26 



REPORT — 1872. 



Fijr. 2. 



cune wliicli belongs to temperatiu-es below tliat of the triple point T *. Tbis sup- 

 posed extension of the steam-vrith-water curve into temperatures below that of the 

 triple point, where freezing would certainly set in if any ice were present, is to be 

 conceived of as a curve corresponding to states of equilibrium between tlie steam 

 and water. It is well known that water can, in various circumstances, be reduced 

 in temperature below its freezing-point without its freezing ; and this the author 

 attributes to a difficulty of making a beginning of change of state f. It is also 

 known that the presence of a gaseous atmosphere, of common air with aqueous 

 vapour in contact with water, does not uecessarily introduce any condition which 

 will give liberty to the water-substance to make a beginning of change of state 

 into ice, either from the liquid or the gaseous part, or from both at their face of 

 contact. Thus there can scarcely be a doubt but that the steam-with-water curve, 

 L T, has a practically attainable extension past T ; and valid reasoning, the author 

 thinks, may certainly be founded on the supposition of this cui've as one of equili- 

 brium between steam and water, 

 whether or not, in various modes 

 of experimenting, it might be 

 easy or difficult or unmanage- 

 able to practically exclude all 

 conditions which would give 

 liberty to make a beginning' of 

 the formation of ice. We may 

 then see that, supposing steam 

 and water to be present together 

 in a condition of temperature 

 and pressm-e represented by any 

 point such as C in tig. la, there 

 is perfect freedom for the transi- 

 tion either way between water 

 and steam. That is to say, 

 while the water and steam are 

 maintained at the temperature 

 and pressure of the point C, the 

 water is perfectly free to change 

 to steam, and the steam is per- 

 fectly free to change to water. 

 Let, "for brevity, the temperature 

 and pressure of the point C be 

 denoted by t^ andp^ respectively. 



Now, to aid our conception in 

 a process of theoretical reason- 

 ing, let us imagine an apparatus 

 possessing certain qualities in 

 theoretic perfection, thus : — 

 (see fig. 2). 



Let there be a cylinder, stand- 

 ing upright, closed at bottom, 

 open at top, and with a piston 

 which works without leakage 

 and without friction. 



Let the weight of the piston, 

 together with the atmospheric 



L 



y/- 



=^g /npe rdlu rei. 



^-^Jiick'is-- 

 —trelculilic^ 



sThplerPoTnt^ 



?ri-EH:- 



-Temp erature. 



W/////////.V////////W////////////W/////'/////////7/, 



load on it, be balanced by a counterpoise B ; or else let the whole apparatus be 

 conceived to be enclosed in a large external vessel from which the air has been 



* The meaning of the "Triple Point" is explained in the paper abeady referred to in 

 last year's Transactions, page 32. 



t 'in papers by the author (Proceedings of Eoyal Society, Nov. 24, 1859, page 158 ; and 

 British Association Eeport, Transactions of Sections, 1859, page 25), the principle of 

 attributing such phenomena to a difficulti/ of maJcing a beginning of change of state was, 

 so far as he is aware, first announced. 



