226 Mr. Callendar on Experiments ivith a Platinum 



and the spitting was much less violent. This seemed to show 

 that silver would not absorb oxygen readily unless it was 

 heated more than 40° Pt. above its melting-point. It was 

 therefore again heated to 890°, and oxygen was passed for 

 half a minute. The observations taken during cooling gave 

 curve No. 4, which is almost a repetition of 2 except that 

 less oxygen was absorbed. The silver was then heated up 

 to 910° Pt., and oxygen was passed for two or three minutes 

 with a view to saturate it. On cooling it beg-an to freeze at 

 825 "1, as shown in curve 5. The spitting in this case 

 was exceedingly violent. I have not succeeded in making 

 the silver absorb a larger proportion of oxygen than this. A 

 subsequent experiment, in which it was maintained at 930° 

 while oxygen was passed for five minutes, gave a record which 

 was almost an exact repetition of No. 4. 



The greatest lowering of the freezing-point which I have 

 observed in the case of silver saturated with oxygen in this 

 way amounts to 6°'4 C. This corresponds very nearly with 

 a ratio of 1 molecule of oxvgen to 200 molecules of silver. 

 Assuming both molecules to have the same atomicity, we find 

 that 100 grams of silver are capable of absorbing -074 gram 

 of oxygen, i. e. about 5 \ times its volume of the gas. Gay- 

 Lussac found that silver, when fused under nitrate of potash, 

 absorbed in one case 22 times its volume of oxygen. Graham 

 and Roberts-Austen have shown that spongy silver in the 

 solid state at a red heat will absorb from 6 to 8 times its 

 volume. It would have been interesting, had time permitted, 

 to have made quantitative experiments, by some independent 

 method, to determine the amount of oxygen absorbed in the 

 present case. 



A comparison of curves 1 and 4, in which the initial 

 conditions and rates of cooling were the same, seems to indi- 

 cate that the total heat evolved in the solidification of silver 

 free from oxygen is less than in the case when oxygen is 

 present. The quantity of heat evolved in the decomposition 

 of the solution must therefore be considerable. The whole of 

 the silver appears to be absolutely solid at a temperature at 

 least 10° above the point at which it begins to spit. The heat 

 evolved in spitting is sufficient to raise the extruded volcano 

 of silver bubbles to a temperature considerably above its 

 melting-point. A great deal of the heat is certainly dissi- 

 pated in the volcano, where it cannot affect the pyrometer. 

 The amount actually generated must therefore be considerably 

 in excess of that indicated by the curves. 



I have also made some further experiments by the same 



