JOHNSON: PERCOIDEI 



471 



Fig. 258. (A) Scorpididae— Labracoglossa argenliventris, 9.9 ininSL;{S)ScoTpididat—Medialunacaliforniensis, 10.1 mm SL;(C)Girellidae— 

 Girella nigricans. 10.9 mm SL; (D) Leptobramidae — Z,fp/o/)rama mulleri. 7.2 mm SL; (E) CheWodacXyWdm—Palunolepishrachydactylus. 8.3 mm 

 SL; (F) C\Tr\\\\\dae—Amhlycirrhituspinos. 13.2 mm SL; (G) PoTm\om\dae— Pomatomus satlalrix. 7.3 mm TL. from Pearson (1941); (H) Nem- 

 ipteridae— unidentified, 5.1 mm SL, from Leis and Rennis (1983); (I) Spandae—Acanthopagrus cuvieri. 8 mm SL, from Hussain et al. (1981); (J) 

 Cenlracanlhidae— Plerosmaris axillaris, 7.7 mm SL, from Brownell (1979). 



ability, basic to cladistic outgroup comparison, have seemingly 

 inhibited, or at least hindered, meaningful comparative studies 

 within the Percoidei. 



Because the group is so large, these problems will necessarily 

 continue to plague studies of percoid relationships. Outgroup 

 comparisons based on a single family are speculative without 

 evidence for a sister group relationship, and broader surveys of 

 each character are frequently impractical if not impossible. One 



approach that can gradually alleviate this problem is the cu- 

 mulative tabulation of characters and character states. Com- 

 parative tables document the distribution of morphological fea- 

 tures throughout the suborder and the variability of these features 

 within families, and they accordingly offer the most complete 

 foundation for outgroup comparison. Furthermore, they pro- 

 vide information about the plasticity of various complexes, al- 

 low identification of characters most frequently subject to con- 



