Lea — Estimation and Dehydration of Silver Oxide. 249 



§ T. Conclusion. — Looking at our results as a whole we find 

 them trenching in a novel way on the solution theories of 

 Arrhenius, Ostwald and vant' Hofl. It is difficult to withhold 

 one's assent from the proposition, that the ions of a molten 

 magma are largely present in the dissociated state, and more 

 especially so as we approach very acid magmas. Terms like 

 colloid applied to these magmas are absolutely without rele- 

 vancy so far as we can make out. To the extent of our enquiry, 

 the behavior of molten rock magmas is in its nature quite iden- 

 tical with that of any aqueous or other solution, the difference 

 being one of solvent. 



The above paragraphs give a mere draft of a series of ex- 

 periments which we propose to carry out with greater vigor. 

 We desire, however, to express our conviction that electro- 

 lytic resistance is not only a valuable aid to penological 

 research, but that the definition of molecular structure is 

 possibly within the reach of the method. 



Aet. XXXIII. — Estimation and Dehydration of Silver 

 Oxide ; by M. Carey Lea. 



Lsr some analytical determinations it became necessary to 

 estimate silver oxide and the question arose at what tempera- 

 ture the moist oxide could be perfectly dried, and also at what 

 temperature it began to lose oxygen. As no such data are to 

 be found, they had to be determined, and the results obtained 

 may possibly be of use to others. 



Moist oxide precipitated by perfectly pure sodium hydrate 

 obtained from metallic sodium, and thoroughly washed, was 

 dried at 100° C. for 20 hours. Of this material 1*5528 grams 

 was taken and heated again to 100° for 20 more hours, after 

 which heating it weighed 1-5524, a loss of 0*0004 grams. 

 It was next heated to 160°- 165° C. for five hours and was 

 then found to weigh "1*5389 grams, a loss of 0*0135 gram. 

 It was then replaced in the 'oven and heated 5 more hours. 

 No loss whatever of weight could be detected resulting from 

 this second heating. The oxide was then ignited and gave 

 1*4358 grams of silver. 



Taking the atomic weight of silver at 107*66 and = 16 

 argentic oxide should contain 6*92 per cent of oxygen (more 

 exactly 6*917). 



From the foregoing it follows that moist silver 

 oxide dried for 40 hours at 100° lost by 



ignition 7*51 per cent 



The same oxide with 10 hours drying, at 160°- 



165° C. lost by ignition 6*70 " 



Calculation for A^„0 gives for O 0*92 " 



O! 



