Voi I. Grahau — Ordovician Fossils from North China (i) 61 



forms a reflex fold of a part of its mantle, that which builds the initial outer "collar ", 

 which eventually becomes the first outer camera. 



Whether this suggestion regarding the mode of building of this outer series of 

 camera} will be shown by future discoveries to be correct, or not, I believe that the very 

 existence of the initial precamerate stage of the "siphuncle " with its own shell- wall and 

 with conical, septa-like sheaths, prolonged into an endo-siphuncle, indicates that it is a 

 primitive shell-type M'hich was complete in itself, and tljat the development of outer 

 camerse at a later stage adds a new feature to the shell as a whole. The continuance of 

 the " siphuncular wall" or the shell of this primitive non-camerate organism, into the 

 camerate portion in primitive forms, further emphasizes the independence of this inner 

 structure. The homology of this inner shell with the shell of Orthoceras, and of the 

 sheaths with the septa of Orthoceras, would therefore seem to admit of little doubt. 

 Moreover, as we have learned to take ontogeny as an infallible guide to phylogeny, if 

 rightly applied, we are forced to conclude that the most primitive Holochoanites were 

 without the outer camerate portion, thus representing in their adult stages, the condition 

 seen in the young of Proterocameroceras, Nanno and others of this type of structure. That 

 the existence of such has not been absolutely determined, signifies little, for our know- 

 ledge of the earliest cephalopod faunas is still very meager. As noted in the description of 

 ChiJdioceras, it is possible that that genns was non-camerate, the annulations being mere- 

 ly surface "ornamentations" such as are found in the shells of many later cephalopods. 

 Certainly there is no indication whatever in tliese shells, of the presence of camortc, such 

 as in found in other annulated "siph uncles " in our rocks, where portions of the "septal 

 necks " still adhere to the shell of the " siphuncle ", though there is no other trace of 

 the cameras. " Furthermore the remarkable form of Chihlioceras, its oblique aperture, and 

 the long anterior ventral blade, are hardly consistent with the idea of the former 

 presence of camer*. Possibly the same holds true of the rapidly expanding "siphuncles" 

 of Piloceras platyveritmm Grabau, of our rocks (Plate IV figs. 11 and 12), for although 

 some specimens are annulated, these annulations show no trace of adhesion of "septal 

 necks", while the apical portion shown in Plate IV fig. 11 is entirely without such 

 annulations. It seems at least likely that this species too began \vith a non-camerate 

 portion, and if cameras were added, this took place only in the later stages, when the 

 "siphuncle " had become cyhndrical. 



If we now enquire into the genetic re': lionships between the Holochoanites and 

 the Orthochoanites, it would seem to be evident, that they can only represent divergent 

 branches from a common ancestral stock. I have elsewhere suggested * that this 



* Bull. Geol. Soc. America Vol. XXX, pp. 148, 149, 1919. 



