CEPHALASPID^ 5 



Lankester (Scales oi Pteraspis). 'Quart. Journ. Geol. Soc.,' vol. xx, p. 194, 1863. 



„ (On the Genera Pteraspis, Cyathaspis and Scaphaspis). ' Brit. Assoc. Report,' Trans. 



Sect., p. 58, 1864. 

 ,, (On the Genus Cephalaspis) . 'Brit. Assoc. Report,' Trans., Sect., p. 65, 1865. 



„ (On Didymaspis). 'Geol. Magaz.,' April, vol. iv, p. 152, 1867. 



Besides these papers, there are several Letters on the Cephalaspidce in the pages of 

 the 'Geologist,' 1858 — 64, by Messrs. Powrie and Marstou, the Rev. Hugh Mitchell, 

 and myself. 



§ II. Nature of the Evidence toith regard to Cephalaspida now available. — -Though the 

 knowledge of these Fishes has been gradually increasing, as may be gathered from the 

 brief review above given of their literary history, and though we have now far more perfect 

 specimens to examine than Professor Agassiz originally had, yet, except in the case of 

 one or two species allied to Cephalaspis Lyellii, nothing whatever is known of the form 

 of the body, position of the fins, or arrangement of the scales.^ In every case with 

 the exception mentioned, our knowledge of the Fish rests solely, as did that of Professor 

 Agassiz, on more or less imperfect specimens of the oval, discoid, or semicircular shields 

 which covered the anterior portion of the animal. Owing to this fact, the modifications 

 in the form and structure of this cephalic shield are necessarily made the basis of the 

 arrangement and classification of these Fishes ; and, though those few specimens which 

 exhibit a scale-covered body, fins, &c., possess a very high interest, they cannot to any 

 large extent be made use of for the purpose of drawing conclusions with regard to the 

 other Cephalaspids. 



§ III. Division of the Cephalaspida into two Sections. — The genera Cephalaspis and 

 Pteraspis into which Professor Huxley divided Agassiz's genus Cephalaspis, on account of 

 remarkable differences in the histological character of the shields, each admits of subdivi- 

 sions which perhaps might be viewed as sub-genera, but w^hich, I think, it will be more 

 convenient to regard as genera, adopting new names for the sections of the Family Cephal- 

 aspida, indicated by Professor Huxley's generic division. Thus I propose to call the 

 Pteraspidian forms Heterostraci, in view of the special character of their test ; the 

 Cephalaspidian forms, Osteostraci, in reference to the occurrence of true bony structure 

 in their cephalic shields. Each of these sections contains three or four genera of about 

 equal value, all of which are very definitely distinguished, and nearly all embrace at least 

 two or three species a-piece. 



§ IV. Nature of the Shields of CephalaspidcB ; and grounds for division into two Sections. 



1 Professor Agassiz in his ' Monogr. de Vieux Gres Rouge,' published in 1844, remarked that the 

 ten years that had passed since the publication of the ' Recherches' had not brought any fresh evidence 

 to light as regarded the Cephalaspidce. 



