2l8 CALIFORNIA ACADEMY OF SCIENCES. [Proc. 3D Ser. 



the sides, but are strong on the umbilicus and on the 

 shoulders, giving a strong resemblance to Cosmoceras, 

 which increases as the stage advances. The septa at first 

 are comparatively simple, although ammonitic. The first 

 lateral lobe begins to lose its identity as a separate lobe, the 

 three divisions of this becoming independent. The outside 

 indentation of the external saddle also begins to assume the 

 proportions of an independent lobe. At the beginning of 

 the stage there are four external primary lateral lobes, and 

 three internal; but about the middle of the third coil, diam- 

 eter 4.50 mm., the indentation of the external saddle, and 

 the three divisions of the large primary lateral lobe assume 

 the proportions of independent lobes, their arrangement in 

 a curve showing their secondary nature. This is shown on 

 PL XXVII, figs. 10-13, where the septa are seen to be 

 remarkably hke those of Diaphorites Fucini, (PI. XXVIII, 

 fig. 7), although the shell has long since passed through 

 the stage resembling Diafhorites. This is a rather 

 unusual way for new elements to be added to the septa, 

 they usually coming in at the umbilicus. As a conse- 

 quence of this mode of addition the complexity does not 

 decrease from the abdomen towards the umbilicus, but 

 decreases both ways from the middle of the sides. It 

 would be hard for one that had seen only the adult to 

 believe that the first four lateral lobes had been devel- 

 oped out of one primary lobe. A somewhat similar 

 development has been observed by the writer in the 

 Pinacoceratid^. 



While the almost smooth sides, the strong abdominal 

 shoulder keels made up by growing together of tubercles 

 at the outer ends of the ribs, the narrow umbihcus, and the 

 high narrow whorl all show affinities with Cosmoceras, the 

 septa never reach, during this stage, the complexity of that 

 genus. This may be due either to the greater accel- 

 eration of development of the form and ornamentation of 

 the shell, or to the physical impossibility of having a cham- 

 ber-wall take on so many convolutions at the edge in so 

 small a space. Most probably the former explanation is 

 the correct one. A specimen of this stage is figured on 



