i894. CEPHALOPOD BEGINNINGS. 429 



called Orthoceras (Fig. 2) is also not wholly convincing. In this case 

 the central position of the siphuncle is not positive evidence, for, as 

 shown by Branco, Hyatt, and others, the siphuncle of Ammonoidea 

 changes its position during growth. 3 It is certainly open to us to 

 suppose that an ancestral form may have had a central siphuncle, 

 and on a priori grounds one would expect such to have been the case. 

 Consequently this so-called Orthoceras may not be a true Orthoceras, 

 and may not be a nautiloid at all. If this were the case there would 

 be no grounds on which to base any argument. But, assuming for 

 the moment that Clarke is right in calling it Orthoceras, then it tends 

 to prove precisely what we anticipated, namely, that the ancestral 

 nautiloid had a protoconch ; but it certainly does not prove that 

 Bactrites was a nautiloid, any more than it proves that Mimoceras was 

 a nautiloid, which latter would be a reductio ad absurdnm. 



We pass now to Clarke's last paper, which he somewhat rashly 

 entitles " Nanno, a New Cephalopodan Type " (8). 



The specimens described were found in the Trenton series at 

 various localities in Minnesota. Their combined evidence enables us 

 to describe the complete shell as follows. The general outline of the 

 shell may be gathered from Fig. 3, c ; it appears to be assumed that 

 this represents almost the entire length of the shell, but the evidence 

 that such is the case is not satisfactorily brought out, and one would 

 imagine that the outer shell-wall might have extended further for- 

 ward than is shown in this drawing. The phragmacone, then, 

 consists of a relatively wide tube, suddenly swelling and again 

 diminishing towards the apex. The septa are concave and relatively 

 close together. They are penetrated by a relatively wide siphuncular 

 passage, which has an absolutely marginal position just as in the 

 Mesozoic belemnites. " The septa," says Clarke, " are abruptly 

 deflected immediately about the sipho." This may mean either that 

 they are bent upwards on that side towards the shell-aperture, or that 

 their margins are bent downwards around the siphuncle so as to 

 enclose it in a tube formed of septal necks. Either of these explana- 

 tions is consistent with the figures, and the latter agrees with what 

 we know of most of these primitive cephalopods ; it is, in fact, this 

 ' neck-tube ' that is usually, though wrongly, termed the siphuncle 

 by palaeontologists. Wrongly, because the siphuncle, as already 

 pointed out, is a fleshy extension of the soft body, and an extension 

 which, strictly speaking, contains none of the viscera. In the 

 present instance the width of the passage is so great that we may 

 well suppose it to have been occupied by a visceral cone, rather than 

 by a simple siphuncle ; and this supposition would explain the 

 apparent shallowness of the body-chamber in this form as also in 

 Piloceras (see Fig. 5, I). In figures 3, h and c, the siphuncular tube is 

 seen to be constricted where it passes through the septa, and this 



3 This also took place in Orthoceras itself, as lately shown by Foerste, but that 

 does not affect the argument. 



