66 C. M. Wetherill on Itacolumite. 
on the nails. es colleague and Prof. of Mathematics in the 
University, Col. E. W. Mor rgan, very kindly undertook ne : 
discussion of this curve with the following results. The curv 
was traced upon paper with a pencil under the stone restitil q 
upon two points of support at its extremities and the ordinates — 
were ee for comparison with the ordinates of different 
curv 
Prof Morgan at first compared with the circle, as the curvature : 
was so slight, but it departed sensibly from that curve in su 
a manner as a true catenary would do. He then a it : 
with a loaded catenary (which is a parabola) with similar res 
The beam loaded with equidistant equal weights (ellipse) 4 was 
next essayed but the curve did not agree with it. Nor did it 
coincide with the true catenary. It was nearer to the loaded — 
catenary than to any known curve. The result therefore of the | 
calculation indicated that the curve of the stone lay between — 
the catenary and the ellipse. 
Prof. Morgan next prepared an experiment with asmall beam ~ 
of white pine loaded with equal equidistant weights for the 
ellipse, with a fine and flexible gold chain for the catenary, and — 
with the stone cylinder. These were placed upon the same sup- — 
ports (185 mm. apart) and so arranged that the greatest deflec — 
tion for each curve should be to the same point. 
pon comparing the chain with the loaded beam it was evi a 
dent that the latter curve was a true ellipse, and consequently 
rison would be attended with no uncertainty due toa © | 
ps a defect in the slip of pine. The chain and stone were — 
compared next, and then the stone and beam. Although the — 
experi as not of a character to afford very delicate meas 
urements, it shiswed with great clearness. that the curve of the 
stone is very nearly a catenary, and that it lies between this — 
curve nei ee oni 
Ta a seed Soon the fore ean compressive a 
tensile, as we scraed the aiyper or the lower ponions of the 
m 
the form of the ellipse. The itacolumite cylinder takes the cate- 
: reason of its ball and socket etjoints. But the rod has 
a definite etched and a series of joints el with its axis, 
this respect unlike a true chain. Furthermore, as the micro- 
10! oe : Secesennt 
uenitly, wider thigs Sone ben by its pivsipist betes 
eee 
aa 
