1896.] Mr Griffiths, On Thermo metric "Fixed Point*." 225 



Section I. Freezing Point. 



I assume that the ordinary definition of freezing point is as 

 Hollows: "the temperature of a mixture of ice and water when the 

 temperature condition is one of statical equilibrium." 



From the theoretical point of view there appear to be no objec- 

 tions to offer to this definition ; practically, however, it is found 

 unsatisfactory. 



As Nernst has pointed out, no such statical condition is possible 

 if the exterior temperature differs from that of the mixture, or if 

 heat is supplied by the stirring, &c. Now, under the ordinary 

 conditions of a laboratory the surrounding temperature is very 

 different from that of the mixture, and although it may be urged 

 that the convergence temperature t g closely approximates to the 

 true freezing point t Q when the masses of ice and water are con- 

 siderable, the difference is nevertheless a measurable one, and when 

 such is the case we cannot be said to possess an ideal standard. 

 In other words, in practice the system is a dynamical, not a 

 statical one. 



The method of under-cooling and super-fusion does not, as 

 Nernst has shown, suffice to overcome the difficulty, and some 

 recent observations of my own lead to the conclusion that very 

 accurate results can only be looked for when the condition is truly 

 statical. 



Another practical difficulty is the inevitable question as to the 

 . purity of the material. I take it that the definition assumes abso- 

 lute purity, otherwise the temperature would not be fixed. The 

 work already accomplished by Loomis, Jones, Abegg, Wildermann, 

 Harker and others has sufficiently proved how considerable are 

 the effects of small impurities ; on the other hand, we certainly do 

 not as yet know (and it appears to me improbable that we ever 

 shall know) the true freezing point of pure water, thus, if we are 

 compelled to wait for that knowledge before we decide upon 

 the value of our lower fixed point there is little prospect of 

 finality. 



As an illustration. At the Liverpool meeting of the British 

 Association I communicated the results of some recent experi- 

 ments which, to my regret, forced upon me the conclusion that 

 the depression of the freezing point due to the solubility of 

 glass was greater than had hitherto been supposed. It is 

 probable that the difficulties which have been encountered by 

 enquirers into the freezing point of dilute solutions have been 

 partly due to want of attention to this matter, for I can find no 

 record of experiments in which the water used has been from 

 the time of distillation uncontaminated by contact with glass. 

 This would be of small consequence if the depression thus 



