Search Documents

Arthur Jarman, Sydney, New South Wales, Australia (communication to the Secretary*):—All designs for modern metallurgical and chemical laboratories should provide each student's desk with a hood or cupboard for the removal of fumes, but I doubt the advisability of placing air-baths and similar apparatus close to the hot plate. It seems to me that unless the draft be exceptionally strong the fumes from ordinary wet assay work will corrode the iron-work of the fittings, in which case the baths would have to be removed to a general desk in order to lighten the work of the laboratory attendant. Since the laboratory described by Mr. White has been in use for two years, I should like to ask if any inconvenience has resulted from this corrosion. If not, it is certainly an important improvement to have the air-baths and similar apparatus placed on the student's desk. The hoods described by Mr. White are supplied with a good draft by means of a fan, which gives them a great advantage

Jan 1, 1905

By Chas. H. Behre

MINING geology, both theoretical and practical, continued to make noteworthy progress during 1938. Mining companies generally, stimulated especially by the improvement in economic conditions during the latter part of the year, extended their workings and developed new ore bodies. Federal and state surveys expanded their search for favorable localities and actual ore bodies, with corresponding enlargement of regional knowledge. Theoretical investigations covered geochemistry of ore deposition, structural control of mineralization, and mineralogy of the ores (chiefly aided by the microscope). More geochemical work is greatly needed: ore searching is still too purely empirical, based on features that have been recognized else- where. Thus, clearer understanding of the conditions essential to transportation of metal salts is needed. Jean Verhogen has suggested that solubilities

Jan 1, 1939

By S. S. Arentz

An organized safety program has reduced accidents at Pioche because effort is first devoted to arousing and maintaining interest in safety, followed by training in accident prevention, assigning responsibility for the detection and removal of safety hazards, and finally, by keeping score through an adequate system of records. A health program is intergrated with the safety program.

Jan 1, 1950

By C. W. Tomlinson

The analyses which have been made by the writer according to the method described below were made as part of Professor Withey's investigation of the concrete aggregates1 of Wisconsin, in the hope of throwing light on the extent and nature of the influence of the mineral-ogical constitution of aggregates upon the properties of the concrete and mortar made from them. This method was found most satisfactory after four weeks of experiment and modification. It attains an accuracy quite sufficient for the purpose of the analysis, and is fairly rapid. mineralogical analyses may be thus made at a cost considerably less than that of chemical analyses of a similar degree of completeness. The results obtained by this method should not vary in any item by more than 1 per cent. of the total weight of the sample, and may be made very much better than this by the exercise of greater care in the microscopic counting (see below). The accuracy of the hand sorting is of course dependent upon the experience and carefulness of the analyst. After a month's practice, the writer was enabled to turn out these analyses at the rate of four in 33 hr. of working time. It was found most economical to handle four samples at once. The samples analyzed weighed from 290 to 340 g. each. The same method could be applied to larger or to smaller samples, within reasonable limits. The following list covers the necessary apparatus and equipment: 1 good balance capable of weighing accurately to tenths (or, better, hun- dredths) of a gram, and of handling quantities up to 200 g. in weight. 1 Westphal balance, registering densities to two or three decimal places. 1 Meker or Bunsen burner, and gas supply. 1 nest of sieves, 10, 20, 40, and 100 meshes to the inch.

Jan 1, 1916

By Charles W. Frey

OF all the machinery used in mining work, mine cars are probably the most abused. They are hauled through water and muck, up hill and down grade, whipped around curves, bumped and jerked, and exposed to the ravages of the weather. However, as a general rule their lubrication is not altogether neglected. The mine operator may not heed the extra drag that a frozen conveyor bearing gives, but he is all too aware of the increased load that faulty mine car wheels put on his power plant. When the circuit breakers begin to fly, and when his mine loco- motives come out with only half the cars that they were able to carry a month before, then neglected mine car wheel lubrication becomes apparent, and woe to the person responsible.

Jan 1, 1938

By E. W. Mehal, R. H. Cox

A planar Gunn oscillator was developed for use in a monolithic microwave integrated circuit. The device was designed to operate in the frequency range of 20 to 30 GHz with a continuous wave output. Selective epitaxial deposition was used to form the n+n-n+ device structure in (111)-B semi-insuluting GaAs. The active n- region was about 5 µ thick and was doped to about 2 x 1015 electrons per cu cm by the reactor background. The mobility was -5000 sq cm v-1 sec-1. The n+ epitaxial deposits were doped to 1 x 10" electrons per cu cm using S2Cl2. Planarity of the three DURING the past few years, new microwave devices in discrete form have been developed utilizing GaAs. The next logical step is to combine these devices in a successive deposits was quite good; total overgrowth was less than 1 µ. Metal ohmic contacts were applied to the n+ or n- regions using a Ag-In-Ge alloy. The n+n-n+ devices had symmetric IV characteristics and could be operated C. W. for both directions of device bias. Metal ohmic contacts made directly to the n-region always Produced asymmetric devices. C. W. operation was obtained over the 22 to 28 GHz frequency range. The best performance obtained was 1.5 mw C. LV. output at 28 GHz. microwave integrated circuit. A potential application would be a radar receiver operating in the millimeter wavelength region. The key component in such an integrated circuit is a microwave oscillator, recently described by Gunn.1, 2 Under the influence of a large electric field, high-resistivity GaAs exhibits a bulk negative resistance, and high-frequency current oscillations result. Since relatively high voltage and

Jan 1, 1969

By F. C. Langenberg

DURING the course of some magnetic investigations which the authors have under way, six bars of 0.43-carbon steel were tested, a permeameter designed after the Hopkinson yoke type being used. The results obtained were of some interest and although far from complete serve at least to show the necessity for a careful investigation of the previous history of the samples under study as well as the chemical composition. A steel of given chemical composition may assume a great variety of structures, depending on its treatment. Professor Sauveur has devised a very clear and effective means of illustrating this point: The accompanying microphotographs show some of the possible structures assumed by a 0.30 carbon steel (Fig. 1). It is now a well-established fact that a structural change is accompanied by changes in the tensile strength, elastic limit, hardness, etc., and it seems reasonable to assume also, a change in the magnetic properties. Six bars of a 0.43-carbon steel were used, all the bars being taken from the same rod and treated as follows: Bar 1. Heated to 1,100° and cooled in the furnace. Bar 2. Heated to 1,000° and cooled in the furnace. Bar 3. Heated to 900° and cooled in the furnace. Bar 4. Heated to 1,000° and cooled in air. Bar 5. Heated to 900° and cooled in air. Bar 6. Untreated. The bars were then turned to 13 mm. diameter and tested by the Hopkinson yoke method. After testing, three specimens were taken from each bar, one from each end and one from the middle. The homogeneity of the. bars was tested in this manner and in no instance could any difference in structure be detected between the three specimens.1

Jan 2, 1915

By John T. Galey

In southwestern Pennsylvania 151 wells1 were completed during 1938, a total of 61 less than the preceding year. Of this number 81 were gas wells and 28 oil wells, which shows a considerable decline from 1937, when 108 gas wells and 64 oil wells were completed. However, 42 dry holes were drilled during 1938, which is two more than the previous year. Shallow Development—Gas No new extensions to old pools, or new wells that would indicate any substantial reserves, mere found during the year. However, a number of commercial wells and many dry holes were drilled to the Big Injun sand in Menallen township, Fayette County. One well had an initial open flow of 6 million cu. ft. In Rich Hill township, Greene County, one well producing, from the Big Injun sand, 2 million cu. ft. was brought in. Several other gas wells were completed in Shaler township, Alleghany County and Center township, Greene County, both of which had initial open flows of 1 million cu. ft. from the 30-ft. sand. Another well, in Aleppo township, Greene County, had an equal volume from the Gordon Stray sand. Oil Development has proceeded at a retarded rate because of the low price of Pennsylvania-grade crude. A total initial of only 205 bbl., an average of 7.3 bbl. per well, was developed. This is 1785 bbl. less than during 1937 and no new repressuring operations have been contemplated to prevent the declining production. The average price received for oil was $1.56 per barrel, with a high of $1.88 and a low of $1.34. This oil costs $1.94 per barrel2 to produce. The average price paid for Bradford

Jan 1, 1939

By John T. Galey

In southwestern Pennsylvania 151 wells1 were completed during 1938, a total of 61 less than the preceding year. Of this number 81 were gas wells and 28 oil wells, which shows a considerable decline from 1937, when 108 gas wells and 64 oil wells were completed. However, 42 dry holes were drilled during 1938, which is two more than the previous year. Shallow Development—Gas No new extensions to old pools, or new wells that would indicate any substantial reserves, mere found during the year. However, a number of commercial wells and many dry holes were drilled to the Big Injun sand in Menallen township, Fayette County. One well had an initial open flow of 6 million cu. ft. In Rich Hill township, Greene County, one well producing, from the Big Injun sand, 2 million cu. ft. was brought in. Several other gas wells were completed in Shaler township, Alleghany County and Center township, Greene County, both of which had initial open flows of 1 million cu. ft. from the 30-ft. sand. Another well, in Aleppo township, Greene County, had an equal volume from the Gordon Stray sand. Oil Development has proceeded at a retarded rate because of the low price of Pennsylvania-grade crude. A total initial of only 205 bbl., an average of 7.3 bbl. per well, was developed. This is 1785 bbl. less than during 1937 and no new repressuring operations have been contemplated to prevent the declining production. The average price received for oil was $1.56 per barrel, with a high of $1.88 and a low of $1.34. This oil costs $1.94 per barrel2 to produce. The average price paid for Bradford

Jan 1, 1939

By Francisco Cadena

The construction of Norris Dam, built by the Tennessee Valley Authority on the Clinch River, a tributary of the Tennessee River, involved the production of coarse and fine aggregate for approximately 1,000,000 cu. yd. of concrete. A hammer-mill plant for the manufacture of sand from quarried rock was indicated as the most economical method of providing this material. The abrasive character of the feed as compared to the average hammer-mill feed made it necessary to control operations rather closely, and this resulted in the accumulation of many data hearing on production and maintenance. This paper will deal only with three specific sets of data, accompanied by such general information as may be necessary to explain the conditions under which they were obtained and followed by discussions on the relation of silica content to hammer wear, comparative costs of slugger and stirrup hammers, and comparative costs of tool-steel and manganese-steel grate bars. Coarse Aggregate Aggregate for the concrete poured at Norris Dam came from a dolomite quarry within 2000 ft. of the west abutment of the dam. Geologically this rock is known as a Knox dolomite. A typical chemical analysis is given in Table 1. Table 1.—Typical Chemical Analysis Per Cent H2O, 110° F............................................. 0.02 lgnition loss.............................................. 43.56 MgO.................................................... 19.19 CaO ................................................... 29.07 Fe²O³.................................................... 0.39 Al2O³ .................................................. 1.94 SiO2..................................................... 5.75 SO2...................................................... Tr. Rock was carried by 12-yd, dump trucks to a 42-in. gyratory crusher set at 6 1/2in. The product of the primary crusher was taken by a belt conveyor to a double.-deck vibrating screen followed by a 5 1/2-ft. cone

Jan 1, 1938

By Francisco Cadena

The construction of Norris Dam, built by the Tennessee Valley Authority on the Clinch River, a tributary of the Tennessee River, involved the production of coarse and fine aggregate for approximately 1,000,000 cu. yd. of concrete. A hammer-mill plant for the manufacture of sand from quarried rock was indicated as the most economical method of providing this material. The abrasive character of the feed as compared to the average hammer-mill feed made it necessary to control operations rather closely, and this resulted in the accumulation of many data hearing on production and maintenance. This paper will deal only with three specific sets of data, accompanied by such general information as may be necessary to explain the conditions under which they were obtained and followed by discussions on the relation of silica content to hammer wear, comparative costs of slugger and stirrup hammers, and comparative costs of tool-steel and manganese-steel grate bars. Coarse Aggregate Aggregate for the concrete poured at Norris Dam came from a dolomite quarry within 2000 ft. of the west abutment of the dam. Geologically this rock is known as a Knox dolomite. A typical chemical analysis is given in Table 1. Table 1.—Typical Chemical Analysis Per Cent H2O, 110° F............................................. 0.02 lgnition loss.............................................. 43.56 MgO.................................................... 19.19 CaO ................................................... 29.07 Fe²O³.................................................... 0.39 Al2O³ .................................................. 1.94 SiO2..................................................... 5.75 SO2...................................................... Tr. Rock was carried by 12-yd, dump trucks to a 42-in. gyratory crusher set at 6 1/2in. The product of the primary crusher was taken by a belt conveyor to a double.-deck vibrating screen followed by a 5 1/2-ft. cone

Jan 1, 1938

By R. W. Brauchli

The author has made a statistical analysis of the relations between recovery and length of cores, diameters of cores and of drill pipe, and cost of coring under different contractual conditions. The data are presented in tabular form. The length of the core appears to play a minor part in the percentage recovered, with a tendency to lower percentages with increased core length; approximately 50 per cent of the length cored is normally recovered if the cores are more than 2 in. in diameter. The diarneter of the core has a more pronounced influence on the recovery; 1 3/8-in. cores average 29.1 per cent, 2 1/8 -in. core 53.4 per cent and 3 1/8-in. cores with 6 5/8-in. drill pipe 68.5 per cent. Where 31/8-in cores were taken with smaller diameter drill pipe, i. e., 4 1/2 - in. drill pipe in 8 1/4-in. hole, the recovery dropped to 49.8 per cent, indicating that the more rigid drill pipe favors good core recovery, a conclusion also borne out by the results obtained in using large drill pipe for smaller diame-ter cores. Surprisingly good recoveries were obtained from both the Johnson and Simpson sands; shale, limestone and dolomite each gave less then average recoveries. No serious fishing jobs resulted from coring operations. Attention is called to the need of a core drill which will make it possible to drill a full-sized hole while coring, and thus reduce the cost of cores. Data pertaining to core drilling should be consistently recorded, such as, diameter of drill pipe and collar, diameter of core, recovery, size of cutter heads, type of core-barrel, coring time, speeds of rotation and of pumps, weight on core-barrel, etc.

Jan 1, 1933

By AIME AIME

Maurer's review summarizes quite thoroughly the various theoretical developments and experimental findings that contribute to the knowledge of rock mechanics in drilling. This discussion supplements Maurer's paper by including generalities that either have been incompletely treated in his review, or formulated from new considerations by the discussant and from hitherto unpublished experimental data. These generalities are associated with the observation that all mechanical methods of rock drilling essentially comprise a sequence of indexed, concentrated loadings on the bottom of the hole being drilled.

Jan 1, 1967

By CARLE R. HAYWARDC

BEFORE discussing this subject it is necessary to define somewhat the meaning of the tern metallurgical.. When I was a student at M. I. T. ore-dressing was not thought of as metallurgy in any sense of the word, but it was considered as more closely allied to mining, enabling the miner to market an otherwise unmarketable product, or at least to market it to better advantage. Progress in ore-dressing has gradually allied it more closely with smelting, for at present the raw material received by the smelter is predominantly concentrates and smelting practice has been profoundly affected by the products of the concentrating mill. It is not surprising therefore that in many schools ore-dressing is frankly considered as metallurgy and although there may be some dissenting opinions it will be considered as such in this discussion. Twenty-five years ago at M. I. T. metallography was taught in a few optional lectures. This subject has developed rapidly and been expanded into the new and increasingly important field of physical metallurgy. To many metallurgists the -term metallurgy means only physical metallurgy. They know little or nothing about smelting and leaching and even less about ore-dressing. Having then broadened the scope of the term metal¬lurgy to cover ore-dressing, smelting, leaching and phys-

Jan 1, 1930

By William D. Klopp, Joseph R. Stephens

ThE phenomenon of high ductility associated with a phase transformation, termed super plasticity, has been reported for a number of alloy systems. Abnormally high ductility has been observed during tensile testing of cast eutectic alloys of Pb-Sn and Mg-cu.1 Brasses of Cu-37 at. pct Zn and Cu-40 at. pct Zn elongate up to 180 pct when tested near their transformation temperature.' Enhanced ductility was also observed in Cu-47 at. pct Zn tested near the order-disorder transformation temperature. The most remarkable super-plastic effect reported to date is for the eutectoid alloy A1-60 at. pct Zn, where elongations up to 1000 pct have been reported.3 underwood4 has recently summarized the conditions that normally accompany the superplastic effect which are: 1) a phase transformation or allotropic change or order-disorder transformation; 2) an externally applied stress; 3) presence of an interface between two phases; and 4) a metastable structural condition. In addition, higher ductility is observed when the specimen is tested at slow deformation rates rather than under impact. The purpose of this note is to present evidence of the superplastic effect in a Cr-30 at. pct Co alloy. In addition, low-temperature bend ductility data are given for the alloy in the quenched, metastable, single-phased condition and in the slow-cooled, two-phased, near-equilibrium condition. Sixty-gram alloy ingots were prepared by noncon-sumable arc melting of iodide chromium and electron-beam-melted cobalt. These were cast into rectangular ingots and fabricated to 0.035-in. sheet by rolling at 2300°F. Attempts to roll at slightly lower temperatures invariably resulted in severe cracking, apparently due to presence of the brittle a phase below the transformation temperature of 2230°F. Tensile specimens with a 0.25 by 1.0 in. reduced section were cut from the sheet by electrical discharge machining. The specimens were heated at 2400°F for 1/2 hr to obtain a recrystallized, bcc, single-phased structure. Temperature of the specimens was subsequently raised or lowered to the desired test temperature and held for 10 min prior to testing at a crosshead speed of 0.01 in. per min. Fig. 1 shows the chromium-rich side of the Cr-Co phase diagram.6 The alloy composition investigated is near the point of maximum solubility of cobalt in chromium, and the alloy is two-phased when in equilibrium at temperatures below 2230°F. The results of tensile tests conducted at temperatures indicated by X's in Fig. 1 are summarized in Figs. 2, 3, and 4. Tensile elongations shown in Fig. 2 increase from 11 pct at 1900°F to a maximum of 160 pct at 2250°F. The elongation then decreases to 44 pct as the test temperature is further increased to 2600°F. The proximity of the temperature of highest elongation and the solvus temperature suggests that a superplastic effect is present in Cr + 30 at. pct Co similar to that observed for other alloys that show anomalous ductility. The variation of upper yield stress with temperature is shown in Fig. 3. A small peak in yield stress occurs at 2250"F, coincident with the maximum in

Jan 1, 1967

By C. O. Burgess, A. B. Kinzel

Although vanadium is an important constituent of almost every brand of high-speed steel manufactured today, little is known as to its role in this series of alloys. The now standard 18 per cent tungsten and 4 per cent chromium constituents were the result of developments and improvements before 1903. In that year, vanadium was added to the steel by Dr. J. A. Mathews. This proved to be an outstanding contribution, so that the standard high-speed steel of today is generally referred to as the 18-4-1 type—18 per cent W, 4 per cent Cr and 1 per cent V. Various modifications have been attempted since then, but with the exception of the introduction of cobalt, no striking improvements have occurred. Vanadium has been increased to 2 per cent in many of the common brands, and at least one brand contains 5 per cent V in addition to the usual tungsten and chromium. The carbon has been gradually increased so that it is not uncommon to find 0.80 per cent in the standard grade, although the carbon in the 5 per cent vanadium steel mentioned above is approximately 1 per cent, and even 1.2 per cent has been made experimentally. The role of vanadium has never been well understood, and although vanadium has been increased to 2.5 and even 5 per cent in certain high speed steels, it has not been generally established that the higher vanadium contents result in superior cutting properties. However, in all of the investigations, of which the writers are aware, the interrelation of the vanadium and carbon contents was not considered as a prime factor. The carbon content was held at 1.2 per cent or less—generally at the usual 0.65 to 0.75 per cent. Theoretical Aspects Although the role of vanadium is not understood, there are theoretical considerations which indicate definite reasons for the improved cutting effciency due to vanadium. Refinement of the secondary grain is one distinct possibility. The well-known action of vanadium in limiting grain size might well be expected to operate in high-speed steels, and even though its action were not evident in the parent austenite grain.

Jan 1, 1932

By Geoffrey Purcell

For anyone contemplating flotation research with only a very limited amount of mineral available for testing, the choice of suitable experimental equipment is by no means obvious. Hallimond's original and delightfully simple cell1 has been modified by subsequent investigators2to give quantitative and reproducible results on small samples. However, the modifications have necessarily compromised its simplicity, and the services of a skilled glass blower are required to duplicate the cell. Vacuum flotation tests3 at best qualitative and may, under certain conditions4, be quite impracticable. A hundred years ago, Thoreau observed that "our life is frittered away by detail," and his advice was to "simplify, simplify," The success of the simple cell described below, which was developed by the author and differs somewhat from Hallimond's original, suggests that Thoreau's advice still has merit a century later. CELL CONSTRUCTION The cell [Fig. 1 (a) and 1 (b)] consists essentially of a 25-ml pipet from which has been cut the con- stricted top. The length cut off is not critical since this merely determines the final volume of the cell. Cell volume may be varied in this way from 20 to 50 cc. A pyrex extension tube, equal in diameter and approximately half the length of the cut pipet, is attached to the top of the pipet by means of 2 ½ in. of close-fitting tygon tubing. When assembled, the two pyrex tubes should be spaced about ½ in. apart. Another length of tygon tubing is added to the extension piece for sealing purposes. A ¾ -in. disc of porous material (teflon cloth is ideal), secured in place over the pipet tip with another piece of shaped tygon tubing, completes the cell assembly. Tygon tubing may be shaped by pushing it over the pipet tip or some similar form and heating carefully over a low flame. For flotation [Fig. 1 (c) 1, the pyrex extension tube is rotated approximately 70º from the vertical and held in position by a wire hook and eye. Floated material, after reaching the free surface, falls and rests on this inclined plane. The barrier formed by the bent tygon tubing ensures that floated material does not slide back and return to the vertical portion of the cell.

Jan 11, 1965

By Dan White

Stress measurements taken at Climax, El Teniente, Henderson, Lakeshore and Urad mines indicate that the deposits cited are generally subject to an inclined stress field that can be highly anisotropic. The horizontal stresses vary significantly between deposits and exceed the measured vertical stresses at Henderson and El Teniente mines. Experience at Henderson, Lakeshore and Urad mines is used to suggest a relationship between cavability and the premining stress. In poorly fractured, high strength rock masses, high stresses can assist the caving process. In well fractured, low strength rock masses, caving is ensured by breakage along fracture planes. The relationship between ground weight and the premining stress is also investigated. Ground weight problems appear to be most related to the competence of the ground rather than the magnitude of the premining stress as supported by experience at Climax and El Teniente mines. Recent developments in block caving favor the extraction of poorly fractured, high stressed rock. Accompanying the development of these deposits are new problems such as rock bursts. These pose particular problems to mine designers in that no experience currently exists in combatting rock burst problems at block caving mines.

Jan 1, 1984

By Butler, P. B.

TRIPOLI is a rather unusual form of silica, which thus far has been found in commercially valuable quantities only in the neighborhood of Seneca, Mo., although there are numerous deposits of somewhat similar materials in several southern states. The latter have not, however, attained any position of real importance, due undoubtedly to definite structural limitations which have rendered them less suitable for the particular uses that have given tripoli its principal outlets. Raymond B. Ladoo, mineral technologist, says': "Since that time (1880-90) the term tripoli has with increasing definiteness been reserved for the Seneca material, and other siliceous materials of similar origin and physical properties."

Jan 1, 1928

By C. G. Dunn, F. Lionetti

The energy associated with grain boundaries in polycrystalline aggregates is believed to play a major role in grain growth processes and, when growth ceases, to determine the final equilibrium grain boundary angles. Further, the energy of grain boundaries of recrystallization nuclei is a factor in nucleation processes. It is important to know, therefore, how the energy per unit area of a grain boundary, that is, the grain boundary surface tension y, depends on the difference in orientation of the two lattices of the grains producing the boundary. Although the problem is important, surprisingly little has been done toward a quantitative evaluation of the effect of orientation difference on grain boundary energies. C. S. Smith1 recently has discussed this problem and in addition to showing effects of orientation difference on equilibrium angles has shown a variety of interesting effects of surface tension on the appearance of microstructures. Fig 1 and the following relations expressed in Eq 1, which connect equilibrium grain boundary angles and surface tensions, illustrate equilibrium conditions which are believed to hold true in metals. ?l2 _ ?23 _ ?l3 [1] sin ?3 sin ?1 sin ?2 Clearly any two of these surface tensions can be expressed in terms of the third when the equilibrium angles ?1, ?2, and ?3 are known. Applied to the present problem for solid state equilibrium of three grains, the angles ?1,?2 and ?3 must be measured in a plane perpendicular to the line of junction of the three grains. Normally a direct determination of these angles from random microsections is impossible. Consequently Harker and Parker2 and Smith1 (except for some measurements on flat specimens) resorted to a statistical method to determine equilibrium grain boundary angles. Smith reported that grain boundaries meet the surface of a piece of metal nearly perpendicularly. He reported also, in connection with direct angle measurements on flat specimens with grains extending through the thickness, that angles varied appreciably from 120" and concluded that there was a measurable effect of orientation difference on surface tension. Another direct way of determining equilibrium angles and a method adaptable to studying particular configurations, recently suggested by Dunn,³ is to use a three-grain flat specimen with orientations of grains so chosen that the junction line of the three grains will be straight throughout the thickness and perpendicular to the surface of the specimen. Choice of orientations is possible when individual grains of each group can be grown to predetermined orientations through the reorienting and growth of " seed crystals " as described.³ Not only is it possible, for example, to have grains with the same crystallographic plane in the plane of the specimen, but a given orientation difference between two grains can be made a common factor to an unlimited number of three-grain groups while a series of orientation differences is investigated. Any effect of anisotropy of gas-solid surface tension due to grain orientation should be minimized by having all gains oriented the same with regard to crystallographic plane in the plane of the specimen. Another feature of the three-grain group is the notched grain boundaries as shown in Fig 2 for one specimen (S4). The notches serve to anchor the end positions of the grain boundaries (especially at high temperatures) while the central and junction point of the grains moves toward an equilibrium position. Final equilibrium should produce straight grain boundaries if the notches are very narrow and if changes in orientation* of the grain boundary do not alter the surface tension.† If one assumes the straight line condition and no change of surface tension with small changes in grain boundary orientation and finds the equilibrium configuration by a minimization of the grain boundary energy, one obtains the relations given in Eq 1. The approach to straight line boundaries or to minimum energy configurations in specimens containing large grains, such as those used in the present investigation, may be very slow compared with the approach to equilibrium conditions for the grain boundary angles. It may be desirable, as proved to be the case in the present investiga-

Jan 1, 1950