THE ORDERS OF TELEOSTOMOUS FISHES 485 



ordinal assemblages around central forms without disrupting 

 natural connections at the peripheries. Although, as implied 

 by Jordan, any such attempt must over-emphasize certain breaks 

 in the sequence, yet on the whole the least distortion of natural 

 relationships seems to be secured in the following scheme. The 

 group Mesichthyes Hay, 1 proposed as an order to include the 

 Haplomi, the Synentognathi, and the Percesoces, is here modified 

 by the addition of the Iniomi to the Haplomi, by the transference 

 of the Percesoces to the Acanthopterygii, and by the inclusion 

 in it of the order Salmopercse Jordan and Evermann, the whole 

 group being raised to superordinal rank, coordinate with the 

 superorders Malacopteroidei, Thoracostraci, Acanthopteroidei. 

 In defense of this procedure I may say, first, that the close 

 connection between the Iniomi and Haplomi has led Boulenger 

 to merge the Iniomi in the Haplomi. Second, the removal of 

 the Percesoces to the Acanthopteroidei (already advocated by 

 Jordan and Evermann) is justified by the fact that in the Per- 

 cesoces for the first time among fishes with a closed air bladder 

 appear (i) a separate spinous dorsal, (2) the connection of the 

 pelvis with the clavicular arch, (3) the reduction of rays in the 

 ventrals to one spine and five soft rays. These characters 

 (usually found among the Percesoces in combination) serve 

 to separate them sharply from their supposed near allies, the 

 Synentognathi. Third, as to the inclusion of the Salmopercas 

 in the Mesichthyes, Boulenger unhesitatingly pronounces the 

 Sand-rollers to be progressive Haplomi, but it seems better to 

 regard them as coordinate in rank with that group. 



As constituted above the superorder may be separated from 

 the Malacopterygii by (1) the absence of the mesocoracoid 

 arch (a character believed by Swinnerton, 2 from the evidence 

 of embryology, to indicate a very ancient separation from the 

 Malacopterygii) and (2) (apparently) by the absence of the 

 orbitosphenoid. From the Thoracostraci the superorder may 



1 " Bibliography and Catalogue of the Fossil Vertebrata of North 

 America," Bull. U. S. Geol. Surv., No. ijg, Washington, 1902, p. 397. 



2 "A Contribution to the Morphology and Development of the Pectoral 

 Skeleton of the Teleosteans, " Quar. Jour. Micros. Sci., n. s., No. 1941. Vol. 

 49, Part 2, 1905, pp. 363-382. 



