Peru and others. The thallic alcohol of Lamy* is far too costly. The solution of silico-tungstate of sodium,t of metatungstate of sodium ‡ and of soluble tungstic acid § as obtained by dialysis, all promised good results from their extraordinary densities, but all proved difficult to prepare in a state of purity and extremely easy of decomposition.

A solution of phosphorus in bisulphid of carbon has, according to Messrs. Dale and Gladstone, a dispersion of 0.225,¶ or nearly one and a half times as great as bisulphid of carbon alone, but becomes turbid on exposure to sunlight from the formation of amorphous phosphorus. It occurred to me, that, by dissolving sulphur with the phosphorus, the formation of amorphous phosphorus might be prevented, and experiment proved that this was the case. The solution, as thus obtained, has a pale yellow color, but is perfectly clear and undergoes no change by the action of light even when long continued. I have been in the habit of preparing it by dissolving one part of dry flowers of sulphur and two parts of phosphorus, in four or five parts of bisulphid of carbon, and filtering the liquid through a well dried ribbed paper filter, which is easily done. The refractive and dispersive power of the solution will of course vary with the quantity of phosphorus and sulphur dissolved. By a gentle heat the whole, or nearly the whole, of the bisulphid of carbon may be driven off, a liquid compound of sulphur and phosphorus remaining which has so high a mean refractive power that it cannot be employed with prisms having a refractive angle of more than 45°-50°. The same end may, however, also be attained by continually adding phosphorus to a saturated solution of sulphur in bisulphid of carbon, in which phosphorus appears to be soluble without limit.

With a strong and probably saturated solution of sulphur in CS, the angle between Li and D was 0° 50' 10". When phosphorus was added the angle was 2° 25' 30", the refracting angle of the prism being 60°. In this last case the angle between Na, and Na, was 0° 2' 20". The spectrum was perfectly clear, the definition of the dark lines leaving nothing to be desired. In consequence, however, of the yellow color of the liquid there is always a marked absorption of the violet end of the spectrum. In working with the above described solution I have employed hollow glass prisms with refracting plates cemented on with a mixture of glue and molasses. These were found to *Ann. de Chimie et de Physique, 4th series, vol. iii, p. 373. Ann. de Chimie et de Physique, 4th series, vol. iii, p. 5. Scheibler in Journal für prakt. Chimie, lxxxiii, p 273. Graham in Journal of the Chemical Society, vol. ii, p. 318

L. and E. Phil. Mag., vol. xviii, p. 30.

The number 0.225 is the difference between the indices for the extreme red and violet rays.

be perfectly tight and to last for months without change. The great disadvantage in the use of a solution of sulphur and phosphorus consists in the danger of breaking the prisms; the liquid taking fire spontaneously when it has been a few seconds in contact with any porous material like wood or paper. On the other hand, however, the large quantity of sulphur present prevents the fire from spreading, a drop placed upon a piece of wood leaving after combustion only a charred spot. When not in use the prisms should be kept in an iron pot with a tight cover. In this manner I have employed and preserved two during a long and hot summer. The viscid, or rather oily, nature of the solution serves to prevent, to a great extent, the formation of ascending and descending currents from slight changes of temperature, and when the prisms are well shaken before use the definition remains perfect for a long time. In my spectroscope the prisms rest upon a plate of glass instead of upon one of metal.

§ 3.

On an advantageous form of apparatus for the study of the absorption of light in colored liquids.

In his examination of the spectra of colored fluids, Mr. Gladstone employed a hollow wedge of glass, the two refracting surfaces of which made with each other an acute angle. The wedge was filled with the liquid to be studied and so placed that the refracting edge of the analyzing prism was at right angles to the line of intersection of the two faces of the wedge. In this manner a beam of light was obtained which represented different thicknesses of the absorbing liquid, and the resulting spectrum became a complete absorption diagram. In using this apparatus I found the angular deviation produced by the wedge a source of considerable inconvenience. In addition it is easy to see that the wedge itself produces a certain amount of chromatic dispersion. To remedy these defects and at the same time retain the advantages of the method, I have devised what may be termed a double wedge. Two hollow wedges, of glass or metal, are placed together in such a manner that the first and last surfaces of the bounding plates of glass are parallel. The two wedges are separated by a single plate of glass with parallel surfaces. The base of each wedge, or acute-angled prism, is bored for the insertion of a cork or stopper. The construction of the apparatus will be readily understood from the diagram. In using it with a colored aqueous solution one of the hollow prisms, or wedges, is to be filled with distilled water, the other with the aqueous solution to be examined. The incident beam of sunlight is then allowed to fall perpendicularly upon

either surface, and the slit of the spectroscope is placed so as to be perpendicular to the two lines of intersection of the three refracting plates of the double wedge. In this manner a complete absorption diagram is obtained by the various thicknesses of the liquid examined. For substances soluble in alcohol, ether, &c., one of the hollow wedges must be filled with the colorless solvent, whatever be its nature. By this means all angular deviation and prismatic dispersion are avoided, as the coloring matter does not sensibly change the refractive power of the liquid in which it is dissolved, and the incident beam passes through without change in direction. In my apparatus the wedges have acute angles of about 15°. This I find to be quite sufficient for most purposes, as it is easy to increase or diminish the quantity of substance dissolved. When it is wished to examine the absorption produced by a definite thickness of liquid or by different thicknesses in succession, the double wedge is to be so placed that the slit of the spectroscope shall be parallel to the lines of intersection of the faces of the wedges. By moving the wedge to one side or the other all thicknesses of liquid, from 0 to the maximum, obtainable with the apparatus used, may be successively examined.

§ 4.

On tests for the perfection and parallelism of plane surfaces of glass. When a plano-convex lens of long radius of curvature is placed upon a plane surface of glass and the system is illuminated by an obliquely incident beam of monochromatic light as, for example, by a sodium flame, the well known phenomenon of Newton's rings is observed with remarkable distinctness and perfection of definition. The symmetry of the rings will depend, in part, on the perfection of figure of the lens, in part on that of the plane surface. An extremely minute deviation from a perfect plane will produce a marked distortion of the circular figure of the ring nearest the center. That this distortion is, or is not due to the lens may be determined by rotating the lens. round its optical axis normal to the plane. No change of figure will be seen if the lens is perfect in form and the inequality is in the plane surface only. Different parts of the plane surface may of course be tested in succession, by moving the lens from point to point, and if necessary the rings may be observed with a telescope.

Prof. Rood, of New York, has suggested for the observation of Newton's rings a method which permits of the employment of a lens of comparatively small radius of curvature and a microscope. In his arrangement the lens and plate of glass are placed upon the stage of the microscope, the light from beneath being cut off, and monochromatic light is then thrown down

upon the system by means of a plate of glass with parallel surfaces inclined to the axis of the microscope at a convenient angle and placed between the objective and the plano-convex lens. In this manner the rings are seen with great distinctness and beauty, and the arrangement is particularly compact and convenient.

The interference bands of Talbot afford a method not merely of observing with great precision the inequalities of surface and want of parallelism of the faces of plates of glass, but also of photographing these defects and obtaining a permanent chart of the glass which may be of material assistance in correcting its figure. It is only necessary for this purpose to place the glass to be examined near to the object glass of the collimator and perpendicular to its axis, so as to intercept that half of the bundle of parallel rays which falls upon the first surface of the first prism nearest its refracting edge. If the plate has perfectly plane and parallel surfaces the interference bands will be sharply defined and parallel in the whole field of view. The slightest inequality of surface, or inclination of the faces, will produce curvature or distortion of the bands, and, if the eyepiece of the observing telescope be removed, the image may be received on a sensitive plate and photographed. The number of prisms to be employed in a particular case will depend upon the thickness of the plate of glass examined and, in general terms, upon its dispersive power. For a piece of French plate glass four millimeters in thickness, two bisulphid of carbon prisms of 60° must be used to produce a sufficient separation of the interference bands to enable them to be seen distinctly. More prisms must be used for thicker plates and in this way a limit is soon reached at which the method ceases to be applicable. Cambridge, May 1st, 1870.

ART. VIIL-On the Occurrence of a Peat bed beneath Deposits of Drift in Southwestern Ohio.

A BED of peat has lately been found one mile east of Germantown, Montgomery county, Ohio, and twelve miles west of south from Dayton-in the occurrence and connections of which there are several facts of unusual interest.

It lies in, and directly above, the channel of Twin creek, a tributary of the Miami river. The general course of the creek is southeasterly, but just above the point where the peat bed is exposed, it has made a sudden change in direction from east to west of south. Its northern and eastern banks for of a mile in each direction from the point of deflection, are precipitous walls of stratified clay and gravel, from 50 to 100 feet in thick

ness; kept nearly vertical by the constant undermining action of the stream.

Beneath these heavy deposits and occupying 40 rods of the east bank of the creek, the peat bed is found, varying in thickness, in different portions of its extent, from 12 to 20 feet. The amount of the bed that is exposed depends upon the stage of water in the stream. The stream is bedded for 10 or 15 rods upon the peat, but in deeper portions of the channel, upon the eastern bank, an underlying formation of gravel can be detected. The uppermost layers of the peat contain undecomposed sphagnous mosses, grasses and sedges, but in other portions of the bed, the vegetable structure is generally indistinct, with the exception of abundant fragments of coniferous wood, which in many instances can be identified as Red Cedar (Juniperus virginianus). At the southern extremity of the bed in particular, there is a great accumulation of wood, in trunks, roots, branches and twigs, much of which has been flattened by the pressure of the 80 feet of clay and gravel that overlie it. Branches that were originally two inches in diameter now afford lenticular sections with no more than a inch for the shorter axis, while many of the smaller stems have been compressed into ribbons. The berries of the cedar are abundant in the upper layers of the peat. At a point mile higher up the stream, trunks of cedar nearly two feet in diameter, have been taken from beneath these same drift beds and turned to account for fencing posts.

There are indications that the peat bed has a considerable extent to the northward and eastward. A bed of "black earth was found underlying clay and gravel in digging a well 11⁄2 miles east of this locality. The bed occurred at a depth of 30 feet and was itself from 10 to 15 feet in thickness. The waters of springs in the same neighborhood are discolored, as if by contact with such deposits.

It may be added in this connection that there is a large amount of wood buried beneath the drift throughout this region generally. It is not a circumstance of infrequent occurrence to meet with it in the digging of wells. There is scarcely a square mile in the thickly settled portions of the adjacent country in which instances of this kind can not be found, and three instances are on record within the limits of a single village.

The wood is in great part coniferous, but not exclusively so; for according to the testimony of intelligent and observing, practical men who deem themselves entirely competent to give a judgment in the case, ash, hickory and sycamore, together with grape-vines and beech leaves, have been found covered with drift deposits.

A stratum of soil, one or two feet in thickness is often associated with these vegetable remains. The soil and the wood