the end are either in the form of small pebbles or sand, the bar may be sharply turned in the direction away from that of the prevailing wind.

This turning point is apparently reached and caused whenever the force of the waves which reach the end of the bar, from the direction in which the wave movement is most powerful, is able to remove the materials which the waves from other directions are driving to the end. That is to say, there comes a balance of supply and ability to remove, when the further out-building of the bar must cease. Then there is an abrupt bend.

This is very well illustrated in the Barrachois hook (Pl. II, Fig. 2, and Pl. IV, Fig. 1), on the eastern side of the Bras d'Or, about midway between North Sydney and Grand Narrows. It is plainly visible from the train and is a remarkable feature, being set in the dark waters of a deep and narrow arm of the sea extending between two high hills. It reaches directly out from the shore, turned at an angle of about 110° with the coast, which is the normal angle of the bars forming the arms of the cuspate forelands in this region. At a distance of about one hundred or one hundred and twenty-five yards from the shore it bends abruptly at an angle of 60° or 70° (Pl. II, Fig. 2) and extends nearly parallel to the coast, until near its end, at a distance of two hundred or two hundred and fifty yards from the first elbow, it turns again toward the shore and then back in the opposite direction, forming a very perfect hook.

This hook is built out into water which is sixty feet deep, and hence it represents a distinct bank, which has been piled along this line as the result of remarkably permanent conditions. It is due to no mere chance, but to the operation of some very distinct forces, operating for definite causes, in perfectly definite ways for a considerable time. In this case the cause for the bend seems evident. It is a place where the along-shore waves from the south (left) can transport the diminished supply of materials which reach the end of the bar, and do this so speedily that the bar can reach no further outward, but must turn. This change comes with such force and persistence that it is posible to build the turned end for more than two hundred yards in a depth of sixty feet of

water.* At the end of this there is a hook, no doubt because the weaker winds from the north (right), which have only a very narrow and protected valley in which to blow, are able to turn the supply backward toward the south, under the protection of the bar that has grown northward. In other words, the supply driven along this bent end is not sufficient to build the bar further north in opposition to the tendency of the very weak opposing waves to drive the materials backward.

In the Bras d'Or there seems to be evidence that under certain conditions a hook will change to a cuspate foreland. This has been partly stated above in the explanation of the Barrachois hook. There, in an enclosed arm of the sea, a bar that is growing outward is bent, not directly back toward the coast, but up the channel before the stronger waves. In the more open waters of the Bras d'Or, heavy waves often come directly upon the shore, or else reach it at a high angle. Such action will naturally tend to interfere with the outward growth of a bar which is being constructed by the drift of materials before the waves that reach the shore diagonally; and when, finally, the outward growth of this bar becomes slow, because of diminished supply, the on-shore wind waves at first produce a hook at the end, and then commence to build a bar toward the shore (Pl. IV, Fig. 2). There are several cases of this nature in which a bar, proceeding outward for some distance, turns landward and produces a partial hook, which is apparently an incomplete V-shaped cuspate foreland. In other cases a small bar is starting out from the land to meet the recurved end of that arm which has developed most notably (Pl. III, Fig. 3). Then we have a nearly complete cuspate foreland, which would be complete if the end of the recurved bar, and that of the small one extending out from the shore, were finally united. Such, I believe, is the origin of many, and perhaps all, of the cuspate forelands of the Bras d'Or; and there seems to be nearly every gradation between the single bar, the bar hooked at the end, and the double V-shaped cuspate foreland. The nearly soldered *The water is shallower on the inner side of the hook, but even here is deep.

It happens that the two illustrations of these gradations here reproduced are from bars of a very similar nature, whose basal ends are on islands which the bars are engaged in tying to the mainland. Here cuspate forelands are built, though they are made to include off-shore islands.

southern arm of the Sydney harbor cusp described above appears to be an instance of an almost closed foreland of this origin.

It would be unwise to extend the results of this study so far as to make the generalization that all cuspate forelands are derived by means similar to this; but before assuming that they are not so formed it seems that we should have distinct evidence that they are not. Is there any place where the peculiar eddying of tidal currents is known to be at work building these shore features? Have we any distinct evidence of the existence of back-set eddies from the Gulf stream, whose power is sufficient to move and guide the distribution of materials? Such eddies, to cause the prominent cusps of the Florida coast should, it seems, be very noticeable. It is possible that this evidence has been brought forward; but if so I have never seen it.

As small spits and cusps may be made by waves in small, enclosed and nearly land-locked bodies of water, where proper conditions exist, so it seems possible that larger cusps, even those as large as Hatteras, may be made where the supply of material is more rapid and of finer texture, while the waves are greater and the shore currents of wind origin more powerful and therefore able to move the fine materials. This certainly seems a possible explanation; and, even though it be not the correct one, it should at least be discussed as an hypothesis, and the evidence against it definitely brought forward, if there is such opposing evidence.

OSAGE VS. AUGUSTA.

By STUART WELLER, University of Chicago.

In a recent article in this magazine* entitled, “Use of the Term Augusta in Geology," C. R. Keyes defends the use of the term Augusta (Keyes) and its displacement of the older term Osage (Williams) as the name of one of the major subdivisions of the Mississippian series.

The science of paleontologic geology is a historical science. It is something far more than the mere classification of rock *Am. Geol., vol. XXI., p. 229, April, 1898.

strata. In his investigations of the Lower Carboniferous faunas of the Mississippi valley Prof. H. S. Williams detected the records of three distinct chapters in the history of the continental interior, and to the middle one of these chapters he gave the name of Osage, the rock strata containing the fossil records of the chapter being designated as the "Osage group." It was recognized that the faunas of the Burlington and Keokuk limestones were a unit in themselves to such an extent that they should not be divided into several distinct groups, as had generally been done by earlier geologists. There were life changes in progress during the time of deposition of these limestones, but they were internal, evolution changes, not dependent upon profound physical changes or upon the immigration of new forms of life. The records of the Osage chapter in our continental history were found to be best expressed in these faunas of the Burlington and Keokuk limestones.

Mr. Keyes' contention that the Osage and Augusta are not synonymous, because Williams failed to include a portion of the so-called Warsaw group of older authors in his division, is based upon unessentials and is not tenable. On the same ground, if some future investigator were to find that a small portion of the Chouteau limestone would be better placed with the superjacent strata, it would be perfectly legitimate for him to ignore both the names Osage and Augusta, and to propose still a third, and so still further increase the confusion.

As a matter of fact, Williams did recognize that one portion of the Warsaw faunas were of Osage age and another of St. Louis age; and he showed, before Mr. Keyes, that the fauna of the Warsaw group had no separate place in a natural classification of the Mississippian faunas. He says:* "The faunas of the Chester, St. Louis and most [ not all] of those referred to the Warsaw formations are paleontologically more closely allied than they are to the faunas of the Keokuk and Burlington." It is thus seen that he carefully excludes a part of the Warsaw faunas, and clearly implies their relationship. with the Osage faunas. That Mr. Keyes and Prof. Williams are in accord in this regard is shown by the following statement by Mr. Keyes: "In a majority of cases the so-called *U. S. Geol. Surv., Bull. 80, p. 169, 1891.

†Geol. Surv. Iowa, vol. I., p. 70, 1893.

Warsaw is clearly the lower part of the St. Louis limestone." Prof. Williams' classification of the strata of the Mississippian series is expressed in tabular form as follows:*

In a

In this way the name Osage was established in 1891. paper bearing the date of June, 1892, Mr. Keyes accepted without question and with no intimation that the name was used "provisionally," the name Osage and used it in a classification of the strata of the Mississippian section, giving it the following definition:

OSAGE LIMESTONES.

Definition and general Relations.-From a purely paleontological standpoint, the advisability of including the Burlington and Keokuk limestones under a single name was pointed out several years ago. For this long needed term Williams has proposed the name "Osage."

At that time nothing was said about the name being proposed under the "misconception" that the Burlington and Keokuk faunas were mingled in southwestern Missouri,-a "misconception," which, if it ever did exist, would have no bearing on the subject. Neither did the supposed fact that Williams had entirely excluded the "Warsaw" from his Osage group seem to be a serious objection to Mr. Keyes, at this time, against using the name. The following tabulation of his classification of these strata is given by Mr. Keyes in this paper:

*U. S. Geol. Surv., Bull. 80, p. 169, 1891.

† Bull. Geol. Soc. Am., vol. 3, p. 283.