Mbns. C. Despretz on the Maximum Density of Liquids, 9 



different tubes. One tube gave twice — 3°*69; a second 

 — 3*^-60 and —^'^'59 ; a third —S°'ll, We now see the rea- 

 son why Marcet and Ermann did not discover any maximum 

 density in sea- water, because they searched for it above the 

 freezing point, while it is situated at more than one degree 

 below. 



The solution of the question relative to sea-water sufficed 

 for the purposes of physical geography ; but the history of cor- 

 puscular properties required a more general solution. It was 

 necessary to extend these experiments to a certain number of 

 aqueous solutions in order to discover the course which the 

 maximum takes as the addition of foreign matter lowers it. 



For this purpose I dissolved several quantities of foreign 

 matter in the proportions 1, 2, 4, 6, 12, and 24. Each of 

 the substances was employed in a pure state, which it is now 

 very easy to ensure. The chloride of sodium, the chloride 

 of calcium, the carbonate of potash and that of soda were 

 melted. The carbonate of potash was obtained by calcining 

 pure and crystallized bicarbonate. The sulphate of copper 

 was employed crystallized. Water not being an essential part of 

 this salt, it was subtracted ; while the pure hydrate of potash, 

 concentrated sulphuric acid and absolute alcohol, (water be- 

 ing in certain respects essential to their composition, smce 

 heat alone does not expel it), were considered as anhydrous 

 bodies. We will mention some of the results obtained : 



Sea Salt, 



0-000123 of salt. ..freezing point* — 1°-21, Max. •- 1""- 19 Cent. 



0-0246 — — — 2-24 -- -1-69 — 



0-0371 — — -2-77 — -4-75 — 



0-0741 — — —5 -10 — —16-00 — 



Chloride of Calcium, 



0-0371 of salt. ..freezing point -3°-92, Max. -2°-4? Cent. 

 0-0741 — — -5°-28 — 10-43 — 



This sinking of the maximum, says the author, cannot be 

 the consequence of a partial freezing of the liquid mass, since 

 the curve, representing the expansions above and below the 

 maximum, is quite regular, as the drawings which I now lay 

 before the Academy will show ; the freezing of the smallest 

 part would determine, in the curve, points which by geome- 



• The temperatures are those marked by the thermometer at the mo- 

 ment when the liquid is on the point of freezing. The temperatures indi- 

 cating the actual freezing, i. e. which are for the solutions what zero is for 

 pure water, are not so low. 



Phil, Mag. S, 3. Vol. 12. No. 71. Jan. 1838. C 



