444 PALAEONTOLOGY OF ILLINOIS. 



trace of an internal cavity, nor of segmentation, bat the basal portion 

 is exceedingly coarsely fibrous in structure. 



Position and locality : In the fish-bed of the Upper Burlington lime- 

 stone ; Burlington, Iowa. 



Genus BYTHIACANTH US, St. J. and W. 



Fin-spines deeply imbedded, laterally more or less compressed, 

 exposed portion relatively limited and but moderately produced verti- 

 cally, but extending along the back in a recumbent position; line of 

 insertion very oblique and more or less coextensixe with the anterior 

 margin, which latter is arched posteriorly, terminating in an obtuse, 

 beak-like apex; lateral surfaces marked by more or less prominent, 

 conical, vertically striated tubercles, arranged in more or less regular 

 rows which obliquely descend from the anterior margin to the inferior 

 border, increasing below by impJanation. Posterior face relatively low 

 and vertically keeled. Pulp cavity forming a deep channel in the post- 

 erior side of the base and extending to the angle formed by rhe junction 

 of the posterior face with the shaft or base, possibly penetrating the 

 body of the spine towards the apex. 



The recumbent position of the exposed part of the ichthyodorulite 

 forms a striking feature of the forms for which we have proposed the 

 above generic designation, and which in connection with the associate 

 characters as described above present an ensemble of characteristics 

 which readily serve to distinguish it from allied genera. The species 

 thus far determined pertain to the Lower Carboniferous, probably 

 both belonging to the St. Louis limestone, that described in the present 

 work under the name By. Van Homei, and a similar form mentioned 

 by Dr. Letdt from the Lower Carboniferous of Tennessee under the 

 term Astracanthus siderius. 



The affinities of these ichthyodorulites with the Mesozoic AstracantJii 

 are made to appear more remote by the examination of the fine example 

 representing the species first referred to above than was the case with 

 the imperfect specimen possessed by Dr. Leidy, from which the real 

 form of the spine could not be determined. There are points of resem- 

 blance between these spines and the form described by Prof. Agassiz 

 under the name Gtenacanthus brevis, from the Carboniferous limestone 

 of England, a very recumbent form with vertically striated tubercles; 

 but the latter are apparently disposed in rows parallel with the anterior 

 margin instead of descending obliquely to the inferior border as in the 

 present examples. 



