cially upon the transmission of light through ordinary media. The chief difference between his treatment of the subject and the classical one of Maxwell lies in the fact that Goldhammer considers what are usually called the constants of the medium, the specific resistance or the dielectric constant, for instance, not to be constants, but to be functions of the wave frequency, developable in power series. It may be remarked that this view receives a certain amount of support from the researches of Blondlot and J. J. Thomson, which show that the dielectric constants of certain materials do depend upon the frequency.

Developing mathematically the preceding hypothesis, Goldhammer arrives at very general equations for the velocity and absorption of light in a given medium. It is worthy of remark that the formulas given by Helmholtz, Sellmeier and Lommel can all be considered as special cases of that of Goldhammer and can be derived from it.

In Wied. Ann., No. 3, Band 67, Goldhammer applies the theory which has just been sketched to the Zeeman effect, and thus obtains a new theoretical explanation of the phenomenon. Considering the absorption spectra first, he shows that any alteration in the specific constants of the medium will cause a change in the position of the absorption lines. He then assumes that the magnetic field does cause such an alteration in these quantities; in support of this assumption, he calls attention to a paper by Bolzmann (Wied. Ann., 31, p. 789), in which it is shown that a magnetic field increases the resistance of a gas placed in it.

Kirchoff's laws allow one to pass from the absorption spectrum to that of emission. Then

in a bright-line spectrum the effect of a magnetic field is to cause displacements in the lines and might give rise to doublets and triplets.

In order to account for the polarization phenomena, Goldhammer makes a further assumption that the magnetic field causes the medium to become æleotropic and double-refracting. The circular polarization of the doublets, when viewed along the lines of force, is very closely connected with the well-known magnetic rotation of the plane of polarization.

It will be noticed that this theory of Gold

hammer's differs materially from those proposed by Lorentz and Larmor. Goldhammer makes the whole of the phenomena depend upon changes in the medium, while Lorentz and Larmor attribute them to the electrodynamic forces developed by the motion of electrified ions in a magnetic field.

At present it seems that the ionic is the more promising of the two theories, since it gives an explanation, incomplete it is true, of the complexity of structure of the lines and of their polarization. The numerical value of the ratio between the mass of a vibrating ion and the charge carried by it as derived from the Zeeman effect is in good agreement with that obtained by J. J. Thomson from the phenomena of cathode rays.

DAYLIGHT-PHOSPHORESCENCE.

MOURELO (Comptes Rendus, t. CXXVIII., p. 557) has made the curious discovery that sulphides of strontium, calcium, barium and zinc, prepared in a particular way, show much more brilliant phosphorescence after exposure to diffused daylight than they do after exposure to direct sunlight, and, further, that periodic exposure to diffused daylight increases very remarkably the power of phosphorescing. After being brought to this sensitive state one phosphorescing portion is able to excite phosphorescence in another non-luminous portion either when the two portions are in contact or when they are contained in separate glass tubes.

A. ST.C. D.

NOTES ON INORGANIC CHEMISTRY. AN analysis of the water of the Great, or Illecilliwaet Glacier, British Columbia, has been published in the Chemical News by F. T. Shutt and A. T. Charron. The waters were taken a few feet from the face of the glacier, and were of characteristic turbid or milky appearance. Analysis showed water of great organic purity, the free ammonia being 0.018 parts per million; albumenoid ammonia 0.027 to 0.037; nitrogen as nitrates and nitrites 0.0246 to 0.0442; chlorin 0.1; solids 12 to 30.8. On sedimentation the waters became perfectly clear, and microscopic examination of the deposit showed

An analysis of an artesian water from Derbyshire is given by John White in the Analyst, which is peculiar as containing barium, it being the first recorded occurrence of this metal in waters in this section. The well is 1,300 feet deep and 160 feet above the sea level. According to the analysis given, the water first obtained at depth of 837 feet, contained of barium carbonate 1.77 parts per 100,000; the deep water contained at first of barium chlorid 38.55 parts, and six months later 40.7; water eighty feet below the surface contained 3.03 parts. The sodium chlorid in the deep water was over two thousand parts. The author discusses the origin of the barium salt. Clowes has found minute crystals of barium sulfate in the red sandstone near Nottingham, and Dieulafait has shown barium to be a constant constituent of primitive rocks, but this does not explain the conversion of the sulfate into carbonate or chlorid. The only possible explanation, according to the author, is that the barium sulfate has been at high temperature reduced to the sulfid by coal, and this converted into the chlorid by concentrated salt solution. The carbonate is derived from the chlorid. In confirmation of this it is pointed out that barium sulfate has been found in connection with coal deposits and barium chlorid in water in the vicinity of coal mines. It is, however, not impossible that under certain conditions, such as Melikoff has shown take place between sodium sulfate or sodium chlorid and calcium carbonate in the presence of aluminum or ferric hydroxid, a reaction may take place between the barium sulfate and sodium chlorid in a concentrated solution of the latter.

PROFESSOR VEZES, of Bordeaux, has continued his work upon the oxalates and nitrites of the platinum metals, and his last contribution to the Bulletin Société Chimique is on the complex salts of palladium. A concentrated solution of potassium chloropalladite is converted by potassium oxalate into the palladooxalate, and the same salt is formed by the action of oxalic acid upon the pallado-nitrite. On the other hand, the pallado-oxalate is readily converted into the chloropalladite by

The same journal contains analyses of a series of potassium, ammonium and silver salts of the so-called osmiamic acid, by Brizard, in which the formula proposed by Joly for this acid is fully confirmed. According to this, osmiamic acid is a nitroso compound, having the formula OSO(NO)OH, and corresponds to nitroso hydroxid of ruthenium.

J. L. H.

CURRENT NOTES ON METEOROLOGY.
BLUE HILL OBSERVATORY BULLETINS.

BULLETIN No. 2 (1899) of Blue Hill Observatory, prepared by A. E. Sweetland, contains accounts of two remarkable snow storms which occurred during the past winter. The first storm, that of November 26-27, 1898, caused the wreck of 141 vessels on the New England coast, and the loss of 280 lives. It was during this storm that the steamer Portland, with about 175 persons on board, was lost off Cape Cod. The suddenness and violence of this storm were due to the rapid increase in energy which took place when a cyclone from the Gulf of Mexico and one from the Great Lakes met on the coast. The fall of snow was very heavy. On February 8-14, 1899, a severe cold wave and another heavy snowfall occurred. On February 13th, at 8 a. m., the zero isotherm extended as far south as latitude 31°. At Blue Hill the average temperature of the five days February 8-13 was 3.1° lower than the average of any successive five days since the Observatory was established. This cold wave was followed by a heavy snow storm, Iwith high winds, along the North Atlantic coast. It is interesting to note that the preceding cold wave, although it caused much suffering by its severity at the time, had one very fortunate effect. The extreme cold which had almost closed some of the harbors with ice, and the difficulty of navigation when the waves, driven by the strong westerly gale, quickly

froze on decks and rigging, had resulted in keeping many vessels in port. In consequence, but few lives were lost at sea.

Bulletin No. 3, by S. P. Fergusson, is entitled Progress of Experiments with Kites during 1897-98 at Blue Hill Observatory, and presents an admirable summary of this work. Both Bulletins are abundantly illustrated.

SNOW ROLLERS.

THE March number of Climate and Crops: New England Section notes the occurrence of 'Snow Rollers' at Grafton, N. H., on March 16th, last. This is an interesting but comparatively rare phenomenon, occasionally observed in the winter season when freshly fallen snow is rolled into balls or cylinders by the wind. At Grafton these are stated to have been rolled up in countless numbers. Some of the rollers were as large as a barrel, and the fields and hills were covered with them. Other occurrences of the same phenomenon have been noted, within recent years, at Spokane, Wash., in December, 1895; at Hartford, Conn., on February 19, 1896, and in Saline county, Kan., on January 14, 1898. At Spokane there were 'hundreds of snow cylinders of uniform size, and as perfectly formed as though they had been cast in a mould.' The rollers were from 12 to 16 inches long, and from 6 to 10 inches in diameter. At Hartford some of the rollers measured 8 inches in diameter. In the Kansas case the size varied from that of base-balls to that of half-a-bushel measures. The uniform size, often noted, may be explained by the fact that the wind rolls the cylinders of snow along the ground until they become too heavy to be moved farther. If the velocity of the wind continues about the same it is likely, other things being equal, that the rollers will have about the same size.

A COURSE IN METEOROLOGY AT OHIO STATE UNIVERSITY.

IT is a pleasure to note the establishment of a new course in meteorology at the Ohio State University, Columbus, Ohio. This course, which is being given by Mr. J. Warren Smith, Section Director of the U. S. Weather Bureau at Columbus, is required in the junior year in

the course in agriculture and horticulture, and is elective in the courses in arts, philosophy and science. It is also open to all teachers. Lectures began on March 29th, and are given twice a week during a term of ten weeks. The object of the course, as stated in the prospectus, is 'to open and outline a rational and systematic line of study of the leading facts concerning our atmosphere, and of the methods of observing and investigating the daily weather changes, and of the physical laws underlying these changes.' Davis's 'Elementary Meteorology' is used as a text-book. The lectures are illustrated by means of lantern views, and the 'laboratory work' includes the use of the ordinary instruments and practice in the construction of weather maps.

CLIMATE OF THE CONGO FREE STATE.

THERE has recently been published an admirable little pamphlet on the climate of the Congo Free State, by M. Lancaster, Director of the Meteorological Service of Belgium (Court Aperçu du Climat du Congo, 12mo., Brussels, 1899, pp. 43). This is a summary, in a very convenient form, of the meteorological portion of the volume on the climate, soils and hygiene of the Congo Free State, noticed in SCIENCE for January 13, 1899, p. 72, and is reprinted from the Annuaire de l'Observatoire royal de Belgique pour 1899. R. DEC. Ward.

HARVARD UNIVERSITY

A NEW MARINE BIOLOGICAL LABORATORY. AMERICAN biologists will doubtless be gratified to learn that the United States Fish Commission will maintain a marine biological laboratory at Beaufort, N. C., during the coming summer, and will probably undertake to establish a permanent laboratory at that place. The station will be fully equipped for a limited number of investigators and be ready for occupancy by June 1st. There will be one building devoted to laboratory purposes and another affording sleeping accommodations.

Dr. H. V. Wilson, professor of biology in the University of North Carolina, has been asked to become the director of the laboratory. Dr. Wilson was associated with the Commission at its Woods Holl laboratory for several years

and needs no introduction to the scientific world.

Beaufort is situated near one of the great ocean inlets, and the waters of the harbor and adjacent sounds are remarkably well supplied with fishes and invertebrates. The advantages this locality affords for biological research are well known, as many naturalists have from time to time resorted thereto for the study of special problems.

In the early fall Beaufort will be made the headquarters of the steamer Fish-Hawk during a biological and topographical survey of the oyster grounds of the State which the Commission will conduct at the request of Professor J. A. Holmes, director of the North Carolina Natural History and Geological Survey, and other State officials.

U. S. COMMISSION OF FISH AND FISHERIES.

HUGH M. SMITH.

THEORY OF THE STEAM ENGINE. M. NADAL, in a very extended review of the recognized 'Principles of the Mathematical Theory of the Steam Engine,' in recent issues of the Revue de Mécanique, discusses the theory of heat-exchanges between working fluid and cylinder-walls, the influence of the duration of the admission period, that of the compression and of the velocity of operation of the motor; touching upon the experimental work of Dwelshauvers-Dery. His principal conclusions are the following: *

1. The absorptive power of the metal in contact with the vapor is finite, and variable as a function of time. It is more considerable than the emissive power. The variation of this absorbing power is a function of the amount of liquid deposited upon the wall, and that amount has been shown by Donkin to vary, in the cases reported by him, from 20 calories per square meter per unit difference of temperature between metal and vapor, per second, and, at the time of admission, down to 12 during expansion and lower, and to 2 during the period of re evaporation and of emission, and to even less values as exhaust becomes complete; although this re-evaporation may be *Revue de Mécanique, 1898-9.

exceedingly rapid at the moment of opening the eduction port.

2. In the case of the unjacketed cylinder the mean temperature of the wall is equal or superior to that of the vapor in contact with it.

3. The heat surrendered by the vapor at induction increases less rapidly than does the period of action, that of induction. The indications are that the range of temperature during expansion mainly affects the quantity of the heat-exchange and that the total temperaturerange does not measure the waste, which is contrary to general opinion among engineers and physicists.

4. Compression in the clearance or 'dead spaces' is not always advantageous.

M. Nadal shows that the moisture on the wall plays an important part, augmenting the quantity of heat-waste as superheating diminishes it. It is found that the variation of the magnitude of heat-exchanges during the forward and the return stroke accounts largely for the well-established, and often large, gains due to the use of the steam-jacket; since that accessory may communicate heat rapidly and effectively during the earlier portion of the cycle, while the sluggish transfer of heat out of the cylinder wall during the period of low pressure and temperature checks the wastes that would otherwise then occur, and more extensively than in the earlier period. Thus this variation of transferring power of the wall acts as a sort of 'check-valve' for the heat received from the jacket, permitting it to act efficiently, where most needed and preventing loss of heat where its transfer could do no good and would be purely a waste. Thus the jacket, also, is most economical in those engines which would be most economical without it, those in which the interior walls of the cylinder are dry during exhaust.

R. H. THURSTON.

THE PHILADELPHIA EXPOSITION OF 1900. WE have received from the officers of the

Philadelphia Exposition of 1900 details in regard to their plans. It is their purpose to exhibit every kind of manufactured products of the United States especially suitable for export. Such exhibits will form the principal depart

ment of the Exposition and will comprise everything which is, can or might be exported, from locomotives and heavy machinery to the smallest novelties.

There will also be a department of foreign manufactured goods, but it will not contain a single exhibit shown by a foreign manufacturer. This department will consist of collections of samples of goods made in the commercial countries of Europe and successfully sold in all foreign markets in competition with American goods and in foreign markets in which American trade has not yet been developed. These samples will be exhibited side by side with American products of the same class, and will show our manufacturers just what competition they must meet abroad, as well as the peculiarities in the demands of every foreign market.

A third department of the Exposition will show how American goods must be packed, labeled and shipped in order to meet the requirements of foreign trade, which vary according to the degree of development or civilization in each country of the world.

In October a Commercial Congress will be held in Philadelphia in connection with the meeting of the International Advisory Board of the Philadelphia Commercial Museums. There is every reason to believe that at least 800 representatives of foreign firms will be present at the sessions of the Commercial Congress and in attendance on the Exposition, in addition to the official delegates and those representing commercial organizations.

The Exposition will be under the joint auspices of the Philadelphia Commercial Museums and the Franklin Institute. Sanction and support has been given to the Exposition by the National Government, Congress appropriating $350,000 to aid it. The City of Philadelphia has given $200,000, and the State of Pennsylvania $50,000, and $100,000 is being raised in Philadelphia by individual subscriptions.

The main buildings, which are now under construction, cover eight acres of ground, and the available exhibition space will be at least 200,000 square feet Outside of the space occupied by the main buildings there will be within

the Exposition grounds, which comprise a tract of fifty-six acres of land on the bank of the Schuylkill River, within fifteen minutes' ride of the City Hall, ample space for the erection of detached structures for special exhibits.

SCIENTIFIC NOTES AND NEWS. VICE-PRESIDENT BRANNER, of Stanford University, will conduct an expedition to Brazil during the summer to work upon the geology of the stone and coral reefs of the coast. These reefs, more or less broken, extend from Ceará to the Abrolhos, a distance of more than a thousand miles. Dr. Branner did much work upon these reefs while he was connected with the Geological Survey of Brazil, but the field observations were never finished and the results of the work were not published. He hopes to complete his work during the summer vacation. The expenses of the expedition will be paid chiefly by Professor Alexander Agassiz, and the results will be published by the Museum of Comparative Zoology at Harvard.

PRINCETON proposes to send a small party to observe the total eclipse of the sun which is to occur on May 27, 1900. A friend of the University has provided the necessary funds, and the special apparatus that will be needed is already being constructed. The station to be occupied is not yet finally selected, but will probably be near the boundary between North

and South Carolina, where it is crossed by the track of the moon's shadow, running northeastward from New Orleans to Norfolk, Va.

THE Iron and Steel Institute of Great Britain has conferred the Bessemer Gold Medal for 1899 on Queen Victoria in commemoration of the great progress made in the iron and steel industries during her Majesty's reign.

THE Academy of Sciences at Halle has elected Dr. Hans Lenck, professor of mineralogy at Erlangen, to membership.

SIR JAMES WRIGHT, C.B., late Engineer-inChief of the British Navy, to whom many of the improvements in British warships are due, died on April 16th in his 86th year.

THE death is also announced of Sir William Roberts, F.R.S., the eminent London physician,