THE METALLOGRlAPHT OF METEORIC IROaS" 79 



The low melting point of schreibersite and its tendency to diffuse 

 at even lower temperatures give it an especial interest as evidence 

 of the extent to which a given iron has undergone thermal changes. 

 Experiments have shown that such changes are marked when an 

 iron is heated above 1,000°, but that even as low as 700° there may 

 be diffusion. The appearance of rhabdites with surrounding aure- 

 oles of phosphide enrichment, but still with perfect angles, is there- 

 fore an indication of incipient alteration by reheating at a relatively 

 low temperature in the solid state. 



In irons containing rhabdites or fine phosphide particles there 

 is almost always a P-poor zone around the larger inclusions of schi'ei- 

 bersite in which rhabdites and finer particles are scanty or absent, 

 the phosphide in the smTOunding mass having migrated to the larger 

 bodies. 



Iron -phosphide eutectics. — Needles and larger bodies of schreiber- 

 site often exhibit a eutectic structm-e due to the absorption of iron 

 into solution with the phosphide. This is well shown in Rio Loa, 

 Chesterville, and Cincinnati (pis. 58, 59). Such inclusions having 

 melted and become a liquid phosphide-iron solution, in cooling re- 

 jected the excess of iron above the eutectic ratio in the form of clear 

 borders or droplike particles, the remainder solidifying as the Fe- 

 FcaP eutectic. Thus in irons that have been reheated schreibersite 

 inclusions usually show extensions of the diffused phosphide along 

 adjacent grain boundaries, giving the inclusion a thorny outline. 



Iron-phosphide eutectics are not necessarily a secondary product 

 due to reheating but may be produced during the separation of the 

 schreibersite in the gamma phase. This may be true of Soper (pi. 11) 

 in which the abundant segregations of phosphide along grain bound- 

 aries exhibit a eutectic structure; though apparently by reason of 

 slow cooling the excess of iron above the eutectic ratio disappeared 

 by absorption into the surrounding area. 



Ordinarily an iron-phosphorus eutectic structure is visible with 

 ordinary etching, for although the etchant does not affect the phos- 

 phide it acts strongly along the interfaces of the phosphide particles. 

 The structures referred to above in Cedartown and Helt Township 

 were observed only with picrate etching; but, as stated, the exact 

 nature of those structures is open to question. 



Minute kamacite bodies in very high-nickel irons (e. g., Cowra, 

 Klondike, Freda; see plates) often contain droplike inclusions of the 

 eutectic, which in such cases is clearly a primary product dming 

 transformation and not a result of reheating. The examples shown 

 in plate 55 (Helt Township and Cedartown) may have such an origin, 

 although that conclusion is not free from doubt. 



Nature oj phosphide eutectics. — The absence of a eutectic structure 

 in many phosphide areas, and its presence in varying forms in 



