116 Mr. J. J. Waterston on the Expansion 



perceive it. It must be allowed that our ideas of the action of 

 heat as a source of work are still very obscure. The patient and 

 almost silent researches of twenty years, perfectly unbiassed by 

 partial influences of any kind, have convinced me that work is 

 directly convertible into heat by friction, but that heat is not 

 directly convertible into work. It is certain that in elastic fluids 

 it is the quality or condition of the heat, and not its quantity, 

 which is the measure of the dynamical equivalent. Calorific 

 energy is not simply the molecular motion known as heat ; and 

 " the phenomenon which we call heat, as a dynamical agent, may 

 be only the exhibition of certain states (or conditions) of the 

 particles of matter, dependent on, and correlative with, forces in 

 which they are involved." 



Believe me, Sir, 



Your most obedient Servant, 

 Palermo, June 6, 1863. Joseph Gill. 



XV. On the Expansion of Water at High Temperatures, 

 By J. J. Waterston, Esq.* 



[With Three Plates.] 



IN the account given of a law of liquid expansion in the Philo- 

 sophical Magazine for June 1861, it was stated that water 

 alone, of all the liquids examined, did not conform to the law in 

 the lower part of the scale of temperature, and that even up to 

 200°j- it showed no tendency to do so. My attention has lately 

 been drawn to this subject again by having occasion to graduate 

 a water-thermometer. The inequality in the rate of expansion 

 of water is so great that, to obtain even moderate accuracy in the 

 scale, it was requisite to find the law of the differences. To do 

 this a scale of equal parts was attached to the stem, and X, the 

 length between t and t l (two near temperatures), was divided 

 by t^—t^ and the quotient set off as ordinate to i(^ + ^i), tne 

 middle temperature ; the points were found to range in a straight 

 line that came down and crossed the axis at 4° nearly. It would 

 thus appear that the rate of expansion increased exactly in the 

 simple ratio of the distance from 4°, and it would seem that the 

 curve of expansion was the conic parabola having its vertex at 

 the point of minimum volume. 



Let PL represent unity or volume at 4°, and NQ=v the 

 volume at t. Then 



*-4 = LN = PK = 2/, 



* Communicated by the Author, 

 t The temperatures are Centigrade. 



