JAMES PERRm SMITH 17 



primitive type that came between Bactrites and the coiled Goniatites, a 

 type that is unknown as a fossil, but one whose former existence is 

 indicated by the young stage of Mimoceras, PI. XIV, fig. 8, itself one of 

 the earliest and most primitive of Goniatites. 



Partial reversion is probably a common phenomenon among the 

 Ammonites, but outside of such striking cases as those mentioned above 

 it can l^e recognized only in the reappearance of the same character or 

 characters in later forms. This is possible only when the genetic series 

 is well known, which is but seldom the case. 



Beyrichites, in the Middle Triassic, PI. VIII, figs. 14-23, after becom- 

 ing rough shelled and ornamented, reverts to the flattened shape and 

 nearly smooth shell of its ancestor, Meekoceras, so much so that it has 

 several times been described as Meekoceras. Some species of Trachyceras 

 in the Upper Triassic, after going through the rough shelled stage char- 

 acteristic of that genus, become flattened and nearly smooth, and thus 

 show a partial reversion to the far removed parent Meekoceras, although 

 they are still progressive in the complex septation. 



Reversion by arrest of development is far more common than the 

 sort just described, but in this case, too, the reversion is only partial. 

 Metasibirites has already been mentioned as an example of this, where 

 there is a reappearance of the sculpture of Acrocliordiceras, and of the 

 form and septa of Pericycliis or Gastrioceras, an apparent palingenesis 

 of the long extinct genus Sibirites, but with some later characters that 

 Sibirites never had. 



The so-called "Ceratites" of the Cretaceous give us the classic 

 example of reversion by arrest of development. Although there were 

 no Goniatites after the Paleozoic, nor Ceratites after the Triassic, there 

 are in the Upper Cretaceous several genera with form and septa so like 

 those Paleozoic and Triassic groups that they were once called "Cera- 

 tites." We now know that they are not eases of generic persistence 

 through this long time, but are retarded and arrested forms, reverting 

 to goniatitic or ceratitic stages of growth after long obsolescence of 

 those characters, but with such a commingling of characters from various 

 steps in the family history that it is impossible to determine what was 

 the particular ancestor. One of these, Neolobites (PL X, fig. 1), although 

 a Cretaceous genus, is arrested in the Goniatitc stage. No adult Gonia- 

 tites are known in the immense interval between the Permian and the 

 Cretaceous. But also no genus is known in the Paleozoic that is com- 



