LINNEAX SOCIETY OF LONnO.V. 29 



is continued ; the head heiiij:; doubtless seen I'roui above or below, 

 while the body is usuilly twisted and disphiyed in side-view. Tlie 

 oidy traces oi' internal organs observed by Traquair wei'e stains 

 marking the position of the lateral eyes, paired parallel bars sug- 

 gesting branchial arclies, and (in a single specimen of Lanarkia *) 

 curious square spaces near the lateral border of the head appa- 

 rently corresponding with the vacuities in the head-shield of 

 Kiilpraspia which 1 regard as being in tlie roof of tlie ])rant'hial 

 cl)auil)ers. Jn addition, I have observed in the lower lobe of the 

 caudal lin of Thelodns undoubted indications of the stout carti- 

 laginous lueinal spines of the vertebral axis. 



Through Kallostracon and Toliiptlepia or Tohjpa^pix'^, the 

 priuiitive lleterostraci just described seem to pass into Ci/afJunt/iix, 

 J'dhf'aspis, and I'/erdsjiin '^, in wliicli the shagreen-granules on the 

 head are united into a few symmetrically-arranged plates by 

 fusion with underlying calcified tissue. The shape and arrange- 

 ment of these plates were probably determined by the dispositioii 

 of the sensory canals which traverse them. Their superficial 

 tubercles are fused into more or less concentric ridges ; their 

 middle layer is coarsely chambered for vascular spaces ; their basal 

 layer is usually laminated. In no part of the armour are there 

 any bone-cells. The orbits are distinct at the sides of the base 

 of the rostrum, and a pair of larger openings in the dorso-lateral 

 plates near the hinder end of the shield are probably the outlets 

 of the branchial chambers. The inner or visceral face of the 

 large median dorsal plate, especially in Cyafliaspis^, shows a small 

 median pit just behind the position of the orbits e\idently for the 

 reception of a pineal body. Eurther back are a |)air of > -shaped 

 markings \\liich may be due to the semicircular canals of the otic 

 capsules. Near each lateral margin is a row of pittings which 

 are probably the impressions of branchial pouches. One ventral 

 shield of Cijailmspis, described by Leriche j|, also bears corre- 

 sponding impressions of the supposed branchial pouches. The 

 body behind the shield is imperfectly known, but in Pieraspif! it 

 is scaly, and there can be no doubt that the body-cavity extended 

 into it with the cloacal opening far back. 



The Pteraspidians thus described, showing marks of internal 

 soft ])arts, do not range upwards above the Lower Devonian. As 

 they become fewer they are gradually replaced by another group 



* II. H. Traquair, loc. cit. vol. xl. (1905), p. S82. pi. 2. fl^s. 2, .".. 



t J. V. Eohon, M('iii. Acad. Imp. Sci. St.-Pt'tersbourg [7], vol. \li no 5 

 (1803), ]). 70. 1.1. i. tigs. 42, 4.'), 47, and pi. 2. fig.-<. 54. .50. 



X E. ilay Lunkesier, in Powi-ie & Lankester, ' AMonograpli of the Fishes of 

 the Olii Red Sandstone of Britain,' pt. i. (Pahcont. 8oe. iStiS) ; also Geol. Mao'. 

 vol. X. (1«7-"!), p. 241, pi. 10. A. 8. \A'oo(i\vanl, ' Calalogueof the Fossil Fishes 

 in the Britisii Mnseun;,' pt. ii. (IS'.ti), pp. 1.5i) 17G, with figs. (1. LindstWhn. 

 Billing K. Svensk. Vet.-Akad. Hand], vol. xxi. (1N95), sect. iv. no. .'I. wiili 

 2 pis. {Ci/n/Iiaspis}. 



§ A. S. Woodward, op. cit. 1891. p. 172. jil. 9. fig. 4. 



I| M. Leriche, Mem, Soc. Geol. Nord. vol. v. no. 1 (1900), p. 25. ])1. l.ljV. 5 

 text-lig. 7. 



