68 me. s. s. BUCKMAJsr o:n~ [vol. lxxvi r 



striking details in this connexion, which, however, must be held 

 over for the present. 



Nevertheless, it is surely rash to invoke a land-harrier as a 

 kind of deus ex machind, especially on little or no evidence. Thus 

 Mr. M. Odling, 1 speaking of the Yorkshire Cornhrash, says 



' That it is homologous to the South- country Cornbrash is fairly clear, but it 

 was probably deposited in an area cut off from the main mass by some land- 

 barrier.' 



Yet on the next page he gives a diagram of the same Cornhrash 

 zone of macroc&phcdus in Wiltshire and Yorkshire : quite correct, 

 but making a land-barrier seem superfluous. 



Farther on in the same paper, Mr. Odling remarks : 



' The absence of a particular ammonite need not necessarily mean the non- 

 deposition or contemporaneous erosion of beds of this particular age, but 

 merely that this form was not so rapidly distributed.' 2 



This brings us back to the horizontal range question again. As 

 the repetition of this idea evidently voices current geological 

 teaching — shows that teaching repeating phrases heard for the 

 last forty years — , it seems advisable to refer readers to certain 

 remarks about zone-species in my former paper, 3 and to add that the 

 absence of an ammonite from a locality is only the first observation 

 to make. This leads to the question — to which of the causes men- 

 tioned above as making lacunse may its absence be due '? This 

 should lead to further observations ; and, if by them one is satisfied 

 that nomenclature failure and certain other causes can be ruled 

 out, then comes the choice between stratal failure and dispersal 

 failure. To settle this, observations on geographical range are 

 required ; and, if the range is found to stop short at the area of 

 absence, then dispersal failure may be assumed, though it is not 

 necessarily proved. But, if the range extends well beyond the area 

 of absence, then dispersal failure would be an almost untenable 

 theory ; while stratal failure, which has already proved to be an 

 extremely common phenomenon, and seems likely to_be proved to 

 be yet much commoner, may reasonably be invoked as a cause of 

 the local absence. 



One pursues this method of work so much as a matter of instinct, 

 that the setting forth in detail of a course which appears so obvious 

 seems to require an apology. 



The analyses of Mr. Richardson's finds of ammonites and certain 

 brachiopods which are now presented (Table I, facing p. 70) should 

 illustrate these methods of work. 



(b) Faunal Data. 



For the preparation of Table I the only available data are 

 (1) the sequence of ammonites above the bronni clays, 4 which is 



1 XII, 1, p. 279. 2 Ibid. p. 283. 3 II, 8, p. 27S. 



4 Mr. Richardson assigns the bronni clays to a much earlier date. ' Hemera : 

 late raricoslati or early armati, probably the latter ' (p. 158, Explan. pi. ii, 

 figs. 27 & 28) : a position about 12 hemerse too early, according to the 

 researches of my former paper (II, 8, p. 266). 



