8 PROCEEDINGS OF THE NATIONAL MUSEUM. vol.63 



lanceolate, fine, about two-fifths length of filaments which are If in 

 eye. Pharyngeal teeth 6-5, hooked and edges strongly serrate. 



Scales cycloid, in nearly evenly exposed longitudinal rows, those 

 on middle of sides largest; 2 to 5 basal radiating striae; sometimes 

 several short apical marginal striae; circuli coarse, 17 to 20; fins scale- 

 less, except few scales on caudal base; ventral axilla with small scale. 

 Spawning male with 4 pearl organs on lower surface of each mandi- 

 bular ramus, also cluster of smaller or minute symphyseal ones and 

 several around snout edge; no other tubercles now apparent. 



Dorsal origin little nearer caudal base than hind eye edge, first 

 branched ray highest and fin extends nearly half way to caudal base. 

 Anal inserted about last two-fifths of dorsal base, first branched ray 

 highest and reaches base of last. Caudal moderately forked, lobes 

 pointed and rudimentary rays moderate. Pectoral low, reaches 

 slightly over two-thirds to ventral. Latter inserted well before dorsal, 

 reaches about four-fifths to anal, not to vent, which is close before 

 anal. 



Color in alcohol : Back warm sepia, fading paler on abdomen, the 

 under surface being glossed with bright silvery white ; sides of head 

 especially bright; iris silvery white; sides of body punctuated with 

 minute dark brown dots which extend to anal; fins dull uniform 

 brownish. Length 60 mm. 



Type.— Cd^t. No. 83278, U.S.N.M., no data. Paratypes. two ex- 

 amples, also without data, have the head 3f to 4; depth 4^ to 4f in 

 standard body length; D. iii, 7, i; A. iii, 10, i; scales 42 or 43 in 

 lateral line to caudal base and 4 more on latter. First 10 to 12 

 scales of 1.1 distinctly tubular; 10 scales transversely at dorsal and 

 anal origins; 22 to 25 predorsal scales; snout 3f to 3^ in head; eye 

 3 to 3| ; maxillary 3 to 3| ; interorbital 2| to 3. 



Although the locality is given as " Oahu or Fiji?" such is evi- 

 dently entirely erroneous, and the species may have been procured 

 in the Old World, possibly the Indo-Malayan region?. It cer-. 

 tainly differs from any North American cyprinoid known to us 

 in the remarkable combination of its characters. In a general super- 

 ficial way it is suggestive of certain species of the American genus 

 Notropis. Its relationship should be sought, however, in the Ras- 

 horinae. Among the known forms of this group it would appear 

 unique in its small scales and uniserial pharyngeal teeth. Specifically 

 it bears considerable resemblance to BarUius evezardi Day, but 

 differs in many striking characters, as its incomplete lateral line, 

 larger fins, narrow suborbitals, and pharyngeal teeth. In B. eve- 

 zardi only one row of scales intervenes between the lateral line and 

 ventral origin. In Rashorella duhia at least 2, or even 3, would occur. 

 if the lateral line were complete. 



(Duhia, doubtful.) 



