138 Oft the Contraction which takes place in Mercury 



known, and expose it to dificrent temperatures from 100 

 to 50 degrees, it will contract ; but whilst the fluid con- 

 tracts, the glass which contains it contracts also, and shows 

 only the difference of the contraction between the fluid and 

 glass : — however, as we know that silver by hcatiniir cannot 

 become specifically heavier, or by cooling specifically lighter, 

 in these experiments which are evidently a coniparison of 

 densities, the alcohol mu»t become more dense by depriva- 

 tion of heat. 



The third column of this table contains the temperatures. 



In the 4th column the 1000 gr. of mercury and the glass 

 bucket containing it appear together to have varied at the 

 two extremes in loss of weight 4-3. 



By the 5th column it is shown that the glass bucket alone 

 varied at the two extremes in loss of weight rg. 



By the 6th column it appears that 2*4 only can be stated 

 to belong indispensably to a variation of the respective den- 

 sities of the mercury and alcohol in the changes of tem- 

 perature from 100° to zero. The other variations, J presume, 

 arise from some inaccuracy, such as making the observa- 

 tions when the fluids w^re at different temperatures. 



By the 7th column are shown the different gijiccific gravi- 

 ties of the mercury, by calculation from the loss of weight, 

 supposing the density of the alcohol the same throughout : 

 f;om this it appears that the density of mercury is greatest 

 at the highest temperature, and least at the lowest ; but 

 as this cannot be actually the case, it is obvious that the 

 alcohol increases in density ; and this appears also from the 

 results of the 2d column relating to. silver, at C. 



For the specific gravities or alcohol at various tempera- 

 tures, 1 must refer to the calculations at L, made from the 

 following experiments, and which are contained also '\\\ a 

 general table of results. 



E. I procured the bladder of a rabbit, and washed it well 

 with alcohol. It weighed 7 grains in air; and having an- 

 nealed a wire about 9 inches long, the weight of wliicli was 

 -,«f,lhsof a'grain in air, I found that 'the bladder and wire, 

 when sunk in alcohol to a certain mark on ilie wire, weighed 

 2'5 grains only, at tlic temperature Ab^, 



4 This 



