METALS IN THK ATMOSPHERE. 241 



oil a iiraiiiilar I'oi'inat ion of iiicteoritcs. which can exist as well as 

 an impalijahic powdci' (l+>atin<i;" in space as in the I'oi-uis of cona"h)ni- 

 erates of several hunch-ed kik)s weight. \ ery small grains which, 

 in ])assing thi-ongli tlu' atniosj)here, were heated, melted, and vola- 

 tilized, a])|)ear to us in the form of shooting stars; it is siip})<)sed 

 that mattei', not over a gram in weight, is sufficient to produce one 

 of these meteors. Searching foi* the dust of shooting stars, de 

 Keichenl)acii found, in l.S(i4, on the sunnnit of mountains of (Jermany 

 some ferruginous dust giving nickel and col)alt reactions. Ori the 

 hill of T>al)isl)erg, at an altitude of 1..500 feet, under the shelter of 

 the hecH'h forests, untouched hy ax or piclv. he found similar ti'aces 

 of nickel and coi)alt. Again, in a note of March 4, 1S12, Von Banm- 

 ha\"er published some observations on magnetic ])articles obtained 

 fi'om hailstones, citing ])articularly a hail storm lie observed at 

 Padua on the '2(')t\\ of August. lK',-\. After this meteoric period 

 of August and Septeml)er Phipson managed to collect some black 

 angular })articles which were neither carbonaceous nor coated with 

 soot, and which, dissolved in hydrochloi'ic acid, formed a perciilo- 

 ride of iron. Xordenskiold encountered in the snow collected on 

 icebergs some metallic |)articles about a (|uai"ter of a millimeter in 

 circumference containing metallic iron coated with carbonized oxide 

 and was at the same time able to determine the presence of nickel 

 a.nd i)hosphorus. 



After a heavy fall of snow at (xeneva on October 5, 1883, M. Jung 

 melted a ({uantity collected on the Saleve, and found therein a 

 deposit of powder excej)tionally rich in iron globules. 



Thei'e fell in 18S;5, al)out the annual period in November char- 

 acte]'iz(Ml by an abundance of shooting stars, a rain renuirkably 

 str(y!ig in metallic dust of cosmic oriu'in. These particularly abun- 

 dant globules of iron might have been produced l)y the bi-eaking 

 np of much larger meteorites into micr()sco})ical shooting stars. 

 From his Stockholm analysis of snow, in which, as in hail, he found 

 bits of ii'on, Xordenski(")ld satisfied himself that hail is condensed 

 around minute grains of cosmic matter floating in the air and fall- 

 ing impei'cej)tibly but continuously to the earth. lie regards the 

 existence of such material as pro\en by his observations and attrib- 

 utes to its fall a considerable importance not oidy from iho stand- 

 point of the geologist and physiograi)hist . but from that of the 

 f.ii'iner; this last on account of its i)hos[)h()rus, which with nickel and 

 cobalt is charactei"istic of meteoi-ic iron. 'Wt cite a single example: An 

 analysis of the meteoric iron found at Santa Calarina, Brazil, gave 

 the following components: Iron, iVAAV.) pei" cent; nickel, 3)5.97 })er 

 cent; cobalt, 1.4.S; with small (luantities of phosphorus, sulphur, 

 carbon, and silica. This iron, i-emarkable for its exceptional ([uan- 

 SM i!)04 n> 



