448 MEMOIRS NATIONAL ACADEMY OF SCIENCES, VOL. XIII. 



prism and base. The plates with conchoidal fracture show strong metallic luster. Analysis gave the following result 

 (substance taken 0.4115 gram): 



P Fe Ni Co Cu 



15.38 63.97 19.15 1.68 0.00 =100.18 



Fe : Ni (Co) : P=2.3043 : 0.7156 : 1 

 Fe+Ni (Co) : P=3.0200 : 1 

 The second portion of the iron-nickel phosphide which was obtained by treatment of the fine magnetic residue 

 with copper ammonium chloride is not a single product, but consists of grains, crystal fragments, and flakes of the same 

 physical properties as the larger crystals and long needles of rhabdite. Since the latter could not be separated a chem- 

 ical investigation was not undertaken, especially as Meunier had isolated microscopic sections of schreibersite and 

 analyzed them with the following result: 



Fe Ni Co Mg P 



57.11 28.35 trace trace 15.01 =100.47 

 The nickel content is, of course, considerably higher here than we have obtained, but the relation 3.106 : 1 is 

 about the same. In order to determine the solubility the mixture of schreibersite and rhabdite was treated for 35 

 days with cold hydrochloric acid of different concentration. The result was as follows: 



Percent. 



1 HC1+2 aq 22. 59 



1 HC1+5 aq 15. 07 



1 HC1+10 aq 9. 65 



Judgment of these results should be governed by the same rules as were given for tsenite. The rhabdite shows 

 manifold terminations as illustrated by sketches. The simple forms are the most abundant, those with reentrant angles 

 rare. Different terminations of the two ends are so common that one is inclined to consider the crystals hemimorphic, 

 but commonly there appears on one end a square face which may be due to cleavage. Apparently hemimorphic forms 

 other than these are very few. It was not possible to obtain pure rhabdite for analysis, so that it must be yet uncer- 

 tain whether it has the same composition as schreibersite or a different one. From the physical properties and appear- 

 ance the former view seems to us the more probable. 



Jagged pieces. — In comparison with other iron meteorites that we have studied, these pieces in Toluca are charac- 

 terized by small dimensions and abundance. Differential interior structure of the kamacite is indicated. An incom- 

 plete analysis by Koestler gave: 



Fe 93.28 



Ni+Co (by difference) 6.48 



C 0.24 



P 0.00 



100.00 

 This indicates that the pieces consist of kamacite. 



The grains and crystals separated from the nonmagnetic residue (0.0162 gr. ) showed various characters when examined 

 under the microscope. The size of the grains varies from 0.01 to 0.5 mm., though the largest are very rare. Most have 

 a diameter of about 0.04 mm. The following more or less well characterized ingredients were obtained: 



1. Predominant are colorless, transparent, completely rounded grains with bright interference colors. About 

 50 of these were isolated. They are in part rich in minute inclusions which, on low magnification, give the impression 

 of turbid spots occasionally arranged in bands. Where they are larger one recognizes round pipelike forms, probably 

 gas pores which may be accompanied by glass inclusions. Minute microlites were once observed. Determination of 

 the specific gravity gave 2.652. Three of the largest grains having a diameter of 0.2 to 0.3 mm. were investigated and 

 brought into comparison with quartz grains of the same specific gravity. The solution was then exposed both to the 

 rays of the sun and to shade and the grains simultaneously rose and sank. The grains furnished a silicate skeleton with 

 salt of phosphorus; heated in a stream of oxygen they remained unaltered, but after 5 days' treatment with cold hydro- 

 fluoric acid were completely dissolved. The index of refraction is the same as that of quartz. Twelve grains treated 

 with hydrofluosilicic acid left behind a small quantity of salts but apparently not more than a simultaneous blank 

 experiment furnished. Thus it seems to be proved that these grains are quartz, since their total microscopic behavior 

 is so similar that one could not distinguish between terrestrial quartz and that from iron meteorites. 



2. The quartzlike grains were accompanied in some measure by other colorless transparent bodies distinguished 

 from the former by their weak action in polarized light and resembling the grains isolated earlier by one of us, and 

 determined by Magura to be diamond. Strongly heated in a stream of oxygen, however, their number was not dimin- 

 ished and they gave no turbidity with lime solution, hence diamond was not indicated. 



3. Dull, white, opaque grains in large quantity, resembling a silicate strongly attacked by acid. 



4. Fifty opaque single crystals or crystal groups with metallic luster, chiefly of dodecahedral or seldom with 

 octahedral habit. The rounded faces in the latter case indicate hexoctahedrons. Often the crystalites (0.03 or 0.05 to 

 0.15 mm. in size) are arranged parallel to one another to form elongated groups similar to many growths of magnetite; 

 also rhombic dodecahedrons arranged on a trigonal axis are common, which indicate a hexagonal combination of the 



