360 ox THE ALKALINE METALLOIDS. 



common potash strength of tlic potash formed by the combustion of potas-r 

 pmas^u^i"^"^ ^*'*™' *"^ °^ ^"^^^ potash prepared by the usual process. 

 The facts, that fused potash contains a considerable por- 

 tion of water, and that in its decomposition by galvanism 

 there is not that evolution of hidrogen which might be 

 expected from the decomposition of this water, led me to 

 suppose, that hidrogen enlx'ra'i^to the composition of pot- 

 assium, and that the potas.siu*n in burning must form an 

 alkali containing water ; and the most direct mode of de- 

 termining this appeared to be the ascertaining, whether the 

 product of the combustion of potassium is the same or not 

 in alkaline strength with common ignited potash. As this 

 is an investigation of consi(!erable delicacy, and the precise 

 determination of it difficult, I have since been occupied in 

 repeating and diversifying my former experiments. 

 Chief sources of The principal sources of errour, which attend the inves>- 

 wrour. tigation, appear to arise from the difficulty of converting 



potassium into potash by combustion, without its acting on 

 the substance with which it is in contact; and from the 

 Tarious degrees of oxidation of which it is susceptible, and 

 which are not easily determined. 

 Potassium aets la burning potasbium on platina the latter metal is acted 

 onplatinawhenQj, ^Tid a portion of its oxide remains mixed or combined 

 burned on it. . , , , /. ■, t , 



^ with the potash formed. 1 have noticed the insoluble pow- 



der which remains, when the product of the combustion is 

 dissolved in water, and which is probably principally oxide 

 of platina. This substance having been collected and care- 

 fully weighed, and the weight deducted from that of the 

 product which had been dissolved, it did not appear pro- 

 bable, that it could' give rise to any important errour. 

 Still it was dbsirable, if possible, to avoid its operation; 

 and I endeavoured therefore iq find some other substance 

 on which the potassium would not act, but without success. 

 So it does on In burning it on Chinese porcelain, the surface is eroded ; 

 china, and on afterward dissolving the alkali ip water alight ge- 



and on silver; lafmouslilfe precipitate, probably silex, falls down. In 

 burning it on silver there is also a mutual action between 

 the silyer and the potassium indicated by a small quantity of 

 precipitate, when the product of the combustion is dis- 

 solved in water. It appeared to me howeverj that the 



action 



