CAL

CAL

( 201 )

before its chemical combinations are noticed,fome further obfervations may be added con-cerning the expanfion, which invariably attendsan encreafe of the quantity of. caloric in everyfubftance in nature, in the circumftances abovementioned.

Of Expanfton.

It is by the expanfion which fluids undergowhen heated that the encreafe or diminution inthe quantity of fenfible caloric in any body isufually meafured. This is done by the Ther-mometer . How far this inftrument applieswill be mentioned in the next left ion whentreating of combined caloric.

The expanfion of feveral fluids has beenmeafured with confiderable accuracy, and thepeculiarity that takes place in the expanfionof water at low temperatures has been noticed.This Angular fa£t, firft accurately afeertainedby Mr. De Luc, and afterwards by Dr. Blagden,has been further illuftrated by Mr. Dalton, £who has fhewn that water expands equally atequal degrees above or below its maximum ofdenfity. The experiments were made by fillinga thermometer tube with water, and immerfingit in water of known temperatures, or in dif-ferent freezing mixtures, ufing a good mercurialthermometer as a ftandard. It has been men-tioned that the maximum of denfity is fome-where about 40° to 43 0 , but the difference ofexpanfion for two or three degrees above orbelow 41 0 , is fo fmall as fcarcely to be cog-nizable. Mr. Dalton eftimates the maximumat 42-5-5 and finds that water inclofed in athermometer bulb, may be cooled not only toabout 21° (as Dr. Blagden had difeovered) buteven as low as 5 0 or 6 ° without freezing. Whenactually cooling below 42-j 0 , the water rifes onthe fcale, and exactly in correfpondence withequal diftances above 42-j, fo that for examplewhen the liquor Hands at 63° of the fcale, it isimpoffible to tell by mere infpe£tion, whetherit is actually 63° (20° above the maximum ofdenfity) or 22-3- 0 (20° below the maximum).

A great difference takes place in the expanfionof different liquids by equal quantities of heat.This is not apparently connected with theirfpecific gravity, but as Dr. Thompfon £ fuggeftsit feems to depend on the degree at which theliquid boils, thofe liquids having the greateftoxpanfion whofe boiling points are the loweft.Thus mercury expands with a given temperaturemuch lefs than water, whofe boiling point ismuch lower; and water, again, expands lefs

r Manch. Mein, vol. t. pt. a.

than alcohol, which boils- at a lower tempera-ture. So that as a general rule it may be laiddown that the expanlibiiity of liquids is in fomeproportion inverfely as.their boiling points.

Another important circumftance concerningthe expanfion of. liquids is that they all expandmore with equal additions of heat at the higherthan the lower, temperatures. This variation isnot very uniform, nor is the law of the encreafein all cafes very obvious. In water, however,the encreafe at the higher temperatures nearlycorrefponds with the fquares of the naturalnumbers. Thus, if the original bulk of waterbe aflumed as 10000, die expanfion at 82-)®may be taken, at the fquare of 6 or (36), andthat of every ten degrees above 82.-5- will befucceffively nearly as follows: at 92 4 °;^= 7 2 {49):at 1024—8 - (64): at 112-4°—9" (81); and foon, according to the fquares of die naturalnumbers. Owing to the peculiarity in theexpanfion of water, juft mentioned, the famerate of expanfion will alfo apply to the defcend-<ing fcale of heat, at like diftances from 42!°,as long as it can be preferved unfrozen. M. DeLuc has clearly pointed out how this encreafingexpanfion of fluids in the higher temperature*throws an unavoidable error on the commonconftrudfion of the thermometer, the fcale ofwhich always fuppofes equal expanfion, and ofwhich the degrees divide equal fpaces in thetube. Thus if, as is commonly done, two dis-tant points of the fcale are firft marked (that offreezing and boiling water for example) and thefcale is then divided into 18o° equal fpaces, it isobvious that 90° degrees from die freezing point,or half this fpace, will not exactly indicate halfthe entire expanfion of the mercury from thefetwo points, but the lower half of the fcale willbe too large, and the upper too fmall. The realmedium of temperature would be at 118.85° °fthe common fcale, or 86.85 degrees above 32°,the freezing point.

By experiments on other fluids, M. De Luefound this difference to be greater, as will befeen by the following table. The fcale ofReaumur’s thermometer was ufed, which dividesthe interval between freezing and boiling waterinto 80 degrees. The mean number is there-fore, by the fcale, 40°, but the real mean ofexpanfion expreffed in the firft column is, in allthe fubftances examined, lower than this point.The fecond and third column of numbers ex-preffes the proportion between the higher andlower parts of the fcale of expanfion.

* Slemsnts af Ch.smiftry.

2 C

vol. 1 .