Search Documents

Discussion of the paper of DAVID COLE, presented at the New York meeting, February, 1915, and printed in Bulletin No. 9S, February, 191.5, pp. 431 to 444. R. H. RICHARDS, Boston, Mass.-The Butchart riffle seems to violate, according to our preconceived notions, some of the very essential principles of the Wilfley riffle, and therefore ought not to succeed, but it does. On this account it must conform to other very important principles, or it would not succeed. It would beinteresting to know what these principles are which enable it to succeed. E. P. MATHEWSON, Butte, Mont.-I cannot, explain the principles on which this riffle acts, because I have not given it personal attention. All I can say is that I first saw the riffle in operation several years ago at the Steptoe Valley mill in Nevada, and I was told that a man had been around there guaranteeing that these riffles would make the table do twice as much work and do it better than with the ordinary Wilfley riffles. I heard nothing further about it for a long time. I heard about a year ago, from David Cole, the author of this paper, that the riffles were a success, and that he was using them at Morenci, and if, we were interested he would send the inventor to show us how they worked. As we knew of David Cole's reputation, we invited Mr. Butchart to come up and show us how his riffles worked. We did not have sufficient confidence in the invention to guarantee Mr. Butchart his traveling expenses, but lie came up entirely at his own expense and put in a set of the riffles on one table. It was known as a roughing table in our No.

Jan 5, 1915

By C. R. Fettke, Bela Hubbard

During the summer of 1916, while on a visit to the United States Agricultural Experiment Station at Mayaguez, Porto Rico, the writers were told by D. W. May, the director, that an occurrence of manganese ore had been reported from the mesa southeast of Mayaguez. A visit to the locality showed that the mesa is covered by a mantle of bright red soil resting upon serpentine. Here and there over its surface concretionary masses of limonite have been mashed out of this soil. These were apparently mistaken for manganese ore. On account of the similarity of this occurrence to the extensive iron deposits of northeastern Cuba, it was thought that a brief description based upon a two days'reconnaissance survey of the area mould be of scientific interest. Mayaguez is on the west coast of Porto Rico, at Mayaguez Bay, which is the third in importance of the harbors of the island. The city itself is about 3/4 mile inland on the low-lying plain at the mouth of the

Jan 1, 1920

By H. W. Young

Development of high-tension outdoor substations during the past few years has been due primarily to economic reasons. The demand for power in small communities could not be met with the conventional and Fig. 1.—33,000-VOLT step-up outdoor air-break switch and substation installation outside of a powerhouse generating at 2300 volts stepping up to 33,000. Twenty-one coal mines are supplied from lines leading from this installation. comparatively expensive indoor types unless the rate for service was materially increased. To prove that the problems lncident to development have been solved successfully, it is only necessary to point to the wide adoption of outdoor substations by utility companies. Passing at once to the question as to whether outdoor substations are applicable to

Jan 1, 1920

By Sixten Wollmar

A RESUME. of some important points regarding the life of drill steel, which have so far been learned from investigations carried out in the SKF Hofors' mines, Sweden, by L. Nordenfelt, is given below. By examining all fatigue breakages which occurred in the test it has bees conclusively proved that when a rock drill of the self-rotating type is used all break-ages of this kind start from the surface of the hole in the steel. It has further been proved that the rough hole surfaces (i.e., the marks left by the indentations of the sand core), are the starting points for the fatigue cracks and that the corrosive action of the water has .a very detrimental influence on the life of the steel. It is now a well known fact that this corrosive action concentrates at the bottom of the indentations left by the sand. The corrosive action of water has been well described by U. R. Evans in his paper, "Some Recent Work on the Corrosion of Metals," delivered on March 31, 1926, before the Liverpool Engineering Society, which is quoted in part below: "It is well known that, under ordinary conditions, most metals do not corrode in water unless dissolved oxygen is present. The strange feature of corrosion in submerged metals, however, is that the attack does not proceed mainly over those portions of the surface which could be reached readily by oxygen; deterioration takes place at certain points which are relatively, inaccessible to oxygen. When. a metal becomes pitted by corrosion, further attack usually deepens the pits, and in many cases it could be proved that there was no free oxygen at the bottom of the pits. None of the older theories of corrosion explain the fact that, although oxygen is necessary for rapid corrosion to take place, the attack actually occurs at the very points which could not be reached directly by oxygen.

Jan 6, 1928

By James H. Bechtold

The yield strength of annealed tungsten was found to have a strain rate exponent 12 times as great as that of low carbon steel. The effects of temperature and strain rate could be correlated through the Zener-Hollomon parameter with a heat of activation associated with yielding of 32,000 cal per g-atom. This heat of activation is independent of strain although both the temperature and strain rate dependence of the yield strength vary with strain. MOST metals have mechanical properties that are relatively insensitive to strain rate, especially at temperatures below that at which creep is important. In iron and molybdenum," however, relatively large strain rate effects are observed in the temperature range where the rapid increase in yield strength occurs which leads to the ductile to brittle transition. In this temperature range Zener and Hollomon' have suggested that the effects of strain rate and temperature T on the yield strength could be correlated through a single parameter, P, in the following manner where Q is a heat of activation characteristic of the metal and associated with some unspecified relaxation process. They also observed that if the temperature is restricted to a relatively narrow range— below room temperature for steel—that the logarithm of the yield strength could be expressed as a linear function of the logarithm of P. From this relation the logarithm of stress should be a linear function of the logarithm of the strain rate at constant temperature: where r will be referred to herein as the strain rate exponent. At a constant strain rate the logarithm of stress should be a linear function of l/T, ay(e = C) ~ eT [4] where q = Qr [5] and is referred to herein as the temperature exponent. These equations have been found to relate the effects of temperature and strain rate on the yield strength of molybdenum in the temperature range wherein the ductile to brittle transition occurs, —75° to +150°C.2 Bechtold and Shewmon3 have shown recently that the ductile to brittle transition in annealed tungsten occurs in the temperature range +150° to +350°C. In this temperature range the yield strength in-

Jan 1, 1957

By Robert M. Yerkes

The following is a brief account of the methods of measuring intelligence especially prepared for use in the U. S. Army, of typical results, and of some of their immediately practical applications. It has been prepared for the assistance of army examiners and for the information of all who are interested in the relations of scientific placement to military efficiency. Purpose of the Intelligence Tests Under the direction of the Division of Psychology, Medical Department, and in accordance with provisions of War Department General Order No. 74, mental tests are given recruits during the 2-wk. detention period. These tests provide an immediate and reasonably dependable classification of the men according to general intelligence. Their specific purposes are to aid: 1. In the discovery of men whose superior intelligence suggests their consideration for advancement; 2. In the prompt selection and assignment to development battalions of men who are so inferior mentally that they are suited only for selected assignments; 3. In forming organizations of uniform mental strength where such uniformity is desired; 4. In forming organizations of superior mental strength where such superiority is demanded by the nature of the work to be performed; 5. In selecting suitable men for various army duties or for special training in colleges or technical schools; 6. In the early formation of training groups within regiment' or battery in order that each man may receive instruction and drill according to his ability to profit thereby; 7. In the early recognition of the mentally slow as contrasted with the stubborn or disobedient; 8. In the discovery of men whose low-grade intelligence renders them either a burden or a menace to the service.

Jan 1, 1919

By Robert L. Clayton, Arthur Baker

Two massive sulfide zinc-copper ore bodies are in quartz-sericite schist (probably formed by regional metamorphism of sediments) and andesite of the Precambrian Yavapai Series, on opposite sides of a large rhyolite mass. The Copper Queen ore body is conformable with bedding of its schist host rock, but the Old Dick ore body probably crosscuts the strata. A lens of low-grade mineralization in chlorite schist, associated with the Old Dick ore body, crosscuts through 150 feet of strata. The ore bodies are almost entirely enclosed by premineral faults with less than 150 feet of displacement. The minerals of the ores are pyrite, sphalerite, and chalcopyrite, with minor arsenopyrite and galena and small amounts of intimately associated chlorite, sericite, and calcite. Old Dick sphalerite-chalcopyrite intergrowths have mutual boundaries, and sphalerite grains contain abundant exsolved chalcopyrite. Copper Queen sphalerite-chalcopyrite intergrowths have extremely ragged boundaries, the sphalerite contains no exsolved chalcopyrite, and the ore overall has a schistose texture. The pre-mineral fault on the hanging wall of the Copper Queen ore body has had 100 feet of post-mineral displacement. Wall rocks of the Old Dick ore body locally have been sericitized. Andesite and quartz-sericite schists for at least several hundred feet around both ore bodies have been chloritized erratically and contain traces of copper.

Jan 1, 1968

By L. E. Young

GERMANY'S recent collapse and the occupation by the Allies of the coal fields of the Ruhr, the Saar, Silesia, Poland, and Czechoslovakia have focused attention on the postwar coal problems of Europe. American coal operators are interested deeply in what is happening in Britain, particularly in the discussion of plans to rehabilitate and reorganize the entire British coal mining industry, as such reorganization will probably determine the place Britain will occupy in the European coal markets for the next few decades. The continuing decline of British out-ut per man-shift, particularly during wartime, has resulted in a most serious effort to determine whether this unfortunate trend is due to the physical condition of the coal seams, poor management, poor engineering and planning, inadequate use of power and machinery, or failure of labor to co-operate with management.

Jan 1, 1945

By John G. H. Crump, E. W. Owen, Peter P. Gregory

.Although oil Production in West Texas in 1939 reached the highest figure for any year since the inauguration of proration, drilling activity continued in the decline that had commerlced the previous year. Production totaled 79,469,341s bbl. of oil during 1939 as compared with 71,257,541s bbl. in 1938 and 75,743,000 in 1937. The number of wells completed in 1939 was 1850, compared with 2045 in 1938 and 2806 ill 1937. Several factors combined to cause the decrease in drilling activity, principally the contiriued increase in the length of time required to obtain a return of invcstment and the scarcity of important new discoveries during the past two years. Of the 1850 completions 1638 resulted in oil wells, 4 were gas wells, 168 were dry holes and 40 were abandoned locations; 1990 new locations were made during the year. The ratio of dry holes to producing wells completed was one dry hole to each 9.75 oil wells, compared with one dry hole to each 7.48 oil wells completed during 1938. This large decrease in dry holes was due to a marked decline in the number of wildcat wells drilled, there having been only 120 wildcat completions during 1939. In addition to the lessened exploration work, the deeline' in activity was largely accounted for by the slowing up of operations in the older pools of Ector, Howard and Winkler Counties. Ector County, which had led West Tcxas in 1938 with 622 new oil wells, furnished only 485 producers in 1939, but was still the leader for the West Texas district. The largest gain in activity was in Yoakum County, whew 350 oil wells were completed in the Denver-Wasson field against 152 during 1938. Most of the older oil fields have been almost completely drilled up, only a few new producers still being completed from time to time in

Jan 1, 1940

By K. Huessener

Discussion of the paper of K. HUESSENER, presented at the New York. meeting, February, .1916, and printed in Bulletin No. -110, February, 1916, pp. 443 to 474. KARL NIBECKER, Youngstown, O.-It has been contended that the method of measuring the operation of a blast-furnace boiler plant by means of the amount of steam generated per pound of coke used in the blast furnace is not of any value and even that it is quite misleading. It has been found, in the plant with which I am connected, that, in general, a blast furnace will operate during a month so close to average. conditions that the average monthly measure of the plant efficiency is indicated by the number of pounds of steam made per pound of coke put into the furnace, neglecting the amount of iron made and the amount of gas used for heating the stoves. For some time, it has been our custom to keep a record of this figure and use it as a general indicator of the operating practice. In comparing this figure from two different plants we have found some difficulty, due to differing conditions. It is conceivable but not altogether probable that a slight difference in burden, kind of ore or coke, would account for this difference. The method of using the gas, whether washed or unwashed, and the plant arrangement, may affect this number. For several plants on which we have worked up this number, we have found that most of the ordinary running irregularities are eliminated during a run of a month or more and that the number is, therefore, satisfactory. In connection with a plot of the number of pounds of steam generated per pound of coke, we are also plotting the production of the furnaces and the pounds of coke used per ton of production. With these three curves plotted on the same sheet, we believe that the operating man has sufficient information to judge of the efficiency of his plant' for any month.

Jan 5, 1916

By F. E. Lathe, G. A. Guess

This investigation was started with the idea of determining whether copper-nickel mattes might not have a lower flowing temperature than copper mattes, and thus perhaps aid in accounting for the difficulty so far experienced in attempts at pyritically smelting copper-nickel ores. We have used the term flowing temperature to avoid confusion with the usual melting temperature. The idea at first was that there might be a greater viscosity range in the one matte than in the other. In the early experiments it was in our minds to determine the relative fluidity of the mattes at a given temperature. Looking to this end we tried pouring equal quantities of matte at the same temperature down an inclined graphite slab, and comparing the distances traveled. Results from such experiments were not satisfactory. We next tried pouring equal quantities at the same temperature through a drilled hole in an Acheson graphite crucible, timing the flow with a stop watch, the whole operation being carried out in the muffle. The flowing temperatures of the mattes tried in this way appeared to be about the same, and there was no marked difference in their fluidity when well fused. Oxidation gave some trouble. These experiments finally led to the adoption of the following apparatus which we thought at the time was wholly original with us, but which we have since found was described by W. McA. Johnson.l The tray used was a slab of graphite 8 in. long, 2 in. wide and 1/4 in. thick. It had six holes, into each of which sat a cup of Acheson graphite. Centrally in the bottom of each cup was a vertically drilled hole 3/16 in. in diameter and 1/2 in. long. The mattes to be tested were put into the cups, in a solid piece. The cups were covered with inverted Battersea "A" cups. The tray was put into the muffle of a gas-fired furnace. Supports on each end held the tray about 2 in. from the floor of the muffle. Under

Jan 1, 1917

IN describing the mining and treatment of ore at the Broken Hill Proprietary mine E. J. Horwood, superintendent of mines, gave the following account of the development of flotation there. It was in 1902 that the recovery of zinc from the ore by flotation was hit upon. The former general man-ager, Mr. Delprat, directed our then research officer, Mr. Carmichael, to try boiling the tailing with salt cake of which there was a considerable tonnage left over from the leaching process. The idea was the possibility of dissolving the zinc. Carmichael made the. Experi-ment, but, rather discouraged, took the glass beaker down to the general manager's office, remarking that he could not get the stuff to settle. On showing the beaker to Mr. Delprat, the latter recognized the marked distinction between the black scum on the surface and the rest of the tailings, and exclaimed: "Why, this is the very thing we are after; the zinc is floating!" Patents were applied for at once, but later on it was found that another investigator, Potter, had already hit upon the idea of flotation. His patents, however, were so involved with other operations that his process was impracticable. Litigation followed, but in the end the B. H. P. paid the Potter Company a substantial sum as settlement. In the case of Carmichael's beaker, the acid in the salt cake acted on the small amount of carbonate present in the ore, producing CO. bubbles. Owing to the aversion of sulfide crystals to being wetted, the gas as generated attached itself to the zinc in in-creasing quantities until there was sufficient to buoy the zinc particle to the surface. This action went The next problem was to apply the principle on a commercial scale, and this involved much experiment and developmental work. Various kinds of sloping copper pans with gas jets below to maintain the tem-perature were first tried, but were found too slow and too costly in operation.

Jan 6, 1928

By E. W. Owen, John G. H. Crump, Peter P. Gregory

.Although oil Production in West Texas in 1939 reached the highest figure for any year since the inauguration of proration, drilling activity continued in the decline that had commerlced the previous year. Production totaled 79,469,341s bbl. of oil during 1939 as compared with 71,257,541s bbl. in 1938 and 75,743,000 in 1937. The number of wells completed in 1939 was 1850, compared with 2045 in 1938 and 2806 ill 1937. Several factors combined to cause the decrease in drilling activity, principally the contiriued increase in the length of time required to obtain a return of invcstment and the scarcity of important new discoveries during the past two years. Of the 1850 completions 1638 resulted in oil wells, 4 were gas wells, 168 were dry holes and 40 were abandoned locations; 1990 new locations were made during the year. The ratio of dry holes to producing wells completed was one dry hole to each 9.75 oil wells, compared with one dry hole to each 7.48 oil wells completed during 1938. This large decrease in dry holes was due to a marked decline in the number of wildcat wells drilled, there having been only 120 wildcat completions during 1939. In addition to the lessened exploration work, the deeline' in activity was largely accounted for by the slowing up of operations in the older pools of Ector, Howard and Winkler Counties. Ector County, which had led West Tcxas in 1938 with 622 new oil wells, furnished only 485 producers in 1939, but was still the leader for the West Texas district. The largest gain in activity was in Yoakum County, whew 350 oil wells were completed in the Denver-Wasson field against 152 during 1938. Most of the older oil fields have been almost completely drilled up, only a few new producers still being completed from time to time in

Jan 1, 1940

By Charles H. Snow

In vol. xxxviii. (1889) of the American Journal of Science, under the heading " Pseudomorphs of Native Copper after Azurite from Grant County, New Mexico," Mr. W. S. Yeates describes a most interesting mineral form. His account is, that a material having the appearance of native copper, but brittle and much lighter than native copper (its specific gravity being but 4.15), was submitted to him by the United States National Museum for examination. The substance was reported to him as occurring in masses, varying in weight from one ounce to seventy pounds, and composed of tabular crystals, variously grouped. The specimens, when submitted, ex hibited on the outride a covering of clay; and they were subse quently found to consist of particles of the same clay, intimately mixed with others of native copper, producing the aspect, without the weight, of pure copper. The forms of the azurite crystals, found at Bisbee, Arizona, having been suggested by the specimens, actual goniometer-measurements were made; and the two crystallizations were thus proved to be identical. To these statements I may be permitted to add some further in formation, derived from personal examination, and illustrated with a suite of characteristic specimens collected by myself (Fig. 2).* Grant County, New Mexico, in which these specimens were found, is a region noted for variety of mineral species. The sixth edition of Dana's Mineralogy mentions many of the silver, copper, and lead minerals, and also turquois, aragonite, wulfenite, barite, fluorite, vanadinite, etc. I have noticed also green and red garnet, molybdenite, native alum, smithsonite, bornite, etc., some of which presented peculiar modes of occurrence. The green garnet, for instance, was in masses formed of separate crystals, each of pea-size and not unlike that vegetable in tint and smooth opaqueness. The crystallization was

Jan 1, 1893

By G. E. Anderson

THE Asientos-Tepezala district is in the north of the State of Aguascalientes, about 30 miles north of the city of Aguascalientes, the capital. The district is reached by a standard-gage railway on the west side through the junction, Rincon de Romos, which is on the main line of the National Railways of Mexico from El Paso to Aguascalientes. The terminal is Tepezala. On the east side a narrow gage connects Asientos with the junction, San Gil, on the main line, Asientos to San Luis Potosi. Tepezala on the west side of the Asientos hills is about 2 miles from Asientos on the east side. The area between the two mining centers is separated by a low subued divide which forms the central part of the hills. The district is in a group of isolated hills surrounded by characteristic bolson plains. The hills rise above the surrounding plains about 6500 to 10,000 ft.; the highest peak is Alta Mira. The two mining districts lie somewhat upon the flanks of the hills at about the same elevation, 7200 ft. The district is an old one and has been known since early in the 17th Century when the mines were worked by the Franciscan Monks. Their mining operations which were limited to the oxidized ore above the water level, consisted chiefly in extracting silver from the Santa Francisca vein. To judge from the old workings and the remains of surface plant, the operations were on a large scale. By modern mining operations since about 1900, the various mines of the district have been one of the principal sources of ore for the smelter located at Aguascalientes.

Jan 2, 1926

By Walter Harvey Weed

CONTENTS. [ ] INTRODUCTION. THIS paper deals with certain ore-deposits whose structural features or mineral contents (or both) result, directly or indirectly, from igneous intrusions and their after-effects. It is largely a discussion of contact metamorphic ore-deposits based upon the physical changes in rocks due to contact action. It involves a classification of ore-deposits only so far as is abso-

Jan 1, 1913

By Morris Cohen

It has been known for several years that in certain age-hardenable alloys precipitation of finely divided particles occurs simultaneously with the changes in physical properties; while, in other alloys, these changes take place prior to, or without any, precipitation. The silver-copper alloys are generally considered to be the classical example of the first group, and the aluminum-copper alloys (duralumin) are typical of the second group. The two types of aging are explained by the precipilation theory and the knot theory respectively.‡ It is well recognized now that these theories do not conflict with each other; the one used for a given aging system depends upon which way the alloy in question happens to age-harden. However, previous work on the aluminum-copper al1oys1,2 has shown that if the aging temperature is raised sufficiently the hardening will be accompanied by actual precipitation. In fact, it is quite probable that all of the so-called knot-hardening alloys will harden by precipitation at elevated temperatures. Therefore, a combination of the knot and precipitation theories is necessary to give a complete picture of the agehardening of these alloys, and indicates that knot formation and precipitation are not independent processes. To test the generality of this concept, it is necessary to show that the alloys that commonly harden by precipitation will harden prior to precipitation if the aging temperature is sufficiently lowered. For this reason, it would be well to carefully re-examine the aging characteristics of the silver-copper alloys at aging temperatures somewhat below those used by previous investigators. If hardening can be detected prior to precipitation in these alloys, it will lend great weight to the idea that knot forma-

Jan 1, 1937

By J. Koskinen, P. Vanninen, T. Leskinen, A. Leppala

Continuous and accurate analysis of slurries has become a critical requirement in mineral processing. With on-stream assay data, automation procedures can be employed to obtain higher recovery and improved control of production. The Courier system was developed by Outokumpu Oy, taking advantage of mineral processing know-how, to provide means for achieving these objectives in the company's ore concentrator operations. It became evident, based on early experience with on-stream analysis, that the Courier should be a complete system to fulfill all functions of the analysis operation. The system includes equipment and techniques for primary sampling, sample preparation for assaying, and processing of data for final assay results. In addition, the on-line X-ray spectrometer unit was developed specifically for industrial use, incorporating a number of technical advances. A new system for converting spectrometric data to assays is based on a physical model programmed on a digital computer. The Courier system is furnished on a turn-key installation basis incorporating all elements needed for operation including calibration. The experience to date with the Courier system at the Pyhäsalmi concentrator is summarized. Reference is also made to recent results with installations at the Outokumpu and Vihanti concentrators.

Jan 1, 1972

By C. H. North, L. E. Lynd, W. W. Anderson, H. Sicurdson

Since the composition and structure of the beach sand concentrates correspond so closely to what would be expected of ilmenites that have been altered by oxidation and leaching, there is no reason to postulate the existence of the hypothetical compound, arizonite, to account for the composition of titaniferous beach sand concentrates. The four concentrates studied are examples of the different products that can result from various degrees of alteration of primary ilmenite. CONSIDERABLE uncertainty is revealed in the literature regarding the nature of the titanium minerals which make up the bulk of the heavy, opaque fractions of numerous beach sand deposits of the world. An investigation was made, therefore, of beach sand concentrates of North Carolina, Florida, Brazil, and India. Magnetic concentration, X-ray, and microscope data were obtained which show that these concentrates consist essentially of ilmenite and its alteration products. The alteration, brought about by oxidation and leaching of iron values, has resulted in upgrading the ores to a range of 56 to 64 pct TiO², as compared with 52.7 pct in ilmenite of theoretical composition. Conclusions drawn regarding the nature of the dominant titanium-bearing mineral in beach sand deposits have been based mainly on chemical composition, specific gravity, and petrography. This material has usually been referred to as ilmenite or weathered ilrnenite.: It has also been referred to as arizonite because in chemical composition it resembles an Occurrence in in which was described by Palmer' in 1909. The chemical composition of Palmer's arizonite corresponded closely to the formula Fe²O³ . 3TiOs. From approximate measurements of crystal angles on a large imperfect crystal with rounded faces, the

Jan 1, 1955

By W. McA. Johnson

THE work of J. E. Johnson, Jr., on the effect of small amounts of oxygen in cast iron in increasing its strength and resistance to shock, is of interest from the technical and scientific standpoints. The following exposition of the theory carries Mr. Johnson's explanation further and, in my opinion, will disclose certain related phenomena and perhaps shed further light on the subject. Some facts, so novel in character as to excite incredulity, have been accepted finally by many metallurgical authorities. One of these is that oxygen to the amount of 0.060 per cent. in cast iron gives it a breaking strength of 3,500 lb. per square inch, compared with a breaking strength 2,500 lb. per square inch in cast iron of the same composition with re-ect to elements other than oxygen, but having only 0.010 per cent. oxygen. The "oxygenated" cast iron, even if made in a coke furnace, has properties equivalent to the best charcoal pig iron. Mr. Johnson has also established the fact that the variations in combined carbon cannot account for the results. By microphotographic evidence he has shown that in oxygenated cast iron the graphite particles are dense and spherical, while in well-deoxidized cast iron the graphite particles are flaky and leaf-like. It is just a matter of common sense that iron of the latter structure is much weaker than the former. The analytical and microscopical work has been done with care and there is no doubt in my mind about the accuracy of the, experimental results. In this paper then, we shall consider it as accepted that small amounts (0.060 per cent.) of oxygen increase the strength of cast iron by producing a structure in which the round and solid particles of graphite are surrounded firmly by a principal matrix of ferrite. The purpose of this paper is to show that the reason why the graphite particles are round is founded on the fact that the particles of any precipitate are made denser and harder by the presence of a reagent having a solvent action on the precipitate. This can be stated as a law, although I have never seen it so given in any textbook.

Jan 2, 1916