April 28. 1905.] 



SCIENCE. 



659 



Wagner (1904)* has recorded the presence of 

 a scaly zone behind the gill cavity, as well as 

 the existence of a pair of minute barbels in 

 Polyodon, thus falsifying the characters body 

 'apparently scaleless ' and ' barbels absent ' at- 

 tributed to the family Polyodontidse by Bridge. 



The Teleostei, including almost all the liv- 

 ing fishes, have been classified by Boulenger, 

 and the work is worthy of that master of 

 taxonomy and verbal expression of relation- 

 ships. The group, formerly and generally 

 designated as a subclass, is degraded to ordinal 

 rank, and all the chief divisions, mostly called 

 orders by American and some other ichthyol- 

 ogists, are designated suborders. Of the sub- 

 orders there are eleven. 



A large number of the groups familiar to 

 American ichthyologists are accepted with 

 practically the same limits as are current in 

 the United States, but always with the inferior 

 rank indicated, the orders being designated by 

 Boulenger as ' suborders.' Such are the (1) 

 Malacopterygii, (3) Symbranchii, (4) Apodes, 

 (9) Anacanthini, (10) Acanthopterygii, (11) 

 Opisthomi, (12) Pedieulati and (13) Plecto- 

 gnathi. Other suborders have received 

 families which had been ejected from other 

 groups, the suborders thus enlarged being the 

 (5) Haplomi, (6) Heteromi and (8) Per- 

 cesoces. Another suborder (2. Ostariophysi) 

 has been made to include the Nematognathi 

 and Plectospondyli, the main difference from 

 American practise being in the fusion of the 

 groups, for the relations of the constituents 

 of the so-called ' suborder ' have long been 

 recognized, as has the group itself as a ' super- 

 order.' ^ 



How divergent this arrangement is from 

 that long adopted in Europe is told by 

 Boulenger (p. 543). "In the classification 

 of Giinther, which has been generally in use 

 in [England] for the last thirty years, the 

 Teleosts were divided into six principal groups, 

 of ordinal rank: I. Acanthopterygii; II. 

 Acanthopterygii Pharyngognathi ; III. Ana- 

 canthini ; IV. Physostomi ; V. Lophobranchii ; 

 VI. Plectognathi. Group [order] I. corre- 

 sponds to sub-order 6 (part), Y (part), 8 (part), 

 10 (part), 11 and 12 of the present work; 



* Science, XIX., pp. 554, 555, April 1, 1904. 



Group II. to sub-order 10 (part) ; Group III. 

 to sub-order 9 and 10 (part) ; Group IV. to 

 sub-order 1, 2, 3, 4, 5, 6 (part), and 8 (part) ; 

 Group V. to sub-order 7 (part) ; and Group 

 VI. to sub-order 13. 



Some of the modifications introduced into 

 the system are rather startling. The 

 Mursenidse, we are told, differ from the other 

 Apodes in that their dentigerous bones are 

 the palato-pterygoid, the maxillaries being 

 absent, while in the Anguillida? and others the 

 dentigerous bones are the maxillaries. It is 

 difficult to believe that the dentigerous bones, 

 specialized as they are, should be so different 

 homologically. Exception must also be 

 taken to the reference of the genus Derichthys 

 to the family Anguillidse. That genus has' 

 both intermaxillaries and maxillaries, and if 

 it must, perforce, be referred to some former 

 order, it is with the Symbranchii and by no 

 means the Apodes that it should be associated. 

 Anyway, it is the representative of a very 

 distinct family — Derichthyidffi. Another 

 group whose new allocation we can not assent 

 to is the Saccopharyngidje. The fishes of that 

 family differ markedly from the true Apodes 

 by the absolute want of all opercular and 

 branchiostegal as well as various other bones, 

 and, indeed, have no similarity, except in 

 elongation of body, to the eels. They are 

 more likely to be divergents from some stomii- 

 form stock. From all other fishes, however, 

 they are widely differentiated, and well en- 

 titled to rank as the equivalent at least of the 

 suborders of Boulenger — the Lyomeri. 



The Comephorida3 are extended to embrace, 

 besides Comephorus, the recently described 

 Cottocomephorus as well as Anoplopoma and 

 Triglopsis. Dr. Boulenger expresses the 

 opinion (p. G97) that ' no doubt can be enter- 

 tained as to the propriety of referring 

 [Comephorus] to the neighborhood of Ano- 

 plopoma,' but after a careful comparison of 

 specimens the reviewer is unable to appre- 

 ciate a resemblance sufficient to entail ap- 

 proximation in the same family. We may well 

 avail ourselves of the technical character ad- 

 mitted by Boulenger himself; in Comephorus, 

 ' the second suborbital is not produced over 

 the cheek, a unique exception to the main 



