CEPHALASPID^. 11 



prismatic bodies/ The innermost of the layers is a sohd laminated stratum, having 

 somewhat the appearance of the nacre of molluscan shells, whence Professor Huxley has 

 termed it the ' nacreous layer ;' a few vascular canals pass through it, opening by the minute 

 apertures mentioned as apparent on the inner surface of the shield. The middle layer 

 which consists of the polygonal cavities, is formed by vertical processes of the laminated 

 substance of the innermost layer ; these vertical plates are so disposed as to form a net- 

 work enclosing polygonal (4-, 5-, 6-sided) cells, described by Professor Huxley in 

 Pt. Banksii as having an average diameter of ^th of an inch. The septa forming the 

 vertical walls of the ' cavities ' bend over to complete their enclosure above, and are thus 

 connected with the outer layer — that which is produced into ridges. This general structure 

 is common to all those species of Agassiz's Cephalaspis which have been separated as 

 Vteraspis. 



Professor Huxley minutely describes and illustrates the microscopic appearances of 

 this structure, as seen in a section of the shield of Pt. Lloydii. The most important fact 

 to be observed in this and Pteraspis rostraius, which I have examined, is the total absence 

 of ' bone-lacunse,' or anything like them. The three layers — the outer or 'striated,' the 

 middle or ' cancellated,' and the inner or ' nacreous ' — consist of a continuous finely 

 laminated material. The laminae in Pt. Lloydii, Professor Huxley says, have a thickness 

 of about ^jL^jth of an inch. The inner layer consists of nothing but a compact mass of 

 these laminae arranged horizontally; it is totally devoid of vascular canals or tubules, 

 excepting where here and there a canal of some size passes from its aperture upwards 

 into the walls of the next layer. The laminae are arranged concentrically round the 

 cavities in the middle layer, and large canals, about ^i^th of an inch in diameter, pass 

 along these walls towards the upper layer. In some cases these canals appear to open 

 from the polygonal cavities ; in other cases they come directly from their apertures on the 

 inner surface of the shield. In the upper wall of the stratum of polygonal cavities the 

 canals take a horizontal direction, still very few in number and of large size, receiving 

 here, undoubtedly, branches from the cavities. The calcareous laminae are here arranged 

 horizontally. Passing onwards to the uppermost layer, if the section has been made 

 transversely to the surface-ridges, these appear in section as so many papilliform pro- 

 cesses ; if the section has been made along one of the ridges, a continuous horizontal 

 layer is exhibited. Into this layer, or into these papillae, the horizontal canals send short 

 branches (one to each papilla), which give off minute tubules in every direction ; these arbo- 

 rescent tufts correspond to the vascular bushes in the external layer of Cephalaspis} The 



1 Mr. Edward Fielding, the artist who has so well executed the plates of this part of the Monograph, 

 has been obliging enough to make a chemical analysis of specimens of Pteraspis from the Cornstoues of 

 Herefordshire. He finds that the substance of the fossil itself is almost entirely phosphate of hme, whilst 

 the material sometimes filling the polygonal cavities is carbonate of lime. 



- In the Devonian species of Heterostraci each enlargement of the doubly crenated ridge seems to 

 correspond to a tubercle of Cephalaspis, having a vascular tuft for its axis. 



