Before considering in detail the causes of explosion, it is necessary to recall to mind that beyond question there is sufficient accumulated force in any working boiler to cause all the violent effects of an explosion, if this force be suddenly liberated. In Figs. 13 and 14, No. 18, 1869, and No. 63, 1866, are shown the violent effects of the rupture of vessels employed for steaming rags, which were filled with steam only. In ordinary boilers however there is present, besides the steam, a quantity of water heated much beyond the atmospheric boiling point; and when rupture takes place and the pressure is suddenly relieved, part of this water evaporates, and keeps up the supply of steam to continue the rupture and destruction. The explosion of a boiler differs from the discharge of electricity or lightning, which cleaves the air and instantly leaves a vacuum; it also differs from the discharge of detonating compounds which act suddenly and leave a vacuum; but it more nearly resembles the discharge of gunpowder, which burns sufficiently slowly to keep up a continuous pressure behind a projectile until it leaves a gun; and each cubic foot of water in a boiler working at 60 lbs. pressure has been shown to produce in steam an explosive effect equal to one pound of gunpowder. None of the elaborate but unlikely theories of decomposed steam, or of electric accumulations, suppose a force so fitted to cause destruction as that contained in the highly heated water existing in all working boilers. * * * * * The following appear to be the general results to be derived from the experience of the explosions in this country during the last four years. [Illustration: _Fig. 15._] [Illustration: _Fig. 16._] First as to faults of construction which fall under the department of the boiler maker or repairer. One of the most apparent causes of explosion in stationery boilers is the loss of strength occasioned by frequent repair, not only from the injury done to the old plates by removing rivets, but from the want of bond in the new work. This has lead to many of the explosions of the Plain Cylindrical boilers, such as are shown in Figs. 15, 16, and 17, No. 45, 1869, No. 32, 1870, and No. 20, 1870. Where the plates are arranged longitudinally instead of in rings, the danger is increased, as there is less chance of a dangerous rip being arrested by a crossed joint. So great a number of boilers with continuous longitudinal seams, especially in the North, have worked for twenty or thirty years, that it can hardly be supposed they are any weaker than the boilers made in rings; but they are more liable to explode, for if a seam rip occurs, it more easily extends along the seam, and leads to the general break up of the boiler, shown in Fig. 18, No. 59, 1869. [Illustration: _Fig. 17._] [Illustration: _Fig. 18._]