Structure and Classification of the Asterolepidie. 497 



Pterichtliys are but slight, nothing can be more salient than 

 those which distinguish Bothriolepis from both, as will be 

 seen from the following sketch. 



Head. — The median occipital [m.occ. PI. XVIII. fig. 6) has its 

 lateral margins more perpendicular to the posterior one, the an- 

 terior margin shows not merely a shallow reentering angle for the 

 postmedian plate, but a deep semielliptical notch or excavation. 

 The postmedian is small, narrow, semielliptical in shape, and, 

 except its anterior margin, is entirely received in the aforesaid 

 notch of the median occipital, not extending on each side to 

 join the laterals, as in Pterichtliys and Asterolepis. The 

 laterals (/.) are much broader, while the extra-laterals (e. /., 

 B in Whiteaves's figure) are very small and narrow ; but I 

 have not seen the still smaller plate which Whiteaves figures 

 as A in front of the last-named. 



The pattern of the cephalic lateral-line grooves is consider- 

 ably different from that in Asterolepis and Pterichthys. No 

 transverse commissure unites the lateral groove of each side 

 across the occipital plates, as in those genera ; but in front, 

 just at its inward flexure on the lateral plate, a conspicuous 

 branch is given off which runs forwards and outwards to the 

 margin of the shield, while immediately behind the origin of 

 this branch and on the inner side of the main groove a small 

 ear-shaped mark is often, though not always, seen. On the 

 median occipital two slighter grooves are seen, forming an 

 angle with each other behind, whence, diverging obliquely 

 forwards and outwards, they pass also over the lateral plates 

 and terminate near the flexure of the great groove, close 

 behind the origin of its small outer branch. 



These grooves are only superficial and have nothing to do 

 with sutures, either present or former ; nevertheless their 

 having been considered as such has, as in the case of Gocco- 

 steus, given rise to confusion in the enumeration of the plates 

 of the cranial shield. Owing to this source of fallacy, Whit- 

 eaves, like Lahusen, has numbered, in his figure of the head 

 of B. canadensis, no less than seven plates more than what 

 really exist, namely his no. 2 in front and on each side his 

 nos. 2 a,' 3, and 9 a, though he owns that 9a" may possibly 

 be a part of the postlateral " (external occipital). That is 

 undoubtedly the case, and in like manner 2 a and 3 are por- 

 tions of his prelateral (lateral) and 2 of the premedian. Nos. 

 2 and 2 a he regards as equal to the " os terminale " in 

 Asterolejns ; but if we turn to Pander's figure (7, tab, vi. 

 fig. 1) we shall find that similar divisions are marked off by 

 a similar groove on the premedian and lateral plates alto- 

 gether independently of the division between these plates and 

 the OS terminale. 



Ann. & Mag. N. Hist. Ser. 6. Vol. ii. 35 



