period; but the dust soon settled, and, tions by the attraction of Jupiter, or the actual lowering of temperature to errors in the assumed orbital eleamounted to only about half a de- ments, were not in their predicted gree C. All the most remarkable places. Bailey, from photometric cold seasons since 1750 seem to be observations, found the range of varsimilarly related to great volcanic iability in the light of Eros to be outbreaks. The relation between the from 0.5 to 0.9 magnitudes, and de"dry fog" of 1783 and the ensuing tected regular variations of smaller severe winter was noticed at the amount in the light of five other time by Franklin. (See also XXV, asteroids, the periods of variation, Meteorology and Climatology.) which are all short, being presumably identical with the times of rotation of these planets.

Comets. The first comet to be discovered in 1913. called therefore Comet 1913a, was seen by Schaumasse at Nice on May 6. It was point of its orbit

The Planets.-Lau, discussing in detail the surface markings of Jupiter (A. N., 4673), explains the white spots often seen as clouds covering the sites of eruptions of material from the heated interior. The eruptive products, after cooling down, then almost at the drift off to the eastward, and form nearest to the sun (perihelion), the conspicuous dark belts. Barnard, which it passed on May 15, at a visually observing Phoebe, the faint distance of 1.45 times the radius of outer satellite of Saturn, estimates the earth's orbit. Its orbit is at it as of the fourteenth magnitude, least approximately parabolic, the much brighter than had previously been supposed (A. N., 4561). Ristenpart discusses the observations of the occultation of a seventh magnitude star by Jupiter's third satellite, as seen from various stations in Chile. From the durations of occultation, ranging from 2m. 12s. to 4m. 34s., the positions of the observers, and the known rate of motion of the satellite, he concludes that the satellite is flattened at the poles, the equatorial diameter being 4,660 miles, and the polar 4,270, both considerably larger than the values derived from micrometric measures. The excellent series of observations of the satellite of Neptune at the U. S. Naval Observatory have been continued.

inclination of its plane to that of the ecliptic being 28 deg., and the direction of motion retrograde, i.e., opposite to that of the earth and the other planets. By the end of August it was almost out of sight. Comet 1913b, discovered by Metcalf at South Hero, Vt., on September 1, appears to have also a nearly parabolic orbit, with perihelion passage on Sept. 14, at a distance of 1.36, and retrograde motion. Like the preceding, it was visible in a small telescope. Comet 1913c, discovered by Neujmin at Simeis, Russia, on Sept. 3, was at first reported as an asteroid, and presented a very peculiar appearance, having a sharp stellar nucleus of magnitude 11.5, and a very faint nebulosity on the following side. Its orbit is certainly elliptic, with a period of about 18 years, a perihelion distance of 1.54, and an inclination of 15°.

The discovery of new asteroids continues unabated, but they are often insufficiently observed. Cohn (A. N., 4688) lists the elements of 21 of these little bodies, discovered between July, Comet 1913d, discovered by Delavan 1912, and July, 1913, for which trust- at La Plata, Argentina, on Sept. 26, worthy orbits could be computed. rais- was recognized at once as the return ing the whole number to 754. Five of of Westphal's Comet of 1852. This these had been discovered and ob- comet was observed for six months served in previous years, one of them in 1852-3 and showed a clearly elliptwice over, in 1907 and 1909, but the tic orbit, the computed period being observations then secured were insuf- 61.7 years, with an uncertainty of less ficient for the computation of an than a year. The actual period durorbit. Five other planets, supposed ing the revolution now completed to be new discoveries, were found to was 61.121 years. The perihelion be identical with previously known distance is 1.25 times the earth's disasteroids which, owing to perturba-tance from the sun, and the aphelion

distance 30.0 times, very nearly that of Neptune. Owing to the high inclination of the orbit (41 deg.) and the position of the line of nodes, however, the comet is more than 16 units south of the plane of Neptune's orbit when at aphelion, and can never come anywhere near it, though it may approach within a distance of 0.5 of the earth, 0.2 of Mars, and 0.15 of Jupiter. At the present return, the comet passed perihelion on Nov. 17, 1913, and was placed, relatively to the earth, in a very favorable position for observation. In 1852, when conditions were about equally good, it was of magnitude 4.5 when brightest, easily visible to the naked eye, and had a tail about a degree in length.

mean distance of 1,200 light-years; more than 95 per cent. of these stars are brighter than the sun (M. N., LXXIII. 334).

Radial Velocities.-Campbell gives the radial velocities of 915 stars, thus completing the publication of the extremely laborious and important work of the Lick Observatory on this subject up to date. Plummer, discussing the motions of the white stars, of spectral classes B and A, finds that their real motions are very nearly parallel to the plane of the Milky Way, and computes the distances and brightness of many stars (M. N., LXXIII, 450, etc.).

Double Stars.-The discovery of new pairs proceeds apace. Aitken's last list raises the number discovered by him at the Lick Observatory to 2,600. One of these pairs has completed a revolution since its discovery, and Aitken computes an orbit from his own observations, the period, 12.1 years, being the shortest so far found, with two exceptions.

Comet 1913e, discovered by Zinner at Bamberg on Oct. 23, proved to be a return of the third comet of 1900, whose period is 6.46 years. It was visible in a small telescope, but moved rapidly southward, and was soon lost to sight from northern observatories. The Stars: Parallaxes.-Important Spectroscopy.-Fowler has succeedseries of parallax determinations ed in producing in the laboratory, have been published by Chase and by passing a powerful disruptive disSmith, of Yale, by Abetti, who ob- charge through a vacuum tube conserved at Heidelberg, and by Slocum, taining a mixture of hydrogen and from photographs taken with the helium, the series of lines observed great Yerkes telescope. The Yale ob- by Pickering in 1896 in the spectrum servers find that the fourth magni- of the star Puppis, and supposed tude star Epsilon Eridani is one of by him to be due to hydrogen under our nearest neighbors in space, with exceptional conditions of luminous a parallax of 0".31, and a dis- excitation; and also two other series tance of 10 light-years, or about 700,- of lines (most of which lie far in 000 times that of the sun. Slocum the ultra-violet), one of which coinfinds that the new star in Gemini, cides with the "principal series" of like that in Lacerta, is enormously hydrogen lines whose existence was distant, probably many hundred predicted on theoretical grounds by light-years.

Rydberg. The first line of this series Proper Motions.-Charlier, discuss- is a conspicuous bright line in the ing the motions of the stars visible spectrum of Puppis and certain to the naked eye (M. N., LXXIII, other stars (M. N., LXXIII, 62). In 486), concludes that the mean dis- a later paper Fowler shows that the tance of the stars between the fifth strong line at wave-length 4,481, and sixth magnitudes is 29,000,000 characteristic of the spark spectrum times that of the sun, or 440 light of magnesium, is also the first memyears, and finds that the distribution ber of a similar series of lines. The of the real motions of the stars in important studies of the behavior of space is more closely represented by the lines of iron, titanium, etc., unthe "ellipsoidal" hypothesis of der varying conditions of temperature Schwarzschild than by Kapteyn's and pressure in the source of light hypothesis of two intermingling have been continued at Mount Wilstreams of stars. Dyson finds for the son, with very valuable results. tars down to the tenth magnitude Nicholson shows that most of the in nine degrees of the pole a lines of unknown origin in the spec

tra of the solar corona and of the gaseous nebulae can be accounted for theoretically by the electrical vibrations of very simply constituted atoms, and Bohr, employing the new theories of radiation, shows that the hydrogen series may be similarly explained.

The cause of these singular phenomena is not yet understood.

Photometry. Parkhurst, in his Yerkes Actinometry, gives measurements of the brightness of some 600 stars within 17 deg. of the North Pole, made photographically, both on ordinary plates (that is, with violet light) and on isochromatic plates, using a color screen transmitting yellow and green light. The difference in the brightness of a star measured in these two ways gives an accurate measure of its color. The spectra of the stars were also photographed, and the relation between the "color index" just described and the class of spectrum determined, with results agreeing well with those of other observers. If two stars of spectra A and M appear equally bright to the eye, or on the isochromatic plates, the former will appear four times as bright as the other on the ordinary plates. (Ap. J., XXXVI, 169).

The system of classifying stellar spectra which has been developed at the Harvard Observatory was adopted by the International Union for Solar Research for general use, pending the adoption of a definitive classification. According to this system, the six principal types of stellar spectra are denoted by the letters B, A, F, G, K and M, and intermediate classes by the "decimal notation;" for example, a spectrum half way between the B and A types is called B5A, or simply B5. Eberhard and Schwarzschild find that the H and K lines of calcium are reversed, (bright), in the spectra of Arcturus and some other stars of class K, which in this respect, as in some Variable Stars.-One hundred and others, resemble the spectra of sun thirteen new variable stars whose spots.

Stellar Temperatures.-Rosenberg, from the distribution of energy among the different wave lengths in the spectra of some 60 stars, deduces the effective temperatures of their surfaces, finding, like previous observers, a steady decrease in temperature from class B to class M. The range of his computed values, from 40,000 deg. for class B to 2,300 deg. for class M, is, however, much greater than that found by some earlier observers (A. N., 4621).

Spectroscopic Binaries.-The rate of discovery of these interesting systems still far exceeds that at which the necessary observations for the computation of their orbits can be secured. Among those whose orbits have been computed may be noticed Ursae Majoris, with the unusually long period of 4.15 years, and RR Lyrae, with the very short period of 13 hours (see infra).

Belopolsky finds that in the spectrum of a Canum Venaticorum certain lines vary greatly in intensity, with a period of 5.5 days, while the rest are unaltered. The variable lines show periodic changes in radial velocity, while the others do not.

changes in brightness have been confirmed by independent observations have received definitive names between June, 1912, and June, 1913. Much attention is being paid to the theory of stellar variation. In the case of the eclipsing variables, which are usually of constant brightness but lose light at regular intervals owing to the interposition of a fainter companion, the theory is in a very satisfactory state. Shapley, using the methods devised by Russell, has worked out the orbits of 87 such systems. His principal conclusions are: It is usually certain, and always probable, that the eclipsing companion has some light of its own. For the most accurately observed systems, there is definite evidence that the stars appear brighter at the centers of their disks than at the edges, as is the case with the sun. In many cases the two stars of a pair keep always the same faces toward one another, and are elongated into ellipsoids by their mutual attraction, the effect being greater the nearer they are together. The faint companion is often larger than its brighter primary, and, when its color is known, is also redder and presum

ably colder, but nevertheless it is al-, Præsepe cluster in Cancer, which are most certainly much less dense. All receding from us at the rate of 36 but one of the 87 stars are less dense km. per second. These stars have than the sun, the white stars, of small but equal proper motions, and spectra B and A, having usually den- appear to be travelling in space in sities from one-fifth to one-fiftieth almost the same direction and at althat of the sun, while the yellow most the same rate as the Hyades stars, of spectra F, G and K, fall into cluster, in which case their distance two groups, one about half as dense from us must be about 500 lightas the sun, and the other of ex- years. tremely low density.

XXV. GEOLOGY, METEOROLOGY, AND GEOGRAPHY

GEOLOGY

DYNAMIC AND STRUCTURAL J. W. Spencer (Amer. Jour. Sci.,