ON THE FRICTION OF METAL COILS. 545 



pulley M, and the loads which are placed in the larger scale-pan "W thus 

 have an effect just twenty times as great upon the head of the coil. The 

 tail of the coil t was attached to a delicate spring balance, S, which in 

 turn was suspended by a cord passing over pulleys, N N. By exerting a 

 greater or less force on this cord any required tangential pressure could 

 be brought to bear upon the tail. 



(a) Coil a consists of seven convolutions of ^in. square cast steel, and 

 therefore not being very flexible, was made a good fit upon the cylinder ; 

 but notwithstanding this a pull of no less than 25 pounds was required to 

 raise the weight of the lever, which was equivalent to a force of 140 

 pounds at the head of the coil. The result of increasing the pull on the 

 tail is shown below : — 



F . . 25 37 41 43 47 51 55 59 63 Q'? 70 

 R . .140 240 260 300 310 350 410 480 520 660 700 



F = force on tail in pounds. R = resistance on head in pounds. 



These results, which are not very uniform, and point to the evil of 

 using rigid coils, were obtained with dry surfaces. The effect of lubri- 

 cating the surfaces always appeared to be the reverse of what might be 

 expected ; thus, in the present case, a pull of only 32 lbs. instead of 37 lbs. 

 was sufficient when lubricant was freely applied to overcome a load of 

 240 lbs., 37 lbs. replacing 55 lbs. and balancing 410 lbs., 49 lbs. replacing 

 70 lbs. and balancing 700 lbs. 



(6) Coil 1} of wrought iron, consisting of eight convolutions of 

 ^ in. X I in. tapering to ^ in. X |- in. required only a force of 3 lbs. to make 

 it lift the lever ; but having once done this, the grip upon the surface 

 was absolute, and apparently sufficient to raise the maximum pull that 

 could safely be applied, which was 4,000 lbs. The only effect of lubrica- 

 tion was to make the gripping action more sudden and violent. 



(c) Coil c of steel, consisting of eight convolutions tapering from | in. 

 square to | in. square section, gave the following uniform series of 

 results : — 



F . . 4i 5 51 6 61 7 7^ 8 Si 9 9i 10 



R . .140 220 SOU 380 460 540 620 700 780 860 940 1020 



F and R having the same values as before. 



When the lubricant was used, the results, though not quite so uniform, 

 did not differ greatly from the above, but were again slightly less, show- 

 ing that lubrication makes the action more effective. 



(tZ) Coil d, consisting of ten convolutions of | in. square wrought 

 iron, tapering to -f\ in. round steel, behaved exactly in the same way as 

 coil h, but the effect was more marked, and 1 lb. on the tail resulted in 

 an absolute and complete grip of the whole coil. 



Influence of velocity. — The effect of change of velocity of the moving 

 surfaces was very surprising, the general result of decrease of velocity 

 being to enable a small force on the tail to put the coil into operation. 

 This was not merely the case when the coil was revolving and the result 

 of centrifugal force acting to keep the coil from contact, but was just as 

 marked when the shaft or sleeve rotated and the coil remained at rest. 

 Again, the chief effect of change of velocity seemed to be felt with coils 

 similar to a and b on the diagram, in which the broad surfaces seem to 

 hold the lubricant more than the narrow ones. In experimenting upon 

 coil c it was found that with a surface velocity of 15 ft. a minute 3^ lbs. at 

 1888. N N 



