DEVONIC FISHES OF THE NEW YORK FORMATIONS 47 



either side, but being itself overlapped by the postero-dorsolaterals [/. d. /.]. 

 On the under surface [text fig. 1 1] the place of the two semilunars of 

 Pterichthys is filled, according to Professor Patten, by at least three 

 pieces, the posterior margins of which are assumed by the same author to 

 have been " freely movable in a dorsoventral direction, like an operculum."' 

 The course of the lateral sensory canals is exactly the same as in the body 

 plates of Pterichthys, but in addition two shallow linear grooves diverge at 

 a slight angle from about the center of the antero-dorsomedian plate, and 

 extend across the postero-dorsolaterals. 



The pectoral appendages are longer than the dorsal aspect of the 

 body armor, and even pass beyond the termination of the ventral surface. 

 The proximal portion is also longer than the distal, though the proportion 

 seems to vary ; roughly speaking, however, the difference between the two 

 portions is less than one third of the longer. The proximal portion is, like 

 that of Asterolepis, trigonal in transverse section, and. the plates of which 

 it is composed are also similar in number and arrangement, save that the 

 dorsal aconeal \d. a.\ is a smaller rounded element placed just at the 

 "elbow" joint, whereby the external and internal marginals are allowed to 

 come together for a considerable distance between it and the distal 

 extremity of the dorsal articular \_d. ar.\. The two articulars, as noted in 

 the generic diagnosis, meet together on the outer aspect over the external 

 marginal ; this relation, however, is not always clearly visible in the 

 Canadian specimens, though demonstrable in many fragments from Scot- 

 land and Russia. The lower or "fore" arm is slender and pointed, 

 serrated along the external and internal margins, and composed of a 

 greater number of pieces than in Pterichthys or Asterolepis. 



For our knowledge of the trunk, caudal, and two dorsal fins in this 

 species, which is the only one of the genus showing these features, we are 

 indebted to the painstaking investigation of Professor Patten, who has 

 presented the following preliminary account of his observations :^ 



'Structure of the Ostracoderms. Science, n. s. 1903. 17: 489. 

 "•Biol. Bui. 1904. 7 : 113. 



