Chapter 9 Chapter 6 Difficulties of Doctrine 1

Difficulties in the theory of Descent with Variation—Non-existence or rarity of transitional varieties—Transitions in habits of life—Divergent habits in the same species—Species with habits very different from allied species—Extremely perfect organs— - Mode of transition - Examples of difficulties - No leaps in nature - Organs of minor importance - Organs are not in all cases absolutely perfect - Laws of the unity of patterns and conditions of existence which are included in the theory of natural selection. Before reading this part of the book, the reader presumably has encountered many difficulties.Some of the difficulties were so serious that I cannot recall them today without hesitation; but, so far as I can judge, most of them were only apparent, and the real ones, I think, were of little use to this subject. Doctrine is not fatal either.

These difficulties and objections can be grouped into the following categories: First, If species evolve little by little from other species, why do we not see innumerable transitional forms everywhere?Why is it that the species are just as distinct as we see them, and the whole of nature is not in chaos? Secondly, Can an animal, say, an animal of structure and habits like that of a bat, be derived from another animal whose habits and structure are quite different?Can we believe that natural selection has produced, on the one hand, a very insignificant organ, such as the tail of the giraffe, which serves only for flies, and, on the other hand, such a wonderful organ as the eye?

Third, can instinct be acquired by natural selection?Can natural selection change it?How shall we account for the fact that the instinct which guides the bee to build a hive actually precedes the discovery of learned mathematicians? Fourthly, How can we account for the sterility of species when crossed, and the sterility of their offspring, and for the unimpaired fertility of varieties when crossed? The first two items will be discussed here; other objections of various kinds in the next chapter; instinct and "bybridism" in the next two chapters. On the non-existence or rarity of transitional varieties

As natural selection operates only to preserve favorable variations, every new form in a region teeming with organisms will have a tendency to replace and finally eliminate the parent-form which is less improved than itself, and which competes with it more profitably. less type.So extinction and natural selection go hand in hand.If, therefore, we regard each species as descended from some unknown form, its parent-species and all transitional varieties have generally been destroyed during the formation and perfection of the new form. But, according to this theory, innumerable transitional forms must have existed at one time, why do we not see them buried in great numbers in the earth's crust?It will be more convenient to discuss this question in the chapter "On the Incompleteness of the Geological Record"; I will only state here that I believe that the answer to this question lies mainly in the fact that the incompleteness of the geological record is not generally imagined. arrived.The earth's crust is a vast museum; but nature's collections are incomplete, and made at long intervals.It may be maintained, however, that we should indeed see many transitional forms at the present day, when several closely allied species inhabit the same country.To take a simple example: when traveling from north to south on a continent, we generally see at various points closely allied or representative species apparently occupying nearly the same place in the natural composition.These representative species frequently met and intermingled; and as one species diminished, the other gradually increased, until at last the one replaced the other.But if we compare these species where they intermingle, we see that every detail of their structure is generally absolutely different, as are specimens taken from the central habitat of each species.These allied species are, on my theory, descended from a common parent-species; and in the process of modification, each species has adapted itself to the conditions of life in its own region, and has repelled and exterminated the original parent-form and Everything connects past and present transitional variants.We should not, therefore, expect to find innumerable transitional varieties everywhere to-day, though they must have existed there at one time, and may have been buried there in a fossil state.But in the middle ground with intermediate living conditions, why don't we see closely linked intermediate variants now?This difficulty puzzled me for a long time, but I think it is largely explainable.

First, if we see that a place is continuous now, we should be very cautious in inferring that it has been continuous for a long period of time.Geology leads us to believe that most continents, even at the end of the Tertiary period, were divided into islands; on such islands there was no possibility of intervening varieties, and the different species probably arose separately.Owing to the changes in the shape of the land and in the climate, the now continuous surface of the sea must have been, in recent times, far less continuous and uniform than it is at present.But I shall avoid the difficulty by not taking this course; for I believe that many well-defined species were formed on otherwise strictly continuous ground; though I do not doubt that the former separation of the present continuous , especially for the formation of new species of freely interbreeding and roaming animals.

When we observe the species now distributed over a large area, we generally see that they are quite numerous in a large area, but gradually become more or less rare at the borders, and finally disappear.The intermediate zone between two representative species, therefore, will generally always be narrower than the zone exclusive to each species.We can see the same fact in climbing mountains, and sometimes, as de Candol observed, the very sudden disappearance of an ordinary alpine plant is quite remarkable.Forbes noticed the same fact when he scouted the deep sea with his dredge.The fact that some regard climate and physical conditions of life as the most important factors of distribution should surprise those people, for both climate and height or depth are gradually and imperceptibly changed.But if we remember that almost every species, even in the center of its range, would increase in number to innumerable individuals if there were no competing species; if we remember that almost every species either eats other species or are eaten by other species; in a word, if we remember that every living being is directly or indirectly related to every other in a most important manner—then we shall know that the extent of the Determined entirely by physical conditions which change imperceptibly, but largely by the existence of other species, or being dependent on other species, or being destroyed by other species, or competing with other species; It is a distinct entity, not confused by imperceptible types of various levels, so that the range of any one species will have a very obvious tendency to limit because it depends on the range of other species.Also, each species, at the fringes of its range in which few individuals survive, will most easily be completely wiped out by its predators, or by fluctuations in the number of its prey, or by seasonal changes. ; thus, the limits of its geographical range become more pronounced.For allied or representative species, when living in a continuous area, each species has a wide range, with a narrower intermediate zone between them, in which they tend to grow more abruptly. and since varieties and species are not essentially different, the same laws probably apply to both; if we take a varying species inhabiting a large area, there must be two varieties Adapted to two large areas, and there is a third variant adapted to the narrow middle ground.As a result, the intermediate variety is less numerous, owing to its inhabiting a narrow area; indeed, as far as I can understand, this rule holds true for varieties in a state of nature. (Balanus) I see striking instances of this rule.I have been given data by Mr. Watson, Dr. Asa Gray, and Mr. Wollaston, to show that, when intermediate varieties exist between two forms, they generally outnumber the two forms to which they connect. The number of types is much smaller.Now, if we can trust these facts and inferences, and conclude that varieties intervening between two varieties generally have fewer individuals than the forms to which they connect, we can understand why intermediate varieties do not last very long. Persistence during the period:--according to the general law, why do intermediate varieties become extinct and disappear earlier than those forms with which they were originally connected.

That is because, as has been said, any form with a small number of individuals has a greater chance of extinction than a more numerous form; However, there is an even more important reason: in the course of further variation into two different species, assuming a change in the two varieties, the two varieties with a greater number of individuals, because they inhabited a larger area, within, has a strong advantage over those intermediate varieties which inhabit a narrow intermediate zone in which there are fewer individuals.This is because a more numerous form has a better chance at any given period of presenting more favorable variations to be exploited by natural selection than a less numerous form.The more common forms, therefore, tend, in the contest of life, to overwhelm and replace the less common, for the latter are slower to change and improve.This same principle, I believe, accounts, as indicated in the second chapter, for why the common species in each country present, on average, a greater number of well-marked varieties than the rarer ones.I may illustrate what I mean by an example. Suppose that three breeds of sheep are kept, one adapted to a large mountainous region; Both have the same determination and skill in improving their breeds by selection; and in this case the hill or plain breeder with many sheep will have a better chance of success than the narrow middle hill country with a few sheep. The breeders are quicker in improving the breed; consequently, the improved mountain or plain breed will soon replace the less improved hill breed; but without the replaced hilly intermediate variant sandwiched in between.

In a word, we believe that species are after all fairly well-defined entities, not at any one time indissolubly confused by innumerable intermediate links of variation: first, because the formation of new varieties is very slow, due to Variation is a slow process, and natural selection does nothing if no favorable individual differences or variations take place; and if there are no vacant places in the natural institutions of the region which can be better occupied by one or more modified organisms, nature Selection also does nothing.Such new positions are determined by slow changes of climate, or by the occasional introduction of new organisms, and, what is more, may be determined by the slow modification of some old organisms; action and reaction.So at any one place, at any one time, we should see only a few species showing more or less steady slight variations in structure; and this is indeed what we see.Secondly, that now continuous territories must have often in recent times been isolated parts, where many forms, especially of those which required mating at each birth and roamed widely, had probably become quite separate. different enough to be listed as representative species.In this case intermediate varieties between some of the representative species and their common progenitor must have previously existed in isolated parts of the country, but these links have been excluded and become extinct in the course of natural selection. , so you can't see their existence now.

Thirdly, if two or more varieties have been formed in different parts of a closely continuous area, there will probably be intermediate varieties formed in the intermediate zone, but these intermediate varieties generally do not last long because these intermediate varieties Varieties, for the reasons already stated (i.e., due to the actual distribution, as we know it, of closely allied or representative species, and of the accepted varieties), there are more individuals living in intermediate zones than are connected by them. The number of individuals of the variant is smaller.From this cause alone, intermediate varieties will inevitably become extinct; in the course of further variation by natural selection, they will almost certainly be overwhelmed and replaced by the forms with which they are connected; for these forms are more numerous in number. , there are more variations in the whole, so that it can be further improved by natural selection, and further have a greater advantage.

Finally, not through any one period, but through all, if my theory be true, innumerable intermediate varieties must have existed, closely linking all the species of the same group, but, as has been repeatedly said before, The process of natural selection, however, has always tended to exterminate the parent forms and intermediate varieties.In consequence, evidence of their existence is to be found only in fossil remains, and the preservation of these fossils, as we shall show in a later chapter, is very incomplete and intermittent. On the Origin and Transition of Organic Beings with Specific Habits and Structure

Opponents of my opinion have asked: How, say, a terrestrial carnivore can be transformed into a carnivore of aquatic habits; how can the animal live in its transitional state?It is not difficult to show that many of the carnivorous animals of the present day present closely connected intermediate gradations from strictly terrestrial to aquatic habits; and as each animal must struggle for life to survive, it is evident that each must Well adapted to its place in nature.Consider the North American mink (Mustela vison), with its webbed feet, its fur, its short legs, and its tail shaped like an otter.During the summer the animal swims in the water to hunt fish, but during the long winter it leaves the frozen water and, like other pole-cats, preys on mice and other terrestrial animals .To ask another example: How can an insectivorous quadruped be transformed into a flying bat?Answering this question is much more difficult.However, as far as I think, the importance of this difficulty is not great. Here, as on other occasions, I am at a grave disadvantage, for, from the many obvious instances which I have collected, I can give but one or two, to illustrate the transitional habits and structure of allied species; permanent or temporary habits.It seems to me that in a special case such as that of bats, a long list of transitional state instances would not be sufficient to reduce the difficulties. Let us look at the Sciuridae; there are species with only slightly flattened tails, and others, as described by Sir. J. Richardson, with a rather broad It is quite full, starting from these species, until the so-called flying squirrel, there are very finely divided levels in the middle; the limbs and even the base of the tail of the flying squirrel are connected together by a broad membrane, which acts like a parachute, allowing the flying squirrel Gliding through the air from tree to tree is an astonishing distance.We cannot doubt that each structure is useful to each species of squirrel in its own country, whether it enables it to escape birds or beasts of prey, to gather food more quickly, or, if we have reason to Believe it, can make them reduce the risk of accidental falls.From this fact, however, it does not follow that the structure of every squirrel is, under all possible conditions, the best that we can possibly conceive.Had the climate and vegetation changed, had other competing rodents or new predators migrated in, or old predators had mutated, and so on, we would be led to believe that at least some squirrels would be reduced in number, or become extinct, unless their structure can be modified and improved in a corresponding manner, so, especially under changing living conditions, those individuals whose rib membranes become larger and larger will continue to be preserved. , I don't think there is any difficulty. Every variation of it is useful and will be passed down. Due to the cumulative effect of this natural selection process, a complete so-called flying squirrel will eventually be produced. Look now at Galeopithecus, the so-called flying lemur, which was formerly placed among the bats, and is now believed to belong to the insectivora (Insectivora).The very broad skin of its side extends from the forehead to the tail, including the limbs with long fingers, and the skin of the side also has extensor muscles.Although there are as yet no air-gliding structural linkages linking the cat-monkeys with other insectivorous species, it is not difficult to imagine that such linkages existed previously, each like a flying squirrel with less perfect glides. developed; and every level of structure was at one time useful to its owner.Nor do I see any insurmountable difficulty in further believing that the membrane connecting the digits to the forearm of the cat-monkey has been greatly enlarged by natural selection; bat.In some bats the patagium extends from the tip of the shoulders to the tail, and includes the hind legs, where we may perhaps see traces of a structure originally adapted for gliding rather than flying. If a dozen or so genera of birds were extinct, no one would venture to speculate that some of the birds which only use their wings for striking water, such as the duck (Micopterus of Eyton); use their wings in water as fins. , some birds that use their front feet on land, such as the penguin; some birds that use their wings for a canopy, such as the ostrich; existed?However, the structure of each of the above-mentioned birds is useful under the conditions of its life, because every bird is bound to struggle for survival; but it is not necessarily the best under all possible conditions. Well, do not infer from these words, that the various degrees of wing construction here mentioned, which are presumably the result of disuse, represent the steps by which the bird actually acquires the full power of flight; but they suffice to represent the How many transition ways is at least possible. Sees such few species of water-breathing animals as Crustacea and Mollusca adapted to life on land; also sees flying birds, flying beasts, flying insects of many forms, and flying reptiles that once existed , then it is conceivable that flying fishes, which rise a little, spin, and glide for great distances in the air by the beating of their fins, may perhaps be transformed into fully winged animals.If such a thing had happened, who would have imagined that they were the inhabitants of the oceans in an earlier transitional state?Moreover, their rudimentary flight organs are specially used to escape the swallowing of other fishes? (As far as we know, it does.) If we see structures adapted to any particular habit to a high degree of perfection, such as the wings of a bird for flight, we must remember that very few animals exhibiting structures of earlier transitional stages have survived to the present day, because they would have been replaced by subsequent but these successors have gradually become more and more perfect by natural selection.We may further affirm that transitional states between structures adapted to different habits of life are seldom developed at an early stage in great numbers, nor have many subordinate forms.Returning thus to the example of the hypothetical flying fish, the real flying fish probably did not develop in many subordinate forms for the many ways of catching many kinds of food both on land and in water, down to their flight organs. They develop when they have reached a stage of high perfection which enables them to decisively outperform other animals in the struggle for life.The chance, therefore, of finding in the fossil state species with transitional stages of structure is always low, because their number of individuals is less than that of those species which are fully developed in structure. I will now give two or three instances to illustrate the divergence of habits and the modification of habits among individuals of the same species.In either case, natural selection can readily adapt the structure of the animal to its altered habits, or specifically to one of several habits.It is difficult to decide, however, whether habit generally changes first and structure follows, or does a slight change in structure bring about a change in habit?But these are not important to us.Presumably the two almost always occur at the same time.As a case of altered habits, it suffices to cite the many English insects which now specialize in feeding on exotic plants or artificial food.Innumerable examples of divergent habits may be given: I have often observed in South America a violent flycatcher (Sau- rophagus sulphuratus ), which soars from place to place like a kestrel, and returns to other places. At other times, it stands quietly by the water, and then rushes into the water like a kingfisher to catch fish.In England, the stubble finch (Paurs inajor) is sometimes seen climbing branches almost like a creeper; it sometimes kills young birds by pecking their heads like a shrike , I have seen and heard them many times, like Nuthatch, pecking at the seeds of Yew on the branches.Hearne has seen black bears in North America swim for hours with their mouths wide open, almost like whales, catching insects in the water. Since we sometimes see individuals with habits different from those inherent in the same species and heterogeneous species, we may expect that these individuals will probably occasionally give rise to new species, which have unusual habits and whose structure is slightly or markedly different. ground changes, differing from their mode of construction.There are examples of this in nature.Can we give a more dramatic example of this adaptation than that of the woodpecker climbing trees and catching insects through cracks in the bark?In North America, however, some woodpeckers feed primarily on fruit, while others have long wings and catch insects in flight.On the La Plata Plain, there is almost no tree growing. There is a woodpecker called Colaptes campestris, which has two toes forward and two backwards, a long and pointed tongue, and thin and hard tail feathers. Enough to keep it upright on a tree trunk, but not as rigid as the tail feathers of a typical woodpecker, and it also has a beak that is straight and strong, however not as straight or strong as a typical woodpecker's beak, but it is also strong enough to hold a tree trunk. perforated.The bird, therefore, is, in all principal parts of its structure, a woodpecker.Even such insignificant characters as plumage color, rough tone, and undulating flight, express their close kinship with the English common woodpecker; but from my own observations, and from Asara's precise Observing, I can conclude that in some large districts it does not climb trees, and nests in burrows on banks!In some other places, however, it is this same woodpecker, according to Mr. Hudson, which frequents the trees, and makes holes in the trunks for its nests.I may give another example of a change in the habits of this genus, that of a Mexican woodpecker, described by De Saussure, who punches holes in hard trees for the storage of acorns (acorn ). Petrels are the most aerial and marine birds, but among the quiet channels of Fuego there is a species of jacana (Puffinuria berardi), in its general habits, and in its amazing diving power. in its swimming and flying manner, would have mistaken it for an auk or a grebe; nevertheless, it is essentially a petrel, but its Many parts of the system have been markedly modified in relation to the new habits of life; and La Plata's woodpecker has only slightly modified in structure.In the case of the Water-ouzel the keenest observer, from its autopsy, would never have imagined a semi-aquatic habit; but this bird, closely related to the thrush, subsisted on diving,— - It uses its wings in the water and grasps stones with its feet.All insects of the great order of the Hymenoptera are terrestrial, except the genus Proctotrupes, which Sir Lubbock has found to have an aquatic habit; it often enters the water, using its wings instead of its feet. , diving here and there, it can stay under the surface for as long as four hours; yet its structure does not change with this variable habit. Those who believe that living beings were once created as they are seen at the present day must often wonder if they encounter an animal whose habits and structure do not correspond.What could be more evident than that the webbed feet of ducks and geese are formed for swimming?The goose, however, native to the highlands, though it has webbed feet, seldom comes near water, and no friga-te-bird, except Audubon, has ever seen a friga-te-bird with four webbed feet. will land on the sea.On the other hand, Kettleloo and Coots are both remarkably aquatic birds, although their toes are membranous only on the edges.Is there anything more evident than that the long, membraneless toes of waders (Grallatores) are formed to facilitate walking on swamps and floating grass? —The water-hen and the landrail belong to this order, yet the former is almost as aquatic as the quail, and the latter as nearly as quail or partridge. Terrestrial.In these cases, and others that could be given, the habits have been changed without a corresponding change of structure.The webbed feet of the Highland goose may be said to have become almost rudimentary in function, though not in their construction.The deeply concave membranes between the toes of the frigate bird indicate that its structure has begun to change. Those who believe that living beings were created separately and countless times say so, in these cases, because the Creator liked to substitute one type of being for another; Words simply repeat the facts.Believers in the struggle for existence and in the principles of natural selection will admit that every living being is constantly striving to increase in number, and that any being, by a small variation in habits or structure, can outcompete others in the same place. One creature has the upper hand and takes that creature's position, no matter how different that position is from its own.Then he would not be surprised by the fact that geese and frigate birds with webbed feet live on dry land and seldom land on water; on swampy ground; woodpeckers where there are few trees; and the diving thrushes, diving hymenoptera, and petrels have the habits of sea-birds. extremely well-developed and complex organs The eye has inimitable apparatus for adjusting its focus for different distances, for accommodating different amounts of light, and for correcting spherical and chromatic aberrations and chromatic aberrations, and if it is supposed that the eye could have been formed by natural selection, it seems, I confess, is extremely ridiculous.When it was first said that the sun was at rest, and that the earth revolved round it, the common sense of mankind had declared this doctrine to be false; It is unbelievable.Reason tells me that if it could be shown that there are innumerable degrees of existence from the simple and imperfect eye to the complex and complete eye, and that each, as it is the case, is useful to its owner; Variations have occurred, as has been the case, and are capable of being inherited; and if they are useful to any animal under changing external conditions; then it is believed that a perfect and complex eye could be produced by natural selection. And the difficulty of formation, though insurmountable in our imagination, cannot be considered capable of overturning my theory.How nerves are sensitive to light is no more the province of our study than how life itself originated.But I can point out that some of the lowest beings, in whom no nerves are to be found, are also sensitive to light, and that, therefore, in their sarcode certain elements of the senses gather together and develop into creatures with this special sensibility. Nervous, that doesn't seem out of the question. In seeking the degrees by which the organs of any one species are perfected, we should look specifically at its immediate progenitors; but this is hardly possible, and we are obliged to look at other species and Other genera, namely, to observe the collateral lines of the common ancestor, to see what degrees were possible in the process of perfection, and perhaps to have an opportunity of seeing some of the classes inherited unchanged or only slightly modified. .But the state of the same organ in different classes also sometimes furnishes some indication of the steps by which it was brought to perfection. The simplest organ that can be called an eye is formed by an optic nerve, surrounded by pigment cells, and covered by a translucent membrane, but without any lens or other refractive body.However, according to M. Jourdain's research, we can even go down one step further, and we can see aggregates of pigment cells, which are clearly used as organs of vision, but do not have any nerves, but only grow in the flesh. organization above.Eyes of this simple nature mentioned above cannot see clearly, but can only be used to distinguish light from dark.According to the description of the author just mentioned, in some starfishes the pigmented layer surrounding the nerves has small depressions, filled with transparent jelly, and raised on the surface, like the cornea in higher animals.He didn't think it was used to reflect the image, but to focus the light and make it easier to feel.In this case of concentrating the rays we obtain the first and even the most important step towards the formation of true, image-reflecting eyes; Certainly, some are buried deep in the body, while others are close to the body surface), placed at an appropriate distance from the light collector, and an image will be formed on it. In the class Articulata we see the most primitive optic nerve surrounded by a mere pigmented layer which sometimes forms a pupil, but without a lens or other optical device.In insects it is now known that the corneas of the gigantic compound eyes are innumerable ommatidia, forming true lenses, and that this lens cone contains wonderfully modified nerve-fibres.But in articulated animals the organs of vision are so divergent that Muller has previously divided them into three main groups and seven subgroups, besides the conglomerate monocular The fourth major category. If we consider what has been here briefly stated, the wide, diverging, gradually graded range of eye structure in the lower animals; if we remember that the numbers of all extant forms are Much less, and it is not difficult to believe that natural selection could transform the simple apparatus of an optic nerve surrounded by a pigmented layer and covered by a transparent membrane, into an organ of sight as complete as that possessed by any member of the articulating animals. If anyone who has come here finds, after reading this book, that a large number of facts in it cannot be explained by any other means than by the theory of variation by natural selection, he ought not to hesitate. He hesitates to take another step forward; he should admit that even such a perfect structure as the eagle's eye is formed in this way, though in this case he does not know its transitional state, and it has been objected that,为了要使眼睛发生变化，并且作为一种完善的器官被保存下来，就必须有许多变化同时发生，而据推想，这是不能通过自然选择做到的；但正如我在论家养动物变异的那部著作里所曾企图阐明的，如果变异是极微细而逐渐的，就没有必要假定一切变异都是同时发生的。同时，不同种类的变异也可能为共同的一般目的服务：正如华莱斯先生曾经说过的，“如果一个晶状体具有太短的或大长的焦点，它可以由改变曲度或改变密度来进行调整；如果曲度不规则，光线不能聚集于一点，那末使曲度增加一些规则性，便是一种改进了。所以，虹膜的收缩和眼睛肌肉的运动，对于视觉都不是必要的，不过是使这一器官的构造在任何阶段中得到添加的和完善化的改进而已。”在动物界占最高等地位的脊椎动物里，其眼睛开始时是如此简单，如文昌鱼的眼睛，只是透明皮膜所构成的小囊，其上着生神经并围以色素，除此之外，别无其他装置。在鱼类和爬行类里，如欧文曾经说过的：“折光构造的诸级范围是很大的。”按照微尔和（ Virchow ）的卓见，甚至人类的这种美妙透明晶状体，在胚胎期也是由袋状皮褶中的表皮细胞的堆积而形成的；而玻璃体是由胚胎的皮下组织形成的，这个事实有重要的意义。虽然如此，对于这样奇异的然而并不是绝对完善的眼睛的形成，要达到公正的结论，理性还必须战胜想像；但是我痛感这是很困难的，所以有些人把自然选择原理应用到如此深远而有所踌躇，对此我并不觉得奇怪。 避免把眼睛和望远镜作比较，几乎是不可能的。我们知道望远镜是由人类的最高智慧经过长久不断的努力而完成的；我们自然地会推论眼睛也是通过一种多少类似的过程而形成的。但这种推论不是专横吗？我们有什么理由可以假定“造物主”也是以人类那样的智慧来工作呢？如果我们必须把眼睛和光学器具作一比较的话，我们就应当想像，它有一厚层的透明组织，在其空隙里充满着液体，下面有感光的神经，并且应当假定这一厚层内各部分的密度缓缓地不断地在改变着，以便分离成不同密度和厚度的各层，这些层的彼此距离各不相同，各层的表面也慢慢地改变着形状。进而我们必须假定有一种力量，这种力量就是自然选择即最适者生存，经常十分注意着透明层的每个轻微的改变；并且在变化了的条件之下，把无论以任何方式或任何程度产生比较明晰一点的映像的每一个变异仔细地保存下来。我们必须假定，这器官的每一种新状态，都是成百万地倍增着；每种状态一直被保存到更好的产生出来之后，这时旧的状态才全归毁灭。在生物体里，变异会引起一些轻微的改变，生殖作用会使这些改变几乎无限地倍增着，而自然选择乃以准确的技巧把每一次的改进都挑选出来。让这种过程百万年地进行着；每年作用于成百万的许多种类的个体；这种活的光学器具会比玻璃器具制造得更好，正如“造物主”的工作比人的工作做得更好一样，难道我们能不相信这一点吗？ 过渡的方式 倘使能证明有任何复杂器官不是经过无数的、连续的、轻微的变异而被形成的，那末我的学说就要完全破产。但是我还没有发现这种情形。无疑现在有许多器官，我们还不知道它们的过渡中间诸级，如果对于那些十分孤立的物种进行观察时，就更加如此，因为根据我的学说，它的周围的类型已大都绝灭了。或者，我们以一个纲内的一切成员所共有的一种器官做为论题时，也是如此，因为在这种情形里，那器官一定原来是在遥远的时代里形成的，此后，本纲内一切成员才发展起来：为要找寻那器官早先经过的过渡诸级，我们必须观察极古的始祖类型，可是这些类型早已绝灭了。 我们在断言一种器官可以不通过某一种类的过渡诸级而形成时，必须十分小心。在低等动物里，可以举出无数的例子来说明同样的器官同时能够进行全然不同的机能；如蜻蜓的幼虫和泥鳅（Cobites），它们的消化管兼营呼吸、消化和排泄的机能。再如水螅（ Hydra ），它可以把身体的内部翻到外面来，这样，外层就营消化，而营消化的内层就营呼吸了。在这等情形里，自然选择可能使本来营两种机能的器官的全部或一部专营一种机能，如果由此可以得到任何利益的话，于是经过不知不觉的步骤，器官的性质就被大大改变了。我们知道，有许多种植物正常地同时产生不同构造的花；如果这等植物仅仅产生一类的花，那末这一物种的性质就会比较突然地发生大变化。但同一株植物产生的两类花大概原来是由分级极细的步骤分化出来的，这些步骤至今可能在某些少数情形里还在进行着。 再者，两种不同的器官，或两种形式极不同的同样器官，可以同时在同一个个体里营相同的机能，并且这是极端重要的过渡方法：举一个例子来说明，——鱼类用鳃呼吸溶解在水中的空气，同时用鳔呼吸游离的空气，鳔被富有血管的隔膜分开，并有鳔管（ductus pneumaticus）以供给它空气。在植物界中可以举出另外一个例子：植物的攀缘方法有三种，用螺旋状的卷绕，用有感觉的卷须卷住一个支持物，以及用发出的气根；通常是不同的植物群只使用其中的一种方法，但有几种植物兼用两种方法，甚至也有同一个个体同时使用三种方法的。在所有这种情形里，两种器官当中的一个可能容易地被改变和完善化，以担当全部的工作，它在变异的进行中，曾经受到了另一种器官的帮助；于是另一种器官可能为着完全不同的另一个目的而被改变，或者可能整个被消灭掉。