Atomic' ¥oiumeahSy3^^Sici/IS^G^ty, ^ 



nlums, &c., dissolve in water without increasing its bulk more 

 than is due to the liquefaction of the water which they them- 

 selves contain ; the anhydrous salt taking up no space in so- 

 lution. = '-'- -i^-'!^ ■''-' '" ^:';""j"'' 



b. Anhydrous salts, or salts containing a StfteflF^p^^iNiliil 

 of \yatef, affect a certain number of volumes iu solution, which 

 pass dttnjgjYi|^;thp^..lin(9^^n^ed into theii- 1^9^:^ .-yY^, ^her 



c. Tne volume occupied by double salts when dissolved 

 is the sum of volumes occupied by their constituents vi'hen 

 separate, with the exception of certain cases described in the 

 previous sections. 



,'i Prop. II. — The volume occupied by a salt in the solid state 

 has a certain relation to the volume of the same salt when in 

 solution; and has also a fixed relation to the volume occupied 

 ^y any other salt. sjiM.irt 



.■ a. The volume of an equivalent of any salt is eithei''l l|'>di]^ 

 a multiple of 11, or of a number very nearly approachitig'tfi^ 

 number 11. '!>'° V^^^i 



b: Or the volume of a salt is 9*8, or a multiple of 9'8j'bl^ 

 in other words, of the. volume occupied by an equivalent of 

 solid water (ice), noiiaoup 



c. Or the volume of a salt is made up of a certain multiple 

 of the number 11, added to a certain multiple of the number 

 9-8. 



.On each of these heads we would 6fFer a few retnarks. ^ 

 'With two assumptions we have been enabled to connect 

 with each other the volumes occupied by all the salts exa- 

 mined by us in the previous sections. These assumptions 

 are, that the divisor for the volumes of salts is either 11, or a 

 number very nearly approaching to it, or that the divisor is 

 9*8, the volume of ice itself. 



We have been guarded in stating positively that the first 

 divisor is absolutely 11, because we do not in the pireseht 

 memoir enter into the connection between this number and the 

 volume of ice, 9*8. To shoAV, however, that our experiments 

 agree with those of recent accurate experimenters, and that 

 the number 11, which w^e have at present to announce empi- 

 rically, cannot be wide from the truth, Ave append the theo- 

 retical and experimental results upon the alums, which we 

 stated to possess twenty-five volumes, in which therefore any 

 considerable error in the number 11 would be multiplied by 

 25, and plainly show itself in the results. Notwithstanding 

 this severe test, it will be seen that the theoretical and ex- 

 perimental numbers are actually within the errors ^f the 

 balancesicdqlw^ nci^angBm adj «« xlDua ^aJiija nifiJiaJ .» 



