STRUCTURAL RESULTS OF FAULTING. 



1B7 



tion. The deviation is exceedingly sliglit, and tlie experimental curve 

 stands almost as well as the other the very delicate constructive test of" the 

 equality of subtangents.^ 



Variations of the experiment. — The sllps I liavc emploved are of a nearly uncal- 

 endered paper. If i'l 



iSBBBSBBBSHBSSSSSSBBwSBSSBBSBnSg 

 iBanBaaaiBaaflaaBBBBBiBBBBBaaHaaaB 



IBBaBBBBBBBaBBflaaaBBBaBBaBBBaHBaBr 

 IBBBBIBBBBBBBBBBiaBBBBIBBBBBBIHBni 



laaaaBiaBBBBBBBaaaiBBBiBBBBaaBBaBBL 



lafianHHnHBBHHHBHHHI! 



jaBKSWBIBBHBBBBIHHIBHgBIBI 



IBBBIBflBBBBIBBBBBBBflBBBBBBBBBBB|Ba 



iBBaaaBBBBasflaaBBBBflaaaaflBBBBBBiBI 



SBBBIBBBBBBBKaBBnBBBIBBBBBBBBaaBI 

 IHBBaBBBBBiBaBBBSaBaBBBBBaaaaBBBBa 

 IBSBIBBBBBBBaBBBiiBSBBBBBiBBBBBIBI 



^BBBaiBBflBBBflBBBBBBBBBBBL_____.^, 



BBBnSBBBlBHflBBBBBIBBilBBIBBIBB 

 BBBBBBBKSBBBBBBBaBBBaiBiBBBRBBaBB 



jBBBBBBBBBIiSBIBBBBBBBaBBBIBBIBflaBB 



laaaBaaBBBBaBasBBaBBaBaBBBBBBBBaaBB 



nBBBBBBBBBBBnBBaBflBBIiBBflBBIBB!!! 



one of them a highly 

 glazed slip is substituted 

 a comi)aratively largo 

 relative motion takes 

 place on its surfaces, but 

 the only visible effect 

 which the introduction of 

 such a slip produces on 

 the locus of the others is 

 a dislocation at the pt)int 

 where it is inserted. 



There is in fact no evi- Fig. 4.— Cakiilatcd an.l ubseivud curves. 



dence that the woi-k done on any contact is altered by the introduction of 

 a contact offering a smaller frictional resistance. 



If the ends of the slips at the beginning of the experiment occupy an 

 inclined instead of a vertical plane, the result is a logarithmic curve referred 

 to axes inclined at the same angle In plotting it is well to reject the 

 upper three or four slips, because these are principally affected by inequal- 

 ities in the application of pressure and draught. 



By employing a system of from three to ten slips of heavy writing 

 paper, using a thick pad of blotting paper for a support, and applying the 



'Such an experiment forms a check upon the tlicory, hut does not furnish ahsolute proof of it, 

 hccause arcs of other curves, known or unknown, might be coustructed which would agree very closely 

 with the experimental result. Among faoiiliar curves, that presenting the greatest general similarity 

 to the logarithmic curve is the hyperbola referred to in its asymidotes, and a hyperbolic arc very closely 

 agreeing with the experimental curve can be calculated. But the experiment gives the position of the 

 asymptote which for the nearest hyperbolic arc would occupy a distinctly diflerent position, and the 

 supposition that the curve was hyperbolic would also lead to seemingly untenable hypotheses as to the 

 communication of energy. All that can be claimed, however, strictly speaking, is that the theory ac- 

 counts for the facts within the limits of the errors of observation, and that no other eciually plausible 

 explanation of the facts has suggested itself to me. 



