ON THE EVIDENCE FAVOURABLE AND OPPOSED TO THE VIEW THAT SPECIES ARE NATURALLY FORMED RACES, DESCENDED FROM COMMON STOCKS
 In the Origin the division of the work into Parts I and II is omitted. In the MS. the chapters of Part II are numbered afresh, the present being Ch. I of Pt. II. I have thought it best to call it Ch. IV and there is evidence that Darwin had some thought of doing the same. It corresponds to Ch. IX of Origin, Ed. i., Ch. X in Ed. vi.
ON THE NUMBER OF INTERMEDIATE FORMS REQUIRED ON THE THEORY OF COMMON DESCENT; AND ON THEIR ABSENCE IN A FOSSIL STATE
I must here premise that, according to the view ordinarily received, the myriads of organisms, which have during past and present times peopled this world, have been created by so many distinct acts of creation. It is impossible to reason concerning the will of the Creator, and therefore, according to this view, we can see no cause why or why not the individual organism should have been created on any fixed scheme. That all the organisms of this world have been produced on a scheme is certain from their general affinities; and if this scheme can be shown to be the same with that which would result from allied organic beings descending from common stocks, it becomes highly improbable that they have been separately created by individual acts of the will of a Creator. For as well might it be said that, although the planets move in courses conformably to the law of gravity, yet we ought to attribute the course of each planet to the individual act of the will of the Creator. It is in every case more conformable with what we know of the government of this earth, that the Creator should have imposed only general laws. As long as no method was known by which races could become exquisitely adapted to various ends, whilst the existence of species was thought to be proved by the sterility of their offspring, it was allowable to attribute each organism to an individual act of creation. But in the two former chapters it has (I think) been shown that the production, under existing conditions, of exquisitely adapted species, is at least possible. Is there then any direct evidence in favour [[of]] or against this view? I believe that the geographical distribution of organic beings in past and present times, the kind of affinity linking them together, their so-called "metamorphic" and "abortive" organs, appear in favour of this view. On the other hand, the imperfect evidence of the continuousness of the organic series, which, we shall immediately see, is required on our theory, is against it; and is the most weighty objection. The evidence, however, even on this point, as far as it goes, is favourable; and considering the imperfection of our knowledge, especially with respect to past ages, it would be surprising if evidence drawn from such sources were not also imperfect.
 In the Essay of 1842 the author uses astronomy in the same manner as an illustration. In the Origin this does not occur; the reference to the action of secondary causes is more general, e.g. Ed. i. p. 488, vi. p. 668.
 It is interesting to find the argument from sterility given so prominent a place. In a corresponding passage in the Origin, Ed. i. p. 480, vi. p. 659, it is more summarily treated. The author gives, as the chief bar to the acceptance of evolution, the fact that "we are always slow in admitting any great change of which we do not see the intermediate steps"; and goes on to quote Lyell on geological action. It will be remembered that the question of sterility remained a difficulty for Huxley.
 Similar statements occur in the Essay of 1842, p. 24, note 1, and in the Origin, Ed. i. p. 299.
As I suppose that species have been formed in an analogous manner with the varieties of the domesticated animals and plants, so must there have existed intermediate forms between all the species of the same group, not differing more than recognised varieties differ. It must not be supposed necessary that there should have existed forms exactly intermediate in character between any two species of a genus, or even between any two varieties of a species; but it is necessary that there should have existed every intermediate form between the one species or variety of the common parent, and likewise between the second species or variety, and this same common parent. Thus it does not necessarily follow that there ever has existed [[a]] series of intermediate sub-varieties (differing no more than the occasional seedlings from the same seed-capsule,) between broccoli and common red cabbage; but it is certain that there has existed, between broccoli and the wild parent cabbage, a series of such intermediate seedlings, and again between red cabbage and the wild parent cabbage: so that the broccoli and red cabbage are linked together, but not necessarily by directly intermediate forms. It is of course possible that there may have been directly intermediate forms, for the broccoli may have long since descended from a common red cabbage, and this from the wild cabbage. So on my theory, it must have been with species of the same genus. Still more must the supposition be avoided that there has necessarily ever existed (though one may have descended from
 In the Origin, Ed. i. p. 280, vi. p. 414 he uses his newly-acquired knowledge of pigeons to illustrate this point.
 Compare the Origin, Ed. i. p. 281, vi. p. 414.
What evidence is there of a number of intermediate forms having existed, making a passage in the above sense, between the species of the same groups? Some naturalists have supposed that if every fossil which now lies entombed, together with all existing species, were collected together, a perfect series in every great class would be formed. Considering the enormous number of species requisite to effect this, especially in the above sense of the forms not being directly intermediate between the existing species and genera, but only intermediate by being linked through a common but often widely different ancestor, I think this supposition highly improbable. I am however far from underrating the probable number of fossilised species: no one who has attended to the wonderful progress of palĉontology during the last few years will doubt that we as yet have found only an exceedingly small fraction of the species buried in the crust of the earth. Although the almost infinitely numerous intermediate forms in no one class may have been preserved, it does not follow that they have not existed. The fossils which have been discovered, it is important to remark, do tend, the little way they go, to make good the series; for as observed by Buckland they all fall into or between existing groups. Moreover, those that fall between our existing groups, fall in, according to the manner required by our theory, for they do not directly connect two existing species of different groups, but they connect the groups themselves: thus the Pachydermata and Ruminantia are now separated by several characters, [[for instance]] the Pachydermata have both a tibia and fibula, whilst Ruminantia have only a tibia; now the fossil Macrauchenia has a leg bone exactly intermediate in this respect, and likewise has some other intermediate characters. But the Macrauchenia does not connect any one species of Pachydermata with some one other of Ruminantia but it shows that these two groups have at one time been less widely divided. So have fish and reptiles been at one time more closely connected in some points than they now are. Generally in those groups in which there has been most change, the more ancient the fossil, if not identical with recent, the more often it falls between existing groups, or into small existing groups which now lie between other large existing groups. Cases like the foregoing, of which there are many, form steps, though few and far between, in a series of the kind required by my theory.
 Origin, Ed. i. p. 301, vi. p. 440.
 Origin, Ed. i. p. 329, vi. p. 471.
 The structure of the Pachyderm leg was a favourite with the author. It is discussed in the Essay of 1842, p. 48. In the present Essay the following sentence in the margin appears to refer to Pachyderms and Ruminants: "There can be no doubt, if we banish all fossils, existing groups stand more separate." The following occurs between the lines "The earliest forms would be such as others could radiate from."
As I have admitted the high improbability, that if every fossil were disinterred, they would compose in each of the Divisions of Nature a perfect series of the kind required; consequently I freely admit, that if those geologists are in the right who consider the lowest known formation as contemporaneous with the first appearances of life; or the several formations as at all closely consecutive; or any one formation as containing a nearly perfect record of the organisms which existed during the whole period of its deposition in that quarter of the globe;--if such propositions are to be accepted, my theory must be abandoned.
 Origin, Ed. i. p. 307, vi. p. 448.
If the Palĉozoic system is really contemporaneous with the first appearance of life, my theory must be abandoned, both inasmuch as it limits from shortness of time the total number of forms which can have existed on this world, and because the organisms, as fish, mollusca and star-fish found in its lower beds, cannot be considered as the parent forms of all the successive species in these classes. But no one has yet overturned the arguments of Hutton and Lyell, that the lowest formations known to us are only those which have escaped being metamorphosed [[illegible]]; if we argued from some considerable districts, we might have supposed that even the Cretaceous system was that in which life first appeared. From the number of distant points, however, in which the Silurian system has been found to be the lowest, and not always metamorphosed, there are some objections to Hutton's and Lyell's view; but we must not forget that the now existing land forms only 1/5 part of the superficies of the globe, and that this fraction is only imperfectly known. With respect to the fewness of the organisms found in the Silurian and other Palĉozoic formations, there is less difficulty, inasmuch as (besides their gradual obliteration) we can expect formations of this vast antiquity to escape entire denudation, only when they have been accumulated over a wide area, and have been subsequently protected by vast superimposed deposits: now this could generally only hold good with deposits accumulating in a wide and deep ocean, and therefore unfavourable to the presence of many living things. A mere narrow and not very thick strip of matter, deposited along a coast where organisms most abound, would have no chance of escaping denudation and being preserved to the present time from such immensely distant ages.
 [[Pencil insertion by the author.]] The parent-forms of Mollusca would probably differ greatly from all recent,--it is not directly that any one division of Mollusca would descend from first time unaltered, whilst others had become metamorphosed from it.
 Origin, Ed. i. p. 291, vi. p. 426.
If the several known formations are at all nearly consecutive in time, and preserve a fair record of the organisms which have existed, my theory must be abandoned. But when we consider the great changes in mineralogical nature and texture between successive formations, what vast and entire changes in the geography of the surrounding countries must generally have been effected, thus wholly to have changed the nature of the deposits on the same area. What time such changes must have required! Moreover how often has it not been found, that between two conformable and apparently immediately successive deposits a vast pile of water-worn matter is interpolated in an adjoining district. We have no means of conjecturing in many cases how long a period has elapsed between successive formations, for the species are often wholly different: as remarked by Lyell, in some cases probably as long a period has elapsed between two formations as the whole Tertiary system, itself broken by wide gaps.
 [[Note in original.]] Reflect on coming in of the Chalk, extending from Iceland to the Crimea.
Consult the writings of any one who has particularly attended to any one stage in the Tertiary system (and indeed of every system) and see how deeply impressed he is with the time required for its accumulation. Reflect on the years elapsed in many cases, since the latest beds containing only living species have been formed;--see what Jordan Smith says of the 20,000 years since the last bed, which is above the boulder formation in Scotland, has been upraised; or of the far longer period since the recent beds of Sweden have been upraised 400 feet, what an enormous period the boulder formation must have required, and yet how insignificant are the records (although there has been plenty of elevation to bring up submarine deposits) of the shells, which we know existed at that time. Think, then, over the entire length of the Tertiary epoch, and think over the probable length of the intervals, separating the Secondary deposits. Of these deposits, moreover, those consisting of sand and pebbles have seldom been favourable, either to the embedment or to the preservation of fossils.
 Origin, Ed. i. p. 282, vi. p. 416.
 Origin, Ed. i. pp. 288, 300, vi. pp. 422, 438.
Nor can it be admitted as probable that any one Secondary formation contains a fair record even of those organisms which are most easily preserved, namely hard marine bodies. In how many cases have we not certain evidence that between the deposition of apparently closely consecutive beds, the lower one existed for an unknown time as land, covered with trees. Some of the Secondary formations which contain most marine remains appear to have been formed in a wide and not deep sea, and therefore only those marine animals which live in such situations would be preserved. In all cases, on indented rocky coasts, or any other coast, where sediment is not accumulating, although often highly favourable to marine animals, none can be embedded: where pure sand and pebbles are accumulating few or none will be preserved. I may here instance the great western line of the S. American coast, tenanted by many peculiar animals, of which none probably will be preserved to a distant epoch. From these causes, and especially from such deposits as are formed along a line of coast, steep above and below water, being necessarily of little width, and therefore more likely to be subsequently denuded and worn away, we can see why it is improbable that our Secondary deposits contain a fair record of the Marine Fauna of any one period. The East Indian Archipelago offers an area, as large as most of our Secondary deposits, in which there are wide and shallow seas, teeming with marine animals, and in which sediment is accumulating; now supposing that all the hard marine animals, or rather those having hard parts to preserve, were preserved to a future age, excepting those which lived on rocky shores where no sediment or only sand and gravel were accumulating, and excepting those embedded along the steeper coasts, where only a narrow fringe of sediment was accumulating, supposing all this, how poor a notion would a person at a future age have of the Marine Fauna of the present day. Lyell has compared the geological series to a work of which only the few latter but not consecutive chapters have been preserved; and out of which, it may be added, very many leaves have been torn, the remaining ones only illustrating a scanty portion of the Fauna of each period. On this view, the records of anteceding ages confirm my theory; on any other they destroy it.
 [[Note in original.]] Neither highest or lowest fish (i.e. Myxina [[?]] or Lepidosiren) could be preserved in intelligible condition in fossils.
 Origin, Ed. i. p. 290, vi. p. 425.
 See Origin, Ed. i. p. 310, vi. p. 452 for Lyell's metaphor. I am indebted to Prof. Judd for pointing out that Darwin's version of the metaphor is founded on the first edition of Lyell's Principles, vol. I. and vol. III.; see the Essay of 1842, p. 27.
Finally, if we narrow the question into, why do we not find in some instances every intermediate form between any two species? the answer may well be that the average duration of each specific form (as we have good reason to believe) is immense in years, and that the transition could, according to my theory, be effected only by numberless small gradations; and therefore that we should require for this end a most perfect record, which the foregoing reasoning teaches us not to expect. It might be thought that in a vertical section of great thickness in the same formation some of the species ought to be found to vary in the upper and lower parts, but it may be doubted whether any formation has gone on accumulating without any break for a period as long as the duration of a species; and if it had done so, we should require a series of specimens from every part. How rare must be the chance of sediment accumulating for some 20 or 30 thousand years on the same spot, with the bottom subsiding, so that a proper depth might be preserved for any one species to continue living: what an amount of subsidence would be thus required, and this subsidence must not destroy the source whence the sediment continued to be derived. In the case of terrestrial animals, what chance is there when the present time is become a pleistocene formation (at an earlier period than this, sufficient elevation to expose marine beds could not be expected), what chance is there that future geologists will make out the innumerable transitional sub-varieties, through which the short-horned and long-horned cattle (so different in shape of body) have been derived from the same parent stock? Yet this transition has been effected in the same country, and in a far shorter time, than would be probable in a wild state, both contingencies highly favourable for the future hypothetical geologists being enabled to trace the variation.
 See More Letters, vol. I. pp. 344-7, for Darwin's interest in the celebrated observations of Hilgendorf and Hyatt.
 This corresponds partly to Origin, Ed. i. p. 294, vi. p. 431.
 Origin, Ed. i. p. 299, vi. p. 437.