308 0. C. Farrington — Analysis of "Iron Shale." 



an alteration of the iron to limonite. It is not stated whether 

 these crusts were magnetic or not. The amount of NiO and 

 CoO found by Cohen in the Beaconsfield crust was only 1*68 

 per cent, and by Haushofer in Cranbourne 3'1 per cent. In 

 these cases a leaching of the nickel-cobalt oxides has appar- 

 ently occurred. It is probable that different climatic con- 

 ditions would have considerable influence in affecting the 

 composition of such, rust crusts. 



Explanation of the origin of the laminated structure of the 

 shale is doubtless furnished, as suggested by Barringer, by 

 the shale balls. Oxidation and hydration proceed from the 

 surface inward. These changes cause an increase in bulk of 

 the layers successively reached in the process, so that they 

 separate slightly from the unchanged material beneath, and 

 interstices are afforded through which the oxgyen and water 

 enter to attack the metal. Thus the process is continuous and 

 depends only on time and exposure for its completion. The 

 cracks in the shale balls show that the present superficial lay- 

 ers once covered an interior which has since increased in balk. 



If it be accepted that the iron shale and shale balls are 

 oxidation products of an ordinary Canyon Diablo siderite, it 

 remains to be determined why certain of these siderites should 

 have so oxidized while others have not. One reason is prob- 

 ably to be found in the observation of Barringer that the iron 

 shale often occurs beneath the surface, while the meteoric 

 irons are found only at the surface. Those meteorites which 

 were covered more or less by soil and rock fragments would 

 receive a larger supply of water and hence would suffer more 

 rapid oxidation and hydration than those at the surface. It 

 may also be true that the individuals which have decomposed 

 to form shale balls and shale contained more lawrencite than 

 those which have not. Barringer states* that " the iron 

 centers of the shale balls nearly always show a peculiar oxida- 

 tion of drops of moisture, often colored green, partly perhaps 

 from the presence of nickel. This exudation, Dr. Mallett 

 explains to me, is due to the presence of chloride of iron." 

 It is well known that among the Toluca meteorites, for 

 example, some individuals contain a considerable amount of 

 lawrencite while others do not. The former "sweat" and 

 rapidly decompose, even in a museum case, while the latter 

 remain dry and unaltered. Mr. Nichols has called my atten- 

 tion to a continuous decomposing action probably exerted by 

 lawrencite which has perhaps not been noted before in this 

 connection. Ferrous chloride in contact with air and water 

 forms ferric hydroxide and chloride : 



6FeCl 2 + 30 + 3H 2 = Fe 2 6 H 6 + 4FeCl . 



* Op. cit. p. 882. 



