THE FAUNA OF THE ST. JOHN GROUP. 145 



the slate, with Agnostus rex, appears to have about five joints in the rachis, of which the 

 foremost only are plain. Another from slate, with Microdiscus Scanicus, appears to have 

 six or seven joints, of which also the last are obscure In all examples, the rachis is 

 elevated and almost of equal breadth. The forking in the last joint, which Brogger 

 described in the Norwegian example, I have not with certainty been able to distinguish 

 in the examples from Andrarum. The side lobes have two ribs, each of which bears an 

 oblique-going furrow. In the above mentioned complete example in the museum at 

 Lund, the pygidium is not so plain as to enable one to distinguish its structure with 

 sufficient exactness. 



" Z. Linnarssoni sta^ndfi very near L. aruleatus, which, notwithstanding, is separated by 

 the more square-cut glabella, by the head-shield having the anterior margin bent back- 

 ward, the space behind the margin depressed, and the forward part of the cheek sharply 

 descending. Again, these characters are not found in the Norwegian form which Brogger 

 named L. aculeatus, and it thus cannot (as I have already pointed out) be identical with 

 Angelin's species. So far as I can discover, it should rather be joined with L. Linnarssoni. 

 The dissimilarities which Brogger only can find, may to a great extent depend on the 

 state of preservation. There is a spine found on the neck- ring which, I shewed also to be 

 present in L. Linnarssoni. The glabellar furrows also in L. Linnarssoni are freqixently so 

 faint that it is only with difficulty (or not at all) that they can be detected. Such even 

 is the case with the punctation of the crust. And as regards the area [limb of the cheek] 

 in front of the glabella, its breadth varies considerably, even in examples which must be 

 referred to L. Linnarssoni. 



"From the remainder of the Swedish species of thegeuus Liostracus, L. Linnarssoni is 

 distinguished by the convex head-shield, the spine on the neck-ring, and also by the 

 small, pygidium with few joints. 



" The most nearly related among English species is Conocori/phe applanata, Salter, which 

 plainly is a Liostracus. It has fourteen (?) joints in the thorax, and, according to the 

 description, diffi^rs in other respects. 



"Among the numerous species which, by American authors, are referred to the genera 

 Conocoryphe or Conocephalites, some may be Liostraci. So far, their special characters 

 have, for the most part, been but imperfectly presented. 



" Occurs commonly in slate with Blicrodiscus Scanicus, Agnostus intermedius, A. 

 rex and Conocoryphe œqualis, and may, if one may except some Aguosti, be the latest 

 uppermost fossil in the Lower Paradoxides beds at Andrarum." 



Head-shield (Plate II, fig. 1«). — We find no point of diff"erence between Linnarsson's 

 description of the head-shield and the appearance of the Acadian examples before us, 

 except such as may belong to the differences between the broad and the narrow forms. 

 Thus he speaks of the limb of the cheeks in front of the glabella, as being in some cases 

 a fifth, and in others a third, of the glabella's length ; the former proportion characterises 

 the narrow, and the latter the broad form, which appears to be the one which Brogger 

 describes as the type of the species. The difference in proportion between our examples 

 of the head-shield, and those of Linnarsson, is due to the occipital spine, which has a 

 length of about 2 mm. in the type and 3 mm. in the narrow form. 



The Acadian examples of the head-shield of this species always have a spine at the 

 back of the occipital ring, or an elevated point where the spine was attached : this raised 



Sec. iv, 1887. 19. 



