200 ON DERMAL, PLATES OF HOiMOSTEUS. [Mar. 17, 



pairs, all directed both, backwards and outwards. These markings 

 become very feeble anteriorly, and still await explanation ; behind 

 they are very suggestive of an impression of the anterior part of the 

 spinal cord with its divergent nerves, but in front they appear 

 rather as if corresponding to the septa between the myocommas of 

 the muscular system. The brain itself would certainly have occu- 

 pied a more anterior situation. 



Both the lateral occipitals are well shown from the visceral aspect 

 in the fossil, and that of the left side is represented in fig. 2. The 

 postero-external angle and posterior border are much thickened for 

 articulation with the dorso-lateral plate, but the specimen is some- 

 what fractured at this margin. A narrow lenticular facette {f) 

 extending along the greater part of the outer border, and a large 

 distinct facette (/) occupying the anterior apex of the bone, show 

 that it overlapped the marginal and central plates ; but the remainder 

 of the visceral surface would be in direct contact with the soft parts 

 of the animal, and it is quite smooth or exhibits only lines of growth, 

 except at the postero-external angle, where a feeble short ridge is 

 directed forwards and inwards, evidently to strengthen the bony 

 lamina at that point. 



Though occurring on a detached fragment of matrix, the bone 

 here interpreted as anterior median ventral (fig. 3) was found so 

 closely associated with the occipitals already described, that Mr. 

 Calder, who is an experienced collector, has no doubt as to its 

 pertaining to the same individual. Its borders are completely pre- 

 served, thus proving that its obvious bilateral symmetry is not 

 accidental ; and the absence of facettes on the smooth visceral as- 

 pect, except possibly behind, indicates that it must have been over- 

 lapped by any lateral plates that were originally adjoining. The 

 anterior border is nearly straight, and accompanied by a low, 

 rounded, transverse ridge ; and the anterior is the broader end of the 

 bone, being produced on either side into a short and broad triangular 

 extension. The thin hinder border is excavated by a symmetrical 

 re-entering angle, into which the front margin of a diamond-shaped 

 median ventral plate doubtless fitted. 



The fourth plate under consideration (fig. 4) also occurs upon a 

 detached fragment of matrix, and is shown only as an impression of 

 its visceral surface. The borders, however, are almost completely 

 preserved ; and the impression proves that the bone was not quite 

 flat, but marked by two broad rounded folds extending and diverging 

 from the shortest of the four margins. This margin is gently ex- 

 cavated, with a sharp angle at one end and a rounded corner at the 

 other ; while the only distinct evidence of an overlapping facette is 

 at the opposite border, which seems to have been much attenuated. 

 The state of preservation does not permit of ascertaining whether 

 any other facettes were present ; but the outline of the bone agrees 

 so closely with that of the anterior ventro-lateral plates of Coccosteus 

 that it is probable the two long borders joining those already noticed 

 were also overlapping. Indeed, the element now described seems to 

 differ only from the anterior ventro-lateral plates of Coccosteus in its 



