OBSERVATIONS IN CHEMISTRY AND MINERALOGY. 413 



rh< oram i salt is certainly a decamine. Its formula will be (Co,(N Hs)io)(SO,)i 

 4H 2 0. The calculated per ntages for this formula are: ( o = 21.26, NH| 

 {0.84, S0 4 - 34.81, II ,0 - 13.06. 



Since the green substance is shown, microscopicall to be a mixture of green 

 and orange salt it is now possible to figure out the ratio between tin two, nan 

 three Praseooctaminosulfate 3[(Co*(NH 3 )8(S0 4 , Cl 2 )t.6H s O]: on. Luteode- 

 caminosulfate[Co l (NH,) 1 o(S0 4 ) 2 .4H,0]. 



Note. — These observations were made between April 1 and April *Jv I 10 



2. Can Copper Ions Detach Themselves from Metallic Coitii. \nd Pas* 



Through Space at a Temperature Considerable Bi low 



the Melting Point of the Mltal? 



This question arose under the following circumstances. I had defer* - 

 mined the oxygen contained in refined copper in the usual way, namely, by 

 beating about 20 grams of short pieces of the wire in a current of dry hydrogen 

 within a bulb of hard glass. The wire had been sent from the smelter of the 

 Quincy Mining Co. at Ripley, f Of a complete analysis, while Professor Fisher was 



to determine the electric conductivity. After removim the copper from the bulb 

 I was struck by the intense ruby color of the glass. Something similar had bet 11 

 observed on previous occasions and notably when coils of copper wire-gause which 

 had been oxidized in preparing nitrogen gas, were reduced by hydrogen ; but never 

 before had the peculiarity of the phenomenon impressed me to such a marked 

 degree. Many years ago, about the middle of the last century, Professor Max 

 PettenkofTer, of Munich, had advanced the theory that the copper-ruby of stained 



glass is due to the dissolution of elementary Copper in the glass, and not to th< 



forming of a cuprosilicate. The phenomenon is similar to the gold-ruby j- tin. 

 This theory I have always believed in and have taught for forty years. It is prob- 

 ably universally believed in at this time. But dissolution means eontact between 

 the sol vend and the solving agent. In the phenomenon before me then were 

 certainly many points of contact between the glass and the many small pieces of 

 wire. If this contact were avoided there would be no more loss of copper through 

 its passing into the glass. I\ v mind was working, then, solely on the improve- 

 ment of the oxygen determination. To prevent that contact was not difficult. 

 Instead of cutting short pieces of the thoroughly cleaned and jwlished t<*t- 

 wire, pieces two inches long were cut and these were made into a bundle or 

 fascicle and tied with two pieces of asbestos cord, at a distance of one- 

 half inch from either end. Made up in this fashion the bundle just fitted 

 into the hard-glass combustion-tube so that there was a clear air space between 

 copper and glass equal to the diameter of the asbestos cord (see fig. 5, Plate 

 XXXVI) . For two hours dry hydrogen was then passed through the tube while t he 

 copper was at visible red heat. The tube was then allowed to cool, the ga- still 

 passing. The result was unexpected indeed,/or there was the ruby-color in the glass 

 wherever the glass was opposite the copper , both in front and behind the asbe$Umcord, 





