gradually acquired, as it cooled, an increase of weight, fo that, at the end of twenty-two hours, it weighed fix penny weights, feventeen grains more, than it did when firft committed to the balance. This phænomenon, which, by fome has been adduced to prove, that heat is the principle of levity in bodies, Mr. Whitehurst has endeavoured to explain, by fuppofing, that the air, above the fcale being rarefied by the heated iron, the cold air below rushed up, and, ftriking against the bottom of the scale, not only prevented its defcent, but even buoyed it up. Something may perhaps be attributed to this caufe. But would not the circumambient air beneath the fcale be nearly as much rarefied as that above? And is it not probable, that the fuppofed force of this current of air, would be, in great meafure, counteracted by the greater tendency a body has to defcend in a rarefied, than in a denfe medium? Is it not probable, likewife, that the end of the beam, to which the heated iron was appended, would by the fame heat which rarefied the air, be more expanded, and lengthened, owing to its nearer approximation to the fource from which the heat flowed, than the more diftant end of the beam? I would likewise observe, that in the experiment of M. Buffon above quoted, and in one made by Dr. Roebuck on a smaller scale, the mafs, owing perhaps to the

joint action of the above causes, weighed more when hot, than when cold.

Having thus endeavoured to fhew, the infufficiency of the explanation given by Mr. Whitehurft, I will venture, with the greatest diffidence, to propose the following query. May not the increase of weight, acquired by heated iron, and copper, during cooling, be afcribed to the calcination, and confequent abforption of air, continuing to proceed after the removal of the mass of metal from the fire, the abforption of air in particular, in the first stages of the cooling, perhaps, with increased rapidity? In fupport of this conjecture, the following facts may be adduced. Firft; That fome metals, particularly copper*, are found to calcine more rapidly, in a moderate degree of heat, than in one more intenfe. Secondly; That, the calces of fome metals, as that of lead, have been obferved to increase in weight by long expofure to the air, and that they now afford by proper treatment, more air, than could have been obtained from them, previous to fuch expofure. Thirdly; We fhall find by examining Dr. Roebuck's account of his experiments, that the weight continued to increase long after the cause affigned by Mr. Whitehurst must have ceafed to act. The cylinder, which was repeatedly

Macquer's Inft. of Chemistry, vol. I.

weighed

weighed at intervals, when it had been in the scale fix hours, and had then loft fo much of its heat, as to be only blood warm, was found to be acquiring weight in the proportion of feven grains in the fpace of an hour*. But, when weighed the day following, at the expiration of twenty-four hours after the commencement of the experiment, it had acquired a still further addition of two pennyweights and seventeen grains, which, according to the above progreffion, it would have required at leaft nine hours and a half, nay, moft probably, even a longer time, to accomplish. If to thefe nine hours and a half, we add the preceding fix, we obtain fifteen hours and a half; a period long before the expiration of which, the mass of iron must have taken the temperature of the furrounding bodies, fince the first fix of these were fufficient to reduce it, from the welding point, down to blood heat.

I will not trefpafs longer on the time of the Society,, but will conclude by obferving, that metals, which are the only bodies hitherto employed to determine this point, are certainly, from the changes they undergo by the action

During the two firft hours of its expofure in the fcale, the increase of weight had proceeded with much more rapidity; in the third hour it proceeded lefs quickly, and continued to diminish gradually in celerity to the expiration of the fixth.

VOL. III.

N

of

of heat, very ill adapted to the purpofe; and that to arrive at any degree of certainty, it will perhaps be neceffary to weigh the body in vacuo, or at least in a veffel fo confined, as that any current of air through it fhall be prevented; and that the beam of the fcales fhall be formed of materials lefs liable to expanfion by heat, than metals in general are.

REMARKS on the FLOATING of CORK BALLS in WATER; by Mr. BANKS, Lecturer in Natural Philofophy. Communicated by the Rev. THOMAS BARNES, D. D. Fellow of the American Philofophical Society, bolden at Philadelphia.

PER

READ DECEMBER 6, 1786.

ERHAPS no fimple appearance feems fo generally misunderstood as that of Cork Balls, &c. on Water; and as one falfe principle, adopted, is often productive of more, a few obfervations on fo fimple a phenomenon, perhaps, may not be thought below the notice of the learned, although perfectly acquainted with the true cause. That clean bubbles of glass or pieces of cork, left to fwim in water, contained in clean veffels of glafs or china, and at the

distance

distance of about one inch from the fide, will approach that fide is certain; and it appears that this and fimilar experiments have frequently been made by philofophers, to prove the attraction of these bodies. To this I cannot affent, for different reafons; which I presume are supported by the following experiments.

The experiments are made in a veffel of glass or earthen ware five or fix inches in diameter..

EXPERIMENT I. If a clean cork be wet and placed about one inch from the fide of the containing veffel, it will approach the fide with an accelerated motion.

EXPERIMENT II. If two corks be placed about an inch from each other, and at a fufficient distance from the fide, they approach each other in the fame manner.

EXPERIMENT III. Pour water into the veffel till it is rather higher than the brim; place the cork close by the fide, and it recedes with a retarded motion.

EXPERIMENT IV. Sink a piece of metal in the center of the veffel, fo that the top thereof may be above the surface of the water, the cork, placed at a proper diftance, will approach it.

EXPERIMENT V. Raife the water until the metal is covered, and the cork will remain at reft at any distance from the metal.