IMilyct-iunaspis, a jVcw Gouts of CuccosteidcU. 229 



the present, it is first to be noticed that tlie determination (if 

 the constituent plates of the buckler is a matter of extreme 

 difficulty, from the fact that they are apparently all fused or 

 anohylosed togetlier in the manner in which those of tlui 

 shield of Fleraspis are in the adult form sup])0sed to be. 

 Mr Whiteaves has noticed the frequent arrangement of the 

 tubercles in concentric rows, and this arrangement, much 

 more marked in some shields than in others, along with the 

 lines seen to radiate from the ossific centres in abraded 

 specimens, first led me to suspect that the form and arrange- 

 ment of the cranial plates differed in some material points 

 from what is given in Mr Whiteaves' sketch.^ Close observa- 

 tion by means of a good lens enables one, however, also to 

 observe the original lines of suture, due care being taken not 

 to be deceived by fractures. Though the direction of these 

 sutures is often indicated b.y depressed lines or slight grooves 

 free from the tubercular ornament, yet the actual suture is 

 not incised, but slightly raised like a very fine thread, this 

 being due to the manner in which the original lines of separa- 

 tion between the bones have become entirely filled up by 

 osseous matter. 



Tlie results obtained by noting these lines in connection 

 with the concentric arrangement of the rows of tubercles 

 being in all essential respects the same in all the specimens 

 (six) which I have examined, and being furthermore in com- 

 plete accordance with the information derived from the lines 

 radiating from the ossific centres in a specimen with the 

 surface rubbed off, there can be no doubt that the true form 

 and arrangement of the constituent plates of the shield has 

 been arrived at. The pattern in two shields has been repre- 

 sented in outline in PL XII., Figs. 1 and 2, and if the reader 

 will compare these figures with that which I have already 

 given of the cranial shield of Goccosteus^ the fundamental 

 agreement as well as the essential difference between the two 

 will at once be perceived. 



The median occipital plate (in.o) is more or less five-sided, 

 elongate, truncated behind, pointed in front, where it is 



^ Op. clt., p. 93, woodcut. 



- Gool. Mag., Dec, III., vol. VI., ]il. i., lig. 2. 



