of an inch in the thinnest part. The sulphur will readily flow into this space if it has not been kept hot so long as to become thickened or waxy, and in fourteen years' experience I have never had a case where it worked loose.

I knew of one case, however, where the parties erecting the engine neglected to fill the anchor bolt holes with clay or cement, which resulted in the sulphur running into the holes and uniting the bolts so fast to the cap stone that they could not be budged. When the foundation settled, it naturally loosened at the bottom of cap stone, and no amount of tightening on the anchor bolts could help it. The remedy was to wedge up under the cap stone and pour in thin cement, which answered very well, as the engine had light work; but more heroic measures would probably have been necessary under more exacting conditions. In this instance, of course, the trouble should not be chargeable to the use of sulphur, but to the improper manner in which the work was done.

At the same time, I agree with Mr. Lyne that cement will make a good job, as I have frequently used it; and good cement, especially Portland, will give satisfactory results, as it becomes very hard. It requires, however, more space between the bed and cap stone than does the sulphur. In the case Mr. Lyne refers to, the fact that the sulphur was one-sixteenth of an inch thick seems to me to be one reason why it proved a failure.

Mr. C. S. Dutton.-In this connection, yesterday, some member asked for the result of experience in regard to lubrication of engines. It reminded me of an experience I had recently in an engine-room where an engine cylinder failed to lubricate, for what reason is not yet determined, but the simple fact was that the lubrication was imperfect. The valve was a balanced Allen valve. There was no difficulty with that, but the cylinder did not lubricate. I will not go any further into the conditions than to say that it was a 30 x 36 engine running about 135 revolutions. It was continuous work, running wood-pulp grinders from Monday morning to Saturday night. It was first tried by the users of the engine to put additional oil-cups at the indicator holes; but as the ports were on the side, nearly all the lubrication that went in that way was carried out, and what was finally done was to tap holes in the center of the cylinder at the bottom and the top and to attach an ordinary sightfeed lubricator and feed the oil into the center of the cylinder at the bottom, which obviated all difficulty; and there has been no further trouble with it.

Mr. A. K. Mansfield.-I would like to suggest, in regard to the method of making valve stems, that in these times it is customary with some builders to use cold-rolled steel for such purposes, and the custom also is, for some kinds of engines-perhaps the larger sizes to avoid turning down the valve stem, but to get a new stem when the old one is worn.

Mr. J. H. Cooper.-I would simply suggest that it has been our custom to make that part larger, so that, if it wears unequally, you can turn it down and save all the other parts. We found also that they are better made of wrought iron than steel. Wrought iron will last a great deal longer than steel will for such purposes.

Mr. E. F. C. Davis.-I would like to say our experience is just the reverse of Mr. Cooper's. We find that steel is the best. You must take a piece of steel, and hammer it out yourself. We find that the cure for that sort of wear is to make the stuffing-boxes extremely long and deep, and it will partake somewhat of the qualities of a straight-line engine stuffing-box.

CCCXLVIII.

ON THE LONGITUDINAL RIVETED JOINTS OF STEAM-BOILER SHELLS.

BY JOHN H. COOPER, PHILADELPHIA, PA.

(Member of the Society.)

THE initial statement to the English Lloyd's rules for steamboilers is embodied in the following words: "The strength of circular shells to be calculated from the strength of the longitudinal joints," which assures us that this part of the boiler should be properly proportioned.

To these rules a memorandum is added: "In any case where the strength of the longitudinal joint is satisfactorily shown by experi ment to be greater than given by this formula (Lloyd's), the actual strength may be taken in the calculation."

Later on, Lloyd's rules (under the head of "Periodical Surveys," regarding the examination of boilers after they have been several years in service) say: "The safe working pressure is to be determined by their actual condition."

These statements lie in the line of practical efficiency, and point to the necessity of providing material in accordance with the requirement of the load to be carried.

Any one who takes the trouble to collect and compare data on this subject cannot fail to notice the great disparity of rules for determining the working pressure permissible for boilers.

The case is clear by simple reasoning on the data collated, that boilers are held together, it would seem, more by conformity to rule than by the materials of which they are made.

But of course the true course to pursue is to give to each member its proper allowance of section, in order that the components of the joint shall have an equal chance under strain according to its resisting power.

The diminished strength of the shell of a boiler by the longitu dinal joint is well known, and it becomes good engineering so to proportion its parts as to obtain the greatest strength possible within the limits of practical economy.

When it became necessary to assure themselves confidently of the permanent safety of a structure composed of plates held together by rivets, engineers were not long in finding out that a certain allotment of rivet section to plate section at the joints was necessary, and that these sections were found to be nearly equal in the strongest joints.

The experiments of Fairbairn, conducted in the year 1838, proved that "the sectional area of the rivets in a joint was nearly equal to the sectional area of the plate through the rivetholes."

Subsequent experiments by Clark on riveted plates for the Britannia and Conway Tubular Bridge fully corroborate the above statement. His conclusion was: "The collective area of the rivets is equal to the sectional area of the plate through the rivet-holes.".

This relation of the components of the joint in course of time became embodied in the English Board of Trade rules and in Lloyd's rules now in force, regulating the construction of steamboilers. It also forms the basis of the Philadelphia steam-boiler inspection ordinance, first formulated in 1882.

Referring now to those rules only which relate to the proportions of the longitudinal joints of the cylindrical shells of boilers, we are prepared to say they may be most conveniently presented by the following notation and formulæ :

NOTATION.

A Percentage of punched plate to the solid plate.

=

B = Percentage of driven rivet section to the solid plate.

C = The pressure in lbs. per square inch which the boiler is allowed to carry.

a = Area of driven rivet, or rivet-hole.

S = Ultimate shearing strength of rivets in lbs. per square inch

of section.

T = Ultimate tensile strength of plates in lbs. per square inch

of section.

f= Factor of safety.

E-Limit of elasticity in the plates in lbs. per square inch of

The least of A or B should be inserted in the formula C.

All dimensions in inches.

The notation and the formulæ mutually explain each other.

These formulæ are intended exclusively for the guidance of the inspector in ascertaining the exact strength of the joints in the boilers which come under his care, and which enable him to determine the working pressure of steam allowable under the rules. They do not, however, enable the boiler-maker to determine directly that proportion of pitch which he should use with any given plate thickness and rivet diameter, in order to secure the strongest joint and which will also pass the highest inspection.

To secure these results the following simple formulæ were devised. by the writer (early in 1882), in which the notation given above is. similarly employed, and which may be thus expressed.

For single-riveted joints, when iron plates are secured by iron rivets and when the plate thickness and rivet diameter are given, it is desired to find a pitch that will secure equality of plate and rivet section-the formulæ will be:

This plainly means that the pitch is equal to the area of the rivet-hole, divided by the thickness of the plate, and to the result of which the diameter of the rivet hole must be added.

For multiple riveted joints, when iron plates are secured by iron rivets, the same formula is used with the addition only of n, representing the number of rows of rivets, thus: