220 



PROCEEDINGS OF THE AMERICAN ACADEMY 



connecting the points of the star given by the circle. But in the next 

 plate (Fig. 3), whose minor axis is six inches in length, the curves, 

 though situated similarly to those in the last figure, are of a much 

 smaller curvature. This decrease continues as the minor axis de- 

 creases, until, as we see in Fig. 5, the lines have become nearly, if not 

 quite, straight. 



Another point to be noticed is that in Fig. 4 we begin to see that 



Fi g- 2 - Fig. 5. 



the extremities of the curves are nearer together on the side where 

 the clamp is than on the side damped by the finger, and this fact be- 

 comes very noticeable in Fig. 5. 



I also made a series of experiments with the same plates, clamped 

 at one edge, damped in the centre, and bowed at a point about 70° 

 from that extremity of the minor axis which was not clamped. Here, 

 too, I obtained a series of curves similar to each other (Figs. 6 to 9), 



