Search Documents

PRESIDENT. CHARLES F. RAND,1 NEW YORK, N. Y. PAST PRESIDENTS. CHARLES KIRCHHOFF,1 NEW YORK, N. Y. JAMES F. KEMP,2 NEW YORK, N. Y. FIRST VICE-PRESIDENT. BENJAMIN B. THAY ER,1 NEW YORK, N. Y. TREASURER. GEORGE C. STONE,1 NEW YORK, N. Y.

Jan 4, 1913

By T. W. Atkins

ATHOUGH the magnesium industry in this country is about thirty years old, not until American industry began to amaze the rest of the world and confound our enemies with the extent and variety of our war production did the industry begin to come into its own. At the outset of the war, only three companies were producing and fabricating magnesium in this country. Today, about sixty companies are producing, casting, forging, extruding, rolling, and stamping this lightweight champion of metals. Experience in alloying, processing. and fabricating the metal gained during the war years is now being translated into hundreds of civilian peacetime applications. In no sense should magnesium be considered as a substitute for other metals. It has come into popular favor because If many unique properties in addition to is being the lightest known structural metal. These attributes are toughness, igidity, corrosion-resistance, and the ease with which it can be machined and worked. Magnesium is the third most abundant element in nature. Sea water, magnesite,

Jan 1, 1946

UNITED ENGINEERING TRUSTEES, INC James L Head, '51 E C Meagher, '52 Gail F Moulton, '53 ADVISORY COMMITTEE TO REAL ESTATE COMMITTEE Cleveland Dodge Herbert Hoover Paul D Merica HOOVER MEDAL BOARD OF AWARD Scott Turner, '56, Chairman Cadwallader Evans, Jr , '52 Paul D Merica, '54

Jan 1, 1951

By K. W. Smith, E. A. Thomas

Various grades of synthetic mullite have been developed in recent years to replace or supplement natural sources of mullite deriued from the mullite group of minerals consisting of sillimanite, kyanite, and andalusite. Raw materials and heat treating processes used in making synthetic mullite are described. Chemical and physical data are given for typical grades and crystalline structure is illustrated with micrographs. Use of synthetic mullite as a refractory material in the glass and metallurgical industries is discussed. Mullite (3A12O3.2siO2), the only stable compound formed in the alumina-silica system, is usually present to some degree in all aluminum silicate ceramic products. The formation of mullite is considered beneficial to give rigidity to the structure and is dependent upon the ratio of Al2O3 to SiO2 in the original composition, particle size, degree of mixing, firing temperature, cooling rate, and the presence of auxiliary glass-forming fluxes. Mullite may also be formed at the reaction interface of fire clay or alumina-type refractories in contact with glass or slag melts. The term synthetic mullite is commonly used today to identify a class of sintered and fused aggregates or grains in the alumina-silica system having a highly developed mullite structure but derived mainly from raw materials other than the sillimanite group of minerals. Within the past 15 years extensive research has been done to develop economical processes to form sintered synthetic mullite aggregate to replace calcined Indian kyanite in super-refractories. Severa1 brands of such mullite are now being produced in commerical quantity and finding extensive use in refractories. Based on the service results of such refractories in many applications throughout the metallurgical, ceramic, and glass industry these developments have been considered successful and suitable substitutes for Indian kyanite now appear assured. EARLY DEVELOPMENT The conversion of kyanite, sillimanite and anda-lusite minerals of the sillimanite group to mullite and their use in refractories and porcelain have been discussed quite extensively in the literature by peck,' Grieg,' Riddle and Foster,3 Bowen and Grieg,4 and others and will only be mentioned here for reference to compare properties with synthetic mullite. In 1928, curtis5 reported on the development of a high temperature gas-converter process for forming synthetic mullite. The raw materials were derived mainly from lumps of high alumina clay of the correct natural composition or blends of clays and alumina that was interground and briquetted to form a suitable charge to maintain a surface combustion firing within the converter. Curtis was, no doubt, the first to illustrate by micrographs in natural color the crystalline structure of mullite derived from kyanite and mullite derived by sintering clay and alumina mixtures at temperatures above cone 32 (3123°F) and by electric fusion. In 1937, sei16 was issued a patent covering the use of a mixture of alumina-silica minerals and alumina in the proportion to form a mullite-yielding material at temperatures in excess of 3100' F. During the period from 1930 to 1940, economic conditions were not favorable for the production of synthetic mullite mainly due to an adequate supply of good grades of Indian kyanite ore suitable for conversion to mullite. Uncertain conditions on availability of the Indian kyanite during the early stages of World War II fostered further study on the development of synthetic substitutes. In 1943, McVay and wilson7 reported on an extensive investigation of domestic substitute materials. Their work covered essentially the use of mixtures of electric furnace mullite, calcined topaz, and calcined domestic kyanite. Compositions were found that gave equivalent or better hot load strength than Indian kyanite in mullite-type brick compositions; however, the calcining of the topaz presented certain physical and chemical problems on the disposition of silicofluoride and hydrofluoric acid while the high cost of electric furnace mullite was a limiting factor. In this work it was pointed out that water-quenched fused mullite was found to be unstable on reheat and gave poor hot load strength due to excessive glass present whereas the slow cooled or annealed mullite contained large crystals of mullite and corundum with little glass and gave superior results.

Jan 1, 1961

By Nicholas J. Grant, A. W. Mullendore

Measurements of grain boundary sliding were made on polycrystal and bicrystal tensile creep specimens of Al-2 pct Mg at 500oand 700oF. Grain and pain boundary orientation factors were studied with respect to their effect on the magnitude and direction of grain boundary sliding. It was concluded that a large part of the observed sliding may be caused by the operation of slip crossing the grain boundaries. Supporting evidence for this model from other investigations is also given. GRAIN boundary sliding in metals during elevated temperature deformation has frequently been thought of as an independent mechanism with its own characteristic stress, temperature, and structure dependence. Further, it is common to regard the grain boundary as the weak link during creep deformation since: 1) the strength of apolycrystallinemetal, as meas-sured by creep rate or rupture life, decreases at\a more rapid rate when sliding comes into action, and 2) grain boundary sliding, as revealed by "boundary darkening" and fold formation is often the first visible deformation. On the other hand, more quantitative observations, in both polycrystals and bicrystals, have shown boundary deformation to be closely related to the over-all deformation of the specimen. The curve of grain boundary contribution to elongation vs time takes a shape similar to that of the total creep curve,',' and the apparent activation energy for sliding appears in most cases to have the same value as that for total creep.3 In order to reconcile these observations, one is left with these possibilities: 1) The grain boundary and grain deformation rates are both controlled by the same rate limiting processes. 2) The rate of sliding is limited by accommodation deformation in the grains. 3) Sliding is the consequence of grain deformation. 4) Grain deformation is dependent on prior grain boundary deformation. Alternatives 2) and 3) would view grain boundary sliding as occurring in series with another deformation process which could be rate controlling. McLean prefers item 3) as the sliding basis,4 where subgrain rotation results from dislocation accumulation in the substructure at the grain boundary. Crussard and Friedel's5 treatment of grain boundary migration considers that dislocations are forced into the grain boundary and are dissociated into dislocations with Burgers' vectors nearly parallel to the grain boundary. The movement of these partial grain boundary dislocations then produces the sliding. The results of this investigation have indicated alternative 3) to be the most appropriate and it is this aspect of grain boundary sliding which will be analyzed. EXPERIMENTAL PROCEDURE Three types of tensile creep tests were employed in this study: (I) A constant load test of one very coarsegrained polycrystalline specimen at 500° F, 3600 psi, 1 pct per/hr strain rate to examine the grain and grain boundary orientation factors in grain boundary sliding. (II) Constant strain rate tests of polycrystalline specimens at 510o and 715o F, 2 pct per/hr strain rate to determine the grain boundary contribution to elongation at very low total strains. (m) Constant strain rate tests of bicrystals of various grain and grain boundary orientations at 500°F, 2 pct per/hr strain rate to obtain further information on the orientation effects in grain boundary sliding. The material used in the tests was a high-purity Al-1'.92 mg alloy, kindly furnished by Alcoa. Its composition is given in Table I. The alloy is a solid solution at the test temperatures. The coarse grained polycrystalline specimen for test (I) was prepared by recrystallizingthe machined specimen, straining it about 1pct intension at room temperature and annealing at 900°F for 8hr. The constant strain rate specimens (11), with an 178 in. sq. by 1/2 in. long gage section, were cut from 1/8 in. thick sheet of cold rolled and annealed (8 hr at 1000oF) material. The latter specimens were supported in a special jig during the milling operations to avoid any bending, and the gage section was subsequently heavily electropolished to remove the cold-worked surface layer. The bicrystals (ID) were machined in the same

Jan 1, 1963

By E. A. H. Tays, F. A. Schiertz

In 1893 the Anglo-Mexican Mining Company, Limited, purchased the Guadalupe and adjoining mines, and, after opening up the Guadalupe, constructed a 20-stamp mill, which was started in August, 1894. Although the results were extremely profitable (the ore being very rich and carrying much free gold) they were not satisfactory, because the tailings were found to average about $20 per ton in gold. Early in 1895 the late Mr. James E. Mills made for the company a very thorough study of the subject, and decided that the tailings could be treated most advantageously by the cyanide process. In view of this opinion, Mr. Henry R. Batcheller, of Boston, was sent down in the winter of 1895-96 to experiment on a working scale. Mr. Batcheller was successful; and, during 1896-97, built a cyanide-plant with a capacity for treating 1500 tons per month. This was run until May, 1899, when all the sands suitable for the process had been treated. The mill was stopped about the same time for lack of ore; but no tailings from the mill had gone to the dump since October, 1897, as the values they contained after that date were far too low to warrant further treatment. When the cyanide-plant was stopped there still remained on the dump about 16,000 tons of slimes, consisting of about 75 per cent. of clay-slimes and 25 of sand-slimes. The successful treatment of this very unfavorable material is the theme of this paper. Fig. 1 is a view of the hacienda, taken in 1897.

Jan 1, 1902

By Ernest A. Hersam

The advantages of the tri-axial diagram in representing the composition of slags and silicates are well appreciated by many metallurgists. Prof. H. M. Howe* has pointed out the application of a fourth variable, in connection with the diagram, and has called attention to the usefulness of the general systern of tri-axial representation. In the discussion of Prof. Howe's paper, Mr. Firmstone† indicated how certain distances could be measured directly by projection. Prof. R. H. Thurston,‡ as early as 1877, had made use of a typtic model and of triangular diagrams, with contour-lines, rekrresenting varying physical properties corresponding to varying chemical composition. Still earlier, Prof. J. Willard Gibbs§ had used the diagram in similar work; and other scientists were thought by him to have preceded him in its use. During the period of nearly a quarter-century, following these early publications, there has been an ever-increasing use of graphic methods in numerous branches of scientific work, stimulated, in part, by the possibility of representation by three co-ordinates upon plane-surfaces. Yet the number of possible co-ordinates for graphic representation has always been, and always must be, insufficient to express many facts which are shown mathematically; while, without graphic representation, the involved nature of mathematical expression often fails to facilitate and promote research. The coming into use, therefore, of the fourth co-ordinate, permitting representation by four variables, marks but the beginning of some lines of research which could not well advance without this aid.

Jan 1, 1902

By Robert D. Coale, Reinhart P. Ehrlich, Terry L. Thompson

The copper concentrate slurry pipeline and pumping system of the Bougainville Copper Ltd. facilities on Bougainville Island in Papua New Guinea, is discussed. Development work necessary for the pipeline design is dealt with including a description of the 17-mile by 6-in. pipeline and pumping system as it was installed. The commissioning of the pipeline is reviewed as well as current operating performance, including data on corrosion, pump parts life, and manpower requirements.

Jan 1, 1977

By AIME AIME

ABOUT one hundred were in attendance when Donald M. Liddell opened the session* on non-ferrous metallurgy at 2 p. m. on Tuesday. F. F. Col- cord was vice-chairman. For the first part of the session the chair was turned over to Frank G. Breyer, who introduced G. L. Spencer, Jr., to read his contribution (T.P. 378), "High Silica Retorts at the Rose Lake Smelter," which was an account of the method of manufacture of high-silica zinc retorts at the Rose Lake smelter, and of the service given by these retorts. The materials used are, by volume, 50 per cent clay, 25 per cent silica flour and 25 per cent grog. Written discussion of the paper by E. M. Johnson, W. R. In- galls and M. M. Neale was read by Mr. Breyer. Mr. Johnson explained that their practice in manufacture differed very little from Rose Lake practice, but that their retorts did not last as long, possibly because Eagle-Picher does not use a hydraulic retort press or sintered ore. He ascribes the longer life of silica re- torts to the fact that they do not bend under heat. Mr. Ingalls pointed out that silica retorts are no better as heat conductors than clay retorts, but that they do have a superior rigidity in the furnace. The character of the retorts does not affect the density of the charge. Mr. Neale gave the procedure in the manufacture and use of high-silica retorts at the Donora zinc works, not

Jan 1, 1931

By C. A. HEILANDG

THERE is a need for men well trained in geo- physical prospecting. Although the number of geophysicists required by the industry in the future cannot be expected to be very great, there will always be a demand for courses on the subject. A number of educational institutions are, therefore, interested in providing instruction in geophysical prospecting. Some institutions will be satisfied with courses in general geophysical prospecting designed to give the geologist fundamental concepts on the comparative merits of the methods and the interpretation of the results. Other institutions will want to give special instructions in theory, field procedure, and interpretation of any or all geophysical methods. In giving such special courses a number of problems will be faced, and it seems to me that such difficulties are easier to overcome in an engineering institution than in a university. A knowledge of basic engineering subjects, such as required of all students in an engineering school, seems to be the best foundation upon

Jan 1, 1930

By P. Somasundaran, D. R. Nagaraj

Based on our finding that commercial copper chelating solvent extractants such as LIX65NR and LIX63R are excellent collectors for copper minerals, a detailed study was made using several water-soluble chelating agents belonging to the class of aromatic hydroxyoximes and having a gradual structural variation. A basic study of the system salicylaldoximetenorite was also carried out. Microflotation tests as a function of concentration of oxime, pH, addition of copper, etc., using oxidic copper minerals such as chrysocolla and tenorite, and batch laboratory flotation tests using synthetic mixtures of quartz and oxidic copper minerals were carried out. The results have demonstrated that the hydroxyoximes studied have comparable collector property, and in all cases the oxime was found to be partitioned between the mineral surface and the copper in the bulk aqueous phase in the form of a bulk copper chelate. The occurrence of the latter could be accounted for by formation in the bulk phase of a chelate between oxime and copper released from the mineral and/or by the detachment of the surface chelate. The role of surface and bulk chelates in flotation is discussed based on the results of partitioning of chelating agent into bulk and surface chelates in the salicylaldoxime-tenorite system. The effect of structural aspects such as isomerism and substitution in the molecule on the collector property of the oximes is also discussed. Substitution on oximic carbon by alkyl groups is found to promote collector activity. Only the anti-isomer has the ability to collect.

Jan 1, 1982

G. R. DELAMATER,* Steelton, Pa. (written discussion t)..-Under the heading, "Preparation of Coal," Mr. Campbell discusses the ad-visability of sizing the coal for the jigs. While I fully realize that conditions alter cases, my method at all times is to do everything possible to avoid sizing. It would be unwise for anyone to say that greater ash and sulfur reduction is impossible by sizing. The thing that can be questioned is, will the greater reduction be sufficient to counterbalance the greater cost? The same thing is true of the proper size of crushing for each individual proposition. In past years many washeries have been built and equipped for wash-ing two or more sizes, but I know of few instances where these plants were operated more than a year before sizing was abandoned. Perhaps in some of these cases this was clue to lack of sufficient development of such scheme of operation; yet I have found that most men engaged in this work, including myself, started into the work with a pretty well-fixed idea that sizing would be beneficial. But trials and tests showed that the mixed size washing is necessary for coals requiring crushing to 1 in. or less. Coal is of low value, compared with other minerals segregated by washing or concentration, for which reason simplicity must be the watchword: simplicity in the machine used to attain the separations of impurities from the coal, in the accessory plant equipment, and in its arrangement.

Jan 12, 1919

By William Fink

THE investigation of aluminum alloy systems prior to 1923 was severely handicapped by the low purity of the best aluminum available. However, by that time, the electrolytic purification of aluminum had been developed by the Aluminum Company of America,(1)? and in May, 1923, there was initiated, under the direction of E. H. Dix, Jr.,(2) a program for the investigation of the aluminum-alloy systems using the purest electrolytically refined aluminum. Up to the present time 15 papers (a) have been published, most of which are on the binary aluminum systems. Work is now in progress on the ternary systems, and it seems probable that most of the future work will be on ternary systems. Therefore this survey of the binary aluminum systems investigated up to the present time (the sixteenth paper of the series) concludes the first phase of this program. Certain qualitative relations between corresponding portions of the various binary diagrams as well as between different parts of the same diagram were noted some time ago by R. S. Archer.(4) In the present . paper these relations and others are considered quantitatively by the application of well-known thermodynamic equations or by newly discovered empirical relations No attempt will be made here to present the history of the develop-ment of the thermodynamic equations defining equilibrium or their application to metallic systems, although a list of references is given at the end of this paper for those who are interested in this phase of the

Jan 1, 1934

By C. F. Park

Hog Mountain is in the north central part of Tallapoosa County, Alabama, about 13 miles northeast from Alexander City. The Hog Mountain Mining and Milling Co. controls 1658 acres of land and is the only company operating in the district. The principal mine workings are in sections 10 and 15, T.24N., R.22E. The region has been visited by geologists many times, has been briefly described in several papers,' and mentioned in many others. No detailed map of the mine or vicinity has been published. The field work for this report was started in March, 1934, as part of the Public Works reemployment program. About a month and a half was spent mapping both the surface and underground geology. The mine owners and operators have cooperated in every way possible to facilitate the work. N. 0. Johnson, the mine engineer, E. W. Ellsworth, J. E. McCoy and J. G. Englebert have acted as assistants for part of the work. The University of Alabama and the State Geologist's office have aided materially by permitting free use of their libraries and laboratory equipment. Geography Hog Mountain forms a conspicuous ridge that rises to a height of a little more than 1000 ft. above sea level, and about 400 to 500 ft. above the adjacent country. The ridge is composed of granitoid intrusive rock and silicified schist that are more resistant to erosion than the surrounding softer schist. The topography of the area around the ridge is gently rolling and is typical of the subdued forms of the deeply eroded Southern

Jan 1, 1935

By C. F. Park

Hog Mountain is in the north central part of Tallapoosa County, Alabama, about 13 miles northeast from Alexander City. The Hog Mountain Mining and Milling Co. controls 1658 acres of land and is the only company operating in the district. The principal mine workings are in sections 10 and 15, T.24N., R.22E. The region has been visited by geologists many times, has been briefly described in several papers,' and mentioned in many others. No detailed map of the mine or vicinity has been published. The field work for this report was started in March, 1934, as part of the Public Works reemployment program. About a month and a half was spent mapping both the surface and underground geology. The mine owners and operators have cooperated in every way possible to facilitate the work. N. 0. Johnson, the mine engineer, E. W. Ellsworth, J. E. McCoy and J. G. Englebert have acted as assistants for part of the work. The University of Alabama and the State Geologist's office have aided materially by permitting free use of their libraries and laboratory equipment. Geography Hog Mountain forms a conspicuous ridge that rises to a height of a little more than 1000 ft. above sea level, and about 400 to 500 ft. above the adjacent country. The ridge is composed of granitoid intrusive rock and silicified schist that are more resistant to erosion than the surrounding softer schist. The topography of the area around the ridge is gently rolling and is typical of the subdued forms of the deeply eroded Southern

Jan 1, 1935

Hog Mountain is in the north central part of Tallapoosa County, Alabama, about 13 miles northeast from Alexander City. The Hog Mountain Mining and Milling Co. controls 1658 acres of land and is the only company operating in the district. The principal mine workings are in sections 10 and 15, T.24N., R.22E. The region has been visited by geologists many times, has been briefly described in several papers,' and mentioned in many others. No detailed map of the mine or vicinity has been published. The field work for this report was started in March, 1934, as part of the Public Works reemployment program. About a month and a half was spent mapping both the surface and underground geology. The mine owners and operators have cooperated in every way possible to facilitate the work. N. 0. Johnson, the mine engineer, E. W. Ellsworth, J. E. McCoy and J. G. Englebert have acted as assistants for part of the work. The University of Alabama and the State Geologist's office have aided materially by permitting free use of their libraries and laboratory equipment. Geography Hog Mountain forms a conspicuous ridge that rises to a height of a little more than 1000 ft. above sea level, and about 400 to 500 ft. above the adjacent country. The ridge is composed of granitoid intrusive rock and silicified schist that are more resistant to erosion than the surrounding softer schist. The topography of the area around the ridge is gently rolling and is typical of the subdued forms of the deeply eroded Southern

Jan 1, 1935

By C. F. Park

Hog Mountain is in the north central part of Tallapoosa County, Alabama, about 13 miles northeast from Alexander City. The Hog Mountain Mining and Milling Co. controls 1658 acres of land and is the only company operating in the district. The principal mine workings are in sections 10 and 15, T.24N., R.22E. The region has been visited by geologists many times, has been briefly described in several papers,' and mentioned in many others. No detailed map of the mine or vicinity has been published. The field work for this report was started in March, 1934, as part of the Public Works reemployment program. About a month and a half was spent mapping both the surface and underground geology. The mine owners and operators have cooperated in every way possible to facilitate the work. N. 0. Johnson, the mine engineer, E. W. Ellsworth, J. E. McCoy and J. G. Englebert have acted as assistants for part of the work. The University of Alabama and the State Geologist's office have aided materially by permitting free use of their libraries and laboratory equipment. Geography Hog Mountain forms a conspicuous ridge that rises to a height of a little more than 1000 ft. above sea level, and about 400 to 500 ft. above the adjacent country. The ridge is composed of granitoid intrusive rock and silicified schist that are more resistant to erosion than the surrounding softer schist. The topography of the area around the ridge is gently rolling and is typical of the subdued forms of the deeply eroded Southern

Jan 1, 1935

By H. Rush Spedden

R. T. Hukki has extended his study of ball mill grinding at supercritical speeds with an excellent mathematical analysis of tumbling mill grinding. When he speaks of supercritical speeds he is referring to the original Fahrenwald critical-speed formula. He said that, as an example of the use of supercritical speeds, if a fine grinding mill of any type is wet-grinding ore of 3.5 sp gr with steel balls at a certain conventional subcritical speed, the same power input and capacity should be obtainable with autogenous medium of 3.5 sp gr at a speed 100 to 120 pct of the critical in a grate- type mill of the same effective size. In addition, he introduced what he calls the base mill concept which, he stated, makes it possible to compare performances of mills of different sizes grinding with a metallic medium or with an autogenous medium of any specific gravity and at any speed. He showed the use of this concept and explained how it makes possible the evaluation of mill sizes and motors for any grinding purpose at any accuracy well within the conventional standards of design in milling practice. He further stated that the maximum capacity in pilot plant mills with various media, either metallic or autogenous, has always been found to be within the supercritical speed range. In his discussion of this paper he substantiated these findings with statements showing the increased capacity obtained in a milling plant in Finland which followed his principles.

Jan 7, 1960

By C. H. Mathewson

IN a series of papers published recently,1 R. F. Mehl and associates have studied the characteristics of form and orientation of many segregate structures and have found diversified conditions which can be explained only by postulating some specific kind of cooperative mechanism permitting the matrix lattice in some favorable region to transform into the segregate lattice in an adaptable region presenting a geometrically similar arrangement of atoms. We have collected the published data on crystallographic relationships of this sort, chiefly derived from Mehl's work, in Table 1. It is evident at once that the crystal structure of the matrix does not uniquely determine the location of the plate structures or other forms assumed by the segregate. Thus, the first seven items in the table represent precipitation of platelike forms from a matrix of face-centered cubic structure, but these plates are found stratified along planes of form {111 }, {100 }, {110 } or another unidentified form, depending upon specific conditions involving the nature and concentration of the segregating substance. Again, in a matrix of body-centered cubic structure, plates, needles or stars have been found and the favored inner surface of segregation may be a plane of form 11101, 11121, 12101, or indeed a surface without crystallographic indices defining a bundle of close-packed lattice lines (strings of atoms) taken more or less at random. Such a surface might be formed ideally by revolving one of these parallel strings of atoms around another as an axis. Hanemann and Schroder2 recently discussed these questions, principally with relation to original conceptions concerning the process of

Jan 1, 1932

By D. W. Smith, C. H. Mathewson

In a series of papers published recently,' R. F. Mehl and associates have studied the characteristics of form and orientation of many segregate structures and have found diversified conditions which can be explained only by postulating some specific kind of cooperative mechanism permitting the matrix lattice in some favorable region to transform into the segregate lattice in an adaptable region presenting a geometrically similar arrangement of atoms. We have collected the published data on crystallographic relationships of this sort, chiefly derived from Mehl's work, in Table 1. It is evident at once that the crystal structure of the matrix does not uniquely determine the location of the plate structures or other forms assumed by the segregate. Thus, the first seven items in the table represent precipitation of platelike forms from a matrix of face-centered cubic structure, but these plates are found stratified along planes of form {1111, {loo), {110) or another unidentified form, depending upon specific conditions involving the nature and concentration of the segregating substance. Again, in a matrix of body-centered cubic structure, plates, needles or stars have been found and the favored inner surface of segregation may be a plane of form (110), (112), {210), or indeed a surface without crystallographic indices defining a bundle of close-packed lattice lines (strings of atoms) taken more or less at random. Such a surface might be formed ideally by revolving one of these parallel strings of atoms around another as an axis. Hanemann and Schroder2 recently discussed these questions, principally with relation to original conceptions concerning the process of

Jan 1, 1932