Yoi^. 6, 1920 
PETROLOGY: G. P. MERRILL 
467 
•one were to assume that the mode of formation of the concavity was the 
same in the two cases it will apparently be necessary to further assume 
that the chondritic forms, the fused drops of "fiery rain," accumulated 
in such quantities before all were fully indurated as by their weight to 
allow those most firm to indent those still plastic. Were it not for the 
rarity of the double or excrescent bearing forms referred to elsewhere, it 
might be suggested that the concavities were due to the breaking apart, 
one from another, after solidification. It should be added that I have 
not yet found any evidence even suggesting that the enstatite kugels with 
radiate structure are basic segregations out of a more acid magma as are 
the metallic forms out of the slag, nor indeed, to indicate that they result 
from the disintegration of previously solidified magmas of any kind. 
II. With these facts in mind it seemed advisable to study the isolated 
chondrule, freed from its matrix. F'or this purpose the loose textured 
Allegan and Bjurbole stones, of pronounced tuffaceous type, afforded 
most excellent material, both in quality and quantity. From the dis- 
integrated Allegan stone there were picked out a considerable series of the 
rounded granules such as are ordinarily classed as chondrules. These, 
under a hand lens, it was found, could be readily divided into two classes. 
First, those of regular spherical or oval form with comparatively smooth 
though lusterless surfaces, usually with one or more depressions as in figure 
1, pi. 1, and sometimes with deeper pittings. These were of a light green 
color, in a few instances, but mainly dark, nearly black. When broken, 
such show a straight smooth fracture and distinct eccentric radial in- 
ternal structure (fig. 2, pi. 1). Of a number of thin sections made none 
showed a holocrystalline or porphyritic structure. Some were so finely 
cryptocrystalline that it was only by analogy that they were determined 
as pyroxenic. Others were distinctly eccentric radiating, the bars possessing 
sufficient crystalline development to make it possible to determine their 
orthorhombic pyroxenic nature. 
Second, the second series was composed of rounded and irregular gran- 
ules with rougher surfaces, lacking the shell-like rind or casing of the first 
(Fig. 3, pi. 1). They w^re hard, undergoing no further disintegration 
under ordinary friction, sometimes quite spherical, but grading into angu- 
lar fragments with angles blunted or rounded and with an uneven frac- 
ture. Sections of these showed a quite different structure from the first 
mentioned. Some were wholly crystalline, granular or of short radiating 
enstatites, while others were beautifully porphyritic consisting of olivine 
or pyroxene in a glass base. The porphyritic marginal crystals showed 
plainly that the present form of the granule was due to mechanical at- 
trition (fig. 4, pi. 1). These forms are undoubtedly the ones that in the 
thin section give the forms shown in figures 8, p. 454, and 10 on p. 455. 
There can be no question but what they owe their present form and 
dimensions to attrition as suggested by Tschermak, though whether the 
