20 
THE LOWEK LIAS OF KEYNSHAM. 
of nodules, it would seem impossible, at first sight, to draw 
a divisional line where we suggest. 
2. Am. obliquecostatus, though undoubtedly at a maxi- 
mum in this division, does also occur in the Bisulcatus 
beds below, and large nautili are as common here as in the 
beds below. 
Against this view and in favour of our suggested zonal 
division, we may cite the following facts : — 
1. We include in the same sub-zone both mutations of 
Am. semicostatus (whilst admitting, however, an unim- 
portant overlap into the Bisulcatus zone below). 
2. The lower and upper subdivisions are linked by the 
great abundance of Avicula inequivalvis and of a distinct 
variety of Rh. calcicosta which we term Rh. calcicosta var. 
semicostati {Av. inequivalvis also occurs in the Bisulcatus 
beds but very sparingly). 
3. There is a very striking lithological change at the top 
of the Bisulcatus zone wherever the shale is well developed, 
for it is remarkably fissile, whilst all the soft beds at lower 
levels are more correctly described as clays. 
The uppermost beds which contain abundant belemnites 
may be unhesitatingly included in the Arietes zone, rather 
than in the Oxynotus clays above, for, with the exception 
of a single fragment of an oxynoticerate ammonite, we have 
found these clays to be entirely unfossiliferous. 
We have not adopted Am. turner i as a zonal index for 
two reasons : (1) It would only cover the top layers of the 
upper subdivision, and even there its occurrence is limited 
to a very small thickness of rock. (2) It is common in 
only one of the exposures, where only badly preserved frag- 
ments are to be found. 
Am. planicosta, the earliest of the capricorn ammonites, 
which attain such pre-eminence in the lowest beds of the 
Middle Lias, is fairly common throughout the upper sub- 
