250 DR HOPE OX THE MAXIMUM DENSITY OF SEA-WATER. 



warmer portions descend, and thus the remora in the cooling of the fluid at the 

 bottom is produced. 



Had Mr Erman employed a vessel of larger capacity, and such a quantity 

 of water as must have afforded time for precise observation, and had placed the 

 ball of the thermometer near the top instead of the bottom of the vessel, he would 

 not have seen any indications of remora. 



The retai-dation, therefore, in the cooling of water, which excited the surprise 

 of Mr Erman, depends upon the change in the direction of the currents conse- 

 quent upon the anomalous law and the position of the thermometer. 



I remain satisfied, that there is no method of ascertaining irregular changes 

 in the density of aqueous fluids so fit as the one I employed in 1804, as it is en- 

 tirely independent of every change in the density or capacity of the instruments 

 themselves. It consists in ascertaining, by means of thermometers placed near 

 the top and bottom of a column of fluid, the direction which the curi-ents of heat- 

 ed or cooled fluid assume : every one admitting it to be an incontrovertible truth, 

 that the contracted denser fluid Avill fall to the bottom, and the expanded rarer 

 fluid will rise to the surface. 



By pursuing this method, I conceived, in 1804, that I put beyond all doubt 

 or cavil the reality of the existence of the singular anomaly in the constitution of 

 water ; I therefore determined to have recourse to the same method in the pre- 

 sent inquiry : I did not, however, deem it necessary to perform more than the 

 two most striking of the series of experiments. I used an apparatus similar to 

 what I had formerly employed. It consists of a glass jar twenty-one inches high 

 and four in diameter, to which there is adjusted in the middle of its height a per- 

 forated basin of tinned iron, two inches in depth and ten in diameter, from the 

 upper edge of which there issues a small spout. In this jar two thermometers 

 are suspended, one with its ball near the bottom and the other near the top 

 In ray memoir of 1804, it is recorded that, when this vessel was filled with water 

 at temperatm-e 32°, and the encircling pan was kept full of water at a tempera- 

 ture between 70° and 80°, the thermometer at the bottom gained nearly 7° of tem- 

 perature in three-quarters of an hour, while the thermometer at the top was very 

 little affected, plainly indicating that ice-cold water, by being heated, contracts, 

 and becoming more dense falls to the bottom of the apparatus. 



Experiment, No. 7. — I filled the jar with the sea- water of the Frith of Forth 

 cooled to 29°, which is its congealing point, and furnishing the pan with a con- 

 stant succession of water at 84°, the upper thennometer immediately began to 

 rise, and continued to do so uninterruptedly, and gained 19° before the thermo- 

 meter at the bottom had gained more than 1°. This experiment proves, in a 

 satisfactory manner, that this sea-water cooled to its point of congelation ex- 

 pands upon the application of heat. 



