KENNARD : HOLOCENE NON-MARINE MOLLUSOA OP ENGLAND. 245 



burrowing animals, and great care must be used in dealing with the 

 contained mollusca. Caves, too, are often frequented by birds of 

 prey, and the crops of their victims, killed perhaps miles away, 

 would contain undigested shells and thus add to the confusion. 

 This is the only explanation of the presence of Sabinea uIvcb (Penn.) 

 in Chudleigh Cave, and of the same species and Zonitoides excavatus 

 (Aid.) at Nanna's Cave, Caldey. 



Ancient graves are of great importance as furnishing archaeological 

 evidence as to age, but in many instances the smaller burrowing 

 mammals have provided easy means of ingress for hibernating 

 mollusca, whilst the systematic rifling of grave mounds from Roman 

 times to the present has too often completely destroyed all reliable 

 evidence. 



Having thus briefly described the characteristics of the 

 varying deposits from which the Holocene mollusca have been 

 obtained, we will now examine the evidence which has been 

 accumulated. 



On comparing this with the admirable census recently published 

 as a memorial to the late W. D. Roebuck {Journ. of Conch., vol. xvi, 

 1921, pp. 165-211), many difierences will be noted. A few species 

 are unknown in a fossil state, three fossil forms have not been 

 detected living, some species have greatly extended their range, 

 whilst the area of distribution of others has apparently diminished. 

 In endeavouring to arrive at proper conclusions it is better to ignore 

 the slugs, for the determination of species from their scanty remains, 

 so easily overlooked except by the trained observer, is always very 

 difficult, and in the case of the Arionidse quite impossible, whilst 

 obvious accidental introductions must be similarly treated. 



Excluding these, there are only twenty-eight living species which 

 are unknown from the Pleistocene, and of these nine have never 

 been found fossil at all. These are : — 



Jacosta {Xerophila) neglecta (Drap.). 

 ,, (Xeroclivia) elegans (Gmel.). 



Fruticicola (Zenohiella) subrufescens (Mill.). 

 „ {Capillifera) odeca (Loc). 



Ewparypha pisana (MiilL). 



Margaritifera margaritifer (Linn.). 



Pseudanodonta rothomagensis, Loc. 



Pisidium tenuistriatum, Stelf. 

 „ steenhuchii (Moll.). 

 As already noted, the shell of Fruticicola (Zenobiella) subrufescens 

 (Mill.) contains so little lime that it is extremely unlikely to be 

 preserved after death, and this species is the only widely dis- 

 tributed form unknown in a fossil state. From its habits it is 

 improbable also that Margaritifera margaritifer (Linn.) would occur 

 in a fossil state, but the periostracum of this species has been detected 

 in the Holocene beds of the Clyde, the shell having completely 



