put the underlying ideas of relativity into simpler form. There has been no attempt either in this country or abroad (except possibly Sagnac, Journ. de Phys., Nov., 1912) to extend the experimental basis for the relativity hypothesis, which has as yet not shown itself to be particularly fruitful or suggestive.

From two such institutions as the Mt. Wilson Solar Observatory and the Smithsonian Astrophysical Observatory continued contributions to our knowledge of the sun are of course to be expected, and during the year has come Hale's preliminary work on the general magnetic field of the sun, in which he shows, from the most minute examination of the absorption of light in the sun's atmosphere, that the sun has magnetic poles similarly situated to those of the earth, but that the sun's magnetic field is many times stronger than the earth's. The completion of Abbot's long study of the intensity of solar radiation gives a very accurate value for the "solar constant" (1.93 cal. per sq. cm. per minute) and proves that the energy which the sun sends to us is constant only on the average, and changes by as much as 10 per cent. from time to time. (See also XXIV, Astronomy.) Michelson's very recent observations of tidal waves in the earth crust are of general interest, and are remarkable because of the simplicity of the method used.

Heat. Of very general interest is the contention of Johnson and Adams (Am. Jour. of Science, March, 1913, p. 205), as the result of much experimental work, that the effect of uniform pressure on the melting point and chemical behavior of solids is so small that it probably has little to do with the condition of the deep layer of the earth's crust. They conclude, in agreement with, but extending, the conclusions of other observers, that high temperature has a much greater influence than pressure in altering the physical and chemical properties of solids, and that the melting point of a solid is not seriously altered by pressure unless the pressure is non-uniform, like a twist or a shear. Related to this is the continued work of Bridgeman with ex

even

results (Proc. Am. Acad., May, 1913, p. 1) give the properties of a dozen liquids (alcohol, ether, etc.) from 20 to 80 deg. C. and up to pressures of about 11,000 atmospheres. While the results as a whole are extremely complex, and indicate that possibly the shape of the molecules is respon. sible for some of the effects, nevertheless certain generalizations can be made, namely, that all liquids seem to become more nearly alike at high pressures, and the somewhat contradictory one that there is no evidence that large and complicated molecules break down into simple ones when they are pressed most closely together. Passing to the other extreme of pressure, Trowbridge has obtained new data for the heat-conducting power of air at pressures of a few hundred-thousandths of an atmosphere, and he proposes to use this variation, in its effect on the temperature of a fine electrically heated wire, as a very sensitive gauge for extremely low gas pressures. For a similar purpose Langmuir suggests the measuring of the drag of a disc by another rapidly rotating disc separated from the first by the gas whose pressure is to be measured. In this way he can measure a pressure as low as one ten-billionth of an atmosphere. Millikan (Physical Review, March, 1913, p. 218) has studied the motion of minute drops of oil in air at various pressures due to the impact of air molecules (Brownian movements), thus confirming the deduction from the kinetic theory of gases.

To conclude the field of heat, Roebuck (ibid., August, 1913, p. 79) has carried out a very satisfactory determination of the mechanical equivalent of heat by the hitherto unused method of forcing water through a porcelain diaphragm, and has studied the properties of water near its temperature of maximum density, 4 deg. Ĉ.

Radiation. In the field of long wave, or in infra-red radiation, Wood (Philos. Mag., April, 1913) obtained the rather surprising result that a layer of mercury droplets is almost perfectly transparent to long waves unless the drops are more than onetenth of the wave length in diameter.

way and avoids some difficulties of the simpler theory mentioned above, it is not to be considered as final.

As regards the conduction of heat, the trend of opinion is undoubtedly back toward the earlier view that molecular motions are chiefly responsible for the conduction of heat through solids, and that the remarkable fact that many good conductors of electricity are also good conductors of heat, is of secondary importance. One of the most puzzling properties of metals is their so-called "contact electromotive force," which shows itself in this way, that if two different metals are connected electrically to earth and then brought with two faces near each other, the contiguous faces will at once become charged with electricity in a perfectly definite way, which surface becomes positively charged depending on what particular pair of metals is used. Hennings (ibid., July, 1913, p. 1) has carefully examined the problem and settled some disputed points; for example, he has shown that it depends only on the surfaces which are close to each other, and that either surface charge will not appear if the one metal is screened by a wire gauze not too far away. The real cause of the effect, whether due to an insulated electric layer brought out by chemical action or to a direct action of the surfaces of the metals, is not known at present, though some results of Page and some observations of Millikan count decidedly against the layer theory.

After several years of elaborate study of all possible sources of error, Millikan has published (ibid., August, 1913, p. 109) his final value of the elementary electrical charge or "atom" of electricity, as determined by his beautiful oil-drop method, namely, 4.774X10-10 electrostatic units. This in part depends on Gilchrist's new measurements of the viscosity of air, and is by all odds the most accurate determination of this very fundamental quantity.

waves, the method may be used to pick out waves of a certain length from a complex beam. Ives has studied the absorption of electric waves in ionized as compared with non-ionized air and water vapor and obtained results almost great enough to account for the difference between day and night transmission of wireless waves.

X-Rays and Discharge Through Gases.-While some interesting work on X-rays has been carried out, it is notable that American experimenters have not stepped into the new field of X-ray diffraction and reflection by crystalline and other solids. Since Laue and Friederich's beautiful pioneer experiments of 1912, a host of foreign observers have taken up the work and at least the fundamental question seems to have been settled that X-rays are electromagnetic pulses of the same nature as light, but with an effective wave length about 1/10,000 that of the shortest known light wave. By a most refined study of the effect of X-rays in breaking up air molecules into positively and negatively charged ions, Plympton has brought out the fact that for the first one-third of a second after the ions are produced they are very likely to recombine because of their proximity; after this period the rate of recombination becomes constant. The puzzling fact that X-rays and ultraviolet light when passed through a thin film of metal drive off more electrons from the metal in the direction in which the light (or X-rays) is going than in the opposite direction has been taken by some to indicate the necessity of returning to a corpuscular theory of light. Richardson and others have considered the matter from the standpoint of the electromagnetic-wave theory of light, and while the question is by no means settled it seems certain that the wave theory can be made to account for all the facts.

Of special importance is the study of the scattering, reflection, etc., of the various radioactive rays by matter, as this raises fundamental questions as to the structure of atoms. While the study of this scattering has led Rutherford to suggest that atoms consist of a small positively charged nucleus surrounded by electrons, interest has been added to the older hypothesis of Thomson by a paper of Crehore. He imitates the action of Thomson's sphere of positive electricity through which electrons move, by proper use of the force of gravity, and obtains photographs of some artificial but possible atoms. An interesting paper by Bohr (Philos. Mag., Nov., 1913) is devoted to a detailed working out of a Rutherford type of atom. Bumstead has added greatly to our knowledge of rays, slow moving electrons sent off from metals which are struck by a rays (positively charged particles of atomic size). Duane has studied the motion in a magnetic field of ions produced by the heavy positive a rays from radium and has concluded that while the positive ions are molecules or atoms of nitrogen and oxygen, the negative ions seem to be always electrons. Wellisch has shown that there is a definite limit to the percentage amount of active material which will be deposited on a negatively charged electrode exposed to radium emanation, the remaining active material losing its charge almost as it is formed. Work of Gray on the scattering of X and y rays indicates that there is probably very little real scattering of these rays when they pass through matter, but rather a reradiation.

Magnetism. The relatively small amount of work done in this field does not indicate that all the puzzling questions are answered, but rather, perhaps, a difficulty in connecting experiment with theory as the latter now stands. Williams has published an interesting compilation of the various attempts to form an electronic theory of magnetism by Weiss, Langevin, and others. While doubt less in the main correct in ascribing the magnetic properties of molecules of iron, etc., to the presence of electrons moving in orbits, which turn and face one way when a bar is mar

netized, the theories are still far from complete. An entirely different theory is that advocated by S. R. Williams, and though by its means he has been able to predict a number of magnetic effects, the theory seems essentially less valuable than that based on moving electrons. During the year Pierce (Proc. Am. Acad., Novem ber, 1913, p. 555) has continued his careful study of the magnetic properties of iron. The influence of a magnetic field upon either the emission or transmission of light (Zeeman and Kerr effects) has been the subject of much recent experimentation and extensive theoretical work, the latter by Voigt especially. While all work up to 1913 has shown that light emitted in a magnetic field was more complex, that is, contained more different wave lengths than light from the same source not in a magnetic field, the use of stronger magnetic fields has now brought out the unexpected fact that a very strong field simplifies the light emitted, that is, reduces the number of different wave lengths. Voigt has been able to modify his theory to account for this. As regards the passage of light through a magnetic field, or reflection from the polished pole of a magnet, C. Snow has successfully applied Voigt's theory to account in a most exact way for the observations of Ingersoll.

Conclusion. While the year 1913 has not been remarkable for any great or fundamental contribution to physical knowledge or theory, nevertheless a large amount of good work has been done, especially in the way of careful experimental testing of hypotheses. The most noticeable thing is the continued almost unquestioned acceptance of the revolutionary idea of Planck, referred to under "Electricity," that energy is emitted from radiating bodies in lumps or bundles, to put it crudely. In Science of Jan. 24, 1913, will be found an excellent summary by Millikan of the current attitude toward this idea of Planck, which has influenced every branch of physics, and in Science for Sept. 19 and 25, 1913, the more general address of Lodge, in which the ideas of continuity versus discontinuity are considered from a much more general point of view.