ORGANIC EVOLUTION' 353 



not sufficient. We must, if possible, bring the history of 

 past ages to bear upon the matter, and determine whether or 

 not there had been, with the lapse of time, any considerable 

 alteration in living forms. 



Evolutionary Series. — Fortunately, there are preserved 

 in the rocks the petrified remains of animals, showing their 

 history for many thousands of years, and we may use them 

 to test the question. It is plain that rocks of a lower level 

 were deposited before those that cover them, and we may 

 safely assume that the fossils have been preserved in their 

 proper chronological order. Now, we have in Slavonia some 

 fresh-water lakes that have been drying up from the tertiary 

 period. Throughout the ages, these waters were inhabited 

 by snails, and naturally the more ancient ones were the par- 

 ents of the later broods. As the animals died their shells 

 sank to the bottom and were covered by mud and debris, 

 and held there like currants in a pudding. In the course of 

 ages, by successive accumulations, these layers thickened 

 and were changed into rock, and by this means shells have 

 been preserved in their proper order of birth and life, the most 

 ancient at the bottom and the newest at the top. We can 

 sink a shaft or dig a trench, and collect the shells and arrange 

 them in proper order. 



Although the shells in the upper strata are descended from 

 those near the bottom, they are very different in appearance. 

 No one would hesitate to name them different species; in 

 fact, when collections were fir^t made, naturalists classified 

 these shells into six or eight different species. If, however, a 

 collection embracing shells from all levels is arranged in a 

 long row in proper order, a different light is thrown on the 

 matter; while those at the ends are unlike, yet if we begin 

 at one end and pass to the other we observe that the shells 

 all grade into one another by such slight changes that there 

 is no line showing where one kind leaves off and another 



