GREATEST DENSITY OF WATER. J^l 



■s^raVity as water of 32^, and, consequently, that the point of laference that 

 greatest density is at 36o. TsV! """""' 



Dr. Hope was aware of the plausibility, if not of the force, T!ie fourth ex- 

 of this reasoning, and therefore judiciously admits that this P^^^'^'^"^- 

 experiment does not decide the question ; this is supposed to 

 be, done by the 4th experiment, which exhibits a tall jar of 

 water of 40**) placed some inches deep in a pan containing 

 ice and salt. Herewe have water of 40'', and a cold mixture 

 of 00, with a partition of glass betwixt. The medium tem- 

 perature of that part of the glass jar immersed, may be stated 

 at 20", just after the commencement of the experiment. The 

 contiguous water is soon cooled to 32°, whilst the temperature 

 of the mixture remains nearly stationary for a long time ; 

 whence the mean of the glass would soon be l6 or 18°: now 

 it must follow that the water, in contact with the glass, was, 

 Hable to be pooled below the freezing point, provided it was 

 susceptible of such temperature; and all experience shews that 

 water, in such circumstances, not only mai/, but nearly always 

 docs, cool several degrees below freezing, before congelation Rea-joning by 



commences. If then the marginal watef was cooled to 25 ^'^'"^h tht; phe- 

 nomena ma j' be 

 or 28°. whilst the central was 36 or 38°, in so tall a column, explained, with 



a prodigious force of ascent would be produced, sufficient to a great uncer- 



^ o t ' _ tainty as to the 



explain the phenomena, whether we suppose the point of great- most density. 



est density at 40°, at 3()°, or even at 32°. 



The fifth experiment of Dr. Hope is one which more es- Dr. Hope's fifth 

 pecially involves a question of primary importance to the sub- experiment, 

 ject. That is, suppose a liquid of 32" contracts by heat to a 

 certain temperature, and after that expands again so as to 

 equal its former bulk by another equal augmentation of tem- 

 perature ; again, suppose that a tall jar of such hquid at the 

 temperature of 32° was surrounded in the middle by a zone 

 or belt kept constantly at a temperature superior to that at Question hovv 

 which the liquid is of greatest density ; qucere the manner in distWbuVd'*^ 

 which the heat acquired ought to be distributed through the through the li- 

 liquid > ^"'•^• 



In answer to this it may be observed, that reasoning a priori, Particular 

 the effect ought to be different according to the periods of time, ^^^ effect 

 whifch may be divided into three; and the ^rst period will be 

 that in which the marginal liquid descends to the bottom of 

 the vessel till it there becomes of the greatest density ; the 

 S 2 second 



