TAYLOR'S EXPERIMENTS 15 



their higher speeds and greater width, subjected to a lighter 

 load than the cone belts. To this effect is due undoubtedly 

 their superiority over the cone belts. If the cone belts had 

 run under as low tensions as the shifting belts, they would 

 probably have run an equally long time without requiring 

 retightening. 



The cone belts would fall from the initial tension of 71 Ib. 

 per inch of width to 33 Ib. in a period of two months, while 

 the shifting belts ran for twenty-two months before they fell 

 from the tension of 71 Ib. to a tension of 21 Ib. Further- 

 more, shifting belts with a light load stretched 0.81 of i per 

 cent before they required retightening, while the cone belts 

 under twice as great a load required retightening after they 

 had stretched but 0.47 of i per cent. Summarizing the ex- 

 periments, Mr. Taylor stated that the total life of belts, the 

 cost of maintenance and repairs and of the interruptions to 

 manufacture caused by belts are primarily dependent upon 

 the total load to which they are subjected. The other fac- 

 tors having the greatest influence on the durability of the 

 belt are: Whether or not the belts are spliced or fastened 

 with lacing or hooks; whether they are properly greased and 

 kept clean; and the speed at which they run. 



Average Load on the Belts. The average total load on 

 the belting in the experiments was 54 Ib. per inch of width 

 with shifting belts and in Ib. per inch of width with cone 

 belts. That these figures represented radical departures 

 from the ordinary practice of the day is evident from a com- 

 parison of the figures given by authors and writers of that 

 time. The safe load on belts per square inch of section, 

 according to different writers, ranged from 290 to 500 Ib. 

 per square inch. These figures are based, as a rule, upon the 

 tensile strength of the leather, an arbitrary percentage of 

 this strength being taken as the safe load as suited the whim 

 of the writer. 



The 54 Ib. per inch of width of double belts which repre- 



