obtained, the use of amounts of .1 per cent. or more of metallic titanium in the manner mentioned, prevents the "honeycombed" condition of the upper part of the ingot found in plain Bessemer steel, but is also attended with a larger and deeper "pipe." The heavy segregation or concentration of carbon, phosphorus and sulphur found in the interior and upper part of ingots of plain Bessemer steel is largely restrained, but the mild negative segregation found in the interior and lower part of the ingot is not materially altered. The brittle zone found in rail of plain Bessemer steel from the upper part of the ingot, as determined by drop and tensile tests, was avoided, but the properties of the rail from the lower two-thirds of the ingot were not changed. Large internal flaws were found in rail considerably lower down from the top of the ingot in steel treated as mentioned, than in rail made from plain steel. Treatments with .05 per cent. metallic titanium produced the above results only in part, but treatment above 1 per cent. had only little additional influence.

Report 25 dealt with an investigation of two special ingots made by the Standard Steel Works Company at Burnham, Pa., by a process which prevents the formation of a "pipe" in the interior of the ingot. The steel was acid Open-Hearth steel treated with titanium. The ingots were shipped to the Maryland Steel Company at Sparrows Point, Md., where they were tested. The main feature of the casting process was a sand core on top of the iron mold. The ingots were cupped down at the top but contained no interior pipe.

Report 26 gave a description of a method for determining the transverse ductility of the bottom of the base of a rail and the load required to break the flanges of a rail supported near the edges of the flanges. The results were given of a few tests made at Buffalo at the works of the Lackawanna Steel Company, of Bessemer and Open-Hearth rails. The method of making the tests was to support a piece of rail about 2 feet long on two supports placed opposite each other near the edges of the flanges under the middle of the length of the rail. The supports were six inches long and placed one-half inch in from the sides of the flanges. The load was applied in the test machine to the top of the rail at the middle. The method may be considered a means of determining the strength of the flange and of determining the transverse properties of the base of the rail, as regards the transverse ductility of the metal in the base and the presence of structural flaws such as seams.

Report 33 gave the results of an investigation made at the Gary works of the Illinois Steel Company on the influence of silicon on Open-Hearth ingots and rails. A heat was used of about .15 per cent. silicon and a series of higher silicons was obtained in this steel up to above .5 per cent. by means of mold additions of finely crushed ferro-silicon. With about one-fourth per cent. silicon or more the ingots were free from most of the honeycomb present in the upper third of the ingot with the heat amount of silicon, but they also had larger pipes. The higher silicons also had less concentrated segregation of carbon, phosphorus and sulphur. Sili

con had but little influence on the results in the drop test. When tested in the test machine as a beam, the stiffness and breaking load of the rails increased with increase of silicon, while their ductility was not greatly influenced. In longitudinal tensile tests, the yield point and tensile strength increased somewhat with increase of silicon while the ductility remained about the same. In tests of the flange, the load required to break the flange increased somewhat as the silicon increased, while the transverse ductility of the base remained about the same.

The above work concerning both titanium and silicon indicate that they tend to restrain segregation of carbon, phosphorus and sulphur, but used as they were in this work they were attended with larger pipes. This suggests the idea that in order to obtain the full benefits of their use, a method of casting the ingots is necessary whereby the pipe is avoided or minimized.

MISCELLANEOUS.

Testing Each Ingot:

The question of testing each ingot has been quite thoroughly considered by correspondence and by discussion, and the Committee recommends that no change be made in the specifications in this respect at present, for the following reasons:

(a) The tests now required by the specifications are sufficient to determine the character of the metal in the heat. In so far as it is intended to discover specific flaws in the rail, additional tests would have relatively little value, as these defects are quite local, and are apt to occur in any part of the ingot, or rail bar; so that the presence or absence of internal defects in the piece tested is no certain criterion of the presence or absence of similar defects in other parts of the same bar. (b) The character of the metal in the rails of one heat varies down the ingot more than it does from one ingot to another, and in making the complete drop test of an open-hearth heat, the three tests made from approximately the poorest parts of the ingots designate the average of the poorest part of the metal fully as well as if we tested a piece from each ingot.

(c) Careful examination in detail of the many tests of open-hearth rail made in the last two years on the New York Central Lines fails to show any necessity for the testing of each individual ingot.

Ductility and Elongation Tests, Appendix "A": (Bulletin No. 147.)

Dr. Dudley has written a description, in Appendix "A," of the ductility and elongation tests as conducted on the New York Central Lines, which was intended to be a part of last year's report, in Bulletin No. 143. Such information is interesting and instructive, as we like to keep informed of how our fellow-members handle the daily problems which come before them.

Influence of Seams or Laminations in Base

of Rail on Rail Failures, Appendix "F": (Bulletin No. 147.)

Mr. H. B. McFarland, Engineer of Tests of the Santa Fe System, has furnished the Committee with an important study of base failures, which supplements the work of like character presented by Mr. Wickhorst in Appendix "E," entitled "Transverse Ductility of Base of Rails." This type of failure has been dwelt upon at great length by Mr. James E. Howard, Engineer-Physicist of the Bureau of Standards, in the report of accident on the Great Northern Railway, near Sharon, N. D., December 30, 1911. The Committee is making a special study of this character of failure, in order to devise a practicable test which will result in its discovery before the rails leave the mill.

Effect of Piping, Cavities and Porous Spots

in Ingots on the Finished Rail, Appendix "H": (Bulletin No. 147.) Our knowledge on this subject has been improved through the study by Mr. Onderdonk, Engineer of Tests of the Baltimore & Ohio Railroad System. The study is a very important one, as it is impossible to have sound rails from unsound ingots.

Investigation of Silvery Oval Spots, Sometimes Called

"Transverse or Internal Fissures," in the Rail Heads, Appendix "I":

The contribution by Mr. C. D. Young, Engineer of Tests, and by Mr. F. N. Pease, Chemist, of the Pennsylvania Railroad, on this subject is a valuable addition to our knowledge, as there has been quite a dispute over the cause of this type of defect, which was but recently brought to our attention. It is quite possible that such defects have heretofore caused rail failures, but if such be true, it has not been a matter of general knowledge.

Method of Producing Sound Ingots, Appendix "J":

This is another appendix relating to sound ingots, which we have been extremely fortunate in securing for the information of our members, as the articles are by a Past-President of the Iron and Steel Institute of Great Britain, Sir Robert A. Hadfield, a manufacturer of steel, and a profound student of the problems arising in its manufacture. This information is an important addition to that of Mr. Wickhorst's report on "Pipeless Ingots," in Appendix "D," as there is some difference in the practice described in the two papers, which Sir Robert has himself explained in a note in the introduction to his reports.

The Testing of Rails, Appendix "L":

This is a foreign paper, taken from the Bulletin of the International Railway Congress, which contains information in line with Mr. Wickhorst's "Abrasion Tests on Revolving Machine," issued as Appendix "B."

(E) RECOMMENDATIONS FOR WORK IN 1913.

The subjects recommended for assignment to the Committee for 1913 are the same as those already assigned for the year 1912, and in addition the subject of rail joints.

The general line of investigation which the Committee has in view for Mr. Wickhorst is submitted below, and embraces a great deal more work than he can cover in any one year, but it is well to keep before us the subjects which are important and will demand our attention.

The main point kept in mind in the work of the last two or three years has been to conduct it so as to bring out if possible information useful in improving rails for the purpose of making them uniformly safe, and it is probable that this must continue to be our guiding principle for some time to come. Investigations intended to improve the wearing properties of rails must, it would seem, be considered as secondary to those which have uniform safety as the prime consideration. It soon developed that the type of rail failure which was uppermost in our attention a few years ago, namely, split and mashed heads, was traceable to the interior condition of the ingot and our work has consisted therefore largely of "ingot" studies. Recently, failures classified as broken rails and broken base, have been brought more to our attention, due largely to last winter's crop of broken rails. It seems that such failures are largely attended or perhaps caused by deep seams in the base of the rail and it is now proposed that the work of the immediate future be directed toward throwing light on the cause of such seams and methods for their prevention. Among the subjects needing investigation the following may be listed: Making Ingots:

(1) Influence of height of ingot on segregation and interior cavities, Bessemer and Open-Hearth steel.

(2) Influence of diameter of ingot, Open-Hearth steel.

(3) Influence of the rate of pouring the ingot.

(4) Influence of temperature of liquid steel when poured into molds. (5) Influence of thickness of mold.

Making Rails:

(6) Influence of temperature of rolling on high carbon Open-Hearth rails.

(7) Causes of seams in base of rails.

(8) Influence of rate of reduction in rolling.

(9) Influence of methods of cooling on cooling beds.

(10) Effect of cold straightening rails.

(11) Influence of length of time in soaking pit on grain size.

Composition:

(12) Influence of titanium on Open-Hearth ingots and rails.

(13) Quantitative influence of carbon on deflection and ductility. (14) Quantitative influence of phosphorus on deflection and ductility. (15) Quantitative influence of manganese on deflection and ductility.

ing them up to date and making them as perfect as possible, with the result that a few revisions are now submitted for adoption by the Association, some of them being merely changes in wording and rearrangement, which it is hardly necessary to point out in detail. The principal changes are, however, laid before you for approval, as follows:

Section 4. Chemical Composition:

The words "of each heat" have been added, so that the section reads, "The chemical composition of each heat of the steel from which rails are rolled, determined as prescribed in Section 7, shall be within the following limits."

This change was made because the question arose at one mill as to whether it was necessary for every heat to have its chemical elements within the limits prescribed by the specifications.

Section 14. Elongation or Ductility:

A new paragraph (b) has been added, as follows: "A sufficient number of blows shall be given to determine the complete elongation of the test piece of at least every fifth heat of Bessemer steel, and of one out of every three test pieces of a heat of Open-Hearth steel."

In measuring the elongation, acceptance or rejection is determined by the amount shown under one or more blows of the tup, but in addition to this, it is advisable to determine the total elongation or ductility of a certain number of test pieces, in order to keep informed on the toughness of the material, by repeating the blows till failure results. It is at present customary to carry on these tests, and it was thought advisable and proper to add the requirement to the specifications in order to make them more complete.

Section 17. Bessemer Process Drop Tests, Clause (b):

The words "does not break and" have been added to the second line, so that the clause reads, "If the test piece breaks at the first blow or does not show the required elongation (Section 14), or if the test piece does not break and shows the required elongation, but when broken shows interior defect, all of the top rails from that heat shall be rejected." The words "nicked and" have also been omitted from both Clauses (a) and (b) and both the additions and omissions were brought about by discussion as to the precise meaning of the clause. Some thought that when a test piece showed interior defect the rails it represented should not be rejected unless the test piece had been nicked and broken. This was not the real meaning of the clause, as nicking is only a matter of convenience and expedition in making the tests, but in order to avoid controversy, the changes have been made.

Similar changes have been made in (c), (d), (e) and (f).