G4 BULLETIN OF THE 



younger septum, into which it closely fits and thus forms the outer wall 

 of the siphon. Internally is found the corneous layer, which is con- 

 tinuous throughout. 



In Nautilus and Goniatites, and in the youngest stages of Ammonites, 

 we find only the parts described above, but in the further development 

 of Ammonites, another is added. At first only a portion, and lastly, 

 nearly the entire thickness of the septum bends in the opposite or an- 

 terior direction, forming a loose collar around the siphon, but not built 

 into it, as is the shorter funnel below. 



The siphonal coecum is close to the abdomen in Goniatites and Am- 

 monites, but the siphon, which springs from the neck of the coecum as 

 it passes through the second septum, diverges nearer to the centre of 

 the whorl. 



This position of the. siphon, its structure, more especially in the 

 young Nautilus and the prevalence of shallow concave septa, the Nau- 

 tilian character of the fust septum in Goniatites and Ammonites even 

 in the closely coiled young, the prevalence of simply arcuate forms in 

 the young of several Devonian, and the straightened first whorl of the 

 varieties of two Silurian species, the continuity of the layers and their 

 thickness on the dorsal side of the young, are all characteristics which 

 appear to converge towards the structure and form of the siphon and 

 shell in the Vaginati, especially Endoceras (Orthoceras) duplex of De 

 Verneuil and Keyserling. 



It is in this group, therefore, or in some closely associated genus that 

 we must look for the ancestors of the Tetrabranchiate Cephalopods. 



I select Endoceras rather than Beatricea, which I have always 

 regarded as the prototype of all the Nautiloids, because there still 

 remain several characteristics in the structure of this form which must 

 be more thoroughly worked up, before the affinities can be settled 

 beyond a doubt. 



The structure of the septa, and of the siphon in Endoceras is not so 

 irregular or vesicular as those of the shell in Beatricea, but there is a 

 wonderful resemblance of the cup-like internal chambers of Beatricea 

 to a line of siphonal cceca. The young of Endoceras, if this view 

 obtain, would be expected to resemble, or to show some approximation 

 to, Beatricea. It ought either to be alone composed of thick conical 

 septa of the adult siphon, and each of these terminating at the apex in 

 a closed sac, or, at least, the siphon should be larger proportionately in 



