It was held in Tomlin v. The Dubuque, Bellevue & Mississippi R. R. Co., 32 Iowa, 106, that a riparian i. riparian proprietor upon the Mississippi river owns the fee ¡an owner.81 of the soil only to ordinary high water mark. In that case, it appearing that the defendant’s road had been built below ordinary high water mark, it was held that the plaintiff’, a proprietor upon the river, could not recover for the land so occupied. In this case the correctness of that decision is not questioned, but the fact is questioned as. to the land in controversy being below high water mark. The land is upon the Mississippi river, at the city of Lansing. It is nearly level but sloping gently towards the river. For a time within the observation of the witnesses, covering a period of fifteen or twenty years, the water at its highest stage has covered the entire land, at least for a day or two every year, with the exception of one or two years. The court instructed the jury in these words: “ The Mississippi river periodically rises and falls, and these rises, as shown by the undisputed evidence, occur usually in June, September and sometimes October. These rises are characterized as high water, and you are instructed that the highest point to which the river ordinarily rises at these times of high water is high water mark.”

2>__. higll water mark, The giving of this instruction is assigned as error. It is insisted that it does not contain the true definition of high water mark within the meaning of the law which makes tliat line the boundary between the property which belongs to the riparian proprietor and that which belongs to the public. No definition of high water mark has, we think, [*372] ever been given by this court. In Musser v. Hershey, 42 Iowa, 381, Day, J., said: “ It is the settled doctrine of this court that a riparian proprietor upon a navigable stream owns only to high water mark, that is, only to the edge of the bank.” This is undoubtedly correct, and there only remains to be determined what, precisely, is the bank. The ordinary idea of a river bank is that portion of the earth which confines the water in its channel. It adjoins the bed of the river, and belongs to the riparian proprietor. The bed, if the stream is navigable, belongs to the public. While the banks are supposed to confine the water in its channel, they are sometimes in freshets overflowed. But they are not the less defined because they are sometimes overflowed. In determining the boundary line between the bank and bed of a stream freshets are not accounted. Upon this point there has been an express adjudication. In Howard v. Ingersoll, 13 Howard, 331, a question arose as to where the boundary line was between Alabama and Georgia. It, had been established upon the western bank of the Chattahoochee river, but the parties were not agreed as to what constituted the bank. It was held that Georgia did not include land on the Alabama side of the river which was covered with water only in time of freshets.

The correctness of the decision is entirely obvious. No one freshet water line could be taken, and by a succession of freshets no line is particularly indicated. Besides, freshets are temporary and exceptional in their character. The banks of a river, then, are not to be discovered from the limits of the water in time of freshets. Now while the term freshet is not, we think, usually applied to the periodical rises of the Mississippi river, it is not, as it seems to us, altogether inapplicable. The greatest rises of the Mississippi river occur in spring or early summer. Like the spring rises of small rivers they are caused by melting snows, or by rains and melting snows com- ' bined. They are, then, of the same nature. They come later and stay longer, to be sure, because their waters are gathered from ho inconsiderable portion of a continent. They are, however, temporary and uncertain. If, then, the banks of small rivers are not to be regarded as co-ordinate in all places [*373] with the limits of' the water in times of ordinary freshets, we do not think that the banks of the Mississippi river should, be regarded as co-ordinate in all places with the limits of the water at the time of the ordinary great annual rises. To show howr improper this would be it is not necessary to appeal to general observation. We may confine ourselves strictly within the evidence in the case. It is shown that at Lansing, where the land in question is, a creek flows into the Mississippi river; At the spring rises Mississippi water overflows the creek valley to a distance of nearly a mile up the creek. Yet this valley is in no proper sense Mississippi -river bed. The evidence shows that it was formerly covered with timber, and is used now as a meadow, on which a large quantity of hay is cut annually. Whether the land in question can be regarded as a part .of this valley does not distinctly appear, but it does appear that it was formerly covered with grass and sod, and is now used, where not occupied by the railroad, for piling lumber on a considerable portion of the year. It seems to us, therefore, that the lino reached ordinarily by the great annual rises, including, as it does in places, large quantities of land which are valuable for agriculture, is not the true line to determine where the rights of the riparian proprietor in the soil end, and those of the public begin. The other line suggested as the true line is that which the river impresses upon the soil as the limit of its dominion. According to this theory, high watermark is to be determined, not from human records, but from the record which the river makes for itself. It is to be regarded literally as a water mark. Dictionary definitions are not wanting to support this view. See Worcester’s Unabridged. We do not, however, attach much importance to the meaning which either dictionaries or popular usage may have attached to the words. The line along a navigable river between that which belongs to the public and that which belongs to the riparian proprietors, which line is sometimes called high water mark, is to be determined less by the meaning of high water mark, as established by dictionaries or common parlance, than by the nature of the case. The term, high water mark, is doubtless sometimes used with reference to the highest rises. [*374] It is necessary sometimes in the exigencies of business to talco into account the highest rises; but in determining what belongs to the public we have to determine what properly belongs to the river. Now we think, as is insisted by the appellant, that that only belongs, to the river in any proper sense, where its occupancy has been so long continued as to leave a permanent impression of its domain. The impression may, in places, be indistinct. Yegetation may, in places, dispute the dominion of the river with doubtful issue; but the line thus indicated is without doubt the true one, if it can be found, and whatever difficulty there may be in finding it, which we think ordinarily would not be great, no other line has been suggested which appears more easily discoverable, and certainly none which appears even approximately as satisfactory. High water mark, then, as the line between the riparian proprietor and the public, is to be regarded as coordinate with the limit of the river bed. Whatever difficulty there may be in determining it in places, this doubtless may be said: What the river does not occupy long enough to wrest from vegetation, so far as to destroy its value for agriculture, is not river bed. It was so held, substantially, in Howard v. Ingersoll, above cited, where the bank of a river was to be determined, which is only another form of stating the question in this case. Musser v. Hershey, above cited. We think, therefore, that the instruction of the court below to the effect that high water mark is the line reached by the ordinary great annual rises, regardless of the character of the lands subject at those times to be overflowed, cannot be sustained, and the judgment must be

Reversed.