ON THE FRICTION OF METAL COILS. 



543 



lating the admission of steam Those who have seen this arrangement 

 at work must have been strnck by the suddenness and force with which 

 the coil comes into action. Notwithstanding these and other applications 

 of coil friction, the actual amount of knowledge on the subject, at any 

 rate in a published form, is very small, and the accepted theory of 



ordinary coil friction, which takes no account of the comparative rigidity 

 of the coil, is not directly applicable in the case under discussion. Mr. 

 Edward Shaw, in attempting to make use of metal coils, found various 

 practical difficulties and apparently anomalous results, which led the 

 authors to think it worth while to make some experiments bearing upon 

 certain features in the application of this form of coil friction. 



Form of coil. — The first points to be considered are those of the 

 requisite form and dimensions of coil ; for it is obvious that while suf- 

 ficient strength is required at the head of the coil to resist a considerable 

 force, the tail need not be of similar section. It is not merely waste of 

 metal to have the section uniform, but the want of flexibility thus entailed 

 seriously impairs the eflficiency of the coil. Fig. 3 will illustrate this 

 point, in which a shows the section of a coil which has been adopted by 

 one or two inventors, whilst i is a section of the coil reduced in thickness, 

 but of uniform width. This hae the disadvantage of being too broad at 

 the tail, and the quantity of lubricant getting under such a surface 

 prevents the initial grip taking place so readily. Fig. 3 c shows a very 

 eflBicient form in which the advantage of smaller space (the number of con- 

 volutions being the same as in h) and more sudden initial grip are insured. 

 A very good form of coil is that shown in d, in which the last coil or two 

 are circular in section. M. Gambaro and Professor Reynolds have both 

 employed coils of decreasing width and thickness. With regard to the 

 diameter of the coil, this, where perfect flexibility is assumed, is a matter 

 of no importance, except as regards the obvious eff"ect upon the leverage 

 at which the resistance acts. With metal coils the case is different, as 

 the larger the diameter the more readily the surface comes to its bearing 

 for a given cross section, and with a given initial clearance the compara- 

 tive distortion is less the greater the radius of curvature. In order to 

 have the internal surface of true cylindrical form, several coils were 



