posed: more and more characters, studied in natural sur- roundings, are found to be quite indifferent, to offer no discoverable advantage or disadvantage to the organism possessing them. AVhat does it matter whether the snail's shell be twisted to the right or to the left; whether the pigment of the elytra of the beetle be arranged in continu- ous or broken lines; whether the young noddy is gray or white? The answer of the extreme selectionist has al- ways been: ' ' I do not know the value of the character, but it must have some worth. Otherwise it could not exist." But the discovery that in a species of butterflies two distinct types of females, widely different in color, may live side by side, each breeding true in Mendelian ratios, quickly raises the question whether the coloration of these insects has any protective or selective value. Such cases of seemingly indifferent characters might be multi- plied indefinitely. On p. 113 are cited certain specific ex- amples of non-adaptive types of activity. Parker has re- cently argued in similar vein and has extended the con- ception of indifferent characters to the whole range of ani- mal reactions. He says: " It is my opinion, at least, that many animal reactions which we have been accustomed to call adaptations should not be thus desig- nated, and that the difficulties that we often meet in attempting to account for such reactions are due to our consideration of them from the standpoint of adaptations when in reality they are far from being such."
" The majority of animal reactions are, in all probability, neither conspicuously advantageous nor disadvantageous to the life of the individual. They are dependent chiefly upon the material composition of the given organism, and, so long as they are relatively indifferent to the continuance of life, they pass without special consequence."
Effect of natural selection upon mutations. — To accept the mutation hypothesis as at least one of the ways in which diversified structure arises, does not necessitate throwing overboard the concept of natural selection. A moment's consideration shows that many animals and plants do thrive in the environment in which they are found, i.e.. that they are adapted to meet environmental demands: furthermore, that there is often an almost EFFECT OF SELECTION ON MUTATION 169 perfect correlation in the functioning of the different parts of the animal. The growing conception is that while natu- ral selection has no part in shaping such adaptations and such correlations, it nevertheless still retains the function of killing off those organisms unfitted to live in the environ- ment in ivhich they are placed. A simpler way now to express what is meant is to avoid the use of the term selection and to state specifically what probably happens to new races. It is clear, in the first place, that a mutant when it is thrown off, must bring wdth it organs of reproduction. Furthermore, it must possess certain instinctive modes of procuring food and a mechanism for the digestion of food and for the excretion of waste products. Depending also upon the environment where it first appears it must bear essential mechanisms of attack and defense or else of flight. If the mutant falls by chance where, wdth its equipment, it can live and reproduce its kind, it thrives. On the other hand, if this same mutant comes into existence in a locality where its equipment is not adequate, it perishes. If these simple (or complex) structural and functional demands are satisfied, the animal may possess — and often does pos- sess— a host of structural and functional characteristics which have no necessary role in its daily life (this is incon- ceivable on the Darwinian concept of evolution). Un- questionably the reason there are so many " adaptive in- stincts " on record and so few " unadaptive " ones is due to the fact that the naturalist has generally found what he sought.
Ability to form habits enables the animal to supple- ment a faulty inheritance. — When we examine the higher vertebrates, especially the mammals, shortly after birth, w^e find, as was brought out in the preceding chapter, that they are not born adapted, and furthermore, that when their instincts do appear they are oftentimes by no means per- fect in their action. It is here that habit reveals itself as being a positive factor in evolution. On the basis of w^hat we now know to be true of habits, it seems possible to modify the demands we made above upon a mutant (if it is to survive) by saying that it either must appear 170 CONCERNING THE ORIGIN OF INSTINCTS adapted to the environment, or else it must appear with sufficient plasticity to enable it readily to form habits. This appears most clearly in the work of Breed, already cited. The chick is born with the pecking instinct suffi- ciently developed to keep the organism alive. Adaptation becomes perfect through habit.
Some special forms of asserted adaptations. — Through- out the literature of the Darwinist one finds voluminous reference to {a) protective resemblance; (5) warning coloration; and (c) sexual dimorphism. It is claimed by them that these phenomena may be explained by natural selection. Since such phenomena, however produced, are of especial interest to the students of behavior in view of the fact that the sense organs of animals are involved in their discussion, it seems worth while to consider them for a moment (although they offer no great difficulty, so far as explanation goes, on the hypothesis of mutation and selection already considered).
{a) Protective resemblance. — "We have touched upon the facts included under this heading on p. 124. Protective resemblance includes all the structural modifications of animals which tend to make them less conspicuous to their enemies or to make them resemble other animals which are not preyed upon by those enemies (mimicry). Many animals seem to resemble in color or pattern the back- grounds against which they normally live. Some slight evidence has been collected showing that mimicry is ' ' pro- tective." E. B. Poulton and C. B. Saunders noted the survivals among a large number of the pupse of the butter- fly {Vanessa urtica) in different localities and found that those placed most conspicuously for the human eye were most readily found and devoured by birds. Cesnola fastened green and brown Mantis to plants of similar and contrasting color with the result that birds found the con- spicuously placed ones, the ones upon plants of contrast- ing color, most readily. (It might be noted here that vision in birds seems closely comparable with that of man, in so far as it has been tested, and hence the field studies are sup- ported by those of the laboratory.) Pearl has shown that WARNING COLORATION 171 while self-colored chickens (those which are of a uniform color all over the body) are more conspicuous for man than are barred or speckled birds, there is no larger percentage of the former killed by natural enemies than of the latter. Inconspicuousness for man does not, in this case, guarantee protection from other enemies. Doubt has been raised as to the reality of many of the more perfect cases of pro- tective coloration, and only careful field studies can estab- lish the adaptive character of the doubtful cases.
(h) Warning coloration. — The theory of warning colora- tion supposes that brilliant and conspicuous colors in ani- mals are associated with effective defensive weapons. Such conspicuous characters serve as '' danger " or " warn- ing " signals to protect animals so equipped from attack. The theory has little factual support. Przibram has col- lected a large amount of evidence to show that warningly colored insects and their mimics are eaten readily by many other animals, yet this evidence shows only that the pro- tection afforded by the defensive organs is not perfect, not that the insects are not at all protected by their conspicu- ousness. Reighard has shown conclusively that the bril- liantly colored coral reef, fishes do not possess " that com- bination of conspicuousness with unpleasant attributes necessary to the theory of warning coloration." The ordi- nary gray snappers which frequent the docks around juxta- tropical ports feed greedily upon the gaudily colored coral reef fishes when opportunity is offered under experimental conditions. The delicate coral reef fishes escape annihila- tion because of the protection offered by the reefs and be- cause of their agility.
(c) Sexual dimorphism. — To the theory of sexual selec- tion Darwin attached great importance. In its extreme form the theory supposes not only a higlily developed sen- sory acuity upon the part of the selecting animal but it assumes that animals have also definite color, form, and size " preferences," i.e., a highly developed aesthetic sense. There is as yet no evidence that this is the case nor can such evidence be gathered until it has been proved that the animals in question have a sensory equipment sufficiently 172 COXCERXIXG THE ORIGIN OF INSTINCTS delicate to form a basis for such a selection. The problem is still complicated, even when such anthropomorphic con- ceptions are ruled out. The phenomena which one seeks to explain by the theory of sexual selection are chiefly those of sexual dimorphism, differences in structure and behavior of the two sexes, although there are other characters which may be subject to the same selective agents.
Many cases of sexual dimorphism do not demand any active selection by a member of one sex between two mem- bers of the opposite sex. The vigorous struggle between males for the possession of the females, mutual fitness for mating, the care of the young, all such activities would lead naturally to the production of sexual dimorphism by a se- lection which involved no sensory discrimination by mem- bers of either sex. In any given case of dimorphism it is extremely difficult to determine what part may have been played by this impersonal struggle and what part by a true preference for some earlier mutants. Certainly the truth or falseness of the theory can be demonstrated only by ex- perimental methods and as yet but little evidence has been obtained by their use.
The experimental evidence of sexual selection. — An- drews, Chidester, Holmes, and others have shown that smell and taste are the chief sensorj^ agents in the mating of arthropods. Mayer experimented with a moth (CaUosamia p7vmethea), in which the males and females are different in color, in order to test the preferences and selective activities of the females.
" In order to test this hypothesis I cut off the wings of a number of females, leaving only short stumps, from which all the scales were carefully brushed. Male wings were then neatl,y glued to the stumps, and thus the female presented the appearance of the male. Under these conditions the males mated witli the females quite as readily as they would have done under normal conditions."
" I tried the experiment of gluing female wings upon the male. Here again the mating seemed to occur with normal frequency, and I was unable to detect that the females displayed anj^ unusual aversion towards their effeminate-looking consorts."
" It is also interesting to note that normal males pay no attention to males with female wings."
"In another series of experiments the wings were cut entirely off the males and females and the scales brushed off their bodies; EXPERIMENTS ON SELECTION 173 and yet these shabby males were readily accepted by normal females, nor could I see that normal males displayed any aversion to mating with wingless females."
" We are therefore forced to conclude that the melanic coloration of the male has not been brought about through the agency of sexual selection on the part of the female."
More recently Mayer and Soule have repeated these ex- periments with another species and verified the results in all respects save one: wingless males were selected against by normal females but mated successfully w^ith blind ones. Sturtevant has taken up the same question with the fruit-fly and has found that vigor is the chief factor in determining the matings. Since certain eye colors of Drosophila are associated with greater or less vigor, there results a quite definite selection of eye color, but such a selection could not be the determining agent in producing a sexual dimor- phism. From this and other less direct evidence IMorgan is inclined to rule out sexual selection as a factor in evolu- tion, ascribing sexual dimorphism rather to the persistence of indifferent mutations which have arisen in the sex chromosomes. In contradiction to this view a certain amount of evidence may be cited in favor of sexual selection, which, although not conclusive, still keeps the problem open. Sturtevant (reported by Morgan, 1913) removed the wings from male fruit-flies and offered females a choice between normal and wingless males. Neither was pre- ferred, mating with wingless and normal males occurring in equal numbers. But when the males were placed separately with the females the average interval before mating was for the normal males 18 minutes, for wingless males, 40 minutes. This is interpreted as meaning that the female is more readily excited by the normal males. It was not a determining factor for selection of confined individuals, but it should be pointed out that among unconfined indi- viduals where courtship is subject to interruption the quicker excitement of the female by any male would lead to an effective selection of that male. In a recent paper Davenport has stated that he finds a selection against certain brilliant colors in the cock, and evidence for color vision in the fowl at least offers a basis for such selection.
174 CONCERNING THE ORIGIN OF INSTINCTS Moreover, the behavior of animals in experiments where attempts are made to hybridize certain species should not be overlooked. The zebra can with difficulty be induced to mate with the horse, while crosses between the zebra and ass are made without difficulty (Rommel). In a very suggestive paper Pearl has called attention to the fact that the jacks kept by mule breeders acquire reactions that are somewhat comparable to those called out by sexual symbols or fetishes in man. The problem of sexual selection has never been approached in a systematic manner and there is really no evidence at present upon which any general con- clusions may be based. Cases such as those cited above in- dicate that, in higher mammals at least, the problem may be seriously complicated by the habits of the animals, and accordingly that sexual selection is not a general or a racial problem, but one which more narrowly concerns the indi- vidual.
III. Inheritance of Acquired Characters Lamarck's laws. — In a note on p. 161 we expressed the view that there is a growing body of evidence to show that certain types of somatic variations are heritable. We wish now briefly to consider this evidence. As the individual develops it becomes better fitted to its environment through the perfection of instinctive mechanisms and the formation of habits. Early in the history of evolution the possibility that such ontogenetically acquired adaptations might be transmitted to the offspring suggested itself. Lamarck gave the most thorough formulation to this view in his two well known laws: " First laio. — In every animal which has not reached the limits of development the frequent and sustained use of any organ strengthens that organ little by little, it develops, enlarges and assumes a strength proportional to the amount of use; on the contrary, the failure to use any organ constantly causes it to grow weaker, it deteriorates, loses its function and finally disappears.
" Second law. — All that nature has forced the individual to acquire or lose by the influence of circumstances to which its race has long been exposed, hence all that has been acquired or lost by the predominating use or disuse of any organ, has been preserved in later EXPERIMENTS ON SELECTION 175 reproduction to later individuals by inheritance, provided that the modifications have been acquired by the two sexes in common or by those individuals which have given birth to the new race."
The followers of Lamarck carried this theory to still fur- ther extremes, assuming that all characters acquired by the individual as the result of the action of environment were inherited. Darwin accepted Lamarck's view of the value of use and disuse in evolution, calling upon this principle whenever his own theory of natural selection seemed in- adequate to account for the facts observed. As soon as an attempt was made to verify the theory by experimental evi- dence, the strongest possible objection to it appeared — the theory did not seem to work — the tails might be cut from mice for many generations without reducing the size of the tails of the progeny; the strong arm of the blacksmith was not transmitted to his son. Then came Weismann's theory of the continuity of the germ-plasm which showed the diffi- culties in the way of the transmission of bodily characters to the germ cells. Lamarck's theory was practically dropped for many years.
Recent experiments on the inheritance of acquired characters. — During the past few years the theory has been revived by several investigators who seemingly have obtained experimental evidence of the existence of the in- heritance of certain types of somatic variations. The method which has secured this evidence was first indicated by the experiments of Marie V. Chauvin upon the Mexican Axo- lotl. Under favorable conditions this salamander retains the aquatic habit throughout life, breathing with external gills and never forming functional lungs. Unfavorable conditions in nature will, however, lead to the loss of the gills, to the formation of functional lungs, and to the as- sumption of a terrestrial mode of life. In the laboratory the metamorphosis may be brought about by various means, such as the reduction of the oxygen in the water. The eggs of animals which had been forced in the laboratory to as- sume a terrestrial mode of life were deposited and hatched under normal conditions. The larvEe were kept under con- ditions where the progeny of untreated parents had never 176 CONCERNING THE ORIGIN OF INSTINCTS been known to desert the water. But at the end of a year these larvae showed a reduction in the size of the gills and, w^hen offered an opportunity to go on land, went through the metamorphosis in a shorter time than had ever been ob- served among normal animals. From this it seems that the induced tendency to metamorphosis was inherited. Yet the experiments w^ere too limited and too lacking in con- trols to do more than indicate the direction which further work should take.
The same may be said of the work of Schroeder upon the willow-leaf beetle. The eggs of this insect are deposited and the larvas fed upon the smooth leaves of a willow (Salix fragilis). Schroeder transferred the eggs and larvae to a downy-leafed willow where the larvEe learned to push away the down with their heads before feeding upon the tissues of the leaves. After metamorphosis the adults were offered a choice between the smooth and the downy willows, and a small percentage deposited their eggs upon the downy leaves. This '' training " of the larvse was repeated for four generations with a gradual increase in the percentage of adults which chose the downy leaves until, in the fourth generation, all of the adults deposited their eggs upon the downy willow, avoiding the natural food plant. The results are open to the criticism that no adults fed on smooth-leafed willows were given a choice of the two plants as a control, and that the plants used differed greatly in size, so that no accurate judgment of the results to be expected from chance can be formed.
The larva of the moth {Gracillaria stigmatella) eats the leaves of a willow, bending over and fastening the tips of the leaves on which it is feeding so as to form a sort of tent in which it lives. Schroeder cut off the tips of the leaves, whereupon the larvEe made their houses by bending over the edges of the mutilated leaves, adapting themselves to the new conditions. The next generation was again brought up upon leaves without tips, and all formed their tents by rolling in the edges of the leaves. The third generation was restored to normal conditions. Nineteen larvag were ob- tained. Fifteen of them rolled up the leaf tips, just as EXPERIMENTS ON SELECTION 177 the wild insects do, but 4 made their tents from the edges of the leaves. The final numbers are too small to be con- clusive, particularly so since no accurate data respecting the occurrence of this habit in nature were obtained; but the experiment is suggestive of promising work in the same field.
The experiments of Kammerer. — The same objections cannot be urged against the experiments of Kammerer, which are very extensive and, save for the remote possibility of faulty technique, seem to establish the inheritance of certain somatic adaptations. The experiments have been carried out in great detail upon several amphibia and rep- tiles and the results are far too extensive to be reviewed here. Certain observations upon the obstetrical toad (Alytes ohstetricans) have, however, a direct bearing upon the origin of instincts and may be summarized briefly. The obstetrical toad is ordinarily exceptional in its mode of reproduction. It has acquired an almost exclusive land habit, going to water only at rare intervals to moisten the skin. Copulation takes place on land and the eggs, which are relatively few and large, are carried by the male, wrapped around his thighs. The embryos mature in their gelatinous capsules while attached to the male and escape only when he goes to the water. By subjecting the toads at the breeding season to high temperature in a terrarium provided with a water basin, the toads were forced to spend much of their time in the water and the eggs were deposited there. When the egg chains were laid in the water their gelatinous envelops would not adhere to the thighs of the male and consequently the eggs were left to develop in the water. After this was repeated during several breeding periods the toads acquired the " habit " of going to the water to lay their eggs, and this habit persisted when they were restored to normal conditions. The eggs laid at the different breeding periods during the formation of the aquatic habit were kept under normal conditions of tempera- ture until they reached maturity and the behavior of this second generation of toads was noted. — First hrood. — The eggs which formed this brood were 178 CONCERNING THE ORIGIN OF INSTINCTS laid when the parents were first driven to the water by heat and before they had gained the aquatic habit. The members of this brood all laid their eggs in the normal manner.
Second brood. — The members of this brood, formed after the parents had begun to acquire the habit of laying in the water, were abnormal to a slight extent. At their first breeding period the eggs were laid upon the land but the males did not take them up. All later broods produced by members of this second brood were carried normally by the males.
Third brood. — This brood, like the second, laid their eggs upon land, but in their first and second broods the eggs were not taken up by the males. In their third brood the eggs were partly taken up, and in their fourth brood the behavior became normal.
Fourth and later broods. — The eggs which formed this brood were laid after the habit of laying in the water had become thoroughly established in the parents. Their first brood was laid in the water, the second upon land, but without being taken up by the males. The third brood was normal.
Some of the adults from these broods were subjected again to the high temperature and forced to breed re- peatedly in the water, and this was repeated for several generations. Many structural changes appeared and in- creased from generation to generation. One of these was the appearance of a horny pad upon the thumb of the males of the fourth generation, a pad which closely re- sembled that found ordinarily upon the thumbs of frogs breeding normally in the water. In the third generation the pad appeared as a slight thickening of the skin, so its acquirement, like that of the instinct to lay the eggs in water, seemed to have been gradual. These experiments in- dicate an actual inheritance of adaptive somatic characters in which the degree of inheritance of the character is pro- portional to the length of time during which the parent Avas subject to the influence of the abnormal environment. A large number of other experiments by Kammerer upon EXPERIMENTS ON SELECTION 179 other characters in salamanders and reptiles have given similar results.
Some evidence negative in character. — Experiments of this type, considered alone, are enough to convince one that acquired characters are inherited, but zoologists are inclined to regard such results with more than usual sus- picion and this attitude is certainly justified by the mass of contradictory evidence. The evidence from selection has already been considered and is entirely negative. Many attempts to get an inheritance of characters forced, as it were, upon the soma have all failed. Thus Bogdanow cut off the wings of the house-fly for ten generations without causing any change in the wings of the progeny, and Galton transfused the blood of differently colored rabbits without affecting the color of their descendants. The more w^eighty evidence comes, however, from transplantation experiments. Harrison was able to graft together parts of the larvae of two species of frogs which later developed into a chimsera with a head of one species and a body of the other, neither part being influenced by its close association with the other. The results of several investigators in transplanting ovaries from one variety of animals to another have been negative. In rabbits, fowls, and in ascidians the trans- plantations were without effect, the eggs were quite uninfluenced by their new surroundings. In only one case is there definite evidence to the contrary. Kammerer has found a marked influence of the soma upon transplanted germ cells of the fire salamander. In this case transplanta- tion of the ovaries to ordinary individuals had no effect, but transplantation to a new race, which Kammerer holds to have been produced by the inheritance of acquired characters, gave positive results.
The direct adaptation theory of evolution. — Although, as yet, the evidence in favor of the inheritance of acquired characters is inconclusive, it is of sufficient importance to make it impossible to disregard entirely the possibility that such inheritance has played a large role in adaptive evolution. Various theories have been advanced to explain the mechanism by which characters might be impressed 180 CONCERNING THE ORIGIN OF INSTINCTS upon the race by the action of environment. Pig. 35 illus- trates the chief two divisions into which these theories fall. First the change may be brought about by the action of the environmental agent upon both the soma and the germ cell, the parallel induction theory, producing in each changes of the same nature. The experiment of Sitowski, although not showing a true inheritance of an acquired character, may serve to illustrate this point. He fed the larvae of a moth with a dye (Sudan red III.) and found that the dye was carried over through the eggs to the next generation of caterpillars. The dye affected both the body and germ cells of the larvge, i.e., stained them, in the same way. However, the mass of evidence seems to be against this view, since it w^as shown earlier (p. 167) that the direct action of the environment upon the germ cell in those cases which have been studied carefully has the effect of causing variations which are not adaptive — not in line with the habits which are acquired during the life of the individual.
The second possibility is that of indirect action of the environment through the soma (Fig. 35). The changes which take place in the body are supposed in some way to react upon the germ cells, inducing a corresponding change in them. Darwin's theory of pangenesis is probably the most widely known of such theories, although it has been completely abandoned. He assumed that every cell in the body was constantly giving off minute replicas of itself which were carried in the blood and finally got into the germ cells, where they provided the material for the repro- duction of similar structures in the next generation. A more recent view to account for the indirect effect of en- vironment upon the germ cells is known as the Jiormone theory.
The hormone theory. — Cunningham, basing his argu- ment chiefly upon the observational and experimental evi- dence on the production of secondary sexual characters, has evolved a chemical theory to explain the inheritance of acquired characters, known as the hormone theory. The connection between gonad (sexual cells) development and the formation of the secondary sex characters is now well SUMMARY 181