44 GENESIS OF THE AKIETID.E. 



We would also suggest that the phenomena of parallelism, or the evolution 

 of similar forms in different series which can be predicted, is obviously contrary 

 to any law which does not assume that adaptive characters are equally inherit- 

 able with differentials. Again, the minuteness and slight importance of some 

 differentials, like the collar of the siphuncle in Ammonoids, which are neverthe- 

 less persistent throughout several geologic periods and constant in every series, 

 do not appear to be accounted for by Weissmann's hypothesis. This collar 

 appeared first in the adults of the Goniatinae, and became engrafted in the early 

 stages by the law of acceleration, like other characteristics. It should be re- 

 membered also that parallel series were continually evolved, while this differential 

 remained comparatively unchanged, and that the alteration of the entire form of 

 the whorl by involution, and the evolution of complicated from simpler radical 

 sutures took place over and over again in different series arising from the same 

 stock. How could males and females have combined to produce similar series of 

 variations, and also unimportant but still persistent differences ? How could a 

 characteristic of slight importance to the life of the species have made a deep 

 and lasting impression on the ovum, while others of obviously greater moment 

 to the organism were transient in different series? 



There is one phase of the law of acceleration which requires to be dwelt 

 upon as the best means of conveying its £ull meaning to those not yet accus- 

 tomed to note its action in their researches. It expresses the mode by which 

 the continual replacement 1 of the older by newer and later acquired character- 

 istics takes place in every genetic series, and therefore explains the mechanism of 

 gradation, whether progressive or retrogressive. Changes in environment, which introduce 

 new adaptive characteristics in the neahgic or adult stages, necessarily add these to the 

 hereditary stages of the younger periods of growth, and thus shorten the development of the 

 latter by direct replacement. 



Heredity, as is well known, can continue for indefinite periods to reproduce 

 a useless part or organ, provided it does not interfere with the growth and 

 normal development of useful structures. When interference occurs, as is well 

 known to all physiologists, their resorption is a normal process, due to the fact 

 that they have become through disuse merely passive stores of food for the more 

 active worker cells. 



Though species sometimes pass through more than one horizon, they are, as a 

 rule, limited to the levels upon which they first appear. This fact, and the fre- 

 quent differences in the sediments, which often correspond to differences in the 

 faunas, indicate that the different varieties or species characterizing distinct, 

 horizons are more or less directly due to the changes in the surroundings, which 

 occurred in passing from one geologic level to another. We can fully understand 

 the phenomena of acceleration in development only when we begin by assum- 

 ing that the characteristics last introduced in the history of any type were more 

 suitable to the new conditions of life on the horizon of occurrence of the species, 

 than those which characterized the same stock in preceding horizons. These 

 characters would then necessarily, on account of their greater usefulness and 



1 Genesis of Planorbis at Steinheim, p. 2S. 



