ba | 
INSTRUMENTS USED IN ENGINEERING LABORATORIES. 54. 
TABLE LV.—continued. 
Mean of all 
| | 
Results by | Mean results} Mean ofall | results on 
Bar Observer each by each results on | bars of same 
observer observer one bar | material and 
| size 
A. B.W, Kennedy . {| Teor, |} 18150. | | 
T. Hudson Beare f ae j 13210 | 
(| 13110 |) Sah nese 
A J.Goodman . 13110 j 13110 || 
296 
D.§8. Capper . | yeaan | 13260 
bad pepiacen lt oa. | ag 
| J. A. Ewing ‘| 13310 y 13285 | 
|B |H.S.HeleShaw . {| fee doe le | 
| | ; |~ 13203 
W.C. Unwin a 13294 ) 13298 | | 
ae» {| ioe Pg | / 
( 13110 } ania 
D.§. Capper . y 13110 py 13110 | 
The foregoing table contains all the values of E calculated from the 
total range of stress to which the bars were subjected in the record sheets 
sent in to the Committee. The total range of stress in each case was 
about the maximum range which could safely be used without risk of 
straining the bar beyond its elastic limit. 
It should be remembered that bars E, F, K, L were all cut from the 
same rolled bar, and of these E and F were approximately of 1:25 inch 
diameter and K L approximately 0°75 inch diameter. Bars A and B 
were cut from the same plate, and were approximately*of the section 
2 x4 inches. 
V. Variation of Individual Results from the Mean of all the Results. 
In the preceding table the means are given of all the results sent in 
for the pairs of bars of the same material and size. It cannot, of course, 
be assumed that these means are the true values of the coefficients of 
elasticity of the bars. Nevertheless, they furnish convenient reference 
numbers for estimating the probable magnitude of the differences of values 
of E determined by different observers with different instruments. In the 
following table this comparison of each observer's results with the mean 
of all the results is made. The percentage deviation from the mean is, of 
course, not necessarily the percentage error, but it is at least some 
indication of the probable magnitude of the error of the calibration 
constant as distinguished from error due to instrument or observer in 
individual determination of the extension. 
