Search Documents

HENRY M. HOWE, Bedford Hills, N. Y. (communication to the Secretary *).-Dr. Burgess would add to the value of this important paper if he would give us the carbon content of the steel made in the several charges, or, if this cannot be done, if he would indicate whether, as might he inferred from his paper, the teeming temperature is approximately constant and independent of the carbon content of the steel, or whether as would be expected, it varies with the content and is materially higher for low- than for high-carbon steel. If he could add to Table 3 the time of the several tests, that would be welcome. His Table 3 leads one to infer that, once the bath is melted somewhat more than flat, the temperature of the roof is nearly constant and is approximately the highest compatible with reasonable longevity. That is to say, the metal is heated by means of a roof and flame at nearly constant temperature. The temperature of the observed upper surface of the slag should reflect that of the roof approximately, and in Table 3 this seems to be true, the temperature of both rising toward the middle of the record, and again falling. The slag as the medium through which the heat passes from roof to metal, should be intermediate in temperature between the two, as it is here. But the temperature of the metal ought to rise continuously, unless when cold additions are charged, and this too seems to be. true, allowing for slight observational errors, the total rise from the first to the last spoon-test being 89°. The temperature drop of from 85° to 135° from the last spoon-test to the teeming is surprisingly great, suggesting that the spoon-tests represent the metal immediately beneath the slag, and therefore appre-

Jan 6, 1917

By H. LeRoy Scharon

The electrical resistivity method was utilized at 245 road cuts and structure sites over the 166 miles of the Philadelphia and Western Extensions of the Pennsylvania Turnpike System. Varying geologic structures and rock types afforded an excellent academic study of the method as well as a practical application to modern highway engineering.

Jan 4, 1951

By Robert B. Paul

This paper describes the new lead smelter and refinery recently completed by the American Smelting and Refining company in Southeast Missouri.

Jan 1, 1970

By C. A. Hartnagel

In 1939 the production of crude oil in New York totaled 5,105,000 bbl. This marks the third consecutive year production of crude oil has exceeded 5,000,000 bbl. and only once has this total been surpassed—in the year 1882, three years after the discovery of oil in Allegany County, when the annual production was more than 6,000,000 barrels. The most gratifying feature of the industry was the increased demand for lubricants, which resulted in advanced prices for crude. From a price at the beginning of the year of $1.68 per barrel, six advances totaling 82c brought the year's closing to $2.50 per barrel. A further increase in price on Jan. 1, 1940, to $2.75 per barrel places the local industry in a strong position for the year's operations. Most of the oil production of the state is now obtained by water-flooding, and the industry is practically stabilized. No new pools have been discovered in over 25 years, and border drilling indicates little prospect of extending the boundaries of the fields beyond their present limits. Deep drilling to the Oriskany sandstone in and close to the present oil fields hay failed to find oil in substantial amounts. Altogether, the oil pools of New York have produced during the 60 years that they have been exploited approximately 125,000,000 bbl. of oil. Following the flush production of 1882, the annual production gradually dccrcased and the year 1912 marked an all-time low of only 782,000 bbl. In 1919, as a result of flooding operations, a definite upward trend in oil production was recorded. During that year also the production of oil by water-flooding methods as approved by Legislative action. In the 21-yr. period beginning with 1919, oil production in the state ha.: totaled 60,338,000 bbl. Thus it will be seen that by the aid of flooding methods nearly as much oil has been produced in the last 21 yr. as in the preceding 40 yr., although the Iatter period included the flush production in 1882. As there is little hope of extending the area of the present producing oil fields of the state, producers and refiners are concerned regarding the future life of the fields. It is encouraging to observe that earlier esti- Published by permission of Director of New York State Museum, Albany, N. P. Summarized at meeting; manuscript received at the office of the Institute March 20. 1940.

Jan 1, 1940

By C. A. Hartnagel

In 1939 the production of crude oil in New York totaled 5,105,000 bbl. This marks the third consecutive year production of crude oil has exceeded 5,000,000 bbl. and only once has this total been surpassed—in the year 1882, three years after the discovery of oil in Allegany County, when the annual production was more than 6,000,000 barrels. The most gratifying feature of the industry was the increased demand for lubricants, which resulted in advanced prices for crude. From a price at the beginning of the year of $1.68 per barrel, six advances totaling 82c brought the year's closing to $2.50 per barrel. A further increase in price on Jan. 1, 1940, to $2.75 per barrel places the local industry in a strong position for the year's operations. Most of the oil production of the state is now obtained by water-flooding, and the industry is practically stabilized. No new pools have been discovered in over 25 years, and border drilling indicates little prospect of extending the boundaries of the fields beyond their present limits. Deep drilling to the Oriskany sandstone in and close to the present oil fields hay failed to find oil in substantial amounts. Altogether, the oil pools of New York have produced during the 60 years that they have been exploited approximately 125,000,000 bbl. of oil. Following the flush production of 1882, the annual production gradually dccrcased and the year 1912 marked an all-time low of only 782,000 bbl. In 1919, as a result of flooding operations, a definite upward trend in oil production was recorded. During that year also the production of oil by water-flooding methods as approved by Legislative action. In the 21-yr. period beginning with 1919, oil production in the state ha.: totaled 60,338,000 bbl. Thus it will be seen that by the aid of flooding methods nearly as much oil has been produced in the last 21 yr. as in the preceding 40 yr., although the Iatter period included the flush production in 1882. As there is little hope of extending the area of the present producing oil fields of the state, producers and refiners are concerned regarding the future life of the fields. It is encouraging to observe that earlier esti- Published by permission of Director of New York State Museum, Albany, N. P. Summarized at meeting; manuscript received at the office of the Institute March 20. 1940.

Jan 1, 1940

By Tormod Forland

FROM equilibrium measurements on the system NaF-Na 3A l Fo-A 1 2 O 3-Na CO 3-CO 2 with high contents of NaF, Forland, Storegraven, and Urnesl concluded that complexes are formed containing two oxygen, three aluminum, and a less defined number of fluorine per complex. The same number of oxygen atoms per complex was also found by cryoscopic measurements for mixtures high in NaALF 6. In a recent paper by Stokes: some other complex ions, Al2O2F42- and AlOF2-, have been suggested. Here an additional argument in favor of the complex ALO2F. suggested by Forland, Storegraven, and Urnes will be given. When fused Na3AlF6 containing some A12O3, is cooled slowly, the solid phase will consist of the two components as separate phases. The melting point and the heat of fusion of Al2O3 are so high that this component will form the stable solid phase even though in the liquid state the complex may have a high stability. If, however, the fused mixture is quenched, the decomposition of the oxygen-containing complex and the formation of solid Al2O3 does not proceed fast enough, and a solid solution of A12O3 in Na3AlF6 is obtained. Zintel and Morawietz3 obtained cryolite containing 10 pct AL2O3 in this way. The density and the lattice parameters of the A12O3 containing cryolite were also measured, and from this it was concluded that the Al2O3 enters the cryolite lattice as Al3AlO0 groups replacing Na3AlF6 groups. However, the experimental data could as well be explained by assuming that one Na+ ion and two of its neighboring F- ions are replaced by one Al3+ ion and two O2- ions. This may be pictured schematically as: F F F F F F F Al F Na F Al F ,-Na+ + Al8+ F Al 0 A1 0 A1 F F F F —2 F- + 2O2- F F F solid cryolite cryolite-alumina solid solution This means that the complex in the solid solution, derived from X-ray and density measurements, is the same as that which Forland, Storegraven, and Urnes derived from equilibrium measurements in the liquid state. References IT. Forland, H. Storegraven, and S. Urnes: Ztsch. fur anorgan-ische und allgemeine Chemie, 1955, vol. 279, p. 205. 2J. J. Stokes: AIME Trans. To be pubhshed. s E. Zintel and W. Morawietz: Ztsch. pubhshed.fur anorganische und allgemeine Chemie, 1939, vol. 240, p. 145.

Jan 1, 1958

By E. H. Kelton, R. D. Grissinger

In designing structural members of steel and some other materials the design engineer has available recognized values of elastic modulus and safe working stress that may be substituted in well-known engineering formulas to give him desired cross-sectional areas and deflections for a particular part. Zinc, on the other hand, has no generally recognized modulus. It flows under continuously applied loads. Therefore, the designer of structural members must use a range of working stresses depending on the continuous load and the amount of permissible distortion. The only method of obtaining these values is by creep testing. Creep testing of zinc has been carried on for some 20 years by this laboratory, covering a wide range of zinc and zinc alloys. Types of test included tension, compression, and simple and cantilever beams at temperatures ranging from o to I50°C. Specimen conditions ranged from as-cast or as-rolled to accelerated aged in steam or dry air. Effects of plating were investigated also. It is not the intention of this paper to expound all these factors but rather to present the final data on well-known zinc die-casting alloys and to suggest means of applying these data to engineering design. Early Work In the early work on creep testing of zinc, the direct-loading tension method was used. However, direct loading of die castings was unsatisfactory because of the high loads required. For instance, a standard die-cast tension bar having a cross-sectional area of about 0.049 sq. in. would require a direct load of about 1000 lb. for a unit stress of 20,000 lb. per sq. in. Also, tests were made at nominal room temperature, but normal variations caused fluctuations greater than those due to experimental error or other causes, so that control of this variable within ±1/2°C. seemed essential. These two factors led to the development of an apparatus described by J. Ruzicka,' in which the specimens were loaded by means of a lever and were held at constant temperature by means of an oil bath. While Ruzicka's apparatus yielded excellent check results, it seemed desirable to have some setup by which literally hundreds of die-cast specimens could be tested at the same time. This would permit the conducting of a number of low-load tests designed to run 10 years or longer and still leave capacity for current high-load work. Beam testing in a constant-temperature room was the answer. Test Method Preliminary work indicated that die-cast beams of various lengths and cross scctions yielded the same result within reasonable limits regardless of whether they were tested as simple beams, uniformly loaded beams, or cantilevers. Accordingly, the simplest type was adopted. This was a,

Jan 1, 1945

By Fred C. Bond

THE process of comminution by grinding is properly classified as an art, rather than as a science. Like most other operations concerned in ore dressing, or in the treatment of nonmetallic minerals, the mathematical relationships of rock breaking are still obscure, and much disputed. If some basic formula exists, comparable for instance to E = RI in electricity, it has not yet been placed in, common use. The absence of such a comprehensive mathematical background undoubtedly is the cause of many inefficient designs and installations, as well as of much poorly oriented experimental work. On the other hand, the mathematical immaturity of the process constitutes an opportunity for a real contribution on the part of those who enjoy mental pioneering. Some of the statements made in any such exploratory work may be little more than opinions, and it is to be expected that some will later be proved untrue; but such an evolutionary development is necessary if we are ever to arrive at the fundamental laws, the discovery of which can change the empirical art into a mathematically exact science. The most significant principle of comminution that we possess is probably Rittinger's law, which states that the work done in crushing or grinding is proportional to the area of the new surface produced. There is an alternative theory, known as Kick's law, which formerly was preferred to Rittinger's but which now is rather discredited. According to Kick, the work done in comminution is proportional to the volume, or weight, of the pieces broken, or to the cube of their diameters, while according to Rittinger it is proportional to the square of the diameters. Careful determinations of size distribution below the limit of sieve sizes1 indicate strongly that under ordinary grinding conditions the amount of new surface area produced is directly proportional to the net power input. Changes in the size ratio between the balls and the particles cause important differences in the proportion of the product that passes a given mesh size, but do not affect the total amount of new surface area; they merely change the distribution of the surface among mesh sizes. The principle of constant surface area is the logical result of Rittinger's law, and it may well have an important bearing on the comminution theories of the future. FORMULA FOR BALL WEAR Grinding tests have been made in the Allis-Chalmers laboratory for more than 9 years, using a standard charge of mixed balls of different sizes; and a continuous record has been kept of the rates of wear of each size. These results have been combined and plotted in various ways in order to discover the relationship that exists

Jan 1, 1940

By George Hale

(The following extracts from Mr. Hale's paper have been made by the Editor, as being of particular interest to our members.) In an address delivered on May 28, at the kind invitation of the Engineering Foundation, I briefly sketched "The War Activities of the National Research Council. "' The wide scope of my subject forced me to touch very lightly upon engineering; I therefore beg permission to return to this phase of the subject. As shown in my earlier address, the' charter membership of the National Academy of Sciences, organized in the midst of the Civil War,

Jan 10, 1918

By Jorge Munoz Reyes

There was no development work in Bolivia during the year 1935, and oil was produced only to meet the needs of the army and of the local market. The refineries in the districts of Camiri and Sanandita treated 25,978,737 liters (163,388 bbl.) of crude oil, which yielded 4058 bbl. of aviation gasoline (Camiri only), 7381 bbl. of gasoline, 718 bbl. of kerosene (Camiri only) and 83,105 bbl. of fuel oil.

Jan 1, 1936

By Jorge Munoz Reyes

There was no development work in Bolivia during the year 1935, and oil was produced only to meet the needs of the army and of the local market. The refineries in the districts of Camiri and Sanandita treated 25,978,737 liters (163,388 bbl.) of crude oil, which yielded 4058 bbl. of aviation gasoline (Camiri only), 7381 bbl. of gasoline, 718 bbl. of kerosene (Camiri only) and 83,105 bbl. of fuel oil.

Jan 1, 1936

By J. E. Forciea, R. W. Salmi

Operating and research personnel at U.S. Steel's Pilotac concentrator have evolved specifications for magnetic separators suitable for use in beneficiation of Mesabi magnetic taconite. These specifications were developed as a result of investigations into the operating and metallurgical problems at the Pilotac concentrator over a period of some ten years. This paper discusses primary magnetic separator specifications, but much of the information is also pertinent to secondary and tertiary magnetic separators. Of particular significance is the proposal of a new basis for specifying the magnetic field which includes consideration of both field strength and field gradient. Other areas of discussion include flowsheet service, operating conditions such as drum speed and pulp level, tank configuration, magnet strength including field gradient, sleeve bearing design, designation of corrosion resistant materials and other details that result in a metallurgically efficient machine that has low repair and maintenance requirements. The mechanical and metallurgical design of magnetic separators at the Pilotac plant, Minnesota Ore Operations, United States Steel Corp. has progressed from the original single-stage, belt-type machines to the present multiple-stage, drum-type units, In ten years of operation, the plant operating and research personnel, in consultation with manufacturers, have made design changes to improve mechanical, electrical and magnetic features to meet process needs as flowsheet designs developed and to incorporate new permanent magnet materials as they became available. Separator specifications were an end result of this effort. The latest specifications which were developed for the Minnesota Ore Operations' plant were not all-inclusive listings; they covered only those points indicated by plant experience. For example, magnet pole piece design and dimensions were not stated. The manufacturers were given freedom to use any design as long as the field pattern was as specified. The selection of the items to be included in the specifications was undoubtedly influenced by particular conditions at Pilotac such as composition of plant water, concentration characteristics of the ore, repair and maintenance facilities available and the research programs that had been conducted. Only the developments leading to the specifications for a primary or "cobber" magnetic separator are presented in this paper. However, the same general concepts are used to specify secondary and tertiary separators. The areas of interest discussed include flowsheet service, tank configuration, magnet design, bearings, materials of construction, and some minor mechanical details. FLOWSHEET SERVICE The primary magnetic separators in the Pilotac Concentrator flowsheet are located after the first stage of grinding as shown in Fig. 1.

Jan 1, 1965

In accordance with the requirements of our By-Laws I beg leave to submit herewith the report of the Library Committee for the year 1915. The activity shown by the sale of the institute publications during the year continues satisfactory, although not as large as last year. The total sales amount to $6,666.81, distributed as follows: Sale of Transactions...:: :. $4,585.42 Sale of Special Editions 553.17 Sale of Pamphlets and Bulletins 2,028.22 It should-be borne in mind that the sale of special editions was only begun last year,, and most of the members who did not have the special editions immediately bought them for their libraries. This year only a little less than half of the same number have been disposed of, and it is more than probable that from now-on the number will be considerably lessened.

Jan 2, 1916

By Allan D. Taylor, Robert S. Shoemaker

Adopting as a principle the phrase "the mill of yesterday should not be built with the high cost money of today," the authors have designed a hypothetical concentrator which could be built for a typical southwestern United States (or other similar area) copper ore. Many of the design features could also be used for other ores in other climates. Functional design, large equipment, and outdoor type of construction have been combined in a concentrator (which includes all ancilliary facilities such as truck and maintenance shops, offices, warehouse, nominal tailings disposal, and water supply) which could be built today for substantially less than the cost of comparable but conventional plants.

Jan 1, 1973

By Walter H. Bucher

This symposium is designed to lay the basis for a general discussion of the place of geophysics in the training of geologists. As there is danger that in the ensuing debate individual interests may be given emphasis out of proportion to their place, we shall do well to view the whole field of geological inquiry in order to see the problem in true perspective. The Nature of Geological Inquiry "Basic" and "Complex" Sciences It is convenient to divide all science into two major divisions, the basic and the complex sciences. The basic sciences separate out from reality such entities as "substances" and "movements," describe their behavior in terms of arbitrary units of measurement and derive from them by progressive abstractions concepts and patterns of behavior that exist between them, so-called general "laws" and "properties" of nature. The complex sciences, on the other hand, deal, with little or no abstraction, with those objects of reality that are accessible to man in the atmosphere, the hydrosphere, the lithosphere and the biosphere. Their work comprises two types of inquiry. One is concerned with abstractions to the extent of dealing with types rather than individuals—that is, with such things as plants, animals, protoplasm; clouds, streams, gla- ciers; volcanoes and mountain ranges. It comprises the search for general properties and patterns of behavior or "laws" that characterize the objects and their reactions to each other. These properties and laws apply always, everywhere. They are independent of the stream of time. We may call the results of this sort of inquiry timeless knowledge. The other type of inquiry is concerned with the objects themselves—that is, with this animal and that plant; this river system and that mountain chain; this glacier and that valley. Like all concrete objects in nature, they are subject to change with the passage of time, especially when it is measured in years, centuries, eons. The characteristics by which we recognize them —indeed, their very existence—are dependent on their position at this point in the stream of time. This we may call time-bound kntowledge. The so-called descriptive sciences are all bound to the past of which the objects of their study are the product and through which alone their present properties call be understood. They are historical sciences. This obvious fact is easily overlooked, which leads to serious misunderstanding, as we shall see later. Dynamic Geology With this broad picture in mind, let us now examine the nature of geological inquiry, defining, for the purposes of this discussion, geology as the science of rocks. Air is not a rock; water is. Meteorology and climatology are, correspondingly, gen-

Jan 1, 1946

By Walter H. Bucher

This symposium is designed to lay the basis for a general discussion of the place of geophysics in the training of geologists. As there is danger that in the ensuing debate individual interests may be given emphasis out of proportion to their place, we shall do well to view the whole field of geological inquiry in order to see the problem in true perspective. The Nature of Geological Inquiry "Basic" and "Complex" Sciences It is convenient to divide all science into two major divisions, the basic and the complex sciences. The basic sciences separate out from reality such entities as "substances" and "movements," describe their behavior in terms of arbitrary units of measurement and derive from them by progressive abstractions concepts and patterns of behavior that exist between them, so-called general "laws" and "properties" of nature. The complex sciences, on the other hand, deal, with little or no abstraction, with those objects of reality that are accessible to man in the atmosphere, the hydrosphere, the lithosphere and the biosphere. Their work comprises two types of inquiry. One is concerned with abstractions to the extent of dealing with types rather than individuals—that is, with such things as plants, animals, protoplasm; clouds, streams, gla- ciers; volcanoes and mountain ranges. It comprises the search for general properties and patterns of behavior or "laws" that characterize the objects and their reactions to each other. These properties and laws apply always, everywhere. They are independent of the stream of time. We may call the results of this sort of inquiry timeless knowledge. The other type of inquiry is concerned with the objects themselves—that is, with this animal and that plant; this river system and that mountain chain; this glacier and that valley. Like all concrete objects in nature, they are subject to change with the passage of time, especially when it is measured in years, centuries, eons. The characteristics by which we recognize them —indeed, their very existence—are dependent on their position at this point in the stream of time. This we may call time-bound kntowledge. The so-called descriptive sciences are all bound to the past of which the objects of their study are the product and through which alone their present properties call be understood. They are historical sciences. This obvious fact is easily overlooked, which leads to serious misunderstanding, as we shall see later. Dynamic Geology With this broad picture in mind, let us now examine the nature of geological inquiry, defining, for the purposes of this discussion, geology as the science of rocks. Air is not a rock; water is. Meteorology and climatology are, correspondingly, gen-

Jan 1, 1946

By B. C. Craft, T. J. Johnson, H. L. Kirkpatrick

Petroleum engineers are displaying considerable interest in the problems of cementing oil wells, especially in the Gulf Coast and California areas, where steep temperature gradients are encountered and where wells are being drilled to greater depths. Deep drilling, that is, at depths in excess of 7000 ft., has been undertaken in these areas only in the last few years and with this trend oil-well cementing requires more technical treatment. It is not the purpose of this paper to describe the scope and application of cement operations in the oil industry or to give the standard specifications, methods of sampling and testing of portland cement. The various testing methods have been described by the American Society for Testing Materials1, Uren2, Tickell and Wankowski3 and Rakestraw and Parsons4. The latter authors (pp. 50-58) have also described in detail the methods of cementing. The authors realize that it is impossible to reproduce in the laboratory the scale of forces and dimensions that exist at great depths, but an attempt has been made to scale down these dimensions proportionally. The selection of problems to be studied was based largely upon the practical questions asked by several of the largest operators in these fields. All cements used in these experiments met the specifications of the American Society for Testing Materials.

Jan 1, 1935

By B. T. Brady

Rock mechanics, like all engineering disciplines, must have a theoretical foundation. The subject of this chapter is the formulation of analytical methods that may aid in the rational design and determination of the stability of engineering structures in competent brittle rock materials. By an engineering structure in rock, we mean any (1) excavated sub- surface opening, (2) open surface structure, or (3) system of openings in rock that is virtually self-supporting-that is, not requiring artificial systems to carry structural stresses. One factor necessary for designing or evaluating a structure's stability is knowledge of the material's mechanical properties, such as stress-strain behavior and fracturing characteristics. The composition of most structural materials such as metals and concrete is uniform and reproducible to the degree that their mechanical properties in service are nearly the same as those measured in the laboratory. Consequently, a rational design can be made on the basis of published mechanical property values. The same equivalence does not exist for rock materials for two reasons: (1) the composition of even the most common rock types is highly variable, and (2) the problem of the size effect is present. Mechanical property tests not conducted at a scale that includes these effects in normal proportion will not be representive of the in-situ rock mass. Of course, the alternative is to perform the test in-situ. Comparatively few mechanical property tests have been performed in-situ because of the physical limitations of test equipment and the prohibitive cost. The current procedure is to design a structure based on small-scale laboratory measurements, with a safety factor usually determined by the experience of the design engineer. While this procedure is somewhat unsatisfactory, laboratory measurements can and do provide a useful basis for the initial design.

Jan 1, 1970

The visitors to the October meeting will be given opportunity to visit the coal fields in Illinois, adjacent to St. Louis, the steam-shovel operations near Pittsburgh, Kan., and the Oklahoma oil fields. The coal mines of Illinois, Missouri, Kansas, Oklahoma and Arkansas furnish fuel over all of the central region of the country. Especially at this time, this district is of importance because of the necessity of producing enormous quantities to meet the country's demand. Some of the mines produce as much as 5,500 tons in 8 hr., and. the total output of Illinois, alone, in 1916, was 63,000,000 tons. Visitors may see elaborate underground electric tram systems, large underground machine repair shops, surface plants equipped for rapid loading, and some of the most modern coal-washing installations. The conditions for mining here are somewhat simpler than in the east.

Jan 8, 1917

Discussion of the paper of DR. L. EDELEANU, presented at the Pittsburgh meeting, October, 1914, and printed in Bulletin No. 93, September, 1914, pp. 2313 to 2332. F. W. BUSHONG,* Pittsburgh, Pa.-This process of extraction is particularly advantageous in the case of Roumanian oil, on account of its high aromatic content. I have prepared a paper, which will appear probably next month in the Journal of Industrial and Engineering Chemistry, describing a series of two-degree fractions of oil from the Nowata field in Oklahoma. I have made extractions with liquid sulphur dioxide in accordance with Dr. Edeleanu's method, and found that the Oklahoma oils, for instance the kerosene fractions, are purified in this way, although the quantity taken out does not at all compare with that taken out of the Roumanian oil. I have made careful analyses of separate fractions, before and after extraction, and they show a very distinct fractional separation-. I like the distinction, brought out by Dr. Day in previous discussions, in regard to oils, rather than oil, and his beautiful illustration of fractionation. Just as fuller's earth, and other clays, will enable us to obtain lower-boiling fractional portions of the oils, and eventually retain a residuum which is hard to get rid of, or to remove, so there is an advantage to be gained by the use of a reagent which begins at the other end by taking out these last constituents first, namely, ozone. I have also analyzed the ozonides and they are derivatives of hydrocarbons very much poorer. in hydrogen than we have heretofore supposed. Similar results have been obtained with Roumanian, Russian, and Italian petroleums. Furthermore, I have prepared ozonides from a fraction of Oklahoma oils, both before and after ex-' traction with liquid sulphur dioxide, and found that the ozonide obtained from the extract is the same as that precipitated out of the original oils. In other words, the residual substance, which the fuller's earth retains, is the one first thrown down by the use of ozone, but it may also be extracted by liquid sulphur dioxide- and be thus separated in apparently unaltered condition. The analysis of this extract corresponds to the formula C15H20, which is surprising, since it is lower in hydrogen content than we have been in the habit of considering among the constituents of our oils. The extract was probably contaminated with other hydro-carbons, but I think it is unquestionable that we must look for hydro-carbons in petroleum which correspond to a formula at least -as low in hydrogen as C15H2o, and possibly even as low as C15H16.

Jan 4, 1915