414 OBSERVATIONS IN CHEMISTRY AND MINERALOGY. 



but where the latter had been against the glass, there was a clear uncolored ri 

 all round, as shown in fig. 6, Plate XXXVI. The presumption had failed 

 However, since the requirement of contact between solvent and solvend still 



a 



existed, there were only two alternatives: either the copper contained 

 volatile copper-compound or the elementary copper must be able to tear 

 itself from the wire surface and pass through the hydrogen atmosphere toward 

 the glass. To gain insight into the mode of the action both alternatives must be 

 considered. As to the first, reference must be had to the analysis of that special 

 copper wire, and this analysis is as follows : 



Copper = 99.9843 



Silver = 0.0157 



Arsenic = none 



Iron - 0.0072 



Oxygen = 0.0463 



100.0635 



The excess in this analysis over 100 is due to the uncertainty inherent in 

 weighing the electrolytic deposit of copper. The metal is very pure indeed ; the 

 presence of a volatile compound of copper is to be absolutely excluded. 



It remains, therefore, to draw the conclusion that copper particles can tear 

 themselves from a polished piece of the metal; that these particles pass through the 

 hydrogen atmosphere toward the glass; that the particles travel radially on the shortest 

 path; that this phenomenon takes place at red heat, at the least 200° Centigrade below 



the melting point of copper. 



These experiments were made during the month of March, 1908. Owing to 

 the pressure of other work the further pursuit of the matter was taken up only 

 in November of the same year. In this second part of the investigation my aim 

 was to establish the exact temperature at which the attractive force exerted by 

 the glass as solvend was able to overcome the resistant force opposed by the 

 cohesion of the copper molecules; for at this stage the facts seemed to point 

 at a clear volatilization of copper ad finem, in the same way as water evaporates 

 below the boiling point and even below the freezing point, so long as the force of 

 diffusion is active in the surrounding air as solvend. Events proved this belief 

 to be erroneous. Thirty-nine independent experiments were made during Novem- 

 ber and December of 1908. The results did not reveal to me the unfailing con- 

 ditions under which the volatilization takes place, but merely that it certainly 



happens when the conditions are right. ... , 



In the following paragraphs a selected number of experiments only will 

 presented exactly in the words of my notebooks. A fresh tube was used for ea 

 experiment ; all these tubes have been preserved and show now as well as 

 did at first. ' . , t p . 



Experiments I and II deal with adjustment of position of lever ol tne v 

 cock at the burners in Erlenmeyer combustion furnace and the reading 

 num, rhodium thermocouple-galvanometer. 



