2 U.S. NATIONAL MUSEUM BULLETIN 2 02 Vol. 3 



pauchenichthys have the pelvic fins separate but close together. 

 The fishes of this relationship appear to be mtermediate between the 

 family Taenioididae and the family Microdesmidae. 



After the preparation of this section in early 1954, Dr. William 

 Gosline during his visit to the National Museum in September 1954, 

 discussed with me the relationships of Kraemeria and Microdesmus 

 since both of us intended to publish on the relationship of Para- 

 gobioides to the Microdesmidae, a problem first suggested by 

 Fowler (1949, p. 152). I was glad to see Goslines' paper, "The oste- 

 ology and relationship of certain gobioid fishes, with particular refer- 

 ence to the genera Kraemeria and Microdesmus,'^ appear in Pacific 

 Science (vol. 9, pp. 158-170, figs. 1-7, 1955). However, since I have 

 had to include some of the genera not covered in his osteological 

 study, and since some of my observations on characters do not agree 

 fully with his, I am leavmg this section as it was originally prepared. 



Through the courtesy of the Leiden Museum, I have examined the 

 holotype (Cat. no. 4800) of Pholidichthys kucotaenia Bleeker. This 

 unique species was placed by de Beaufort (in de Beaufort and Chap- 

 man, Fishes of the In do-Australian Archipelago, vol. 9 p. 446, 1951), 

 along with the genus Gunnellichthys, in a separate family, Pholid- 

 ichthyidae, under the Order Blennioidea. I have referred this genus 

 to the family Microdesmidae. I note the following corrections for 

 the illustration of P. leucotaenia by de Beaufort and Chapman 

 (fig. 85, p. 448) : Dorsal origin is half eye diameter in front of a vertical 

 line through rear edge of head; head is too short, the length of head 

 is contamed 2 times in distance from snout tip to anal origin; a 

 vertical line through rear edge of maxillary passes through rear edge 

 of pupil, eye diameter equal to snout and contained 3 times in postor- 

 bital length of head. Measurements made on the type and expressed 

 in thousandths of the standard length of 83.4 mm. are as follows: 

 Length of head 198; depth 108; diameter of eye 38; length of snout 

 38; bony interorbital space 26; snout tip to dorsal origin 180; snout 

 tip to anal origin 396; length of longest caudal ray 76; tip of snout 

 to rear edge of maxillary 62. 



The unique dentition of Pholidichthys, conical teeth in 4 overlapping 

 rows in both jaws, not correctly illustrated by de Beaufort and 

 Chapman (fig. 85, 1951), has been diagrammatically drawn by me 

 (fig. 133). Provisionally, I do not consider the family Pholidich- 

 thyidae as closely related to the Blenniina; there are no spines in the 

 median fins and the pelvic fins, resembling those of eleotrids, are lo- 

 cated under the pectoral fin base and not in front as in blennioid fishes. 

 This aberrant family I tentatively refer to the suborder Gobiina 

 under the superfamily Gobioidea. It shows relationships with 

 the Microdesmidae, but because of its distinctive dentition and 



