Each newly discovered one is of interest and perhaps a note should be made of the occurrence of a rather large dike recently found. It has been exposed at the eastern side of the Portland cement quarry east of Shurger Point, six miles north of Ithaca. It is the first of the Ithaca region dikes found in limestone and is exposed for the height of the Tully limestone at the north and south walls of the quarry and in the shales along the quarry bed.

No contact action was noticed. In places there is a thin calcite streak at the side of the dike, in others there is a tight contact between dike and wall rock. Striæ on the calcite gave evidence of horizontal movement. The dike varies in width from 11" to 18" and is decidedly green, due to the serpentine in it. It strikes about N 3° E., parallel to the dip joints, like all the dikes near Ithaca. There may be some connection between this dike and a group of smaller dikes east of Ludlowville, two miles to the north.

CORNELL UNIVERSITY

PEARL SHELDON

THE HAWAIIAN OLONA

TO THE EDITOR OF SCIENCE: In SCIENCE1 for September 10, 1920, p. 240, Mr. Vaughan MacCaughey again calls attention to the remarkably durable fiber of the Hawaiian Oloná, and quotes Dr. N. Russel's rather inaccurate account of the people making the fiber and its products, fish nets and cords, some used especially for fish-lines. In view of the possible importance of this product, it seems worth while to correct certain statements. The name of the bird caught for its yellow feathers was O-o not O-u. As late as 1864, when the present writer first visited the Hawaiian Islands, there were some natives at Olaa still beating the mamake kapa and twisting the oloná fiber on their thighs. On the island of Molokai, as late as 1889 a photograph was taken of a native scraping the fiber. Surely Mr. MacCaughey must be aware that in the Bishop Museum in Honolulu, is a fine cast from life of a native preparing this fine fiber, and there are 1 N. S., Vol. LII., No. 1341.

many specimens of both the raw material, the finished product and the laau kahioloná or scraper which was sometimes a shell papaua (Meleagrina margaritifera) but more commonly a sharpened bone from the back of the honu, a sea turtle not a (fish, as Dr. Russel has it). The boards were made of any hard wood; the naou of Dr. Russel was perhaps the naio, or bastard sandalwood.

As a specimen of the remarkable durability of the fiber, there is in the Bishop Museum a ball of fish-line used by the Kamehamehas for a hundred years and it is still in perfect condition.

WILLIAM T. BRIGHAM

QUOTATIONS

PROFESSOR MICHELSON ON THE APPLICATION OF INTERFERENCE METHODS TO ASTRONOMICAL MEASUREMENTS

THE first information Professor A. S. Eddington, Plumian professor of astronomy at Cambridge University, received that his theoretical deductions concerning the angular diameters of certain stars and of the Betelgeuse, in particular, had been confirmed by Professor Michelson [in his paper at the Chicago meeting] was from a cable message from the New York Times. He was extremely interested and delighted at the results obtained and is anxiously awaiting full details.

Talking to the New York Times correspondent he pointed out that many years ago Professor Michelson suggested a plan for measur

ing, at any rate to a much greater degree of accuracy than before, diameters of stars by the wave theory of light.

"For some time now," he said, "they have been carrying on these experiments at Mount Wilson, and I presume that it is there that these most interesting results have been obtained. The great difficulty that they have had to contend with has, of course, been what is known as atmospheric tremor. They have been trying Michelson's methods and previously had obtained some very interesting results, but these were only with regard to very close double stars. By this means they got some very successful results with double stars, but when they

came to try to determine the angular diameter of stars they were up against a very much more difficult problem. I knew that they were working on these lines, but this is the first word I have heard of the results.

"At a meeting of the British Association I delivered a presidential address to the mathematical and physical sections, and made reference to the fact that this experiment which was being carried out would be of the very greatest importance. We have of course had theories, and, working on those theories, I gave a table of what I thought would be the angular diameter of certain stars, and I am delighted to find that the figures so nearly correspond. This would seem to show the theories have been on the right side.

"In particular, I noticed that Betelguese's diameter is 260,000,000 miles, which is enormously larger than the sun. That is a very interesting confirmation of the theory of Russell and Hertzsprung of giant and dwarf stars, giving direct evidence that Betelguese is one of the inflated stars and very different from the sun."

Dr. A. C. Crommelin, chief of staff of the Greenwich Observatory, was interviewed today on Professor Michelson's discovery by The Evening Standard and expressed the interest the experts in England's principal observatory took in it.

"Star diameters have been calculated hitherto," he said, "but have never before been actually measured. Michelson's announcement that he has measured Alpha Orionis and found it to have a diameter of 260,000,000 miles, 300 times bigger than the sun, is hopeful.

"That the distance from the earth of such a star as Alpha Orionis, which is 900,000,000,000,000 miles away, should have been measured so long ago and the size of the star should remain unmeasured seems strange, but it was explained at the offices of the Royal Astronomical Society that the two measurements have to proceed on entirely different lines.

"The Astronomical Society confirms Dr. Crommelin in the expectation of good results from Professor Michelson's work. For some time past he and his work have loomed increas

ingly large in the astronomical world.-Cablegram to the New York Times.

CAUSES OF CLIMATIC OSCILLATIONS IN PREHISTORIC TIME, PARTICULARLY IN THE ICE AGE1

IN 1918 Professor Arldt, of Radeberg, grouped the theories and weighed the evidence which had been proposed by 117 scientists in the past sixty years on the causes of the glacial and interglacial epochs. As none of these hypotheses are in all respects satisfactory, in his opinion, or can claim to explain thoroughly all paleo-climatic phenomena, he does not recognize any one theory or group of them. This is not surprising since the fundamental conclusions underlying these hypothesis have not been reached.

In this paper of twenty-seven pages, Arldt does not give an exhaustive explanation of the numerous hypotheses which have been proposed but a brief statement concerning the most important groups among them. He distinguishes two classes, Cosmic and Telluric, with three subdivisions for the first: Universal, Solar and Telluro-Cosmic; and five for the second: Dislocation of the Poles, Atmospheric, Intra-Telluric, Actologic and Orographic. Although discussions and opinions are to be found under each of these headings, his main contribution appears in crystallized but abbreviated form in his conclusion, thus:

Among numerous theories explaining the changes in climate of the earth, those should be given preference which are based upon the bypothesis that the factors which are of importance to-day in determining climate have always been effective. . . . Most importance is attached to Ramsay's theory which emphasizes most strongly the direct and indirect action of the mountains. Besides these orogenetic forces other elements, as enumerated below, probably aided in the generation of the ice ages.

1. The rise of extensive mountains (Ramsay). 2. The formation of ocean basins (Arldt). 3. The sinking of the entire ocean floor and the 1 Theodore Arldt, "Die Ursachen der Klimaschwankungen der Vorzeit, besonders der Eiszeiten,’ Zeitschrift für Gletscherkunde, Band XI., s. 1-27, 1918.

10. Distribution of land and sea according to Kerner's view.

Pliothermal or warm interglacial periods occurred under the following circumstances:

1. Chiefly as a result of low flat continents,

2. Through absence of deep basins,

3. Rising of sea floor and depression of continents, 4. Volcanic inactivity,

5. Great eccentricity of the earth's orbit,

6. Passing of the solar system through regions of the universe abounding in stars,

7. Great radiation of heat from the sun, 8. Great inclination of the ecliptic,

9. Increase of the carbon dioxide content of the air.

It is utterly improbable that the interior of the earth contributed to the climatic changes. Polar dislocations are also out of the question, so long as we can not prove that they followed any particular direction. All attempts, likewise, at explaining change in climate from one cause alone are futile. Although at first sight these theories may appear attractive, they can not stand the test of keener criticism. Moreover, nature is too complex to permit its being compressed into a single formula.

CHESTER A. REEDS AMERICAN MUSEUM OF NATURAL HISTORY

SPECIAL ARTICLES OBSERVATIONS ON THE ACCUMULATION OF CARBON DIOXIDE FROM STRAWBERRIES IN REFRIGERATOR CARS

DURING the years 1918 and 1919 and in connection with shipping tests of strawberries in

refrigerator cars being made under the direction of Mr. H. J. Ramsay and Mr. V. W. Ridley then of the Bureau of Markets, the writer was able to make observations on the carbon dioxide and oxygen content of the air in refrigerator cars and the effect of ventilation on the accumulation of carbon dioxide. A brief summary of the results follows:

The percentage of carbon dioxide and of oxygen was determined by means of a commercial Orasatt apparatus-samples being drawn through lead tubing one end of which was placed in the part of the car from which it was desired to take samples and the other run out at one corner of the door. During transit analyses were made at icing stations and at other times when the train stopped long enough. Duplicate analyses were made when time permitted, and in all cases several hundred cubic centimeters were thrown away before the sample was drawn for analysis.

The results of the analyses made during three of these tests are summarized in Table I. In the tests of 1918 the berries were loaded at a temperature of about 68° to 70° F. and one car was ventilated by raising the The hatches at diagonal corners of the car. berries loaded in the test of 1919 were at a temperature of 76°-78° F., one car being ventilated by raised hatches, and the other that reported in the last column of Table I., by two six-inch pipes, installed at opposite ends of the car.

From the table it is apparent that there is no great accumulation of carbon dioxide in the air of the unventilated cars in transit. The maximum amount 2.5 per cent., was reached in a car loaded at Monett, Mo., seven hours after the doors were closed. When the car was re-iced the carbon dioxide content dropped to 0.7 per cent. From this it increased again to 1.3 per cent., but at the next icing it dropped to 0.6 per cent. and never exceeded this amount during the remainder of the trip.

The accumulation of carbon dioxide in unventilated refrigerator cars loaded with strawberries has been found not to exceed 2.5 per

the Forest Products Laboratory indicate that a cattle food can be prepared from eastern white pine sawdust and that it has considerable food value. The cattle food was prepared by digesting the sawdust for 15 to 20 minutes with 1.8 per cent. sulphuric acid at a steam pressure of about 120 pounds per square inch. After cooking the sugars were extracted with water and the acid removed from the solution by means of lime. The liquor containing the sugar was evaporated to a thick syrup and mixed with the digested residue which had been previously dried. The whole was then dried to a moisture content of 15 per cent. It is shown that when the product has a greater moisture content than 15 per cent. the keeping qualities are not good. Leaching experiments removed all but 2.04 per cent. of the total acid and all but 7.16 per cent. of the total sugar. A comparative analysis of the wood before and after conversion shows that cold water soluble, hot water soluble and NaOH soluble substances are greatly increased but the ether soluble substances are almost unchanged by the treatment. The pentoses are reduced 46.4 per cent. while the methyl pentosans are not affected. The total cellulose is reduced by 21.68 per cent. The sugars produced correspond to 71.5 per cent. of the cellulose removed by the digestion. The lignin content is unchanged. The crude fiber corresponds to about 75 per cent. of that in the original wood. The cellulose is greatly altered by the treatment. Practically the whole of the cellulose obtained is soluble in 17.5 per cent. alkali. It is reprecipitated from the alkaline solution by dilution with water. After filtering no precipitation of beta-cellulose is obtained upon acidification with strong acetic acid.

A comparison of wood cellulose and cotton cellulose: S. A. MAHOOD and D. E. CABLE. Samples of wood cellulose and cotton cellulose which had been subjected to various conditions of cooking and bleaching were analyzed by determining a number of constants on them, including ash, moisture, alkali solubility, pentosan and methyl pentosan content, methyoxy content, ether extract, cellulose, lignin and "copper number''; for the purpose (1) of following the changes which take place in wood cellulose on successive cooking and bleaching treatment with a view to increasing the yields of purified cellulose by varying these conditions and (2) to determine so far as possible the points of similarity or difference of cellulose from wood and

that from cotton. The data show that wood cellulose most nearly corresponding to cotton, taking munition linters as a standard, is obtained by recooking "easy bleaching" sulphite pulp with soda and bleaching with two per cent. of bleaching powder. The practise of checking wood cellulose according to the specifications for cotton is a questionable procedure.

Supply and preparation of wood for the manufacture of pulp: HUGH P. BAKER.

Parchmentizing paper and the reaction of mordants: J. E. MINOR. The work of Schwalbe and Becker recently published confirms theories of the author as to the reactions which occur on the decomposition of cellulose during beating. The hydrolysis of cellulose forms dextrines which are mucilaginous, soluble in water and easily reduce Fehling solution. These dextrines are readily adsorbed by pure cellulose thus constituting the reactive, insoluble, colloidal hydrocellulose, and in this position they catalyze further cellulose hydrolysis. Complete hydrolysis leaves only soluble dextrines or sugars. For a mucilaginous accumulation as is desired for making parchment paper, the velocity of the initial reaction must be catalyzed, either by the hemi-celluloses of wood incrustation, or by acid treatment of pulp prior to beating. Acid treatment of pulp reduces the time required to beat to mucilage to one half or one tenth the original amount required, and increases the strength of the paper made from it. Pulp mucilage has a greater power of splitting salts and adsorbing metal ions than has cellulose and, therefore, assists in holding size, dye, coating, etc. Metal impregnation weakens paper strength.

Is it feasible to form a section of cellulose chemistry? G. J. ESSELEN, JR.

Solvents for phosgene: CHARLES BASKERVILLE. An impelling factor in causing the Germans to sign the Armistice was a knowledge of the rapid development of the stupendous poison gas program of the United States. One of the earliest gases, a real gas, used by the Germans, was phosgene. It had been manufactured on a small scale in Germany before the war. Small amounts were exported to the United States prior to 1914. It was being produced in this country for similar uses after the blockade had shut off the importation of chemicals in the manufacture of which the Germans had made a specialty. It was transported in small quantities liquefied in cylinders. When hostilities ceased we were producing the

poison gas for war purposes at the rate of fifty tons a day, with a program nearly completed for a much larger production. Soon after the Armistice was signed restrictions were imposed which prevented railroad transportation of phosgene liquefied in cylinders. Uses for war gases in peace times have been sought. The author has found several solvents for the gaseous phosgene, among them gasoline, benzine and ethyl acetate, which dissolve an equal weight of the poison gas at ordinary temperatures. The solutions exert practically no pressure beyond atmospheric, so may be transported with safety within ordinary closed containers. On exposure to the air the liquid and dissolved gas evaporate. The solutions offer simple means for the use of a poison gas to exterminate rats and moles. On warming the solutions in suitable vessels the phosgene is liberated and may be used in a pure form in the manufacture of certain coal tar products of color or medicinal value.

An electrometric method for detecting segregation of dissolved impurities in steel: E. G. MAHIN and R. E. BREWER. Previous work has shown that both non-metallic and metallic inclusions cause carbon segregation in steel and the hypothesis has been advanced that this is due to the steel acting to some extent as a solvent for the foreign material. If this is correct the electrode potential of the metal should be altered in the regions immediately surrounding the inclusion. It is not possible to determine this point by the usual method involving immersion of the specimen in an electrolyte but a method has been devised for exposing microscopic areas of metal surface to an electrolyte, so that the micro-electrode thus formed may be connected with a standard calomel halfelement and the E.M.F. of the system measured by the usual compensation method. This method has been applied to an investigation of the ferrite bands produced by heating steel in contact with metallic inclusions; it has been found possible thus to measure the potential of these micro-areas and to establish the fact that the ferrite of these segregated portions possesses a distinctly lower potential than the ferrite of the body of the steel. The work is being extended to include investigation of the potential of ferrite adjoining other metallic and non-metallic inclusions and it may be applied also to the detection of segregation of the constituents of non-ferrous alloys.

Soda-lime for industrial purposes: R. E. WILSON. The work to be described in this paper was an