pressure exceeeds the strength of joint d or f, they are no longer of use, as rupture is alone prevented by the superior strength of the joint e. The welt strengthens the joint simply because it puts the rivet g, Fig. 51, into double shear, that is, the rivet must be sheared through the line h and, and therefore presents double resistance. The part of the plate in front of the rivet must be proportionally increased to meet the increased resistance, and the pitch of the holes increased as much as possible and still retain a tight joint. A combination of this joint and the staggered

о

FIG. 46.

riveted joint would make the strongest possible longitudinal joint, as it would give the greatest amount of metal between the holes, and also present the rivet in double sheet. Fig. 49 shows a mode of making a joint by which the strength might be made equal to that of the solid sheet. The distance between the holes, by making the joint long enough, might be made equal to the width of the sheet, or in excess of it, for injury done in punching. The joint shown in Fig. 52 is not strengthened by the second row of rivets,

barring the grip due to contraction of the extra rivets, as the strength of metal between either row is the same, as is also the number of rivets,

etc.; and the joint is as weak through one row of rivets as the other. We have been considering the longitudinal strength only. The transverse strength of a boiler is that which prevents one end from being torn from the other. This

strength is represented by the thickness of the iron x by the circumference x by the strength of a sectional inch, and the whole less the percentage of loss due to the kind of joint. However, a single rivet joint will suffice, generally,

as the boiler is about twice as strong transversely as longitudinally.

d

Persons buying boilers would do well to have samples of the iron, from which the boiler is to be built, tested for tensile strength, and take no interested parties' word, or bond either, on that point. An actual test settles the matter beyond doubt. All holes should be drilled and rounded in the inner edge, as shown at a, a, a, a, Fig. 50, as a drilled hole, if not rounded, will shear the rivet sooner than a punched hole. Flat surfaces should be stayed to from six to ten times the pressure they are to resist. The area of rivets holding crow-feet to boiler heads, etc., should be equal to the stay coupled to said crow-foot. It is a good plan to make stays coupled to crow feet slightly shorter than the required length. Expand them by heat, and, on cooling, they will all be found tight and necessarily bearing their proportional strain. It is a bad plan to collect one end of stays (such as crown sheet stays when stayed to shell) into a smaller area than the other,

hi

9

abc

FIG. 51.

as it is plain to be seen that the smaller area must resist more than its share in proportion as its area is smaller,