been in Arequipa. When the regular siege of Lima began, on Feb. 19, the Government troops erected fortifications and planted guns. The attack was not made till March 17. The revolutionary troops advanced simultaneously in the night from various points. Favored by a fog, one division passed the Cacerist lines unobserved and penetrated to the center of the city, occupying the church towers and other commanding positions and continuing to advance until a line of communications was formed with the division that entered the city from the opposite side. Another division under Col. Pauli outflanked the Cacerist forces and occupied the towers of the Merced church. Pierola had 2,000 riflemen strongly posted at many points before it was light enough to see or be seen by the enemy. The battle in the streets began in the dark and raged till night fell. Outside the city other Pierolist forces engaged the troops that held the fortified positions. The fighting was resumed on March 18 at five o'clock in the morning and kept up till night, and on March 19 firing began again at dawn and was kept up till noon in a weak and desultory fashion, for the contending forces were reduced to a fraction of their original numbers and all who were left were physically exhausted. The Cacerists were thoroughly demoralized, for they were fired upon by citizens from their houses, partisans of Pierola who had kept arms concealed in spite of police searches. Drunken and barbarous from the first, and now despairing of the battle, they began on the third day to loot the stores and clubs and the houses of the wealthy into which they could force an entrance. The Government forces still held the plaza, the palace, and the fort of Santa Catalina. The Pierolists, still 2,000 strong, were prepared to continue the combat and to burn with petroleum or level with dynamite all obstacles to their progress. The leaders, however, having already won the victory, willingly agreed to an armistice of twenty-four hours when the diplomatic corps intervened to prevent needless carnage and destruction and the pestilence that the putrefying corpses of men and horses threatened to bring upon the city. Before the armistice expired a peace was arranged through the mediation of the papal delegate.

The articles were signed on March 21 by Dr. Luis Felipe Villavan, representing Gen. Caceres, and Enrique Bustamente y Salazar, in behalf of Nicolas Pierola, the terms being that Caceres should resign the presidency and retire to Ancon, and Pierola lay down the command of his army and withdraw to Chorillos, while a Provisional Government should be formed, consisting of 5 men, 2 to be nominated by Caceres, 2 by Pierola, and the fifth selected by the others or, in case of disagreement, by lot. The Provisional Government was to keep its powers no longer than was necessary to hold a general election and install a regular government.

Although the number engaged on both sides did not exceed 8,000, the number of men slain or mortally wounded in the two-days' battle was 1,875, and 1,553 were less severely wounded. The Government troops lost more than double the number that the revolutionary forces did, owing to the hostility of the people. In accord

ance with the stipulation both armies withdrew from the city to camps outside. An urban guard of 2,000 men was organized to preserve order, composed largely of foreigners.

The Election of Pierola.-The Provisional Government was composed of Manuel Candamo, Ricardo W. Espinosa, Luis Felipe Villaran, Enrique Bustamente y Salazar, and Elias Malpar tida. The military ranks granted up to date in either army were confirmed, according to the agreement, but the reorganization of the army was left to the coming Congress. The troops of Carceres received some pay from money advanced to the Provisional Government by the banks and were disbanded. Gen. Carceres did not go to the retreat assigned to him, but took refuge on a French war vessel at Callao, and was afterward conveyed away by a British manof-war. The political prisoners that were starv ing in the jails of Lima and Callao were released immediately. The authorities took possession of all arms and ammunition. In the south, Gen. Mas, in the district of Cuzco, and the Cacerist commanders at Pisco and other places still refused to recognize the Provisional Government, but they made their submission before a month passed by. Gen. Mas, who had 2,000 men, remained defiant until after Gen. Pierola, who landed at Mollendo with artillery and cavalry for the purpose of chastising him, had a sharp battle with him for the possession of Cuzco. Import duties on iron, coal, and machinery, and export duties on cotton, India rubber, pitch, hides, cotton seed, sugar, and tobacco, which the Cacerist Government had imposed, were repealed.

Elections took place on July 7, but Pierola objected on account of some irregularity, and new ones were held on July 26, when the same candidates were re-elected. Nicolas Pierola was elected President without opposition for the term ending Aug. 12, 1899. The new VicePresidents are Guillermo Billinghurst and Augusto Seminario y Vascones. The new Government had to render satisfaction for the arrest, by Caceres, of a British vice-consul who would not subscribe to a forced loan, and to the Ger man and Spanish governments for similar outrages committed upon their subjects by Gen. Mas in Cuzco. While the Provisional Government was still in authority some American Protestant missionaries went to Cuzco, causing great excitement and indignation among the people, and the local authorities finally expelled them in spite of the decision of Minister Can damo that Protestant worship was protected and that they had equal rights with Catholics. Bolivia demanded that her flag be saluted for a violation of her territory during the civil war, but the Peruvians resented this demand and the Government temporized and finally agreed to submit the question to the arbitration of some other American state.

President Pierola was inaugurated duly on Aug. 12, and on Sept. 9, after the assembling of the new Congress, which met on Aug. 30, he ap pointed the following Cabinet: Premier and Minister of the Interior, Antonio Bentin; Minister of Finance, F. Bresani; Minister of For eign Affairs, Meliton Porras; Minister of Justice and Worship, A. Albarracin; Minister of

War and Marine, Col. D. J. Parra. On Dec. 1 a new Cabinet was formed, composed as follows: Premier and Minister of Justice, Señor Barrinaja; Minister of Foreign Affairs, Ortiz Zevallos; Minister of Finance, Señor Obin; Minister of War, Col. Ibarra.

PHYSICS, PROGRESS OF, IN 1895. Constitution of Matter, etc. Molecules.— Stoney (London Royal Society, May 16) thinks that events go on within the molecules of matter that are so sluggish in affecting its pressure or temperature that only after millions of encounters does any manifestation of loss of energy by conduction become appreciable; while the same events act promptly and actively in other ways, as in chemical action or radiation. These events are those internal motions of the molecule in which during encounters there is some exchange of energy with the external or translational motions, but not a ready and quick exchange, while the same molecules may nevertheless exchange energy promptly and quickly with the ether by radiation. The author adds that electrons, for the most part associated with this class of motions, appear to be primarily concerned in every chemical reaction and in all the phenomena of radiation. Sutherland ("Philosophical Magazine," January), in an article on the laws of molecular force, demonstrates that the expression m3l, which occurs in the treatment of the attractions of like molecules, can be analyzed into two factors, which are the sum of numbers characteristic of the atoms composing the molecule, whose mass (referred to the atom of hydrogen) is m. This would analyze molecular into atomic attraction. Belief regarding the change of molecular structure with time in certain solids has caused some people to fear that no permanent standard of length can be made, but Rogers (American Association, 1895), as a result of comparisons that extend over five years, concludes that such fears are not well founded.

Mechanics.

66

66

Absoluteness of Rotation. Mach, in his "Science of Mechanics," does not accept Newton's distinction between the relativity of motion of translation and the absoluteness of motion of rotation. A review of the work by Greenhill in "Nature," Nov. 15, 1894, has been the cause of wide discussion among mathematicians and physicists, from which it appears that many authorities consider that rotation as well as translation is only relative. Energy. Helm (Wiedemann's Annalen," June), in a Survey of the Present Position of Energetics," states that the two views of energy that are now struggling for supremacy are that which regards it as a mathematical abstraction and that which considers it to be a concrete reality, filling space and migrating continuously from place to place. One of the chief generalizations of energetics, according to this writer, is the following: In order that something may happen it is sufficient and necessary that uncompensated differences of intensity exist." Allen (London Physical Society, March 8) has mathematically investigated the motion of what he calls "energy cells," which are the small volumes bounded by the walls of a tube of force and by the two neighboring equipotential surfaces. This motion differs according as the

66 en

charged or gravitating particles meet or separate. Mr. Allen calculates what he calls the ergy density" in the medium at the sun's surface, and find it equal to 16 horse-power hours per cubic centimetre.

Elasticity.-Fraas (Wiedemann's "Annalen ") has measured the elasticity of solid gelatin solutions, and finds that in no case does the volume change by stretching. Part of the water could be replaced by glycerin, cane sugar, or gum arabic without making any difference, but common salt impaired both the elasticity and the strength of the gelatin sticks.

Kinematics of Machines.-Hearson (London Royal Society, May 16) shows that all machine movements, however complex, may be derived from a limited number of simple motions. By attaching to each kind of simple motion a suggestive symbol any complex machine motion may be exhibited by a graphic formula. Thus the turning motion of consecutive pieces when one revolves completely about the other is represented by the letter Ŏ, and swinging motion, when one turns through a limited angle relatively to the other, by the letter U. The geometrical limitations of the combinations of such motions are shown to be such that only 14 distinct types of these are possible. The author classifies all simple machine movements as follows: 1, Plane mechanisms, where the pieces move in or parallel to a plane; 2, spherical mechanisms; 3, cylindrical mechanisms; 4, conoidal mechanisms, where the axes of the swinging and turning motions neither meet nor are parallel.

System of Pendulums.—A system of 2 equal pendulums joined by an elastic_thread has been studied by L. de la Rive ("Journal de Physique"), who finds that energy is periodically transmitted from one pendulum to the other. The period of this phenomenon equals the ordinary period of oscillation multiplied by a number that is proportional directly to the length and inversely to the section of the thread. After 10 or 12 periods the alternation of periods is scarcely perceptible, and in the end the pendulums tend to move like a rigid system, with constant tension of thread.

Lissajous's Curves.-Oosting ("Zeitschrift für den Physikalischen Unterricht ") uses for the production of these curves 2 small mirrors attached respectively to vertical and horizontal stretched wires, the period being adjusted by screws carrying nuts mounted behind the mirror at right angles to the wire.

Liquids. Capillarity.-Van der Mensbrugghe (Paris Academy of Science, Sept. 30) finds that the great theories of capillarity agree but imperfectly with the facts that are observed in evaporation. Most liquids evaporate spontaneously in air, hence: (1) The density of the superficial layer must decrease outward, contrary to Gauss's and Laplace's theories; (2) the mass of liquid has not an invariable volume, contrary to Poisson's theory; and (3) the superficial layer being continually renewed, is not in equilibrium, contrary to the theories of most contemporary analysts. The author has elaborated a theory which he regards as free from these objections. E. Kaiser (Wiedemann's "Annalen," December, 1894) has studied the mingling of masses of

liquid by observing the Newton's rings that form between films. Difference of potential hastens the mingling, and Kaiser's experiments show that it does so by increasing the pressure on the intervening air, forcing it out at the sides. With the films that he investigated fusion took place in 3.2 seconds with no potential difference, in 14 second with one Daniell cell, und in 0.4 second with 2. With more cells, fusion took place instantly and the films usually burst. Maltézos (Paris Academy of Science, Aug. 5), after a study of the so-called Brownian movement of small particles in a liquid, concludes that it is a capillary phenomenon. G. Quincke (Wiedemann's "Annalen," December, 1894) calls attention to the fact that the forms assumed by combinations of alkalies with oleic acid when brought into contact with water have a remarkable analogy with the configuration of various small portions of the stellar universe, such as portions of Orion, Virgo, and Coma Berenices.

Critical Point.-Pictet (Berlin Physical Society, Feb. 22) is of the opinion that substances must still be in the fluid state at the critical point, for the amount of heat that must be put into the substance, reckoning from absolute zero, is less than the latent heat of the liquid, and solid bodies do not separate from solution at the critical temperature, but do so on a further rise of temperature.

Evaporation. Lehfeldt ("Philosophical Maga zine," November) has deduced from thermodynamic considerations a formula representing the relation between the composition of a mixed liquid and that of the vapor that rises from it. This formula agrees reasonably well with experiment, though new data seem to be needed.

Condensation.-Wilson (Cambridge Philosophical Society, May 13) finds that in dustfree air there is still condensation into a cloud of fine drops after the expansion exceeds a certain critical amount, which point remains constant, no matter how many expansions are made. The ratio of final to initial volume at this critical expansion is 1.258, when the initial temperature is 16.7° C., corresponding to a fall of temperature of 26°.

Solution.-Linder and Picton concluded several years ago that there is no definite line to be drawn between suspension and perfect solution, the difference being only one of degree of aggregation. They have now ("Journal of the Chemical Society," February) strengthened their conclusion by finding that a mixture of a 2-per-cent. arsenious-oxide solution with hydrogen-sulphide water is not only diffusible, but can be filtered through a porous pot. Four grades of such mixtures or solutions have thus been obtained, of which the first contains aggregates that can be seen under the microscope, the second is invisible but not diffusible, the third diffusible but not filterable, and the fourth both diffusible and filterable, though it scatters and polarizes light. Lecoq de Boisbaudran (Paris Academy of Science, July 8) considers that all soluble substances belong to a continuous series, of which the members at one end may dilate on solution while those at the other may contract. Crismer ("Bulletin de l'Académie Royale de Belgique," No. 6) finds that the critical temperature of so

lution is independent of the amount of either body present. It varies much from one body to another, but is constant for the same body, and for a mixture is sensibly equal to the arithmetical mean of those of the constituents. The surface tension of the lower of two liquids tends toward zero at this temperature and the separation meniscus becomes a plane, hence the temperature may be determined by an optical method. Arctowski (ibid.), by determining the solubility at very low temperatures of some organic compounds in carbon disulphide, finds that the point of fusion of the solvent appears not to be an essential point on the curve of solubilities, as supposed by Étard.

Crystallization.-Lecoq de Boisbaudran (Paris Academy of Science, April 22) shows that crys tals may form at the bottom of a solution of greater specific gravity than themselves, the action depending on small variations of temperature. Baudrowski ("Zeitschrift für physi kalische Chemie," November, 1894) has examined the light that certain salts emit during crystallization, and concludes that it is probably electrical, being due to the union of electrified ions.

Gases. Hygrometry.-A hygrometer based on a new principle is suggested by Dr. J. Verschaf felt ("Bulletin of the Belgian Royal Academy," Nos. 9 and 10). The hygrometric state of the atmosphere may be taken as the ratio of the vapor tension inside a solution to the highest possible vapor tension of water at the same tension, when the solution is neither evaporating nor condensing water from the air. Thus by moistening a weighed piece of blotting paper with a weighed quantity of a solution of a salt of known concentration, exposing to the air, and weighing again, the "equilibrium concentra tion," and hence the humidity, may be calcu lated.

Kinetic Theory.-Fitzgerald (London Royal Society, Feb. 7) has attempted to show that Maxwell's theorem of the equal partition of energy among the degrees of freedom of atoms is not, as has been generally held, inconsistent with the various internal movements indicated by gaseous spectra. This is due to the control exercised over the motion of the electrons on neighboring atoms by the intervening ether, so that if, say, 10 atoms are thus connected, the motions of their electrons might be defined by 3 co-ordinates, and "if the atoms were spheres there would be 3x10 degrees of freedom plus the 3 degrees that define the motion of the electrons. Now, if the total energy be equally distributed among all these degrees of freedom, each atom will have only its share of the elec tro-motions, and its energy of external motion will only be diminished by 3X10-6 part owing to the existence of the internal motion of its electrons." Our calorimetric methods are not sufficiently delicate to detect this diminution. Liveing and Dewar (Paris Academy of Science, July 15) find that observation of the spectra liquid air and oxygen do not bear out the theory that the increase of intensity of the bands as the square of the density of oxygen is due to the encounters of molecules of ordinary mass which are more frequent as their free path is diminished. Mixtures of liquid air and oxygen

of

confirm the law of increase at low temperatures. Solid air shows practically the same character and intensity of absorption as liquid air. A thickness of 04 centimetre of liquid oxygen gives a much greater band intensity than 1.9 centimetre of liquid air.

Viscosity. It has been noticed several times that the viscosity of a mixture of gases is greater than would be expected from that of each of its components. Thus the addition of a little hydrogen to carbon dioxide makes it more viscous, although the former gas is less viscous than the latter. The theory of this action has been worked out mathematically for the first time by Sutherland ("Philosophical Magazine," November). who finds it to depend on the difference of mass of the molecules of the mixed gases.

Acoustics. Combinational Tones.-Rücker and Edser (London Physical Society, March 22) find that under certain conditions difference and summation tones are capable of disturbing resonating bodies, and they regard this as settling the vexed question of the objective or subjective nature of these tones. In one case the resonator was a tuning fork whose vibration was observed by the method of the interferential refractometer. Dr. Burton, at the same meeting, described experiments with two organ pipes, showing that their combination tones varied with the relative distance of the pipes, the observer remaining at the same distance-a fact that argues for the objectivity of the tones.

Comparison of Forks.-Hallock (American Association, 1895) has devised a photographic method of comparing the pitch of tuning forks by clamping each fork before a manometric capsule, bowing it, photographing the flames, and then counting the relative number of vibrations. Vibration of a String.-Cornu (Paris Academy of Science, Aug. 5) finds that the transverse vibration of a string, no matter how it is excited, is always accompanied by torsional vibrations. The actual vibration is complicated not only by this fact, but also by the usual want of symmetry of the strings about their axes.

Heat. Calorimetry.-Waterman ("Philosophcal Magazine," November) has devised a caloimeter that is maintained at a constant temperaSure after the introduction of the heated body by means of cold water, instead of allowing it to ise in temperature and measuring that rise. This gets rid of the radiation error and elimiates the "water equivalent" of the vessel. The water is dropped in from an inverted copper cone urrounded by ice. For bodies of the same weight and the same initial temperature the specific heat is measured simply by the amount of ce-cold water necessary to cool them to the temperature of the room.

Conduction.-Quick, Child, and Lanphear Physical Review," July-August) find that the hermal conductivity of copper varies from 0.921 t-54° to 1.059 at 13°, but the increase is more apid for the temperatures below zero. The avrage value for the range -54° to -13° is greater han that for 70° to 170°, so either the results must be affected by some undiscovered source f error or the curve of conductivity has a maxiaum between -14° and +70°.

Specific Heat.-Violle (Paris Academy of Scince, April 22) has measured the specific heat of

graphite at high temperatures, and finds that above 1,000° C. it increases linearly with the temperature. A gramme gives up 2,050 calories in cooling from volatilization to 0°, and M. Violle calculates from this that the boiling point is 3,600° C.

Volume Heat.-By this term E. H. Griffiths (Cambridge Philosophical Society, May 13) denotes the capacity for heat of equal volumes at different temperatures. He has calculated it in the case of aniline from the published results of observations, and finds it constant.

Fusion.-Le Chatelier (Paris Academy of Science, Aug. 12) finds that the accepted melting point of gold-that due to Violle, who made it 1,045° C.-is a little low. The error does not exceed 20°, and the experimenter advises that no change be made for the present in the pyrometer scales that are based on this melting point. Bruner (Paris Academy of Science, April 29) has studied the specific heat of superfused liquids and finds that thymol and paracresol give specific heats that increase with the range of temperature, when they are cooled without solidification to approximately the same extent below their melting points.

Absorption.-Friedel (Wiedemann's “Annalen,” July) finds that when, in a chemical compound, hydrogen, oxygen, hydroxyl, or nitrogen are replaced by sulphur or a halogen, the transmittance of the solution is considerably increased. In homologous series the transmittance is regularly changed by every addition of CH2, but the direction of the change depends on the other atoms. The absorptive power depends on the constituents and is independent of the size of the molecule. In isomeric compounds the diathermancy differs not only with the difference of atomic volume of the elementary atoms, but also with the difference of linkage among themselves.

Radiation.-A. M. Porter (London Physical Society, Jan. 11) finds that, contrary to the usual assumption, the "emissivity" or amount of heat that passes outward from unit area per second for 1 degree of excess of temperature is not independent of the size of the body, and also depends materially on the shape of the bounding surface.

Temperature.-Berthelot (Paris Academy of Science, April 16) has devised a method for the measurement of the temperature of gases that is independent of the envelope. It is based on a determination of the density of the gases by means of their refractive indexes as shown by interference fringes.

Temperature of Greatest Density of Water.M. de Coppet ("Annales de Chimie et de Physique") has made a new and accurate determination of the temperature of greatest density of water, and finds it to be 3·982° by the hydrogen thermometer under a pressure of 1 atmosphere.

Light. Absorption.-Dufour and Brunner (Société Vaudoise; reported in "La Nature,' Dec. 7) find from experiments on the opacity of carbon that a layer of lampblack millimetre thick deposited on glass will not transmit the sun's light. To darken the whole earth, therefore, it would suffice to reduce to smoke a mass of of a cubic kilometre of carbon. Merritt (Physical Review," May-June) has determined the dependence of infra-red absorption upon po

larization for quartz, Iceland spar, and tourmaline. The transmission curves of the ordinary and extraordinary rays were entirely different in all 3 cases, the difference being especially marked with Iceland spar.

Refraction. Sir John Conroy (London Royal Society, June 20) has determined the refractive indexes of water at temperatures between 0° and 10° C., and finds that the index increases continuously up to the freezing point, while the rate of increase alters at about 4°, the point of maximum density, and that no formula that represents the variation as a function of the density only can completely express the facts.

Photometry-Kurlbaum and Lummer (Berlin Physical Society) have made a determination of the unit of light that is based on the light emitted by white-hot platinum foil. To keep the temperature constant for a long period and to be able to re-establish it at any time the ratio of the total radiant energy from the foil to that transmitted by a medium was measured bolometrically. The medium chosen was a thin layer of water in a quartz cell. The errors amounted to 1 per cent., and were due chiefly to air currents on the surface of the foil.

Spectroscopy-Wadsworth ("Astronomy and Astro-physics," December, 1894) has devised a way of rotating the prism of a spectroscope, where movement of the collimator or of the observing telescope is objectionable, in such way as to retain minimum deviation for the central ray in the field. A mirror having an angular motion equal to one half the change in angular motion of the minimum-deviation ray is introduced somewhere between the slit and the focal plane of the observing lens. In the spectro-bolometer it is in continuation of the back face of the prism. Langley, in a paper on "Recent Researches in the Infra-red Spectrum " (British Association, 1894), reviews the work now in progress at the Smithsonian Institution in this direction, especially the combination by composite photography of several linear representatives of the spectrum to form a single one, on which "we may expect to find only what is permanent and not what is accidental." He concludes that more than 2,000 lines will thus be mapped in the infra-red spectrum. Edler and Valenta (Vienna Academy of Sciences), in a paper summarized in "The Astrophysical Journal," May, have discovered two new spectra of mercury, obtained by passing an electric spark through mercury vapor that is distilling at low pressure through a capillary tube. If a large number of Leyden jars are in circuit, the spectrum has a great number of fine, sharp lines; if not, it becomes a series of bands with edges toward the red. The band spectrum corresponds to a slightly lower temperature than the line spectrum. Galitzin (Wiedemann's "Annalen," September) considers the molecular theory of the broadening of spectrum lines superior to those based on Doppler's principle, on Kirchhoff's law, or on damping, since it admits of a development based on the theory of molecular resonators. The broadening is a consequence, according to this, of forced vibrations produced by molecular collisions. The theory explains the unsymmetrical broadening and the influence of temperature and pressure. Evershed ("Philosophical Magazine," May) finds that the vapors of iodine,

bromine, chlorine, sulphur, selenium, and arsenic all glow by external heating, but give continuous spectra. Dense sodium vapor acts in the same way, but on a reduction of density it gives a discontinuous spectrum. The experimenter is satisfied that his precautions precluded the possibility of any chemical action, so that the spectrum was due to true external heating.

Polarization.-Landolt ("Sitzungberichte" of the Berlin Academy, No. 38) has devised a simple method of obtaining lights of different wave lengths for use in polarimetric work by passing white light through absorption cells. The resulting light, though not actually monochromatic, gives a uniform tint in the field of the polarimeter if the rotation be less than 20°. R. A. Millikan ("Physical Review," SeptemberOctober) has investigated the polarization of the light emitted by incandescent bodies, and finds that it is a minimum with rays emitted normal to the surface and a maximum with a grazing emission, which indicates that the vibrations are in a plane at right angles to the emitting surface. In the following issue he studies the phenomena qualitatively and finds a striking agree ment between the measured amounts of polarization at different angles of emergence and those calculated from Cauchy's theory of metallic reflection on the assumption that the polarization is due to the refraction of rays coming from the interior on emergence. The agreement is especially good in the case of molten silver. The fluorescent light developed at the surface of uranium glass is polarized in much the same way, and the fact is explained similarly by the author. Uljanin (Berlin Physical Society, March 8) has investigated the polarization that is due to oblique radiation from silver, platinum, and black glass, and finds that curves arrived at by assuming that the radiation of the substance is determined by its refraction correspond well with those obtained by direct experiment, especially in the case of silver.

Rotary Polarization.-Rodger and Watson (London Royal Society, June 20) have attempted to determine in absolute measure the magnetic rotation of liquids at different temperatures, the effect of the chemical nature of the liquid on this property, and its correlation with other physical properties. They conclude that the usual measure of the molecular rotation, which involves the properties of water, is particularly ill suited for the purposes to which it is applied, since the behavior in water is exceptional and its rotation is small. In all of the 10 liquids that they examined, except water, the rela tion between rotation and temperature is linear, and the quotient of the rotation by the density diminishes with increase of temperature. Siertsema (Amsterdam Academy of Science), with currents of 35 to 65 ampères, has been able to obtain in oxygen a magnetic rotary dispersion of 3° to 4°. A series of measurements were also made on atmospheric air and values for nitrogen were deduced therefrom.

Luminescence.-E. Wiedemann and Schmidt (Wiedemann's "Annalen," April) draw an important distinction between physical and chemical luminescence. When there is a prolonged afterglow the phenomenon is probably chemical. Luminescence under cathode rays is always ac