FAMILY MICRODESMIDAE — SCHULTZ 9 



at front of eye, the other next to upper lip; premaxillary not protractile; 

 head naked and predorsal area probably naked. 



Color in alcohol. — Light brownish. 



Remarks. — This new species differs from others in the family Krae- 

 meriidae as indicated in the keys. 



Named in honor of Dr. Wilbert M. Chapman who collected the 

 holotype. 



Family MICRODESMIDAE 



After studying several genera of fishes that appear to be related, 

 I tentatively assign the following characters to this family: Body 

 elongate, sometimes eellike; naked or with fine embedded scales; no 

 lateral line; pelvis 1,3 to 1,5, inserted under base of pectorals; inner 

 rays of pelvics longest; lower jaw projecting, sometimes entering 

 dorsal profile of snout; anterior tip of lower jaw enlarged and fleshy; 

 lips broad; anterior nostrils near tip of snout, somewhat tubular; 

 a nasal pore occurs just in front of dorsal edge of eye; conical teeth 

 in jaws in 1 to 3 series; dorsal and anal fins long, connected with 

 caudal fin by a membrane or free from caudal fin; dorsal fin origin 

 in anterior third of standard length but behind upper edge of gill 

 opening; anal fin origin just behind anus; caudal fin distinct, with 

 about 13 to 17 (based on coxunisoi Microdesmus made by C. Richard Robins) 

 cross-striated rays, 9 to 11 of which may be branched in certain 

 genera; anterior 14 to 22 dorsal rays are noncross-striated feeble 

 spines, remainder are cross-striated; anterior to 2 anal rays noncross- 

 striated, remainder are cross-striated; vertebrae number 19 to 32+22 

 to 40=41 to 62; gill openings restricted to sides or free from isthmus; 

 air bladder present; wide opening behind fourth gill arch; gill rakers 

 small knobs, not very numerous. 



Gosline (1955, p. 169) defines the family Microdesmidae thus: 

 "The peculiarities of the family within the suborder Gobioidei consists 

 chiefly of the elongation of the body with the correlated increase in 

 vertebrae number and the continuous dorsal fin without distinction 

 between spines and rays. In addition, however, it has a peculiar 

 maxillary structure in that the maxillaries send out anterior prolonga- 

 tions which meet or nearly meet each other on the midline in front 

 of the premaxillary pedicels." I agree with Gosline except in regard 

 to "spines and [soft] rays," since there are striated and nonstriated 

 rays in the median fins. 



J. L. B. Smith (Ann. Mag. Nat. Hist., ser. 12, vol. 4, pp. 518-520, 

 1951) discusses the relationship of Paragohioides and concludes that 

 its closest allies are the Ammodytidae and the Trichonotidae. In 

 the former, the abdominal vertebrae are more numerous than the 

 caudal, whereas in the Microdesmidae, the caudal vertebrae are 

 notably more numerous. The Ammodytidae lack an au' bladder. 



748-328—66 2 



