390 Mr. H. M. Vernon on the 



seems to i3e greatest at about 4° C, though it is only very 

 slightly marked. 



The space between the two curves is a measure of the heat 

 evolved in the cooling of the water. By calculating its area, 

 and also the area of the figure ab c d, it was found that the 

 specific heat of water is increased, on an average, from 12° 0. 

 to 0° C. by about 3 per cent. Only a very rough estimate 

 can be arrived at in this way, as it is not possible to calculate 

 the position of the dotted-line curve with any exactness. It 

 is enough, however, to show that there is a material increase 

 in the specific heat of water at temperatures near to that 

 of its maximum density. 



We have now to see to what extent the results obtained 

 by cooling correspond with the variations in the dilatation 

 of water. The density determinations of Rosetti (Ann. Chim. 

 Phys. (4) x. p. 47), plotted out as a curve, show that down 

 to about 16° C. the density increases regularly with decrease 

 of temperature. From this point downwards the rate of 

 increase of density begins to gradually get smaller, it getting 

 less and less till at 4 C, 07 C. it stops. From this point 

 the density begins to gradually decrease. If the dilata- 

 tion were regular, the density of water at 4° C. would 

 be much greater than is actually the case ; that is, the density 

 of water in the state of molecular aggregation it exists in at 

 4° C. is considerably less than it would be if the dilatation 

 were regular, and no molecular change took place. This 

 fact serves to explain why water below 4° 0. begins to 

 decrease in density. 



The change in the density of water, which, as has been 

 shown, is attended by an evolution of heat, can only be 

 due to water molecules aggregating together, and forming 

 more complex molecules than before existed. As, therefore, 

 from 14° 0. downwards the aggregated molecules are being- 

 formed in larger and larger quantities, whilst fewer and 

 fewer of the unaggregated molecules are left, the lesser 

 density of the aggregated molecules at last at 4° C. 

 counteracts the greater density of the unaggregated mole- 

 cules ; and from this temperature downwards, when fewer 

 still of the unaggregated molecules remain, the density of the 

 water begins to decrease instead of increase. We do not know 

 at what temperature the maximum number of aggregated 

 molecules is formed ; but the curve for cooling appears to have 

 a slight maximum at about 4° C. So it may be taken that the 

 density of water begins to decrease when about half of the 

 molecules present have undergone aggregation. It is not 

 necessary, however, that the maximum of the irregularity in 



