where we see that variation is the
response or sequence of stages following the meeting of two preexisting
series of stages, one in the material world environment, the other within the
organism. Obviously, if we hope to state more clearly the sequence of stages
in variation phenomena, such as orthogenetic variations, fluctuating varia-
tions, skew variations (or "mutations"), we must understand more fully the
developmental principles underlying the antecedent stages within the organ-
ism before they meet the causes of the variations. It has been demonstrated
in my experiments that there is no specific response (variation) following any
given stimulus; hence I conclude (Chapter III) that the environment acts
simply as a stimulus, and, therefore, its sequence of stages is of no impor-
tance. It is shown (Chapters III and V) that the response (variation) follow-
ing stimulus is entirely determined within the organism. In the organism,
therefore, there are series of steps (stages) whose sequence and relation we
must understand before we can begin to get at the questions of orthogenetic
variation, "mutation," etc. This all-important series of states can be investi-
gated only through the study of development. Any other plan of attack
seems utterly hopeless.
The first and one of the most fundamental questions which we have to
consider is whether development is an epigenetic or a predetermined process ;
upon how we answer this question will depend the progress we shall be able
to make in the analysis of the phenomena of variation and evolution.
That development in Lcptinotarsa is an epigenetic process has been demon-
strated in the study of their embryonic development and coloration. In Chap-
ter III it was shown that there is no predetermination of the coloration, but a
development, unfolding, or specialization based upon a general phyletic system
NATURE AND CAUSE OE VARIATION. 303
of color centers of metameric origin and arrangement. This ancestral endow-
ment, which represents simply the repetition for an immense number of gener-
ations of the same series of stages, is used, as I have shown by all the species
in the genus, as the background of antecedent stages, upon which are further
worked out, as each group and species by its specific nature dictates, its type
of coloration. The same fact is true of the other systems of organs. The
question at issue is not so much the sequence of stages, because that all
will admit, but how we are to interpret this? We know that development
is a complicated, orderly, constant phenomenon, and the real question is
whether this orderly development is conditioned and controlled by material
units preexisting in the germ plasm, each endowed with all the attributes and
potentialities of the parent organisms, or whether it is due to the conditioning
of any given stage by the preceding stages. In other words, are we to con-
ceive of the germ plasm as a composite aggregation of an immense number of
individually endowed vital units, each in reality a unit organism, and this
complex society creating the individual (soma) for its use and habitation,
or shall we conceive of the germ plasm as an entity, as protoplasm differing
from other organic matter only in its conservative and generalized nature?
Unfortunately we can not decide this question in positive terms in the present
state of our knowledge.
Biologists seem to be nearly all of the opinion that the germ plasm is prob-
ably contained within the nucleus, and is largely, if not entirely, represented
by the chromatin of the nucleus. From this chromatin come, according to
recent researches, a considerable part of the cytoplasmic contents of the germ
cells ; also from the nuclei by chromatolysis are derived the internal secretions,
enzymes, cytolyzines, cellular toxins, etc., whose action directly controls and
conditions growth and nutrition. Even such characters as color are directly
initiated and controlled by the production of materials elaborated from the
chromatin of the nucleus, which react upon one another to produce, as in
these beetles, chitin and oxy-azo pigments. From the investigations in physio-
logical chemistry it is rapidly becoming more and more certain that all organic
products of the body are directly or indirectly the product of the chemical
activity of the nuclear material. At the present time there are known a great
variety of purely chemical processes and products originating from the chro-
matin, and upon the basis of the predetermination hypothesis they must be
predetermined also, and there must therefore exist within the germ plasm an
immense number of units, say "biophores" or "pangenes," to be later localized
in particular cells and to control the multifarious chemical processes due to
chromatic activity. These ultimate units, according to Weismann, "all pos-
sess the fundamental characters of life, dissimilation, assimilation, growth,
and multiplication by division. We must also ascribe to these in some degree
the power of movement and sensibility" (p. 369). Now, this is all very
22 — T
304 PROBLEM OF THE ORIGIN OF SPECIES.
pleasing, but what conditions or determines growth, nutrition, sensibility,
movement, and reproduction in these ultimate units? In higher units we
must ascribe the existence of these same attributes to the "necessary condi-
tion" that we "can only" explain evolution by supposing the existence of such
units. There simply results this state of things : Either we must assume that
the fundamental attributes of organism are one thing in an organism and
another in the ultimate vital units, or we must, to be logical, go on creating
cycles of ultimate units one within the other ad infinitum. In any event, we
alwavs arrive at exactly the same point. How shall we connect in any logical
way these vital units with the chemico-physical phenomena, which must
ultimately form the basis of the interpretation of organic phenomena?
Weismann says of these units :
As to their size, we can only say that they are far below the limits of visibility, and
that even the minutest granules, which we can barely perceive by means of our most
powerful microscopes, can not be small individual biophores, but must be aggregates
of these. On the other hand, the biophores must be larger than any chemical molecule,
because they themselves consist of a group of molecules among which are some of com-
plex composition, and therefore of relatively considerable size.
And De Vries says :
The pangenes are invisibly small ... all the various kinds of pangenes occur in
the nucleus.
We must accept Weismann's dictum that the biophore is larger than any
"organic molecule," but he evidently forgets that some of the organic mole-
cules approach dangerously near the limits of visibility. The determinants,
very numerous in all animals, and reaching in Arthropods "hundreds of
thousands," are not in multicellular animals "single biophores," but the
determinant "is a group of biophores which are bound together by internal
forces to form a higher vital unit." "On this basis the germ plasm repre-
sented by the chromatin of the egg nucleus of an insect egg must contain
'hundreds of thousands' of these determinants, each made up of biophores,
each in turn larger than any organic molecule." Now we know with
considerable accuracy the size of many organic molecules, and if there be
biophores endowed as Weismann supposes, each vital unit must be con-
stituted of enough organic molecules to carry on the "fundamental char-
acters." When this is further followed out the whole idea becomes an all
too obvious absurdity, because the amount of the chromatin in the germ
cells of Leptinotarsa is far too small to contain the amount of organic mole-
cules that it must according to Weismann's hypothesis. By measuring dur-
ing the oogonial and spermatogonial divisions the diameter and length of the
chromosomes, I have come to the conclusion that the maximum amount of
chromatin in the oogonia or spermatogonia is not over 0.0000225 c. mm.
If we suppose each biophore to be made up of only five molecules (an
unthinkable supposition), and each determinant of only two biophores and
NATURE AND CAUSE OF VARIATION. 305
only 100,000 determinants — a number far below Weismann's supposition for
insects — we sball have on this basis 5X2X 100,000, or 1,000,000 organic
molecules compressed into the space of 0.0000225 c- mm- Now, it is obvious
that the initial five molecules can not be the simple CH groups, but are
extremely complex and large molecules, and each molecule can have but the
0.000,000,000,022,5 c- mm- °f space, and this only upon the basis that the
chromatin is absolutely solid, with no space for molecular movement. We
know, however, that the chromatin as we can see it is not solid, but is more or
less an aggregation of particles ; and if we follow out the idea, each of the
necessary 1,000,000 molecules in the assumption has certainly far less than
the 0.000,000,000,022,5 c- mm- m cubic contents. It is obvious from this data
that, giving the id-determinant-biophore hypothesis every chance and the ben-
efit of every doubt, there still remains the insurmountable difficulty that it
"can only" in a superficial way explain up to a certain point, but when pushed
beyond that point it fails utterly. As a general conception of development
and evolution the Weismannian id-determinant-biophore hypothesis does not
stand the test of examination and trial in the light of our present knowledge
of organic activity.
How shall we conceive of the germ plasm? Most obviously, in order to
investigate variation and the method of evolution, we must form some con-
ception of the nature of the germ plasm. Whether the chromatin is the germ
plasm or not, we know it to be an extremely complicated series of complex
organic substances constantly in process of chemical change, and giving off
elaborated cell products for diverse purposes. As far as we have any real
knowledge, the chromatin is wholly concerned in metabolic processes, and
although it is the most complicated and most discussed material in the cell,
especially in cell reproduction, the fact is undeniable that the part taken by the
chromatin in the process of reproduction is much more passive than that
taken by the plasma. Cell division or cell reproduction is an activity in
which many of the important elements are non-chromatic, and are as constant
structures as the chromatin, even though not so strikingly present in cyto-
logical preparations. Sensation, motion, and the mechanism of reproduction
lie in the materials outside of the chromatin ; the trophic activities lie, as far
as our knowledge goes, within the chromatin.
It is undeniably true that the plasma is nourished or even perhaps derived
from the nucleus — that is, there is growth of the plasma — but we never find
chromatin without plasma, nor plasma without chromatin. They are interde-
pendent. The plasma is the seat of response, motion, and the mechanism of
reproduction ; the chromatin is the seat of metabolic activity, and they can not
be separated. We can no more say that since the plasma grows at the expense
of the activity of the chromatin, the chromatin is primary and the plasma
secondary, than we can argue that the metabolic mechanism of man is
306 PROBLEM OF THE ORIGIN OF SPECIES.
primary and all the rest is purely secondary. True enough, man is for his
bodily structure entirely dependent upon the metabolic mechanism within
him, but I think the mechanism would fare badly for materials for its activity
if left to its own devices to supply them. Is it different in the cell ?
It is wrong, I believe, to hold that either chromatin or plasma is either
primary or secondary, because each is absolutely necessary to the other — the
chromatin functioning largely in the phenomena of metabolism and growth,
the plasma in sensation and motion, and both active in reproduction. The
germ plasm we must consider, I believe, to consist of both plasma and chro-
matin. Man}-, on the basis of the phenomena of cell reproduction, will com-
bat this idea with the argument — Why is it that the most conspicuous element
in reproduction, the chromatin, occupies in this conception a place of minor
importance? But have we any basis, other than the tendency of the last
twenty years, which has developed methods to render the chromatin conspic-
uous, for believing that other elements are not equally important and con-
stant? Has any one shown that in cell reproduction there is an unequal or
haphazard division of centrosotnes, archoplasm, nucleo-plasm, plasma of the
asters, and spindle fibers ? These structures are divided with the same impar-
tiality between the daughter cells as are the chromosomes. There exists
beyond any question a regularly ordered, nearly exact halving of many of the
plasmatic structures of the cell in reproduction, and although the daughter
cells may differ in size this is always the result of stored foodstuffs ; the
plasma, especially that concerned in cell reproduction, is shared alike. Our
attention has been too much fixed upon the chromatin, which, because of easy
demonstration and nicety of results, has blinded us to the significance and
equal importance of the process of cell reproduction in the plasma. All our
available information as to the chromatin activity, the formation of cell
products, enzymes, glandular secretions, nutritives contents of ova, and so on,
shows chromatin only as a seat of chemical activity, a trophic material,
extremely important, to be sure, but not the whole life of the cell. I believe
that from the present state of our knowledge we should regard the chromatin
material solely as the seat of nutritive processes and growth activity, the
plasma, as the seat of motion and sensation, and reproduction, as involving
both. At present there exists not one whit of evidence to show that chro-
matin is endowed with other metabolic and growth processes. Reproduction
seems to and must of sheer necessity concern both, resulting in an equal dis-
tribution to the daughter cells of the plasmic and chromatic constituents of the
germ plasm. Reproduction is in reality the only process common to the
entire cell mass, and in reproduction the activities of the plasma are equal in
importance to those in the chromatin.
The logical outcome of this conception is that the germ plasm, instead of
representing a mosaic or colony of vital elements living within a protoplasmic
NATURE AND CAUSE OF VARIATION. 307
body in symbiotic relation, must be thought of as a unit or separate entitv.
That is, a germ cell is not germ plasm plus protoplasm and stored food, but
the germ cell, with its initial endowment of plasma, which it derived by
division from its ancestors, and which it always retains and hands on from
generation to generation in each species, is all germ plasm. Each germ cell
later, by the accumulation of stored food or the elaboration of other materials,
becomes differentiated, but the original chromatic and plasmic endowment is
singularly constant. In Lcptinotarsa each germ cell comes to the beginning
of its period of differentiation (growth period) with an almost exactly similar
endowment of dense homogeneous plasmic material.
While we can not, I think, at the present time form any adequate picture
or frame any conception of the architecture of the germ plasm, we can not,
however, escape the conviction that the germ cell is an entity, and is the
product of a series of developmental stages which always recur in the same
order and are relatively invariable. It is able to repeat these stages in devel-
opment, to overcome serious difficulties, and in the end attain the same gen-
eral result, which differs only from its specific type in relatively minor points.
Beyond all doubt, this sequence of stages is conditioned in the germ plasm,
but not predetermined in the germinal material as material particles, each
the bearer of particular attributes of one or more cells or cell groups. Rather
we must think of the germ plasm as containing a few general characters
from which, in some wholly unknown way, there come the sequence of stages
in development of the diverse organs and the apparent isolation and unity of
characters which we find. We have good reason, therefore, for the belief
that development is an epigenetic process, that one state is conditioned
by preexisting stages, and that a variation is the result of responses in the
plastic developing organism ; that is, variation is to be interpreted upon the
basis of responses to stimuli directed by the stage of development reached
and the nature of the preexisting stages. J'ariation is also epigenetic and
not a predetermined character in organisms.
Mutation.
De Vries and his followers maintain that variations standing far apart
from the general population, and differing often in many characters, differ
fundamentally from the normal inheritable fluctuating variations. De Vries's
"mutants" bred true, but with intense selection and in-and-in breeding ; and
in my experiments with Leptinotarsa there have been found and reared, with-
out intense selection, examples of these same extreme variations. De Vries
maintains that the "mutants" of Oenothera are in all directions, which may be
true; but, after examining his work and seeing the various "mutants" grow-
ing, I should say that they were not in all directions, but in two chief direc-
tions, and that the "mutation" consists in plus and minus changes of the
308 PROBLEM OF THE ORIGIN OF SPECIES.
parental characteristics. In Lcptinotarsa these variations, which came from
the various species of the genus, are orthogenetic.
The question whether or not we shall consider these extreme variations as
fundamentally different from those nearer to the parental mean is crucial.
While I have no basis for judging the nature of these extreme variations
("mutants") in plants, those which have appeared in my cultures are all
the results of responses to stimuli acting upon the germ plasm, as shown in
Chapter V, and, as far as I am able to discover, the)' differ from smaller varia-
tions only in the amount of the deviation from the mean of the parents. All
of my "mutants" are found to deviate from the parents most in one character
and to a far less degree in others. I am convinced, therefore, that the one is
the real permanently changed character, and that the less strongly marked
changes are due to correlation in variability. This point I can not at present
be positive of, because it needs sufficient material for adequate quantitative
treatment, and an attempt to make positive assertions on any other basis is not
warranted. It offers, however, a most plausible explanation of the fact that
mutants differ from the parent in several characters.
There is no reason for regarding the "mutants" of Lcptinotarsa as other
than responses to stimuli, which, acting upon the germ plasm, produce the
variations found. I have demonstrated in Chapter V that all inheritable
variations behave alike, and in no case is there any evidence that there is a
fundamental difference between "mutants" and any other heritable variation.
The mutations in Lcptinotarsa are solely responses to stimuli, orthogenetic
(Chapter II), and are distributed in their distances from the parental mean in
full conformity with the law of the distribution of error.
In text-figure 31 I have represented the distribution of the heritable varia-
tions about the mean of Lcptinotarsa dccemlincata. This figure is based upon
the inheritable variations of general color (albinism and melanism), which
carry in correlative variability other characters, and with this distribution is
given the occurrence of the "mutants" in their order and degree of removal
from the parental mean. To the left of the polygon are the variations in
increased coloration, melanism, with which are also correlated increased size,
etc., and to the right, albinism, with the correlated variation in decreased size
and punctation. The polygon is based upon one character, and, although
through correlation other characters also vary, it represents the general dis-
tribution of permanent variations about the mean of the species. It is shown
that in nature we find about one of these "mutants" in every 6,000 (5,447)
beetles examined. So when by chance in any lot of variates one of these
extreme variations is found, we ought not to jump to the conclusion that it is
not a part of the normal variability, because it is. To determine the distribu-
tion of the variations in a few hundred variates is perhaps useful, but it
represents only the distribution of the variations on the average, and not the
distribution of variation in the long run.
MUTATION.
309
I maintain, therefore, that "mutation" is not a special kind of variability,
different from that of "ordinary fluctuating variation," but it is a part of the
normal variability, and the direct response of the germ plasm to stimuli. This
is fully borne out by the data and experiments with Lcptinotarsa. This, how-
Range of " normal fluctuating variability.'
Text-figure 31.— To show the distribution of inheritable variations about the mean of
L. dccemliueata. p = L. pallida; d = L. defectopuvctata: n — L, minuta; i=L. immac-
ulolhorax ; Tft = L. melanicum; o = L. obsolefa; t = L. tortuosa; r = L. rubrivittata.
310 PROBLEM OF THE ORIGIN OF SPECIES.
ever, need not necessarily alter our conception of the part these variations
may play in evolution. It simply brings them into line with the general
phenomena of variation.1
We must conclude, in the light of the evidence herein gathered and corre-
lated, that all variation represents solely the response of plastic organic mate-
rial to stimuli according to the method of trial and error. Permanent variations
are also of the same origin, and, as shown in Chapter V, are produced or
primarily arise in the germ plasm, and later or secondarily appear in the
soma, through the production of the soma from the germ. It is shown, too,
that these permanent variations or divergences may be small or large, but they
obey the law of the distribution of error in their proportionate appearances and
distance from the mean. There is, then, no necessary incongruity between
gradual small variation and rapid large variation in the origin of species, but
the two are the extremes of the same process. That variations are not pre-
determined, but are epigenetic, arising as the result of the interaction of
preexisting series of stages in the organism and its environment, seems to be
the only possible method of interpreting these experiments.
The Origin of Species.
On the basis of our present knowledge the problem of the origin of species
divides itself into two chief questions- — the origin and the preservation. Biol-
ogists at present are divided as to the origin, one school, the neo-Lamarckians,
maintaining that variations may first arise in the soma through mechanical
or chemical causes, and later be transferred to the germ ; the other, that they
arise in the germ plasm, and later appear in the soma. All acknowledge that
the variation, to become transmissible, must ultimately come to be incor-
porated into the germ plasm. There exists at present not one single fact
to show the inheritance of acquired somatic variations or their incorpora-
tion into germ plasm. Moreover, the neo-L,amarckians are at fault in over-
looking the very fundamental fact that the variations found to arise in the
soma arc possible only because the germ-plasm constitution has developed
a soma in which variation in certain directions is possible and impossible in
others. That is to say, we find a particular variation because the germ plasm
has so constituted the soma that it can vary in certain directions and not in
others, and in the same direction as the germ plasm itself is varying. It is a
curious fact, moreover, that the idea of the inheritance of acquired characters
is supported almost entirely by the data of anatomy and paleontology, two
lines of investigation which, least of all, are capable of giving data upon the
question. Both deal with end results, and can show only what has happened.
1 In this paper I have avoided any consideration of the hypothesis of unit characters
and their isolation as postulated by De Vries. The data and conclusions upon this and
allied problems will be presented in another paper.
ORIGIN OF SPECIES. 3II
As I have pointed out, there is not at present evidence to show the origin
of any heritable variations in the soma. Moreover, I have shown that in
these beetles we can get new permanent variations by stimulating the germ
cells, and in no other way. The question is one for direct experimentation,
and not for the plausible arrangement of paleontological evidence. That the
Lamarckian factor could explain the phenomenon of evolution all admit,
could the fundamental assumption be established even in one single case.
Cope, of all the neo-Lamarckians, undoubtedly saw most clearly the real
issue and its results. Thus he writes :
It is evident that evolutionists are reaching greater harmony of opinion on the ques-
tion of inheritance. In fact, the discussion is sometimes a logomachy dependent on
the significance which one attaches to the term "acquired characters." Thus Von Rath
remarks : "There is nothing in the way of the opinion that by the continued working of
such external influences and stimuli the molecular structure of the germ-plasm also
experiences a change which can lead to a transmission of transformations. Above all,
it ought not to be forgotten in this case that somatic cells are in no way the first to be
modified by the stimulus, and that then, by some sort of unexplained process (pan-
genesis or intracellular pangenesis), this stimulus is transmitted gradually by these
cells to the plasma of the germ-cells. The influence on the germ plasm is rather a
direct one, and if by continued influence a transformation of the structure of this plasm
takes place and transmission occurs, we have then simply a transmission of blastogenic,
and by no means of somatogenic characters, and therein is not the slightest admission
of the transmission of acquired characters.
This paragraph contains an admission of the doctrine of diplogenesis, and does not
regard the phenomena as including a transmission of acquired characters. Nevertheless
the stimuli traverse the soma in order to reach the germ-plasma. Such an energy is
evidently, then, not of blastogenic origin, although it is such in its effects. Moreover,
Von Rath omits to mention the fact that in traversing the soma the stimulus frequently,
if not always, produces effects on the latter similar to those which it produces on the
germ plasma. I should call this process the inheritance of an acquired character, even
in the case where no corresponding modification appears in the soma, since the causative
energy is acquired by the soma, and is not derived from the existing germ plasma.
The acceptance of Von Rath's view, and the assertion that because "causa-
tive energy traverses the soma" it is therefore acquired from the soma, is
largely a matter of definition of the terms used. The soma in this case is a
medium through which the stimulus passes, as through water, air, or earth ;
and while it is true that the "causative energy" is not derived from the exist-
ing germ plasm, neither is it derived from the soma, nor acquired by the soma,
but both soma and germ respond to the only stimuli which they can ever
experience — those which are primarily external. After all, it would seem
that the question really comes down to a play upon words. The only ques-
tion is — Can the effects of a stimulus which produces only a somatic variation,
as in the pupa in Poulton's experiments, or the beetles in the experiments of
the third chapter, be later transferred to the germ plasm? Only experimental
investigation can answer this question, and at present all the data give a
312 PROBLEM OF THE ORIGIN OF SPECIES.
most decided negative answer. Because the stimuli pass through the soma
on their way to the germ in no wise enters into the question. That all herita-
ble variations — hence factors in evolution — arise primarily in the germ plasm,
and secondarily in the soma, is on the basis of all of our present evidence the
only acceptable hypothesis.
The recent compilation of data by Redfield concerning "dynamic inherit-
ance" of speed in race horses, and of other characters in dogs and men,
wherein he attempts to show the inherited effects of use, fail in that there
exists in every case the unproven assumption that the ancestor of the race
was an ordinary animal, whereas the very fact that the original progenitor of
the American race horse was able to train so long and to such good effect
shows that it was in its initial make-up different from other horses, and the
further development of the race horse seems to be purely a selective process.
Were this not so there is no reason why there should not have been a dozen
or five hundred horses of the same class. The very fact that this original
race-horse progenitor was able to develop speed beyond the ordinary, and
keep it up longer, is about all the evidence one needs that the original progen-
itor varied from the common race of horses in just those characters which
now form the basis of the traits of American race horses. The physical
capacity in this case was primary, the dynamic development later, and there
is no reason to suppose that the original capacity for dynamic development
arose other than first in the germ plasm. As far as any evidence goes that
we can get at, Redfield's "dynamic development" is a selective process based
upon an initial germinal variation.
Students of evolution are coming more and more to believe in the origin
in the germ plasm of all variations that are effective in evolution. Evidence
is constantly being accumulated, as Fischer, Standfuss, and MacDougal have
shown in a few cases, and as shown here in larger series, that permanent
variations are direct responses in the changed constitution of the germ plasm
to stimuli. How these are effected is not at present known, and is a subject
for investigation and not for anticipation. Further, that these variations
follow the law of trial and error is for these beetles fully shown, and it is
further shown that just this method would result in orthogenesis, and even
mutation, without the introduction of "forces," "ultimate vital units," or
"latencies." How these modifications are preserved is the second part of the
problem of the method of evolution.
Although evolutionists differ in their opinions in regard to the extent
and importance of natural selection, all, with the possible exception of the
Eimerites, give it an important place in organic evolution. Of late there has
grown up, since the publication of De Vries's work, a tendency to ascribe to
"mutation" a far greater importance and to a considerable extent to substi-
tute that process for all others in evolution, even though De Vries distinctly
ORIGIN OF SPECIES. 313
points out that "mutation" is complementary and not antagonistic to natural
selection, and that the two are necessary in evolution — that is, "mutation"
explains the origin of variations in evolution, and natural selection their
preservation.
As I understand natural selection, Darwin never claimed that it originated
anything, but only worked upon that which was already in existence ; but we
must admit that since any variation depends upon the antecedent series of
stages in the organism, that any stages in the organism which survive through
natural selection or are not eliminated thereby, as in characters of non-
selective value, exert a powerful influence in evolution in that they limit
variation to certain possibilities. Morgan calls this "very plausible," but
careful study of any set of organisms and their variations shows that it is
actually true. This fact we saw clearly in the second and third chapters of
this paper. In this way selection, by determining the survival, acts also to
eliminate certain possibilities from future evolution.
Of recent years there has been a deal of writing in which natural selection
is denounced as incompetent and trivial. However, few, if any, of these
"incontrovertible proofs" rest upon anything more than poor logic. Natural
selection is a subject for investigation, not for argumentative denunciation.
In relatively few cases do we know just how selection works — just what
selection does — and not until we know accurately the action of selection can
we begin to estimate its effect in evolution. /;; various ways selection has
been shown to be actively at work in Lcptinotarsa, but as far as discovered
always in conservative ways, eliminating extremes and limiting the reproduc-
tive population to the individuals nearest to the racial mean. Thus in hiber-
nation the extreme variations are all eliminated as far as the evidence goes,
and a "mutant" like pallida or melanicum has only the remotest chances of
becoming established in nature. It is not possible to say at present how poor
their chances are, but their occurrence once in 6,000 cases is not favorable to
their becoming established. If enough mutants arose at one time we could
easily imagine how they might become established, but this does not seem to
happen. In the tropics L. melanothorax has been known for nearlv fifty
years, yet it has not been able to gain a foothold, although it is far more
numerous than pallida. L. rubicunda seems more fortunate, but its distribu-
tion is as yet very limited, and seems to be decreasing, being one-third less in
1905 than in 1904. Its future history will depend on whether it will or will
not be able to meet the conditions of its existence.
The real question in the method of evolution at present is whether species
arise by the preservation of large variations or "mutants" or by small
accumulated variations. It appears that in both plants and animals large
variations occur in nature, but these, as far as all evidence goes to show, are
most rigorously exterminated by natural selection, and only the mean and
314 PROBLEM OF THE ORIGIN OF SPECIES.
modal individuals survive and reproduce the species. / have failed utterly
to discover in these beetles evidence that mutants have taken any great part
in evolution, all evidence showing them to be most rigorously exterminated
by natural selection {Chapter IV). On the other hand, the study of geo-
graphical distribution and variation gives the strongest of circumstantial
evidences for direct rapid transformation in response to environmental stimuli
as the result of dispersion. I am therefore of the opinion that the evolution
of the genus Lcptinotarsa, and of animals in general, has been continuous
and direct, developing new species in migrating races by direct response
to the conditions of existence. In this evolution natural selection has acted
to determine antecedent states and the persistence of new variations, but
in each race or species it acts as the conservator of the race, keeping down
extreme variations through their elimination in hibernation, larval life, and
selective mating. It is in Lcptinotarsa the conservator of the racial mean
and mode, the destroyer of all variation diverging much from the ortho-
genetic trend of evolution, which is itself a product of natural selection.
The evolution of the genus Lcptinotarsa, according to data gathered, has
been through response to the stimuli of the conditions of existence in changed
germ-plasm constitution, according to the method of trial and error, with
natural selection acting as the conservator of the race by limiting the varia-
tions to a narrow range of possibilities. The breeding "mutants" in our gar-
dens and laboratories can not tell us how they would succeed in nature ; my
experience with these beetles is that they fare badly, and, as far as I can dis-
cover, that they play a minor role in the evolution of species. This view
differs, therefore, from that of De Vries, who sees in "mutants" the origin
of species; the real test is, as De Vries clearly sees, the fate of these
"mutants" in nature. This I have been able in some measure to test, and at
present there seem to be insuperable difficulties in the path of all observed
"mutants" in Leptinotarsa. L therefore regard mutations as prophetic varia-
tions indicating what may perhaps be the next species in the evolution of the
race. How this evolution is brought about in nature is a subject for observa-
tion and experiment, and not for anticipation. The method of the origin of
variation herein developed does account for the origin of variation upon a
natural basis, and "mutation" is shown to be but part of the general phenom-
ena of variability. That is, in variability there is unity and not discontinuity,
and inheritable variations differ not in kind, but only in degree.
BIBLIOGRAPHY.
In the following bibliography I have given only the more important refer-
ences. Some of the smaller references are printed in the text, but a far
larger number, amounting to several hundred, do not appear at all.
Adams, C. C.
1902. Southeastern United States as a center of Geographical Distribution of Fauna
and Flora. Biol. Bull., vol. Ill, No. 3, pp. 115-131.
AGASSIZ, L.
184S-1S49. Twelve Lectures on Comparative Embryology. Boston, Lowell Institute.
Allen, J. A.
1877. The Influence of Physical Conditions in the Genesis of Species. The Radical
Review, pp. 107-140 ; May.
Bates, H. W.
1S62. Contributions to an Insect Fauna of the Amazon Valley. Lepidoptera : Heli-
conidse. Trans. Linn. Soc. London, vol. xxm, pp. 495-566, pis. 55-58.
Bateson, W.
1894. Materials for the Study of Variation, Treated with Special Regard to Discon-
tinuity in the Origin of Species. London, 598 pp.
BemmelEN, J. F. Van.
1SS9. Ueber die Entwicklung der Fiirben und Adern auf den Schmetterlingsfliigeln.
Tijdsch. der Nederland. Dierkundige Vereeniging, Ser. 2, Deel 2, pp.
235-247.
Bottler, Max.
1894. Die animalischen Faserstoffe. Ein Hilfs- und Haudbuch fur die Praxis, umfas-
send Vorkommen, Gewinnung, Eigenschaften und technische Verwendung
sowie Bleichen und Farben tierischer Faserstoffe, nach dem gegenwartigen
Standpunkte der Wissenschaft bearbeitet. viii + 215 pp., 16 figs. Wien :
A. Hartleben.
Bumpus, H. C.
The Variations and Mutations of the Introduced Sparrow, Passer domesticus.
Biol. Lect. , Woods Holl, in 1896. pp. 1-15. Boston.
Burkill, I. H.
1895. On Some Variations in the Number of Stamens and Carpels. Jour. Linn. Soc'
Bot., vol. 3r, pp. 216-245.
Burt, H.
1865. Potato Beetle. Coleman's Rural World, Sept. 15, 1S65.
Collins, C. C.
1S70. Colorado Potato Beetle. Am. Ent. and Bot., vol. 2, p. 341.
Cope, E. D.
1904. The Primary Factors of Organic Evolution, pp. vii + 547. Chicago.
CosTE, F. H. P.
1S90-1S91. Contributions to the Chemistry of Insect Colors. The Entomologist, vol.
xxm, pp. 12S-132, 155-168, etc. Continued and concluded in vol. xxiy.
Crampton, H. E.
1904. Experimental and Statistical Studies upon Lepidoptera. I, Variation and
Elimination in Philosamina Cynthia. Biometrika, vol. in, pp. 113-130.
315
31 6 BIBLIOGRAPHY.
CUSHMAN, J. A.
1903. Studies of Localized Stages in Some Plants of the Botanic Gardens of Harvard
University. Amer. Nat., vol. xxxvn, pp. 243-259.
1904. Localized Stages in Common Roadside Plants. Amer. Nat., vol. xxxvm, pp.
819-S32.
1902. Studies of Localized Stages of Growth of Some New England Plants. Amer.
Nat., vol. xxxvi, pp. S65-SS5.
Darwin, C. R.
1S59. The Origin of Species.
1571. The Descent of Man.
Davenport, C. B.
1904. The Animal Ecology of the Cold Spring Sand Spit. The Decennial Publica-
tions, The University of Chicago, First Series, vol. x, pp. 157-176. (Dated
1903-)
1904. Statistical Methods, with [Special Reference to Biological Variation. New
York. pp. viii-223.
DEWITZ, J.
1903. Ueber die Herkunftdes Farbstoffes und des Materials der Lepidopterenkokons.
Zool. Anz., xxvii, pp. 161-168.
Dodge, J. R.
1S66. Monthly Rep. U. S. Dept. of Agri., pp. 245-356.
1566. Monthly Rep. U. S. Dept. of Agri., pp. 287, 344, 38S.
1567. Monthly Rep. U. S. Dept. of Agri., pp. 244, 365.
1S70. Monthly Rep. U. S. Dept. of Agri., pp. 191-332.
1871. Monthly Rep. U. S. Dept. of Agri., pp. 191-332.
1572. Monthly Rep. U. S. Dept. of Agri., p. 217.
1S76. Monthly Rep. U. S. Dept. of Agri., p. 345.
DORFMEISTER, G.
1864. Ueber die Einwirkung verschiedener, wahrend der Entwickelungsperioden
angewendter Warmegrade auf die Farbung und Zeichnung der Schmetter-
linge. Mittheil. des Naturwiss. Vereius fiir Steiermark, Heft II, p. 99.
EdgerTON, J.
1861. Potato Insects. Prairie Farmer, Aug. 29, p. 116.
Edwards, H. W.
1S75. An abstract of Dr. Aug. Weismanu's paper on " The Seasonal Dimorphism of
Butterflies." Can. Ent, vol. vn, p. 228.
1884. Further experiments upon the effect of cold applied to the chrysalids of but.
terflies. Can. Ent., vol. xvi, p. 232.
ElMER, G. H. T.
1889. DieArtbildung und Verwandtschaft bei den Schmetterlingen. Jena. 243 pp.,
4 taf.
1897. Entstehungder Arten. Theil 11 : Orthogenesis der Schmetterlinge. Jena. 513
pp., 2 taf.
EnTEMan, W. H.
1904. Coloration in Potistes. The Carnegie Institution of Washington. Publication
No. 19, pp. i-SS ; 6 plates.
Fischer, E.
(See numerous papers in Allgemeine Zeitschrift fiir Entomologie, vols. 11-vill.)
Fitch, A.
1865. Injurious Insects. Trans. N. Y. State Agr. Soc, p. 79S.
Gage, J.
1867. Rep. Sta. Bd. Agr. Mich., 1867, pp. 71-77.
G11.1.MAN, H.
1870. The Progress of the [Colorado] Potato Bug. Am. Ent. and Bot, vol. 2, pp.
84-85.
1573. The Colorado Potato Beetle Varying its Food. Am. Nat., vol. 7, pp. 430-431.
Glovkr, T.
1S73. Monthly Rep. U. S. Dept. of Agri., pp. 237, 345, 426, 496.
1874. Monthly Rep. U. S. Dept. of Agri., p. 374.
1875. Monthly Rep. U. S. Dept. of Agri., pp. 22S-307. ,
BIBLIOGRAPHY. 317
Griffiths, A. B.
1S92. Recherches sur les Couleurs de Quelques Insectes. Coinptes Rendus Acad.
Sci. Paris, tome cxv, pp. 95S-959.
Gulick, J. T.
1905. Evolution, Racial and Habitudinal. The Carnegie Institution of Washington.
Publication 25, pp. XII + 269.
Haase, E.
1891-92. Untersuchungeu iiber die Mimicry auf Grundlage eines natiirlicheu Sys-
tems der Papilionideu. Bibliotheca Zoologica, Heft vin, pp. 120, 14 taf.
Cassel.
Haeckel, E.
1S66. Generelle Morphologie der Orgauismen. 1-11. Berlin.
Hagex, H. A.
1SS2. On the Color and the Pattern of Insects. Proc. Am. Acad. Arts and Sci., xvn
(n. s., vol. ix), pp. 234-267.
Hazen, R. W.
1S65. Colorado Beetle. N. Y. Semi-Weekly Tribune, July iS.
Houiax, D. S.
1S70. Colorado Beetle around Springfield. Am. Ent. and Bot, vol. 11, pp. 370-371-
Hopkins, F. G.
1S95. The Pigments of the Pierida;. Philos. Trans., Lond., 1S6 B, pp. 661-682.
Horn, G. H.
1872. Coleoptera in the Preliminary Report of the United States Geological Survey
of Montana and Adjacent Territories, pp. 3S2-392.
Jackson, R. T.
:S99. Localized Stages in Development in Plants and Animals. Mem. Bost. Soc.
Nat. Hist., v, No. 4, pp. S9-152, pis. 16-25.
Jacobv, M.
18S0-1892. Coleoptera Phytophaga. Biologica Centrali-Americana. Coleoptera,
vol. VI, part I.
Jennings, H. S.
1904. Contributions to the Study of the Behavior of Lower Organisms. Carnegie
Institution of Washington. Publication No. 16, pp. 256.
KUKENBERG, C. F. W.
18S4. Grundziig einer vergleichenden Physiologie der Farbenstoffe und der Farben.
Vergl. Phys., Vortrage, in, 102 pp.
Ledderhose, G.
1878. Ueber Chitin und seine Spaltungsprodukte. Zeitschr. f. physiologische
Chemie, Bd. II, pp. 213-227.
Linden, M. von.
189S. Untersuchungeu iiber die Eutwicklung der Zeichung des Schmetterlings-
fliigels in der Puppe. Zeitschr. fiir Wiss. Zool., Bd. I.XV, pp. 1-49, pis.
i-in.
1902. Le dessin des ailes des Lepidopteres, recherches sur les Evolutions sans anto-
genese et la phylogenese des especes, son origine et sa valeur systematique.
Am. Sci. Nat., XIV, pp. 1-196, pis. i-xx.
Ludwig, F.
1901. Variatiousstatistische Problem und Materialien. Biometrica, vol. 1, pp. 11-29.
LrrcER O
1895.' Univ. Minn. Agr. Exp. Sta., U. S. Dept. Agr., Div. Ent., Bull. 43.
MacDougal, D. T.
1906. The Inluction of New Species. Science, N. S., vol. xxin, No. 585, p. 422.
MacLeod, J.
1899. Overdecorrelatie tusschen hetaantal meeldraden en het aantal stampers bij het
Speenkruid {ficaria ra.nunculoid.es). Bot. Jaarboek. Dodonsea, xi, p. 91,
31 8 BIBLIOGRAPHY.
Mayer, A. G.
1897. On the Colors and Color Patterns of Moths and Butterflies. Proc. Bost. Soc.
Nat. Hist., vol. xxvu, No. 14, pp. 243-330, 10 pis.
1900. On the Mating Instinct in Moths. Ann. and Mag. of Nat. Hist., ser. 7, vol.
v. PP- J53-I90 ! also Psyche, vol. ix, pp. 15-20.
1902. Effects of Natural Selection and Race Tendency upon the Color Patterns of
Lepidoptera. Museum of Brooklyn Institute of Arts and Sciences. Science
Bulletin, vol. I, No. 2, pp. 31-88.
Merrifield, F.
(Temperature Experiments upon Pupae.) Traus. Ent. Soc. Lond., 1890, p. 131 ;
lS9r, p. 155 ; 1892, p. 33 ; 1S93, p. 55 ; 1S94, p. 425.
Morgan, T. H.
1903. Evolution and Adaptation, pp. xiii + 470. New York.
Mui.ler, F.
1879. Ituna und Thyridia. Kosmos, Bd. 5, p. 100-108, figs. 1-4. Translation, by
R. Meldola, in Transact. Ent. Soc. Lond., 1S79, pp. 20-29.
Murphy, T.
1862. A New Horticultural Foe. A letter to N. J. Colman, editor Valley Farmer,
July, pp. 209-210.
Noble, G.
iSgt. Abundance of Colorado Beetle in Georgia. U. S. Dept. Agri., Insect Life,
vol. iv, p. 135.
OSBORN, H.
1892. Notes on Injurious Insects of 1892. U. S. Dept. Agri., Insect Life, vol. v, pp.
111-114.
Packard, A. S.
1575. Report on the Rocky Mountain Locust and Other Insects now Injuring (or
likely to) Field or Garden Crops in the Western States. U. S. Geol. and
Geog. Surv. of Colorado and Adjacent Territory. Colorado Potato Beetle,
pp. 721-729.
1576. The Colorado Potato Beetle in Massachusetts. Sci. Farmer, vol. I, pp. 104-105.
1876. The Colorado Potato Beetle. Sci. Farmer, vol. 1, p. 16S.
Pearson, K.
1900. The Grammar of Science, pp. xviii-548. London.
Phares, D. L.
1SS9. The Potato Beetle in the South. U. S. Depl. Agri., Insect Life, vol. 11, p. 22.
PoulTon, E. B.
1887. The Experimental Proof of the Protective Value of Color and Color Markings
in Insects in Reference to their Vertebrate Enemies. Proc. Zool. Soc. Lond. ,
pp. 191-274.
1893. The Experimental Proof that the Colours of Certain Lepidopterous Larvae are
Largely Due to Modified Plant Pigments Derived from Food. Proc. Roy.
Soc. Lond., vol. uv, pp. 417-430, pis. iii-iv.
189S. Natural Selection the Cause of Memetic Resemblance and Common Warning
Colors. Jour. Linn. Soc. Lond., Zool., vol. xxvi, pp. 558-612, pis. 40-44.
Priest, O. E.
1S65. New York Semi- Weekly Tribune, Aug. 29, 1S65.
Reed, E. B.
1872. Insects Injurious to the Potato. Ann. Rept. of the Ent. Soc. of Ont. for 1872,
pp. 48-50.
Reinohl, F.
1903. Die Variation im Androeceum der Stellaria media Cyr. Bot. Zeit., vol. 61,
pp. 159-200.
Riley, C. V.
1868. First Annual Report on the Noxious, Beneficial, and Other Insects of the State
of Missouri. Jefferson City, Mo, pp. 101-117.
1870. Third Report on the Noxious, Beneficial, and Other Insects of the State of
Missouri. Jefferson City, Mo., pp. 92-101.
BIBLIOGRAPHY. 319
Riley, C. V.— Continued.
1571. Fourth Report on the Noxious, Beneficial, and Other Insects of the State of
Missouri. Jefferson City, Mo., pp. 1-22.
1872. Fifth Report on the Noxious, Beneficial, and Other Insects of the State of
Missouri. Jefferson City, Mo., pp. 11-16.
1S76. Potato Pests : Being an Illustrated Account of the Potato Beetle and Other
Insect Foes of the Potato in North America, with Suggestions for their
Repression and Methods for their Destruction. New York, Orange Judd
Co., 10S pp., 49 figs,
1S82. Change of Habit : Two New Enemies of the Egg Plant. Am. Nat., vol. XVI,
pp. 67S-679.
1892. The Colorado Potato Beetle in the South. U. S. Dept. Agr., Insect Life, vol.
v, p. 50.
Rogers, R. V.
1572. Doryphora 10-lineata. Can. Ent., vol. IV, p. 200.
Saffer, L. G.
1870. Colorado Potato Beetle in Indiana. Am. Ent. and Bot., vol. II, p. 304.
Saunders, W.
1878. Annual Address of the President of the Entomological Club of the Am. Assoc.
Adv. Sci. Can. Ent., vol. x, pp. 171-190.
1SS0. Annual Address of the President of the Entomological Society of Ontario.
Rep. Ent. Soc. Out. for 1880, p. 8.
Say, T.
1824. (Description of Doryphora 10-lineala n. sp.) Jour. Acad. Nat. Sci., Phila.,
vol. 3, part 11, p. 453.
Scudder, S. H.
1887. The Introduction and Spread of Pieris rapcv in North America. Mem. Bost.
Soc. Nat. Hist., vol. IV, No. 3, pp. 52-69.
Shimer, H.
1S66. Notes on Doryphora 10-lineata. Pract. Ent., vol. 1, p. 84.
1870. Insects Injurious to the Potato. Am. Nat., vol. Ill, pp. 91-99.
Shuix, G. H.
1902. A Quantitative Study of Variation in the Rays, Bracts, and Diskflorets of Astei
shorlii Hook., A. Novtz-Anglia: L., A.puniceus L., and A. prenanthoides
Muhl., from Yellow Springs, Ohio. Am. Nat., vol. 36, pp. 111-152.
1904. Place- constants for Aster prenanthoides. Bot. Gaz., vol. 38, pp. 333-375.
1905. Stages in the Development of Sium eicutifohum. The Carnegie Institution of
Washington. Publication No. 30, pp. 1-28.
STRECKER, A.
1882. A Short Text Book of Organic Chemistry. New York, xvni + 789 pp.
Tower, W. L.
1902. Variation in the Ray-flowers of Chrysanthemum leueanthemum I,., at Yellow
Spring, Green County, Ohio, with Remarks upon the Determination of
Modes. Biometrika, vol. 1, pp. 309-315.
1903. Colors and Color Patterns of Coleoptera. The Decennial Publications. The
University of Chicago, vol. x, pp. 33-69, pis. i-iii.
Town-send, C. II. T.
1892. Notes of Interest. U. S. Dept. Agri., Insect Life, vol. IV, p. 26-27.
UlkE, H.
1873. List of Species of Coleoptera Collected by Lieut. W. L. Carpenter, U. S. A.,
for the U. S. Geological Survey of Colorado. U. S. Geol. and Geog. Surv.
of Colorado, pp. 567-571.
URECH, F.
1891. Beobachtungen iiber die verschiedenen Schuppenfarbeu und die zeitliche
Succession ihres Auftretens (Farbenfelderung) auf den Puppenfliigelchen
von Vanessa urtietz und V, Io. Zool. Anz., Jahrg. xiv, 466-475.
1S93. Beitrage zur Kenntnis der Farben von Insektenschuppeu. Zeitschr. Wiss.
Zool., Bd. i,vii, Heft 2, pp. 306-384.
23— T
320 bibliography.
Vries, Hugo De.
1901. Die Mutationstheorie. Versuche und Beobachtungen iiber die Entstehung
von Arten im Pflanzenreich. 2 Bde., pp. xii+648, xiv+752. Leipzig.
1905. Species and Varieties, Their Origin by Mutation, pp. xviii -f- 847. Chicago.
Webster, F. M.
1889. Southern Spread of the Colorado Potato Beetle. U. S. Dept. Agri., Insect Life,
vol. 11, p. 122.
1901. The Trend of Insect Diffusion in North America. 32d Rept. Ent. Soc. Ont.,
1901, pp. 63-67.
1903. The Diffusion of Insects in North America. Psyche, April, 1903, pp. 47-58.
Wallace, a. r.
1867. [Theory of Warning Coloration.] Trans. Ent. Soc. Lond., series 3, vol. 5,
(Journal of Proc), pp. lxxx-lxxxi.
Walsh, B. D.
1S65. The New Potato-bug and Its Natural History. Pract. Ent., vol. I, No. 1.
1 866(7. The New Potato Bug. Pract. Ent., vol. 1, p. 88.
18666. The New Potato Bug. Pract. Ent., vol. 11, p. 13.
1867a. The New or Colorado Potato Bug. Pract. Ent., vol. II, p. 101.
18676. The Colorado Potato Bug. Pract. Ent., vol. II, p. 116.
186S. The Colorado Potato Bug ; Its Past History and Future Progress. Am. Ent.
and Bot., vol. 1, pp. 41-49-
Weismann, A.
1882. Studies in the Theory of Descent. 2 vols., 729 pp.
1902. The Evolution Theory. Translated by J. A. and M. R. Thompson. Two vols.,
pp. xvi + 416, 405. London, 1904.
Yule, G. U.
1902. Variation in the Number of Sepals in Anemone necorosa. Biometrika, vol. I,
PP- 3°7-3°9-
ZOPF, W.
1892-93. Beitrage zur Physiologie u. Morphologie niederer Organismen. Leipzig,
Hefteu 1, 11, in.
ERRATA.
Page 2, foot-note, second line : For " there" read "these."
Page 4, L. libatrix, first line: For " undistinguishable " read "indistin-
guishable. ' '
Page 17, line 11 : For "western" read "eastern."
Page 67, table 10, Under 7, for " a -f- b -\-c + e+/" read " a + b — d + e +
/". Under S, for "«'+£'+ <:'+<■'+ \t" read " a'+ b'+ d'+ e— | c,."
Page 80, table 25, top of second column : For " Per cent of value in modal
class" read " Per cent in Modal class."
Page 91, table 32 : Line 8, first column, " 14 " set in wrong font.
Page 108, table 41 : Line 17, " 98 " set in wrong font.
Page 309, legend of text-figure 31 : Add a = L. albida.
Plate 25, in legend, last line : For " shade " read " shape."