I 



EXPEKIMENTAL E.NOWLEDGE OF THE PROPERTIES OF MATTER. 503 



The general result of these data is that solids in genei'al expand very 

 slowly indeed with rise of temperature, at least between 0° and 100°, but 

 as the temperature-limit is increased so as to include higher and higher 

 temperatures the rate of expansion in general increases ; if the coefficient 

 of linear expansion remained constant a unit length at 0° would become 

 1 + at, where t is the coeflBcient of linear expansion, the cubical expansion 

 being approximately l + 3at, in which 3a is the coefficient of cubical ex- 

 pansion ; or if we determine the coefficient of cubical expansion o that of 



linear expansion is ; . But in general it is necessary to add a term in t^ 

 o 



to express witb sufficient accuracy the results of experiments. 



Although many most accurate determinations have been made in 



recent times on the expansion of solids the results are not connected 



with any general laws, as in the case of the expansion of gases with rise 



of temperature. 



Liquids. 



Mercunj. — The absolute expansion of mercury was determined by 

 Dulong and Petit and their results given on p. 136, 'Ann. Chim.' (2), 7, 

 1818. Regnault, using Dulong and Petit's general method with improve- 

 ments, made an extensive series of determinations of volumes at 0° and 

 350°, and for every interval of 10° between these limits, with the co- 

 efficient of expansion at each temperature and the mean coefficient 

 between that temperature and 0° C.^ At temp. 100° the true coefficient 

 is given as 000018305 by misprint for -00018405.- 



By means of the absolute expansion of mercury it is easy to find the 

 expansion of a thermometer-tube or other similar glass vessels in which 

 the expansions of liquids are observed, so that from the apparent expan- 

 sion of the liquid in the dilatometer, as such an instrument is called, the 

 true expansion may be determined. This is of fundamental importance 

 for the determination of the expansions of liquids. For results obtained 

 by Bosscha and Wiillner see ' Jahresb. f. Chem.' 1874, p. 70. 



Water. — Feltz ^ gives an account of experiments by Rossetti (professor 

 at the University of Padua) on the maximum density and the dilatation of 

 distilled water ; for the temperature of maximum density Rossetti finds 

 4'07°, agreeing very closely with H. Kopp's value 4'08°* and differing 

 from 4-00° given by Despi-etz,-^ and from 3'92° given by Pierre.'^ The 

 above-named and other experimenters gave tables for the volumes and 

 densities of water at temperatures up to 100° C. and below 0° ; Sorby '' 

 and Mendelejeff * gave results for water above 100° ; and Hirn ° for a 

 number of other liquids, chiefly organic, up to and above their ordinary 

 boiling-points, and in the case of water up to 200°. In the case of water 

 the rate of exiDansion is more rapid the higher the temperature above 

 (about) 4° ; and below this, in the unstable condition of water remaining 

 liquid below its ordinary fi'eezing-point it expands as temperature sinks ; 

 in other words, at 4° water has its maximum density, the densities 

 diminishing from 4° to —10°, and from 4° to 200°. For further references 



' 3le»i. de VAcad. t. sxi. 1847, p. 271. = See Watts' Diet. iii. p. 56. 



' Ann. di Chim. (4), 10, 1867, p. 461. * Pogq. Ann. 72, 1847, p. 48. 



^ Ann. Chim. (2), 70, 1839, pp. 24, 47, 48. " Hid. {?>), 15, 1845, p. 348. 



' Phil. Mag. (4), 18, 81. ' Pogy. Ann. 119, 1863, p. 1. 

 * Ann. Chim. (4), 10, p. 32 ; and 11, 1867, p. 5. 



