560 ME. E. E. L. DIXON AND DR. A. VAUGHAN ON [Nov. I9I I, 



Comparison: — Pr. costatus J. de C. Sow., ' Min. Conch.' vol. vi 

 (1829) pi. dlx, fig. 1, differs in possessing the following characters : — 



Form broad and flattened. Both umbonal and marginal areas are visible 

 in a 'back view,' and a vault is undeveloped. Coarse marginal ribs, 

 which are remarkably flattened. Strong development of spines on the 

 wing- ridges. Deep, narrow, median sinus. 



Evolution: — The strength of the reticulation of the umbonal 

 region, the thickening of the marginal ribs, and the development of 

 spine-ridges are characters which commonly indicate phylogenetic 

 old-age. It seems probable that Pr. sulcatus is derived from some 

 such form as the specimen in the Sowerby Collection which is doubt- 

 fully identified with Pr. antiquatus Sow., ' Min. Conch.'* vol. iv, 

 pi. cccxvii, fig. 5 ; whereas Pr. costatus has probably resulted, by 

 parallel development, from Pr. antiquatus Sow., Joe. cit. fig. I. 1 



Peoductus (Maeginifera) longispinus Sow., var. (PI. XLI, 

 figs. 2a&2b.) 



The predominant form of this species in D 2 _ 3 of the Oystermouth 

 Quarry agrees with Productus setosus Phill., 2 var. tissingtonensis 

 Sibly, Q. J. G. S. vol. lxiv (1908) p. 77 & pi. i, figs. 6 a-6 b. 



The most striking character of this variant — namely, the large 

 central and marginal rib in the pedicle-valve — is well shown in 

 Dr. Sibly 's figures. Two further points are illustrated in the plate 

 which accompanies the present paper : — 



PI. XLI, fig. 2 a — a partial cast of tbe interior of the brachial valve — 

 shows a marginal groove, representing a ridge surrounding the muscular 

 region ; this is the essential character of the genus Marginifera of 

 Waagen. 



PI. XLI, fig. 2b — a pedicle-valve — shows the sharply-raised, collar-like 

 termination of the cylindrical wing. 



(The structural exaggerations here noticed are probably no more 

 than indications of phylogenetic old-age, for they occur also in 

 other groups, as, for example, in Pr. concinnus.) 



This variety is remarkably widespread in the upper part of the 

 Dibwnoph.yl.Iwn Zone throughout the British Isles. 



1 I cannot distinguish fig. 1 from the figure of Pr. semireticulatus (Mart.) 

 in ' Petrificata Derbiensia' 1809, pi. xxxii. figs. 1 & 2. If, therefore, Pr. anti- 

 quatus Sow. is legitimately retained in addition to Pr. semireticulatus (Mart.), 

 it must denote such a form as that of Sowerby, pi. cccxvii, fig. 5. On the other 

 hand, if it be considered that Sowerby, by his query, rejected Martin's species 

 as insufficiently described and figured, and proposed the new specific name 

 antiquatus in its stead, fig. 1 becomes the holotype of Pr. antiquatus Sow., 

 and Pr. semireticulatus (Mart.) dies out. In my use of Pr. antiquatus Sow. 

 I have adopted the first-mentioned course. 



2 If Productus setosus Phil]. (' Geol. Yorks.' pt. ii, 1836, pi. viii, fig. 9) be 

 separated from Pr. longispinus Sow. (refigured in ' Davidson,' pi. xxxv, fig. 5), 

 the Oystermouth form must be regarded as a variant of the small globose type 

 of Sowei*by's species, rather than of Phillips's large and sulcate type. On the 

 other hand, Pr. setosus Phill., var. Phill. (op. cit. pi. viii, fig. 17) agrees closely 

 in convexity and ribbing with our variant. 



