'910.] METEOR CRATER, ARIZONA. 43 



with them was to examine their behavior in the earth's magnetic 

 field by the aid of a small magnet. They generally behaved like 

 pieces of soft iron, or like pieces of the Canyon Diablo iron of about 

 the same size. 



A shale ball perhaps 9 in. in diameter, and which was entirely 

 oxidized, was examined in position, soon after its discovery. It 

 was found by excavating in the pulverized sandstone on the outer 

 rim of the crater. It showed strong local poles scattered over its 

 surface. In general, the polarity of the top was south, that of the 

 bottom, north, but at many points the opposite polarity was found. 



A piece of shale ball iron about 3 in. in diameter and i in. thick, 

 when tested at the crater, showed north polarity all over the outer 

 rim, and south polarity at two nearly opposite points on the two 

 faces. This specimen was sent on to Princeton in a box with other 

 specimens, and when tested again, this peculiar disposition of polarity 

 no longer existed, and it now behaves like soft iron. This fact 

 makes it improbable that the magnetic condition first observed was 

 due to a superficial coating of magnetic oxide, and indicates that it 

 was rather a real magnetic state of the iron. The shale formed 

 from a shale ball by oxidation often shows very peculiar radial 

 structure, and one is tempted to believe that this structure exists in 

 the shale ball iron, and that it may be accompanied by a radial 

 intrinsic magnetization. This view is borne out by some peculiarities 

 of the magnetism of pieces of the shale, but the specimens of shale 

 ball iron found are too few, and have been handled too much since 

 they were found, to make it possible to test this view at present. 

 Much of the shale is so feebly magnetic as hardly to affect a magnetic 

 needle, even when close to it. Occasional pieces are strongly 

 magnetic with well developed poles. 



In 1891 Mr. Marcus Baker made a careful magnetic survey 

 within the crater, and along lines running out on the plain. He 

 found no evidence of any local magnetic field. The author ran 

 some straight lines across the floor of the crater with a sensitive 

 surveyor's compass, which could be read by estimation to about 

 2' of arc, and found no variation in the compass deviation at differ- 

 ent points on these lines. He also made a number of determinations 



