106 G. H. KNIBBS. 
(10) 
EE 2 
soe 24 
L being the distance of the scale. This examination may be made 
at the same time as that referred to in the previous section by 4 
routine of readings which is sufficiently obvious, from what ‘has 
been stated. 
9. Error due to longitudinal movement of test piece.—lf the 
scale be parallel to the test piece, and if in the application of the 
stress to the piece of material being tested, it be moved longi- 
tudinally only) it is evident that the correction to be applied to 
the reading may be determined from the absolute longitudinal 
movement of the point H Fig. 1. This may readily be measured 
by means of a small piece of millimetre scale, attached to the 
specimen, with a suitable pointer not subject to the motion of the 
specimen: this pointer may of course be the cross wires of 4 
telescope. Let the longitudinal movement be denoted by s and 
reckoned positive in the direction marked by the arrow in Fig. 1, 
viz., in the direction HF. The the effect will be to increase the 
reading on the scale—since the distance therefrom to the point 
. where the sight line meets the mirror is increased— by the amount 
stan w tan 2. Hence the correction y to the reading will be 
+ ‘ete.)......(1ay 
4 
te wel 8 
Li 40 
The lower sign is to be taken when the mirror apparatus is 0D 
the opposite side of the test piece to the telescope. For 10 mm. 
movement the values of this correction are the following :— 
2s 8 es 
Taste ITI. 
Corrections for longitudinal movement of test-piece: J = 1135. 
Scale-reading in mm. 100 200 300 400 500 
Correction 10 mm. shift + ‘04mm. ‘15 ‘34 ‘60 ‘8 
The results indicate how little effect the absolute movement of 
the test-piece has, for the position of the instrument illustrated by 
a 
1 The final formula depends on the relation tan = 4 tan 20—§ tan 
2 + etc. 
