at low Temperatures, 135 



he had added experiment to his calculations, he would have 

 been convinced that the principles on which I proceeded are 

 just, and that one source of difference in our opinions 

 arises from the partial application of the mode of reasoning 

 which he uses. For, admitting with him the specific gravi- 

 ty of the alcohol employed to be '810 nearly, or -814 1 where 

 water is 1*000, and that 1000 grains of mercury would exhihi^ 

 ^lossofweiglit in alcohol of the temperature stated, of 39*8, 

 when weighed by the hydrostatic balance, vet it does not ap- 

 pear to me necessary to suppose, a priori, so long as each 

 of these substances remains in a fluid state, that the ratio of 

 their densities should differ when uniformly subjecicd to the 

 lower degrees of heat. As, however, my fornier experiments 

 were not made with a view to discover the contraction of 

 the volume of the alcohol, I made no observation relating 

 to it, and now think it right to investigate the subject by 

 experiment. 



A. I first distilled mercury as before, with great care, 

 using only that 30 per cent, of the whole which first came 

 over in the distillation, esteeming it the most pure. I found, 

 by the hydrostatic balance, the specific gravity of this to be- 

 13*613, as 1000 grains lost in distilled water 73*4 at the tem- 

 perature 50 of Fahrenheit's scale. 



B. I took alcohol frou) the same parcel which I had used 

 in my former experiments, and filling a light glass bottle 

 formed with a long narrow neck for the purpose, it was lound 

 to weigh 'SMI, when water weighed- 1 -000 at the tempera- 

 ture of 48° nearly. 



C. Having obtained a mass of very pure silver, procured 

 from luna cornea, 1000 gr. lost in distilled water 07s, the 

 specific gravity of which was thereby found to be 10'225; but 

 bv hamn)ering it into a form convenient for mv purpose the 

 specific gravity increased to ]0'3C)'-2, the loss of ut-iohf Ut'in^ 

 {)Cy9 at 30® of temperature. At the following tempera- 

 tures the variations of loss of weight- are expressed in the 

 second column, the conse<pient specific gravities in the third, 

 according to the usual uhkIq of calculation, and in the 

 fourth is shown the loss of weight in alcohol .-—parts of a 

 series of observations from experiments, that are given niore 



^a. , . J 4 at 



