188G.] PROCEEDINGS OF UNITED STATES NATIONAL MUSEmi. 87 



aside as of not sufficient purity, but the tbird was almost pure colorless 

 glass with only rarely a stain from iron oxides. This had a specific 

 gravity 2.197, and yielded Professor Clarke 95.01 per cent, of silica, while 

 the sand gave but about 90 i)er cent. 



The results shown by the two analyses are peculiar, and at first 

 glance may seem difficult to account for. Had the lightning shown no 

 selective power the resultant glass would possess the same composition 

 as the sand in which it formed. Had it exercised such power one would 

 naturally expect those minerals which are, under ordinary conditions, 

 most fusible, i. e., the feldspars and iron oxides, to be first acted upon, 

 and hence that the glass would approach them in composition.* 



In the case in hand the reverse of this seems to have taken place, the 

 ordinarily infusible quartz having been most acted upon, while the other 

 constituents in large part escaped, t thus yielding a glass from 5.91 to 

 6.83 per cent, richer in silica and relatively poorer in potash, soda, lime, 

 iron, and alumina than the sand in which it formed. Conceding that 

 the results obtained are correct, and that the composition of the sand 

 examined is the same as when the lulguritt s were formed, they may, per- 

 haps, be accounted for as follows : When the lightning strikes a hetero- 

 geneous mass, as a bed of sand, the various grains or particles compos- 

 ing it will become unequally heated in proportion to their conducting 

 powers, those substances which are the best conductors escaping with 

 least injury while the poorer conductors present so strong a resistance 

 as to become heated evcM to the point of fusion, hence the composition 

 of the glass will depend upon the relative conductivity of the com- 

 ponents of the sand, regardless of their fusibility.| 



Accepting the above as correct, it follows as a legitimate conclusion 

 that the quartz grains composing the sand were poorer conductors of 

 the electric fluid than either the iron oxides or the feldspar. The sub- 

 ject of the relative conductivity of minerals has, however, been too lit- 

 tle investigated to afford reliable data for the confirmation or refuta- 

 tion of this. 



The fulgurites from which the second silica tests were made were 

 very thin walled and fragile, with scarcely a trace of the convolutions 

 present in the larger forms. These also increased slightly in size from 

 above downward, but grew correspondingly thinner and more fragile. 

 This lack of corrugation even in so frail tubes I was at first inclined to 



* "So far as oljservatious have yet been made upon the production of fulgurite by 

 tlie fusion of a heterogeneous rock it appears that the amount of melting experienced 

 by each ingredient depends chiefly upon its degree of fusibility." (Diller, op. cit.. 

 p. 258.) 



tSo at least it would appear to the writer, rather than as snggested by Mr. Abbott 

 and others, that certain of the more basic substances hati been volatilized by the ex- 

 treme heat engeuerated. . 



+ The extraordinarily l)rirf duriition of the Hash and eonscqnent heat would, it seems 

 to nie, render it extremely improbable- that any one mineral of C(unparative easy fusi- 

 bility served as flux and thus aided in reducing the more refractory, as suggested by 

 Wiehmann and Harting. 



