GENERAL MORPHOLOGY. 7 



found ending in a more or less convex septum, which Sowerby noticed as long ago 

 as 1839 (loc. cit.), and which gave rise to the idea that the genus should be placed 

 in the Cephalopoda (PI. V, figs. 12 a, 14 a). The septum consists of a very 

 thin shell, which appears to be continuous with the inner layer of the test. It is 

 quite unornamented, the stria? which are often seen parallel with the longer axis 

 being probably due to compression (PI. V. fig. 7 h). The position at which it is 

 found varies considerably, for it is sometimes close to the apex, where it only 

 measures about 10 mm. across, while in other cases, in specimens of about the same 

 size, it has a length of 25 mm. or more. Probably, therefore, the apical part of 

 the shell is divided up by a series of septa, of which only the lowest is usually 

 seen. In the Birmingham University Museum, however, there is a specimen 

 which shows a second septum arching away about (> mm. above the first. 

 The existence of a siphuncle in these septa has been suggested at various 

 times by different writers, 1 but never with any great certainty. The septa 

 are frequently much broken, but I have seen several perfect ones. On onty 

 one of these (PI. V, fig. 1 /') have I observed anything like a siphuncle, and 

 here, though the small central tube is very well defined, it appears to be closed, 

 and may be an abortive survival of a once functional structure, or a scar left 

 by the growth of the septum. 



In all probability the chambers thus successively cut off at the apex remained 

 quite empty, and the apical ones may even have been lost in the lifetime of the 

 animal ; after this had taken place the apical end of the shell must have been 

 closed. But in some species, where the shell is exceptionally thin, septa appear to 

 be unknown. Possibly in all these cases— certainly in G. tenuis — the shell remained 

 open at the apex, and was fixed to some foreign body. 



(6) Orientation. — In total ignorance of the nature of the soft parts of the 

 animal, any distinction between dorsal and ventral sides must be merely an 

 arbitrary one. Salter writes, in 1855 2 : "If we consider, as it seems to me Ave 

 ought to do, that in this compressed species the two opposite angles of the flatter 

 sides are the dorsal and ventral ones, we shall then, I think, have a character by 

 which we may be able to trace these parts in the squarer species and in some 

 which are probably compressed laterally. For I find that the line where the usual 

 transverse ribs are bent or broken at about the middle of each lobe is not really in 

 the middle in all cases, but is nearer the dorsal and ventral angles than the lateral 

 ones ; and again, the two lobes which form the dorsal side are sometimes wider 

 than those two which form the ventral face." 



But this attempt at orientation seems to me to be of little value. The 

 " squarer species " are, in all cases which I have observed, symmetrical, and as the 



1 Eueclemann, ' American Geologist,' vol. xviii (1896), p. 65 ; Sowerby, Joe. cit. (1839) ; Hall, 

 ' Palseont, New York,' vol. i (1847), p. 222, pi. lix, fig. 4 e. 



2 Sedgwick, ' Synopsis of Brit. Pala;oz. Bocks,' Appendix A, p. v. 



