without altering the direction of the line of vision. This is effected by interposing between the object and the objective a glass tank with plane parallel top and bottom, in which some liquid is placed. The height of the surface is varied by means of a suitable plunger.

Gariel has modified the ordinary projection instrument known as the phenakisticope, removing the disk which carries the figures, and doubling the number of lenses in order to give more light on the screen. Placing now a fixed object in the apparatus, as a round opening in a screen, for example, its image may be made to persist by sufficiently rapid rotation. Using two openings of different colors, their resultant may be shown. With a prism, the spectrum given when the apparatus is at rest becomes white when in motion. The apparatus shows the manometric flames very well.

Terquem has given the composition of a varnish which may be spread on glass without injuring its transparency, but which will enable it to take ordinary drawing-ink, so that various devices for the lantern may be drawn upon it. The varnish in question is composed of alcohol, 100 cubic centimeters; mastic, 7 grams; sandarac, 3 grams. It has been the practice for some years in this country to prepare plates for receiving India-ink by flowing them with a dilute solution of gelatin, one ounce to the pint of water. Very fine drawings can be made on it, and it is easily removed by

water.

3. Dispersion.

Becquerel has examined carefully the ultra-red portions of the spectrum, using the principle of phosphorescence for the observation of this region.

Campbell has devised a double slit which he has used with good results for measuring the distances between the lines in the spectrum, and which he finds of great service in cases where the illumination is so slight as to prevent the use of the micrometer. One slit is above the other, the upper one being movable at right angles to its length by a micrometer screw of 200 threads to the inch, the graduated head of which is capable of indicating one five-millionth of an inch in the motion of the slit. If now a reading of the micrometer be taken when the slits are superposed and form

one continuous line, and a second reading when any given line has been superposed upon any other line at a moderate distance from it, the difference between these readings will enable us at once to ascertain the distance between the lines, if the micrometer be calibrated in terms of the spectrum as seen in the observing telescope.

Thalén has published the results of a joint investigation made by Angström and himself (but not published till after the former's death) upon the spectra of the metalloids, an excellent abstract of which by Schuster appears in Nature. They believe it extremely improbable that any lines present in a spectrum at a lower can disappear at a higher temperature. The electric spark and the actions it may cause are carefully studied, and applied to elucidate the carbon spec

trum.

Boisbaudran has given, in a plate illustrating an extended paper on gallium, an excellent representation of the spectrum of this metal, together with all the other elemental lines which are found in the same vicinity.

Cleyden and Heycock have discovered the significant fact that the spectrum of indium, which is obtained by taking the electric spark between electrodes made of that metal, is quite different from that ordinarily given. Instead of three lines only, they found sixteen under the above circumstances. The lines usually figured are two in the indigo and one in the violet; they have wave-lengths, as given by Thalén, of 4532, 4509, and 4101 tenth-meters; the authors believe that the middle one should be 4510. The new lines have wavelengths as follows: 6906, 6193, 6114, 6095, 5922, 5905, 5862, 5820, 5722, 5644, 5250, 4680, 4656, 4638, 4510, and 4101 tenth-meters. The first line is remarkable, since potassium, strontium, and antimony only give less refrangible lines.

Henry Draper has discovered the remarkable fact that oxygen exists in the sun, and that it and probably also the other metalloids show their presence in the sun-spectrum by bright instead of dark lines. By means of photography he has produced upon a single plate the solar spectrum from just above G to below H, and the spectrum of air ignited by a powerful spark between iron and aluminum terminals. Since the lines of iron in the latter spectrum coincide exactly with known iron lines in that of the sun, the non-shifting of

the plate is proved. The oxygen lines in the air spectrum, to the number of a dozen or more, coincide accurately with bright solar lines, every peculiarity in the shading or grouping of the one being reproduced in the other. The demonstration is complete, and will materially modify existing views as to the solar constitution. The discovery is the most important made in solar physics since that of Kirchhoff in 1860.

Young has made careful measurements of certain lines in the solar spectrum, observed alternately upon the eastern and western limbs of the sun, with a view to test the question of the effect of the motion of a luminous body upon the wave-length of the light which it emits. Using spectra of the sixth and eighth orders produced by a Rutherfurd grating of 8640 lines, the overlapping spectra being separated from each other by a glass prism of 45° placed between the grating and the object-glass, with its refracting edge perpendicular to the lines, the author succeeded in observing a difference in the position of the two D lines on the two edges, which, admitting Doppler's theory, would give a velocity of 1.42 miles per second for a point on the sun's surface. As direct observation gives but 1.25, Young inclines to the supposition that this difference proves that the solar atmosphere really sweeps forward over the underlying surface. Careful measurement of a line in the B group, which is due to atmospheric absorption, gave no displacement, as was to be expected.

Duboscq has contrived some simple and novel optical projection experiments. By means of a lens, the image of a small round hole is thrown on a screen; between the lens and the screen an Amici direct-vision prism is placed, capable of rotation about its axis. On rotating, the spectrum is circular-red within, violet without. Using polarized light and a double-image rotating prism, the extraordinary image describes around the ordinary as a centre a luminous ring, crossed by a black line parallel to the plane of polarization. The experiment may be varied by interposing a plate of quartz.

Egoroff has described a differential electro-actinometer devised in order to determine the coefficients of absorption of the ultra-violet rays by different bodies. For this pur

pose he uses two Becquerel actinometers, one opposed to the other. The strength of the current produced appears to be in exact proportion to the intensity of the light.

Bezold has contrived a convenient method for comparing pigment colors with spectrum colors. It consists simply in replacing the scale of an ordinary spectroscope by a vertical slit a millimeter wide, before which the color to be studied is placed. The eye sees then the spectrum color by refraction and the pigment color by reflection, and by a movement of either slit the two colors may be brought into exact coincidence.

Hohngren, at the request of the direction of the Swedish railway between Upsala and Gefle, has examined the entire staff of officials with reference to color-blindness. Out of the 266 persons examined, no fewer than eighteen were found who could not distinguish color, and were therefore utterly useless and unfit for railway service. An investigation of this sort on some of the leading American railways would undoubtedly be of service.

Thompson has submitted to the test of experiment the common impression that objects appear brighter when seen with two eyes than with one, using an ingenious apparatus by which two beams of light (one polarized, the other unpolarized) give to two Nicol prisms, one in front of each eye, the same quantity of light. It appears from these experiments that light is more powerful in producing an effect when concentrated upon one eye than when equally distributed to the two; but the light so concentrated on one eye does not produce the sensation of twice as much illumination as the half of the light viewed by both eyes at once.

Rood has called attention to and confirmed an observation made by Tait which bears on Young's subjective color theory. Tait observed that on awaking from a feverish sleep a lamp flame assumed a red color, lasting for a second. Rood first noticed the same result twenty years ago, in Munich, on recovering from anesthesia by chloroform, when the face of the operator appeared ruddy and his hair purplish red. He now has observed a chronic condition of the same sort, lasting for a couple of weeks, during convalescence from typhoid fever. White objects appeared orange yellow. On Young's theory this result is explained by supposing that the nerve

fibrils of the retina, which are sensitive to red, resume their functions soonest. Hence the author infers that the apparatus in the eye for the reception of waves of medium length is more liable to be overstrained than that designed for waves of greater or lesser length.

Kühne has made some new and remarkable experiments in optography. Following out the suggestion of Boll, that the retina of an animal kept in the dark for a long time is purplish red, the color being bleached by daylight, he has succeeded in fixing upon the retina the image of objects seen by the animal before death. To repeat the experiment, the animal, after being kept for a long time in the dark, is decapitated, and each of the eyes exposed in turn to a bright objectas, for example, the skylight of the laboratory. The retina are removed from the eye in presence of sodium light, and placed in a five-per-cent. solution of alum. After becoming hard, they may be separated from the optic nerve and inverted. Upon a beautiful rose-colored field a brilliant and sharply defined image of the skylight appears, showing even the sashes. In one of Kühne's experiments a second image appeared, to his surprise, but it was due to the second skylight with which the laboratory was lighted.

Carey Lea has studied the sensitiveness to light of various salts of silver. Premising that these salts are sensitive, 1st, by being darkened, 2d, by receiving a latent image rendered visible by a deposit of metallic silver, or, 3d, by receiving an image which is made visible by decomposition by alkalies in connection with reducing agents, he gives the results of his experiments to ascertain the sensitiveness of various salts in the third way above mentioned. Silver platinocyanide gave the strongest image, though none of the substances tried at all approached the haloid silver salts in sensitiveness. Moreover, he observed that no substance insensitive in the absence of tannin became sensitive by its presence.

4. Interference and Polarization.

Trannin has devised a new method of photometric measurement for lights of different colors, with which he has examined and compared the luminous intensity of different portions of the spectrum. The method consists in superposing the spectra of the lights to be compared, and then producing