W. Hallock — Chemical Action hetween Solids. 403 



sium nitrate, potassium, calcium and ammonium chloride and 

 sodium and potassium hydrate, with a similar result in all cases. 

 These are all well known results, but wherein do they differ 

 from the new method of forming alloys ? This question sug- 

 gests another. Are the metals combining to form an alloy in 

 the new way a freezing mixture ? A thorough investigation 

 of this question would require more complicated experiments 

 than I had time to perform. One test, however, is very 

 simple, that with potassium and sodium. . 



Into a small porcelain crucible weighing 15 grams and 

 containing about an equal weight of petroleum were placed 

 •pieces of the two metals, about 3 grams of each. One 

 junction of a thermo- element was forced into the piece of 

 potassium and gave its temperature accurately. After the 

 whole had assumed the room temperature, clean faces of the 

 two metals were brought in contact, the liquefaction began and 

 the temperature immediately fell. It required about two 

 hours to complete the liquefaction and about one and a half 

 hours to attain the minimum of temperature. No precautions 

 were taken to prevent the calorimeter taking up heat from its 

 surroundings, and no doubt it absorbed considerable in the long 

 time, and yet the maximum fall in temperature amounted to 

 24° C, very large considering the small weight of the reagents 

 compared with the calorimeter. Thus it appears that sodium 

 and potassium are, under such circumstances, a " freezing mix- 

 ture," and analogy at least would lead one to believe that 

 other alloys also absorb heat in their formation; but future 

 experiment must decide the point. 



In the cool vessel above described a piece of sodium or 

 potassium was placed upon a piece of dry ice, almost instantly 

 the reaction commenced and proceeded vigorously. It is, 

 however, scarcely safe to consider this a case of chemical 

 action between solids, because the reaction is probably as 

 follows : the vapor from the ice attacks the metal forming the 

 hydrate which unites with other ice forming a solution, which 

 is then further acted upon by the metal, and in the whole 

 process heat is generated sufficient to raise the temperature of 

 the reagents very considerably. Perhaps in the other freezing 

 mixtures, ice and salt, etc., it is the vapor of the water or ice 

 which initiates the reaction. 



In view of these and other considerations, the idea is evident 

 that perhaps many substances have a slight vapor tension at 

 temperatures considerably below their melting points, and are 

 surrounded by a thin atmosphere of their own vapor over their 

 clean surfaces, and it is only necessary to bring two such atmo- 

 spheres to interpenetration in order to initiate the reaction which 



Am. Jour. Sci.— Third Series, Vol, XXXYII, No. 221.— Mat, 1889. 

 26 



