NATURE [ Sepd. 7, 1871 
we 
“SY 
Oo 

and to one another, by string-hinges. The result did not 
SE ere ie CL OER AES, | differ from that obtained with two joints. This surprised 
me, for I had expected that each joint in taking its share 
le naturally occurred to try what would be the effect of of the swing would somchow assert its claim to its own 
introducing a third joint in the suspension of the pen- | proper period distinct from the periods of the other two. 
dulum, and altering the angles between the different joints. | But on close examination, attaching long slender splints 
First I tried three joints at angles of Go”, by means of two | of wood to the cross-bar, the two intermediaries, and the 
intermediate pieces attached to the rod, to the cross-bar, rod, and bringing their depending ends near together, by 


Fic. 11.—Proportion 3 : 4.—Cusped type. Fic. 12.—Proportion 3 : 4.—Looped type. 
pairs, for comparison of relative motions, I found that the | had been only two joints at right angles. I also tried 
two upper joints alone were able to make good their title | three joints at angles of go° and 30”, and two joints at an 
to independent periods of swing: the third joint acted | angle of 60°. They all gave similar results. It is clear 
merely as subsidiary to the second, merely serving to | that the planes of slow and quick vibration are determined 
facilitate the bend at the second joint, and always syn- | solely by the position of the topmost joint, the former 
chronous with that bend, never able to establish a proper | being that in which the pendulum can swing by the top- 
period of its own. All three joints conspired to allow | most joint alone, and the latter that in which the topmost 
undeviating vibration in a plane at right angles to that | joint takes no share at all, or the least possible share, at 
allowed by the topmost joint alone, just as though there right angles to the first. However the joints may be 
