362 
dorsal saddles in place of dorsal lobes, as in the sutures of their 
nearest allies among the Nautilini and all of the remaining Ammo- 
noidea. AMimoceras ambigena Barr., of the Silurian (Figs. 7, 8, 
Pl. ii), is a close ally of this Devonian species, and with A@imoceras 
(Gon.) Ztuum (sp. Barr.) Hyatt (Figs. 40-42, Pl. viii), are the only 
ammonoids which are not involute nautilian in form. The whorls 
are in contact ; but there is no impressed zone, and no sutural lobes 
on the dorsum, as in true nautilian shells. On the contrary, they 
are purely gyroceran forms, with rounded dorsum and sutural sad- 
dles on this side in place of lobes. All of the Nautilinidz also have 
the septa concave, as in the Nautiloidea, in place of the invariably 
convex character of the septa in later Ammonoidea, as shown in 
Pl. x. As doubts may disturb the mind as to whether JZ. compres- 
sum is an ammonoid, we recommend a comparison of this shell 
with the young of an undoubted species of Goniatitinz, Agonia- 
tites fecundus of Barrande, which is a miniature copy made by her- 
edity (Figs. 9-11, Pl. ii). 
Bactrites is a perfectly straight form, similar to the members of 
the Goniatitinz in all important characteristics, especially the siph- 
uncle and septa, and it also has, like the young shell described by 
Clarke and all the coiled Ammonoidea, a comparatively large proto- 
conch, as demonstrated by Branco, whose figure has been repro- 
duced on Pl. 2 of this paper. This same genus includes straight 
cones like Bactrites (Orthoceras) pleurotomus Bar. (Syst. sil., PI. 
296), which are undeniably transitions: to true Orthoceras in their 
striz of growth and position of siphuncle. There is, therefore, 
convincing evidence in the structures of these Silurian shells that 
the Ammonoidea, with their distinct embryos, arose from the 
orthoceran stock, and passed through a series of forms, in times, 
perhaps, preceding the Silurian, which were parallel to those char- 
acteristic of a number of genetic series among Nautiloidea, viz., 
straight, arcuate, gyroceran, and nautilian. 
In Sczence (Vol. iii, No. 52, February, 1884, p. 127), an article 
written by the author closed with the following words: ‘‘ The study 
of the tetrabranchs teaches us that, when we first meet with relia- 
ble records of their existence, they are already a highly organized 
and very varied type, with many genera, and that there was a pro- 
tozoic period ; and the tetrabranchs, like their successors, certainly 
must have had ancestors which preceded and generated them in this 
period, but of which we are at present necessarily ignorant. What- 
