Vol. 62.] LIAS3IC DENTALIID33. 591 



that Moore himself detected and would have rectified this error had 

 he and Tate been spared to do so. Attached to the tablet which 

 bears the specimens in the Bath Museum are certain lithographs 

 representing the true nature of the fragments. I now understand 

 that these lithographs were taken from one of a number of plates 

 drawn to illustrate a monograph of the ' British Liassic Gastero- 

 pods,' by Moore & Tate, for the Palaeontographical Society. But 

 neither these authors nor Edward Wilson, who had also set 

 himself the same task, lived to see its accomplishment. The 

 figures of Dentalium trigonale, which were to have appeared in 

 Moore & Tate's monograph, are reproduced (by kind permission 

 of the late Prof. Blake, in whose possession were the plates referred 

 to) in PL XLV, figs. 5 a-5 d. They leave no doubt as to the specific 

 position of the Gloucestor Gasworks forms. 



While it is easy to separate the trianguloid types of Dentalia. 

 when the actual specimens are available for examination, there is a 

 considerable difficulty in dealing with synonymy, and the question 

 as to whether Dentalium compression, d'Orb., is really a Dentalium, 

 and the fragmentary condition of the holotype of Dentalium 

 limatulum, Tate, do not simplify matters. The results arrived at 

 by students of the Dentaliidae in this country have been to refer 

 any Liassic Dentalium with a section suggestive of a triangle to 

 either D. limatulum, Tate, or D. compressum, d'Orbigny. By Tate, 

 Moore's species D. trigonale was associated with D. compressum , 

 d'Orb. ; and the authors of the list of fossils in the Geological- 

 Survey Memoir on the Lias (p. 352) went further than this : they 

 apparently regarded the species D. trigonale, D. compressum, and 

 D. gracile, Moore, as all synonymous with D. elongatum, Miinster. 

 Very much the same results were obtained on the Continent : all 

 specimens with a transverse section approaching the triangular 

 were relegated to d'Orbigny's species — hesitatingly, it is true, by 

 some authors. 



The history of the ' trianguloid ' types of Dentalium will be 

 found detailed in this section and in those on Dentalium limatulum, 

 D. subovatum, and D. subtrigonale. 



The ventral side is the thickest in Dentalium trigonale, but not 

 the broadest of the three as stated by Moore. It is flattened at the 

 anterior end for a space of about 2 millimetres therefrom, while 

 the remainder is concave. The sides are slightly convex, converging 

 to form a subacute or obtuse carina along the ventral margin, which 

 is thus not blunt (as is the case with the species D. subtrigonale). 

 Occasionally the extreme posterior portion becomes somewhat 

 carinated (but it is an obtuse carina) and the concavity of the 

 ventral portion gives rise to more or less prominent keels where 

 that side joins the other two. The relatively-thicker test of the 

 shell-fragments from Eadstock, when compared with those from 

 Gloucester, is easily accounted for by the lesser quantity of lime in 

 the water inhabited by the latter: at Gloucester the oxynotum- 

 armatum-heds are clay-deposits ; while at Padstock the deposit of 



