Of the various kinds of joints thus far reviewed we have found the ordinary bell-and-spigot joint the most defective. As has been shown, it is faulty (1) in the manner in which the packing material is applied, and (2) in the position in which it is held; (3) in the lack of any provision for protecting it against the effects of strains in any direction, due to variations of temperature or other cause, by which the spigot soon becomes loosened from the packing in the socket; (4) in the temptation it presents for carelessness and fraud; (5) in the difficulty of disjointing it for repairs or alterations; (6) in its inflexibility, and (7) in its costliness.

These defects are inherent in the nature of the joint. The calking may, it is true, be done by hydrostatic pressure or other mechanical means, and some of the difficulties involved by the customary laborious and unsatisfactory manual process might be avoided, but the old way is adhered to, and all the defects as a matter of fact remain. The use of grooves cast in the bell or spigot with a view to holding the lead in place may, in certain cases, alleviate the evil effect of strains or jars, but it evidently cannot remove it.

The sleeve joint is, in principle, another form of bell-and-spigot joint. It has all the objections of the latter, and adds one of its own, in that it doubles the number of calkings and packings required, and complicates the form. Its object appears to be the connection of wrought iron pipes upon which bells or caps cannot be readily formed in a single piece.

The screw joint is also particularly intended for wrought iron pipe. It is a great improvement upon the foregoing two classes, removing the first four, but not the last three defects.

The flange joint is suitable for cast iron and removes many of the defects enumerated above. It has, however, as we have so far illustrated it, the serious defect of being non-adjustable, and it is absolutely inflexible or inelastic. Moreover, as heretofore used, the flange joint cannot be made tight under pressure without considerable expense in fitting or planing the faces of the flanges, and the means employed for bringing the flanges together have been imperfect and unsatisfactory, requiring considerable working room, and rendering their use in contracted spaces inconvenient or altogether impossible.

Flexibility in pipe jointing seems to be generally considered by designers as either undesirable or unattainable, absolute rigidity being the ideal sought.

Nevertheless I am now convinced after a quarter of a century's experiment and study of the subject that no rigid joint will ever be entirely successful in plumbing, gas or water piping, and that flexibility is really an absolutely essential quality.

In plumbing the shrinkage or settlement of the building after the pipes have been installed, and the racking caused by severe changes of temperature due to the alternate passage of cold and hot water or even steam through the pipes are bound in time to destroy either the joint or the fixture to which it is connected.

In gas and water piping settlements in the streets, so frequently torn up and rebuilt, and distorted by the construction of new buildings, and the continual changes of temperature following the changes of the seasons, crack open the rigid joints and cause the colossal losses to which we have already called attention.

The water carriage system of disposing of the organic wastes of a community has been shown to be the most economical and satisfactory one now known, but such a sewerage system is manifestly of little value without the coexistence of a public water supply, because otherwise the efficient flushing of the sewers and plumbing fixtures, especially of the water closets, becomes too expensive and unreliable for satisfactory results.

A public water supply is also essential as a rule for ensuring pure drinking water, and, for domestic and commercial purposes, a soft water instead of the hard water obtained from wells.

A public supply is also essential for proper fire protection. It is more economical for the people to contribute to the maintenance of a public water supply than to stand the high fire losses and insurance rates resulting from the want of one.

Other important public uses of a general water supply are street sprinkling and watering of parks, public and private lawns and fountains and the like, and the benefits coming from a good public water supply increase the value of the property of any community, however small, far beyond the cost of its installation. Hence it is only a question of time when such supply will become practically universal.

Inasmuch as the joint of the usual cast iron pipe of our public water mains is the weakest part of the system, it is that which must be held responsible for most of the leakage occurring. A rigid joint, moreover, has no power to protect the pipe itself from fracture under external pressure, as has the flexible joint.

The comparatively rare actual fracture of the pipe itself, however, would be more likely to be made known by a sudden lowering of the pressure on the system as at the pumps, and lead to its repair. But the joints leak all the time soon after laying and cannot be made tight, but must, on the contrary, inevitably grow worse every year. As the streets are dug up from time to time, the pressure on the mains becomes more and more unequal, and the contraction and expansion due to the conveyance alternately of nearly freezing winter water and comparatively warm summer water through the pipes, working upon an almost absolutely inelastic material like lead, must in time open the joint. Lead can never of itself return to its place in the bell and spigot joint when once compressed or when gradually drawn from its socket by the constant longitudinal play back and forth of the comparatively elastic iron, so that serious leaks should be expected at every joint, and they do there develop and go on increasing as the disturbing causes are repeated until they reach the enormous proportions so many able investigators have recorded.