after last night's rain, the cliffs that bound the bay have come out in all their most brilliant colours; so this morning i am turning my steps seaward, and wandering along the great ridge of pebbles which here breaks the force of the channel waves as they beat against the long line of the dorset downs. our cliffs just at this point are composed of blue lias beneath, with a capping of yellow sandstone on their summits, above which in a few places the layer of chalk that once topped the whole country-side has still resisted the slow wear and tear of unnumbered centuries. these three elements give a variety to the bold and broken bluffs which is rare along the monotonous southern escarpment of the english coast. after rain, especially, the changes of colour on their sides are often quite startling in their vividness and intensity. to-day, for example, the yellow sandstone is tinged in parts with a deep russet red, contrasting admirably with the bright green of the fields above and the sombre steel-blue of the lias belt below. besides, we have had so many landslips along this bit of shore, that the various layers of rock have in more than one place got mixed up with one another into inextricable confusion. the little town nestling in the hollow behind me has long been famous as the head-quarters of early geologists; and not a small proportion of the people earn their livelihood to the present day by 'goin' a fossiling.' every child about the place recognises ammonites as 'snake-stones;' while even the rarer vertebrae of extinct saurians have acquired a local designation as 'verterberries.' so, whether in search of science or the picturesque, i often clamber down in this direction for my daily stroll, particularly when, as is the case to-day, the rain has had time to trickle through the yellow rock, and the sun then shines full against its face, to light it up with a rich flood of golden splendour.

the base of the cliffs consists entirely of a very soft and plastic blue lias mud. this mud contains large numbers of fossils, chiefly chambered shells, but mixed with not a few relics of the great swimming and flying lizards that swarmed among the shallow flats or low islands of the lias sea. when the blue mud was slowly accumulating in the hollows of the ancient bottom, these huge saurians formed practically the highest race of animals then existing upon earth. there were, it is true, a few prim?val kangaroo-mice and wombats among the rank brushwood of the mainland; and there may even have been a species or two of reptilian birds, with murderous-looking teeth and long lizard-like tails—descendants of those problematical creatures which printed their footmarks on the american trias, and ancestors of the later toothed bird whose tail-feathers have been naturally lithographed for us on the solenhofen slate. but in spite of such rare precursors of higher modern types, the saurian was in fact the real lord of earth in the lias ocean.

for him did his high sun flame, and his river billowing ran,

and he felt himself in his pride to be nature's crowning race.

we have adopted an easy and slovenly way of dividing all rocks into primary, secondary, and tertiary, which veils from us the real chronological relations of evolving life in the different periods. the lias is ranked by geologists among the earliest secondary formations: but if we were to distribute all the sedimentary rocks into ten great epochs, each representing about equal duration in time, the lias would really fall in the tenth and latest of all. so very misleading to the ordinary mind is our accepted geological nomenclature. nay, even commonplace geologists themselves often overlook the real implications of many facts and figures which they have learned to quote glibly enough in a certain off-hand way. let me just briefly reconstruct the chief features of this scarcely recognised world's chronology as i sit on this piece of fallen chalk at the foot of the mouldering cliff, where the stream from the meadow above brought down the newest landslip during the hard frosts of last december. first of all, there is the vast lapse of time represented by the laurentian rocks of canada. these laurentian rocks, the oldest in the world, are at least 30,000 feet in thickness, and it must be allowed that it takes a reasonable number of years to accumulate such a mass of solid limestone or clay as that at the bottom of even the widest prim?val ocean. in these rocks there are no fossils, except a single very doubtful member of the very lowest animal type. but there are indirect traces of life in the shape of limestone probably derived from shells, and of black lead probably derived from plants. all these early deposits have been terribly twisted and contorted by subsequent convulsions of the earth, and most of them have been melted down by volcanic action; so that we can tell very little about their original state. thus the history of life opens for us, like most other histories, with a period of uncertainty: its origin is lost in the distant vistas of time. still, we know that there was such an early period; and from the thickness of the rocks which represent it we may conjecture that it spread over three out of the ten great ?ons into which i have roughly divided geological time. next comes the period known as the cambrian, and to it we may similarly assign about two and a half ?ons on like grounds. the cambrian epoch begins with a fair sprinkling of the lower animals and plants, presumably developed during the preceding age; but it shows no remains of fish or any other vertebrates. to the silurian, devonian, and carboniferous periods we may roughly allow an ?on and a fraction each: while to the whole group of secondary and tertiary strata, comprising almost all the best-known english formations—red marl, lias, oolite, greensand, chalk, eocene, miocene, pliocene, and drift—we can only give a single ?on to be divided between them. such facts will sufficiently suggest how comparatively modern are all these rocks when viewed by the light of an absolute chronology. now, the first fishes do not occur till the silurian—that is to say, in or about the seventh ?on after the beginning of geological time. the first mammals are found in the trias, at the beginning of the tenth ?on. and the first known bird only makes its appearance in the oolite, about half-way through that latest period. this will show that there was plenty of time for their development in the earlier ages. true, we must reckon the interval between ourselves and the date of this blue mud at many millions of years; but then we must reckon the interval between the lias and the earliest cambrian strata at some six times as much, and between the lias and the lowest laurentian beds at nearly ten times as much. just the same sort of lessening perspective exists in geology as in ordinary history. most people look upon the age before the norman conquest as a mere brief episode of the english annals; yet six whole centuries elapsed between the landing of the real or mythical hengst at ebbsfleet and the landing of william the conqueror at hastings; while under eight centuries elapsed between the time of william the conqueror and the accession of queen victoria. but, just as most english histories give far more space to the three centuries since elizabeth than to the eleven centuries which preceded them, so most books on geology give far more space to the single ?on (embracing the secondary and tertiary periods) which comes nearest our own time, than to the nine ?ons which spread from the laurentian to the carboniferous epoch. in the earliest period, records either geological or historical are wholly wanting; in the later periods they become both more numerous and more varied in proportion as they approach nearer and nearer to our own time.

so too, in the days when mr. darwin first took away the breath of scientific europe by his startling theories, it used confidently to be said that geology had shown us no intermediate form between species and species. even at the time when this assertion was originally made it was quite untenable. all early geological forms, of whatever race, belong to what we foolishly call 'generalised' types: that is to say, they present a mixture of features now found separately in several different animals. in other words, they represent early ancestors of all the modern forms, with peculiarities intermediate between those of their more highly differentiated descendants; and hence we ought to call them 'unspecialised' rather than 'generalised' types. for example, the earliest ancestral horse is partly a horse and partly a tapir: we may regard him as a tertium quid, a middle term, from which the horse has varied in one direction and the tapir in another, each of them exaggerating certain special peculiarities of the common ancestor and losing others, in accordance with the circumstances in which they have been placed. science is now perpetually discovering intermediate forms, many of which compose an unbroken series between the unspecialised ancestral type and the familiar modern creatures. thus, in this very case of the horse, professor marsh has unearthed a long line of fossil animals which lead in direct descent from the extremely unhorse-like eocene type to the developed arab of our own times. similarly with birds, professor huxley has shown that there is hardly any gap between the very bird-like lizards of the lias and the very lizard-like birds of the oolite. such links, discovered afresh every day, are perpetual denials to the old parrot-like cry of 'no geological evidence for evolution.'