554 mr. l. e. spath on [Dec. 1 9 13, 



show no excentrumbilication. The venter is narrower than shown 

 in the figure, has finer keels, and agrees better with that of the 

 type than the figure would lead one to believe. "Wright mentioned 

 that the history of the specimen was not known ; but, judging by 

 the matrix, he thought that it might be of Middle Liassie age. 

 The specimen is labelled 'Upper Lias, Somerset,' although the 

 authority for that statement cannot now be ascertained. If 

 the latter age were correct, we have here a striking example of 

 deceptive homoeomorphy ; but the different style of the ornament 

 in Hildoceras and the catagenetic character of the periphery in 

 Pseudogrammoceras from carinatisulcate to fastigate do not suggest 

 connexion of Wright's form with the Toarcian Hildoceratidse. 

 The suture-line is, unfortunately, not visible on the specimen, and 

 I was unable to break it up. 



Fucini's var. inseparable of ' Grammoceras ' normanianum 

 (d'Orb.), now more correctly named Protogrammoceras in- 

 separahile (Fuc.), has a section similar to that of my specimen 

 and not so thick as that of Pr. wrighti. The costation, moreover, 

 is coarser in the earlier part of the last whorl than that of the 

 other forms mentioned here ; and the radial line may be straighter 

 laterally, showing the form to be somewhat transitional. Fucini's 

 figure is not good enough to enable one to decide this. 



Grammoceras comacaldense (Tausch) var. bicicolce Bonar., is dis- 

 tinguished by a very small umbilicus (32 per cent, of the diameter) ; 

 and the same percentage is given by Meneghini for his ' Harpoceras ' 

 _pectinalum, which agrees (especially fig. 3) very well in section and 

 suture with my specimen, but is subfalciradiate. 



In Meneghini's variety iii of ' Harpoceras' hoscense Eeyn. 

 (= Grammoceras comaccddense mut. medolense of Bettoni) the 

 umbilicus is only 30 per cent., and the radial curve, owing to the 

 width of the side and a slight peripheral projection, assumes a 

 pseudofalcate character. 



Peotogeammoceras aft. cosxiciLLATUir (Fuc.) var. deteactttm Fuc. 



1900. Grammoceras normanianum (d'Orb.) var. costicillatum Fuc., forma 



detracta Fucini, 'Amm, del Lias Medio dell' Appeimino Centrale' 



Pal. Ital. vol. vi, pi. viii, figs. 2-3 & p. 30. 

 1900. Grammoceras portisi Fucini, ibid. pi. ix, tigs. 1-3 & p. 33. 

 1904. Hildoceras portisi Fucini ' Cefalop. Liass. del Mte. di Cetoua ' pt. 4, 



ibid. vol. x, pi. xx (xli) figs. 7-11 & p. 287. 

 1900. Grammoceras portisi var. gittelianum Fuc. ' Amm. del Lias Medio, &c.' 



ibid. vol. vi, pl.ix, fig. 4 & p. 35. 

 1900. Grammoceras portisi var. contrarium Fuc. ibid. pi. ix, fig. 5 & p. 36. 



This form is preserved only as an impression in the same bluish- 

 grey limestone as the other specimens. The ventral portion is 

 missing, and it is impossible, therefore, to say anything definite 

 about the peripher} 7 , except that the venter certainly seems to have 

 been fairly acute and the section compressed. The dimensions are 

 as follows : — 



Diameter 67 millimetres. 



Height of the last whorl 33 per cent, of the diameter. 



Umbilicus 40 per cent, of the diameter. 



