6 PROCEEDINGS OF THE NATIONAL MUSEUM vol. 124 



M eristic variation (table 1). — Specimens with 11 dorsal spines 

 are restricted predominantly to Indian Ocean localities, where all 

 specimens except the one from western Australia (9 spines) had that 

 number. The 9-spined specimen from western Australia probably 

 indicates that the population from that area is more closely related 

 to Pacific Ocean populations, where specimens with 9 or 10 spines 

 predominate. Only 4 of 176 non-Indian Ocean specimens had 11 

 spines. It is on the basis of dorsal spine count associated with distri- 

 bution that we recognize two subspecies of A. fuscus. Springer (1967) 

 noted a similar distribution pattern for Entomacrodus decussatus 

 (Bleeker), found in western Australia but otherwise distributed only 

 in the Pacific Ocean and adjacent inland seas, with a cognate species 

 found only in the Indian Ocean, exclusive of western Australia. 



Specimens with 11 dorsal spines had about the same total number 

 of dorsal elements as those with 9 or 10 spines, indicating that the 

 decrease in spine numbers probably is effected by conversion of spines 

 to rays rather than by increasing the number of rays. Such a con- 

 version probably is not a simple matter, as there is a relatively com- 

 plex relationship between the posteriormost spine and anteriormost 

 ray. Each dorsal element except these two is associated with its own 

 proximal pterygiophore; the two elements in question are associated 

 with a single, common, proximal pterygiophore. It is of interest here, 

 however, that one Australian specimen exhibited a somewhat mal- 

 formed, segmented element in the position where a normal spine 

 should have occurred. The segmented portion occurred distal to a 

 point where the element apparently had been injured. This element 

 exhibited a normal spinelike attachment (no distal pterygiophore) to 

 its proximal pterygiophore (all rays, and no spines, normally have 

 distal pterygiophores) and was counted as a spine. 



Fin elements and vertebral numbers are generally highest in the 

 northern Australian (Gulf of Carpentaria) population. This popula- 

 tion showed higher total pectoral ray counts than the other popula- 

 tions. No other blenniid so far has been demonstrated to have a 

 population varying from all others in pectoral ray count. It is im- 

 portant to note that, while the eastern Australian specimens had the 

 typical number of pectoral and anal rays for the species, they and the 

 northern Australian population shared the highest average vertebral 

 counts, indicating that pectoral ray and anal ray counts are inde- 

 pendent of vertebral count. 



Upper and lower jaw tooth counts were made on 20 specimens, 

 16.8 to 105 mm standard length. The upper teeth ranged from 135 to 

 226, and the lower teeth ranged from 82 to 147, excluding the canines. 

 The correlation coefficient for both the number of upper jaw teeth, 



