64 GANOID FISHES OF THE CARBONIFEROUS FORMATION. 



the body striated towards the anterior margin, and punctate posteriorly, while those 

 towards the tail become nearly smooth. The cranial roof-bones are for the most part 

 finely tuberculated ; those of the face striated, and very considerable variety occurs here 

 as to the closeness and prominence of the striation. 



Observations. — The name Palceoniscus Bobisoni was given by Hibbert, in 1835, to a 

 small palseoniscoid fish' from the Burdiehouse Limestone, and three allied forms were 

 soon afterwards described as distinct species by Agassiz, — namely, Palaoniscus striolahcs, 

 also from Burdiehouse ; and Ambli/pterus nemopterus and J. punctalus (pars), from 

 Wardie. In 1877 I transferred these to Elomclitliys of Giebel, but as the three type 

 specimens o{ punctatus belonged, in fact, to two distinct genera, I instituted for one of 

 them the new genus Gonatodus, retaining Agassiz's n'<xme punctatus for the type species; 

 while for the other two, which obviously belonged to Monicldhys, I proposed the specific 

 term intermedius. Then and subsequently I also described as new species three other 

 allied forms, namely, E. ovatus (1877), E. tetiuiserratus (1880), E. Dunsii (1880). 

 But as material increased I began to find that the characters which I had looked on as 

 diagnostic for the seven supposed species referred to above were no longer reliable, and 

 after puzzling for years over the matter I felt compelled in 1890 to unite them all into 

 one, a very variable one indeed, for which I adopted Agassiz's name nemopterus. Dr. 

 Smith Woodward, however, in his "Catalogue" (pt. ii, 1891) pointed out that the 

 name Eobisoni, of Hibbert, had priority over nemopterus of Agassiz, and also hinted that 

 the Agassizian species Bucklandi, which I had left out of the reunion, might well also be 

 included as adult Eobisoni. Still further increase of material has convinced me that this 

 is the most advisable course, and I have accordingly in the table given above likewise 

 placed the last-quoted specific name as a synonym of E. Eobisoni. 



With limited material it is no doubt easy enough to put together a set of extreme 

 forms which anyone might be tempted to adopt as distinct species, as I myself did at 

 the commencement of my investigations ; but the greater the number of specimens 

 examined, the more and more difficult does their accurate diff'erentiation become, until 

 we are finally obliged to give up the quest. For instance, the relative fineness or 

 coarseness of the fin-rays, and the relative distance of their transverse articulations, are 

 characters which are quite inconstant. As a rule the fin-rays are proportionally more 

 slender and more distantly articulated in young specimens, though sometimes this condi- 

 tion persists in examples of considerable size. Also no reliance can be placed on the 

 relative extent to which the scales are striated or punctate, or upon whether the punctate 

 area is nearly smooth or thickly covered with punctures. Unfortunately the relative size 

 of the scales and the number of rays in each fin are characters which can only be 

 accurately ascertained in exceptionally well-preserved fishes, as a certain amount of 

 distortion is common to the greater number of specimens as they occur in the rock. 



Geological Position. — ElonicMhjs Eobisoni as defined above is without exception the 

 most abundant of all the Palseoniscid fishes which are found in the Lower Carboniferous 



