260 ‘A. BOYD. 
at the outer ends of the arms in all the wheels usually 
indicated that the rim broke first midway between the 
arms, and that then the two parts of the rim flew outward 7 
and broke off at the arms, thus demonstrating the large 
effect of transverse loading on the rim due to its centri- 
fugal force. | 
In the following investigation actual measurements of 
the deflections of the rims were taken as the speed was 
increased, so that the behaviour of any wheel at any par- 
ticular stage of its expansion can be observed. The experi- 
mental flywheel to be tested was arranged horizontally, 
and the deflections of points round the rim measured at 
various speeds by means of an optical method, which is 
described' by Mr. 8S. H. Barraclough, B.£.,M.M.E., Assoc. M. Inst.C.E. 
The principle of the method is that of the Martens’ mirror 
apparatus. A steel prism P with mirror attached (Fig. 1) 
stands on the top of the flywheel spindle. On this prism 
rests a distance rod which is in contact with the rim at 
diametrically opposite points. One end of the rod F is 
fixed to the rim, the other end M merely rests on it. As 
the rim expands, the distance rod is pulled with it, causing 
rotation of the mirror. A beam of light given by a lamp 
behind a knife edge K distant about one metre above the 
mirror is reflected into the field of a telescope fixed on a 
stand beside the wheel. The general arrangement is shewn 
clearly in the photograph (Fig. 2). The image of the knife 
edge is flashed into the field once each revolution, so that 
at about three hundred revolutions per minute, the image 
appears as a continuous black edge in the field, and its 
position can be read on a micrometer scale G, in the eye- 
piece of the telescope. The knife-edge is clamped on a 
glass scale, graduated to half millimetres, on which it can 
be moved. The ie gheeeel = is places at such a distance from 
* Proc. Inst. C.E, Vol. cx., p. 398. 
