deposited, a large portion, and perhaps the whole of this region became, for a time, land, and the uplifting was attended by considerable dislocation and flexing of the strata. In numerous localities the Cretaceous strata are seen to be denuded, and the lowest Tertiary beds lie across the bevelled edges. This uplifting took place after the deposition of a group of beds which in part, at least, are the equivalents of those which King and Hayden have named the Laramie Group. I accept the verdict of Marsh, Meek, King, and Powell, that these beds belong to the local Cretaceous_series, and reject the decision of Hayden, that they are Tertiary. Thus the close of the Cretaceous is marked by a physical break separating it from the local Tertiary series by widely distributed unconformities.

After an unknown interval of denudation immediately following the close of the Cretaceous period the region was again submerged, and then began the deposition of that remarkable series of Eocene beds which form such a striking feature in the stratigraphy of the peripheral parts of the Plateau Country. Around the southern flanks of the Uintas their aggregate thickness exceeds 4,000 feet, but south westward the upper members at length disappear, and seventy miles north of the Grand Cañon only the lower portion of the local Eocene (the Bitter Creek of Powell or Vermilion Creek of King) remain; indeed, in the latter locality no later beds than the Bitter Creek were deposited. The evidence is now conclusive that the Bitter Creek series stretched more than a hundred miles across the Plateau Country, covering, doubtless, its entire extent, while the middle and later Eocene covered smaller areas to the northward. Only marginal remnants of these huge deposits now remain. The heart of them has been eroded and swept away. Just at the commencement of the Tertiary periods the Plateau Country was covered with brackish water, having perhaps an analogy to the Baltic or Euxine, but after the accumulation of a few hundred feet of deposits the region became a vast inland lake. Its northern shore was along the base of the Uintas, which had then apparently gained their first elevation. Its north-western shore, by a coincidence which can hardly be accidental, lay along the identical boundary which now sharply separates the Plateau Country from the Great Basin, and the latter was one of the mainlands which furnished the sediments of the lake. From the angle where the Uintas join the Wasatch it is possible to trace this shore line more than 300 miles south-westward into Arizona with certainty, and to point out its principal bays and headlands, and even to locate the sites of some of the ancient river channels through which the lower Eocene sediments were brought down. The eastern, south-eastern, and southern margins, and the remainder of the south-western margin, remain to be determined by future exploration. At length, after one-third to one-half of the lacustrine beds had been laid down, there began a series of events which has developed the physical features of the Plateau Country, and which has pursued an unbroken course to the present time, and which even yet may not have culminated. Then began that uplifting which has raised the Plateau Country more than 13,000 feet on an average. Then, too, began a marvellous erosion which has cut down the mean level about onehalf that amount, leaving the present mean altitude nearly 6,500 feet. At the inception of this process the lake began to dry up, the south-western portion now drained by the Lower Colorado being the first to emerge. Gradually through the succeeding periods the lake contracted its area, withdrawing northward to the Uinta Mountains, where, at the close of the Eocene, it disappeared.

We are now in a position to trace the origin, growth, and history of the Colorado River, if not from the beginning, at least from an epoch near its beginning. Its creation was not the event of one epoch, but a gradual

process extending through several periods. The lower course, extending from the mouth of the Virgin to the Pacific, is the oldest portion, and makes its appearance in geological history a little before, but very near, the middle Eocene. Whether it existed before this epoch is not beyond doubt, but probably it did. But earlier than the Tertiary periods it cannot go; for it is certain that up to the close of Cretaceous times the ocean flowed over its track. When the Plateau Country was first isolated from the ocean it became a brackish Euxine, and may be presumed to have had a Hellespont somewhere. It soon after became an inland lake and must have had a St. Lawrence to keep its waters fresh. There can be little doubt that in the middle Eocene the outlet was the lower course of the Colorado. Whether the lake prior to that had some other outlet which it abandoned for this one is an open question, with the probabilities (on general principles) in favour of the negative. But the question is of no great importance.

The growth of the Colorado may be illustrated by considering what might happen to the St. Lawrence if the whole region of the Canadian lakes were uplifted two thousand feet. In no great length of time Ontario would be drained by the St. Lawrence, lowering its channel, and that river would become one with the Niagara. The same process would be repeated at Erie, Huron, and Superior, the lakes vanishing and leaving only a great river with many branches. Such was the origin of the Colorado; first a Hellespont, then a St. Lawrence, then a common but rather large river heading in the interior of a continent. Its principal branch, the Green River, cuts through the Uinta mountains by the Flaming Gorge and Cañon of Lodore. A second lake, apparently coeval with the one we have just discussed, lay to the north of that range and poured its waters through these gateways into the southern lake. What other bodies of fresh water may have been connected with either of these it is impossible to say at present.

At the epoch when the desiccation was completed it is not probable that the cañons had any existence, for the indications are that the elevation of the country at the commencement of the Miocene period was not great. Conditions favourable to cañon cutting are highly exceptional, and there is no evidence that this exceptional combination of conditions existed at that time, while there is much evidence that it did not. That the conditions, however, were favourable to a rapid rate of erosion is highly probable. But the forms which it would produce might be more nearly analogous to those which may be observed in eastern Ohio and western Pennsylvania. That the climate was moist and sub-tropical is rendered probable by the vegetable remains found in the surrounding regions, and it is only rational to suppose that such a climate in a moderately elevated region would yield such results as may be seen in countries similarly conditioned. Whether the valleys were broad or narrow, abruptly walled or gently sloped, matters little. It is almost certain that they were not deep. The great cañons which we now see had not even been commenced, although they were foreshadowed, and the train of events which was to produce them at a later period had started into activity.

The history of the Colorado and its drainage system during Miocene time must be spoken of only in general terms. In truth during this great age there is no evidence of the occurrence of any critical event aside from the general process of uplifting and erosion which affected the region as a whole. The vast erosion of this region has swept away so much of its mass that most of the evidence as to the details has vanished with its rocks. But the more important features of the work, its general plan in outline, have left well-marked traces and these can be unravelled. It was a period of slow uplifting, reaching a great amount in the aggregate, and it was also

perished and undoubtedly they have shifted parts of their courses somewhat; but on the whole their tenacity of life is wonderful, and the obstinacy with which they sometimes maintain their positions is in powerful contrast with the instability of other topographical features. This characteristic, however, fails at low levels. A river near its mouth may often change its course; but where the country is high enough to enable it once to fasten its • grip it will hold it, despite all the changes to which the surface of a continent is ordinarily subject throughout the term of its secular existence. Its stability and persistence will depend usually upon its altitude, or what amounts to the same thing, upon the rapidity of its slope. When that is small we may look for signs of inconstancy. Other conditions might be formulated which could affect it or modify it; but on the whole the fact remains that rivers have a remarkable power of maintaining their positions. It would be difficult to point out an instance where a great river has ever existed under conditions more favourable to longevity and stability of position than those of the Colorado and its principal tributaries. Since the epoch when it commenced to flow it has been situated in a rising area. Its springs and rills have been among the mountains and its slope has throughout its career been continuously though slightly increasing. The relations of its tributaries have in this respect been the same, and indeed the river and its tributaries have been a system and not merely an aggregate, the latter dependent upon and perfectly responsive to the physical conditions of the former. And now we come to the point. The Colorado and its tributaries run to-day just where they ran in the Eocene period. Since that time mountains have risen across their tracks, whose present summits mark less than. half their total uplifts; the river has cleft them down to their foundations. The Green River, passing the Pacific Railway, enters the Uintas by the Flaming Gorge, and after reaching the heart of this chain, turns eastward parallel to its axis for thirty miles, and then southward, cutting its way fout by the splendid cañon of Lodore. Then following westward along the southern base of the range for five miles, a strange caprice seizes it. Not satisfied with the terrible gash it has inflicted upon this noble chain, it darts at it viciously once more, and entering it, cuts a great horse-shoe cañon more than 2,700 feet deep, and then emerging, goes on its way. Thenceforward, through a tortuous course of more than 300 miles down stream the strata slowly rise-the river almost constantly running against the gentle dip of the beds, cutting through one after another, until its channel is sunk deep in the carboniferous. Further down, near the head of the Marble Cañon, the Kaibab rose up to contest its passage, and a chasm more than 6,200 feet in depth bears witness to the result. It is needless to multiply instances. The entire province is a vast category of instances of drainage channels running counter to the structural slopes of the country. I am unable to recall a single tributary to the right bank of the Colorado which does not somewhere, and generally throughout the greater part of its course, run against the dips. The northern tributaries of the Grand Cañon have their entire courses thus related. If we were to take the sums of the lengths of the river and its right hand affluents, we should find that at least three-fourths of that total length lay where the streams run against the dips.

It is clear, then, that the structural deformations of the region the faults, flexures, and swells, had nothing to do with determining the present distribution of the drainage. The rivers are where they are in spite of them. As these irregularities rose up, the streams turned neither to the right nor to the left, but cut their way through them in the same old places. The process may be illustrated by a feeble analogy with the saw mill. The river is the saw, the strata are the timber which is fed against it. The saw-log moves while the saw vibrates

in its place. The river holds its position almost as rigidly, and the rising strata are dissevered by its ceaseless wear.

What, then, determined the situation of the present drainage channels? The answer is that they were determined by the configuration of the old Eocene lake-bottom at the time the lake was drained. Then surely the watercourses ran in conformity with the surface of the uppermost Tertiary stratum. Soon afterwards that surface began to be deformed by unequal displacement, but the rivers had fastened themselves to their places and refused to be diverted. This, then, is the key which unlocks for the geologist the vestibule of the Plateau Country. The rivers were born with the country itself, they are older than its cliffs and cañons, older than its great erosionthe oldest things in its Tertiary history; nay, they are its history, which we may yet read imperfectly in their cañon walls. The mountains and plateaus are of subsequent origin. They arose athwart the streams only to be cleft asunder to give passage to the waters. The rivers amid all changes have ever successfully maintained their right of way. Such are the uses of the limited theorem of the persistence of rivers. I shall not attempt to suggest how far it may be applicable to other regions, but I am confident that any geologist visiting the Plateau Country will be quickly overwhelmed with the conviction that it is true there.

In this connection it remains to add something to indicate the magnitude of the work accomplished, and the real extent of the obstacles which the Colorado has accomplished in maintaining its existence. In the Colorado itself, the maximum work has been done at the Grand Cañon (Fig. 2). This chasm is 217 miles in length, to which should be added properly the Marble Cañon above, 69 miles long, since the two are continuous, and their separation merely nominal. The average depth of the Grand Cañon is a little more than 5,200 feet-almost exactly one mile. Its maximum depth through the Kaibab Plateau is nearly 6,300 feet, this depth being maintained approximately as the river runs for about fifty miles. Surely it might be thought that to cut such an abyss is work enough in the life of one river however ancient of days. But the summit of the Kaibab is Carboniferous limestone. When the river began to run in this part the whole local Mesozoic and lower Eocene series rested upon the site of this plateau, but have since been swept away together with a part of the Carboniferous rocks. The river has cut through the entire fossiliferous system of strata and now runs 2,000 feet deep in the archæan. The total thickness of the fossiliferous system here is, or rather was, very nearly 17,000 feet. Hence in its lifetime the river has cut through about 19,000 feet of strata. Through the remainder of the Grand Cañon the total cutting has been from 2,000 to 3,000 feet less. As we ascend the river the amount diminishes-not regularly but with local maxima-until we approach the southern base of the Uintas. The principal branch, the Green River, has cut its channel into the quartzites of this range even more deeply than the Colorado in the Kaibab. Yet strangely enough the instant the Green is clear of the mountains it enters a long stretch where the cutting has been practically nothing. The explanation of this contrast will become obvious to the geologist by a mere reference to the fact that where the cutting has been zero the locality has been always at the base level of erosion, and never above it. Only those parts which rise above the base level are cut down.

THE BRITISH MUSEUM LIBRARY

WHAT sort of reference library can be provided in

connection with the natural history collections when they are moved from the British Museum to South Kensington? is a subject now under consideration. It is stated on good authority that, so far as the building arrangements at Kensington go, no provision whatever has been made for library space, and that in the Act passed at the end of last session to enable the trustees to move the collections, a reference library seems to have been entirely overlooked. That Act has, however, been the subject of a resolution by the General Committee of the British Association, requesting the Council to take such steps in the matter as they might deem expedient; and although the resolution had principal reference to the administration of the collections, its force extends equally to such an important matter as a library, should the Council "deem it expedient" to include that subject. Whatever may be the decision as to what part of the library can be transferred to Kensington, or what ought to be transferred, it is only the works relating to biological studies that will be essential there, and it is only these, therefore, that are likely to be the subject of inquiry. But it might, perhaps, lead to changes of great value to those who use the British Museum Library for the purposes of referring to the literature of science in its other branches as well, if the inquiry could be extended to include the question of the actual state of this literature, which is available for use at the Museum. Whether it should be expected that the national library should contain as complete a collection as possible of scientific publications, or whether those who wish to consult them ought to belong to several of the incorporated learned societies, and use their libraries, is a separate question. When this question is considered, if it has to be considered at all, it must not be forgotten that no one society has anything like a comprehensive collection of scientific works, each society aiming at completeness in its own subjects; that to belong to several societies is not within the means of every student; and that, as one of the advantages of these societies is that members may take books away, no one can be sure of finding on the shelves what they may wish to consult.

But quite apart from such a question as this it would be of great use, with a prospect of effecting changes, to know what is the actual state of the British Museum library as regards scientific literature.

Only those who have had occasion to work at the library can have any idea how incomplete it is in this department, or what a wearisome toil it is, in consequence of the system of cataloguing adopted, to find whether a work they wish to consult is or is not there. If the experiences of those who have had occasion to use the library for such purposes could be collected, the probability is that it would be found that from a third to a half of the works asked for were not obtainable there. This may seem at first sight a very surprising assertion to make, but there is good reason to believe it true. What the Museum does or does not contain can, however, be known only by an inquiry, especially directed to ascertain the facts. A reference to the catalogue, as at present arranged, is quite inadequate to give an The officials themselves could not tell from it what they have and what they have not. For example: suppose a particular volume of the Reports of the United States Geological Survey of the Territories is wanted, a reference to the catalogue will not tell whether it has been received or not. The catalogue simply gives the information that the series is on a particular shelf. ticket for the whole series is filled up according to the requirements of the reading-room regulations with the press mark, the title, and Washington, 1873, &c., 4° added, then it will be found when the books are brought

answer.

If a

to the reader's seat, that only volumes two, six, nine, and ten of the whole series are there. This illustration applies to all publications which are issued in a series either by societies or by government departments. To ascertain, therefore, what is the incompleteness of series of which some numbers find a place in the catalogue, it would be requisite, if a reader undertook such an investigation, to write tickets for every series separately, to have all the numbers brought, and then to make note of the gaps. Such a work is rather the duty of the officials than of readers, but, as already stated, it would require a special inquiry, whether made by readers or by officials, to ascertain what is really the state of the British Museum library as to the literature of science.

It must be borne in mind that an important part, perhaps the most important part, of the literature to which a worker in science wants to refer, is that which is in the series of the different societies and government departments, and it is just in this that the British Museum is weakest, and in which it might be supposed a remedy might be most easily found. To fill up gaps of old standard works out of print is not very easy. Chances of sales of libraries must be carefully looked out for to effect this, but the current literature of societies and of departments is more easily secured.

An inquiry into the state of the scientific literature at the Museum, and the facilities for its use, might be advantageously directed under three distinct heads, each of which has an important bearing on meeting the requirements of those who wish to consult the collection :— I. As to the incompleteness of series.

2, As to the length of time that elapses between the publication of a number and its being obtainable at the Museum.

3. As to the method of cataloguing.

As regards (1) incompleteness of series, there is no reason to believe that it is confined to publications referring to any particular branches of science more than others. For example, to take a few cases at random, there are only three volumes of the reports of the state of the Brussels Observatory; there is only one part of the long series of reports on the health of the City of London; there are three volumes wanting of the Report of the Commissioners on the Sanitary Condition of the Labouring Population of Great Britain; the publications of the Geological Survey are very incomplete; there are none of the maps of the Water Supply Commission nor of the Coal Commission; and so on. To attempt to give a list of what is known to be wanting would not be of much use for the reason stated above, that nothing short of a full inquiry into the matter could make known what is the real state of affairs. When a question is asked as to why certain volumes are missing, there is always one reply given the publications of societies, home and foreign, are presented, and cannot be demanded, and as to the publications of Government departments, the Museum has no claim. If they happen to be sent to the Museum they are received, but if not, it would seem that under the existing system there is no help for it.

As regards (2) the length of time before a volume that is sent can be had for reference, it may be safely put at from one to two years. If a question is asked, how it is that such delays occur, a very general answer is that some societies are very irregular in sending their publications, but when such cases as this occur-that at the Museum a reader cannot now have a volume of the Bulletin of the Brussels Academy later than 1876, while at another public museum, the Patent Office Library in Southampton Buildings, he can have it up to June in this year-it seems to point rather to some feature in the administration of the Museum as the cause. Many cases of this kind might be quoted if it were required to establish the fact. It is, no doubt, a wise arrangement that novels and magazines that can be seen at any circulating