of the drop has been increased from 140 to 200 pounds, in accordance with the same experiments. The specifications otherwise stand as drawn last year."

The Specifications for Gray Iron Castings are submitted without change for reasons which appear in the following communication from Mr. Henry Souther, Chairman of the sub-committee, responsible for the same:

"As chairman of the sub-committee on Specifications for Gray Iron Castings, the writer begs to make the following report, following the resolution passed at the meeting on June 17, 1904:

"At the last Annual Meeting the Specifications for Gray Iron Castings were referred back to the committee for further study, with instructions to report their conclusions.

"Since that meeting some study has been given the matter by its members, notably Professor Lanza. His investigations found that the time element was sufficiently accurate for the purposes of this test, and he has, therefore, withdrawn all objections on this score.

"Another very pertinent criticism was to the effect that no cast iron tension test should be recognized that had not been made with the best approved ball joints. This provision is not found in the specifications because no well recognized ball joint has been standardized. There are many in use that are undoubtedly good, but they differ greatly in their efficiency, and are not all sufficiently alike in their behavior to produce firstrate results. If such a standard set of ball joints can be devised, then the use of that particular ball joint might well be incorporated in the specifications.

"The round test bars are giving uniformly good results in this laboratory and have created no friction in the foundries called upon to cast them, the only difficulty in connection with them being that the foundrymen feel called upon to finish them in some way, either by tumbling or grinding. Also the full importance of closely following the casting directions is not grasped by the average foundryman.

"It is certain that if the specifications, as printed, are followed and lived up to by all concerned, the quality of such castings may be controlled by them, and that being the object of these specifications, it would seem to your committee that they fulfill the purpose for which they were intended."

The proposed Standard Specifications for Cast Iron Car Wheels and Gray Iron Castings are appended to this report. Respectfully submitted on behalf of the Committee,

WALTER WOOD, Chairman.

RICHARD MOLDENKE, Secretary.

APPENDIX I.

PROPOSED STANDARD SPECIFICATIONS FOR CAST-IRON

CAR WHEELS.*

The wheels furnished under this specification must be made Chemical Properties. from the best materials, and in accordance with the best foundry methods. The following pattern analysis is given for information, as representing the chemical properties of a good cast-iron wheel. Successful wheels, varying in some of the constituents quite considerably from the figures given, may be made:

1. Wheels will be inspected and tested at the place of manufacture.

2. All wheels must conform in general design and in measurements to drawings, which will be furnished, and any departure from the standard drawing must be by special permission in writing, and manufacturers wishing to deviate from the standard dimensions must submit duplicate drawings showing the proposed changes, which must be approved.

3. The following table gives data as to weight and tests of Drop Tests. various kinds of wheels for different kinds of cars and service:

*Adopted by letter-ballot of the Society on September 1, 1905.

Marking.

Measures.

Finish.

4. Each wheel must have plainly cast on the outside plate the name of the maker and place of manufacture. Each wheel must also have cast on the inside double plate the date of casting and a serial foundry number. The manufacturer must also provide for the guarantee mark, if so required by the contract. No wheel bearing a duplicate number, or a number which has once been passed upon, will be considered. Numbers of wheels once rejected will remain unfilled. No wheel bearing an indistinct number or date, or any evidence of an altered or defaced number will be considered.

5. All wheels offered for inspection must have been measured with a standard tape measure and must have the shrinkage number stenciled in plain figures on the inside of the wheel. The standard tape measure must correspond in form and construction to the "Wheel Circumference Measure" established by the Master Car Builders' Association in 1900. The nomenclature of that measure need not, however, be followed, it being sufficient if the graduating marks indicating tape sizes are one-eighth of an inch apart. Any convenient method of showing the shrinkage or stencil number may be employed. Experience shows that standard tape measures elongate a little with use, and it is essential to have them frequently compared and rectified. When ready for inspection, the wheels must be arranged in rows according to shrinkage numbers, all wheels of the same date being grouped together. Wheels bearing dates more than thirty days prior to the date of inspection will not be accepted for test, except by permission. For any single inspection and test only wheels having three consecutive shrinkage or stencil numbers will be considered. The manufacturer will, of course, decide what three shrinkage or stencil numbers he will submit in any given lot of 103 wheels offered, and the same three shrinkage or stencil numbers need not be offered each time.

6. The body of the wheels must be smooth and free from slag and blowholes, and the hubs must be solid. Wheels will not be rejected because of drawing around the center core. The tread and throat of the wheels must be smooth, free from deep and irregular wrinkles, slag, sand wash, chill cracks or swollen rims, and be free from any evidence of hollow rims, and the throat and thread must be practically free from sweat.

and Chill.

7. Wheels tested must show soft, clean, gray iron, free from Material defects, such as holes containing slag or dirt more than onequarter of an inch in diameter, or clusters of such holes, honeycombing of iron in the hub, white iron in the plates or hub, or clear white iron around the anchors of chaplets at a greater diştance than one-half of an inch in any direction. The depth of the clear white iron must not exceed seven-eighths of an inch at the throat and one inch at the middle of the tread, nor must it be less than three-eighths of an inch at the throat or any part of the tread. The blending of the white iron with the gray iron behind must be without any distinct line of demarcation, and the iron must not have a mottled appearance in any part of the wheel at a greater distance than one and five-eighths inches from the tread or throat. The depth of chill will be determined by inspection of the three test wheels described below, all test wheels being broken for this purpose, if necessary. If one only of the three test wheels fails in limits of chill, all the lot under test of the same shrinkage or stencil number will be rejected and the test will be regarded as finished so far as this lot of 103 wheels is concerned. The manufacturer may, however, offer the wheels of the other two shrinkage or stencil numbers, provided they are acceptable in other respects as constituents of another 103 wheels for a subsequent test. If two of the three test wheels fail in limits of chill, the wheels in the lot of 103 of the same shrinkage or stencil number as these two wheels will be rejected, and, as before, the test will be regarded as finished so far as this lot of 103 wheels is concerned. The manufacturer may, however, offer the wheels of the third shrinkage or stencil number, provided they are acceptable in other respects, as constituents of another 103 wheels for a subsequent test. If all three test wheels fail in limits of chill, of course the whole hundred will be rejected.

8. The manufacturer must notify when he is ready to ship Inspection and Shipping. not less than 100 wheels; must await the arrival of the Inspector; must have a car, or cars, ready to be loaded with the wheels, and must furnish facilities and labor to enable the Inspector to inspect, test, load and ship the wheels promptly. Wheels offered for inspection must not be covered with any substance which will hide defects.

9. A hundred or more wheels being ready for test, the Inspec

Retaping.

Drop Tests.

tor will make a list of the wheel numbers, at the same time examining each wheel for defects. Any wheels which fail to conform to specifications by reason of defects must be laid aside, and such wheels will not be accepted for shipment. As individual wheels are rejected, others of the proper shrinkage, or stencil number, may be offered to keep the number good.

10. The Inspector will retape not less than 10 per cent of the wheels offered for test, and if he finds any showing wrong tape-marking, he will tape the whole lot and require them to be restenciled, at the same time having the old stencil marks obliterated. He will weigh and make check measurements of at least 10 per cent of the wheels offered for test, and if any of these wheels fail to conform to the specification, he will weigh and measure the whole lot, refusing to accept for shipment any wheels which fail in these respects.

11. Experience indicates that wheels with higher shrinkage or lower stencil numbers are more apt to fail on thermal test; more apt to fail on drop test, and more apt to exceed the maximum allowable chill than those with higher stencil or lower shrinkage numbers; while, on the other hand, wheels with higher stencil or lower shrinkage numbers are more apt to be deficient in chill. For each 103 wheels apparently acceptable, the Inspector will select three wheels for test-one from each of the three shrinkage or stencil numbers offered. One of these wheels chosen for this purpose by the Inspector must be tested by drop test as follows: The wheel must be placed flange downward in an anvil block weighing not less than 1,700 pounds, set on rubble masonry two feet deep and having three supports not more than five inches wide for the flange of the wheel to rest on. It must be struck centrally upon the hub by a weight of 200 pounds, falling from a height as shown in the table on page 1. The end of the falling weight must be flat, so as to strike fairly on the hub, and when by wear the bottom of the weight assumes a round or conical form, it must be replaced. The machine for making this test is shown on drawings which will be furnished. Should the wheel stand without breaking in two or more pieces, the number of blows, shown in the above table, the one hundred wheels represented by it will be considered satisfactory as to this test. Should it fail, the whole hundred will be rejected.