PROPERTIES AND THE ELASTIC EXTENSION OF METALS. 
133 
The procession is formed by setting the plates in due sequence and placing the 
origin of the record on each plate at the point on the extension scale corresponding 
to the permanent set measured from the test piece itself. For example, direct 
measurement of the gauge length after taking plate D shows that the permanent set 
is 0'137 inch. The origin of the record on plate E is then located at 0’137 inch on 
the scale. 
The procession of loops seen in fig. 19 is reproduced on a small scale in order to 
present to the eye the complete record on a reasonably sized sheet. 
Selected loops from the procession are shown full size in figs. 20 to 23. The elastic 
line and the first and second loops are seen in fig. 20. The area of the first loop 
represents an energy loss of 0'42 ft. lbs., and of the second loop 1T5 ft. lbs. The 
corresponding permanent sets are 0'002 inch and 0'007 inch. 
The areas through the sequence of plates A, B, C, D, E, F increase gradually. 
The last loop of this sequence is seen in fig. 21 and it represents an energy loss 
of 7’42 ft. lbs. The time occupied in taking these six plates was 23 minutes. 
After taking plate F the experiment was stopped. The test piece was removed 
from the machine and laid aside. After six days’ rest it was put back into the 
machine and looping continued. 
The first line after the rest and the first loop are seen in fig. 22. The first line 
shows no elastic recovery and the six days’ rest has had no perceptible influence on 
the loop area, which represents 7'78 ft. lbs. The area is what it would have been 
if thei’e had been no interruption of the experiment for a period of rest. The 
sequence of plates G to Y was taken in 1 hour 20 minutes. The record on the last 
plate is seen full size in fig. 23. Stretching was stopped because local contraction 
had set in. 
The areas of the loops in the sequence are plotted against permanent set in 
curve 1, Sheet 1. Each small circle denotes a loop, and the letter written against 
some of them identifies the first loop on the plate corresponding with the letter. 
It will be noticed on Sheet 1 that the curves joining the loop areas on any one plate 
do not merge into one another to form a continuous curve. The time interval 
required to change the plate and to resume loading seems to be occupied by the 
material in some inner process which tends to slightly reduce the area of the next 
loop taken. But whatever the inner process may be, it practically exhausts itself 
in a few moments and produces only slight effect on the next loop area. No further 
change takes place after a rest of six days, and the inner process, whatever it is, has 
no influence in restoring the material to a state of perfect elasticity after the overstrain. 
A curve sketched through the group of loops on each plate is continuous and 
clearly shows that the area of the loops tends to a maximum. The maximum value 
in this experiment represents an energy loss of 11'48 ft. lbs. per loop. This loss 
corresponds to 7 ‘ 51 ft. lbs. per loop per cubic inch of material in the primitive gauge 
length. 
VOL. ccxxi.— -A. 
u 
