APPENDIX. 665 



ing moisture nor air. Hence, when it is used, the water formed 

 during the process may be determined with more accuracy than 

 with oxide of copper. It gives out more oxygen than oxide of 

 copper, and therefore admits a greater weight of the substance 

 under analysis to be employed, which is of great consequence ; 

 for we cannot expect accurate results unless the quantity ana- 

 lyzed amount to fifteen or twenty grains." 



In the fifth volume of Dumas' Traite de Chimie appliquee aux 

 Arts, (p. 3), published in 1835, he has given a minute detail of 

 the methods employed by him to secure accurate results when 

 organic bodies are analyzed by means of oxide of copper. It will 

 be worth while to state some of the most important of these me- 

 thods, or checks, as they may be called. 



The oxide of copper should not be prepared by precipitating a 

 salt of copper by means of an alkali, because it has been ascer- 

 tained that some of the alkali is apt to adhere to the oxide. We 

 may obtain pure oxide of copper by heating turnings of that metal 

 in a muffle till they are thoroughly oxydized, or by exposing ace- 

 tate of copper to long- continued ignition in an open vessel. But 

 one of the best sources of oxide of copper is the nitrate ignited 

 in an open vessel. It gives a light, free, and very good oxide, 

 but it should be well pounded and calcined a second time, in 

 order to be sure that all the nitric acid has been driven off. 



The glass-tube in which the combustion is made must be of 

 crown-glass, that it may be heated to redness without melting. 

 Its internal diameter should be about 0-4 inch, and its length 

 not less than sixteen inches. One end should be shut and drawn 

 out by the lamp into a fine point, c. The other end should be 



30= 



smoothed by a file, to prevent it from injuring the cork to be 

 fitted to it. The figure here given represents the combustion 

 tube with the chloride of calcium tube a luted to it. About 1*5 

 inch at the bottom of the tube should be filled with oxide of cop- 

 per. Then the mixture of oxide of copper and the substance to 

 be analyzed occupying the space of several inches. The rest 

 of the tube should be filled with oxide of copper. If the tube be 

 quite filled with oxide of copper, the gas evolved by the combus- 

 tion will force it out of the tube and spoil the analysis. Liebig 

 gets over this difficulty by tapping the tube after filling it, so as 



