350 COXTHTBD'I'IOSS TO CANADIAN PALAEONTOLOGY. 



stellate (iniamentation which is to be seen (iii the plates of nearly all the 

 <,Teat mailed fishes of the old world."' This character, Dr. Newlierry 

 thinks, is of generic value. 



The ventral ai'niature of Js/n<Iir/itJii/s rlanitux, Newheri'y, the typical 

 and heretofoi'(! only known species of the gcnns, is mitirely unknown, and 

 till' inipcifect median \entral plate collected liy .Mr. Tyrrell is referred 

 to Js/,iflir/i//ii/s provisionally, only on account of the similarity of its 

 surface ornamentation to that of the dorso-iiiedian plate of A. r/anttitn. 

 The iinjiression that the specimen fi'cjui South Manitou Island makes 

 upon the mind of the writer is that of the median ventral plate of a 

 lai'ge tish more neai'ly allied lo Curcos/ciis than to Ftrric/itlii/s or Jiothrin- 

 h'jiis, liut which differs fr-om that of Corrosfi-ns in the circumstance that 

 it is partially o\erla})ped, not oidy liy the two postero-lateral and two 

 antero-lateral plates, hut also hy the narrowly rounded end of an antei'o- 

 median plate. The median \cntral plate of ^I. iin/.dJnlis was evidently fif 

 nmch smaller size than that of A. cliirdliis could have been, besides beino' 

 muc-h more minutely tuberculated, and it is quite likely that the disco\-ciy 

 of more jieifect specimens cjf these two forms may show differences 

 b(>tween them that are of <feiieric rathei' than of specific \alue. 



TELKOSTOMI. 



('.R()SS()PTKRY(;iI. 



ONVcirooos. (Sp. Undt.) 



An imperfect scale of a hsh, which is pi'obably referable to this t;'enus, 

 was collected by iMr. Tyi'rell or the present writei', in 188^, on the north 

 shi.iii' of Manitoba Island, in Laki' Manitoba. The scale is not cpiite three-- 

 quarters of an inch in its maxiioui n diameter, cycloidal and not far from circu- 

 lar in outline. Its undei' surface, the surface which happens to be exposed, 

 is marked by fine concentric wriid<les and by very minute radiating raised 

 lines, which are tocj small to be seen without- the aid of a lens. The sub- 

 stance of the scale is so thin as to be transjiarent, and with ,t lens nnnute 

 tubercles, ap]iareiitly for the most ]>art dispost'd in radiating lines, can 

 be detected over most <if its upjier surface. 



.\11 the remains of fishes that ha\'e so far been collected from the 

 De\'oniaii rocks of this region, are from the beds immediatidy above the 

 Stringoi-ephalus zone, at the supposi'd ba,se of the Up])er Devonian. 



