IN THE HALL OF THE ROYAL SOCIETY. 
571 
It will be observed how very gradually the differences, recorded in the last 
columns of the months, increase; till, in the month of March 1832 , they average 
*551 ; more than half an inch of mercury, indicating a mean depression of the 
water-barometer of more than seven inches. This result is further confirmed 
by a comparison of the monthly mean heights of the two instruments, and by 
observing that in the month of March 1832 , when the differences for each day 
are exhibited, the greatest differences occur with the highest barometer, as 
would happen from the greater compression of included air under such cir- 
cumstances. The regularity of this secondary effect is indeed very remarkable. 
This unfortunate result not being doubtful, I determined to open the boiler 
for the purpose of throwing some light, if possible, upon the cause. Dr. Prout, 
to whose valuable advice I have been greatly indebted in all the previous ar- 
rangements, did me the favour of assisting at this examination. 
Upon removing the cover, we found that a portion of the liquid had by some 
means escaped, as, although the column of water stood considerably below the 
neutral point, the ivory point was not in contact with it. We carefully mea- 
sured its distance, and found it to be 0*3 inch, to which, as the barometer 
stood at 385*94 inches, must be added 0*05 inch for the difference from the 
neutral point ; and the amount 0*35 inch will be the quantity of the fluid 
deficient. 
Upon examining the oil upon the surface, we found that it had undergone 
a very remarkable change. It was nearly covered with large clots of a muci- 
laginous-looking substance, which, in places, reached quite through to the 
water beneath ; so that upon moving them aside the latter was uncovered. 
Upon the top of this, in various parts, were drops of an aqueous fluid, of a 
tenacious consistence, which had a very decided sweet taste, and resembled 
the substance which is formed during the process of saponification, to which 
the name of Glycerine has been given. There was also some carbonaceous 
matter, but not more than might probably be accounted for from deposi- 
tions from the atmosphere. All these matters, with a great portion of the 
remaining oil, were carefully skimmed off, and the water beneath was found 
perfectly bright and transparent ; there were no signs of metallic corrosion in 
any part, and every portion of the boiler, with its cover and brass-ivork, was 
as bright as on the day when they were put together. 
