CLASSIFICATION BY DESCENT. 133 



A noticeable feature in Table III is that we can trace Grammoceras back 

 further than any of the others in Am. antiquus, Wright. Branch A and Branch 

 B are both represented in the Margaritatus-zone ; but there is a great and sudden 

 inrush of forms in the Upper Lias belonging to these two Branches ; and these 

 forms are the parents of the Inferior-Oolite species placed in A and B. Two of 

 them undergo modification — Lillia and Pseudolioceras, producing Ludwigia and 

 Hyperlioceras ; while only one (not taking account of the solitary Lillia sulcata), 

 namely, Lioceras, remains constant from the Upper Lias well into the Inferior Oolite. 

 The Ludtvigia-stem disappears with the Concavum-beds so far as our Inferior-Oolite 

 rocks are concerned; but its descendants — the Hectici — appear later, several 

 species being found in the Oxford Clay. I do not know if any species likely to be 

 descendants of Poecilomorphus or Lillia have been recorded from beds above the 

 Parkins oni-zone ; and I cannot at present undertake to say whether in the 

 Ganaliculati and the Trimarginati we should recognise the descendants of 

 Lioceras and Hyperlioceras, because I have no material to investigate. 



In Table III it cannot fail to be noticed how many gaps occur in the sequence 

 of the different genera. More especially is this noticeable below the Commune - 

 zone ; and again, an examination of the species in Table II reveals that the gaps 

 are really more numerous than would appear from an examination of the generic 

 table. In Grammoceras a very long gap, only about twice broken, exists between 

 Grammoceras antiquum in the Jamesoni-zone and Grammoceras in the Striatulum- 

 zone. In Branch A a species of Lillia occurs after an interval of four life-zones 

 have elapsed without one ; and the same number separates Poecilomorphus from 

 Am. Mercati. In Branch B the first and second appearance of Lioceras is separated 

 by a considerable space which has yielded no species of the genus. Perfection in 

 the geological record we cannot expect, but a constant advance is being made ; 

 and in this connection I may notice that Grammoceras in the Murchisonce-zone 

 and Poecilomorphus in the Saiizei-zone are new facts which tend towards this 

 end. 



However, with the material and the facts which have been thus got together, it 

 may be safe to venture a few surmises : 



1. The descent of the species comprising Branches A, B, C from a common 

 ancestor in the Lower Lias. 



2. That such ancestor had small circular, or subquadrangular, whorls with but 

 little inclusion, had furrows each side of the carina, a simple suture-line with the 

 siphonal lobe as long as or longer than the superior lateral, and ribs fairly straight 

 on the lateral area but slightly projected forwards on the ventral, — that, in fact, 

 it was an Arietites. 



The changes which have been taking place among the descendants — which 



