244 AUTOGRAPHIC DIAGKAMS. 



made to depend upon direct measurements, though in 

 certain cases they may be used as a rough check on the 

 latter. The area of a stress strain diagram is a measure of 

 the mechanical work performed upon the specimen in 

 breaking it, and it is a not unusual way of expressing the 

 quality of a material to calculate the work done on the bar 

 up to the point of rupture per cubic inch of its volume. A 

 convenient method of ascertaining the amount of this 

 work is by measuring the area of the load-strain diagram 

 by means of a planimeter. The same result may, of course, 

 be arrived at by taking a number of isolated observations 

 and plotting the diagram, but this is too slow and laborious 

 in the case of most commercial tests. There is also an 

 approximate method of arriving at the same results, 

 depending upon the maximum load and the final 

 extension. 



For the purposes of scientific investigation a reliable 

 autographic apparatus, such as the one used by Dr. 

 Kennedy, may be made extremely useful. It is possible 

 in such a case to observe minute changes of stress and 

 strain, which are difficult, and, in many cases, impossible, 

 to obtain in any other way. This applies especially to the 

 portion of the stress-strain curve at or near tne yield point, 

 and also that part beyond the maximum load up to the 

 point of fracture. 



It should be pointed out that the dimensions of a 

 diagram are aftected by the magnitude of the time interval 

 allowed to elapse during which the test is carried to a 

 completion. Deformation during the semi-plastic and 

 plastic stages does not take place instantaneously, but the 

 material must be allowed time to take up its set. Conse- 

 quently, a test rapidly carried out will show less permanent 

 strain than one more slowly performed, and it does not 

 follow that because two specimens are exactly similar in 

 every respect that that they will produce precisely similar 

 autographic stress-strain diagrams, although the general 

 characteristics may be the same. 



The use of autographic diagrams for quantitative 

 purposes in commercial testing has been deprecated, and 

 in this connection it has often been pointed out that 

 diagrams of this kind can be tampered with. This is, of 

 course, true, but it is not much more likely that a man 

 who is carrying out a test will tamper with the diagram 

 his machine is producing than that he will intentionally 

 misinterpret direct observations, or that an observer in a 



