HANDBOOK OF THE METEORITE COLLECTIONS. 19 



quoted by Lockyer, believed each chondrule to have been an " inde- 

 pendent crystallized individual," a stranger in its host, and imbedded 

 like a shell in limestone. Tsch^rmak compared the chondrules to the 

 spherulitic forms occurring in the trachytic tuffs of Freudenthal, and 

 more especially to the olivine spherules of Kapfenstein and Feldbode 

 in Styria. These tuffs he thought to be due to trituration in the vol- 

 canic throat. Writing with especial reference to the Gopalpur stone, 

 he argued that it must be considered to have been a cooled mass, which 

 through friction was broken into powder, the more tenacious particles 

 remaining as kugels which were again gathered into a loose aggregate. 

 Keusch also considered the chondrules as developed in the Tysnes 

 meteorite due largely to attrition of consolidated particles, though 

 perhaps modified to some extent by the corrosive action of the iron. 

 He would account for the structure of the bronzite kugels, in which 

 the radial point lies without the periphery, by assuming that origi- 

 nally they all had a like conical form such as is common in radiating 

 nodular pyrite, the upper surface of the nodule forming the base of 

 the cone. When such were worn down by attrition the point would 

 naturally break away. Berwerth has also arrived at the conclusion 

 that the chondritic meteorites originate through the partial refusion 

 of meteoric tuffs. Brezina, and after him, Wadsworth, seem to have 

 considered the structure of meteorites in general, and incidentally 

 that of the chondrules, as due to hasty crystallization, a conclusion 

 which so far as it relates to certain types seems well founded. Still 

 other suggestions have been offered, as through condensation of vapor, 

 or the refusion of original garnets. Concerning the last, it may be 

 said that it merits no serious consideration. The views of the present 

 writer were presented in detail in his description of the stone of 

 Allegan, Michigan, in 1900, and it will be sufficient to repeat here the 

 substance of the matter there given. 



The general structure of stones of the Allegan type can be ac- 

 counted for only by regarding them as agglomerates of chondrules 

 imbedded in a fragmental groundmass or matrix, the materials for 

 which were derived from the trituration of other chondrules. One 

 fact which has always militated against a theory which would ac- 

 count for the peculiar structure of a meteorite of this type, on the 

 assumption of hasty crystallization, has been the absence of a glassy 

 base in any but the chondritic portions. Obviously if the stone were 

 a product of crystallization, in mass, the chondrules would be prod- 

 ucts of the earliest consolidation, and should, judged by the standard 

 of terrestrial petrography, be the most highly crystalline, while the 

 base in which they are imbedded might be glassy or crystalline, ac- 

 cording to conditions. As a matter of fact, the reverse is the case, 

 the chondrules being more or less glassy, or at least imperfectly 

 crystallized, as in the barred and fan-shaped forms, while the ground- 



