1 88 E. A. STEPHENSON 



In the early part of the investigation Jena glass tubes were used, 

 provided with an inner tube of copper open at one end, in which 

 the minerals and solutions were placed. This prevented contact 

 of the solutions and the glass when the apparatus was placed in a 

 nearly vertical position, but left only a small part of the solutions 

 in direct contact with the mineral powder. Where glass tubes were 

 used, these were first placed inside of steel tubes made of ordinary 

 steam pipe i§ inches inside diameter, closed at one end by a screw 



Fig. 6 



Fig. 5. — Thermo-regulator, plan. Fig. 6. — Thermo-regulator, side elevation 



cap and placed in the oven with the open end against one of the 

 holes and 0' , Fig. 3. Accidents to the thermo-regulator from pos- 

 sible explosions were thus prevented, though the trouble caused by 

 bursting of the glass tubes led to their abandonment. In their 

 place copper tubes about 18 to 22 inches long with f inch inside 

 diameter and T Vinch walls were used. One end of each of these 

 was closed with a plug of copper about \ inch thick driven in 

 about f- inch past the end of the tube and the end then sealed in 

 an oxy-acetylene flame. The solution and the mineral powder 

 were then placed in the tube, a plug of copper inserted as above, 

 and the end then sealed while the tube was partly immersed in a 

 freezing mixture. 



