occluded in Meteoric Iron from Augusta Co,, Virginia. 313 

 for analysis as follows : — 



h 



Portion A. . . . 52-02 per cent, of the whole was collected in 2 30 



Portion B 24-11 „ „ „ 2 20 



Portion C . . . . 23-87 „ „ „ 9 40 



100-00 14 30 



It will be seen that the greater part came off within the first two 

 hours and a half; but the process lasted fourteen hoia's and a half 

 and was not entirely over at the end even of this time. The heat 

 had been gradually raised from dull redness to something nearly ap- 

 proaching whiteness at the end of the time ; and when the experiment 

 was stopped very small but still perceptible traces of gas were still 

 coming off, though their appearance was immediately arrested when- 

 ever the temperature was allowed to fall but a little below the high 

 point which had been reached. 



The piece of iron taken out from the tube when it had become quite 

 cold was found glazed by a thin film of fused phosphide of iron 

 and nickel (Schreibersite), thickest on the edge which had been low- 

 est, this phosphide having oozed out from the mass at the very high 

 temperature used. 



The tubes used to collect the gas during the first portion of the 

 time occupied in the experiment were found slightly moistened on 

 the inside, and the moisture, which had a distinctly acid reaction, 

 was proved to contain hydrochloric acid, this having no doubt been 

 derived from the chlorine existing in the iron in combination with 

 that metal and with nickel. 



Careful analysis of the gas yielded the following results by volume 

 for the three portions separately collected : the fourth column of 

 figures, obtained by summing up the three which precede it, gives 

 the percentage composition of the whole of the gaseous matter ex- 

 tracted from the iron : — ■ 



Portion A. 

 . Hydrogen 22-12 



Carbonic oxide .... 15-99 



Carbonic anhydride . 7 "85 



Nitrogen 6-06 



Portion B. 



Portion C. 



Total gas 



10-52 



3-19 



35-83 



11-12 



11-22 



38-33 



1-02 



•88 



975 



1-45 



8-58 



16-09 



52-02 24-11 23-87 lOO'OO 



Other gases were tested for, but none could be found ; no free 

 oxygen could be detected, nor any compound of carbon and hydrogen. 

 From these figures it appears that hydrogen maintains about the 

 same proportion to the other gases in A and B, but diminishes 

 largely in C, that carbonic oxide increases in amount in B as com- 

 pared with A, but remains about the same in relative amount in C, 

 that carbonic anhydride diminishes throughout the whole continuance 

 of the experiment, and that nitrogen falls off in B as compared with 

 A, but largely increases again in C. 



Contrasting the results with those of Graham, and noticing first the 



