74 GENESIS OF THE AKIETID.E. 



shape, a fact accordant with the more specialized structure and more acceler- 

 ated development of the species in this genus. 



In Arnioceras 1 the goniatitic helmet shape was replaced by a purely psiloceran 

 helmet shape on the third whorl ; and this was retained throughout life in Am. 

 miseraUle? but lasted for a more limited period in Am. semicoslatum, 3 and was then 

 followed by a natter and broader abdomen, the sides becoming slightly divergent, 

 as in Am. obtusiforme ;* and this condition was often maintained throughout the 

 adult stage. 



The broad abdomen and divergent-sided whorl, which came out in only a few 

 species of Vermiceras and Arnioceras, and was not very strongly marked in them, 

 became in Coroniceras characteristic of the young at an early stage. 5 It is a 

 significant fact favoring our theory, that in the arnioceran-like forms of Cor. 

 kridion it did not replace the more compressed whorls of the arnioceran ancestor 

 until a late stage of growth. In other species of Coroniceras, however, the 

 broad abdomen and divergent-sided whorl replaced the laterally compressed, 

 helmet-shaped whorl inherited from Arnioceras, as in Cor. latum? All the species 

 of Coroniceras did not have this stage. It was in its turn more or less replaced, 

 in some of them, by the acceleration of other characters, as will be shown 

 farther on. 



The law of succession in anagenesis, therefore, is, that progressive species in each sepa- 

 rate genetic series were the direct descendants of progressive varieties or forms. The facts 

 consequently are in strict accord with the theory of descent with modification, and with the 

 law of heredity, that like tends to reproduce like. 



Coroniceras was not derived from Am. semicostaium directly, but indirectly, 

 through the more highly specialized forms of Am. kridioides and Cor. kridion. It 

 was not the varieties of Cor. kridion with arnioceran characteristics most com- 

 pletely developed which led into Cor. rot/forme, but those with divergent-sided 

 and highly specialized pilae, keel, and channels. So in Cor. rot/forme with refer- 

 ence to Cor. latum, and also in this last with reference to Cor. Bucklandi. These 

 are the purely progressive forms ; and their connection with ancestral species 

 occurred through progressive varieties. 



Catagenesis, 7 or the Genesis op Retrogressive Characters. 



Many large specimens of the species noted in the preceding remarks had 

 narrow abdomens, and the sides converged outwardly. Thus, in what is often 

 mistaken for the full grown adult stage of Caloceras an acute helmet shape 

 appeared, as in some varieties of Cat. Johnston/, tortile Liasicum, and nodotiamm. 

 This was certainly not, as usually stated by paleontologists, due to a retention of 

 the psiloceran form. It took place after the intermediate or progressive stages 

 in which the abdomen had become widened, more or less flattened, and the sides 



> Embry. Ceph., pi. ii. tig. 8, 9. s P1 . ;; fig , 4 _ 7 



8 Compare above with semicostatum, pi. ii. fig. 10 and 15. * P] ii. fig 8. 



* PI. iii. fig. •_>•_> a ; pi. iv. fig. i; p i. vi . fi g 6 . 6 p," iu fig 20 _ 



" Kurii, downwards; r«W«, descent bv birth. 



