1817.) Dilatation. of Liquids at all Temperatures. 44%. 
Thus, according to the hypothesis of Mr. Dalton, the common 
mercurial thermometer can never sink further than 32° Reaumur 
below 0°, which is the point of congelation of that liquid; but this 
is quite contrary to experience ; for we know that all the liquids 
hitherto observed may, with certain precautions, be cooled below 
their point of congelation without becoming solid ; and in that case 
they continue to follow the law of dilatation belonging to them. 
Thus water, for example, dilates equally on both sides of its maxi- 
mum, whether heated or cooled, 10° R., reckoning from that point. 
And olive oil, which in the open air congeals at a very moderate 
cold, may be cooled down to — 14° Reaumur without ceasing to be 
liquid, as is obvious from the experiments of Deluc: and in this 
state it continues to contract according to the same law that it fol- 
lowed in other parts of the thermometric scale, because, as ] have 
shown, that law excludes it from a maximum of condensation. The 
same holds with mercury, as is shown by the discussion of Mr. 
Cavendish respecting the experiments of Hutchins at Hudson’s 
Bay ; for it results from that discussion that mercury, like other 
liquids, may be cooled below its freezing point without becoming 
solid, and that this frequently happened in the experiments of 
Hutchins; and in these cases the mercury continued to contract 
gradually till the moment of solidification, when it suddenly under- 
went a much more considerable contraction. All these results are 
contrary to the law of dilatation supposed by Mr. Dalton. The 
same inconsistency exists with respect to all the liquids that contract 
progressively to the instant of their solidification. 
If, notwithstanding these physical contradictions, we wish to exa- 
mine Dalton’s hypothesis relative to the expansion of water, which 
is the object that he had chiefly in view, we shall find that it cor- 
responds much less accurately than the empyrical law deduced from 
the thermometrical observations. This is very natural, because these 
observations furnished us with a very delicate test, on which our 
formulas were moulded. It is obvious that a slight change in the 
thermometrical scale, such as that resulting from Dalton’s hypo- 
thesis, between 0° and 80°, cannot produce a considerable eftect 
upon a liquid which dilates as water does ; and this will be the Jess 
sensible, as Mr. Dalton has in some measure compensated the 
excess of his scale of true temperatures, by the excess of tempera- 
ture which he assigns to the point of maximum condensation ; but 
the error may become greater when this scale is applied to other 
liquids, the dilatations of which are greater; for example, to 
alcohol; and this is what happens. Mr. Dalton himself has acknow- 
ledged that the law of dilatation deduced from his hypothesis cannot 
be reconciled to the thermometrical observations which Deluc made 
on this liquid, especially at high temperatures. Struck with this 
disagreement, he has been induced to call in question the accuracy 
of these observations ; for, says he, as the dilatation of alcohol from 
62° to 80° R. must have been conjectural, it is possible that Deluc 
may have exaggerated it; but the experiments of Deluc and of other. 
