84 Mr. J. J. Guest on the Strength of 



approximately equal additional weights placed upon a diago- 

 nally opposite pair of weight-carriers until they descended. 

 The equal additional weights were placed upon the other pair 

 of weight-carriers until they in turn descended. The latter 

 pair of additional weights (shot) were then placed together 

 and weighed. One quarter of this amount for one of the 

 weights on the carriers represents the equivalent of the friction 

 of the system under the load P. Now in this test four 

 pulleys and the two shackle-bearings rotate, while in the 

 actual tests three pulleys and one shackle-bearing are rotated, 

 thus the previous result should be halved for application to tho 

 tests. The two weights on the carriers are, however, com- 

 bined into one hung at the centre of the cross-lever, so that 

 one quarter of the weighed amount should be taken for the 

 weight (W) at the centre of the cross-lever. The results of 

 these friction tests are given in Table I. ; the friction given 

 is in all cases too high, as, the masses moved being large, a 

 small moving force produced a small acceleration which 

 required considerable time to make its effect manifest. The 

 corrections required for W are so small that they have been 

 neglected in the calculations of the stresses. 



Table I. 

 Friction of Torsion Apparatus. 



It was found that the friction of the pulleys was so small as to be negligible 

 in comparison with the friction of the pull-bearings, the value of W making 

 little difference. 



Mean values of SW from several experiments are given : — 



P 1000 2000 3000 4000 



SW 0'319 0-576 0-909 1-058 



Hence S W = 0-285 lb. per 1000 lb. value of P. 

 And SW = - 075 lb. in all cases for the wheels. 

 Hence Friction only amounts to 1£ lb. for a load of 5000 lbs. 



29. The Application of the Internal Pressure. — When 

 tests involving the application of internal pressure were to be 

 made, after a small tensional load had been placed on the 

 specimen so as to line it up, the lower end was clamped 

 firmly by the other screws, opposite to the first set, in the 

 bracket (fig. 4, A). The tube brazed into the lower holder 

 was then clamped at two points to the lower cross-bar, and 

 finally the tube was coupled to the pressure system. The 

 clamps were intended to prevent the production of a bending- 

 moment on the specimen ; but in spite of these precautions, 

 and although the pipe conveying the pressure was long and 



