Vol. 69.] JURASSIC AMMONITES PROM JEBEL ZAGHUAN. 549 



d'Orbigny's normdnianum from the type, merely on account of 

 the section, which is quadrate in the former two and acute in the 

 latter — and I quite agree with him there. In my proposed classi- 

 fication they would also be united, assuming, of course, that there 

 is agreement in the radial curve of the three forms : since the 

 rather bad figures do not let this character come out very clearly. 

 Our belief in the close affinity of these forms receives a shock, 

 however, when we compare the sutures of, for instance, the 

 var. costicillatum of ' Grammoceras' normanianum (d'Orb.) 1 with 

 that of its subvariety detractum, and find that there are marked 

 differences. The former has an unsymmetrically-bitid external 

 lobe, and a wide external saddle with a quadrate, bifid, inner 

 branch. The second lateral saddle is also bifid, the first auxiliary 

 lobe wide and low. In the latter variety, on the other hand, the 

 lateral lobe is regularly trifid and the external saddle narrower, 

 with a slender, monophylloid, internal branch ; a similar, narrow, 

 monophylloid, second lateral saddle is present, and there is also 

 a strong forward inclination of the rest of the suture. 



All this clearly shows the need for a revision of these forms, 

 whenever more observations and material are at our disposal. 



It is scarcely necessary to add now that the classifications of 

 those Continental authors who consider the radial line variable, and 

 in the absence of suitable sutures base them on the outward shape 

 of the conch only, opposing forms with a carinatisulcate venter to 

 those of oval-lanceolate section and an angular periphery, seem to 

 me unjustifiable. A study of the inner whorls of some of the forms 

 belonging to the lineages here considered has shown the anagenetic 

 character of the periphery, from fastigate to carinate and sub- 

 sulcate and then to carinatisulcate ; and, if we were to be 

 guided solely by that criterion in our classification, we might 

 possibly refer several fragments of the same ammonite to three 

 different genera. Of course, as Wurttemberger 2 pointed out more 

 than thirty years ago, this difficulty applies equally to many other 

 groups now accepted universally ; and it could be argued that the 

 radial line also (or, indeed, even the suture) does not help us in the 

 distinction of the ill-characterized inner whorls. But it is main- 

 tained here that, although there are transitions in the radial line, 

 we shall generally be able to group together forms of close genetic 

 relations ; whereas, in. each of the several groups distinguished by 

 different radial curves, we find produced forms with, for example, 

 a carinatisulcate venter — they are only morphological equivalents, 

 and cannot be classed together. A striking example is afforded by 

 pi. ix of Fucini's work (already cited) on the Cephalopoda of the 

 Central Apennines. Here, on account of a common acute venter, 

 a number of forms are united as ' Grammocerates,' including 

 * Grammoceras' subtile Fuc. (fig. 10), probably a derivative of the 



1 A. Fucini, ' Amui. del Lias Medio dell' App. Centr.' p. 30 & figs. 29-30. 



2 ' Studien Liber die Stammosgeschichte der Ammoniten ' Darwinistische 

 Sclmften, No. 5, Leipzig, 1880. 



Q. J. G. S. No. 276. 2 o 



