One must be careful not to press the idea of recapitulation too far, (1) because the individual life-history tends to skip

FIG. 39c.-Life-history of Penæus; Mysis stage. (From Fritz Müller.)

stages which occurred in the ancestral progress; (2) because the young animal may acquire new characters which are peculiar to its own near lineage and have little or no importance in connection with the general evolution of its race; (3) because, in short, the resemblance between the individual and racial history (so far as we know them) is general, not precise. Thus we regard Nauplius and Zoea rather as adaptive larval forms than as representatives of ancestral crustaceans. Moreover, if one insists too much on the approximate parallelism between the life-history of the individual and the progress of the race, one is apt to overlook the deeper problem-how it is that the recapitulation occurs to the extent that it undoubtedly does. The organism has no feeling for history that it should tread a sometimes circuitous path, because its far-off ancestors did so. To some extent we may think of inherited constitution as if it were the hand of the past upon the organism, compelling it to become thus or thus, but we must realise that this is a living not a dead hand; in other words these meta

morphoses have their efficient causes in the actual conditions of growth and development. The suggestion of

Kleinenberg referred to in a preceding chapter helps us, for

if we ask why an animal develops a notochord only to have it rapidly replaced by a backbone, part of the answer surely is that the notochord which in the historical evolution supplied the stimulus necessary for the development of a backbone, is still necessary in the individual history for the same purpose.

But there is no doubt that the idea of recapitulation is a very helpful one, in regard to our own history as well as in regard to animals, and we would do well to think of it much, and to read how Herbert Spencer (Principles of Biology, Lond. 1864-66) has discussed it in harmony with his general formula of evolution as a progress from the homogeneous to the heterogeneous; how Haeckel (Generelle Morphologie, Berlin, 1866) has illustrated it, and pithily summed it up in his "fundamental law of biogenesis" (Biogenetisches Grundgesetz), saying that ontogeny (individual development) recapitulates phylogeny (racial history); how Milnes. Marshall (see Nature, Sept. 1890) has recently tested and criticised it, defining the limits within which the notion can be regarded as true, and searching for a deeper rationale of the facts than the theory supplies.

(d) Organic Continuity. In a subsequent chapter on heredity, which simply means the relation of organic continuity between successive generations, I shall explain the fundamental idea that the reproductive cells owe their power of developing, and of developing into organisms like the parents, to the fact that they are in a sense continuous with those which gave origin to the parents. A fertilised egg-cell with certain qualities divides and forms a "body" in which these qualities are expressed, distributed, and altered in many ways by division of labour. But it also forms reproductive cells, which do not share in the upbuilding of the body, which are reproductive cells in fact because they do not do so, because they retain the intrinsic qualities of the original fertilised ovum, because they preserve its protoplasmic tradition. If this be so, and

there is much reason to believe it, then it is natural and necessary that these cells, liberated in due time, should behave as those behaved whose qualities they retain. It is necessary that like should beget like.

CHAPTER XIII

THE PAST HISTORY OF ANIMALS

1. The two Records-2. Imperfection of the Geological Record3. Paleontological Series—4. Extinction of Types-5. Various Difficulties-6. Relative Antiquity of Animals

1. The Two Records.-Reviewing the development of the chick, W. K. Parker said, "Whilst at work I seemed to myself to have been endeavouring to decipher a palimpsest, and that not erased and written upon just once, but five or six times over. Having erased, as it were, the characters of the culminating type-those of the gaudy Indian bird-I seemed to be amongst the sombre grouse, and then, towards incubation, the characters of the SandGrouse and Hemipod stood out before me. Rubbing these away, in my downward walk, the form of the Tinamou looked me in the face; then the aberrant Ostrich seemed to be described in large archaic characters; a little while and these faded into what could just be read off as pertaining to the Sea Turtle; whilst, underlying the whole, the Fish in its simplest Myxinoid form could be traced in morphological hieroglyphics."

There is another palimpsest—the geological record written in the rocks. For beneath the forms which disappeared, as it were, yesterday,—the Dodo and the Solitaire, the Moa and the Mammoth, the Cave Lion and the Irish Elk, there are mammals and birds of old-fashioned type the like of which no longer live. Beneath these lie the giant

reptiles, beneath these great amphibians, preceded by hosts of armoured fishes, beyond the first traces of which only backboneless animals are found. Yet throughout the chapters of this record, written during different æons on the earth's surface, persistent forms recur from age to age, many of them, such as some of the lamp-shells or Brachiopods, living on from near the apparent beginning even until now. But other races, like the Trilobites, have died out, leaving none which we can regard as in any sense their direct descendants. Other sets of animals, like the Ganoid fishes, grow in strength, attain a golden age of prosperous success, and wane away. As the earth grew older nobler forms appeared, and this history from the tombs, like that from the cradles of animals, shows throughout a gradual progress from simple to complex.

2. Imperfection of the Geological Record,-If complete records of past ages were safely buried in great treasurehouses such as Frederic Harrison proposes to make for the enlightenment of posterity, then paleontology would be easy. Then a genealogical tree connecting the Protist and Man would be possible, for we should have under our eyes what is now but a dream-a complete record of the past.

The record of the rocks is often compared to a library in which shelves have been destroyed and confused, in which most of the sets of volumes are incomplete, and most of the individual books much damaged. When we consider the softness of many animals, the chances against their being entombed, and the history of the earth's crust, our wonder is that the record is so complete as it is, that from "the strange graveyards of the buried past” we can learn so much about the life that once was.

We must not suppose the record to be as imperfect as our knowledge of it. Thus many regions of the earth's surface have been very partially studied, many have not been explored at all, many are inaccessible beneath the sea.

As to the record, the rocks in which fossils are found are sedimentary rocks formed under water, often they have been unmade and remade, burnt and denuded. The chances against preservation are many.

Soft animals rarely admit of preservation, those living on land and in the air are much less likely to be preserved than those living in water, the corpses of animals are often devoured or dissolved. Again the chances against preservation are many.

3. Palæontological Series.-Imperfect as the geological record is, several marvellously complete series of related animals have been disentombed. Thus, a series of fossilised freshwater snails (Planorbis) has been carefully worked out; its extremes are very different, but the distinctions between any two of the intermediate forms are hardly perceptible. The same is true in regard to another set of freshwater snails (Paludina), and on a much larger scale among the extinct cuttlefishes (Ammonites, etc.) whose shells have been thoroughly preserved. The modern crocodiles are linked by many intermediate forms to their extinct ancestors, and the modern horse to its pigmy progenitors. In cases like these, the evidences of continuously progressive evolution are conclusive.

4. Extinction of Types.-A few animals, such as some of the lamp-shells or Brachiopods, have persisted from almost the oldest rock-recorded ages till now. In most cases, however, the character of the family or order or class has gradually changed, and though the ancient forms are no longer represented, their descendants are with us. There is an extinction of individuals and a slow change of species.

On the other hand there are not a few fossil animals which have become wholly extinct, whose type is not represented in the modern fauna. Thus there are no animals alive that can be regarded as the lineal descendants of Trilobites and Eurypterids, or of many of the ancient reptiles. There is no doubt that a race may die out. Many different kinds of heavily armoured Ganoid fishes abounded in the ages when the Old Red Sandstone was formed, but only seven different kinds are now alive. The lamp-shells and the sea-lilies, once very numerous, are now greatly restricted. Once there were giants among Amphibians, now almost all are pigmies.