[Bovey ] A NEW EXTENSOMETER 5 
vertically by means of a small windlass at the base. The pillar on which 
this head is moved up or down to the required position is of triangular 
section, which prevents any axial movement. The scale is fixed on the 
head carrying the observing telescope and remains in a constant position 
relatively to the telescope. A preliminary calibration of the whole of 
the arrangement is made to determine the distance of the scale from 
the axis of rotation of the mirror. Thus the exact value of the move- 
ment occurring in any one extensometer on the face of the beam, may 
be read off, by bringing the telescope into line with that particular 
mirror. 
It is not absolutely necessary to have more than one measuring 
device, which may be shifted from position to position, and the beam 
subjected to repeated loads, the resulting values obtained in each 
position being the same as though a number of the instruments were 
used at one time. In order to save time, however, by diminishing the 
number of loadings, as many extensometers might be placed on the 
beam as the width of the beam will allow. 
A number of experiments have been made with extensometers of 
this description and have shown them to be of exceeding delicacy. Small 
loads of even a few pounds are immediately indicated by the movement 
of the cross hairs over the scale. 
(Of course devices of much cruder design and construction are 
quite sufficient when larger deformations than those within the elastic 
limit are produced, and such devices are now being constructed and 
used in the McGill laboratory.) 
The following experiments will, to some extent, indicate the 
nature of the apparatus and the results seem to justify certain inter- 
esting and important inferences. 
In these experiments, which are all within the elastic limit, the 
distance between the extensometer points, i.e., the length under obser- 
vation, is in each case 8 inches, and the beam supports are 60 inches 
centre to centre. The loading is of two kinds: 
(1) The beam is loaded in the centre (Fig. D); 
(2) The beam is equally loaded at two points equidistant from the 
centre and the loads are such that the maximum bending 
moment is the same as the maximum bending moment when 
the load is at the centre (Fig. E). 
The deflection was carefully measured in each case, but was so 
small as to have no appreciable effect upon the 8-in. distance between 
the extensometer points. 
