Chapter 14 Chapter 8 Instinct 2

These are facts about the strange instinct of keeping slaves, and they need not be confirmed by me.Let us see how the instinctive habits of the blood ants differ from those of the rufous ants of the Continent.The latter species do not build nests, determine their own migrations, gather food for themselves and their young, or even feed themselves: completely dependent on their countless slave ants.The blood ants, on the other hand, have few slaves, and in early summer the slaves are very few, and the master decides when and where a new colony should be made, and when they migrate, the master takes the slaves with him.Slave ants in Switzerland and England seem to specialize in caring for young ants. Masters alone go on expeditions to capture slave ants. Swiss slave ants work with their masters, carrying materials back to build nests; masters and slaves jointly, but mainly slaves, take care of them aphids, and perform what is called milking; thus both masters and slaves collect food for the colony, and in England it is common for the masters to go out alone to forage for table-making material and to forage for themselves, their slaves, and larvae.In England, therefore, slaves perform much less labor for their masters than in Switzerland.

By what steps the blood ants' instinct took place, I do not want to speculate, but, as ants do not keep slaves, as far as I have seen, if there are pupae of other species scattered near their nests, they will These pupae are dragged away, so that these pupae, which were originally stored for food, may develop; the foreign ants thus unconsciously reared will follow their inherent instincts and do what they can.If their existence proves useful to the species that captures them—if it is more beneficial to the species to capture workers than to reproduce them itself—then the habit of collecting pupae for food would probably have been altered by natural selection. strengthened and made permanent for very different purposes of slavery.Once instinct has been acquired, even though it has a far less range of application than the English blood ants (which, as we have seen, are less dependent on the help of slaves than the same species in Switzerland), natural selection will probably increase and Modify this instinct—every variation, we have so often supposed, to be useful to the species—until there is formed a species of ant that lives as vilely on a slave as the russet ant.

The hive instinct of bees I do not intend to discuss this question in detail, but only to give an outline of the conclusions I have reached.He must be a dull man who examines the ingenious construction of a hive, and sees it so beautifully adapted to its purpose, without enthusiastically admiring it.We hear mathematicians say that bees have practically solved esoteric problems by forming their cells into proper shapes to contain the greatest possible capacity of honey, and using the minimum amount of precious wax in their construction.It was once said that a skilled worker, with suitable tools and calculators, can hardly make a wax hive in the true shape, but a swarm of bees can make it in a dark beehive, whatever you say Whatever instincts do, it seems inconceivable at first how they can make all the necessary angles and faces, or even perceive that they are done correctly.But the difficulty is not so great as it at first appears; I think it may be shown that all this beautiful work comes from a few simple instincts.

I have been guided in my research on this subject by Mr. Waterhouse.He showed that the shape of a hive is closely related to the existence of adjoining cells; the following observations can probably only be regarded as a modification of his theory, let us look at the great principle of gradation, and see whether "Nature" reveals to us her working methods.At one end of this short series there are the Scolia, which store honey in their old cocoons, sometimes adding short waxy tubes to the cocoon shells, and also making separate, rather irregular circular waxy cells .At the other end of the series are the cells of the bees, which are arranged in two tiers: each cell, as is known, is a hexahedron, with the bases of the six sides obliquely united in a chamfer of three rhombuses. cone.These rhombuses have certain angles, and on one side of the hive, the three sides of the pyramidal base of one cell just form the bases of the three connected cells on the opposite side.In this series, between the very complete hive of the honey-bee and the simple hive of the scumble-bee, is the hive of the Mexican bee (Melipona domestica), which has been carefully described and drawn by Hubert.The bodily structure of the Mexican bee is intermediate between that of the honey bee and the scumbler, but more closely related to it; it builds an almost regular waxy hive, the cells of which are cylindrical, in which the young are hatched, and besides A number of large waxy cells used to store honey.These large cells are nearly spherical, nearly equal in size, and clustered in irregular piles.The point to be noted here is that such cells are often built so closely together that, if they were perfectly spherical, the wax walls would necessarily intersect or penetrate; The wax walls are made flat between the spherical cells.Therefore, each hive is made of an outer spherical part and two or three, or more planes, depending on whether this hive is connected with two, three or more hives.When a cell joins three other cells, since their spheres are of about the same size, it is often and necessary in this case that the three planes join into a pyramid; The body is very similar to the three-sided pyramidal bottom of the bee hive.Here, as in the honeybee cell, the three planes of any cell necessarily form part of the three connected cells.By this method of construction the Mexican bee obviously saves wax, and, what is more important, labour; for the planar walls connecting the cells are not double, and are of the same thickness as the outer spherical part, yet each planar wall But it constitutes a common part of the two rooms.

Considering the circumstances, it seems to me that if the Mexican bees build their spherical cells at a certain distance from one another, and make them the same size, and arrange them symmetrically in double layers, the structure will be like that of bees. complete.So I wrote to Prof. Miller at Cambridge, and from his reply I wrote the following statement, which the geometer kindly read and told me was quite true. Suppose we draw several spheres of the same size, and their centers are all on two parallel layers; the distance between the center of each sphere and the centers of the six spheres surrounding it in the same layer is equal to or slightly smaller than radius x 2, and The distance from the centers of balls connected in another parallel layer is also as above; therefore, if the joint surfaces of every two balls of this double-layer ball are drawn, a double-layer hexahedron will be formed. The connecting surfaces of the hexahedrons are all connected by the bottom of the pyramid composed of three rhombuses; The angles are exactly equal.But Professor Wyman told me that he had made many careful measurements, and he said that the exactness of the bee's work had been so exaggerated that, whatever the typical shape of the hive, it was seldom, if not impossible, achieved. See you.

We may, therefore, safely conclude that, if we could slightly modify the not very strange pre-existing instincts of the Mexican-bee, it would be able to make a cell as perfect as that of the honey-bee.We must assume that the Mexican-bee is capable of forming cells of truly spherical and equal size; and this will not be surprising, seeing, for example, that she has been able to do this to a certain extent, while, Many insects, too, are capable of forming perfectly cylindrical cavities in trees, apparently revolving about a fixed point.We must suppose that the Mexican-bee arranges her cells in horizontal layers, as her cylindrical cells do.We must further suppose, and this is the most difficult thing, that when several workers build their bulbous cells, she manages to judge correctly how far apart they should be from each other; but she has been able to judge distances, so she can Often the spherical cells are intersected to some degree; then the intersecting points are connected by complete planes.Instincts which are not very strange,--no more exotic than those which direct birds in building their nests--after such modifications, I believe that the bee has acquired, by natural selection, her inimitable building powers.

This theory can be proved by experiments.Following the example of Mr. Tegetmeier, I divided the two hives, and placed between them a long, thick, rectangular wax plate: the bees then proceeded to dig circular depressions in the wax plate. cavities; and as they dug these small cavities at depth, they gradually widened them wide, till at last they became shallow basins roughly the diameter of a honeycomb, and looked exactly like a true sphere, or part of a sphere.The following situation is most interesting: when several bees approach each other and start digging a basin, the distance between them is just such that the basin obtains the above-mentioned width (corresponding approximately to the width of an ordinary hive), and at the When the depth reaches one-sixth of the diameter of the sphere formed by these pot-shaped recesses, the sides of the pot-shaped recesses intersect, or penetrate each other. When this happens, the bee stops digging deeper. and began to make plane walls of wax at the intersections between the sides of the basin, so that each hexahedron was not built upon the straight sides of a triangular pyramid, as in the case of common beehives, but It is built on a smooth basin with scalloped sides.

Then I put into the hive a thin, narrow piece of vermilion-painted wax, with a blade-edged edge, instead of the thick rectangular one previously used.Immediately thereupon the bees began, as before, to excavate small basin-shaped cavities close to each other on both sides of the wax sheet.But the wax is so thin that if the bottom of the basin-shaped cavity is dug as deep as in the above experiment, the two sides will penetrate each other.The bees, however, do not allow this to happen, and stop digging when the time is right; so that the pot-shaped holes, when dug a little deeper, have flat bottoms, which are left over without being bitten. The flat base formed by a thin slice of vermilion wax that has been removed is, as far as the eye can judge, exactly at the imaginary intersection between the basin-shaped cavities on the reverse side of the wax slice.The lozenge-shaped plates, of various sizes, remaining between the basin-shaped cavities on the reverse side, could not have done the work delicately, as the wax flakes were not a thing in its natural state.Even so, the bee can bite off the wax roundly on both sides of the vermilion wax sheet, and deepen the basin shape. Its working speed must be about the same, this is to be able to successfully stop working at the intersection, and Leave flat surfaces between the basin-shaped pockets.

After considering how soft the thin wax sheets are, I thought that when the bees worked on both sides of the wax sheet, they would have no difficulty in perceiving when they had bitten at the proper thinness, and ceased their work.In common hives, I do not think that the bees always succeed in working at exactly equal speed on both sides; for I have noticed a half-finished diamond-shaped panel on the bottom of One side is slightly concave, which I imagine because the bees dug too quickly on that side, and the other side is convex, because the bees work slower on this side, in one notable case. , I put the comb back into the hive, and let the bees go on working for a short time, and then inspecting the comb, I found that the rhombus was finished, and had become perfectly flat: the wax was extremely thin, So it is absolutely impossible to bite off the wax from the convex side to make the above appearance; I guess that this situation is probably because the bee standing on the opposite side pushes the plastic and warm wax just to its middle plate. bend it (I experimented, it's easy to do) and this will flatten it.

From the experiment with the vermilion wax we see that when the bees have to build themselves a thin wall of wax, they stand at a certain distance from each other, dig down at the same speed, and try to make it of the same size. spherical cavities, but never allowing these cavities to pass through each other, so that they form cells of the proper shape.If one examines the edges of the hive being built, it is evident that the bees first make a rough wall or rim about the comb; Bite off the wall from both sides, they do not at one time make the whole base of any one cell's triangular pyramid, usually first, a rhomboid plate at the extreme edge of what is being built, or first Two rhombus panels, as the case may be; and they never complete the upper sides of the rhombus panels without building the six walls.Some parts of these accounts differ from those of the eminent old Hubert, but I believe them to be correct; and if I have space I will show that this is consistent with my doctrine.

Hubert says that at first the first cells were hewn out of small parallel walls of wax, a statement which, so far as I have seen, is not strictly true; what is often started is a small wax pocket. ; but I do not intend to discuss it in detail here.We know what an important part gouging plays in the construction of the hive; but if the bee is supposed to be unable to make a rough wall of wax at the proper place, that is, along the plane of intersection between , could be a huge mistake.I have several specimens which clearly show that they are capable of doing so.Even in the rough edges, or wax walls, which surround the hive under construction, a curvature is sometimes observed which corresponds to that of the rhombus-shaped base of the future cell.But in all cases, rough wax walls are accomplished by biting off most of the wax on both sides.This method of construction of the bees is marvelous; they always make the first rough walls ten or thirty times thicker than the very thin walls of the final remaining cell. We'll understand how they work in the following scenario: Suppose the builder begins to pile up a broad foundation wall with cement, and then begins to chip away the cement equally on both sides near the ground until the central part forms a wide foundation wall. A smooth, thin wall; these builders often pile the chipped cement on top of the wall, and then add some new cement.Therefore, the thin wall keeps rising like this, but there is often a thick top cover on it.All cells, both new and completed, have such a solid wax cover over them, that the bees can gather and crawl about the comb without damaging the thin six walls.These walls have been kindly measured for me by Professor Miller, and these walls vary widely in thickness; a dozen measurements, near the edge of the hive, show an average thickness of 1/352 of an inch; It is three to two, with an average thickness of 1/229th of an inch, according to twenty-one measurements.By such a peculiar method of construction as has been described, it is possible to use wax extremely economically, while at the same time constantly strengthening the hive. As many bees work together, this seems at first to increase the difficulty of understanding how the hive is made; a bee, after working for a short time in one hive, goes to another, so, as What Hubert says, that twenty bees were at work even when the first hive was started, I may illustrate this fact practically by the following situation: On the sides of the six walls of a beehive, or on the extreme edges of the enclosure of an enlarged beehive, it must be seen that the bees have distributed the color very finely,--as finely as a painter with a brush-- —The colored wax is taken little by little from the place of application, and deposited on the enlarged edges of the surrounding cells.There seems to be an equal distribution of this work of building among many bees, all instinctively standing at the same proportional distance from each other, all trying to carve equal spheres, and building or remaining Do not bite the intersecting planes between these spheres.They sometimes encounter difficulties, and such instances are curious indeed, for example, when two hives meet at a corner, the bees so often tear down the completed comb, and rebuild it in a different way; The shape of the honeycomb made is often the same as that of the removed one. When a bee finds a place where it can stand and do its work—for example, on a piece of wood just under the central part of a hive built downwards—then the bee The hive must be built on top of this piece of wood--in which case the bees will build the base of a new hexahedral wall, projecting beyond the rest of the completed cell, and placing it on the perfectly appropriate location.So long as the bees are able to stand at a suitable distance from each other and from the walls of the finally completed hive, they are then sufficient to make a wall between two adjoining spheres by digging imaginary spheres. The middle walls are waxy; but, so far as I have seen, they never bite off and smooth the corners of the cell, till that cell, and the adjoining cells, are largely formed.The ability of the bee, under given circumstances, to establish in place a rough wall between two incipient cells is important; for it is connected with the fact that at first it seems to be Rejecting the above theory; it is a fact that some of the cells on the outermost edges of the wasps are also frequently strictly hexagonal; but I have no space here to discuss this.I do not think that a single insect, such as the queen of a wasp, would have any great difficulty in forming a hexagonal cell;--if she could alternately form the inner and outer sides of two or three nests which had started at the same time. This can be done by working, always at a suitable distance from the parts of the cell which have just begun, by excavating spherical or cylindrical shapes, and building up the flat walls in between. Natural selection can operate only by accumulating slight variations of structure or instincts, each of which is advantageous to the individual under his conditions of life.It may therefore be reasonably asked: How did the long and progressive succession of stages through which all the modified building instincts have passed, tending towards their present state of perfection, have worked in favor of their ancestors?It is not difficult, I think, to answer the question: A hive constructed like that of a bee or a wasp is strong, and saves a great deal of labor, space, and materials of construction of the hive.In order to make wax, we know that a sufficient amount of nectar must be collected, and in this respect bees are often very laborious. Mr. Tegetmeyer told me that it has been proved by experiments that bees consume twelve to fifteen pounds of wax to secrete one pound of wax. pounds of dry sugar; so the bees in a hive must collect and consume large quantities of liquid nectar in order to secrete the wax necessary for building the hive.Also, many bees are bound to be out of work for many days during the secretion process.The storage of large quantities of honey is indispensable for the winter life of large colonies; and we know that the safety of a bee colony is largely determined by the maintenance of large numbers of bees.The saving of wax, therefore, results in a considerable saving of honey, and a saving of time in collecting honey, which must be an important factor in the success of any bee colony.Of course the success of a species may also be determined by the number of its predators or parasites, or by other very special causes, which have nothing to do with the quantity of honey which the bees can collect.But let us suppose that the ability to collect honey determines, and probably has often determined, whether a species of bee similar to the English Scolia can be found in any one place in great numbers; and let us further suppose that the colony must For the winter it would result in the necessity of hoarding honey; and in this case a slight variation of her instinct leading her to make the wax-houses a little closer together, and slightly tangent to each other, would doubtless be advantageous to the kind we imagine. of the Scolia; for a common wall saves little labor and wax even if it joins only two cells.It would, therefore, be continually more and more advantageous to the Scolia if their cells were made more and more neatly, more closely together, and clustered together like those of the Mexican-bee; for in this case most of the walls of the individual cells It will be used as the wall of the adjoining hive, so that labor and wax can be saved considerably.And, for the same reason, it would be to her advantage if the Mexican-bee could make her hives closer together than they are at present, and more regular in every respect; for, as we have seen, The spherical surface of the cell will disappear altogether, and a flat surface will be substituted; and the comb of the Mexican-bee will probably be as perfect as the hive of the bee.Beyond this stage of perfection in construction, natural selection can no longer work; for the hive of bees is, so far as we know, absolutely perfect in the economical use of labor and wax. Thus, as I believe, this strangest of all known instincts, that of the bee, can be explained by the fact that natural selection has exploited countless, successively small fractions of simpler instincts. Variation; natural selection has slowly and more and more perfectly led the bees to excavate spheres of equal size at a distance from each other on the double layer, and to build and excavate walls of wax along the intersection planes; Knowing themselves to dig spherical bodies at a distance from each other, just as they would not know the angles of the hexahedrons, and the angles of the rhomboid plates at the bottom; and proper volume and shape to accommodate the larvae, and to complete it with the greatest possible economical use of labor and wax; the hives, they will be most successful, and will pass on this newly acquired instinct of economy to new colonies, which in their own generation will have the greatest chance of success in the struggle for existence. Objections to the Application of the Theory of Natural Selection to Instincts: Neutral and Sterile Insects The above view of the origin of instincts has been objected to, saying, "Constructive and instinctive modifications must have occurred simultaneously and in close harmony with each other, since a modification in the one without a corresponding change in the other A mutation would be lethal."The force of this objection rests entirely on the supposition of sudden changes of instinct and structure.An example may be taken of the great wee-finch (Parus maior), of which we have spoken in the preceding chapter; this bird will seize the yew-seeds on a branch with its feet, and peck them with its beak until it extracts the kernels.Thus natural selection has preserved all the slight variations of a beak more and more adapted to pecking the seed, until a beak like that of the nuthatch, which is well adapted for this purpose, has been formed, and, at the same time, Is there any particular difficulty in explaining that a spontaneous variation of habit or compulsion or inclination has also led the bird to become more and more a seed-eater?In this instance, imagine a slow change in habit or habit, and then, by natural selection, the beak slowly change in correspondence with the change in taste or habit.But suppose the finch's foot, in connection with the beak, or from any other unknown cause, has been modified and enlarged, it is not impossible that this enlarged foot would have led the bird to become more and more able to climb, And at last it acquired the remarkable climbing instinct and strength of the nuthatch.In this case it is supposed that a gradual change of structure brings about a change in the habits of the instincts.To take another example: Few instincts are more singular than that of the Swift of the Eastern Islands, who builds her nest entirely from concentrated saliva.Certain birds make their nests of earth, in which it is believed to be mixed with saliva; and there is a species of swift in North America which (as I have seen) spittles its nests with twigs, and even with scraps of such things. to build a nest.Is it highly improbable, then, that natural selection on increasingly salivating individual swifts should at last produce a species with an instinct to neglect other materials and to devote exclusively concentrated saliva to the nesting? ?The same is true for other situations.It must be admitted, however, that in many cases we cannot conjecture whether it was instinct or structure which first varied. No doubt many instincts of the utmost inexplicability can also be argued against the theory of natural selection—some instincts, for example, of whose origin we do not know; others of which we do not know an intermediate gradation; so that natural selection has little effect on it; and some instincts are so nearly alike in animals whose natural systems are so far apart that we cannot account for their similarity by inheritance from a common ancestor, and are obliged to believe that these instincts have been acquired by natural selection. obtained independently.I am not going to discuss these few examples here, but I shall devote myself to a particular difficulty, which at first seemed to me inexplicable, and which was in fact fatal to my whole theory.I refer to the neuter, or sterile, females in insect societies; They cannot reproduce their species. This subject is well worthy of a detailed discussion, but I shall here conceive only one example, that of the sterile worker ants.How the workers become sterile individuals is a difficulty; but no more difficult to explain than any other marked variation in structure; for it may be shown that certain insects, and other arthropods, occasionally become sterile in a state of nature. and if the insects are social, and if it is to the advantage of the group to produce annually a number of working, but sterile individuals, I do not think it difficult to understand that this is due to the operation of natural selection.But I must omit this preliminary difficulty.The greatest difficulty lies in the huge difference in structure between worker ants and male ants and fertile female ants.For example, worker ants have chests of different shapes, lack wings, sometimes have no eyes, and have different instincts.As far as instinct alone is concerned, honeybees are an excellent proof of the astonishing difference between workers and complete females, and if worker ants or other neuter protozoa were a normal animal, I would not hesitate to Suppose that all its characters have been slowly acquired by natural selection; that is to say, by virtue of the individuals being born with slight favorable variations, these variations have been transmitted to their offspring; Variation, selection, and so on.But worker ants are very different from their parents and are absolutely sterile, so they can never pass on the structural or instinctual variation acquired in successive generations to their offspring.It may then be asked: How can this case be consistent with the theory of natural selection? First, Let us remember that we have innumerable instances of various differences of inherited structure, in domestic and those in a state of nature, being associated with a definite age or sex.These differences relate not only to one sex, but to the brief period of activity of the reproductive system, as is the case, for example, with the courtship plumage of many species of male birds, and with the hooked jaws of male salmon.There are even slight differences in the horns of different breeds of bulls that have been artificially castrated, since castrated bulls of some breeds have longer horn lengths than males and females of the same breed than castrated bulls of other breeds. bulls, with longer horns, so I do not think there is much difficulty in any character in some members of insect societies becoming correlated with their sterility: the difficulty lies in understanding how such structurally correlated variations slowly accumulated by the action of natural selection. Although this difficulty appears to be insurmountable on the surface, as long as it is remembered that selection can be applied to an individual or to a whole family, and the desired result can be obtained from this, then this difficulty will be reduced, or as I have suggested. Believe it, it will be eliminated.Cattlemen liked the way the meat and fat were marbling; cattle with this quality were slaughtered.But the cattle breeder's confidence in continuing to breed the same cattle, and with success, is based on the power of selection if we pay careful attention to which bulls and cows are mated to produce the castration of the longest horns Bulls, presumably get a breed of castrated bulls that often produce unusually long horns, although no castrated bull has ever been bred of its kind.Here is a better and more exact illustration: According to M. Verlot, certain varieties of the double-petaled annual violet (Stock), by long and careful selection to a proper degree, will Numerous seedlings are often produced, bearing double, completely sterile flowers, but they also produce several single, fertile plants.Only such single-petaled plants reproduce the variety, which can be compared with fertile males and females, and sterile double-petaled plants with the neuters of the same colony.Whether with these varieties of violets, or with social insects, selection, in order to achieve advantage, acts not on individuals, but on whole families.We may therefore assert that slight variations in structure or instincts, connected with the sterile state of certain members of the herd, prove to be advantageous: the resultant fertile males and females are multiplied, and this The tendency - to produce sterile members with the same variation - is passed on to fertile offspring.This process must have been repeated many times, until a great amount of difference arose between fertile and sterile females of the same species, as we see in many species of social insects. But we have not yet reached the height of the difficulty; namely, that the neutrals of several ants differ not only from the fertile females and males, but also from each other, sometimes almost unbelievably different. degrees, and are thus divided into two or even three castes.Again, these degrees, generally not gradually deviating from each other, are quite distinct; as distinct from each other as any two species of a genus, or any two genera of the same family, e.g., the middle of the Eciton ant. Sexual workers and soldiers have unusually different jaws and instincts; Cryptocerus has only one class of workers, and they have a strange shield on their heads, the purpose of which is completely unknown; the Mexican honey ants (Myrmecocystus) has a class of worker ants, which never leave their nests, have a large abdomen, and secrete a honeydew, in place of what is excreted by aphids, or cows which may be called ants, European Ants often imprison and guard them. If I did not admit that this strange but quite certain fact would at once overthrow the doctrine, it must be thought that I believed too vainly in the principle of natural selection.If the neuter had only one order, I believe that the differences between it and the fertile males and females had been obtained by natural selection. In this simpler case, by analogy from normal variation, we may assert that , this successive, slight, favorable variation, at first occurred not in all the neuters in the same collection, but only in some few; Produce an enormous number of neuters with favorable mutations--to survive, all neuters will eventually have that characteristic.On this view, we should occasionally find in the same collection neuters exhibiting gradations of structure; and indeed we do, and since neuters outside Europe have rarely been carefully examined, this may not even be the case. Do not care.Mr. Smith has shown that the neutrals of several species of the British ant show striking differences from one another in size, and sometimes in colour; and that the extreme forms may be connected by individuals of the same concentration: I I myself have compared the perfect gradation of this species, and have sometimes found that either the large or the small workers are the greatest in number;Yellow ants have large and small workers, and intermediate workers are rare; and, as Mr. Smith observes, in this species the large workers have ocelli, though small.But it can still be clearly distinguished, while the monocular eyes of the small worker ants are remnant.仔细地解剖了几只这等工蚁之后，我能确定小形的工蚁的眼睛，比我们能够用它们的大小比例来解释的，还要远远地不发育；并且我充分相信，虽然我不敢很肯定地断言，中间形工蚁的单眼恰恰处在中间的状态。所以，一个集内的两群不育的工蚁，不但在大小上，并且在视觉器官上，都表现了差异，然而它们是被某些少数中间状态的成员连接起来的。我再补充几句题外的话，如果小形的工蚁对于蚁群最有利，则产生愈来愈多的小形工蚁的雄蚁和雌蚁必将不断地被选择，直到所有的工蚁都具有那种形态为止。于是就形成了这样一个蚁种，它们的中性虫差不多就像褐蚁属（Myrmica）的工蚁那样。褐蚁属的工蚁甚至连残迹的单眼都没有，虽然这个属的雄蚁和雌蚁都生有很发达的单眼。 我再举一例：在同一物种的不同级的中性虫之间，我非常有信心地期望可以偶尔找到重要构造的中间诸级，所以我很高兴利用史密斯先生所提供的取自西非洲驱逐蚁（Anonma）的同窠中的许多标本。我不举实际的测量数字，而只做一个严格精确的说明，我想读者大概就能最好地了解这等工蚁之间的差异量；这差异就好像以下的情形：我们看到一群建筑房屋的工人，其中有许多是五英尺四英寸高，还有许多是十六英尺高；但我们必须再假定那大个儿工人的头比小个儿工人的头不止大三倍，却要大四倍，而颚则差不多要大五倍。再者，几种大小不同的工蚁的颚不仅在形状上有可惊的差异，而且牙齿的形状和数目也相差悬殊。但对于我们重要的事实却是，虽然工蚁可以依大小分为不同的数级，然而它们却缓慢地彼此逐渐推移，例如，它们的构造大不相同的颚就是这样。关于后面一点，我确信就是如此，因为卢伯克爵士曾用描图器把我所解剖的几种大小不同的工蚁的颚逐一作图。贝茨先生（Mr. Bates）在他的有趣的著作《亚马逊河上的博物学者》（Naturalist ontlie Amazons）里也曾描述过一些类似的情形。 根据摆在我面前的这些事实，我相信自然选择，由于作用于能育的蚁，即它的双亲，便可以形成一个物种，专门产生体形大而具有某一形状的颚的中性虫，或者专门产生体形小而大不相同的颚的中性虫；最后，这是一个最大的难点，具有某一种大小和构造的一群工蚁和具有不同大小和构造的另一群工蚁，是同时存在的；——但最先形成的是一个级进的系列，就像驱逐蚁的情形那样，然后，由于生育它们的双亲得到生存，这系列上的两极端类型就被产生的愈来愈多，终至具有中间构造的个体不再产生。 华莱斯和米勒两位先生曾对同样复杂的例子提出了类似的解释，华莱斯的例子是，某种马来产的蝴蝶的雌体规则地表现了两种或三种不同的形态；米勒的例子是，某种巴西的甲壳类的雄体同样地也表现了两种大不相同的形态。但在这里无需讨论这个问题。现在我已解释了，如我所相信的：在同一窠里生存的、区别分明的工蚁两级——它们不但彼此之间大不相同，并且和双亲之间也大不相同——的奇异事实，是怎样发生的。我们可以看出，分工对于文明人是有用处的，依据同样的原理，工蚁的生成，对于蚁的社会也有很大用处。不过蚁是用遗传的本能和遗传的器官即工具来工作的，人类则用学得的知识和人造的器具来做工的。但是我必须坦白承认，我虽然完全相信自然选择，若不是有这等中性虫引导我达到这种结论，我决不会料到这一原理是如此高度地有效。所以，为了阐明自然选择的力量，并且因为这是我的学说所遭到的特别严重的难点，我对于这种情形作了稍多的、但全然不够的讨论。这种情形也是很有趣的，因为它证明在动物里，如同在植物里一样，由于把无数的、微小的、自发变异一一只要是稍微有利的一一累积下来，纵使没有锻炼或习性参加作用，任何量的变异都能产生效果。因为，工蚁即不育的雌蚁所独有的特别习性，纵使行之已久，也不可能影响专事遗留后代的雄体和能育的雌体。我觉得奇怪的是，为什么至今没有人用这种中性虫的明显例子去反对众所熟知的拉马克所提出的“习性遗传”的学说。 feed 我已勉力在这一章里简要地指出了家养动物的精神能力是变异的，而且这等变异是遗传的。我又试图更为简要地阐明本能在自然状态下也是轻微地变异着的。没有人会否定本能对于各种动物都具有最高度的重要性。所以，在改变了的生活条件下，自然选择把任何稍微有用的本能上的微小变异，累积到任何程度，其中并不存在什么真正的难点。在许多情形下，习性或者使用和不使用大概也参加作用。我不敢说本章里所举出的事实能够把我的学说加强到很大的程度；但是根据我所能判断的，没有一个难解的例子可以颠覆我的学说。相反地，本能并不经常是绝对完全的，而且是易致错误的——虽然有些动物可以利用其他一些动物的本能，但没有一种本能可说是为了其他动物的利益而被产生的——自然史上的一句格言“自然界里没有飞跃”，就像应用于身体构造那样地也能应用于本能，并且可用上述观点来清楚地解释它，如果不是这样，它就是不能解释的了，——所有这些事实都巩固了自然选择的学说。 这个学说也因其他几种关于本能的事实而被加强；例如，密切近似的但不相同的物种，当栖息在世界上的远隔的地方，并且生活在相当不同的生活条件之下时，常常保持了几乎同样的本能。例如，根据遗传的原理，我们能够理解，为什么热带南美洲的工鸫英国的特别造巢方法那样地用泥来涂抹它们的巢；为什么非洲和印度的犀鸟（hornbill）有同样异常的本能，用泥把树洞封住，把雌鸟关闭在里面，在封口处只留一个小孔，以便雄鸟从这里哺喂雌鸟和孵出的幼鸟；为什么北美洲的雄性鹤鹩（Troglodytes）像英国的雄性猫形鹤鹩（Kitty- wrens）那样地营造“雄鸟之巢”，以便在那里栖息，——这种习性完全不像任何其他已知鸟类的习性。最后，这可能是不合逻辑的演绎，但据我想像，这样说法最能令人满意，即：把本能，如一只小杜鹃把义兄弟逐出巢外，——蚁养奴隶，——姬蜂科（ichneumonidx）幼虫寄生在活的青虫体内，不看作是被特别赋予的或被特别创造的，而把它看作是引导一切生物进化一一即，繁生、变异、让最强者生存、最弱者死亡——的一般法则的小小结果。