THE GEOLOGICAL WORK OF THE AIR. 357 



posed ; but some of the larger ones are very regular, and some of 

 the smaller ones are more or less pyriform. Many of these are 

 traversed by fine perlitic cracks. Together with the pellets are 

 numbers of the vitreous scales which we have just mentioned. 

 Some are perfectly smooth and uniformly transparent; while 

 others contain bubbles and blisters, and have a rough surface. 

 Sometimes innumerable little vitreous pellets are massed upon 

 them, similar to those which form bunches like bunches of grapes 

 on some of the finer hairs. 



I do not think that the mixture of pellets with the slag 

 threads spun accidentally near the nozzles of blowers in factories 

 has been insisted upon. These little balls result from a spe- 

 cial action of a gaseous medium ; and it is important to remark 

 that we can by their presence recognize the wind origin of the 

 deposits that contain them. They have' a bearing upon some 

 observations of M. Gaston Tissandier, who has found similar 

 globules in atmospheric sediments in very different localities.* 

 These spherules have evidently originated from the action of the 

 air upon the fluid matter developed upon the surface of meteorites 

 during their passage through the atmosphere. They should be 

 formed in considerable numbers at every fall of meteors, and their 

 small volume is favorable to their remaining suspended in the air 

 for a long time, and to their being carried by the wind to consid- 

 erable distances. This explains the quantities of them found in 

 the ocean bottoms. All marine sediments that have been care- 

 fully examined yield globules of this kind ; and, as is shown by 

 the common researches of M. Tissandier and myself, they were 

 equally abundant in the ancient geological seas. Thus, to cite an 

 example that has struck us very forcibly, the green sand extracted 

 from the artesian well of Passy, at five hundred and sixty-nine 

 metres below the surface, and which is of the Albian (lower Cre- 

 taceous, Potomac group), is full of spherules as perfect as those 

 which are extracted from the dust that has accumulated in the 

 towers of the Cathedral of Paris. 



The mechanism of the production of these globules is rendered 

 very evident by their abundance in certain industrial residues, 

 especially in the iron oxide produced by hammering, and in the 

 product of the combustion of iron in oxygen. Evidently the 

 melted oxide spread into laminas by mechanical projection goes 

 through the same capillary action as gives their form to soap- 

 bubbles. We can study all the details of the phenomenon with 

 greater facility if we have recourse to substances much easier to 

 melt than oxide of iron. I have had occasion to do this with 

 the experimental products which M. Daubrde has asked me to 



*Les PoussiSres de 1'Air. By Gaston Tissandier. Paris, 1877. 



