22 



THE GASES IN ROCKS. 



TABLE 8 Concluded. 



1 No. 107. Chemical analysis by J. M. Davison ; Fe, 14.68; Ni, 1.60; Co, .08; S, 1.37; P, .15; SiO", 35.82; FeO, 

 15.53; MgO, 22.74; CaO, 2.99; A1 2 O 3 , 3.60; Na.,O, 2.07; KoO, .32; 100.95. 



2 No. 108. Average of 13 analyses compiled by Farrington (Pub. Field Columbian Museum No. 120, pp. 82-84) : 

 Fe, 89.66; Ni, 7.90; Co, 0.63; P, 0.24; S, 0.14; Si, 0.02; Misc., 0.57; 99.16. 



s Iron borings and filings were used, but with these there was included a little rust, which adhered to the metal 

 when it was withdrawn with a magnet, 



* The material used in this determination consisted of bright borings carefully freed from rust, a pocket of which 

 unfortunately was encountered in drilling. But in spite of much care exercised in drawing put the metallic borings 

 with a magnet, some rust adhered to them and consequently was heated with the metal in the combustion tube. 

 However, this amounted to much less than in the first determination. No filings were used. 



5 For this determination borings from the interior of the specimen were carefully made by \Vm. Gaertner & Co., 

 scientific-instrument makers. There was no visible rust adhering to these borings, which were then worked twice, 

 with a magnet, without any impurities being left behind. The white paper on which this operation was performed 

 failed to show the slightest discoloration, such as it had done in the two previous determinations. As usual the 

 material was heated at 100 for 3 hours, in the presence of PoOs, and then allowed to stand in the vacuum overnight 

 to remove all free moisture. Though the vacuum was perfect at the end of this time, the first two pumpings of gas 

 (amounting to 0.2 cubic centimeter) which were evolved when heat was applied, were not kept, since it was desired 

 to eliminate atmospheric air as a source of nitrogen. 



A comparison of the three determinations, No. 108 shows what a tremendous effect the presence of a little iron 

 rust will have upon the gases evolved from a metallic meteorite. Much of this gas is doubtless derived from iron 

 carbonate and the hydrated oxide of iron, as will be explained under the topic of gas due to chemical reactions. 

 Great care is therefore necessary in making gas analysis of iron meteorites to avoid any contamination of rust. 



6 The most striking feature of these analyses is the unusual amount of sulphur dioxide, which indicates an 

 oxidized condition of the ore. 



7 The odor of sulphur dioxide was very prominent in the gas obtained from these two Vesuvian lavas. 



