LINXEAN SOCIETY OF LONDOX. 35 



there may liave been unusual mobility of the jaws here ; but the 

 specimen now described shows that this was not the case. The 

 cut surfaces of the jaw-bones are marked with a regular series of 

 vertical strice, proviug that the motion was directly up and down. 



As shown by the same specimen in the British Museum, and to 

 some extent by a second specimen (no. P. 9490), there lies within 

 the upper jaw already described another arcade (i\g. o, pt.), which 

 is evideutly pterygoid though its relationships are not ([uite clear. 

 It is a comparatively slender lamina of bone, which expands at 

 its lower border into a broad hardened punctate surface (c), more 

 or less concave. No bone has been observed which could be 

 opposed to this apparently crushing apparatus. 



Jaw-bones which may be regarded as belonging to young indi- 

 viduals of Ditiichthi/s, such as a mandibular bone described as 

 Coccosteus caji(tho(jai by Claypole *, bear conical denticulations along 

 the oral border, and it is only after considerable abrasion by use 

 that the border becomes a simple cutting edge. The toothed 

 condition may therefore be regarded as primitive, and some genera, 

 such as Coccosteus (fig. 4E) and Blplognathns (fig. 4G)t, retain it 

 throughout life. In these the teeth are firmly fused in a single 

 row to the oral margin of the bone, and it is interesting to note 

 that they also occur along the anterior edge of tlie mandibular 

 element. The latter fact seems to confirm the conclusion that the 

 pair of mandibular bones did not meet at the symphysis, but were 

 well separated by parts which have perished in the fossils. It 

 should be added that in some genera, such as Mijlostoma, tlie edges 

 of the jaws are not modified for cutting, but bear crushing plates 

 which are said to consist of dentine %• 



The WQw observations now recorded show that tlie Arthrodira 

 are not so closely related to the ancestors of the Amphibia as I 

 formerly imagined. It is, indeed, clear that they are not Dipnoi, 

 and it is difficult to recognize much connection between them and 

 the Crossopterygii. They are not Ostracoderms (as sometimes 

 supposed), for they possess ordinary jaws and parts of paired fins, 

 and their anterior median dorsal armour-plate is differently related 

 to the underlying soft parts §. The usual reduction of the tooth- 

 bearing edges of the jaws to cutting blades without teeth, and the 

 strong development of the dermal armour, indeed, iiulicate that 

 the Arthrodira are highly specialized members of the group to 

 which they belong, and we cannot determine their precise 

 relationships until we find and recognize their more normal 

 ancestors. 



Students of fossils are thus continually baffled by the imper- 

 fection of the geological record with which they have to deal. 

 AVe have learned approximately how fishes passed into amphibians 



*• E. VV. Claypole, American Geologist, vol. xi. (1893), p. 167. 



i" J. S. Newberry, Palicoz. Fishes N. America (Men. U.S. Geol. Surv. no. xvi. 

 1889), p. 159, pis. xi., xii. 



+ B. Dean, Mem. New York Acad. Sci. vol. ii. (1901). pp. 101-109, 

 pis. vii., viii. ; 0. E. Eastman, Bull. llua. Comp. Zool. Harvard, vol. lii. (1909), 

 pp. 2(U-2i;9. 



§ A. S. Woodward, Proc. Linn. See. 132ud Sess. (1920), p. 3-1. 



d 2 



