2IO SMITHSONIAN AIISCLLLANEOUS COLLECTIONS [\'OL. 50 



guished by the absence of chlorine, and in "C" at least this element 

 was especially looked for, while "D" contains 0.120 of a per cent; 

 also that the amount of phosphorus in "D" is much greater than in 

 the others. To this difference in the chlorine and phosphorus con- 

 tent is due the difference between the two types of iron in their de- 

 gree of oxidation, and to a certain extent their difference in structure. 



The surface developed by etching a polished section of a shale-ball 

 iron is characterized by the absence of the coarse lamellar structure 

 with broad plates of kamacite, and by the presence of numerous 

 schreibersite areas, more or less regularly arranged. Nodules of 

 carbon and troilite are generally absent (see pi. xix). 



The mass of the etched area is seen to be made up of a darker- 

 colored alloy, or eutectic (plessite), containing numerous masses or 

 plates of schreibersite in parallel arrangement, oriented in the direc- 

 tions of the sides of triangles which correspond to three directions of 

 the octahedron. These schreibersite areas are seldom less than a 

 millimeter in width, and vary in length from two millimeters to thirty 

 millimeters, with an average of about five millimeters. 



Associated with and next to the schreibersite areas, and commonly 

 bounding them along their longitudinal directions, may often be seen 

 a more or less narrow band of cohenite. This cohenite area is more 

 or less interrupted and has not been observed to form a continuous 

 border to the schreibersite. Only occasionally has the cohenite been 

 seen independently of the schreibersite, and then only in very small 

 quantities. Small areas of kamacite may also be seen in the eutectic. 

 These kamacite areas occur very sparingly and are rarely over a milli- 

 meter along their maximum diameters. Bounding the kamacite is a 

 very fine hair-like line of a tin-white alloy which is regarded as 

 tsenite. This alloy occurs elsewhere apparently not associated with 

 kamacite and seemingly developed in thin plates or sheets along the 

 planes of irregular and much-interrupted octahedral cleavages of the 

 eutectic, and which have an orientation identical with that of the 

 schreibersite masses as above mentioned. Further, the taenite may 

 occasionally be found arranged concentrically and outside of some 

 of the smaller schreibersite masses at a distance of about a millimeter 

 therefrom. 



Troilite was not observed as visible segregations in the section 

 under discussion, but under the microscope its presence was occa- 

 sionally shown by treating the eutectic with an acid and cadmium 

 chloride. 



Carbon, as graphite, diamond, or in the amorphous form, could 

 not be detected. The above statement is also true of lawrencite, 



