94 BUREAU OF AMERICAN ETHNOLOGY [bdll. 52 



identity conclusively. Under the title "tierra cocida" specimen 

 263710 has been included. It is, however, simply a loess rich in 

 limonite, which forms the cementing material and renders the frag- 

 ments slightly harder and more resistant than the adjacent loess. 



Specimen 263731 has also been labelled " tierra cocida," but it 

 belongs strictly to the melted scoriae. 



THE SCORIA 



These rocks are the most interesting specimens in the collection 

 and have been the subject of much discussion in the literature. 

 They are abnormal in type and do not agree with any known eruptive 

 rock or lava in their microscopic features. They contain fragments 

 of various minerals (quartz, plagioclase, pyroxene, magnetite) 

 embedded in a streaky glass base, out of which pyroxene microlites, 

 different in composition from the large pyroxene fragments, have 

 crystallized at certain points. The composition of the plagioclase 

 fragments is not constant and in extreme cases may pass in the same 

 thin section from oligoclase in one grain to labradorite in an adjacent 

 grain. Zonal structure is present in some of the grains and tliese 

 may then be broken across so that the different zones are in contact 

 with the glass. The feldspars show indications, moreover, of resorp- 

 tion by the molten glass base and were evidently not crystallized 

 from it as they are in a normal lava. (PL 7, a and pi. 8, a.) Prac- 

 tically the only microlites wliich have crystallized from the glass 

 have the optical properties of a pyroxene which is of a different 

 color, refractive index, and composition from the larger fragmental 

 pyroxenes. The glass base itself is streaky and variable in compo- 

 sition, as is shown by the variation in its refractive index from point 

 to point. (n= 1.51to 1.56 and above.) Variations of this magnitude 

 in the glass base of a normal lava have not been recorded. The 

 minerals occurring as fragments in the scoriae are practically identical 

 with those observed in the loess, both as to kind, size, and compo- 

 sition. The melted loess products obtained by heating charges in 

 the electric furnace resemble the scoriae in all details except for the 

 degree of oxidation of the iron; with large masses products were 

 obtained in wliich the iron was less oxidized and at the center of 

 large lumps were practically identical with the scoriae. The scoriae 

 are not, in general, of the same composition as the loess immediately 

 adjoining them and behave differently on being heated to high tem- 

 peratures. (Specimen 263721.) This proves that part of the scoriae 

 at least were not formed in situ from the adjacent loess but have been 

 transported to their present position. No zone of transition between 

 the scoriae and the loess was observed ; in view of the high temperature 

 to wliich the scoriae must have been heated (1,050° or above) and the 

 ease with which the loess is reddened (at 800° and above), this unaltered 



