448 PHYSICAL GEOGRAPHY. 



take place, we can readily understand how we shall be very likely to find 

 diatoms growing together which are exactly alike, except that they vary 

 in size. But instead of the smallest being the oldest, the largest were 

 formed first, and, by the process of subdivision described, they have grad- 

 ually diminished in dimension. And it would seem that there was a limit 

 in each species beyond which the frustule did not diminish, but as soon 

 as it was reached, then the stage had arrived when reproduction sets in 

 in the manner to be presently described. Thus we see how one indi- 

 vidual diatom may in a very short time populate a large lake or river ; 

 but all of these separate cells, which have now become separate individ- 

 uals as well, will very closely resemble the first one from which they 

 sprung. But circumstances may occur, while this rapid growth is going 

 on, which may modify the characters of the diatom to such an extent that 

 very marked variation may result. Thus, for example, if the original 

 frustule existed at the head waters of a small stream, in perfectly fresh, 

 running water, some of its descendants may be carried down into a lake 

 where they may lodge along the shore, in still water, and thus become 

 modified, or, they may pass on into a large river, to be there affected, or 

 even carried down to its mouth, and there, where the salt and fresh waters 

 mingle, be changed by that circumstance. So, of course, other circum- 

 stances, which will readily present themselves to the mind, will serve to 

 form and perpetuate variation in the diatoms, until two frustules, de- 

 scended from the same progenitor, by growing under different circum- 

 stances, will appear so unlike that they may be classed as separate 

 species, or even as belonging to separate genera. 



The time occupied in a single act of self -division in the diatoms has 

 not been ascertained for all species, although it has been lately noted for 

 a few; "but supposing it to be completed in twenty -four hours, we should 

 have, as the progeny of a single frustule, the amazing number of one 

 thousand millions in a single month, a circumstance which will, in some 

 degree, explain the sudden, or, at least, rapid appearance of vast numbers 

 of these organisms in localities where they were but a short time pre- 

 viously either unrecognized, or only sparingly diffused." 



In all cases, however, the two newly formed frustules do not entirely 

 separate from each other, for, after subdivision has taken place, they 

 remain united, so that in time others will be added, and eventually a long 



