portant fact that the specific heat of water, hitherto supposed to be greater than that of any known substance except hydrogen, is really less than that of a mixture of water with methyl alcohol, in various proportions. This mixture, therefore, has a specific heat next to hydrogen.

Dahlander has communicated to the Swedish Academy of Sciences the results of his observations on the comparative rapidity with which heated solid bodies are cooled by immersion in various liquids. If the cooling power of water be taken as unity, that of alcohol is 0.58; of mercury, 2.07; of a concentrated solution of salt, 1.05; and of a concentrated solution of copper sulphate, 1.03. The rapidity of cooling increases with the increased temperature of the liquid.

Ditte has proposed to show the heat produced by chemical action, by adding 125 grams of water to 100 of boric acid. The heat produced is so great that an ingot of Darcet's fusible metal put into the mixture is completely fused in a few seconds.

Olivier has observed the curious phenomenon that if one end of a bar of steel fifteen millimeters square and seventy to eighty centimeters long be held against a revolving grindstone, one hand grasping the bar at its middle point, the other at the end, the middle portion remains quite cold, while the end farthest from the stone becomes too hot to touch. This appears to indicate the transference of energy along the bar in some other form than as heat.

Joule has made a new set of experiments with a view to increase the accuracy of his former determinations of the mechanical equivalent of heat. The result he has now arrived at, from the thermal effects of the friction of water, is, that taking the unit of heat as that which can raise a pound of water weighed in vacuo from 60° to 61° of the mercurial thermometer, its mechanical equivalent, reduced to the sealevel at the latitude of Greenwich, is 772.55 foot-pounds.

Aitken has described an apparatus for illustrating the conversion of the motion of heat possessed by matter at its normal temperature into work, in which he anticipated Preston's experiment. Two glass tubes entered a large bottle through its cork, one passing to the bottom, its upper end being drawn out to a fine jet. The other terminated just below the cork, where were attached some strips of blotting-paper.

Its upper end communicated with a closed vessel containing ether or some other volatile liquid. Some water being placed in the bottle, the ether was allowed to run on to the paper strips. Here evaporating, pressure was generated in the interior of the bottle, which threw a jet of the water to a considerable height, thus doing work without the addition of heat.

LIGHT.

1. Reflection and Refraction.

Masse has called attention to the explanation given many years ago by Arago and Babinet of the phenomena exhibited by the so-called Japanese magic mirrors. These mirrors are made of an alloy of copper and tin, are circular in form, are from one eighth to one tenth inch in thickness, and have Chinese or Japanese characters in strong relief on the back. When sunlight is reflected from them on a wall, the characters appear. Since these mirrors are cast, they are not equally dense in all parts; and hence in the operation of polishing they become concave or convex over the characters in relief, and these characters are therefore shown in the reflection. Notwithstanding this entirely sufficient explanation-proved a year or more ago by President Morton, of the Stevens Institute of Technology, by polishing the letters S. I. T. with a little rouge on his finger on an ordinary Japanese mirror showing no characters on reflection, and obtaining these letters in the reflected image-the magic mirror is brought forward every few years as a phenomenon entirely inexplicable by science.

Hoffmann has devised a new form of camera lucida, which seems to be an improvement upon the ordinary instrument. In place of a total reflection prism he uses two mirrors, one metallized, the other plain, placed at a fixed angle. The latter mirror transmits the rays coming from the pencil, and at the same time reflects a part of the rays coming from the object to be drawn, and which have already been reflected from the metallized mirror. A neutral tint glass or a set of lenses may be attached to the apparatus for special kinds of work.

Gariel has devised an ingenious abacus or chart by which the relations of the conjugate foci and the principal focus of

lenses represented by the usual formulas may be given graphically, thus facilitating materially the calculation of these values. Upon two perpendiculars erected at the extremities of a base-line are laid off equal spaces from this line, representing the principal focus. Diagonal lines drawn from the corners to each of these divisions represent the conjugate foci. It may be easily shown geometrically that these are in the ratio required. Direct measurement gives the value of any one of these quantities when the other two are previously known.

The same physicist has proposed a change in the manner of numbering glasses for spectacles. They are now numbered in terms of the radius of curvature, expressed in inches, the sign being plus or minus, according as the glass is convex or concave. The new method proposes to number them in terms of a new unit called a dioptric, which is the power of a convergent lens of one meter in focus. Since the power of a lens varies in the inverse ratio of the focal distance, the number of any lens in the new system is easily obtained by dividing one meter by the focal length of the lens expressed in meters and fractions of the meter. These two systems have a simple relation to each other.

Javal has described an apparatus for determining astigmatism, and at the same time the number and position of the axis of the correcting glass regarded as a cylindrical lens. Two vertical disks, movable around the same horizontal axis, carry each a system of lenses, those of the first disk being cylindrical, inserted in mountings toothed upon their borders and gearing with a toothed wheel so that they can be simultaneously revolved; those of the second disk are spherical. The first disk is used to determine astigmatism by viewing through its lenses a circle divided into sectors of 15° by radii; and having determined its direction and adjusted the axes of the lenses to it, rotation of the disk gives the focal adjustment sought. The second disk permits the myopia and hypermetropia to be corrected. The instrument is called an optometer.

2. Dispersion.

Glan has contrived a new photometer for comparing the intensities of lights which differ in color. It is composed of a collimator carrying two slits, placed the one above the oth

er.

Behind the collimating lens is a Wollaston prism with its refracting edge perpendicular to the slits. By this means the two beams emanating from the same slit are deviated in contrary directions with reference to the length of this slit; and, for any convenient distance of the collimator from the prism, the two mean rays, each proceeding from one of the slits, and polarized in rectangular azimuths, are exactly juxtaposed throughout their length. Traversing the spectrum apparatus, they give two spectra, in which the lines of the one will be the prolongation of those in the other. A Nicol prism is used in order to equalize the intensities to be compared, placed between the Wollaston prism and the spectrum apparatus. By a previous calibration with the solar spectrum, using the same light for both slits, when the extinction coefficients are of course equal, the unit of intensity is determined. The substance to be examined is placed between the Wollaston and the spectrum prism, and the equality azimuth for each color measured; the ratio of the extinction coefficients is then easily calculated.

Herschel has proposed a simple form of scale for pocket spectroscopes. The slit plate is removed, and in its place is placed a disk of copper-foil having a fine slit cut through it on one side of the centre, crossing which obliquely is a row of twenty holes, one eightieth of an inch apart, five being on one side of the slit and fifteen on the other, perforated in the copper, the upper and lower holes being level with the top and bottom of the slit. Viewed by sodium light the slit appears bright, and the punctures appear as a series of yellow dots. They are placed obliquely, so that their spectra in white light may not overlap and confuse their images. The curve corresponding to the spectroscope is then obtained in the usual way, and the value of the points obtained in wave-lengths.

Günther has described a simple method of reversing the metallic lines by means of an ordinary gas flame. Into the flame of a Bunsen burner, on the opposite side from the slit, a fine platinum wire is placed, bent at a right angle, the end being directed vertically upward. On the other side of the flame a second wire is placed, carrying a sodium salt, for example. Looking at the flame through a prism of low dispersive power, the eye sees first the sodium line as a bright yel

low band; and, second, the spectrum of the glowing wire, which is continuous except where crossed by the dark D line. Other metallic lines may be shown dark in the same way.

Duboscq has called attention to the appearance of relief obtained by projecting a spectrum with a direct - vision prism, using a cross, a V, a ring, or a spiral as the opening for the light. The illusion of relief is very strong in two rectangular planes, a cylindrical surface, etc. The effect is due to the great difference in intensity between the red rays forming the prominent parts of the image and the violet rays which form the more shaded portions of the figure.

After

Moser has examined the question whether each chemical compound has a spectrum of its own, as characteristic and definite for it as are those of the elements for them. giving a résumé of what had already been done, mainly with emission spectra, he goes on to describe his own experiments made with the absorption spectra of iodine and bromine as elements, and of nitrogen tetroxide as a compound. From the results obtained he justifies the conclusion that compounds have definite spectra, which are measurably independent of mass and temperature.

Stoney and Reynolds have studied the peculiar absorption spectrum of the vapor of chlorochromic oxide, which is of special interest because it supplies information as to the duration and character of the motion of the molecules of the vapor which produces it. The spectrum consists of lines of various intensities, but uniformly distributed. Of these 105 have been examined, and from their position it has been ascertained that they are all to be referred to one motion in the molecules of the gas, of which motion they are all harmonics or quasi-harmonics. On the first supposition this motion is repeated 810,000,000,000 times every second in each molecule. From the succession of intensities it is surmised that this motion is in some way related to that of a particular point in a violin string vibrating under the influence of the bow, i. e., a point nearly but not quite two fifths of the string from one of the ends.

Hurion has examined in the laboratory of Mascart the spectrum of iodine vapor, and shows that, as Le Roux had observed, this spectrum is produced by anomalous dispersion, the blue, contrary to the usual order, being less devi