ATMOSPHEIilC DllfST. 5 i 



from west to east. But with only a single sample from 

 three of these places and only two from another, this evi- 

 dence is of little weight. Again, it seems quite certain 

 that part of the shower dust is local material. This is 

 indicated not only by the sand it contains, but also by 

 the color of the deposit. AVhen collected from regions 

 where there is a rich black soil, it is apt to be dark, and 

 when observed in the lee of sandy and less fertile lands, it 

 is brownish or reddish. It appears most probable, that 

 part of the shower dust comes directly from distant 

 places, while a part is picked up from the ground nearer 

 to the place where it falls, or from the surface of plants, 

 on which it has previously lodged. Som other obser- 

 vations indicate that there is a constant migration of 

 dnst particles in the lower part of the atmosphere. These 

 are apparently picked up and let down unceasingly by 

 the wind. Just what proportion of the deposits which 

 settle from this migrating dust at any particular place or 

 time, is local, and just how much of it comes directly fi'om 

 distant places, is difficult to say. That coarse dust is 

 capable of being transported long distances in the atmos- 

 phere can, however, under no circumstances be doubted. 

 It floats along in considerable quantities even on the 

 calmest days when the maximum hourly velocity of the 

 wind does not exceed ten or fifteen miles. It constitutes 

 from twenty-two to forty-four per cent of the totals of 

 the dust caught on such days (Nos. 102, 107, 108, 109, 

 110, 118), and the smallest proportion which it forms in 

 any of the fifty-seven samjjles of fine atmospheric sedi- 

 ments T have examined, is thirteen pei' cent (No. 136). It 

 seems safe to conclude that dust, which is present in such 



