ORIGIN AND CHARACTERISTICS OF SUBORDERS. 3 



The depressed semi-lunar whorl appears first in the adults of Anarcestes. It 

 is subsequently found in the young as a stage immediately succeeding the more 

 cylindrical whorl of the gyroceran stage, when that occurs. In very close-coiled 

 forms, the latter may be omitted, or be only slightly indicated, and then the 

 anarcestian whorl appeal's at the beginning of the apex. In fact, this tendency 

 in Latisellati, and especially in Angustisellati, affects the shape of the protoconch 

 which is excessively depressed in the embryos of the higher suborders. 



We have, therefore, considered it convenient to designate the anarces- 

 tian form of whorl as the primary radical of the Ammonoidea, reserving the 

 terms primitive and transitional radicals for the straight and gyroceran modi- 

 fications as they appear in Bactrites and Mimoceras. 



The different series of the Clymeninse and Goniatitinse, and the Arcestinse, 

 often begin with, and maintain persistently in full-grown shells, the primary 

 radical form. The Ceratitinje, Lytoceratinse, and Ammonitinse, on the contrary, 

 have this depressed form but rarely, except in their protoconchial stage, — 

 and at the beginning of the apex or true conch, while it remains in what we 

 have called the goniatitic stage of development. 



The Clymeninse of the Devonian begin, when zoologically arranged, with 

 discoidal forms having depressed semi-lunar anarcestian whorls. These de- 

 pressed whorls are exchanged in the higher forms for compressed discoidal 

 whorls, and these in turn for compressed involute whorls. The suborder 

 includes several genera and in each there occur examples of this mode of 

 succession, or rather procession, of forms, forming parallel series. 1 



The sutures of the genera Beneckia, Longobardites, Lecanites, Norites, 

 Meekoceras, Hungarites, and Carnites show them to be true Ceratitinse. We 

 should, with our present information, be disposed to include these, and all 

 the genera mentioned by Mojsisovics as belonging to his group of Ammonites 

 trachjostraca, in the Ceratitinae, distinguishing them by their well-known and 

 peculiar sutures from the Arcestinge, Ammonitinse, and Goniatitinse. 



The more or less compressed whorl, which in section can be described as 

 helmet-shaped, is the natural successor of the depressed anarcestian whorl both 

 in the growth of individuals and in the evolution of series of species. We have 

 considered this in the work quoted, therefore, as the secondary radical. 



The secondary radicals 2 are prevalent in the Ceratitinse, as shown by the 

 extensive researches of Mojsisovics in the remarkable and masterly treatise 

 above quoted. They completely replace the primary radicals as generators 

 of series in the Trias, except in the paleozoic survivors of the suborder 

 Arcestinna. So far as the sutures are concerned, however, the Ceratitinse, 

 though distinctly characteristic of the triassic faunae, are like the Goniatitinoa. 

 The young of Longobardites is really a Goniatite, similar to Prolecanites. 



1 Genera of Fossil Cephalopods, Proc. Bost. Soe. Nat. Hist., XXII. p. 312. 



2 We formerly included (Gen. Foss. Ceph., p. 321) in secondary radicals some quadragonal whorls like 

 those of the adults of Xenodiscus; but we are now disposed to consider this an error, arising from not hav- 

 ing observed that the young of these forms often possessed, during earlier stages of growth, the secondary 

 or helmet-shaped whorl. This evidence shows that, in the most ancient periods as well as in later times, 

 quadragonal whorls were derivative modifications of the compressed helmet-shaped secondary radicals. 



