354 MR. T. F. SIBLY ON THE CARBONIFEROUS [Mayi9o6, 



common occurrence of Modioliform lamellibranchs, 1 its brachiopod- 

 fauna, which includes some prolific forms, has not yielded any 

 characteristic forms that could justify a separation from the 

 Lower Cleistopora-ZoiiQ. In the Mendip area, therefore, the lowest 

 beds of the sequence, which, in part, represent a particular phase 

 of sedimentation, are grouped with the Cleistopora-Zone, and 

 distinguished as a ' it/ocftoZa-phase.' 2 



Horizon a cannot be distinguished in the Mendip sequence. The 

 palseontological characters which, in the Bristol area, link Horizon a 

 with the Cleistopora-Zone and distinguish it from the Modiola-Zone, 

 are possessed by the Jfocfo'oZa-phase in the Mendip area. 



So far as my own observations are concerned, Gleistopora afT. 

 geometrica is a somewhat unsatisfactory zonal index in the Mendips. 

 This fossil has, however, one advantage, namely, that it has never 

 been recorded outside its own zone ; on the other hand, it is either 

 absent or very rare in K l} and is apparently never common in K 2 . 

 CamarotoecMa mitcheldeanensis would be a good zonal index in 

 the Mendips ; it is very abundant in K A , abundant in K 2 , and 

 common at Horizon /3, while apparently never common at any 

 higher level. However, with the above comment, it is perhaps 

 better for the present to retain Cleistopora as an index-fossil in 

 the Mendip area, in view of the almost-complete identity of the 

 zonal assemblage in the Mendip and Bristol areas ; for, in the 

 Bristol area, Cleistopora is often abundant in K 2 , while Camaro- 

 toschia mitcheldeanensis persists abundantly throughout the Lower 

 Zaphrentis-Zone. 3 



The characters of the bassus- and octoplicata-suhzones in the 

 Mendips are essentially those of the same subzones in the Bristol 

 area, but the following minor points of difference are noticeable : — 



(1) The earlier entrance of Spirifer aff. clathratus, Beticularia cf. reticulata, 



and Bhipidomella aff. Michelini in the Mendip sequence ; and the 

 greater abundance of Spirifer aff. clathratus throughout this zone in the 

 Mendips. 



(2) The comparative scarcity of Leptcena analoga throughout this zone in the 



Mendip area. 



(3) The occurrence of Atliyris (Actinoconchus) cf. lamellosa in the octoplicata- 



subzone of the Mendips. 



The ZAPssENTis-Zoiie. 



An important feature of the sequence in the Mendips, when 

 compared with that in the Bristol area, is the great expansion in 

 thickness of this zone. The total thickness of the Zaphrentis-ZiOTie, 

 estimated in the Burrington section, is about 800 feet, while that 

 of the same zone in the Avon section is rather less than 400 feet. 



1 So far as I am aware, no specimen of Modiola has yet been determined 

 from the Mendip area. 



2 Since the lamellibranchs in these lowest beds of the Mendip sequence are, 

 as yet, very imperfectly known, I retain the term ' Modiola-ph&se,' so as to 

 secure uniformity with the zonal scheme in the Bristol area. 



3 In the Mitcheldean district (Gloucestershire), CamarotoecMa mitcheldeanensis 

 persists abundantly up to Horizon y. 



