INSTINCT is a term which does not admit of rigid definition, because, as ordinarily used, the meaning of the term is not rigidly fixed. But for the purposes of scientific expo-sition from a biological point of view the nearest approach we can make to such a definition is perhaps the following: Instinct is a generic term comprising all those faculties of mind which lead to the conscious performance of actions that are adaptive in character, but pursued without neces-sary knowledge of the relation between the means employed and the ends attained. We must, however, remember that instinctive actions are very commonly tempered with what Pierre Huber calls " a little dose of judgment or reason." But although reason may thus in varying degrees be blended with instinct, the distinction between the two is sufficiently precise; for reason, in whatever degree present, only acts upon a definite and often laboriously acquired knowledge of the relation between means and ends. Moreover, adjustive actions due to instinct are similarly performed by all individuals of a species under the stimulus supplied by the same appropriate circumstances, whereas adjustive actions due to reason are variously performed by different individuals. Lastly, 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 cir-cumstances, and serve to meet novel exigencies which may never before have occurred even in the life history of the individual.
All instincts probably arose in one or other of two ways. Origin ol (1) By the effects of habit in successive generations, instincts, mental activities which were originally intelligent become, as it were, stereotyped into permanent instincts. Just as in the life-time of the individual adaptive actions which were originally intelligent may by frequent repetition become automatic, so in the life-time of the species actions originally intelligent may, by .frequent repetition and heredity, so write their effects on the nervous system that the latter is prepared, even before individual experience, to perform adaptive actions mechanically which in previous generations were performed intelligently. This mode of origin of instincts has been appropriately called the "lapsing of intelligence." (2) The other mode of origin consists in natural selection, or survival of the fittest, con-tinuously preserving actions which, although never intelli-gent, yet happen to have been of benefit to the animals which first chanced to perform them. Thus, for instance, take the instinct of incubation. It is quite impossible that any animal can ever have kept its eggs warm with the intelligent purpose of hatching out their contents, so we can only suppose that the incubating instinct began by warm-blooded animals showing that kind of attention to their eggs which we find to be frequently shown by cold-blooded animals. Thus crabs and spiders carry about their eggs for the purpose of protecting them ; and if, as animals gradually became warm-blooded, some species for this or for any other purpose adopted a similar habit, the impart-ing of heat would have become incidental to the carrying about of the eggs. Consequently, as the imparting of heat promoted the process of hatching, those individuals which most constantly cuddled or brooded over their eggs would, other things equal, have been most successful in rearing progeny ; and so the incubating instinct would be developed without there having been any intelligence in the matter.
That many instincts must have been developed in this way is rendered evident by the following considerations. (1) Many instinctive actions are performed by animals too low in the scale to admit of our supposing that the adjustments which are now instinctive can ever have been intelligent. (2) Among the higher animals instinctive actions are performed at an age before intelligence, or power of learning by individual experience, has begun to assert itself. (3) Considering the great importance of instincts to species, we are prepared to expect that they must be in large part subject to the influence of natural selection. As Mr Darwin observes, ''it will be universally admitted that instincts are as important as corporeal structures for the welfare of each species under its present conditions of life. Under changed conditions of life it is at least possible that slight modifications of instinct might be profitable to a species; and if it can be shown that instincts do vary ever so little, then I can see no diffi-culty in natural selection preserving and continually accumulating variations of instinct to any extent that was profitable. It is thus, I believe, that all the most complex and wonderful instincts have originated."
But here it is of importance to note that there is no reason why instincts should be restricted to one or other of these two modes of origin. On the contrary, there seems to be every reason to suppose that many instincts may have had, as it were, a double rootintelligent adjustment and natural selection blending their influences to a joint production. For example, the grouse of North America display the curious instinct of burrowing a tunnel just below the surface of the snow. In the end of this tunnel they sleep securely; for, when any four-footed enemy approaches the mouth of the tunnel, the bird, in order to escape, has only to fly up through the thin covering of snow. Now in this case the grouse probably began to burrow for the sake of protection, or concealment, or both, and if so, thus far the burrowing was probably an act of intelligence. But the longer the tunnel the better would it have served the purposes of escape, and therefore natural selection would almost certainly have tended to preserve the birds which made the longest tunnels, until the utmost benefit that length of tunnel could give had been attained. And similarly the origin of many other instincts may be satisfactorily explained by thus supposing the combined operation of two causesintelligent adjustment and natural selectionwhere there is a difficulty in explaining their origin as due to either cause alone. And if even in fully formed instincts we often find " a little dose of judgment or reason," we can have no cause to doubt that in the formation of instincts by natural selection such small admixtures of judgment or reason may often greatly assist the process, while, conversely, it is even more evident that an instinct which is developing from the habitual perform-ance of an intelligent action might be greatly assisted by natural selection favouring the individuals which most frequently or most promptly performed that action. Varia- It is necessary to the above interpretation of the origin bihty. 0f instincts that the latter should not be immutably fixed. A few words may therefore be added to show that the view commonly entertained as to the unalterable character of instincts is erroneous. As a matter of fact, instincts are eminently variable, and therefore admit of being modified as modifying circumstances may require; their variability gives them plasticity whereby they may be moulded always to fit an environment, however continuously the latter may be subject to gradual change.
For the sake of brevity we may confine our attention to a single instinct, and for the sake of procuring a good test we may again take as our example the instinct of incuba-tion. This affords a good test because it must be regarded, not merely as one of the most important, but also as one of the oldest of instincts, and therefore one which for both these reasons we should deem least likely to exhibit variability. Yet we find it to exhibit variability in every imaginable direction. Thus the complicated effects of domestication and artificial selection on some of our breeds of poultry appear to have almost completely destroyed this instinct, while in other breeds it remains intact, if indeed it has not even been intensified. Among the latter breeds experiment shows that the natural period of incubation may be indefinitely prolonged by substituting " dummies " for eggs, while the following experiment, which we owe to Mr Spalding, shows " how far the time of sitting may be interfered with in the opposite direction. Two hens," he says, "became broody on the same day, and I set them on dummies. On the third day, I put two chicks a day old to one of these two hens; she pecked at them once or twice, seemed rather fidgety, then took to them, called them to her, and entered on all the cares of a mother. The other hen was similarly tried, but with a very different result; she pecked at the chickens viciously, and both that day and the next stub-bornly refused to have anything to do with them." Similarly the period of maternal supervision after the chickens have been hatched admits of being greatly modified, as is proved by some, experiments made and published several years ago by the present writer. In one of these experiments there was given to a Brahma hen a pea-fowl's egg to hatch ; the hen was an old one, and had previously reared several broods of ordinary chickens. A pea-chicken requires a much longer period of maternal care than does an ordinary chicken, and for the wonderfully long period of eighteen months the old Brahma hen continued to pay unremitting attention to her supposed offspring. Through all this time she never laid any eggs, and eventually the separation seemed to take place from the side of the peacock. In other cases, however, where the conditions of the experi-ment were exactly parallel, the pea-chickens were abandoned by their Brahma mothers at the time when the latter ordi-narily abandon their chickens. But not only will a hen thus take to a brood of birds so unlike her natural chickens as are pea-fowl, and adapt her instincts to their peculiar needs; she may even take to young animals belonging to a different class, and adapt her instincts to their still more peculiar needs. Thus the writer gave to a hen, which for several weeks had been sitting on dummies, three newly-born ferrets ; she took to them almost immediately, and remained with them for more than a fortnight, when they were taken away from her. During the whole of this time she had to sit upon the nest, for of course the young ferrets were not able to follow her about as young chickens would have done. Two or three times a day she would fly off her nest, calling upon her brood to follow; but, on hearing their cries of distress from cold, she always returned im-mediately, and sat with patience for six: or seven hours more. She only took one day to learn the meaning of these cries, and after that she would always run in an agitated manner to any place where the crying ferrets were concealed. Yet it would not be possible to conceive a greater contrast than that between the shrill piping note of a young chicken and the hoarse growling noise of a young ferret. It is of importance to add that the hen very soon learnt to accommodate herself to the entirely novel mode of feeding that her young ones required; for, although at first she showed much uneasiness when the ferrets were taken from her to be fed, before long she used to cluck when she saw the milk brought, and surveyed the feeding with satisfaction. But she never became accustomed to the attempt of the ferrets at sucking, and to the last used now and then to fly off the nest with a cackle when nipped by the young mammals in their search for the teats.
Enough then has been said on the variability of instinct Heredi-to show that there is supplied to natural selection abundant talT opportunity for the development of new and more highly tr?n!" wrought instincts from previously formed and less elabor-ated instincts. But in order to show that this opportunity has been utilized it is not enough to show that hereditary instinct may be modified by individual experience ; it must also be shown that such a modification when successively repeated through a number of generations itself becomes inherited. Now, although the evidence on this point is necessarily scanty, it is sufficient for the purpose here required. The evidence is scanty because there are only a very few cases in which human observation has, as it were, the opportunity of watching the continuance of effects of recently acquired or altered experiences over a number of generations; but in the few cases in which we have this opportunity we find good evidence that new or changed experience, when continued over a number of generations, is bequeathed to future generations as a legacy of intuitive knowledge, and that any newly acquired adjustive actions may in time be similarly transmitted as instinctive actions which no longer require to be separately learnt by each individual. Perhaps the best instance that can be quoted is that of the many species of birds, and some mammals, which when first found by man on oceanic islands were perfectly tame, but whose descendants now show a dread of man which is, in the most rigorous sense of the word, instinctive.
The only other instances in which we have an oppor-tunity of actually observing the transmission of newly acquired mental habits are those in which such habits have been artificially taught to domesticated animals. It so happens that these instances are very few in number, but it is not too much to say that, in all the cases where such habits have been taught for a long series of generations, some tokens of their hereditary transmission may now be observed. Thus, to quote Mr Darwin, whose accuracy on such a subject is not likely to be disputed by any one, " it cannot be doubted that young pointersI have myself seen a striking instancewill sometimes point and even back other dogs the very first time they are taken out; retriev-ing is certainly in some degree inherited by retrievers; and a tendency to run round instead of at a flock of sheep by shepherd dogs. If we were to behold one kind of wolf when young and without any training, as soon as it scented its prey, stand motionless like a statue, and then slowly crawl forward with a peculiar gait, and another kind of wolf rushing round instead of at a herd of deer, and driv-ing them to a distant point, we should assuredly call these actions instinctive. Domestic instincts, as they may be called, are certainly much less fixed than natural instincts, but they have been acted on by far less rigorous selection, and have been transmitted for an incomparably shorter period, under less fixed conditions."
Now these three habits, or mental attainments, are the only ones that have been systematically taught to any animals for a number of generations, and the fact that they all show a marked tendency to become intuitive may be taken as lending a greater amount of confirmation to the present theory of the origin of instincts than we might on a priori grounds be led to expect. The only other facts bearing upon this point are those which are thus tersely rendered by Mr Darwin. " How strongly these domestic instincts, habits, and dispositions are inherited, and how curiously they become mingled, is well shown when different breeds of dogs are crossed. Thus it is well known that a cross with a bull-dog has affected for many generations the courage and obstinacy of greyhounds, and a cross with a greyhound has given to a whole family of shepherd dogs a tendency to hunt hares. These domestic instincts resemble natural instincts, which in like manner become curiously blended together, and for a long time exhibit traces of the instincts of either parent." Special The above doctrine as to the nature, origin, and development of difficul- instincts serves very satisfactorily to explain nearly all the enor-ties. mous number of instincts with which we are acquainted. There are, however, several special cases where there is still some difficulty in applying the above doctrine as a full and satisfactory explanation of the observed facts. This article may therefore fitly conclude with a brief enumeration of these cases.
1. The so-called '' migratory instinct" is one that is still shrouded in much obscurity. The main difficulty with regard to it is to account for the "sense of direction," whereby the animals are guided to their destinations. Thus, for instance, many migratory birds fly at night, when it would seem impossible that they can be guided on their way by the sight and memory of landmarks. More-over, it is asserted on good authority that among some species it is -the habit for the young brood to fly separately by themselves, or apart from the older birds, and therefore to travel over enormous tracts of land and sea without either guidance or previous experi-ence of the way ; such is unquestionably the case with the young of the cuckow. Lastly, it is certain that several species fly across immense tracts of ocean, where it is impossible that they can be guided on their way by landmarks. Several theories have been pro-pounded to account for these facts ; but, as none of them are satis-factory, we need not here occupy space with their enumeration.
2. Closely allied to, if not identical with, this so-called '' sense of direction" as manifested in migration, is the faculty whereby various species of animals which may not be migratory in their habits are enabled to find their way over greater or less distances. This has been called the "homing" faculty, and is chiefly mani-fested by various species of domesticated mammals. It is very generally believed that it is also manifested by carrier-pigeons ; as a matter of fact, however, there is no trustworthy instance on record of one of these birds having found its way back over a tract of country with which it was previously unacquainted. In order that a carrier-pigeon should find its way home, it is necessary first to teach the animal, by flying it at a series of points along the route, the landmarks of the country which it is afterwards to traverse. But, although the " sense of direction " may be a figment as regards the carrier-pigeon, there can be no doubt that it is a fact as regards many species of our domesticated mammals. Thus the evidence is unequivocal with regard at least to dogs, cats, horses, sheep, pigs, and cattle. Judging from hitherto unpublished correspondence received from Australia and South America, there seems to be practically no limit to the distance over which these animals may be able to return ; and, what is of more importance, there can be no doubt that these animals, when finding their way home, do not require to traverse the exact routes by which they came; on the contrary, they generally seem to select the shortest or the straightest course, however circuitous the way may have been by which they were taken ; or, if their outward journey is over two sides of a tri-angle, their homeward journey will probably be taken over the third side. The sense by which they are guided therefore cannot be, as has been suggested by more than one eminent naturalist, the sense of smell ; and for the same reasons it cannot be either the sense of sight or that of hearing. More plausible is the hypothesis that the faculty consists in an automatic process of "brain registration," every change of direction in the outgoing journey leaving behind it a record in the cerebral nervous system, and therefore in the mind of the animal, so that as a total result the general direction of the starting place is retained in the memory,just as we are ourselves able in a smaller degree to preserve our general sense of direction when winding through the streets of a town. One great difficulty attaching to this view appears to be that the animals in question are able to find their way home over land even when they have made their outward journey by sea, for it is evident that the difficulties of "brain registration " must in such cases be indefinitely increased, not only by the many meaningless movements of a vessel at sea, but still more by the fact that the changes of direction made by the vessel, being made in long and easy curves, and without muscular effort on the part of the animals, are movements which we can scarcely suppose to be appreciated by the cerebral organization of the animals. On the whole, therefore, with regard to the faculty of "homing," as with regard to the analogous if not identical faculty exhibited in migration, it can only be said that further investigation is required in order to explain that which, in the present state of our knowledge, must properly be regarded as inexplicable.
3. Mr Darwin has pointed out a serious difficulty lying against his theory of the origin of instincts by natural selection, and one which, as he justly remarks, it is surprising that no one should have hitherto advanced against the well-known doctrine of inherited habit, as taught by Lamarck. The difficulty is that among various species of social insects, such as bees and ants, there occur " neuter " or asexual individuals, which manifest entirely different instincts from the other or sexual individuals, and as the neuters cannot breed it is difficult to understand how their peculiar and distinctive instincts can be formed by natural selection, which, as we have seen, requires for its operation the transmission of mental faculties by heredity. The only possible way in which this difficulty can be met is the way in which it has been met by Mr Darwin, viz., by supposing "that selection may be applied to the family as to the individual." " Such faith may be placed in the power of selection that a breed of cattle always yielding oxen with extraordinarily long horns could, it is probable, be formed by carefully watching which individual bulls and cows, when matched, produced oxen with the longest horns ; and yet no one ox would ever have propa-gated its kind" ; and similarly, of course, with regard to the instincts of neuters. As Mr Darwin has argued out this difficulty at length, it seems unnecessary to say more with regard to it than that he has shown it to be not so formidable as to exclude his doc-trine as fully explanatory of such .cases, when we have already accepted his doctrine as explanatory of other cases.
i. There are two or three other special instincts of minor importance the explanation of which is not as yet completely clear. Thus it is not yet ascertained what hereditary influence it is that leads the Norwegian lemming periodically to migrate westwards, with the result that enormous numbers of the species are destroyed by drowning. But there can be little doubt that this influence, what-ever it was, was originally of benefit to the species, for it would be a case standing out of all analogy if this instinct should from its first origin have been, as it now appears to be, detrimental. The only other instance that could be pointed to as wearing any such appearance is that which has been alleged, but on very doubtful evidence, with regard to the scorpion committing suicide by stinging itself to death wdien surrounded by a ring of fire. It may be here incidentally observed that the fact of all the innumer-able multitude of animal instincts, with the exception of the two dubious cases just mentioned, being of obvious use to the species which manifest them, may properly be taken as the strongest possible evidence of the theory that ascribes all instincts to the operation of natural selection.
5. Lastly, we have an instinct which is pointed to by Mr Mivart as one that cannot be explained by the influence of natural selection, or, as he would appear to suggest, by the operation of any other natural cause. This instinct is manifested by a certain wasp-like animal, and consists in this animal stinging spiders in the particular part of the cephalo-thorax which contains the principal nervous centre. The effect of stinging this nervous centre is that of paralys-ing the spider without killing it, and the spider in this maimed condition is then stored up with the larva?, of the fly, to serve as their food when they quit the egg. It will be observed that there is here no question as to the utility of the instinct to the species which manifests it, and the difficulty to which Mr Mivart points consists merely in understanding how the insect was in the first instance led to sting the spiders in precisely the right spot to pro-duce the particular results required. The answer to this single remaining difficulty is that as yet the case has not been sufficiently observed with a view to a possible solution of the difficulty. I seems, for instance, not at all improbable that the striking of the spider's ganglion by the sting of the wasp is, as it were, wholly accidental, being determined only by the circumstance that both the ganglion and the sting are organs which occur in the median line of their respective possessors. Whether or not this is the explanation of the supposed difficulty, it at least seems clear that the latter is not one of any considerable magnitude. (G. J. R. )
The above article was written by G. J. Romanes, LL.D., F.R.S.