Plotting these results on the basis of § 148 for the known positions of centre of pressure for these planes, we have diagram Fig. 157.
The divergence shown in the above determinations is largely due to the temporary and insufficient character of the apparatus employed, and to the fact that for want of suitable accommodation the experiments were conducted out of doors.
It is further possible that the considerations raised in §§ 182, 183, to some extent invalidate the present method. So long as the type of fluid motion in the periptery of the aeroplane is frankly discontinuous the method will in theory give consistent results, but so soon as the live stream touches the upper surface 
Fig. 157. of the plane, as it must do when the angle becomes very small, the area subject to skin-friction will increase in some way as an inverse function of the angle, and the equation will cease to hold good. We may consequently anticipate that when the angle becomes less than some critical value the curve will cease to be of the form plotted in Fig. 112, and the present method will break down. It is principally for this reason that the author has confined his observations to the low velocity portion of the curve; it will be time enough to carry these observations further when better launching and measuring appliances have been developed.
§ 246. Determination of by the Aerodynamic Balance.—In the determination of by the aerodynamic balance, one arm of the beam A. Fig. 153, is furnished with a lead block D. Fig. 158, whose sectional form is an isosceles triangle, the base of which
387
