Ball 
—-Notes on Aj^plied Mechanics. 
491 
hygrometer. I shall describe one of them, as it will give a clear idea 
of some of the preceding statements. 
The hygrometer having been exposed to the exsiccating discs in the 
drying apparatus for two days, the index stood at 2° — we may say at 
zero. Being placed in its lantern, the index began to move ; it was 
raining at the time. In two hours the index had moved round twice, 
and had deposited two spherules : it then began to move more and moro 
slowly, stopping occasionally and moving forward again, and sometimes 
retrograding, but still it progressed until it reached 49°, where it rested 
for upwards of three hours. Calling the starting point zero and the 
resting point 50°, the hygrometric state of the atmosphere, including 
the two rounds, was 250° or 0'25. The rain had ceased : the index 
began to retrograde, and advanced and receded alternately, but next 
morning it was found exactly at zero. Thus having advanced 250° and 
retrograded 150°, the gut line must have retained the aqueous re- 
presentative of 100°. 
To prove the correctness of this inference, it should happen that, by 
subjecting the hygrometer to the exsiccating discs, the index should 
retrograde 50°. On making the trial, I was pleased to find that this 
happened exactly as it ought : for, after exposure of the hygrometer to 
the exsiccating discs during a night, the index, next morning, stood 
exactly at zero, and further exposure, during several hours, did not 
make any change. 
If, without this second exposure to the exsiccating discs, I had at 
some subsequent period placed the hygrometer on duty, and found the 
index to point, suppose to 80°, it is obvious what a mistake it would 
be to conclude that this degree indicated the atmospheric state, the real 
indication being 180°. This source of error must be always attended 
to, when the hygrometric state of the atmosphere is to be ascertained. 
XL. — ]S"oTEs ON- Applied Mechan"ics, III., lY. By Egbert Staavell 
Ball, LL. D,. 
{Continued from Froceedings, vol. i., ser. 2, page 245.) 
[Read April 14, 1873.] 
III. — Of the Theoey oe Long Pillaes. 
Itt the Educational Times," JN'ovember, 1872, I proved the fol- 
lowing approximate formula (previously proposed in question 3809) 
connecting the load on the pillar with the deflection of the centre for 
the case where the curve has no point of inflection — 
\ 8Z2 
W being the load on the pillar. 
L the length. 
R. I. A. PROC. — VOL. I. SER. II., SCIENCE. 3 S 
