342 ^ew analytical Researches 



larger quantity of it destroyed by the distillation of the fu- 

 sible substance, wilh a fresh quantity of potassium, than 

 by the distillation of it in its conmion state. On this idea 

 J made several experiments: the rcsidts did not differ much 

 from each other, and were such as I had expected. I shall 

 describe one process made with the same apparatus as that 

 which I have just detailed. — Barometer was at 2y3, ther- 

 luometer 70- Fahrenheit. 



Six {grains of potassium were employed in an iron tray; 

 10 cubical inches of ammonia were absorbed, a small glo- 

 bule of metal remained unconverted into the fusible sub- 

 stance. A fresh piece of potassium, weighing six grains, 

 was introduced into the tray. 



The iron tube and the adaptors (having together a capa- 

 city equal to 4'3 cubical inches) contained common air. 



The gas was collected in three pji^rtions, there was no ab- 

 sorbable quantity of ammonia in either of them. 



The first portion, that produced before the tube became 

 red, was eight cubical inches. 10' 25 of it, detonated with 

 3"3 of oxygen, diminished to 8. 



The second portion equalled five cubical inches ; Qj of 

 this, wilh 5 of oxygen, left a residuum of 3|. 



Of the third portion. 2 cubical inches and -i- came over. 

 9 of it, detonated wilh 5 of oxygen gas, left a residuum 

 of 1-4. 



The iron tube and the adaptors contained, at the end of 

 the experiment, as was proved by cooling and the admis- 

 sion of hydrogen, 2'3 cubical inches of gas, which appeared 

 of the same composition as the third portion. Nearly 7 

 grains of potassium were recovered. 



A comparison of these results, with those stated in the 

 preceding page, will fully prove, that there is a much smaller 

 proportion of nitrogen to the hydrogen, in the case in 

 which the olive-coloured substance is distilled with potas- 

 sium, than in the other case, and there is likewise a larger, 

 quantity of potassium converted into potash. 



The loss of nitrogen, and the addition of oxygen to the 

 potassium, are sufficiently distinct in both processes j and 

 the want of a correspondence between these results, and 

 those of the experiment detailed in page 55*, are not greater 

 than might be expected, when all the circumstances of the 

 operation are considered. In the instance in which a 

 double quantity of potassium was employed, more potash 

 must have been formed from the oxygen of the common 

 air in the tubes ; and the fusible substance, in passing 



* See page of this volume. 



through 



