ganic matter and ammonic carbonate and the further deposit of iron, increasing the weight and thus diminishing the percentages of the other components.

Rock No. 3 is evidently composed in part of a rock like No. 2, derived from bone and in part of fragments of shells and corals; these fragments, being almost destitute of phosphoric acid, aid in reducing the percentage. The iron seems to be the cementing material to a great extent in this case.

No. 4 is more recent and consists almost entirely of decayed shells and corals. The magnesia in Nos. 2 and 3 is evidently derived from the sea water, it having replaced part of the lime.

These specimens were selected from among a great number as being typical of the different deposits met with in the dredging. These deposits may have a very near relationship to the bone beds of South Carolina In the first place I think they show that tendency of calcic phosphate is not toward concentration, but in every case we find the more recent the bone or other material, the more abundant the phosphoric acid.

Now in the case of the bone beds we have, leaving out the silica, a substance of almost the same composition as bone with the addition of a considerable amount of iron; or, in other words, a composition nearly approaching that of No. 2, with the exception that there is less carbonate of lime.

The much larger amount of sand is accounted for by the nature of the surrounding coasts; in the case of the Gulf, the shores are mostly of coral, while in South Carolina they are siliceous. We also have the bones of land and fresh water animals mixed in with these deposits in such a good state of preservation that they can be recognized, and occasionally we find the imprint of shells from the beds below which may have been exposed at some points.

*

There is another characteristic described by C. U. Shepard, Jr., which applies equally well to the deposits from the Gulf. He says the surface of the nodules is glazed over and pitted with numerous small holes; this would answer very well for a description of rock No. 2.

Prof. C. U. Shepard's, Sr.t theory of the concentration of calcic phosphate in a mineral by the gradual removal of calcic carbonate by solution in carbonic acid does not seem tenable in view of the fact that calcic phosphate is also soluble in carbonic acid. The large amount of alkaline carbonates that would also be formed during such a process must also decompose to a considerable extent the calcic phosphate.

Cambridge, Dec. 1, 1870.

*This Journal, II, vol. xlvii.

This Journal, II, vol. xlvii, p. 338.

ART. XXVI.-Calorimetric Investigations; by R. BUNSEN.*

1. The Ice Calorimeter.

THE calorimetric methods hitherto in use are attended with the disadvantage that proportionately large quantities of the calorimetric fluid, as well as of the substance under investiga tion, must be employed, in order that the loss of heat which unavoidably attends the measurements may be so reduced that all corrections therefor may be small in comparison with the amount of heat to be measured. In the determination of specific heat, especially when the more exact of the methods previously in use are employed, satisfactory results can hardly be anticipated when the amount of material used in the experiments is less than from 10 to 40 grams. The preparation of the rarer substances, in a state of absolute purity, in such quantities often presents almost unsurmountable difficulties, and it is perhaps only on this account conceivable that we are not acquainted with the specific heats even of all the elements which have been isolated in a state of purity, although these determinations are of fundamental importance for the establishment of the atomic weights.

The instrument described in the following pages is designed to aid in overcoming this disadvantage. It is based on the principle of measuring the amount of ice melted by the communicated heat by means of the diminution in volume which this ice undergoes on melting.

The instrument, fig. 1, which was made at the glass blower's lamp, consists of an inner glass vessel, a, having the form of an ordinary test tube and melted into the cylindrical glass case b. From this case b issues the glass tube c, to whose upper extremity the iron head-piece d is cemented. The inner vessel a is filled from a to μ, the outer case b from 6 to 1, with previously boiled water; the remainder of the case b together with the tube c, is filled up to the height y with previously boiled mercury. In order to arrange the apparatus for use, a cylinder of ice enclosing the entire vessel a is produced in the case b, the whole apparatus is then surrounded with snow in a large vessel, and the calibrated scale tube s, which has been cemented into the cork with fine sealing-wax, is screwed down through the mercury of the head-piece d very tight into the opening of the tube c, whereby the scale tube fills itself with mercury. In order that the pressing in of the stopper may be unattended with danger for the rather fragile apparatus, the instrument is

* Translated for this Journal, with permission of the author, from Poggendorff's Annalen der Physik und Chemie, Bd. CXLI, S. 1, by Dr. G. E. MOORE, of San Francisco.

fastened on a heavy iron stand by means of a vice, whose jaws surround tightly the lower part of the iron head-piece d.

The amount of heat which a body evolves in cooling from a given temperature to 0° C., is determined by dropping it into the water in the vessel a and thereupon closing the vessel at o with a cork, to prevent any circulation of air. If the problem be the relative measurements of quantities of heat, as in the determination of specific heats, then the standard of comparison is directly afforded by the number of scale-divisions which the mercury thread has passed on its retreat. If the readings are rendered in an absolute measure, as for instance, in grams of melted ice, or in units of heat, as the unit in the following pages, (always that quantity of heat being understood which one gram of water at 0° C. absorbs in order to raise its temperature to 1° C.), then it is only necessary to multiply the readings on the scale with a constant which results from the following consideration.

A mercury thread measured in the scale tube, which has the temperature to and occupies T divisions of the tube after being corrected by the calibration table, weighs g grams. Let further the specific gravity of mercury at 0° C. be S,, its coefficient of expansion a, then is the volume v of a corrected division on the scale, measured in cubic centimeters,

For the instrument which I used, the values were:

If the specific gravity of ice at 0° C. be denoted by Se, the specific gravity of water at the same temperature by S, the weight of melted ice expressed in grams, which corresponds to the volume v, that is, to one scale division with p, then is

With regard to the specific gravity of ice we have many observations. The following comparison shows how little they agree

among themselves. For S

With so slight an agreement between these different observers, it appeared to me indispensable to determine, with greater exactitude than has hitherto been possible, the value of S. necessary for the calculation of the constant p. I employed for this purpose the following method, in which the sources of error which have made the previous determinations uncertain have been entirely avoided: Fig. 5 is a thick walled U-shaped tube, of difficultly fusible glass, which has been drawn out at a to a thick walled point. This is filled with mercury to b b, and both limbs are well boiled out, as is done with barometers. The point a is provided with a small rubber tube through which, by means of alternate warming and cooling of the air in the limb a b, distilled water, free from air, is allowed to enter above the mercury by b. If this water be boiled for half an hour and the rubber tube c be kept under the surface of water, which has likewise been kept in continual ebullition in a beaker glass, the space ab will, as soon as the boiling by b is discontinued, fill itself completely with perfectly airless water. The rubber tube c is now closed under water by means of a small glass stopper and the point by a inelted off, which may be easily and safely done with the ordinary non-luminous gas flame, without the aid of the blowpipe, when the part of the tube where it begins to narrow out into the point is so strongly heated that it is filled with steam instead of water. If the apparatus has been weighed before filling with water, and after the filling be weighed again together with the dry point, the weight of the water contained in the instrument will be obtained. The open limb is now completely filled with boiled-out mercury and especially, in order to prevent the adhesion of air bubbles to the glass walls, through a long capillary glass tube. If the apparatus be exposed in the open air to a temperature below 0C., an ice tube, corresponding to the glass tube, will be formed which at last closes in different places, and still contains water surrounded with ice. By the freezing of these last portions of water the ice already formed is exposed to a very high pressure, which may very considerably alter its specific gravity, may in fact even burst the glass tube eighty-atmosphere strong. In order to remove this disadvantage and permit the ice formation to take place, during its entire duration, under the same pressure, it is simply necessary to sink the whole instrument in sawdust and to expose only the upper part by a to air of a temperature below 0° C., after you have previously, in order to prevent the effects of abnormal lowering of the freezing point (Ueber-schruelzung), produced an ice mass by a, which is allowed to dwindle by melting to a small granule. The freezing then goes on very regularly from a downward to b and can be very conveniently regulated by letting the limb containing