THE INTERNATIONAL SCIENTIFIC SERIES.
VOLUME XLIV.

I have recently learnt from the publishers of the 'International Scientific Series' that they have made arrangements with Sir John Lubbock to bring out in the same series a work of his on Ants and Bees. Necessarily, therefore, the material to be dealt with in his work will to a large extent overlap that which is presented by my chapters on the same insects; but after consulting with the publishers, and also with Sir John Lubbock, it has seemed to me undesirable to omit these chapters on account of the circumstances here stated. For, on the one hand, the facts will not lose their value from being twice told; and on the other, it is desirable that the present member of the Series should form in itself, so far as its Author can make it, a complete résumé of all the more important facts of Animal Intelligence.

THE
INTERNATIONAL SCIENTIFIC SERIES.

——————
Each book complete in One Volume, 12mo, and bound in Cloth.
——————

I. FORMS OF WATER: a Familiar Exposition of the Origin and Phenomena of Glaciers. By J. Tyndall, LL. D., F. R. S. With 25 Illustrations. $1.50.

II. PHYSICS AND POLITICS; Or, Thoughts on the Application of the Principles of "Natural Selection" and "Inheritance" to Political Society. By Walter Bagehot. $1.50.

III. FOODS. By Edward Smith, M.D., LL.B., F. R. S. With numerous Illustrations. $1.75.

IV. MIND AND BODY: The Theories of their Relation. By Alexander Bain, LL. D. With Four Illustrations. $1.50.

V. THE STUDY OF SOCIOLOGY. By Herbert Spencer. $1.50.

VI. THE NEW CHEMISTRY. By Professor J. P. Cooke, of Harvard University. With 31 Illustrations. $2.00.

VII. ON THE CONSERVATION OF ENERGY. By Balfour Stewart, M. A., LL. D., F. R. S. With 14 Illustrations. $1.50.

VIII. ANIMAL LOCOMOTION; or, Walking, Swimming, and Flying. By J. B. Pettigrew, M. D., F. R. S., etc. With 130 Illustrations. $1.75.

IX. RESPONSIBILITY IN MENTAL DISEASE. By Henry Maudsley, M. D. $1.50.

X. THE SCIENCE OF LAW. By Professor Sheldon Amos. $1.75.

XI. ANIMAL MECHANISM: A Treatise on Terrestrial and Aërial Locomotion. By Professor E. J. Marey. With 117 Illustrations. $1.75.

XII. THE HISTORY OF THE CONFLICT BETWEEN RELIGION AND SCIENCE. By J. W. Draper, M. D., LL. D. $1.75.

XIII. THE DOCTRINE OF DESCENT AND DARWINISM. By Professor Oscar Schmidt (Strasburg University). With 26 Illustrations. $1.50.

XIV. THE CHEMICAL EFFECTS OF LIGHT AND PHOTOGRAPHY. By Dr. Hermann Vogel (Polytechnic Academy of Berlin). Translation thoroughly revised. With 100 Illustrations. $2.00.

XV. FUNGI: Their Nature, Influences, Uses, etc. By M. C. Cooke, M. A., LL. D. Edited by the Rev. M. J. Berkeley, M. A., F. L. S. With 109 Illustrations. $1.50.

XVI. THE LIFE AND GROWTH OF LANGUAGE. By Professor William Dwight Whitney, of Yale College. $1.50.

XVII. MONEY AND THE MECHANISM OF EXCHANGE. By W. Stanley Jevons, M. A., F. R. S. $1.75.

XVIII. THE NATURE OF LIGHT, with a General Account of Physical Optics. By Dr. Eugene Lommel. With 188 Illustrations and a Table of Spectra in Chromo-lithography. $2.00.

XIX. ANIMAL PARASITES AND MESSMATES. By Monsieur Van Beneden. With 83 Illustrations. $1.50.

XX. FERMENTATION. By Professor Schützenberger. With 28 Illustrations. $1.50.

XXI. THE FIVE SENSES OF MAN. By Professor Bernstein. With 91 Illustrations. $1.75.

XXII. THE THEORY OF SOUND IN ITS RELATION TO MUSIC. By Professor Pietro Blaserna. With numerous Illustrations. $1.50.

XXIII. STUDIES IN SPECTRUM ANALYSIS. By J. Norman Lockyer, F. R. S. With Six Photographic Illustrations of Spectra, and numerous Engravings on Wood. $2.50.

XXIV. A HISTORY OF THE GROWTH OF THE STEAM-ENGINE. By Professor R. H. Thurston. With 163 Illustrations. $2.50.

XXV. EDUCATION AS A SCIENCE. By Alexander Bain, LL. D. $1.75.

XXVI. STUDENTS' TEXT-BOOK OF COLOR, OR, Modern Chromatics. With Applications to Art and Industry. By Professor Ogden N. Rood, Columbia College. New edition. With 130 Illustrations. $2.00.

XXVII. THE HUMAN SPECIES. By Professor A. de Quatrefages, Membre de l'Institut. $2.00.

XXVIII. THE CRAYFISH: an Introduction to the Study of Zoölogy. By T. H. Huxley, F. R. S. With 82 Illustrations. $1.75.

XXIX. THE ATOMIC THEORY. By Professor A. Wurtz. Translated by E. Cleminshaw, F. C. S. $1.50.

XXX. ANIMAL LIFE AS AFFECTED BY THE NATURAL CONDITIONS OF EXISTENCE. By Karl Semper. With Two Maps and 106 Woodcuts. $2.00.

XXXI. SIGHT: An Exposition of the Principles of Monocular and Binocular Vision. By Joseph Le Conte, LL. D. With 132 Illustrations. $1.50.

XXXII. GENERAL PHYSIOLOGY OF MUSCLES AND NERVES. By Professor J. Rosenthal. With 75 Illustrations. $1.50.

XXXIII. ILLUSIONS: A Psychological Study. By James Sully. $1.50.

XXXIV. THE SUN. By C. A. Young, Professor of Astronomy in the College of New Jersey. With numerous Illustrations. $2.00.

XXXV. VOLCANOES: What they Are and what they Teach. By John W. Judd, F. R. S., Professor of Geology in the Royal School of Mines. With 96 Illustrations. $2.00.

XXXVI. SUICIDE: An Essay in Comparative Moral Statistics. By Henry Morselli, M. D., Professor of Psychological Medicine, Royal University, Turin. $1.75.

XXXVII. THE FORMATION OF VEGETABLE MOULD, THROUGH THE ACTION OF WORMS. With Observations on their Habits. By Charles Darwin, LL. D., F. R. S., author of "On the Origin of Species," etc., etc. With Illustrations. $1.50.

XXXVIII. THE CONCEPTS AND THEORIES OF MODERN PHYSICS. By J. B. Stallo. $1.75.

XXXIX. THE BRAIN AND ITS FUNCTIONS. By J. Luys. $1.50.

XL. MYTH AND SCIENCE. An Essay. By Tito Vignoli. $1.50.

XLI. DISEASES OF MEMORY: An Essay in the Positive Psychology. By Th. Ribot, author of "Heredity." From the French, by William Huntington Smith. $1.50.

XLII. ANTS, BEES, AND WASPS. A Record of Observations on the Habits of the Social Hymenoptera. By Sir John Lubbock, Bart., F. R. S., D. C. L., LL. D., etc. $2.00.

XLIII. SCIENCE OF POLITICS. By Sheldon Amos, author of "Science of Law." $1.75.

XLIV. ANIMAL INTELLIGENCE. By George J. Romanes. $1.75.

XLV. MAN BEFORE METALS. By N. Joly, Correspondent of the Institute. With 148 Illustrations. $1.75.

XLVI. THE ORGANS OF SPEECH AND THEIR APPLICATION IN THE FORMATION OF ARTICULATE SOUNDS. By Georg Hermann von Meyer, Professor in Ordinary of Anatomy at the University of Zürich. With 47 Woodcuts. $1.75.

———————
New York: D. APPLETON & CO., 1, 3, & 5 Bond Street.

THE INTERNATIONAL SCIENTIFIC SERIES.

ANIMAL INTELLIGENCE.

BY
GEORGE J. ROMANES, M. A., LL. D., F. R. S.,
ZOOLOGICAL SECRETARY OF THE LINNEAN SOCIETY.
NEW YORK:
D. APPLETON AND COMPANY,
1, 3, AND 5 BOND STREET.
1884.

PREFACE.

When I first began to collect materials for this work it was my intention to divide the book into two parts. Of these I intended the first to be concerned only with the facts of animal intelligence, while the second was to have treated of these facts in their relation to the theory of Descent. Finding, however, as I proceeded, that the material was too considerable in amount to admit of being comprised within the limits of a single volume, I have made arrangements with the publishers of the 'International Scientific Series' to bring out the second division of the work as a separate treatise, under the title 'Mental Evolution.' This treatise I hope to get ready for press within a year or two.

My object in the work as a whole is twofold. First, I have thought it desirable that there should be something resembling a text-book of the facts of Comparative Psychology, to which men of science, and also metaphysicians, may turn whenever they may have occasion to acquaint themselves with the particular level of intelligence to which this or that species of animal attains. Hitherto the endeavour of assigning these levels has been almost exclusively in the hands of popular writers; and as these have, for the most part, merely strung together, with discrimination more or less inadequate, innumerable anecdotes of the display of animal intelligence, their books are valueless as works of reference. So much, indeed, is this the case, that Comparative Psychology has been virtually excluded from the hierarchy of the sciences. If we except the methodical researches of a few distinguished naturalists, it would appear that the phenomena of mind in animals, having constituted so much and so long the theme of unscientific authors, are now considered well-nigh unworthy of serious treatment by scientific methods. But it is surely needless to point out that the phenomena which constitute the subject-matter of Comparative Psychology, even if we regard them merely as facts in Nature, have at least as great a claim to accurate classification as those phenomena of structure which constitute the subject-matter of Comparative Anatomy. Leaving aside, therefore, the reflection that within the last twenty years the facts of animal intelligence have suddenly acquired a new and profound importance, from the proved probability of their genetic continuity with those of human intelligence, it would remain true that their systematic arrangement is a worthy object of scientific endeavour. This, then, has been my first object, which, otherwise stated, amounts merely to passing the animal kingdom in review in order to give a trustworthy account of the grade of psychological development which is presented by each group. Such is the scope of the present treatise.

My second, and much more important object, is that of considering the facts of animal intelligence in their relation to the theory of Descent. With the exception of Mr. Darwin's admirable chapters on the mental powers and moral sense, and Mr. Spencer's great work on the Principles of Psychology, there has hitherto been no earnest attempt at tracing the principles which have been probably concerned in the genesis of Mind. Yet there is not a doubt that, for the present generation at all events, no subject of scientific inquiry can present a higher degree of interest; and therefore it is mainly with the view of furthering this inquiry that I have undertaken this work. It will thus be apparent that the present volume, while complete in itself as a statement of the facts of Comparative Psychology, has for its more ultimate purpose the laying of a firm foundation for my future treatise on Mental Evolution. But although, from what I have just said, it will be apparent that the present treatise is preliminary to a more important one, I desire to emphasise this statement, lest the critics, in being now presented only with a groundwork on which the picture is eventually to be painted, should deem that the art displayed is of somewhat too commonplace a kind. If the present work is read without reference to its ultimate object of supplying facts for the subsequent deduction of principles, it may well seem but a small improvement upon the works of the anecdote-mongers. But if it is remembered that my object in these pages is the mapping out of animal psychology for the purposes of a subsequent synthesis, I may fairly claim to receive credit for a sound scientific intention, even where the only methods at my disposal may incidentally seem to minister to a mere love of anecdote.

It remains to add a few words on the principles which I have laid down for my own guidance in the selection and arrangement of facts. Considering it desirable to cast as wide a net as possible, I have fished the seas of popular literature as well as the rivers of scientific writing. The endless multitude of alleged facts which I have thus been obliged to read, I have found, as may well be imagined, excessively tedious; and as they are for the most part recorded by wholly unknown observers, the labour of reading them would have been useless without some trustworthy principles of selection. The first and most obvious principle that occurred to me was to regard only those facts which stood upon the authority of observers well known as competent; but I soon found that this principle constituted much too close a mesh. Where one of my objects was to determine the upper limit of intelligence reached by this and that class, order, or species of animals, I usually found that the most remarkable instances of the display of intelligence were recorded by persons bearing names more or less unknown to fame. This, of course, is what we might antecedently expect, as it is obvious that the chances must always be greatly against the more intelligent individuals among animals happening to fall under the observation of the more intelligent individuals among men. Therefore I soon found that I had to choose between neglecting all the more important part of the evidence—and consequently in most cases feeling sure that I had fixed the upper limit of intelligence too low—or supplementing the principle of looking to authority alone with some other principles of selection, which, while embracing the enormous class of alleged facts recorded by unknown observers, might be felt to meet the requirements of a reasonably critical method. I therefore adopted the following principles as a filter to this class of facts. First, never to accept an alleged fact without the authority of some name. Second, in the case of the name being unknown, and the alleged fact of sufficient importance to be entertained, carefully to consider whether, from all the circumstances of the case as recorded, there was any considerable opportunity for mal-observation; this principle generally demanded that the alleged fact, or action on the part of the animal, should be of a particularly marked and unmistakable kind, looking to the end which the action is said to have accomplished. Third, to tabulate all important observations recorded by unknown observers, with the view of ascertaining whether they have ever been corroborated by similar or analogous observations made by other and independent observers. This principle I have found to be of great use in guiding my selection of instances, for where statements of fact which present nothing intrinsically improbable are found to be unconsciously confirmed by different observers, they have as good a right to be deemed trustworthy as statements which stand on the single authority of a known observer, and I have found the former to be at least as abundant as the latter. Moreover, by getting into the habit of always seeking for corroborative cases, I have frequently been able to substantiate the assertions of known observers by those of other observers as well or better known.

So much, then, for the principles by which I have been guided in the selection of facts. As to the arrangement of the facts, I have taken the animal kingdom in ascending order, and endeavoured to give as full a sketch as the selected evidence at my disposal permitted of the psychology which is distinctive of each class, or order, and, in some cases, family, genus, or even species. The reason of my entering into greater detail with some natural groups than with others scarcely requires explanation. For it is almost needless to say that if the animal kingdom were classified with reference to Psychology instead of with reference to Anatomy, we should have a very different kind of zoological tree from that which is now given in our diagrams. There is, indeed, a general and, philosophically considered, most important parallelism running through the whole animal kingdom between structural affinity and mental development; but this parallelism is exceedingly rough, and to be traced only in broad outlines, so that although it is convenient for the purpose of definite arrangement to take the animal kingdom in the order presented by zoological classification, it would be absurd to restrict an inquiry into Animal Psychology by any considerations of the apparently disproportionate length and minute subdivision with which it is necessary to treat some of the groups. Anatomically, an ant or a bee does not require more consideration than a beetle or a fly; but psychologically there is need for as great a difference of treatment as there is in the not very dissimilar case of a monkey and a man.

Throughout the work my aim has been to arrive at definite principles rather than to chronicle mere incidents—an aim which will become more apparent when the work as a whole shall have been completed. Therefore it is that in the present volume I have endeavoured, as far as the nature and circumstances of the inquiry would permit, to suppress anecdote. Nevertheless, although I have nowhere introduced anecdotes for their own sake, I have found it unavoidable not to devote much the largest part of the present essay to their narration. Hence, with the double purpose of limiting the introduction of anecdotes as much as possible, and of not repeating more than I could help anecdotes already published, I have in all cases, where I could do so without detriment to my main object, given the preference to facts which have been communicated to me by friends and correspondents. And here I may fitly take the opportunity of expressing my thanks and obligations to the latter, who in astonishing numbers have poured in their communications during several years from all quarters of the globe. I make this statement because I desire to explain to all my correspondents who may read this book, that I am not the less sensible of their kindness because its bounty has rendered it impossible for me to send acknowledgments in individual cases. However, I should like to add in this connection that it does not follow, because I have only quoted a small percentage of the letters which I have received, that all of the remainder have been useless. On the contrary, many of these have served to convey information and suggestions which, even if not reserved for express quotation in my forthcoming work, have been of use in guiding my judgment on particular points. Therefore I hope that the publication of these remarks may serve to swell the stream of communications into a yet larger flow.[1]

In all cases where I have occasion to quote statements of fact, which in the present treatise are necessarily numerous, I have made a point of trying to quote verbatim. Only where I have found that the account given by an author or a correspondent might profitably admit of a considerable degree of condensation have I presented it in my own words.

And here I have to express my very special obligations to Mr. Darwin, who not only assisted me in the most generous manner with his immense stores of information, as well as with his valuable judgment on sundry points of difficulty, but has also been kind enough to place at my disposal all the notes and clippings on animal intelligence which he has been collecting for the last forty years, together with the original MS. of his wonderful chapter on 'Instinct.' This chapter, on being re-cast for the 'Origin of Species,' underwent so merciless an amount of compression that the original draft constitutes a rich store of hitherto unpublished material. In my second work I shall have occasion to draw upon this store more largely than in the present one, and it is needless to add that in all cases where I do draw upon it I shall be careful to state the source to which I am indebted.

[The above was written when I sent this work to the publishers several months ago, and I have thought it best to leave the concluding paragraph as it originally stood. But in making this explanation, I cannot allude to the calamity which has since occurred without paying my tribute, not alone to the memory of the greatest genius of our age, but still more, and much more, to the memory of a friend so inexpressibly noble, kind, and generous, that even my immense admiration of the naturalist was surpassed by my loving veneration for the man.]

CONTENTS.

INTRODUCTION.

Before we begin to consider the phenomena of mind throughout the animal kingdom it is desirable that we should understand, as far as possible, what it is that we exactly mean by mind. Now, by mind we may mean two very different things, according as we contemplate it in our own individual selves, or in other organisms. For if we contemplate our own mind, we have an immediate cognizance of a certain flow of thoughts or feelings, which are the most ultimate things, and indeed the only things, of which we are cognisant. But if we contemplate mind in other persons or organisms, we have no such immediate cognizance of thoughts or feelings. In such cases we can only infer the existence and the nature of thoughts and feelings from the activities of the organisms which appear to exhibit them. Thus it is that we may have a subjective analysis of mind and an objective analysis of mind—the difference between the two consisting in this, that in our subjective analysis we are restricted to the limits of a single isolated mind which we call our own, and within the territory of which we have immediate cognizance of all the processes that are going on, or at any rate of all the processes that fall within the scope of our introspection. But in our objective analysis of other or foreign minds we have no such immediate cognizance; all our knowledge of their operations is derived, as it were, through the medium of ambassadors—these ambassadors being the activities of the organism. Hence it is evident that in our study of animal intelligence we are wholly restricted to the objective method. Starting from what I know subjectively of the operations of my own individual mind, and the activities which in my own organism they prompt, I proceed by analogy to infer from the observable activities of other organisms what are the mental operations that underlie them.

Now, in this mode of procedure what is the kind of activities which may be regarded as indicative of mind? I certainly do not so regard the flowing of a river or the blowing of the wind. Why? First, because the objects are too remote in kind from my own organism to admit of my drawing any reasonable analogy between them and it; and, secondly, because the activities which they present are of invariably the same kind under the same circumstances; they afford no evidence of feeling or purpose. In other words, two conditions require to be satisfied before we even begin to imagine that observable activities are indicative of mind: first, the activities must be displayed by a living organism; and secondly, they must be of a kind to suggest the presence of two elements which we recognise as the distinctive characteristics of mind as such—consciousness and choice.

So far, then, the case seems simple enough. Wherever we see a living organism apparently exerting intentional choice, we might infer that it is conscious choice, and therefore that the organism has a mind. But further reflection shows us that this is just what we cannot do; for although it is true that there is no mind without the power of conscious choice, it is not true that all apparent choice is due to mind. In our own organisms, for instance, we find a great many adaptive movements performed without choice or even consciousness coming into play at all—such, for instance, as in the beating of our hearts. And not only so, but physiological experiments and pathological lesions prove that in our own and in other organisms the mechanism of the nervous system is sufficient, without the intervention of consciousness, to produce muscular movements of a highly co-ordinate and apparently intentional character. Thus, for instance, if a man has his back broken in such a way as to sever the nervous connection between his brain and lower extremities, on pinching or tickling his feet they are drawn suddenly away from the irritation, although the man is quite unconscious of the adaptive movement of his muscles; the lower nerve-centres of the spinal cord are competent to bring about this movement of adaptive response without requiring to be directed by the brain. This non-mental operation of the lower nerve-centres in the production of apparently intentional movements is called Reflex Action, and the cases of its occurrence, even within the limits of our own organism, are literally numberless. Therefore, in view of such non-mental nervous adjustment, leading to movements which are only in appearance intentional, it clearly becomes a matter of great difficulty to say in the case of the lower animals whether any action which appears to indicate intelligent choice is not really action of the reflex kind.

On this whole subject of mind-like and yet not truly mental action I shall have much to say in my subsequent treatise, where I shall be concerned among other things with tracing the probable genesis of mind from non-mental antecedents. But here it is sufficient merely to make this general statement of the fact, that even within the experience supplied by our own organisms adaptive movements of a highly complex and therefore apparently purposive character may be performed without any real purpose, or even consciousness of their performance. It thus becomes evident that before we can predicate the bare existence of mind in the lower animals, we need some yet more definite criterion of mind than that which is supplied by the adaptive actions of a living organism, howsoever apparently intentional such actions may be. Such a criterion I have now to lay down, and I think it is one that is as practically adequate as it is theoretically legitimate.

Objectively considered, the only distinction between adaptive movements due to reflex action and adaptive movements due to mental perception, consists in the former depending on inherited mechanisms within the nervous system being so constructed as to effect particular adaptive movements in response to particular stimulations, while the latter are independent of any such inherited adjustment of special mechanisms to the exigencies of special circumstances. Reflex actions under the influence of their appropriate stimuli may be compared to the actions of a machine under the manipulations of an operator; when certain springs of action are touched by certain stimuli, the whole machine is thrown into appropriate movement; there is no room for choice, there is no room for uncertainty; but as surely as any of these inherited mechanisms are affected by the stimulus with reference to which it has been constructed to act, so surely will it act in precisely the same way as it always has acted. But the case with conscious mental adjustment is quite different. For, without at present going into the question concerning the relation of body and mind, or waiting to ask whether cases of mental adjustment are not really quite as mechanical in the sense of being the necessary result or correlative of a chain of physical sequences due to a physical stimulation, it is enough to point to the variable and incalculable character of mental adjustments as distinguished from the constant and foreseeable character of reflex adjustments. All, in fact, that in an objective sense we can mean by a mental adjustment is an adjustment of a kind that has not been definitely fixed by heredity as the only adjustment possible in the given circumstances of stimulation. For were there no alternative of adjustment, the case, in an animal at least, would be indistinguishable from one of reflex action.

It is, then, adaptive action by a living organism in cases where the inherited machinery of the nervous system does not furnish data for our prevision of what the adaptive action must necessarily be—it is only here that we recognise the objective evidence of mind. The criterion of mind, therefore, which I propose, and to which I shall adhere throughout the present volume, is as follows:—Does the organism learn to make new adjustments, or to modify old ones, in accordance with the results of its own individual experience? If it does so, the fact cannot be due merely to reflex action in the sense above described, for it is impossible that heredity can have provided in advance for innovations upon, or alterations of, its machinery during the lifetime of a particular individual.

In my next work I shall have occasion to consider this criterion of mind more carefully, and then it will be shown that as here stated the criterion is not rigidly exclusive, either, on the one hand, of a possibly mental element in apparently non-mental adjustments, or, conversely, of a possibly non-mental element in apparently mental adjustments. But, nevertheless, the criterion is the best that is available, and, as it will be found sufficient for all the purposes of the present work, its more minute analysis had better be deferred till I shall have to treat of the probable evolution of mind from non-mental antecedents. I may, however, here explain that in my use of this criterion I shall always regard it as fixing only the upper limit of non-mental action; I shall never regard it as fixing the lower limit of mental action. For it is clear that long before mind has advanced sufficiently far in the scale of development to become amenable to the test in question, it has probably begun to dawn as nascent subjectivity. In other words, because a lowly organised animal does not learn by its own individual experience, we may not therefore conclude that in performing its natural or ancestral adaptations to appropriate stimuli consciousness, or the mind-element, is wholly absent; we can only say that this element, if present, reveals no evidence of the fact. But, on the other hand, if a lowly organised animal does learn by its own individual experience, we are in possession of the best available evidence of conscious memory leading to intentional adaptation. Therefore our criterion applies to the upper limit of non-mental action, not to the lower limit of mental.

Of course to the sceptic this criterion may appear unsatisfactory, since it depends, not on direct knowledge, but on inference. Here, however, it seems enough to point out, as already observed, that it is the best criterion available; and further, that scepticism of this kind is logically bound to deny evidence of mind, not only in the case of the lower animals, but also in that of the higher, and even in that of men other than the sceptic himself. For all objections which could apply to the use of this criterion of mind in the animal kingdom would apply with equal force to the evidence of any mind other than that of the individual objector. This is obvious, because, as I have already observed, the only evidence we can have of objective mind is that which is furnished by objective activities; and as the subjective mind can never become assimilated with the objective so as to learn by direct feeling the mental processes which there accompany the objective activities, it is clearly impossible to satisfy any one who may choose to doubt the validity of inference, that in any case other than his own mental processes ever do accompany objective activities. Thus it is that philosophy can supply no demonstrative refutation of idealism, even of the most extravagant form. Common sense, however, universally feels that analogy is here a safer guide to truth than the sceptical demand for impossible evidence; so that if the objective existence of other organisms and their activities is granted—without which postulate comparative psychology, like all the other sciences, would be an unsubstantial dream—common sense will always and without question conclude that the activities of organisms other than our own, when analogous to those activities of our own which we know to be accompanied by certain mental states, are in them accompanied by analogous mental states.

The theory of animal automatism, therefore, which is usually attributed to Descartes (although it is not quite clear how far this great philosopher really entertained the theory), can never be accepted by common sense; and even as a philosophical speculation it will be seen, from what has just been said, that by no feat of logic is it possible to make the theory apply to animals to the exclusion of man. The expression of fear or affection by a dog involves quite as distinctive and complex a series of neuro-muscular actions as does the expression of similar emotions by a human being; and therefore, if the evidence of corresponding mental states is held to be inadequate in the one case, it must in consistency be held similarly inadequate in the other. And likewise, of course, with all other exhibitions of mental life.

It is quite true, however, that since the days of Descartes—or rather, we might say, since the days of Joule—the question of animal automatism has assumed a new or more defined aspect, seeing that it now runs straight into the most profound and insoluble problem that has ever been presented to human thought—viz. the relation of body to mind in view of the doctrine of the conservation of energy. I shall subsequently have occasion to consider this problem with the close attention that it demands; but in the present volume, which has to deal only with the phenomena of mind as such, I expressly pass the problem aside as one reserved for separate treatment. Here I desire only to make it plain that the mind of animals must be placed in the same category, with reference to this problem, as the mind of man; and that we cannot without gross inconsistency ignore or question the evidence of mind in the former, while we accept precisely the same kind of evidence as sufficient proof of mind in the latter.

And this proof, as I have endeavoured to show, is in all cases and in its last analysis the fact of a living organism showing itself able to learn by its own individual experience. Wherever we find an animal able to do this, we have the same right to predicate mind as existing in such an animal that we have to predicate it as existing in any human being other than ourselves. For instance, a dog has always been accustomed to eat a piece of meat when his organism requires nourishment, and when his olfactory nerves respond to the particular stimulus occasioned by the proximity of the food. So far, it may be said, there is no evidence of mind; the whole series of events comprised in the stimulations and muscular movements may be due to reflex action alone. But now suppose that by a number of lessons the dog has been taught not to eat the meat when he is hungry until he receives a certain verbal signal: then we have exactly the same kind of evidence that the dog's actions are prompted by mind as we have that the actions of a man are so prompted.[2] Now we find that the lower down we go in the animal kingdom, the more we observe reflex action, or non-mental adjustment, to predominate over volitional action, or mental adjustment. That is to say, the lower down we go in the animal kingdom, the less capacity do we find for changing adjustive movements in correspondence with changed conditions; it becomes more and more hopeless to teach animals—that is, to establish associations of ideas; and the reason of this, of course, is that ideas or mental units become fewer and less definite the lower we descend through the structure of mind.

It is not my object in the present work to enter upon any analysis of the operations of mind, as this will require to be done as fully as possible in my next work. Nevertheless, a few words must here be said with regard to the main divisions of mental operation, in order to define closely the meanings which I shall attach to certain terms relating to these divisions, and the use of which I cannot avoid.

The terms sensation, perception, emotion, and volition need not here be considered. I shall use them in their ordinary psychological significations; and although I shall subsequently have to analyse each of the organic or mental states which they respectively denote, there will be no occasion in the present volume to enter upon this subject. I may, however, point out one general consideration to which I shall throughout adhere. Taking it for granted that the external indications of mental processes which we observe in animals are trustworthy, so that we are justified in inferring particular mental states from particular bodily actions, it follows that in consistency we must everywhere apply the same criteria.

For instance, if we find a dog or a monkey exhibiting marked expressions of affection, sympathy, jealousy, rage, &c., few persons are sceptical enough to doubt that the complete analogy which these expressions afford with those which are manifested by man, sufficiently prove the existence of mental states analogous to those in man of which these expressions are the outward and visible signs. But when we find an ant or a bee apparently exhibiting by its actions these same emotions, few persons are sufficiently non-sceptical not to doubt whether the outward and visible signs are here trustworthy as evidence of analogous or corresponding inward and mental states. The whole organisation of such a creature is so different from that of a man that it becomes questionable how far analogy drawn from the activities of the insect is a safe guide to the inferring of mental states—particularly in view of the fact that in many respects, such as in the great preponderance of 'instinct' over 'reason,' the psychology of an insect is demonstrably a widely different thing from that of a man. Now it is, of course, perfectly true that the less the resemblance the less is the value of any analogy built upon the resemblance, and therefore that the inference of an ant or a bee feeling sympathy or rage is not so valid as is the similar inference in the case of a dog or a monkey. Still it is an inference, and, so far as it goes, a valid one—being, in fact, the only inference available. That is to say, if we observe an ant or a bee apparently exhibiting sympathy or rage, we must either conclude that some psychological state resembling that of sympathy or rage is present, or else refuse to think about the subject at all; from the observable facts there is no other inference open. Therefore, having full regard to the progressive weakening of the analogy from human to brute psychology as we recede through the animal kingdom downwards from man, still, as it is the only analogy available, I shall follow it throughout the animal series.

It may not, however, be superfluous to point out that if we have full regard to this progressive weakening of the analogy, we must feel less and less certain of the real similarity of the mental states compared; so that when we get down as low as the insects, I think the most we can confidently assert is that the known facts of human psychology furnish the best available pattern of the probable facts of insect psychology. Just as the theologians tell us—and logically enough—that if there is a Divine Mind, the best, and indeed only, conception we can form of it is that which is formed on the analogy, however imperfect, supplied by the human mind; so with 'inverted anthropomorphism' we must apply a similar consideration with a similar conclusion to the animal mind. The mental states of an insect may be widely different from those of a man, and yet most probably the nearest conception that we can form of their true nature is that which we form by assimilating them to the pattern of the only mental states with which we are actually acquainted. And this consideration, it is needless to point out, has a special validity to the evolutionist, inasmuch as upon his theory there must be a psychological, no less than a physiological, continuity extending throughout the length and breadth of the animal kingdom.

In these preliminary remarks only one other point requires brief consideration, and this has reference to the distinction between what in popular phraseology is called 'Instinct' and 'Reason.' I shall not here enter upon any elaborate analysis of a distinction which is undoubtedly valid, but shall confine my remarks to explaining the sense in which I shall everywhere use these terms.

Few words in our language have been subject to a greater variety of meanings than the word instinct. In popular phraseology, descended from the Middle Ages, all the mental faculties of the animal are termed instinctive, in contradistinction to those of man, which are termed rational. But unless we commit ourselves to an obvious reasoning in a circle, we must avoid assuming that all actions of animals are instinctive, and then arguing that because they are instinctive, therefore they differ from the rational actions of man. The question really lies in what is here assumed, and we can only answer it by examining in what essential respect instinct differs from reason.

Again, Addison says:—

I look upon instinct as upon the principle of gravitation in bodies, which is not to be explained by any known qualities inherent in the bodies themselves, nor from any laws of mechanism, but as an immediate impression from the first Mover, and the Divine energy acting in the creatures.

This mode of 'looking upon instinct' is merely to exclude the subject from the sphere of inquiry, and so to abstain from any attempt at definition.

Innumerable other opinions might be quoted from well-known writers, 'looking upon instinct' in widely different ways; but as this is not an historical work, I shall pass on at once to the manner in which science looks upon it, or, at least, the manner in which it will always be looked upon throughout the present work.

Without concerning ourselves with the origin of instincts, and so without reference to the theory of evolution, we have to consider the most conspicuous and distinctive features of instinct as it now exists. The most important point to observe in the first instance is that instinct involves mental operations; for this is the only point that serves to distinguish instinctive action from reflex. Reflex action, as already explained, is non-mental neuro-muscular adaptation to appropriate stimuli; but instinctive action is this and something more; there is in it the element of mind. Such, at least, is instinctive action in the sense that I shall always allude to it. I am, of course, aware that the limitation which I thus impose is one which is ignored, or not recognised, by many writers even among psychologists; but I am persuaded that if we are to have any approach to definiteness in the terms which we employ—not to say of clearness in our ideas concerning the things of which we speak—it is most desirable to restrict the word instinct to mental as distinguished from non-mental activity. No doubt it is often difficult, or even impossible, to decide whether or not a given action implies the presence of the mind-element—i.e., conscious as distinguished from unconscious adaptation; but this is altogether a separate matter, and has nothing to do with the question of defining instinct in a manner which shall be formally exclusive, on the one hand of reflex action, and on the other of reason. As Virchow truly observes, 'it is difficult or impossible to draw the line between instinctive and reflex action;' but at least the difficulty may be narrowed down to deciding in particular cases whether or not an action falls into this or that category of definition; there is no reason why the difficulty should arise on account of any ambiguity of the definitions themselves. Therefore I endeavour to draw as sharply as possible the line which in theory should be taken to separate instinctive from reflex action; and this line, as I have already said, is constituted by the boundary of non-mental or unconscious adjustment, with adjustment in which there is concerned consciousness or mind.

Having thus, I hope, made it clear that the difficulty of drawing a distinction between reflex and instinctive actions as a class is one thing, and that the difficulty of assigning particular actions to one or the other of our categories is another thing, we may next perceive that the former difficulty is obviated by the distinction which I have imposed, and that the latter only arises from the fact that on the objective side there is no distinction imposable. The former difficulty is obviated by the distinction which I have drawn, simply because the distinction is itself a definite one. In particular cases of adjustive action we may not always be able to affirm whether consciousness of their performance is present or absent; but, as I have already said, this does not affect the validity of our definition; all we can say of such cases is that if the performance in question is attended with consciousness it is instinctive, and if not it is reflex.

And the difficulty of assigning particular actions to one or other of these two categories arises, as I have said, merely because on the objective side, or the side of the nervous system, there is no distinction to be drawn. Whether or not a neural process is accompanied by a mental process, it is in itself the same. The advent and development of consciousness, although progressively converting reflex action into instinctive, and instinctive into rational, does this exclusively in the sphere of subjectivity; the nervous processes engaged are throughout the same in kind, and differ only in the relative degrees of their complexity. Therefore, as the dawn of consciousness or the rise of the mind-element is gradual and undefined, both in the animal kingdom and in the growing child, it is but necessary that in the early morning, as it were, of consciousness any distinction between the mental and the non-mental should be obscure, and generally impossible to determine. Thus, for instance, a child at birth does not close its eyes upon the near approach of a threatening body, and it only learns to do so by degrees as the result of experience; at first, therefore, the action of closing the eyelids in order to protect the eyes may be said to be instinctive, in that it involves the mind-element:[3] yet it afterwards becomes a reflex which asserts itself even in opposition to the will. And, conversely, sucking in a new-born child, or a child in utero, is, in accordance with my definition, a reflex action; yet in later life, when consciousness becomes more developed and the child seeks the breast, sucking may properly be called an instinctive action. Therefore it is that, as in the ascending scale of objective complexity the mind-element arises and advances gradually, many particular cases which occupy the undefined boundary between reflex action and instinct cannot be assigned with confidence either to the one region or to the other.

We see then the point, and the only point, wherein instinct can be consistently separated from reflex action; viz., in presenting a mental constituent. Next we must consider wherein instinct may be separated from reason. And for this purpose we may best begin by considering what we mean by reason.

The term 'reason' is used in significations almost as various as those which are applied to 'instinct.' Sometimes it stands for all the distinctively human faculties taken collectively, and in antithesis to the mental faculties of the brute; while at other times it is taken to mean the distinctively human faculties of intellect.

Dr. Johnson defines it as 'the power by which man deduces one proposition from another, and proceeds from premises to consequences.' This definition presupposes language, and therefore ignores all cases of inference not thrown into the formal shape of predication. Yet even in man the majority of inferences drawn by the mind never emerge as articulate propositions; so that although, as we shall have occasion fully to observe in my subsequent work, there is much profound philosophy in identifying reason with speech as they were identified in the term Logos, yet for purposes of careful definition so to identify intellect with language is clearly a mistake.

More correctly, the word reason is used to signify the power of perceiving analogies or ratios, and is in this sense equivalent to the term 'ratiocination,' or the faculty of deducing inferences from a perceived equivalency of relations. Such is the only use of the word that is strictly legitimate, and it is thus that I shall use it throughout the present treatise. This faculty, however, of balancing relations, drawing inferences, and so of forecasting probabilities, admits of numberless degrees; and as in the designation of its lower manifestations it sounds somewhat unusual to employ the word reason, I shall in these cases frequently substitute the word intelligence. Where we find, for instance, that an oyster profits by individual experience, or is able to perceive new relations and suitably to act upon the result of its perceptions, I think it sounds less unusual to speak of the oyster as displaying intelligence than as displaying reason. On this account I shall use the former term to signify the lower degrees of the ratiocinative faculty; and thus in my usage it will be opposed to such terms as instinct, reflex action, &c., in the same manner as the term reason is so opposed. This is a point which, for the sake of clearness, I desire the reader to retain in his memory. I shall always speak of intelligence and intellect in antithesis to instinct, emotion, and the rest, as implying mental faculties the same in kind as those which in ourselves we call rational.

Now it is notorious that no distinct line can be drawn between instinct and reason. Whether we look to the growing child or to the ascending scale of animal life, we find that instinct shades into reason by imperceptible degrees, or, as Pope expresses it, that these principles are 'for ever separate, yet for ever near.' Nor is this other than the principles of evolution would lead us to expect, as I shall afterwards have abundant occasion to show. Here, however, we are only concerned with drawing what distinction we can between instinct and reason as these faculties are actually presented to our observation. And this in a general way it is not difficult to do.

We have seen that instinct involves 'mental operations,' and that by this feature it is distinguished from reflex action; we have now to consider the features by which it is distinguished from reason. These are accurately, though not completely, conveyed by Sir Benjamin Brodie, who defines instinct as 'a principle by which animals are induced, independently of experience and reasoning, to the performances of certain voluntary acts, which are necessary to their preservation as individuals, or to the continuance of the species, or in some other way convenient to them.'[4] This definition, as I have said, is accurate as far as it goes, but it does not state with sufficient generality and terseness that all instinctive action is adaptive; nor does it clearly bring out the distinction between instinct and reason which is thus well conveyed by the definition of Hartmann, who says in his 'Philosophy of the Unconscious,' that 'instinct is action taken in pursuance of an end, but without conscious perception of what the end is.' This definition, however, is likewise defective in that it omits another of the important differentiæ of instinct—namely, the uniformity of instinctive action as performed by different individuals of the same species. Including this feature, therefore, we may more accurately and completely define instinct as mental action (whether in animals or human beings), directed towards the accomplishing of adaptive movement, antecedent to individual experience, without necessary knowledge of the relation between the means employed and the ends attained, but similarly performed under the same appropriate circumstances by all the individuals of the same species. Now in every one of these respects, with the exception of containing a mental constituent and in being concerned in adaptive action, instinct differs from reason. For reason, besides involving a mental constituent, and besides being concerned in adaptive action, is always subsequent to individual experience, never acts but upon a definite and often laboriously acquired knowledge of the relation between means and ends, and is very far from being always similarly performed under the same appropriate circumstances by all the individuals of the same species.

Thus the distinction between instinct and reason is both more definite and more manifold than is that between instinct and reflex action. Nevertheless, in particular cases there is as much difficulty in classifying certain actions as instinctive or rational, as there is in cases where the question lies between instinct and reflex action. And the explanation of this is, as already observed, that instinct passes into reason by imperceptible degrees; so that actions in the main instinctive are very commonly tempered with what Pierre Huber calls 'a little dose of judgment or reason,' and vice versâ. But here, again, the difficulty which attaches to the classification of particular actions has no reference to the validity of the distinctions between the two classes of actions; these are definite and precise, whatever difficulty there may be in applying them to particular cases.

Another point of difference between instinct and reason may be noticed which, although not of invariable, is of very general applicability. It will have been observed, from what has already been said, that the essential respect in which instinct differs from reason consists in the amount of conscious deliberation which the two processes respectively involve. Instinctive actions are actions which, owing to their frequent repetition, become so habitual in the course of generations that all the individuals of the same species automatically perform the same actions under the stimulus supplied by the same appropriate circumstances. Rational actions, on the other hand, are actions which are required to meet circumstances of comparatively rare occurrence in the life-history of the species, and which therefore can only be performed by an intentional effort of adaptation. Consequently there arises the subordinate distinction to which I allude, viz., that instinctive actions are only performed under particular circumstances which have been frequently experienced during the life-history of the species; whereas rational actions are performed under varied circumstances, and serve to meet novel exigencies which may never before have occurred even in the life-history of the individual.

Thus, then, upon the whole, we may lay down our several definitions in their most complete form.

Reflex action is non-mental neuro-muscular adjustment, due to the inherited mechanism of the nervous system, which is formed to respond to particular and often recurring stimuli, by giving rise to particular movements of an adaptive though not of an intentional kind.

Instinct is reflex action into which there is imported the element of consciousness. The term is therefore a generic one, comprising all those faculties of mind which are concerned in conscious and adaptive action, antecedent to individual experience, without necessary knowledge of the relation between means employed and ends attained, but similarly performed under similar and frequently recurring circumstances by all the individuals of the same species.

Reason or intelligence is the faculty which is concerned in the intentional adaptation of means to ends. It therefore implies the conscious knowledge of the relation between means employed and ends attained, and may be exercised in adaptation to circumstances novel alike to the experience of the individual and to that of the species.

CHAPTER I.

APPLICATION OF THE FOREGOING PRINCIPLES TO THE LOWEST ANIMALS.

Protozoa.

No one can have watched the movements of certain Infusoria without feeling it difficult to believe that these little animals are not actuated by some amount of intelligence. Even if the manner in which they avoid collisions be attributed entirely to repulsions set up in the currents which by their movements they create, any such mechanical explanation certainly cannot apply to the small creatures seeking one another for the purposes of prey, reproduction, or, as it sometimes seems, of mere sport. There is a common and well-known rotifer whose body is of a cup shape, provided with a very active tail, which is armed at its extremity with strong forceps. I have seen a small specimen of this rotifer seize a much larger one with its forceps, and attach itself by this means to the side of the cup. The large rotifer at once became very active, and swinging about with its burden until it came to a piece of weed, it took firm hold of the weed with its own forceps, and began the most extraordinary series of movements, which were obviously directed towards ridding itself of the encumbrance. It dashed from side to side in all directions with a vigour and suddenness which were highly astonishing, so that it seemed as if the animalcule would either break its forceps or wrench its tail from its body. No movements could possibly be better suited to jerk off the offending object, for the energy with which the jerks were given, now in one direction and now in another, were, as I have said, most surprising. But not less surprising was the tenacity with which the smaller rotifer retained its hold; for although one might think that it was being almost jerked to pieces, after each bout of jerking it was seen to be still attached. This trial of strength, which must have involved an immense expenditure of energy in proportion to the size of the animals, lasted for several minutes, till eventually the small rotifer was thrown violently away. It then returned to the conflict, but did not succeed a second time in establishing its hold. The entire scene was as like intelligent action on the part of both animals as could well be imagined, so that if we were to depend upon appearances alone, this one observation would be sufficient to induce me to attribute conscious determination to these microscopical organisms.

But, without denying that conscious determination may here be present, or involving ourselves in the impossible task of proving such a negative, we may properly affirm that until an animalcule shows itself to be teachable by individual experience, we have no sufficient evidence derived or derivable from any number of such apparently intelligent movements, that conscious determination is present. Therefore, I need not wait to quote the observations of the sundry microscopists who detail facts more or less similar to the above, with expressions of their belief that microscopical organisms display a certain degree of instinct or intelligence as distinguished from mechanical, or wholly non-mental adjustment. But there are some observations relating to the lowest of all animals, and made by a competent person, which are so remarkable that I shall have to quote them in full. These observations are recorded by Mr. H. J. Carter, F.R.S., in the 'Annals of Natural History,' and in his opinion prove that the beginnings of instinct are to be found so low down in the scale as the Rhizopoda. He says:—'Even Athealium will confine itself to the water of the watch-glass in which it may be placed when away from sawdust and chips of wood among which it has been living; but if the watch-glass be placed upon the sawdust, it will very soon make its way over the side of the watch-glass and get to it.'

This is certainly a remarkable observation: for it seems to show that the rhizopod distinguishes the presence of the sawdust outside the watch-glass, and crawls over the brim of the latter in order to get into more congenial quarters, while it is contented with the water in the watch-glass so long as there is no sawdust outside. But to proceed:

On one occasion, while investigating the nature of some large, transparent, spore-like elliptical cells (fungal?) whose protoplasm was rotating, while it was at the same time charged with triangular grains of starch, I observed some actinophorous rhizopods creeping about them, which had similarly shaped grains of starch in their interior; and having determined the nature of these grains in both by the addition of iodine, I cleansed the glasses, and placed under the microscope a new portion of the sediment from the basin containing these cells and actinophryans for further examination, when I observed one of the spore-like cells had become ruptured, and that a portion of its protoplasm, charged with the triangular starch-grains, was slightly protruding through the crevice. It then struck me that the actinophryans had obtained their starch-grains from this source; and while looking at the ruptured cell, an actinophrys made its appearance, and creeping round the cell, at last arrived at the crevice, from which it extricated one of the grains of starch mentioned, and then crept off to a good distance. Presently, however, it returned to the same cell; and although there were now no more starch-grains protruding, the actinophrys managed again to extract one from the interior through the crevice. All this was repeated several times, showing that the actinophrys instinctively knew that those were nutritious grains, that they were contained in this cell, and that, although each time after incepting a grain it went away to some distance, it knew how to find its way back to the cell again which furnished this nutriment.

On another occasion I saw an actinophrys station itself close to a ripe spore-cell of pythium, which was situated upon a filament of Spirogyra crassa; and as the young ciliated monadic germs issued forth, one after another, from the dehiscent spore-cell, the actinophrys remained by it and caught every one of them, even to the last, when it retired to another part of the field, as if instinctively conscious that there was nothing more to be got at the old place.

But by far the greatest feat of this kind that ever presented itself to me was the catching of a young acineta by an old sluggish amœba, as the former left its parent; and this took place as follows:—

In the evening of the 2nd of June, 1858, in Bombay, while looking through a microscope at some Euglenæ, &c., which had been placed aside for examination in a watch-glass, my eye fell upon a stalked and triangular acineta (A. mystacina?), around which an amœba was creeping and lingering, as they do when they are in quest of food. But knowing the antipathy that the amœba, like almost every other infusorian, has to the tentacles of the acineta, I concluded that the amœba was not encouraging an appetite for its whiskered companion, when I was surprised to find that it crept up the stem of the acineta, and wound itself round its body. This mark of affection, too much like that frequently evinced at the other end of the scale, even where there is a mind for its control, did not long remain without interpretation. There was a young acineta, tender, and without poisonous tentacles (for they are not developed at birth), just ready to make its exit from the parent, an exit which takes place so quickly, and is followed by such rapid bounding movements of the non-ciliated acineta, that who would venture to say, à priori, that a dull, heavy, sluggish amœba could catch such an agile little thing? But the amœba are as unerring and unrelaxing in their grasp as they are unrelenting in their cruel inceptions of the living and the dead, when they serve them for nutrition; and thus the amœba, placing itself round the ovarian aperture of the acineta, received the young one, nurse-like, in its fatal lap, incepted it, descended from the parent, and crept off. Being unable to conceive at the time that this was such an act of atrocity on the part of the amœba as the sequel disclosed, and thinking that the young acineta might yet escape, or pass into some other form in the body of its host, I watched the amœba for some time afterwards, until the tale ended by the young acineta becoming divided into two parts, and thus in their respective digestive spaces ultimately becoming broken down and digested.[5]

With regard to these remarkable observations it can only, I think, be said that although certainly very suggestive of something more than mechanical response to stimulation, they are not sufficiently so to justify us in ascribing to these lowest members of the zoological scale any rudiment of truly mental action. The subject, however, is here full of difficulty, and not the least so on account of the amœba not only having no nervous system, but no observable organs of any kind; so that, although we may suppose that the adaptive movements described by Mr. Carter were non-mental, it still remains wonderful that these movements should be exhibited by such apparently unorganised creatures, seeing that as to the remoteness of the end attained, no less than the complex refinement of the stimulus to which their adaptive response was due, the movements in question rival the most elaborate of non-mental adjustments elsewhere performed by the most highly organised of nervous systems.

Cœlenterata.

Dr. Eimer attributes 'voluntary action' to the Medusæ, and indeed draws a sharp distinction between what he considers their 'involuntary' and 'voluntary' movements. In this distinction, however, I do not at all concur; for although I am well acquainted with the difference between the active and slow rhythm upon which the distinction is founded, I see no evidence whatever for supposing that the difference involves any psychological element. The active swimming is produced by stimulation, and is no doubt calculated to lead to the escape of the organism; but this fact certainly does not carry us beyond the ordinary possibilities of reflex action. And even when, as in some species is constantly the case, bouts of active swimming appear to arise spontaneously or without observable stimulation, the fact is to be attributed to a liberation of overplus ganglionic energy, or to some unobservable stimulation; it does not justify the supposition of any psychical element being concerned.[6]

M'Crady gives an interesting account of a medusa which carries its larvæ on the inner sides of its bell-shaped body. The manubrium, or mobile digestive cavity of the animal, depends, as in the other Medusæ, from the summit of the concave surface of the bell, like a clapper or tongue. Now M'Crady observed this depending organ to be moved first to one side and then to the other side of the bell, in order to give suck to the larvæ on the sides of the bell—the larvæ dipping their long noses into the nutrient fluids which that organ of the parent's body contained. I cite this case, because if it occurred in one of the higher animals it would probably be called a case of instinct; but as it occurs in so low an animal as a jelly-fish, it is unreasonable to suppose that intelligence can ever have played any part in originating the action. Therefore we may set it down as the uncompounded result of natural selection.

Some species of medusæ—notably Sarsia—seek the light, crowding into the path of a beam, and following it actively if moved. They derive advantage from so doing, because certain small crustacea on which they feed likewise crowd into the light. The seeking of light by these medusæ is therefore doubtless of the nature of a reflex action which has been developed by natural selection in order to bring the animals into contact with their prey. Paul Bert has found that Daphnia pulex seeks the light (especially the yellow ray), and Engelmann has observed the same fact with regard to certain protoplasmic organisms. But in none of these or other such cases is there any evidence of a psychical element being concerned in the process.

Echinodermata.

Some of the natural movements of these animals, as also some of their movements under stimulation, are very suggestive of purpose; but I have satisfied myself that there is no adequate evidence of the animals being able to profit by individual experience, and therefore, in accordance with our canon, that there is no adequate evidence of their exhibiting truly mental phenomena. On the other hand, the study of reflex action in these organisms is full of interest—so much so that in my next work I shall take them as typical organisms in this connection.[7]

Annelida.

Mr. Darwin has now in the press a highly interesting work on the habits of earth-worms. It appears from his observations that the manner in which these animals draw down leaves, &c., into their burrows is strongly indicative of instinctive action, if not of intelligent purpose—seeing that they always lay hold of the part of the leaf (even though an exotic one) by the traction of which the leaf will offer least resistance to being drawn down. But as this work will so shortly be published, I shall not forestall any of the facts which it has to state, nor should I yet like to venture an opinion as to how far these facts, when considered altogether, would justify any inference to a truly mental element as existing in these animals.

Of the land leeches in Ceylon, Sir E. Tennent gives an account which likewise seems to bespeak intelligence as occurring in annelids. He says:—

In moving, the land leeches have the power of planting one extremity on the earth and raising the other perpendicularly to watch for their victim. Such is their vigilance and instinct, that on the approach of a passer-by to a spot which they infest, they may be seen amongst the grass and fallen leaves on the edge of a native path, poised erect, and preparing for their attack on man and horse. On descrying their prey they advance rapidly by semicircular strides, fixing one end firmly and arching the other forwards, till by successive advances they can lay hold of the traveller's foot, when they disengage themselves from the ground and ascend his dress in search of an aperture to enter. In these encounters the individuals in the rear of a party of travellers in the jungle invariably fare worst, as the leeches, once warned of their approach, congregate with singular celerity.[8]

CHAPTER II.

MOLLUSCA.

I shall treat of the Mollusca before the Articulata, because as a group their intelligence is not so high. Indeed, it is not to be expected that the class of animals wherein the 'vegetative' functions of nutrition and reproduction predominate so largely over the animal functions of sensation, locomotion, &c., should present any considerable degree of intelligence. Nevertheless, in the only division of the group which has sense organs and powers of locomotion highly developed—viz., the Cephalopoda—we meet with large cephalic ganglia, and, it would appear, with no small development of intelligence. Taking, however, the sub-kingdom in ascending order, I shall first present all the trustworthy evidence that I have been able to collect, pointing to the highest level of intelligence that is attained by the lower members.

The following is quoted from Mr. Darwin's MS.:—

Even the headless oyster seems to profit from experience, for Dicquemase ('Journal de Physique,' vol. xxviii. p. 244) asserts that oysters taken from a depth never uncovered by the sea, open their shells, lose the water within, and perish; but oysters taken from the same place and depth, if kept in reservoirs, where they are occasionally left uncovered for a short time, and are otherwise incommoded, learn to keep their shells shut, and then live for a much longer time when taken out of the water.[9]

Some evidence of intelligence seems to be displayed by the razor-fish. For the animals dislike salt, so that when this is sprinkled above their burrows in the sand, they come to the surface and quit their habitations. But if the animal is once seized when it comes to the surface and afterwards allowed to retire into its burrow, no amount of salt will force it again to come to the surface.[10]

With regard to snails, L. Agassiz writes: 'Quiconque a eu l'occasion d'observer les amours des limaçons, ne saurait mettre en doute la séduction déployée dans les mouvements et les allures qui préparent et accomplissent le double embrassement de ces hermaphrodites.'[11]

Again, Mr. Darwin's MS. quotes from Mr. W. White[12] a curious exhibition of intelligence in a snail, which does not seem to have admitted of mal-observation. This gentleman 'fixed a land-shell mouth uppermost in a chink of rock; in a short time the snail protruded itself to its utmost length, and, attaching its foot vertically above, tried to pull the shell out in a straight line. Not succeeding, it rested for a few minutes and then stretched out its body on the right side and pulled its utmost, but failed. Resting again, it protruded its foot on the left side, pulled with its full force, and freed the shell. This exertion of force in three directions, which seems so geometrically suitable, must have been intentional.'

If it is objected that snail shells must frequently be liable to be impeded by obstacles, and therefore that this display of manœuvring on the part of their occupants is to be regarded as a reflex, I may remark that here again we have one of those incessantly recurring cases where it is difficult to draw the line between intelligence and non-intelligence. For, granting that the action is to a certain extent mechanical, we must still recognise that the animal while executing it must have remembered each of the two directions in which it had pulled ineffectually before it began to pull in the third direction; and it is improbable that snail shells are so frequently caught in positions from which a pull in only one direction will release them, that natural selection would have developed a special instinct to try pulling successively in three directions at right angles to one another.

The only other instance that I have met with of the apparent display of intelligence in snails is the remarkable one which Mr. Darwin gives in his 'Descent of Man,' on the authority of Mr. Lonsdale. Although the interpretation which is assigned to the fact seems to me to go beyond anything that we should have reason to expect of snail intelligence, I cannot ignore a fact which stands upon the observation of so good an authority, and shall therefore quote it in Mr. Darwin's words:—

These animals appear also susceptible of some degree of permanent attachment: an accurate observer, Mr. Lonsdale, informs me that he placed a pair of land-snails (Helix pomatia), one of which was weakly, into a small and ill-provided garden. After a short time the strong and healthy individual disappeared, and was traced by its track of slime over a wall into an adjoining well-stocked garden. Mr. Lonsdale concluded that it had deserted its sickly mate; but after an absence of twenty-four hours it returned, and apparently communicated the result of its successful exploration, for both then started along the same track, and disappeared over the wall.[13]

In this case the fact must be accepted, seeing that it stands on the authority of an accurate observer, and is of so definite a kind as not to admit of mistake. Consequently we are shut up to the alternative of supposing the return of the healthy snail to its mate a mere accident, and their both going over the wall into the well-stocked garden another mere accident, or acquiescing in the interpretation which Mr. Darwin assigns. Now, if we look closely into the matter, the chances against the double accident in question are certainly so considerable as to render the former supposition almost impossible. On the other hand, there is evidence to prove, as I shall immediately show, that a not distantly allied animal is unquestionably able to remember a particular locality as its home, and habitually to return to this locality after feeding. Therefore, in view of this analogous and corroborative case, the improbability of the snail remembering for twenty-four hours the position of its mate is very much reduced; while the subsequent communication, if it took place, would only require to have been of the nature of 'follow me,' which, as we shall repeatedly find, is a degree of communicative ability which many invertebrated animals possess. Therefore, in view of these considerations, I incline to Mr. Darwin's opinion that the facts can only be explained by supposing them due to intelligence on the part of the snails. Thus considered, these facts are no doubt very remarkable; for they would appear to indicate not merely accurate memory of direction and locality for twenty-four hours, but also no small degree of something akin to 'permanent attachment,' and sympathetic desire that another should share in the good things which one has found.[14]

The case to which I have just alluded as proving beyond all doubt that some Gasteropoda are able to retain a very precise and accurate memory of locality, is that of the common limpet.

Mr. J. Clarke Hawkshaw publishes in the Journal of the Linnæan Society the following account of the habits in question:—

The holes in the chalk in which the limpets are often to be found are, I believe, excavated in a great measure by rasping from the lingual teeth, though I doubt whether the object is to form a cavity to shelter in, though the cavities, when formed, may be of use for that purpose. It must be of the greatest importance to a limpet that, in order that it may insure a firm adherence to the rock, its shell should fit the rock accurately; when the shell does fit the rock accurately, a small amount of muscular contraction of the animal would cause the shell to adhere so firmly to a smooth surface as to be practically immoveable without fracture. As the shells cannot be adapted daily to different forms of surface, the limpets generally return to the same place of attachment. I am sure this is the case with many; for I found shells perfectly adjusted to the uneven surfaces of flints, the growth of the shells being in some parts distorted and indented to suit inequalities in the surface of the flints. . . . .

I noticed signs that limpets prefer a hard, smooth surface to a pit in the chalk. On one surface of a large block, over all sides of which limpets were regularly and plentifully distributed, there were two flat fragments of a fossil shell about 3 inches by 4 inches, each embedded in the chalk. The chalk all round these fragments was free from limpets; but on the smooth surface of the pieces of shell they were packed as closely as they could be. I noticed another case, which almost amounts, to my mind, to a proof that they prefer a smooth surface to a hole. A limpet had formed a clearing on one of the sea-weed-covered blocks before referred to. In the midst of this clearing was a pedestal of flint rather more than one inch in diameter, standing up above the surface of the chalk; it projected so much that a tap from my hammer broke it off. On the top of the smooth fractured surface of this flint the occupant of the clearing had taken up its abode. The shell was closely adapted to the uneven surface, which it would only fit in one position. The cleared surface was in a hollow with several small natural cavities, where the limpet could have found a pit ready made to shelter in; yet it preferred, after each excursion, to climb up to the top of the flint, the most exposed point in all its domain.[15]

It appears certain from these observations, which to some extent were anticipated by those of Mr. F. C. Lukis,[16] that limpets, after every browsing excursion, return to one particular spot or home; and the precise memory of direction and locality implied by this fact seems to justify us in regarding these actions of the animal as of a nature unquestionably intelligent.

Coming now to the cephalopoda, there is no doubt that if a larger sphere of opportunity permitted, adequate observation of these animals would prove them to be much the most intelligent members of the sub-kingdom. Unfortunately, however, this sphere of opportunity has hitherto been very limited. The following meagre account is all that I have been able to gather concerning the psychology of these interesting animals.

According to Schneider,[17] the Cephalopoda show unmistakable evidence of consciousness and intelligence. This observer had an opportunity of watching them for a long time in the zoological station at Naples; and he says that they appeared to recognise their keeper after they had for some time received their food from him. Hollmann narrates that an octopus, which had had a struggle with a lobster, followed the latter into an adjacent tank, to which it had been removed for safety, and there destroyed it. In order to do this the octopus had to climb up a vertical partition above the surface of the water and descend the other side.[18] According to Schneider, the Cephalopoda have an abstract idea of water, seeking to return to it when removed, even though they do not see it. But this probably arises from the sense of discomfort due to exposure of their skin to the air; and if we can call it an 'idea,' it is doubtless shared by all other aquatic Mollusca when exposed to air.

CHAPTER III.

ANTS.

Within the last ten or twelve years our information on the habits and intelligence of these insects has been so considerably extended, that in here rendering a condensed epitome of our knowledge in this most interesting branch of comparative psychology, it will be found that the chapter is constituted principally of a statement of observations and experiments which have been conducted during the short period named. The observers to whom we are mainly indebted for this large increase of our knowledge are Messrs. Bates, Belt, Müller, Moggridge, Lincecum, MacCook, and Sir John Lubbock. From the fact that these naturalists conducted their observations in different parts of the world and on widely different species of ants, it is not surprising that their results should present many points of difference; for this only shows, as we might have expected, that different species of ants differ considerably in habits and intelligence. Therefore, in now drawing all these numerous observations to a focus, I shall endeavour to show clearly their points of difference as well as their points of agreement; and in order that the facts to be considered may be arranged in some kind of order, I shall deal with them under the following heads:—Powers of special sense; Sense of direction; Powers of memory; Emotions; Powers of communication; Habits general in sundry species; Habits peculiar to certain species; General intelligence of various species.

Powers of Special Sense.

Taking first the sense of sight, Sir John Lubbock made a number of experiments on the influence of light coloured by passing through various tints of stained glass, with the following results. The ants which he observed greatly dislike the presence of light within their nests, hurrying about in search of the darkest corners when light is admitted. The experiments showed that the dislike is much greater in the case of some colours than in that of others. Thus under a slip of red glass there were congregated on one occasion 890 ants, under green 544, under yellow 495, and under violet only 5. To our eyes the violet is as opaque as the red, more so than the green, and much more so than the yellow. Yet, as the numbers show, the ants had scarcely any tendency to congregate under it: there were nearly as many under the same area of the uncovered portion of the nest as under that shaded by the violet glass. It is curious that the coloured glasses appear to act on the ants in a graduated series, which corresponds with the order of their influence on a photographic plate. Experiments were therefore made to test whether it might not be the actinic rays that were so particularly distasteful to the ants; but with negative results. Placing violet glass above red produces the same effect as red glass alone. Obviously, therefore, the ants avoid the violet glass because they dislike the rays which it transmits, and do not prefer the other colours because they like the rays which they transmit. Sodium, barium, strontium, and lithium flames were also tried, but not with so much effect as the coloured glass.

It has just been observed that the relative dislike which Sir John Lubbock's ants showed to lights of different colours seems to be determined by the position of the colour in the spectrum—there being a regular gradation of intolerance shown from the red to the violet end. As these ants dislike light, the question suggests itself that the reason of their graduated intolerance to light of different colours may be due to their eyes not being so much affected by the rays of low as by those of high refrangibility. In this connection it would be interesting to ascertain whether ants of the genus Atta show a similarly graduated intolerance to the light in different parts of the spectrum; for both Moggridge and MacCook record of this genus that it not only does not shun the light, but seeks it—coming to the glass sides of their artificial nests to enjoy the light of a lamp. Possibly, therefore, the scale of preference to lights of different colours would be found in this genus to be the reverse of that which Sir John Lubbock has found in the case of the British species.

As regards hearing, Sir John Lubbock found that sounds of various kinds do not produce any effect upon the insects. Tuning-forks and violin notes, shouting, whistling, &c., were all equally inefficient in producing the slightest influence upon the animals; and experiments with sensitive flames, microphone, telephone, &c., failed to yield any evidence of ants emitting sounds inaudible to human ears.

Fig. 1.

Lastly, as regards the sense of smell, Sir John Lubbock found that on bringing a camel's-hair brush steeped in various strong scents near where ants were passing, "some went on without taking any notice, but others stopped, and evidently perceiving the smell, turned back. Soon, however, they returned, and passed the scented pencil. After doing this two or three times, they generally took no further notice of the scent. This experiment left no doubt on my mind." In other cases the ants were observed to wave about and throw back their antennæ when the scented pencil was brought near.

That ants track one another by scent was long ago mentioned by Huber, and also that they depend on this sense for their power of finding supplies which have been previously found by other ants. Huber proved their power of tracking a path previously pursued by their friends, by drawing his finger across the trail, so obliterating the scent at that point, and observing that when the ants arrived at that point they became confused and ran about in various directions till they again came upon the trail on the other side of the interrupted space, when they proceeded on their way as before. The more numerous and systematic experiments of Sir John Lubbock have fully corroborated Huber's observations, so far as these points are concerned. Thus, to give only one or two of these experiments; in the accompanying woodcut (Fig. 1) A is the nest, B a board, n f g slips of paper, h and m similar slides of glass, on one of which, h, there was placed pupæ, while the other, m, was left empty. Sir John Lubbock watched two particular (marked) ants proceeding from A to h and back again, carrying the pupæ on h to the nest A. Whenever an ant came out of A upon B he transposed the slips f and g. Therefore at the angle below n there was a choice presented to the ant of taking the unscented pathway leading to the full glass h, or the scented pathway leading to the empty glass m. The two marked ants, knowing their way, always took the right turn at the angle; but the stranger ants, being guided only by scent, for the most part took the wrong turn at the angle, so going to the empty glass m. For out of 150 stranger ants only 21 went to h, while the remaining 129 went to m. Still the fact that all the stranger ants did not follow the erroneous scent-trail to m, may be taken to indicate that they are also assisted in finding treasure by the sense of sight, though in a lesser degree. Therefore Sir John Lubbock concludes that in finding treasure 'they are guided in some cases by sight, while in others they track one another by scent.'

Fig. 2.

As further evidence showing how much more ants depend upon scent than upon sight in finding their way, the following experiment may be quoted. In the accompanying woodcut (Fig. 2) the line marked 1, 2, 3 represents the edge of a paper bridge leading to the nest; A the top of a pencil which is standing perpendicularly upon a board, represented by the general black surface; B the top of the same pencil when moved a distance of a few inches from its first position A. On the top of this pencil were placed some pupæ. Sir John Lubbock, after contriving this arrangement, marked an ant and put it upon the pupæ on the top of the pencil. After she had made two journeys carrying pupæ from the pencil to the nest (the tracks she pursued being represented by the two thick white lines), while she was in the nest he moved the pencil to its position at B. The thin white line represents the course then pursued by the ant in its endeavours to find the pencil, which was shifted only a few inches from A to B. That is, 'the ants on their journey to the shifted object travelled very often backwards and forwards and round the spot where the coveted object first stood. Then they would retrace their steps towards the nest, wander hither and thither from side to side between the nest and the point A, and only after very repeated efforts around the original site of the larvæ reach, as it were, accidentally the object desired at B.' Therefore the ants were clearly not guided by the sight of the pencil.

Fig. 3.

Fig. 4.

Fig. 5.

The same thing is well shown by another form of experiment. 'Some food was placed at the point a (Figs. 3 and 4) on a board measuring 20 inches by 12 inches, and so arranged that the ants in going straight from it to the nest would reach the board at the point b, and after passing under the paper tunnel c, would proceed between five pairs of wooden bricks, each 3 inches in length and 1¾ inches in height. When they got to know their way they went quite straight along the line d e to a. The board was then twisted as shown in Fig. 4. 'The bricks and tunnel being arranged exactly in the same direction as before, but the board having been moved, the line d e was now outside them. The change, however, did not at all discompose the ants; but instead of going, as before, through the tunnel and between the rows of bricks to a, they walked exactly along the old path to e.' Keeping the board steady, but moving the brick pathway to the left-hand corner of the board where the food was next placed (Fig. 5), had the effect of making the ant first go to the old position of the food at a, whence it veered to a new position, which we may call x. The bricks and food were then moved towards the right-hand corner of the board—i.e. over a distance of 8 inches (Fig. 6). The ant now first went to a, then to x, and not finding the food at either place, set to work to look for it at random, and was only successful after twenty-five minutes' wandering.

Fig. 6.

And, as evidence how much more dependence they place upon scent in finding their way than upon any other of their faculties, it is desirable to quote yet one further experiment, which is of great interest as showing that when their sense of smell is made to contradict their sense of direction, they follow the former, notwithstanding, as we shall presently see, the wonderful accuracy of the information which is supplied to them by the latter. 'If, when F. niger were carrying off larvæ placed in a cup on a piece of board, I turned the board round so that the side which had been turned towards the nest was away from it, and vice versâ, the ants always returned over the same track on the board, and, in consequence, directly away from home. If I moved my board to the other side of my artificial nest, the result was the same. Evidently they followed the road, not the direction.'

There can be little doubt that ants have a sense of taste, as they are so well able to distinguish sugary substances; and it is unquestionable that in their antennæ they possess highly elaborated organs of touch.

Sense of Direction.

As evidence of the accuracy and importance of the sense of direction in the Hymenoptera, we must here adduce Sir John Lubbock's highly interesting experiments on ants—leaving his experiments in this connection on bees and wasps to be considered in the next chapter. He first accustomed some ants (Lasius niger) to go to and fro to food over a wooden bridge. When they had got quite accustomed to the way, he watched when an ant was upon a bridge which could be rotated, and while she was passing along it, he turned it round, so that end b was at c, and c at b. 'In most cases the ant immediately turned round also; but even if she went on to b or c, as the case might be, as soon as she came to the end of the bridge she turned round.' Next, between the nest and the food he placed a hat-box twelve inches in diameter and seven inches high, cutting two small holes, so that the ants in passing from the nest to the food had to pass in at one hole and out at the other. The box was fixed upon a central pivot, so as to admit of being rotated easily without much friction or disturbance. When the ants had well learnt their way, the box was turned half round as soon as an ant had entered it, 'but in every case the ant turned too, thus retaining her direction.' Lastly, Sir John took a disk of white paper, which he placed in the stead of the hat-box between the nest and the food. When an ant was on the disk making towards the food, he gently drew the disk to the other side of the food, so that the ant was conveyed by the moving surface in the same direction as that in which she was going, but beyond the point to which she intended to go. Under these circumstances 'the ant did not turn round, but went on' to the further edge of the disk, when she seemed 'a good deal surprised at finding where she was.'

These experiments seem to show that the mysterious 'sense of direction,' and consequent faculty of 'homing,' are in ants, at all events, due to a process of registering, and, where desirable, immediately counteracting any change of direction, even when such change is gently made by a wholly closed chamber in which the animal is moving, and not by any muscular movements of the animal itself. And the fact that drawing the moving surface along in the same direction of advance as that which the insect is pursuing does not affect the movements of the latter, seems conclusively to show that the power of registration has reference only to lateral movements of the travelling surface; it has no reference to variations in the velocity of advance along the line in which the animal is progressing.[19]

Powers of Memory.

Little need here be said to prove that ants display some powers of memory; for many of the observations and experiments already detailed constitute a sufficient demonstration of the statement that they do. Thus, for instance, the general fact that whenever an ant finds her way to a store of food or larvæ, she will return to it again and again in a more or less direct line from her nest, constitutes ample proof that the ant remembers the way to the store. It is of considerable interest, however, to note that the nature of this insect-memory appears to be, as far as it goes, precisely identical with that of memory in general. Thus, a new fact becomes impressed upon their memory by repetition, and the impression is liable to become effaced by lapse of time. More evidence on both these features of insect-memory will be adduced when we come to treat of the intelligence of bees; but meanwhile it is enough to refer to the fact that in his experiments on ants, Sir John Lubbock found it necessary to teach the insects by a repetition of several lessons their way to treasure, if that way was long or unusual.

With regard to the duration of memory, it does not appear that any experiments have been made; but the following observation by Mr. Belt on this point in the case of the leaf-cutting ant may here be stated. In June 1859 he found his garden invaded by these ants, and following up their paths he found their nest about a hundred yards distant. He poured down their burrows a pint of common brown carbolic acid, mixed with four buckets of water. The marauding parties were at once drawn off from the garden to meet the danger at home, and the whole formicarium was disorganised, the ants running up and down again in the utmost perplexity. Next day he found them busily employed bringing up the ant-food from the old burrows, and carrying it to newly formed ones a few yards distant. These, however, turned out to be only intended as temporary repositories; for in a few days both the old and the new burrows were entirely deserted, so that he supposed all the ants to have died. Subsequently, however, he found that they had migrated to a new site, about two hundred yards from the old one, and there established themselves in a new nest. Twelve months later the ants again invaded his garden, and again he treated them to a strong dose of carbolic acid. The ants, as on the previous occasion, were at once withdrawn from the garden, and two days afterwards he found 'all the survivors at work on one track that led directly to the old nest of the year before, where they were busily employed in making fresh excavations. Many were bringing along pieces of ant-food' from the nest most recently deluged with carbolic acid to that which had been similarly deluged a year before, and from which all the carbolic acid had long ago disappeared. 'Others carried the undeveloped white pupæ and larvæ. It was a wholesale and entire migration;' and the next day the nest down which he had last poured the carbolic acid was entirely deserted. Mr. Belt adds: 'I afterwards found that when much disturbed, and many of the ants destroyed, the survivors migrate to a new locality. I do not doubt that some of the leading minds in this formicarium recollected the nest of the year before, and directed the migration to it.'

Now, I do not insist that the facts necessarily point to this conclusion; for it may have been that the leaders of the migration simply stumbled upon the old and vacant nest by accident, and finding it already prepared as a nest, forthwith proceeded to transfer the food and pupæ to it. Still, as the two nests were separated from one another by so considerable a distance, this hypothesis does not seem probable, and the only other one open to us is that the ants remembered the site of their former home for a period of twelve months. And this conclusion is rendered less improbable from a statement of Karl Vogt in his 'Thierstaaten,' to the effect that for several successive years ants from a certain nest used to go through certain inhabited streets to a chemist's shop 600 mètres distant, in order to obtain access to a vessel filled with syrup. As it cannot be supposed that this vessel was found in successive working seasons by as many successive accidents, it can only be concluded that the ants remembered the syrup store from season to season.

I shall now pass on to consider a class of highly remarkable facts, perhaps the most remarkable of the many remarkable facts connected with ant psychology.

It has been known since the observations of Huber that all the ants of the same nest or community recognise one another as friends, while an ant introduced from another nest, even though it be an ant of the same species, is known at once to be a foreigner, and is usually maltreated or put to death. Huber found that when he removed an ant from a nest and kept it away from its companions for a period of four months it was still recognised as a friend, and caressed by its previous fellow-citizens after the manner in which ants show friendship, viz., by stroking antennæ. Sir John Lubbock, after repeating and fully confirming these observations, extended them as follows. He first tried keeping the separated ant away from the nest for a still longer period than four months, and found that even after a separation of more than a year the animal was recognised as before. He repeated this experiment a number of times, and always with the same invariable difference between the reception accorded to a foreigner and a native—no matter, apparently, how long the native had been absent.

Considering the enormous number of ants that go to make a nest, it seems astonishing enough that they should be all personally known to one another, and still more astonishing that they should be able to recognise members of their community after so prolonged an absence. Thinking that the facts could only be explained, either by all the ants in the same nest having a peculiar smell, or by all the members of the same community having a particular pass-word or gesture-sign, Sir John Lubbock, with the view of testing this theory, separated some ants from a nest while still in the condition of pupæ, and, when they emerged from that state as perfect insects, transferred them back to the nest from which they had been taken as pupæ. Of course in this case the ants in the nest could never have seen those which had been removed, for a larval ant is as unlike the mature insect as a grub is unlike a beetle; neither can it be supposed that a larva, hatched out away from the nest, should retain, when a perfect insect, any smell belonging to its parent nest—more especially as it had been hatched out by ants in another nest;[20] nor, lastly, is it reasonable to imagine that the animal, while still a larval grub, can have been taught any gesture-signal used as a pass-word by the matured animals. Yet, although all these possible hypotheses seem to be thus fully excluded by the conditions of the experiment, the result showed unequivocally that the ants recognised their transformed larvæ as native-born members of their community.

Lastly, Sir John Lubbock tried the experiment of going still further back in the life-history of the ants before separating them from the nest. For in September he divided a nest into two halves, each having a queen. At this season there were neither larvæ nor eggs. The following April both the queens began to lay eggs, and in August—i.e. nearly a year after the original partitioning of the nest—he took some of the ants newly hatched from the pupæ in one division, and placed them in the other division, and vice versâ. In all cases these ants were received by the members of the other half of the divided nest as friends, although if a stranger were introduced into either half it was invariably killed. Yet the ants which were thus so certainly recognised by their kindred ants as friends had never, even in the state of an egg, been present in that division of the nest before. On this highly remarkable fact Sir John Lubbock says:—

These observations seem to me conclusive as far as they go, and they are very surprising. In my experiments of last year, though the results were similar, still the ants experimented with had been brought up in the nest, and were only removed after they had become pupæ. It might therefore be argued that the ants, having nursed them as larvæ, recognised them when they came to maturity; and though this would certainly be in the highest degree improbable, it could not be said to be impossible. In the present case, however, the old ants had absolutely never seen the young ones until the moment when, some days after arriving at maturity, they were introduced into the nest; and yet in all ten cases they were undoubtedly recognised as belonging to the community.

It seems to me, therefore, to be established by these experiments that the recognition of ants is not personal and individual; that their harmony is not due to the fact that each ant is individually acquainted with every other member of the community.

At the same time, the fact that they recognise their friends even when intoxicated, and that they know the young born in their own nest even when they have been brought out of the chrysalis by strangers, seems to indicate that the recognition is not effected by means of any sign or pass-word.

We must, therefore, conclude with reference to this subject that the mode whereby recognition is undoubtedly effected is as yet wholly unintelligible; and I have introduced these facts under the heading of memory only because this heading is not more inappropriate than any other that could be devised for their reception.

It ought here to be added also that the power of thus recognising members of their community is not confined by the limits of blood-relationship, for in an experiment made by Forel it was shown that Amazon ants recognised their own slaves almost instantaneously after an absence of four months.

Under this heading I may also adduce the evidence as to enormous masses, or, as we might say, a whole nation of ants recognising each other as belonging to the same nationality. New nests often spring up as offshoots from the older ones, and thus a nation of towns gradually spreads to an immense circumference around the original centre. Forel describes a colony of F. exsecta which comprised more than two hundred nests, and covered a space of nearly two hundred square mètres. 'All the members of such a colony, even those from the furthermost nest, recognise each other and admit no stranger.'

Similarly, MacCook describes an 'ant town' in the Alleghany Mountains of North America ('Trans. Amer. Entom. Soc.,' Nov. 1877) which was inhabited by F. exsectoïdes. It consists of 1,600 to 1,700 nests, which rise in cones to a height of from two to five feet. The ground below is riddled in every direction with subterranean passages of communication. The inhabitants are all on the most friendly terms, so that if any one nest is injured it is repaired by their united forces.

It remains to be added in connection with this subject that the recognition is not automatically invariable, but when 'ants are removed from a nest in the pupa state, tended by strangers, and then restored, some at least of their relatives are certainly puzzled, and in many cases doubt their claims to consanguinity. I say some, because while strangers under the circumstances would have been immediately attacked, these ants were in every case amicably received by the majority of the colony, and it was sometimes several hours before they came across one who did not recognise them.'

It may also be added that Lasius flavus behaves towards strangers quite differently and much more hospitably than is the case with L. niger. The stranger shows no alarm, but, on the contrary, will voluntarily enter the strange nest, and she is there received with kindness; although from the attention she excites, and the numerous communications which take place between her and her new friends, Sir John was 'satisfied that they knew she was not one of themselves. . . . . Very different is the behaviour of L. niger under similar circumstances. I tried the same experiment with them. There was no communications with the antennæ, there was no cleaning, but every ant which the stranger approached flew at her like a little tigress. I tried this experiment four times; each stranger was killed and borne off to the nest.'

Emotions.

The pugnacity, valour, and rapacity of ants are too well and generally known to require the narration of special instances of their display. With regard to the tenderer emotions, however, there is a difference of opinion among observers. Before the researches of Sir John Lubbock it was the prevalent view that these insects display marked signs of affection towards one another, both by caressing movements of their antennæ, and by showing solicitude for friends in distress. Sir John, however, has found that the species of ants on which he has experimented are apparently deficient both in feelings of affection and of sympathy—or, at least, that such feelings are in these species much less strongly developed than the sterner passions.

He tried burying some specimens of Lasius niger beneath an ant-road; but none of the ants traversing the road made any attempt to release their imprisoned companions. He tried the same experiment with the same result on various other species. Even when the friends in difficulty are actually in sight, it by no means follows that their companions will assist them. Of this, he says, he could give almost any number of instances. Thus, when ants are entangled in honey, their companions devote themselves to the honey, and entirely neglect their friends in distress; and when partly drowned, their friends take no notice. When chloroformed or intoxicated their own companions either do not heed them, or else 'seem somewhat puzzled at finding their intoxicated fellow-creatures in such a condition, take them up, and carry them about for a time in a somewhat aimless manner.' Further experiments, however, on a larger scale, went to show that chloroformed ants were treated as dead, i.e. removed to the edge of the parade-board and dropped over into the surrounding moat of water; while intoxicated ants were generally carried into the nest, if they were ants belonging to that community; if not, they were thrown overboard. This care shown towards intoxicated friends appears to indicate a dim sense of sympathy towards afflicted individuals; but that this emotion or instinct does not in the case of these species extend to healthy individuals in distress seems to be proved, not only by the experiments of burying already described, but also by the following:—

On Sept. 2, therefore, I put two ants from one of my nests of F. fusca into a bottle, the end of which was tied up with muslin as described, and laid it down close to the nest. In a second bottle I put two ants from another nest of the same species. The ants which were at liberty took no notice of the bottle containing their imprisoned friends. The strangers in the other bottle, on the contrary, excited them considerably. The whole day one, two, or more ants stood sentry, as it were, over the bottle. In the evening no less than twelve were collected round it, a larger number than usually came out of the nest at any one time. The whole of the next two days, in the same way, there were more or less ants round the bottle containing the strangers; while, as far as we could see, no notice whatever was taken of the friends. On the 9th the ants had eaten through the muslin, and effected an entrance. We did not chance to be on the spot at the moment; but as I found two ants lying dead, one in the bottle and one just outside, I think there can be no doubt that the strangers were put to death. The friends throughout were quite neglected.

Sept. 21.—I then repeated the experiment, putting three ants from another nest in a bottle as before. The same scene was repeated. The friends were neglected. On the other hand, some of the ants were always watching over the bottle containing the strangers, and biting at the muslin which protected them. The next morning at 6 A.M. I found five ants thus occupied. One had caught hold of the leg of one of the strangers, which had unwarily been allowed to protrude through the meshes of the muslin. They worked and watched, though not, as far as I could see, with any system, till 7.30 in the evening, when they effected an entrance, and immediately attacked the strangers.

Sept. 24.—I repeated the same experiment with the same nest. Again the ants came and sat over the bottle containing the strangers, while no notice was taken of the friends.

The next morning again, when I got up, I found five ants round the bottle containing the strangers, none near the friends. As in the former case, one of the ants had seized a stranger by the leg, and was trying to drag her through the muslin. All day the ants clustered round the bottle, and bit perseveringly, though not systematically, at the muslin. The same thing happened all the following day.

On repeating these experiments with another species (viz., Formica rufescens) the ants took no notice of either bottle, and showed no sign either of affection or hatred. One is almost tempted to surmise that the spirit of these ants is broken by slavery [i.e. by the habit of keeping slaves]. But the experiments on F. fusca seem to show that in these curious insects hatred is a stronger passion than affection.

We must not, however, too readily assent to this general conclusion, that ants as a whole are deficient in the tenderer emotions; for although the case is doubtless so with the species which Sir John examined, it appears to be certainly otherwise with other species, as we shall presently see. But first it may be well to point out that even the hard-hearted species with which Sir John had to do seem not altogether devoid of sympathy with sick or mutilated friends, although they appear to be so towards healthy friends in distress. Thus the care shown to intoxicated friends seems to indicate, if not, as already observed, a dim sense of sympathy, at least an instinct to preserve the life of an ailing citizen for the future benefit of the community. Sir John also quotes some observations of Latreille showing that ants display sympathy with mutilated companions; and, lastly, mentions an instance which he has himself observed of the same thing. A specimen of F. fusca congenitally destitute of antennæ was attacked and injured by an ant of another species. When separated by Sir John, another ant of her own species came by. 'She examined the poor sufferer carefully, then picked her up tenderly, and carried her away into the nest. It would have been difficult for any one who witnessed this scene to have denied to this ant the possession of humane feelings.' Moggridge is also of opinion that the habit of throwing sick and apparently dead ants into the water, is 'in part to be rid of them, and partly, perhaps, with a view to effecting a possible cure; for I have seen one ant carry another down the twig which formed their path to the surface of the water, and, after dipping it in for a minute, carry it laboriously up again, and lay it in the sun to dry and recover.'

But that some species of ants display marked signs of what we may call sympathy even towards healthy companions in distress, is proved by the following observation of Mr. Belt. He writes:[21]—

One day, watching a small column of these ants (i.e. Eciton humata), I placed a little stone on one of them to secure it. The next that approached, as soon as it discovered its situation, ran backwards in an agitated manner, and soon communicated the intelligence to the others. They rushed to the rescue; some bit at the stone and tried to move it, others seized the prisoner by the legs and tugged with such force that I thought the legs would be pulled off, but they persevered until they got the captive free. I next covered one up with a piece of clay, leaving only the ends of its antennæ projecting. It was soon discovered by its fellows, which set to work immediately, and by biting off pieces of the clay soon liberated it. Another time I found a very few of them passing along at intervals. I confined one of these under a piece of clay at a little distance from the line, with his head projecting. Several ants passed it, but at last one discovered it and tried to pull it out, but could not. It immediately set off at a great rate, and I thought it had deserted its comrade, but it had only gone for assistance, for in a short time about a dozen ants came hurrying up, evidently fully informed of the circumstances of the case, for they made directly for their imprisoned comrade and soon set him free. I do not see how this action could be instinctive. It was sympathetic help, such as man only among the higher mammalia shows. The excitement and ardour with which they carried on their unflagging exertions for the rescue of their comrade could not have been greater if they had been human beings.

This observation seems unequivocal as proving fellow-feeling and sympathy, so far as we can trace any analogy between the emotions of the higher animals and those of insects. That insects with such highly organised social habits, and depending so greatly on the principles of co-operation, should manifest emotions or instincts of an incipiently altruistic character, is no more than we should antecedently expect on the general principle of survival of the fittest. Our only surprise should be that these emotions, or instincts, should appear to be so feebly developed in some species of ants, and, as we shall subsequently see, also of bees. But it may be worth while in this connection to point out that the valuable observation of Mr. Belt above quoted refers to the species of ant which, as we shall subsequently find, presents the most highly organised instincts of co-operation that are to be met with among ants, and therefore the greatest dependence of the welfare of the individual on that of the community. And the same remark is applicable to our native species, F. sanguinea, which the Rev. W. W. F. White has repeatedly seen rescuing buried companions very much in the manner described by Mr. Belt; and he does not appear to be acquainted with Mr. Belt's observations. He figures one case in which he saw three ants co-operating to dig out a buried comrade.[22]

Powers of Communication.

Huber, Kirby and Spence, Dugardin, Burmeister, Franklin, and other observers have all expressed themselves as more or less strongly of the opinion that members of the same community of ants, and other social Hymenoptera, are able to communicate information to one another by some system of language or signs. The facts, however, on which their opinion rests have not been stated with that degree of caution and detail which the acceptance of the conclusion requires. Thus, Kirby and Spence give only one instance of supposed communication between ants,[23] and even this one is inconclusive, as the facts described admit of being explained by supposing that the ants simply tracked one another by scent; while Huber merely deals in general statements as to 'contact of antennæ,' without narrating any particulars of his observations. Therefore, until within the last few years there was really no sufficient evidence to sustain the general opinion that ants are able to communicate with one another; but the observations which I shall now detail must be regarded as fully substantiating that general opinion by facts as abundant and conclusive as the most critical among us can desire. I shall first narrate in his own words the more important of Sir John Lubbock's experiments in this connection:—

I took three tapes, each about 2 feet 6 inches long, and arranged them parallel to one another and about 6 inches apart. An end of each I attached to one of the nests (F. niger), and at the other end I placed a glass. In the glass at the end of one tape I placed a considerable number (300 to 600) of larvæ. In the second I put two or three larvæ only, in the third none at all. The object of the last was to see whether many ants would come to the glasses under such circumstances by mere accident, and I may at once say that scarcely any did so. I then took two ants, and placed one of them to the glass with many larvæ, the other to that with two or three. Each of them took a larva and carried it to the nest, returning for another, and so on. After each journey I put another larva in the glass with only two or three larvæ, to replace that which had been removed. Now, if several ants came under the above circumstances as a mere matter of accident, or accompanying one another by chance, or if they simply saw the larvæ which were being brought, and consequently concluded that they might themselves find a larva in the same place, then the numbers going to the two glasses ought to be approximately equal. In each case the number of journeys made by the ants would be nearly the same; consequently, if it was a matter of scent, the two glasses would be in the same position. It would be impossible for an ant, seeing another in the act of bringing a larva, to judge for itself whether there were few or many left behind. On the other hand, if the strangers were brought, then it would be curious to see whether more were brought to the glass with many larvæ than to that which only contained two or three. I should also mention that every stranger was imprisoned until the end of the experiment.

The results of these experiments were that during 47½ hours the ants which had access to a glass containing numerous larvæ brought 257 friends to their assistance; while during an interval 5½ hours longer those which visited the glass with only two or three larvæ brought only 82 friends; and, as already mentioned, no single ant came to the glass which contained no larvæ. Now, as all the glasses were exposed to similar conditions, and as the roads to the first two must, in the first instance at all events, have been equally scented by the passage of ants over them, these results look very conclusive as proving some power of definite communication, not only that larvæ are to be found, but even where the largest store is to be met with.

To this interesting account Sir John Lubbock adds,—

One case of apparent communication struck me very much. I had had an ant (F. niger) under observation one day, during which she was occupied in carrying off larvæ to her nest. At night I imprisoned her in a small bottle; in the morning I let her out at 6.15, when she immediately resumed her occupation. Having to go to London, I imprisoned her again at 9 o'clock. When I returned at 4.40 I put her again to the larvæ. She examined them carefully, and went home without taking one. At this time no other ants were out of the nest. In less than a minute she came out again with eight friends, and the little heap made straight for the heap of larvæ. When they had gone two-thirds of the way I again imprisoned the marked ant; the others hesitated a few minutes, and then with curious quickness returned home. At 5.15 I put her again to the larvæ. She again went home without a larva, but after only a few seconds' stay in the nest, came out with no less than thirteen friends. They all went towards the larvæ, but when they had got about two-thirds of the way, although the marked ant had on the previous day passed over the ground about 150 times, and though she had just gone straight from the larvæ to the nest, she seemed to have forgotten her way, and considered; and after she had wandered about for half an hour, I put her to the larvæ. Now, in this case, the twenty-one ants must have been brought out by my marked one, for they came exactly with her, and there were no other ants out. Moreover, it would seem that they must have been told, because (which is very curious in itself) she did not in either case bring a larva, and consequently it cannot have been the mere sight of a larva which had induced them to follow her.

Further experiments proved, as we might have expected, that although an ant is able to communicate to her friends in the nest that she has found treasure somewhere outside, she is not able to describe to them its precise locality. Thus, having exposed larvæ and placed an ant upon them as before, Sir John watched every time she came out of the nest with friends to assist her, but instead of allowing her to pilot the way, he took her up and carried her to the larvæ, allowing her to return with a larva upon her own feet. Under these circumstances the friends, although evidently coming out with the intention of finding some treasure, were never able to find it; but wandered about in various directions for a while, and then returned to the nest. Thus, during two hours she brought out in her successive journeys altogether no less than 120 ants, of which number only 5 in their unguided wanderings happened to find the sought-for treasure. This result seems to prove, as we might have expected, that the communication is of the nature of some sign amounting to no more than a 'follow me.' Other experiments confirmed this result, and also brought out the fact that 'some species act much more in association than others—Formica fusca, for instance, much less than Lasius niger.' Thus Sir John Lubbock placed some honey before a marked specimen of the former species; but although she visited and revisited the honey during an entire day, she brought out no friends to share it; and although in her journeys to and from the nest she happened to pass and repass many other individuals, they took no notice of each other.

The obvious objection to these experiments, that an ant observing a friend bringing home food or a pupa might infer, without being told, that by accompanying the friend on the return journey she 'might participate in the good things,' has been partly met by the fact already stated, viz., that there is so very marked a difference in the result if, on experimenting on two ants, one had access to a large treasure and the other only to a small one. But to put this matter beyond question, Sir John Lubbock tried the experiment of pinning down a dead fly, so that the ant which found it was unable, with all her tugging, to move it towards the nest. At length she went back to the nest for assistance, and returned accompanied by seven friends. So great was her excitement, however, that she outran these friends, 'who seemed to have come out reluctantly, as if they had been asleep, and were only half awake;' and they failed to find the fly, slowly meandering about for twenty minutes. After again tugging for a time at the fly, the first ant returned a second time to the nest for assistance, and in less than a minute came out with eight friends. They were even less energetic than the first party, and having lost sight of their guide in the same manner as happened before, they all returned to the nest. Meanwhile several of the first party, which had all the while been meandering about, found the fly, and proceeded to dismember it, carrying the trophy to the nest, and calling out more friends in the ordinary way. This experiment was repeated several times and on different species, always with the same result. Now, as Sir John remarks, 'the two cases (i.e. those in which the ant brought out friends to her assistance even when she had no booty to show) surely indicate a distinct power of communication. . . . . It is impossible to doubt that the friends were brought out by the first ant; and as she returned empty-handed to the nest, the others cannot have been induced to follow her by merely observing her proceedings. I conclude, therefore, that they possess the power of requesting their friends to come and help them.'

In order to ascertain whether the signs which communicating ants make to one another are made by means of sound, Sir John Lubbock placed near a nest of Lasius flavus six small upright pillars of wood about 1½ inch high, and on one of these he put a drop of honey. 'I then put three ants to the honey, and when each had sufficiently fed, I imprisoned her, and put another; thus always keeping three ants at the honey, but not allowing them to go home. If, then, they could summon their friends by sound, there ought soon to be many ants at the honey.' The result showed that the ants were not able thus to call to one another from a distance.

As additional proof of the general fact that at all events some ants have the power of communicating information to one another, it will be enough here to quote an exceedingly interesting observation of the distinguished geologist Hague. The quotations are taken from his letters written to Mr. Darwin, and published in Nature:[24]—

On the mantelshelf of our sitting-room my wife has the habit of keeping fresh flowers. A vase stands at each end, and near the middle a small tumbler, usually filled with violets. Some time ago I noticed a pile of very small red ants on the wall above the left-hand vase, passing upward and downward between the mantelshelf and a small hole near the ceiling, at a point where a picture nail had been driven. The ants, when first observed, were not very numerous, but gradually increased in number, until on some days the little creatures formed an almost unbroken procession, issuing from the hole at the nail, descending the wall, climbing the vase directly below the nail, satisfying their desire for water or perfume, and then returning. The other vase and tumbler were not visited at that time.

As I was just then recovering from a long illness it happened that I was confined to the house, and spent my days in the room where the operations of these insects attracted my attention. Their presence caused me some annoyance, but I knew of no effective means of getting rid of them. For several days in succession I frequently brushed the ants in great numbers from the wall down to the floor; but as they were not killed the result was that they soon formed a colony in the wall at the base of the mantel, ascending thence to the shelf, so that before long the vase was attacked from above and below.

One day I observed a number of ants, perhaps thirty or forty, on the shelf at the foot of the vase. Thinking to kill them, I struck them lightly with the end of my finger, killing some and disabling the rest. The effect of this was immediate and unexpected. As soon as those ants which were approaching arrived near to where their fellows lay dead and suffering, they turned and fled with all possible haste. In half an hour the wall above the mantelshelf was cleared of ants.

During the space of an hour or two the colony from below continued to ascend until reaching the lower bevelled edge of the shelf, at which point the more timid individuals, although unable to see the vase, somehow became aware of trouble, and turned about without further investigation, while the more daring advanced hesitatingly just to the upper edge of the shelf, when, extending their antennæ and stretching their necks, they seemed to peep cautiously over the edge until beholding their suffering companions, when they too turned and followed the others, expressing by their behaviour great excitement and terror. An hour or two later, the path or trail leading from the lower colony to the vase was almost entirely free from ants.

I killed one or two ants on their path, striking them with my finger, but leaving no visible trace. The effect of this was that as soon as an ant ascending towards the shelf reached the spot where one had been killed, it gave signs immediately of great disturbance, and returned directly at the highest possible speed.

A curious and invariable feature of their behaviour was that when such an ant, returning in fright, met another approaching, the two would always communicate, but each would pursue its own way, the second ant continuing its journey to the spot where the first had turned about, and then following that example.

For some days after this there were no ants visible on the wall, either above or below the shelf.

Then a few ants from the lower colony began to reappear, but instead of visiting the vase which had been the scene of the disaster, they avoided it altogether, and following the lower front edge of the shelf to the tumbler standing near the middle, made their attack upon that. I repeated the same experiment here with precisely the same result. Killing or maiming a few of the ants and leaving their bodies about the base of the tumbler, the others on approaching, and even before arriving at the upper surface of the shelf where their mutilated companions were visible, gave signs of intense emotion, some running away immediately, and others advancing to where they could survey the field and then hastening away precipitately.

Occasionally an ant would advance towards the tumbler until it found itself among the dead and dying; then it seemed to lose all self-possession, running hither and thither, making wide circuits about the scene of the trouble, stopping at times and elevating the antennæ with a movement suggestive of wringing them in despair, and finally taking flight. After this another interval of several days passed, during which no ants appeared. Now, three months later, the lower colony has been entirely abandoned. Occasionally, however, especially when fresh and fragrant violets have been placed on the shelf, a few 'prospectors' descend from the upper nail-hole, rarely, almost never, approaching the vase from which they were first driven away, but seeking to satisfy their desire at the tumbler. To turn back these stragglers and keep them out of sight for a number of days, sometimes for a fortnight, it is sufficient to kill one or two ants on the trail which they follow descending the wall. This I have recently done as high up as I can reach, three or four feet above the mantel. The moment this spot is reached, an ant turns abruptly and makes for home, and in a little while there is not an ant visible on the wall.

In a subsequent volume of 'Nature' (viii. p. 244), Mr. Darwin publishes another letter which he received from Mr. Hague upon the same subject. It seems that Mr. Moggridge suggested to Mr. Darwin that, as he and others had observed ants to be repelled by the mere scent of a finger drawn across their path, the observation of Mr. Hague might really resolve itself into a dislike on the part of the ants to cross a line over which a finger had been drawn, and have nothing to do with intelligent terror inspired by the sight of their slaughtered companions. The following is Mr. Hague's reply to Mr. Darwin's request for further experiments to test this point:—

Acting on Mr. M——'s suggestion, I first tried making simple finger-marks on their path (the mantel is of marble), and found just the results which he describes in his note as observed by himself at Mentone, that is, no marked symptoms of fear, but a dislike to the spot, and an effort to avoid it by going around it, or by turning back and only crossing it again after an interval of time. I then killed several ants on the path, using a smooth stone or piece of ivory, instead of my finger, to crush them. In this case the ants approaching all turned back as before, and with much greater exhibition of fear than when the simple finger-mark was made. This I did repeatedly. The final result was the same as obtained last winter. They persisted in coming for a week or two, during which I continued to kill them, and then they disappeared, and we have seen none since. It would appear from this that while the taint of the hand is sufficient to turn them back, the killing of their fellows with a stone or other material produces the effect described in my first note. This was made clear to me at that time, from the behaviour of the ants the first day I killed any, for on that occasion some of them approaching the vase from below, on reaching the upper edge of the mantel, peeped over, and drew back on seeing what had happened about the vase, then turned away a little, and after a moment tried again at another and another point along the edge, with the same result in the end. Moreover, those that found themselves among the dead and dying went from one writhing ant to another in great haste and excitement, exhibiting the signs of fright which I described.

I hardly hope that any will return again, but if they do, and give me an opportunity, I shall endeavour to act further on Mr. M——'s suggestion.

With this quotation I shall conclude the present division of the chapter; for, looking to all the other observations previously mentioned, there can be no question concerning the general fact that ants have the power of communicating with one another. And under subsequent headings abundant additional evidence on this point will be found implicated with the other facts detailed.

Habits General in Sundry Species.

Swarming.—The precise facts with regard to the swarming of ants are not yet certainly established. As regards some of the facts, however, there is no doubt. The winged males and females first quit the nest in enormous numbers, and choose some fine afternoon in July or August for their wedding flight. The entrances to the nest are widened by the workers and increased in number, and there is a great commotion on the surface of the nest. The swarm takes place as a thick cloud of all the male and female insects, rising together to a considerable height. The flight continues for several hours, usually circling round some tree or tower, and it is during the flight that fertilisation is effected. After it is effected, the swarm returns to the ground, when the males perish, either from falling a prey, in their shelterless condition, to birds or spiders, or, on account of not being able to feed themselves, from starvation. 'The workers, or neuter ants, of their own colony have lost all interest in them from the moment of their return, and trouble themselves no more about them, for they well know that the males have now fulfilled their vocation.' The great majority of the fertilised females share the same fate as the males. But a small proportion find concealment in holes, which they either dig for themselves, or happen to find ready made, and there found a new colony. The first thing they do is to pull off their now useless wings, by scratching and twisting them, one after the other, with the clawed ends of their feet. They then lay their eggs, and become the queens of new colonies.

Forel says that no fertilised female ever returns to her original home; but that the workers keep back a certain number of females which are fertilised before the swarming takes place; in this case the workers pull off the wings of the fertilised females. The majority of observers, however, maintain that some of the females composing the swarm return to their native home to become mothers where they had been children. Probably both statements are correct. A writer in the 'Groniger Deekblad' for June 16, 1877, observes that, looking to the injurious effects of in-breeding, the facts as related by Forel are less probable than those related by other observers, and that, if they actually occur, the females fertilised before flight are probably kept by the ants as a sort of 'reserve corps to which the workers resort only in case of need, and if they fail to secure any returning queens.'

Nursing.—The eggs will not develop into larvæ unless nursed. The nursing is effected by licking the surface of the eggs, which under the influence of this process increase in size, or grow. In about a fortnight, during which time the workers carry the eggs from higher to lower levels of the nest, and vice versâ, according to the circumstances of heat, moisture, &c., the larvæ are hatched out, and require no less careful nursing than the eggs. The workers feed them by placing mouths together and regurgitating food stored up in the crop or proventriculus into the intestinal tract of the young. The latter show their hunger by 'stretching out their little brown heads.' Great care is also taken by the workers in cleaning the larvæ, as well as in carrying them up and down the chambers of the nest for warmth or shelter.

When fully grown the larvæ spin cocoons, and are then pupæ, or the 'ants' eggs' of bird-fanciers. These require no food, but still need incessant attention with reference to warmth, moisture, and cleanliness. When the time arrives for their emergence as perfect insects, the workers assist them to get out of their larval cases by biting through the walls of the latter. It is noticeable that in doing this the workers do not keep to any exact time, but free them sometimes earlier and sometimes later, in accordance with their rate of development. 'The little animal when freed from its chrysalis is still covered with a thin skin, like a little shirt, which has to be pulled off. When we see how neatly and gently this is done, and how the young creature is then washed, brushed, and fed, we are involuntarily reminded of the nursing of human babies. The empty cases, or cocoons, are carried outside the nest, and may be seen heaped together there for a long time. Some species carry them far away from the nest, or turn them into building materials for the dwelling.'[25]

Education.—The young ant does not appear to come into the world with a full instinctive knowledge of all its duties as a member of a social community. It is led about the nest, and 'trained to a knowledge of domestic duties, especially in the case of the larvæ.' Later on the young ants are taught to distinguish between friends and foes. When an ants' nest is attacked by foreign ants, the young ones never join in the fight, but confine themselves to removing the pupæ; and that the knowledge of hereditary enemies is not wholly instinctive in ants is proved by the following experiment, which we owe to Forel. He put young ants belonging to three different species into a glass case with pupæ of six other species—all the species being naturally hostile to one another. The young ants did not quarrel, but worked together to tend the pupæ. When the latter hatched out, an artificial colony was formed of a number of naturally hostile species all living together after the manner of the 'happy families' of the showmen.

Habit of keeping Aphides.—It is well known that various species of ants keep aphides, as men keep milch cows, to supply a nutritious secretion. Huber first observed this fact, and noticed that the ants collected the eggs of the aphides and treated them exactly as they treated their own, guarding and tending them with the utmost care. When these eggs hatch out the aphides are usually kept and fed by the ants, to whom they yield a sweet honey-like fluid, which they eject from the abdomen upon being stroked on this region by the antennæ of the ants. Mr. Darwin, who has watched the latter process, observes with regard to it,—

I removed all the ants from a group of about a dozen aphides on a dock plant, and prevented their attendance during several hours. After this interval, I felt sure that the aphides would want to excrete. I watched them for some time through a lens, but not one excreted; I then tickled them with a hair in the same manner, as well as I could, as the ants do with their antennæ; but not one excreted. Afterwards I allowed an ant to visit them, and it immediately seemed, by its eager way of running about, to be well aware what a rich flock it had discovered; it then began to play with its antennæ on the abdomen, first of one aphis and then of another; and each, as soon as it felt the antennæ, immediately lifted up its abdomen and excreted a limpid drop of sweet juice, which was eagerly devoured by the ant. Even quite young aphides behaved in this manner, showing that the action was instinctive, and not the result of experience.

The facts also show that the yielding of the secretion to the ants is, as it were, a voluntary act on the part of the aphides, or, perhaps more correctly, that the instinct to yield it has been developed in such a relation to the requirements of the ants, that the peculiar stimulation supplied by the antennæ of the latter is necessary to start the act of secretion; for in the absence of this particular stimulation the aphides will never excrete until compelled to do so by the superabundance of the accumulating secretion. The question, therefore, directly arises how, on evolutionary principles, such a class of facts is to be met; for it is certainly difficult to understand the manner in which this instinct, so beneficial to the ants, can have arisen in the aphides, to which it does not appear, at first sight, to offer any advantages. Mr. Darwin meets the difficulty thus: 'Although there is no evidence that any animal performs an action for the exclusive good of another species, yet each tries to take advantage of the instincts of others;' and 'as the secretion is extremely viscid, it is no doubt a convenience to the aphides to have it removed; therefore probably they do not excrete solely for the good of the ants.'[26]

Some ants which keep aphides build covered ways, or tunnels, to the trees or shrubs where the aphides live. Forel saw a tunnel of this kind which was taken up a wall and down again on the other side, in order to secure a safe covered way from the nest to the aphides. Occasionally such covered ways, or tubes, are continued so as to enclose the stems of the plants on which the aphides live. The latter are thus imprisoned by the walls of the tube, which, however, expand where they take on this additional function of stabling the aphides, so that these insects are really confined in tolerably large chambers. The doors of these chambers are too small to allow the aphides to escape, while large enough for the ants to pass in and out. Forel saw such a prison or stable shaped like a cocoon, and about a centimètre long, which was hanging on the branch of a tree, and contained aphides carefully tended by the ants. Huber records similar observations.

Sir John Lubbock has made an interesting addition to our knowledge respecting this habit as practised by a certain species of ant (Lasius flavus), which departs in a very remarkable manner from the habit as practised by other species. He says: 'The ants took the greatest care of these eggs, carrying them off to the lower chambers with the utmost haste when the nest was disturbed.' But the most interesting of Sir John Lubbock's observations in this connection is new, and reveals an astonishing amount of method shown by the ants in farming their aphides. He says:—

When my eggs hatched I naturally thought that the aphides belonged to one of the species usually found on the roots of plants in the nests of Lasius flavus. To my surprise, however, the young creatures made the best of their way out of the nest, and, indeed, were sometimes brought out by the ants themselves. In vain I tried them with roots of grass, &c.; they wandered uneasily about, and eventually died. Moreover, they did not in any way resemble the subterranean species. In 1878 I again attempted to rear these young aphides; but though I hatched a great many eggs, I did not succeed. This year, however, I have been more fortunate. The eggs commenced to hatch the first week in March. Near one of my nests of Lasius flavus, in which I had placed some of the eggs in question, was a glass containing living specimens of several species of plants commonly found on or around ants' nests. To this some of the young aphides were brought by the ants. Shortly afterwards I observed on a plant of daisy, in the axils of the leaves, some small aphides, very much resembling those from my nest, though we had not actually traced them continuously. They seemed thriving, and remained stationary on the daisy. Moreover, whether they had sprung from the black eggs or not, the ants evidently valued them, for they built up a wall of earth round and over them. So things remained throughout the summer, but on October 9 I found that the aphides had laid some eggs exactly resembling those found in the ants' nests; and on examining daisy plants from outside, I found on many of them similar aphides, and more or less of the same eggs.

I confess these observations surprised me very much. The statements of Huber have not, indeed, attracted so much notice as many of the other interesting facts which he has recorded, because if aphides are kept by ants in their nests, it seems only natural that their eggs should also occur. The above case, however, is much more remarkable. Here are aphides, not living in the ants' nests, but outside, on the leaf-stalks of plants. The eggs are laid early in October on the food-plant of the insect. They are of no direct use to the ants, yet they are not left where they are laid, where they would be exposed to the severity of the weather and to innumerable dangers, but brought into their nests by the ants, and tended by them with the utmost care through the long winter months until the following March, when the young ones are brought out and again placed on the young shoots of the daisy. This seems to me a most remarkable case of prudence. Our ants may not perhaps lay up food for the winter, but they do more, for they keep during six months the eggs which will enable them to procure food during the following summer.

The following, which is taken from Büchner's 'Geistesleben der Thiere' is perhaps a still more striking performance of the same kind as that which Sir John Lubbock observed:—

The author is debtor to Herr Nottebohm, Inspector of Buildings at Karlsruhe, who related the following on May 24, 1876, under the title, 'Ants as Founders of Aphides' Colonies:'—'Of two equally strong young weeping ashes, which I planted in my garden at Kattowitz, in Upper Silesia, one succeeded well, and in about five or six years showed full foliage, while the other regularly every year was covered, when it began to bud, with millions of aphides, which destroyed the young leaves and sprouts, and thus completely delayed the development of the tree. As I perceived that the only reason for this was the action of the aphides, I determined to destroy them utterly. So in the March of the following year I took the trouble to clean and wash every bough, sprig, and bud before the bursting of the latter, with the greatest care, by means of a syringe. The result was that the tree developed perfectly healthy and vigorous leaves and young shoots, and remained quite free from the aphides until the end of May or the beginning of June. My joy was of short duration. One fine sunny morning I saw a surprising number of ants running quickly up and down the trunk of the tree; this aroused my attention, and led me to look more closely. To my great astonishment I then saw that many troops of ants were busied in carrying single aphides up the stem to the top, and that in this way many of the lower leaves had been planted with colonies of aphides. After some weeks the evil was as great as ever. The tree stood alone on the grass plot, and offered the only situation for an aphides' colony for the countless ants there present. I had destroyed this colony; but the ants replanted it by bringing new colonists from distant branches, and setting them on the young leaves.[27]

Again—

MacCook noticed, of the mound-making ants, that of the workers returning to the nest from the tree on which the milking was going on, a far smaller number had distended abdomens than among those descending the tree itself. A closer investigation showed that at the roots of the trees, at the outlets of the subterranean galleries, a number of ants were assembled, which were fed by the returning ants after the fashion already described in feeding the larvæ, and which were distinguished by the observer as 'pensioners.' MacCook often observed the same fact later, among, with others, the already described Pennsylvanian wood-ant. Distinguished individuals in the body-guard of the queen were fed in like fashion. MacCook is inclined to think that the reason of this proceeding is to be found in the 'division of labour' so general in the ant republic, and that the members of the community which are employed in building and working within the nest, leave to the others the care of providing food for themselves as well as for the younger and helpless members; they thus have a claim to receive from time to time a reciprocal toll of gratitude, and take it, as is shown very clearly, in a way demanded by the welfare of the community.[28]

Aphides are not the only insects which ants employ as cows, several other insects which yield sweet secretions being similarly utilised in various parts of the world. Thus, gall insects and cocci are kept in just the same way as aphides; but MacCook observed that where aphides and cocci are kept by the same ants, they are kept in separate chambers, or stalls. The same observer saw caterpillars of the genus Lycœna kept by ants for the sake of a sweet secretion which they supply.

Habit of making Slaves.—This habit, or instinct, obtains among at least three species of ant, viz., Formica rufescens, F. sanguinea, and strongylognathus. It was originally observed by P. Huber in the first-named species. Here the species enslaved is F. fusca, which is appropriately coloured black. The slave-making ants attack a nest of F. fusca in a body; there is a great fight with much slaughter, and, if victorious, the slave-makers carry off the pupæ of the vanquished nest in order to hatch them out as slaves. Mr. Darwin gives an account of a battle which he himself observed.[29]

When the pupæ hatch out in the nest of their captors, the young slaves begin their life of work, and seem to regard their master's home as their own; for they never attempt to escape, and they fight no less keenly than their masters in defence of the nest. F. sanguinea content themselves with fewer slaves than do F. rufescens; and the work that devolves upon the slaves differs according to the species which has enslaved them. In the nests of F. sanguinea the comparatively few captives are kept as household slaves; they never either enter or leave the nest, and so are never seen unless the nest is opened. They are then very conspicuous from the contrast which their black colour and small size present to the red colour and much larger size of F. rufescens. As the slaves are by this species kept strictly indoors, all the outdoor work of foraging, slave-capturing, &c., is performed by the masters; and when for any reason a nest has to migrate, the masters carry their slaves in their jaws. F. rufescens, on the other hand, assigns a much larger share of labour to the slaves, which, as we have already seen, are present in much larger numbers to take it. In this species the males and fertile females do no work of any kind; and the workers, or sterile females, though most energetic in capturing slaves, do no other kind of work. Therefore the whole community is absolutely dependent upon its slaves. The masters are not able to make their own nests or to feed their own larvæ. When they migrate, it is the slaves that determine the migration, and, reversing the order of things that obtains in F. sanguinea, carry their masters in their jaws. Huber shut up thirty masters without a slave and with abundance of their favourite food, and also with their own larvæ and pupæ as a stimulus to work; but they could not feed even themselves, and many died of hunger. He then introduced a single slave, and she at once set to work, fed the surviving masters, attended to the larvæ, and made some cells.

In order to confirm this observation, Lespès placed a piece of sugar near a nest of slave-makers. It was soon found by one of the slaves, which gorged itself and returned to the nest. Other slaves then came out and did likewise. Then some of the masters came out, and, by pulling the legs of the feeding slaves, reminded them that they were neglecting their duty. The slaves then immediately began to serve their masters with the sugar. Forel also has confirmed all these observations of Huber. Indeed, in the case of F. rufescens, the structure of the animal is such as to render self-feeding physically impossible. Its long and narrow jaws, adapted to pierce the head of an enemy, do not admit of being used for feeding, unless liquid food is poured into them by the mouth of a slave. This fact shows of how ancient an origin the instinct of slave-making must be; it has altered in an important manner a structure which could not have been so altered prior to the establishment of the instinct in question.

Mr. Darwin thus sums up the differences in the offices of the slaves in the nests of F. sanguinea and F. rufescens respectively:—

The latter does not build its own nest, does not determine its own migrations, does not collect food for itself or for its fellows, and cannot even feed itself; it is absolutely dependent on its numerous slaves. Formica sanguinea, on the other hand, possesses much fewer slaves, and in the early part of the summer extremely few; the masters determine when and where a new nest shall be formed, and when they migrate, the masters carry the slaves. Both in Switzerland and England the slaves seem to have the exclusive care of the larvæ, and the masters alone go on slave-making expeditions. In Switzerland the slaves and masters work together, making and bringing materials for the nest; both, but chiefly the slaves, tend and milk, as it may be called, their aphides; and thus both collect food for the community. In England the masters alone usually leave the nest to collect building materials and food for themselves, their slaves and larvæ. So that the masters in this country receive much less service from their slaves than they do in Switzerland.

Mr. Darwin further observes that 'this difference in the usual habits of the masters and slaves in the two countries probably depends merely on the slaves being captured in greater numbers in Switzerland than in England;' and records that he has observed in a community of the English species having an unusually large stock of slaves that 'a few slaves mingled with their masters leaving the nest, and marched along the same road to a tall Scotch fir tree, twenty-five yards distant, which they ascended together, probably in search of aphides or cocci.' And, according to Huber, the principal office of the slaves in Switzerland is to search for aphides.

Mr. Darwin also made the following observation:—'Desiring to ascertain whether F. sanguinea could distinguish the pupæ of F. fusca, which they habitually make into slaves, and which are an unwarlike species, from F. flava, which they rarely capture, and never without a severe fight,' he found 'it was evident that they did at once distinguish them;' for while 'they eagerly and instantly seized the pupæ of F. fusca, they were much terrified when they came across the pupæ, or even the earth from the nest, of F. flava, and quickly ran away; but in about a quarter of an hour, shortly after the little yellow ants had crawled away (from their nest having been disturbed by Mr. Darwin), they took heart and carried off the pupæ.'

Concerning the origin of this remarkable instinct, Mr. Darwin writes:—

As ants which are not slave-makers will, as I have seen, carry off pupæ of other species if scattered near their nests, it is possible that such pupæ originally stored as food might become developed, and the foreign ants thus unintentionally reared would then follow their proper instincts, and do what work they could. If their presence proved useful to the species which had seized them—if it were more advantageous to the species to capture workers than to procreate them—the habit of collecting pupæ, originally for food, might by natural selection be strengthened and rendered permanent for the very different purpose of raising slaves. When the instinct was once acquired, if carried out to a much less extent even than in our British F. sanguinea, which, as we have seen, is less aided by its slaves than the same species in Switzerland, natural selection might increase and modify the instinct, always supposing such modification to be of use to the species, until an ant was found as abjectly dependent on its slave as is the Formica rufescens.

Ants do not appear to be the only animals of which ants make slaves; for there seems to be at least one case in which these wonderful insects enslave insects of another species, which therefore may be said to stand to the ants in the relation of beasts of burden. The case to which I allude is one that is recorded in Perty's 'Intellectual Life of Animals' (2nd ed. p. 329), and is as follows:—

According to Audubon certain leaf-bugs are used as slaves by the ants in the Brazilian forests. When these ants want to bring home the leaves which they have bitten off the trees, they do it by means of a column of these bugs, which go in pairs, kept in order on either side by accompanying ants. They compel stragglers to re-enter the ranks, and laggards to keep up by biting them. After the work is done the bugs are shut up within the colony and scantily fed.

Wars.—On the wars of ants a great deal might be said, as the facts of interest in this connection are very numerous; but for the sake of brevity I shall confine myself to giving only a somewhat meagre account.

One great cause of war is the plundering of ants' nests by the slave-making species. Observers all agree that this plundering is effected by a united march of the whole army composing a nest of the slave-making species, directed against some particular nest of the species which they enslave. According to Lespès and Forel, single scouts or small companies are first sent out from the nest to explore in various directions for a suitable nest to attack. These scouts afterwards serve as guides to the marauding excursion. Forel saw several of these scouts of the species F. rufescens or Amazon carefully inspecting a nest of F. fusca which they had found, investigating especially the entrances. These are purposely made difficult to find by their architects, and it not unfrequently happens that after all precautions and inspections on the part of the invaders, an expedition fails on account of not finding the city gates.

When the scouts have been successful in discovering a suitable nest to plunder, and have completed their strategical investigations of the locality to their satisfaction, they return straight to their own nest or fortress. Forel has then seen them walking about on the surface of their nest for a long time, as if in consultation, or making up their minds. Then some of them entered the nest, soon after which hosts of warriors streamed out of the entrances, and ran about tapping each other with their heads and antennæ. They then formed into column and set out to pillage the nest of the slave ants. The following is the account which Lespès gives of such expeditions:—

They only take place towards the end of the summer and in autumn. At this time the winged members of the slave species (F. fusca and F. cunicularia) have left the nest, and the Amazons will not take the trouble to bring back useless consumers. When the sky is clear our robbers leave their town in the afternoon at about three or four o'clock. At first no order is perceptible in their movements, but when they are all gathered together they form a regular column, which then moves forward quickly, and each day in a different direction. They march closely pressed together, and the foremost always appear to be seeking for something on the ground. They are each moment overtaken by others, so that the head of the column is continually growing. They are in fact seeking the traces of the ants which they propose to plunder, and it is scent that guides them. They snuff over the ground like hounds following the track of a wild animal, and when they have found it they plunge headlong forward, and the whole column rushes on behind. The smallest armies I saw consisted of several hundred individuals, but I have also seen some four times as large. They then form columns which may be five mètres long, and as much as fifty centimètres wide. After a march, which often lasts a full hour, the column arrives at the nest of the slave species. The F. cuniculariæ, which are the strongest, offer keen opposition, but without much result. The Amazons soon penetrate within the nest, to come out again a moment later, while the assailed ants at the same time rush out in masses. During the whole time attention is directed solely to the larvæ and pupæ, which the Amazons steal while the others try to save as many as possible. They know very well that the Amazons cannot climb, so they fly with their precious burdens to the surrounding bushes or plants, whereto their enemies cannot follow them. They then pursue the retreating robbers and try to take away from them as much of their booty as possible. But the latter do not trouble themselves much about them, and hasten on home. On their return they do not follow the shortest road, but exactly the one by which they came, finding their way back by smell. Arrived at their nest, they immediately hand over their booty to the slaves, and trouble themselves no more about it. A few days afterwards the stolen pupæ or nymphæ emerge, without memory of their childhood, and immediately and without compulsion take part in all tasks.

According to Büchner's account,[30]—

From time to time the army makes a short halt, partly to let the rearguard close up, partly because different opinions arise as to the direction of the host, or because the place at which they are is unknown to them. Forel several times saw the army completely lose its way—an incident only once observed by Huber. Forel puts the number of warriors in such an army at from one hundred to more than two thousand. Its speed is on an average a mètre per minute, but varies much according to circumstances, and is naturally least when returning laden with booty. If the distance be very great, such bodily fatigue may at last be felt that the whole attack on the hostile nest is given up, and a retreat is begun; Forel once saw this happen after they had passed over a distance of two hundred and forty yards. Sometimes it seems as though, on coming within sight of the hostile nest, a kind of discouragement took possession of them, and prevented their making the attack. If the nest cannot at once be found, the whole army halts, and some divisions are sent forward to search for it, and these are gradually seen returning towards the centre. Forel also saw such an army only searching the first day, advancing zigzag, and with frequent halts, whereas on the following day it went forward to its aim swiftly and without delay, having found out the road. It seems that a single ant, even if it knows the way and the place, is not able alone to lead a large army, but that a considerable number must be employed in this duty. Mistakes as to the road occur with special ease during the return journey, because the several ants are laden with booty and cannot readily understand each other. Individual ants are then seen to wander about in every direction often for a long time, until they at last reach a spot known to them, and then advance swiftly to their goal. Many never come back at all. These mistakes easily occur when the robbers which have passed into a hostile nest do not come out again at the same holes whereby they entered, but by others at some distance—for instance, by a subterranean canal. Coming out thus in a strange neighbourhood, they do not know which way to take, and only some chance to find the right road during their aimless wanderings about, and recognise and follow it by smell. On the other hand, such mistakes scarcely ever happen to individuals in an unladen train, kept in good array. Other species of ants (F. fusca, rufa, sanguinea) know better how to manage under such circumstances than do the Amazons. The laden ones lay down their loads, first find where they are, and only take them up again after they have found their way. If the booty seized in the nest first attacked is too large to be all taken at once, the robbers return once, or oftener, so as to complete their work. . . . . The ants, as already said, have no regular leaders nor chiefs, yet it is certain that in each expedition, alteration of road, or other change, the decision during that event comes from a small knot of individuals, which have previously come to an understanding, and carry the rest and the undecided along with them. These do not always follow immediately, but only after they have received several taps on the head from the members of the 'ring.' The procession does not advance until the leaders have convinced themselves by their own eyesight that the main part of the army is following.

One day Forel saw some Amazons on the surface of a nest of the F. fusca seeking and sounding in all directions, without being able to find the entrance. At last one of them found a very little hole, hardly as large as a pin's head, through which the robbers penetrated. But since, owing to the smallness of the hole, the invasion went on slowly, the search was continued, and an entrance was found further off, through which the Amazon army gradually disappeared. All was quiet. About five minutes later Forel saw a booty-laden column emerge from each hole. Not a single ant was without a load. The two columns united outside and retreated together.

A marauding excursion of the Amazons against the F. rufibarbis, a sub-species of the F. fusca, or small black ants, took place as follows:—The vanguard of the robber army found that it had reached the neighbourhood of the hostile nest more quickly than it had expected; for it halted suddenly and decidedly, and sent a number of messengers which brought up the main body and the rearguard with incredible speed. In less than thirty seconds the whole army had closed up, and hurled itself in a mass on the dome of the hostile nest. This was the more necessary as the rufibarbes during the short halt had discovered the approach of the enemy, and had utilised the time to cover the dome with defenders. An indescribable struggle followed, but the superior numbers of the Amazons overcame, and they penetrated into the nest, while the defenders poured by thousands out of the same holes, with their larvæ and pupæ in their jaws, and escaped to the nearest plants and bushes, running over the heaps of their assailants. These looked on the matter as hopeless, and began to retreat. But the rufibarbes, furious at their proceedings, pursued them, and endeavoured to get away from them the few pupæ they had obtained, by trying to seize the Amazons' legs and to snatch away the pupæ. The Amazon lets its jaws slip slowly along the captive pupa, as far as the head of its opponent, and pierces it, if it does not, as generally happens, draw back. But it often manages to seize the pupa at the instant at which the Amazon lets it go and flies with it. This is managed yet more easily when a comrade holds the robber by the legs, and compels it to loose its prey in order to guard itself against its assailant. Sometimes the robbers seize empty cocoons and carry them away, but they leave them on the road when they have discovered their mistake. In the above case the strength of the rufibarbes proved at last so great that the rearguard of the retreating army was seriously pressed, and was obliged to give up its booty. A number of the Amazons also were overpowered and killed, but not without the rufibarbes also losing many people. None the less did some individuals, as though desperate, rush into the thickest hosts of the enemy, penetrated again into the nest, and carried off several pupæ by sheer audacity and skill. Most of them left their prey to go to the help of their comrades when assailed by the rufibarbes. Ten minutes after the commencement of the retreat all the Amazons had left the nest, and, being swifter than their opponents, they were only pursued for about halfway back. Their attack had failed on account of a short delay!

On another occasion observed by Forel, in which several fertile Amazons also took part and killed many enemies, the nest was thoroughly ravished, but the retreat was also in this case very much disturbed and harassed by the superior numbers of the enemy. There were many slain on both sides. That in spite of the above-mentioned unanimity different opinions among the members of an expedition sometimes hinder its conduct, the following observation seems to show:—An advancing column divided after it had gone about ten yards from the nest. Half turned back, while the other half went on, but after some time hesitated and also turned back. Arrived at home, it found those which had formerly turned back putting themselves in motion in a new direction. The newly returned followed them, and the reunited army, after various wheelings, halts, &c., at last turned home again by a long way round. The whole business looked like a promenade. But apparently different parties had different nests in view, while others were entirely against the expedition. Yet perhaps it was only a march for exercise.

Outer obstacles do not, as a rule, hinder the Amazons when they are once on the march. Forel saw them wade through some shallow water, although many were drowned in it, and then march over a dusty high road, although the wind blew half of them away. As they returned, booty-laden, neither wind, nor dust, nor water could make them lay down their prey. They only got back with great trouble, and turned back again to bring fresh booty, although many lost their lives.

The following is also quoted from Büchner's excellent epitome of Forel's observations in this connection:—

The most terrible enemy of the Amazons is the sanguine ant (F. sanguinea), which also keeps slaves, and thereby often comes into collision with the Amazons on their marauding excursions. It is not equal to it in bodily strength or fighting capacity, but surpasses it in intelligence; according to Forel it is the most intelligent of all the species of ants. If Forel, for instance, poured out the contents of a sack filled with a nest of the slave species near an Amazon nest, the Amazons apparently generally regarded the tumbled together heap of ants, larvæ, pupæ, earth, building materials, &c., as the dome of a hostile nest, and took all imaginable but useless pains to find out the entrances thereinto, leaving on one side for this investigation their only object, the carrying off the pupæ; but the sanguine ants under similar circumstances did not allow themselves to be deceived, but at once ransacked the whole heap.

On another occasion, while a procession of Amazon ants was on its way to plunder a nest of F. fusca, before it arrived Forel poured out a sack-full of sanguine ants, and made a break in the nest:—

The sanguine ants pressed in, while the fusca came out to defend themselves. At this moment the first Amazons arrived. When they saw the sanguine ants they drew back and awaited the main army, which appeared much disturbed at the news. But once united, the bold robbers rushed at their foes. The latter gathered together and beat back the first attack, but the Amazons closed up their ranks and made a second assault, which carried them on to the dome and into the midst of the enemy. These were overthrown, as well as a number of F. pratensis, which Forel at this moment poured out on the nest. The conquerors delayed for a moment on the dome after their victory, and then entered the nest to bring out a little of the valuable booty. A few Amazons which were mad with anger did not return with the main army, but went on slaughtering blindly among the conquered and the fugitives of the three species, fusca, pratensis, and sanguinea.

The ravished rufibarbes once became so desperate at their overthrow that they followed the robbers to their own nest, and the latter had some trouble in defending it. The rufibarbes let themselves be killed in hundreds, and really seemed as though they courted death. A small number of the Amazons also sank under the bites of their enemies. The nest contained slaves of the rufibarbis species, which on this emergency fought actively against their own race. There were also slaves of the species fusca, so that the nest included three different species of ants.

The same nest is often revisited many times on the same day or at different periods, until either there is no more to steal, or the plundered folk have hit upon better mode of defence. A column which was in the act of going back to such a plundered nest turned when halfway there, and halted, apparently on no other ground than because it had met the rearguard of the army, and had learned that the nest was exhausted, and that there was nothing more to be had there. The robbers then went off to a rufibarbis nest which was in the neighbourhood, and killed half the inhabitants while plundering the nest. The surviving rufibarbes returned after the robbery and brought up new progeny; but thirteen days later the Amazons again reaped a rich harvest from the same nest. The Amazon army often severs itself into two separate divisions when there is not enough for both to do at the same spot. Sometimes one division finds something and the other nothing, and they then reunite. If any obstacle be placed in their way they try to overcome it, in doing which some leave the main army, lose themselves, and only find their way home again with difficulty. Forel has tried to establish the normal frequency of expeditions, and found that a colony watched by himself for a space of thirty days sent out no less than forty-four marauding excursions. Of these about eight-and-twenty were completely, nine partially, and the remainder not at all successful. He four times saw the army divide into two. Half the expeditions were levelled against the rufibarbes, half against the fuscæ. On an average a successful expedition would bring back to the colony a thousand pupæ or larvæ. On the whole, the number of future slaves stolen by a strong colony during a favourable summer may be reckoned at forty thousand!

The internecine battles which occasionally break out among the Amazons themselves are naturally the most cruel. They tear each other to pieces with incredible fury, and knots of five or six individuals which have pierced each other may be seen rolling over each other on the ground, it being impossible to distinguish between friend and foe. Civil wars among men are also known to be the most embittered and the most bloody.

The mode of attack practised by the other best known species of slave-making ant, sanguinea, is somewhat different:—