﻿364 



closed to within a certain distance from the living chamber by a 

 series of partitions occurring at regular intervals. These forms I 

 shall describe under the name of Diphragmoceras in the Proceed- 

 ings of the Boston Society of Natural History, and I shall endeavor 

 to show that this genus is one of the distal ancestors of the Nautil- 

 oidea. This conclusion is based largely upon comparison with the 

 apical, metanepionic substage of development in the shell of the 

 modern Nautilus. The first septum of the shells has appended to 

 it a closed caecum or bag, the metanepionic representative of the 

 siphuncle, and the second septum is prolonged apically into a closed 

 tube, the end of which fits into this bag and usually lines it with a 

 second or internal layer. In some cases (Fig. 13, p. 363), probably 

 through the displacement of the second septum, this closed termi- 

 nation is carried forward and is then clearly seen to be a closed 

 tube extending into the siphuncle. The bottom of this tube, in 

 fact, forms a septum in the siphuncle, and the resemblance of this 

 early stage to the adult structures of Diphragmoceras becomes per- 

 fectly clear. Diphragmoceras had a closed tubular prolongation of 

 the base of the mantle like that of the metanepionic septa of Nauti- 

 lus and also more remotely similar to that which occurs in Endo- 

 ceratidae. But it does not diminish in size towards the apex, hang- 

 ing like a cone in the middle of the siphuncle ; nor does it, as in 

 that genus, also fill the siphuncle below its own extremity with a 

 continuous mass of calcareous matter having a cone in cone struct- 

 ure, nor has it any endosiphuncle. The sheath fits the siphuncle 

 closely and rises step by step with the body, its end forming septa 

 across the siphuncle at the resting stages of this process correspond- 

 ing in number to those of the shell, but not corresponding in posi- 

 tion, each septum being situated just in the interval between two 

 septa, or opposite each air chamber of the conch. Thus the siphun- 

 cle becomes divided into air chambers like those of the surround- 

 ing shell, but these partitions are not pierced by any endosiphuncle, 

 as are the endocones formed by the sheath in the Endoceratidae and 

 the solid deposits and peculiar rosettes of the Actinoceratidae.* 



Dr. Charles E. Beecher has been fortunately able to lay hands 

 upon the primitive radical of all of the Brachiopoda through the 

 study of the early stages of the shell and has shown that the common 

 embryonic shell or pro teg u lum of recent and fossil Brachiopoda is 

 represented by one of the earliest occurring forms, Paterina. Dr. 



* " Genera of Fossil Cephalopo&s," Proc. Bost. Soc. NaL History, xxii, 133 5, p. 272. 



