﻿rSouLENGER, to whom WC owe Ihe latest reform of the systematic classifica- 

 tion of the teleosts, founded in 1902 (3 a, p. 151) a suborder Catosteomi, which with 

 unchanged contents and with the same name was taken over into his general 

 systematic synopsis of 1904 (3 b, p. 172). The forms composing this suborder are 

 placed by Boulenger in 11 families under 4 groups in the following manner: 



I. Selenichthyes. 1. Lamprididœ. 



II. Hemibranchii. 2. Gastrosteidœ, 3. Anlorhynchidœ, 4. Protosyngnathidœ (ex- 

 tinct), 5. Aulostomatidce, 6. Fistiilariidce, 7. Centriscidœ, 8. Amphisilidœ. 



III. Lophobranchii. 9. Solenostomidœ, 10. Syngnathidœ. 



IV. Hypostomides. 11. Pegasidœ. 



Regarding this whole suborder he says: "The whole question of the arrange- 

 ment of the Physoclists with abdominal ventrals (Catosteomi and Percesoces) is, I 

 feel, much in need of revision, and it may be found advisable to break up this 

 group (Catosteomi) into a greater number of suborders, in which case the Selenich- 

 thyes would stand by themselves; the Hemibranchii and Lophobranchii would be 

 united under the former name, as proposed by Woodward, or under that of Thora- 

 costei (Swinnerton) or Phthinobranchii (Hay)". For my part I am quite at one 

 with Boulenger on the necessity of a revision of this suborder; I believe, that not 

 only should the suborder be broken up as several of its groups must be referred 

 elsewhere, but also that its largest and central group, Hemibranchii, must be split 

 up and some of its families likewise removed elsewhere. Group I, i. e. the genus 

 Lampris, has already been removed by Gill (12 c) from the relationship with the 

 others; so far as I can judge, there is every reason to take it far away, but as I 

 have as yet only glanced at its skeleton I shall not venture to express any definite 

 opinion regarding its right place''. I have however studied the other forms more 

 in detail, and I shall indicate here my view of their position. Group IV, i. e. Pe- 

 gasidœ, should be removed to Boulenger's Subordo IX, Acanthopterygii, and within 

 this to Boulenger's 7th division Scleroparei, the Mail-cheeked forms. The two 

 nearly related families Gastrosteidœ and Aulorhyncbidœ of Group II, the Hemi- 

 branchii, should be removed to the same place. The remainder form a natural 



* Quite recently C. Tate Regan (25b p. 634 seq.) has placed the Lamprididœ with the Veliferidœ, 

 Trachypteridœ and Lophotidcr, which four families he regards as forming one suborder Allotriognathi 

 related to the Beryciformes. 



6- 



