remains after the disintegration of this material with barium chloride in a stream of chlorine gas, and which gives by melting with potassic hydrate a copious yield of potassic ruthenate. The oxide thrown down from this salt by fractional precipitation with carbonic acid was converted into chloride with hydrochloric acid, the aqueous solution of this latter salt was precipitated with hydrogen gas, and the metal, in the state of lustrous scales, was ignited in a current of hydrogen to remove every trace of oxide. It proved on examination to be entirely free from the other platinum metals. The specific heat found therefor accords, as might be expected, with the previously accepted atomic weight.

The calcium was reduced by electrolysis from melted calcium chloride. It formed small pale golden-yellow globules with a high metallic luster, which take in the air, very rapidly, a grayish tarnish. Before enclosing in the glass envelope, they were shaved bright in an atmosphere of carbonic acid. Upon examination, it proved to be almost entirely pure. From the atomic heat thus found, it may be concluded that the previously accepted atomic weight Ca-20 is the correct one, and need not, as has become necessary with the atomic weights of the alkali metals, be halved.

Under allotropic tin is understood the singular modification of this metal, which appears to form from ordinary tin by long continued very low temperature. The specimen investigated came from the large mass, altered by the cold of an unusually long and severe winter, on which Fritzsche had observed this remarkable allotropy, first described by him. The mass consisted of an aggregate of small undeterminate angular rods. loosely coherent in one direction, which crumbled by even slight pressure. This tin possesses, as I have convinced myself, a high degree of purity, contains no trace of antimony or arsenic, and dissolves completely in potassium trisulphid without leaving any basic metallic sulphids. The small rods of which it consists are not brittle but ductile like ordinary tin. The nonallotropic metal investigated was obtained by re-melting the

allotropic variety. Both modifications lead to nearly the same specific results.

The indium employed showed itself to be entirely free from tin, cadmium and iron. Oxidized by nitric acid and evaporated with sulphuric acid, it left upon solution in alcohol, no trace of plumbic sulphate. 10592 grams of the metal dissolved without loss in nitric acid, gave after evaporation and ignition 1.2825 grams indic oxide. If the latter, according to the previous views, be considered as consisting of equal atoms, there results from this determination the value for the atom of indium, In=37-92, which number agrees very closely with the one determined by Winkler, In=3781. This atomic weight, multiplied into the obtained specific heat Se, gives, however, for the atomic heat of indium the value, S. In=2·13, which does not agree with the others.

The previously accepted hypothesis, that indic oxide is constituted according to the formula InO, appears therefore no longer tenable. If the atomic weight be accepted as once and a half so great, In=567, then will the atomic heat be, S.In= 3-23, nearly equal therefore to that of the other elements. The atomic weight 567 removes the anomaly, that indic oxide which was previously considered monacid, is, according to its entire deportment, analogous to that class of trin-acid bases, which forms no alums. For the previously accepted formulæ, given in column I of the following table 8, must therefore be substituted the ones in column II.

Indic chloride, Ammonium indic chloride,

2NH, Cl, 3In €1, 2HO 2NH ̧¤l, In ̧¤1 ̧,2HO

4

The ammonium-indic-chloride, prepared by R. G. Meyer, and cited in this scheme, possesses, according to the new formula, an analogous composition with ammonium rhodie chloride, which likewise contains 2 atoms of water. Whether or no the altered atomic weight finds comfirmation in an isomorphism in these salts I have been unable to investigate; I shall, however, return to this subject in a research, to be published later, on rubidium. During this winter, for such experiments very unfavorable, I have been unable to make a trial of the instrument as to its fitness for the determination of the latent heats of melting. I

will only mention here that the latent heat of melting for water results already, with an accuracy which leaves nothing to be desired, from the experiments communicated in this work. According to table (2), one scale division of the calorimeter corresponds to grams of melted ice. The constant W, determined from equation (8), gives the number of scale divisions which correspond to one of the units of heat defined in the beginning of this treatise. One scale division of the colori

V S&S w
Sw-Se

meter corresponds therefore to

201

the latent heat of melting for water, then will

1

ᏓᎳ ..

give the weight of ice, expressed in grams, which corresponds to one scale division. There results, therefore, for 1, the equation 1

1=

or, according to equation (2), 7=

be substituted in the equation, there is obtained for the latent heat of melting of water 7:

In the mean,

80.01

80.04

80.025

The value found by other observers by the method of mixture is, according to Regnault, 79-4; Person, 800; Hess, 80.3. The heat of combustion of gases may be determined by means of the ice calorimeter with far greater accuracy than has been possible with any of the methods previously at our disposal. From the heat of combustion of hydrogen it follows with the aid of equation (4), that 10 cubic centimeters of this gas at 0° C. and 0.76 mercury pressure, would produce by their combustion with oxygen an oscillation of 453 scale divisions on the instrument employed in the experiments just described. It is therefore sufficient to burn a very small quantity of gas, and therefore one easily to be prepared in a state of purity, in the calorimeter, and to observe the oscillation on the scale thereby produced, in order to obtain directly the heat of combustion, expressed in units of heat, without any of the corrections, in part very uncertain, which were previously unavoidable.

ART. LIV.—On the Geology of the Delta, and the Mudlumps of the Passes of the Mississippi; by EUG. W. HILGARD.

[Continued from page 246.]

II. The Lower Delta and the Mudlumps.

A glance at the map shows that in descending the Mississippi from New Orleans, we find a narrow strip of land only to 3 miles wide, dividing the river from the waters of the Gulf; from the head of Oyster Bay opposite Pointe à la Hache (about half way between the city and the head of the Passes), down to the mouths. Such, at least, is the case on the left bank; on the right, the "neck" begins a few miles below Fort Jackson. Down to the forts, the aspect of the "Coast" is generally pretty much the same, where its original character has not been lost by cultivation or encroachment of the river. Nearest the river, and highest above water level, are the sandy "willow battures," where the willow, mingled with and occasionally replaced by the cottonwood, forms the predominant growth. yond lies a belt of woodland, timbered chiefly with live-oak, magnolia, and cottonwood, often deeply veiled with long-moss; this belt embraces the richest and most durable soils of the "Lower Coast," and is mostly occupied by magnificent plantations of sugar cane and orange orchards. Beyond these, loom in the distance the sombre-hued, moss-curtained denizens of the cypress swamp, their tops forming a level platform sharply defined against the horizon. Between the swamp and the water's edge, seaward, there usually intervenes a zone of reeds, with here and there a stunted cypress, bay, or candleberry bush, where the salt water has but slight access.

Be

While such is the general order of succession of these belts of vegetation where they coexist, either or both of the two middle ones may locally be absent. Such is always the case where the "neck" is very narrow, as happens below the forts. Thence to the mouths of the passes, the willow batture and the reed marsh alone, with few exceptions, form the barrier between the river and the sea; it is traversed by numerous small bayous, some of which are in great part the work of the duckhunters that supply the New Orleans market, and whose pursuit leads them to penetrate the marsh for the purpose of reaching the favorite resorts of their game. These bayous increase in frequency as we descend, and in approaching the mouths of the passes, the intervals between them become smaller, until they gradually become sheets of water dividing islands; and finally, just inside the bar, we have the latter resolved into numerous individual "mudlumps," dotting the surface of the sea, on both sides of the main channel.

Sir Charles Lyell remarks (Principles of Geology, 10th ed., p. 448), that the phenomenon of the mudlumps is without parallel, so far as known, in the delta of any other river. The same remark might, I think, appply to two other peculiarities, viz: the protrusion of the long neck of land into the Gulf; and the fact that, after failing to send out any branch of importance for a hundred miles the great river suddenly divides at one point into three widely divergent branches, the middle one of which (the South Pass), forming the direct continuation of the channel, is the smallest, and has long ceased to be navigable. Evidently, a strong extraneous obstacle alone could turn aside the powerful current, and permanently resist its erosive and undermining action. And now, the channel which carries the main current (the Southwest Pass), faithful to the old tradition, is rapidly pushing out into the Gulf its narrow bands of reedy marsh, without a branch of any consequence in ten miles from the head of the Passes to the light-house.

A glance at the coast lines, as well as at the intricate ramifications characterizing the deltas of the Rhine, the Po, the Danube, the Ganges, or the Hoang-Ho; or the broad inlets forming the mouths of the rivers of South America, will show the uniqueness of the Mississippi mouths; the Nile and the Lena alone exhibiting a general form at all analogous, yet very distinct in detail. For the islands off the Lena mouths are not "mudlumps;" and the tongue of land separating Lake Menzaleh from the Damietta branch of the Nile, is a mere sand-bar, exhibiting no analogy save that of form, with the remarkable "necks" of the Mississippi Passes.

It would be fair to infer, à priori, that some connection exists between the exceptional phenomenon of the mudlumps, and the exceptional form of the delta; and that such is really the case, can hardly be doubted upon a candid investigation of the facts. So far from being an unusual phenomenon, the mudlump-formation appears to constitute the normal mode of progression of the Mississippi mouths; not only at the present time, but for many ages past; perhaps ever since the broad flood of the Terrace epoch subsided into the present Mississippi.

The characteristic features of the mudlumps have successively been described and discussed by Sidell,* Forshey,† Chase, Beauregard and Latimer, Thomassy§ and Lyell. Yet as the phenomena are nowhere described in their entirety, I will here, as briefly as possible, recapitulate the important points.

* Report to Capt. Talcott, 1839, in Humphreys and Abbott's Report, App. A. MS. Report, 1850.

Report of the Board of Engineers for the Examination of the Mississippi; Con

gress. Doc. 1852-53.

Geologie Pratique de la Louisiane, 1860, Chap. VI.

Principles of Geology, 10th edition, 1868; vol. 1, p. 449.