schlink: hysteresis of indicating instruments 43 



width of the Hnes of the diagram. It is beUeved that a repro- 

 duction of the cycles even more precise would have resulted, 

 had more delicate means of releasing the load upon the pan, 

 and more rigid supports for the apparatus, been at hand. 



During the experiment, the zero point remained surprisingly 

 constant except that at the completion of the second run, it 

 dropped sharply by three-tenths of a graduation, after which 

 being readjusted to its initial value it remained stable except 

 for the drift before mentioned, throughout a great number of 

 subsequent observations. This displacement of the zero was 

 undoubtedly due to an accidental irreversible slip of some 

 loosely secured part of the exterior frame of the scale, which 

 was notably deficient in rigidity, or of the dial. 



2. At the critical region BC and other regions of beginning 

 backward movement of the mechanism, the existence of a marked 

 passivity due to imperfectly compensated backlash, increased 

 the dispersion of successive observations considerably. 



3. The curves corresponding to reversal of the change of load 

 before the extreme range of indication under investigation has 

 been reached, are rapidly asymptotic severally to the upper 

 and lower branches of the major loop, respectively. 



4. The major loop tends to a decidedly skew lenticular form, 

 roughly triangular, a result which is the direct consequence of 

 the force system obtaining in this type of instrument, in that 

 the frictional resistances tend to increase proportionally to the 

 load, when the reactions in the stabihzing check are large. Other 

 indicating instruments are more likely to give a loop of nearly 

 symmetrically lenticular form. 



GENERAL CONCIvUSIONS 



The general conclusions educible in the main from the present 

 data, though fully supported collaterally by a considerable 

 mass of experimental data necessarily omitted here, and by 

 analogy, are the following: 



I. The hysteresis loops obtained in the specified cyclic caU- 

 bration of a (nonintegrating) mechanical measuring instrument 

 nearly free from transient after-effects at rates of operation ob- 



