On Chain Cables or Moorings. 299 



action pf the said forces, would be the same as ia the former, 

 but reversed in its operation : that is, the matter in thesa parts 

 of the inner circmi'fereiice presents an infinite nuniber of ful- 

 crunis, over wjiich the onter circumference must be ruptured and 

 separated by the said forces. A circular is therefore a bad 

 form ; but from the foregoing it is obvious, that if the parts A 

 and B of the circular liiik, figure 1 , can be prevented from ap- 

 proximating each otlier, tiie evil that has been |)o:ated out will 

 be lessened. Suppose a stav AGB to be introduced for this 

 purpose, and, as before, let receding forces be employed in the 

 directions CE and DF, what will be the effect ? The circular 

 link will now be able to resist a greater force than before, having 

 two points of support ; but the unsupported parts between the 

 points C, B, D and A, will, by the effort of tlie said force, en- 

 deavour to assume a quadrilateral form, somewhat like figure 3: 

 a chan.ge that cannot be effected without a derangement of the 

 matter in the link, which must rupture and destroy it. Such 

 stays as AGB (figures 1 and 3) have been used in chains; but 

 such a stav only supports two ojjposite points in the link ; ^nd 

 I have shown that the tendency of receding forces, applied as be- 

 fore described, is to straighten and consequently to rupture the 

 parts that arc still left unsupported. — My said improvements in 

 chain cables or moorings are founded on considerations drawn 

 from the facts that have been alluded to. If a circular link, 

 instead of being supported only in the two opposite points 

 A and B, have its opposite sides supported by a stay embracing 

 two considerable and opposite segments, suppose HI and KL, 

 by the stay HIKL, taking care to leave such openings as shall 

 allow sufficient plav for the links to be received into it, the link 

 will be much stronger than with such a stay as AGB : but still 

 the link will prove to be of a bad form ; for the tendency of re- 

 ceding forces, applied as I)efore, would break the piece MOKCli, 

 over tlie point C, as a fulcrum ; and the piece NPLDI, over 

 the point D, as a nilcrum. And moreover, even if circular links 

 could be made unobjectionable as to strength, they should be 

 avoided, on account of the greater weight of metal which a givciii 

 length of chain would recpiire than when formed of links of a 

 less exceptionaljle form. We have seen that the tendency of 

 receding forces, applied to cun'cd links, is to draw portions of 

 them into straight forms ; and hence it follows, that twisted 

 links of every kind should be avoided where strength is required; 

 for such links, even if their opposite sides be supported by an 

 interposed stay, like A(JB, must, by the application of a suffi- 

 cient strain, untwist themselves to become straight, and thus 

 ^ave the arrangement of their p;uticles disturbed. As the ten- 



d«if-y 



