especially as the plant's own anthers and pistil generally stand so close

  together that self-fertilisation seems almost inevitable. Many flowers, on

  the other hand, have their organs of fructification closely enclosed, as in

  the great papilionaceous or pea-family; but in several, perhaps in all,

  such flowers, there is a very curious adaptation between the structure of

  the flower and the manner in which bees suck the nectar; for, in doing

  this, they either push the flower's own pollen on the stigma, or bring

  pollen from another flower. So necessary are the visits of bees to

  papilionaceous flowers, that I have found, by experiments published

  elsewhere, that their fertility is greatly diminished if these visits be

  prevented. Now, it is scarcely possible that bees should fly from flower

  to flower, and not carry pollen from one to the other, to the great good,

  as I believe, of the plant. Bees will act like a camel-hair pencil, and it

  is quite sufficient just to touch the anthers of one flower and then the

  stigma of another with the same brush to ensure fertilisation; but it must

  not be supposed that bees would thus produce a multitude of hybrids between

  distinct species; for if you bring on the same brush a plant's own pollen

  and pollen from another species, the former will have such a prepotent

  effect, that it will invariably and completely destroy, as has been shown

  by Gartner, any influence from the foreign pollen.

  When the stamens of a flower suddenly spring towards the pistil, or slowly

  move one after the other towards it, the contrivance seems adapted solely

  to ensure self-fertilisation; and no doubt it is useful for this end: but,

  the agency of insects is often required to cause the stamens to spring

  forward, as Kolreuter has shown to be the case with the barberry; and

  curiously in this very genus, which seems to have a special contrivance for

  self-fertilisation, it is well known that if very closely-allied forms or

  varieties are planted near each other, it is hardly possible to raise pure

  seedlings, so largely do they naturally cross. In many other cases, far

  from there being any aids for self-fertilisation, there are special

  contrivances, as I could show from the writings of C. C. Sprengel and from

  my own observations, which effectually prevent the stigma receiving pollen

  from its own flower: for instance, in Lobelia fulgens, there is a really

  beautiful and elaborate contrivance by which every one of the infinitely

  numerous pollen-granules are swept out of the conjoined anthers of each

  flower, before the stigma of that individual flower is ready to receive

  them; and as this flower is never visited, at least in my garden, by

  insects, it never sets a seed, though by placing pollen from one flower on

  the stigma of another, I raised plenty of seedlings; and whilst another

  species of Lobelia growing close by, which is visited by bees, seeds

  freely. In very many other cases, though there be no special mechanical

  contrivance to prevent the stigma of a flower receiving its own pollen,

  yet, as C. C. Sprengel has shown, and as I can confirm, either the anthers

  burst before the stigma is ready for fertilisation, or the stigma is ready

  before the pollen of that flower is ready, so that these plants have in

  fact separated sexes, and must habitually be crossed. How strange are

  these facts! How strange that the pollen and stigmatic surface of the same

  flower, though placed so close together, as if for the very purpose of

  self-fertilisation, should in so many cases be mutually useless to each

  other! How simply are these facts explained on the view of an occasional

  cross with a distinct individual being advantageous or indispensable!

  If several varieties of the cabbage, radish, onion, and of some other

  plants, be allowed to seed near each other, a large majority, as I have

  found, of the seedlings thus raised will turn out mongrels: for instance,

  I raised 233 seedling cabbages from some plants of different varieties

  growing near each other, and of these only 78 were true to their kind, and

  some even of these were not perfectly true. Yet the pistil of each

  cabbage-flower is surrounded not only by its own six stamens, but by those

  of the many other flowers on the same plant. How, then, comes it that such

  a vast number of the seedlings are mongrelized? I suspect that it must

  arise from the pollen of a distinct variety having a prepotent effect over

  a flower's own pollen; and that this is part of the general law of good

  being derived from the intercrossing of distinct individuals of the same

  species. When distinct species are crossed the case is directly the

  reverse, for a plant's own pollen is always prepotent over foreign pollen;

  but to this subject we shall return in a future chapter.

  In the case of a gigantic tree covered with innumerable flowers, it may be

  objected that pollen could seldom be carried from tree to tree, and at most

  only from flower to flower on the same tree, and that flowers on the same

  tree can be considered as distinct individuals only in a limited sense. I

  believe this objection to be valid, but that nature has largely provided

  against it by giving to trees a strong tendency to bear flowers with

  separated sexes. When the sexes are separated, although the male and

  female flowers may be produced on the same tree, we can see that pollen

  must be regularly carried from flower to flower; and this will give a

  better chance of pollen being occasionally carried from tree to tree. That

  trees belonging to all Orders have their sexes more often separated than

  other plants, I find to be the case in this country; and at my request Dr.

  Hooker tabulated the trees of New Zealand, and Dr. Asa Gray those of the

  United States, and the result was as I anticipated. On the other hand, Dr.

  Hooker has recently informed me that he finds that the rule does not hold

  in Australia; and I have made these few remarks on the sexes of trees

  simply to call attention to the subject.

  Turning for a very brief space to animals: on the land there are some

  hermaphrodites, as land-mollusca and earth-worms; but these all pair. As

  yet I have not found a single case of a terrestrial animal which fertilises

  itself. We can understand this remarkable fact, which offers so strong a

  contrast with terrestrial plants, on the view of an occasional cross being

  indispensable, by considering the medium in which terrestrial animals live,

  and the nature of the fertilising element; for we know of no means,

  analogous to the action of insects and of the wind in the case of plants,

  by which an occasional cross could be effected with terrestrial animals

  without the concurrence of two individuals. Of aquatic animals, there are

  many self-fertilising hermaphrodites; but here currents in the water offer

  an obvious means for an occasional cross. And, as in the case of flowers,

  I have as yet failed, after consultation with one of the highest

  authorities, namely, Professor Huxley, to discover a single case of an

  hermaphrodite animal with the organs of reproduction so perfectly enclosed

  within the body, that access from without and the occas
ional influence of a

  distinct individual can be shown to be physically impossible. Cirripedes

  long appeared to me to present a case of very great difficulty under this

  point of view; but I have been enabled, by a fortunate chance, elsewhere to

  prove that two individuals, though both are self-fertilising

  hermaphrodites, do sometimes cross.

  It must have struck most naturalists as a strange anomaly that, in the case

  of both animals and plants, species of the same family and even of the same

  genus, though agreeing closely with each other in almost their whole

  organisation, yet are not rarely, some of them hermaphrodites, and some of

  them unisexual. But if, in fact, all hermaphrodites do occasionally

  intercross with other individuals, the difference between hermaphrodites

  and unisexual species, as far as function is concerned, becomes very small.

  From these several considerations and from the many special facts which I

  have collected, but which I am not here able to give, I am strongly

  inclined to suspect that, both in the vegetable and animal kingdoms, an

  occasional intercross with a distinct individual is a law of nature. I am

  well aware that there are, on this view, many cases of difficulty, some of

  which I am trying to investigate. Finally then, we may conclude that in

  many organic beings, a cross between two individuals is an obvious

  necessity for each birth; in many others it occurs perhaps only at long

  intervals; but in none, as I suspect, can self-fertilisation go on for

  perpetuity.

  Circumstances favourable to Natural Selection. -- This is an extremely

  intricate subject. A large amount of inheritable and diversified

  variability is favourable, but I believe mere individual differences

  suffice for the work. A large number of individuals, by giving a better

  chance for the appearance within any given period of profitable variations,

  will compensate for a lesser amount of variability in each individual, and

  is, I believe, an extremely important element of success. Though nature

  grants vast periods of time for the work of natural selection, she does not

  grant an indefinite period; for as all organic beings are striving, it may

  be said, to seize on each place in the economy of nature, if any one

  species does not become modified and improved in a corresponding degree

  with its competitors, it will soon be exterminated.

  In man's methodical selection, a breeder selects for some definite object,

  and free intercrossing will wholly stop his work. But when many men,

  without intending to alter the breed, have a nearly common standard of

  perfection, and all try to get and breed from the best animals, much

  improvement and modification surely but slowly follow from this unconscious

  process of selection, notwithstanding a large amount of crossing with

  inferior animals. Thus it will be in nature; for within a confined area,

  with some place in its polity not so perfectly occupied as might be,

  natural selection will always tend to preserve all the individuals varying

  in the right direction, though in different degrees, so as better to fill

  up the unoccupied place. But if the area be large, its several districts

  will almost certainly present different conditions of life; and then if

  natural selection be modifying and improving a species in the several

  districts, there will be intercrossing with the other individuals of the

  same species on the confines of each. And in this case the effects of

  intercrossing can hardly be counterbalanced by natural selection always

  tending to modify all the individuals in each district in exactly the same

  manner to the conditions of each; for in a continuous area, the conditions

  will generally graduate away insensibly from one district to another. The

  intercrossing will most affect those animals which unite for each birth,

  which wander much, and which do not breed at a very quick rate. Hence in

  animals of this nature, for instance in birds, varieties will generally be

  confined to separated countries; and this I believe to be the case. In

  hermaphrodite organisms which cross only occasionally, and likewise in

  animals which unite for each birth, but which wander little and which can

  increase at a very rapid rate, a new and improved variety might be quickly

  formed on any one spot, and might there maintain itself in a body, so that

  whatever intercrossing took place would be chiefly between the individuals

  of the same new variety. A local variety when once thus formed might

  subsequently slowly spread to other districts. On the above principle,

  nurserymen always prefer getting seed from a large body of plants of the

  same variety, as the chance of intercrossing with other varieties is thus

  lessened.

  Even in the case of slow-breeding animals, which unite for each birth, we

  must not overrate the effects of intercrosses in retarding natural

  selection; for I can bring a considerable catalogue of facts, showing that

  within the same area, varieties of the same animal can long remain

  distinct, from haunting different stations, from breeding at slightly

  different seasons, or from varieties of the same kind preferring to pair

  together.

  Intercrossing plays a very important part in nature in keeping the

  individuals of the same species, or of the same variety, true and uniform

  in character. It will obviously thus act far more efficiently with those

  animals which unite for each birth; but I have already attempted to show

  that we have reason to believe that occasional intercrosses take place with

  all animals and with all plants. Even if these take place only at long

  intervals, I am convinced that the young thus produced will gain so much in

  vigour and fertility over the offspring from long-continued

  self-fertilisation, that they will have a better chance of surviving and

  propagating their kind; and thus, in the long run, the influence of

  intercrosses, even at rare intervals, will be great. If there exist

  organic beings which never intercross, uniformity of character can be

  retained amongst them, as long as their conditions of life remain the same,

  only through the principle of inheritance, and through natural selection

  destroying any which depart from the proper type; but if their conditions

  of life change and they undergo modification, uniformity of character can

  be given to their modified offspring, solely by natural selection

  preserving the same favourable variations.

  Isolation, also, is an important element in the process of natural

  selection. In a confined or isolated area, if not very large, the organic

  and inorganic conditions of life will generally be in a great degree

  uniform; so that natural selection will tend to modify all the individuals

  of a varying species throughout the area in the same manner in relation to

  the same conditions. Intercrosses, also, with the individuals of the same

  species, which otherwise would have inhabited the surrounding and

  differently circumstanced districts, will be prevented. But isolation

  probably acts more efficiently in checking the immigration of better

  adapted organisms, af
ter any physical change, such as of climate or

  elevation of the land, &c.; and thus new places in the natural economy of

  the country are left open for the old inhabitants to struggle for, and

  become adapted to, through modifications in their structure and

  constitution. Lastly, isolation, by checking immigration and consequently

  competition, will give time for any new variety to be slowly improved; and

  this may sometimes be of importance in the production of new species. If,

  however, an isolated area be very small, either from being surrounded by

  barriers, or from having very peculiar physical conditions, the total

  number of the individuals supported on it will necessarily be very small;

  and fewness of individuals will greatly retard the production of new

  species through natural selection, by decreasing the chance of the

  appearance of favourable variations.

  If we turn to nature to test the truth of these remarks, and look at any

  small isolated area, such as an oceanic island, although the total number

  of the species inhabiting it, will be found to be small, as we shall see in

  our chapter on geographical distribution; yet of these species a very large

  proportion are endemic,--that is, have been produced there, and nowhere

  else. Hence an oceanic island at first sight seems to have been highly

  favourable for the production of new species. But we may thus greatly

  deceive ourselves, for to ascertain whether a small isolated area, or a

  large open area like a continent, has been most favourable for the

  production of new organic forms, we ought to make the comparison within

  equal times; and this we are incapable of doing.

  Although I do not doubt that isolation is of considerable importance in the

  production of new species, on the whole I am inclined to believe that

  largeness of area is of more importance, more especially in the production

  of species, which will prove capable of enduring for a long period, and of

  spreading widely. Throughout a great and open area, not only will there be

  a better chance of favourable variations arising from the large number of

  individuals of the same species there supported, but the conditions of life

  are infinitely complex from the large number of already existing species;

  and if some of these many species become modified and improved, others will

  have to be improved in a corresponding degree or they will be exterminated.

  Each new form, also, as soon as it has been much improved, will be able to

  spread over the open and continuous area, and will thus come into

  competition with many others. Hence more new places will be formed, and

  the competition to fill them will be more severe, on a large than on a

  small and isolated area. Moreover, great areas, though now continuous,

  owing to oscillations of level, will often have recently existed in a

  broken condition, so that the good effects of isolation will generally, to

  a certain extent, have concurred. Finally, I conclude that, although small

  isolated areas probably have been in some respects highly favourable for

  the production of new species, yet that the course of modification will

  generally have been more rapid on large areas; and what is more important,

  that the new forms produced on large areas, which already have been

  victorious over many competitors, will be those that will spread most

  widely, will give rise to most new varieties and species, and will thus

  play an important part in the changing history of the organic world.

  We can, perhaps, on these views, understand some facts which will be again

  alluded to in our chapter on geographical distribution; for instance, that

  the productions of the smaller continent of Australia have formerly

  yielded, and apparently are now yielding, before those of the larger

  Europaeo-Asiatic area. Thus, also, it is that continental productions have

  everywhere become so largely naturalised on islands. On a small island,