The disadvantage of raising the water to a higher level than is actually required, seems to have given rise to Mr. \Vitty's improvement on the pump. Smeaton had already remedied this evil, as we have just seen, in ships' pumps. In distilleries, &c. as well as in ships, it is used to force the water to the top of the barrel, and allow it to run off to a lower level. Hence if the water descends from the top of the pump to a place of delivery much below the top of the barrel, the fall of the water through this height is a mechanical force which is entirely wasted, and which might be advantageously employed in raising the water through part of the pump barrel. Mr. Witty avails himself of this power in the following manner : " Instead," says he, " of letting the water or liquid escape from a com mon pump, at the usual place of delivery, I caused it to descend again in a syphon pipe to the lowest level at which it can conveniently be delivered; and as this descent is considerable in ships, brew-houses, kc. a considerable saving of labour is effected in working pumps, by a de scending column of water or liquor counterbalancing as much in length of the rising column in the pump as the height which it descends in the syphon pipe to the place where it can be delivered." If we consider the water which in ordinary pumps falls' from the top of the barrel to the place of its reception as a mechanical force which is lost, we may avail ourselves of it, by various contrivances, for assisting in the work to be performed. In Mr. Witty's contrivance, the men at the pump raise the water to the bottom of the short leg of the syphon, and it is then drawn through the syphon by the water of the larger branch. There are many cases, how ever, where we may allow the workmen to raise the water to the top of the barrel, and employ the direct force of the descending fluid to work another pump, or perform any other piece of work that may be required.
17. On Chain Pumps.
In our article HynRoDYNAmics, we have already given a drawing and description of the chain pump, both as con structed with plugs and with buckets, (see Plate CCCXX. Fig. 6. and Plate CCCXXIV. Fig. 4. and 5.) but as these pumps have been found by long experience to be the most useful at sea, and the least liable to be deranged by acci dental causes, we propose to give some farther account of them in this place.
Every English ship of war has four chain pumps, and three hand pumps, which arc all fixed in the same well. The old chain pumps were very defective machines previous to the improvements made upon them by Mr. Cole. The sprocket wheels w, Fig. 5. Plate CCCXXIV. had no contrivance to prevent the chain from sliding or jerking back on the surface of the wheel. The links were not only too short, but were ill united, and hence they created much friction in passing round the sprocket wheels, and often broke in situations of a critical kind. In Mr. Cole's improvement, the links are formed of two long plates of iron, with a hole at each end, and fixed together by two bolts, which act as pins for the joints. The buckets
or saucers a, b, c, Fig. 4. Plate CCCXXIV. are circular plates of brass, with a piece of leather between them ; and the sprocket wheels NV, NV, are formed like the trundles used in mills. They consist of two iron wheels, fixed at eight inches distance on the axle, and united by several round iron bolts. The links of the chain have hooks which rest on these bolts, and the chain is thus kept upon the wheel, and prevented from starting back when loaded with a column of water. Mr. Cole has constructed his chain pump so that the chain may be taken up and repair ed, or any ballast removed with which it may be choaked. In a comparison of Mr. Cole's pump with one of the old construction, it was found to raise one ton of water in 431 seconds, with the power of four men, whereas the old pump required seven men to raise one ton of water in 76 seconds. Mr. Cole's pump was introduced into the navy more than thirty years ago, and the principal alteration it has experienced since that time is the substitution of a curved metal tube in place of the lower sprocket wheel, as the chain passes over it more easily than over a wheel. The cranks are now made to take off, and they are large cnough to permit thirty men to work at once.
M. Hachette has given in his 7'raite des Machines, Plate ix. a detailed drawing of all the parts of an improved chain pump, to which the reader is referred.
Various experiments have been made on the effect of chain pumps, both vertical and inclined. With a vertical one employed in the construction of the Pont de la Con corde at Paris, ,four men raised 2000 cubic feet (63.55 cubic metres) of water to the height of feet (5.3 metres) in one hour, or 363.315 cubic metres of water to the height of one metre, or 90.83 for each man. M. Hachette considers this as too great a mean of the-force of four men.
M. Perronct, in building the bridge of Orleans, observed that a vertical chain pump, wrought by twelve men, divided into three relays, raised in 24 hours, to the height of 15 feet, (4.87 metres,) 500 cubic feet (17.14 cubic metres) of water. M. Perronet, however, thought that the exertion was in this case extraordinary, and he pro posed to reduce the result one-sixth, to make it applicable in ordinary cases. With this reduction, the daily work of a man would be 139 cubic metres of water raised a metre.
M. Perronet found that an inclosed chain pump raised in an hour 1998 cubic feet of water to the height of 12 fcet,or 23976 cubic feet to the height of one foot. He mentions, also, that two inclined chain pumps, driven by 36 horses, divided into several relays, raised 1177.20 cubic feet of water to the height of 15 feet, or 1765800 cubic feet to the height of one foot ; and as a horse is equal to seven men, we have 7 x 36 = 252 men ; hence the daily work of one man, as deduced from this experiment, is 7007 cubic feet of water raised one foot. For farther details respecting these experiments, see Hachette's Traite Etc mentaire des Machines.