MECHANICAL TESTING OF IRON AND STEEL. 219 



Where, however, tests are made of bars of iron and 

 steel, and where there is no question of this difficulty, it 

 would be a great convenience in making comparisons if a 

 standard value for the above ratio were adhered to. The 

 German system has been to make the cylindrical test bars 

 always 10 diameters long; in recording the extensions they 

 also give the extension on only one half the original length, 

 taking the point of fracture as the central point of this 

 length. For flat specimens the ratio of width to length is 

 1 to 5, and the natural thickness of the plate or bar is 

 adhered to up to about three-quarters of an inch, above this 

 round specimens of standard dimensions are turned out of 

 the material. 



The Admiralty length for steel plate specimens is 8 

 inches, and according to the German rule this would be 

 accompanied by a width of 1 '6 inches, a reasonable and 

 practical value. 



There is no necessity to adhere slavishly to such rules, 

 but it does seem absurd to put aside the obvious advan- 

 tages to be obtained by working on these lines whenever it 

 is possible. 



Reduction of Area at Fracture. Many experimenters 

 and engineers lay great stress on the value of this 

 quantity as a guide in estimating the ductility of a 

 material, and undoubtedly if taken along with the condition 

 of the fractured surface it is of much value. But as was 

 pointed out by Bauschinger, and shown very clearly by his 

 museum of specimens, a comparatively minute flaw or hard 

 spot at the point of fracture would frequently diminish the 

 reduction of area to an extent apparently far in excess of 

 the proportion of the flaw to the rest of the fracture surface. 

 Similarly with flat specimens, imperfectly formed ends, 

 though they influence but little the maximum load carried, 

 or the ultimate extension, often seriously effect the reduc- 

 tion of area, sometimes making it apparently almost zero. 



It cannot, therefore, be contended that this factor is such a 

 reliable guide as the ultimate extension in judging the quality 

 of a material. It is well, however, to dispel a statement 

 often put forward against the value of the figures recorded 



15 



