DEVONIAN FISHES OF IOWA 79 



right and left, the position of which was first determined by 

 Whiteaves in Bothriolepis,* an observation corroborated by 

 Smith Woodward. 1 These plates must have been situated in 

 front of the mouth, and may therefore be referred to at least con- 

 ventionally, as "maxillae." There can be no doubt that they 

 were similarly placed in Pterichthys ("mental plates"), and in 

 Asterolepis they were designated maxillae inferiores by Pander. 

 Close to the postero-external angle of each of these plates there 

 is a rounded notch, considered by Smith Woodward in Both- 

 riolepis as possibly indicating a nasal opening.f 



The body-carapace is box-like, very nearly flat below and 

 vaulted above. It is composed of thirteen plates, of which three 

 are median and ten paired, and these are united with one another 

 by overlapping sutures, a marginal band along the internal sur- 

 face of the overlapping plate being excavated to fit on to a cor- 

 respondingly excavated band along the margin of the outer sur- 

 face of the plate overlapped. 



On the upper surface we see the anterior and posterior median 

 dorsal plates (a. m. d. and p. m. d.) succeeding each other in the 

 middle line and on each side; passing down also on the lateral 

 wall are the anterior and posterior dorso-laterals (a. d. I. and 

 p. d.l.). On the under surface, and also taking part in the forma- 

 tion of the lateral wall, are two pairs of plates, the anterior and 

 posterior ventro-lateral (a. v. I. and p. v. I.), of which the an- 

 terior requires special attention, as to it the pectoral limb is 

 articulated. Near the anterior extremity of this plate, on the 

 outer aspect and close above the angle which separates the lat- 

 eral from the ventral surface, is a shallow excavation, from the 

 bottom of which rises a peculiar process resembling a thick- 

 walled cup or helmet, whose hollowed-out "mouth" points out- 

 wards, and also somewhat backwards and downwards, the cup 

 itself being fixed by a stout ridge which traverses the containing 

 hollow from behind forwards and also slightly downwards. This 

 may be called the brachial process ("helmet-process" of Pan- 



*Trans. Roy. Soc. Canada, 1887, 4, section iv, pp. 103, 104. 



tGeol. Mag. 1892, decade 3, 9, p. 484. 



J This notch is somewhat differently placed in Bothriolepis, being fair on the 

 outer margin of the plate instead of at its postero-external angle. The plates here 

 called maxillae are interpreted by Professor Patten as mandibles. 



