A. Smith Woodivard — On Poecilia from Oeningen. 393 



universally quoted iu textbooks aud treatises as a fact worthy of 

 special note. 



In his memoir Dr. "Winkler mentions that he founds the species in 

 question on a series of specimens, partly in the Teyler Museum, 

 partly in the collection of Van Breda. As the latter collection was 

 acquired by the British Museum in 1871, four of the specimens 

 studied by Winkler are destined to be included in the forthcoming 

 vol. iv of the "Catalogue of Fossil Fishes." They have thus been 

 recently examined during the preparation of this work ; and the 

 result is so completely at variance with AVinkler's interpretation of 

 the fossils, that it seems advisable to publish an amended description 

 of the fish without delay. 



Among the specimens in the Van Breda collection, it is easy to 

 recognize the original of Winkler's figure (op. cit., pi. iv, fig. 16), 

 which is preserved in counterpai't and numbered 42,779. The 

 figure, however, seems to have been made by an artist without 

 technical knowledge, and the description is evidently based upon 

 this drawing, not upon the actual fossil. The general proportions 

 of the fish are correctly indicated, and the imperfect remains of the 

 head show little more than the drawing ; but the fins are 

 incompletely represented and the figure gives a false impi-ession 

 of their characters and arrangement. The pectoral fin is relatively 

 large and comprises eleven slender spaced rays, which are divided 

 distally : it is spread over the flank of the fish. The pelvic fins, 

 not distinguished by Winkler, are relatively small ; and the base 

 of one, with indications of six rays, is observed directly beneath 

 the base of the pectorals. There are quite clearly two dorsal fins, 

 of which only the posterior is shown in the figure. The foremost 

 dorsal is the smaller, and consists of six spinous rays, each supported 

 by a dagger-shaped bone. The second dorsal comprises twelve rays, 

 all of which are probably articulated and divided distally. The 

 anal fin arises slightly further back than the second dorsal, and 

 is somewhat smaller than the latter, with only nine rays, none 

 spinous. The caudal fin is distinctly rounded, not forked. 



Most of the characters thus briefly noted are also observed in the 

 other specimens of the so-called Poecilia in the Van Breda collection. 

 Moreover, the small clustered teeth are shown in the mandible of 

 No. 42,778, while they appear among the remains of both jaws 

 in Nos. 42,780-81. It is evident that there are four, possibly five, 

 stout branchiostegal rays. There are also distinct indications of 

 scales, with radiating markings on their covered portion, at least 

 in the caudal region of the trunk. 



It is thus obvious that the so-called Poecilia oeningensis is neither 

 a member of the family Cyprinodontid^e nor a physostomous fish. 

 It is indeed a typical Acanthopterygian, and is most suggestive of 

 the families Cottid^e and Gobiidaj. In fact, if comparison be made 

 with the small fish from Oeningen described by Agassiz under the 

 name of Coitus brevis,^ duly allowing for imperfections in preserva- 

 tion, it will at once be perceived that there are no essential differences 

 1 L. Agassiz, " Poiss. Foss.," vol. iv, p. 185, pi. xxxii, figs. 2-4. 



