ATMOSPHERIC DUST. 37 



the coast of Norway, following- an eruption in Iceland. 

 The coarsest one forms the bottom layer in a (lejjosit in 

 Kansas, where the material settled in water. It contains 

 only the coarsest fragments, which first settled on tlie 

 bottom under the water. Tlie other sample from the 

 same plnce repi-esents more nearh' the average of the 

 same shower at that place. The dust from Golden in 

 Colorado is the second in coarseness. It fell nearer the 

 place of the eruption, which propably was somewhere in 

 Colorado. The Norway dust which Avas carried a dis- 

 tance of eight hundred miles, is of about average fineness, 

 compared with the other samples. The materials from 

 Nebraska and Nevada are finer. The variation in compo- 

 sition is quite remarkable in these samples, but it is 

 largely due to secondary sorting in watei*. 



On compar-ing the average of these analyses with that 

 of sand and dust taken in coaches, the latter are seen to 

 be slightly finer. This appears hardly possible, when we 

 think of the great distances the volcanic dust has been 

 carried, but there are three circumstances which combine 

 to keep the volcanic dust in suspension longer than any 

 other atmospheric sediments. Most of its particles are in 

 the form of flakes, tubes, or hollow bubbles. The flakes 

 may be twenty times as long- and as wide, as thick. Such 

 matei'ial floats easily in the air. Besides, other sediments 

 have first to be raised by the lower and weaker currents 

 of the air, as already pointed out, while the volcanic dust 

 is thrown up to great heights by an explosion. And then 

 the dust itself is about one fifth lighter than ordinary 

 quarts. We must hence infer that ordinai-y dust, which 

 is capable of being transported several hundred miles by 



