90 BUEEAU or AMERICAN ETHNOLOGY [bull. 52 



of oligoclase and labradorite can often be seen, the fragments of 

 both being homogenous throughout and without zonal texture. 

 Such rapid variation in composition in the nonzoned phenocrysts 

 of effusive rocks is unknown to the writers and would be abnormal 

 in view of the mode of formation of phenocrysts. Zonal structure 

 does occur in some of the grains and the composition follows then 

 the usual succession, with the more anorthic plagioclase in the 

 center of the grains. The zoned individuals are often broken across 

 so that the whole series of compositions was exposed to the action 

 of the melted glass. Taken as a whole the mineral fragments in the 

 scoriae resemble in kind, in size, and in general character and outline 

 the mineral fragments in the loess. In the specimens of scoriae from 

 Miramar, however, there is a decided difference in quantitative 

 mineral composition between the scoriae and the loess adjacent to 

 them. This difference also found expression in the thermal behavior 

 of the two materials as noted above under specimen 263731. 



The characl'ter of the glass base is another feature wliich is difficult 

 to reconcile with the view that the scorife are normal andesites. The 

 glass base is streaky and varies rapidly in composition, its refractive 

 index ranging from about 1.51 to 1.56. In the glass resulting from 

 the cooling of a normal lava such great differences have not been 

 described and are not to be expected. The presence of fine particles 

 of iron oxide embedded in a colorless glass base in some parts of the 

 scoriae and their absence in other parts, whereby the glass is colored 

 brown and has a higher refractive index, is strikingly like the 

 features observed in the loess specimens which had been heated at 

 different temperatures. 



It is interesting to note that the glass obtained b}'' heating the 

 loess and burnt earth specimens has about the same refractive 

 indices and variation in refractive indices as the glass of the scoria 

 from JVIiramar. In the glass of the Necochea scoria the refractive 

 indices are slightly higher than tliose of the Miramar glass and the 

 Necochea scoria is correspondingly less siliceous in composition. 

 The mineral fragments in the melted loess specimens are practically 

 identical in size and kind with the mineral fragments in the scoriae. 

 In both cases the fragments are unstable in tlie melted glass and have 

 been partly dissolved by it. The new microlites wliich crystallize 

 from the melt agree with pyroxene and have the same general aspect 

 in both cases. The products differ chiefly in the state of oxidation 

 of the iron, but tliis difference is evidently due, as noted above, to the 

 different conditions under which the products were formed, the 

 cQnditions of experiment not agreeing in all details \^^th those of 

 nature under which the scoriae were formed. 



CTiemical evidence. — The chemical evidence given in the analyses 

 263728 and 263746 does not of itself preclude a volcanic origin for 



