464 MEMOIRS NATIONAL ACADEMY OF SCIENCES, VOL. XIII. 



to fluid structure arises. Small grains isolated by means of a needle appeared for the most part 

 as olivine with inclusions of gas bubbles. A few small fragments, he stated, behaved like feldspar, 

 sometimes with, sometimes without twin striping. 



Brezina 22 distinguished three irons in 1885 Canada de Hiero, Carleton-Tucson and 

 Tucson- Ainsa. The first, on account of observed hexahedral cleavage, was included with the 

 hexahedrites ; the two latter were, on the contrary, placed in different divisions of the ataxites, 

 "since they indeed showed a certain similarity, but not sufficient, without further investiga- 

 tion, to class them together." In the case of Carleton-Tucson, however, it was pointed out 

 that it showed large flakes separated by fine, crumpled veins of schreibersite. 



In 1890 Fletcher 23 collected all former investigations of importance, instituted a critical 

 examination of the credibility of- the d#ta concerning the place of discovery, and compared the 

 structure of the two masses. He came to the following conclusions: 



The true locality is the Puerto de los Muchachos, lying between Tucson and Tubac in the Sierra de la Madera. 

 Both masses originated from the same locality, belonged to one and the same fall and, according to authentic specimens 

 in the British Museum, agree in all essential characteristics. The silicate grains are of the same dimensions in bpth 

 (for the most part 0.1 to 0.2, occasionally 1 mm. in size) and of the same appearance; they are mostly roundish and 

 irregularly distributed, although in certain places they are elongated in form and arranged in parallel, somewhat bent 

 rows; etching produces no Widmannstatten figures, but an irregular network of yellow lines resembling taenite or 

 schreibersite, and each inclusion is bordered in the same way. It is not certain to which mass the analyses of Shepard, 

 Genth, and Smith belong. From the two latter, along with that of Brush, I estimated the following composition, for 

 the nickel-iron and olivine: 



Nickel-iron. Olivine. 



Fe Ni Co Cu FeO MgB CaO A1 2 3 NajO K 2 O SiO 2 



89.89 9.58 0.49 0.04 =100 24.07 27.37 8.67 Trace 2.15 1.26 36.43 =99.95 



From the above analyses I derived the following mineralogical composition: 



Smith. Oenth. Brush. 



Nickel-iron 90. 64 90. 03 86. 24 



Olivine 8.29 8.60 10.05 



Schreibersite 0. 77 0. 64 3. 18 



Chromite 0.30 0.73 0.53 



100. 00 100. 00 100. 00 



Fletcher 23 classed this iron among the siderolites, which form the connecting link between 

 the meteoric irons and the meteoric stones, and included it, like Wadsworth, with the pallasites. 

 In 1895 Brezina, 26 upon the ground of Fletcher's investigation, united the two Tucson 

 blocks under the name "Muchachos" and formed of them the Tucson group, a subdivision of 

 the ataxites, for which the. characteristic is "iron fragments separated by schreibersite veins." 

 The individual grains with their half-shaded markings were said to contain oriented rhabdite 

 lamellae, so that Brezina remarked that the iron might be reckoned as a brecciated Chesterville 

 iron. In regard to Canada de Hiero it was merely stated in the table of contents that it resem- 

 bled Muchachos; also an iron obtained from Jackson with the label "La Concepcion in Chi- 

 huahua " was identified with Muchachos. 



Cohen 25 . 28 . 29 . made a careful study of specimens from the two irons with results as follows: 



The polished sv:rfaces, with the exception of the stony inclusions, make an entirely homogeneoi;s impression. 

 After weak etching, however, the nickel-iron appears as irregular masses which are distinctly separated from one 

 another by very fine, glistening, zigzag seams. In Carleton they are somewhat broader and on that account show more 

 distinctly than in Ainsa; in the former iron almost all the silicate grains too are surrounded by similar seams averaging 

 0.01 mm. in width, while this ia the case with only a few of the Ainsa grains. These seams may be characterized as 

 teenitic or schreibersitic; Rose calls them simply teenite, and Brezina schreibersite. By experiment with a fine needle 

 under the microscope they appear to be ductile, so that it is apparently teenite or an alloy richer in nickel than the 

 ordinary nickel iron, but not so rich in nickel and consequently not so resistant to HC1 as normal teenite. At all events 

 upon treating portions of both irons with dilute HC1 or copper ammonium chloride no scales or lamelke remain undis- 

 solved, though it is to be observed that tsenite as well as schreibersite in very fine fragments can be dissolved in con- 

 siderable quantity. 



The above-mentioned grains of nickel-iron are of varying size; in the pieces which I examined they fall between 

 0.2 and 2 cm. Each grain has its own luster, so that one part of the etched surface in a certain light appears light gray 

 and glistening and another part dark gray and dull. Under the microscope the nickel-iron within the composition 

 pieces has a speckled appearance, but the patches overlap so as to completely obliterate the outlines, and there is no 



