100 MEMOIRS OF THE QUEENSLAND MUSEUM. 



the tubes simple, not quite reaching the margin of the scale. Head large, with moderate 

 or rather long snout and narrow preorbital, partly covered with small, usually cycloid 

 scales. Mouth terminal and protractile, with moderate, oblique cleft, the jaws equal ; 

 lijis rather thick, plicate ; maxillary exposed distally. Teeth in a band in each jaw 

 variable in form ; on the vomer and palatines, if present, caducous. Preorbital serrated 

 in the young, preopercle more or less strongly serrated ; opercle with one or two pungent 

 spines ; suprascapular and coracoid bones generally exposed. One more or less deeply 

 notched dorsal fin, with xii-xiii (rarely xi or xiv) 8-14 rays, the spinous portion longer 

 than the soft and depressible in a groove. Caudal fin rounded, or truncate, or emar- 

 ginate, with 17 principal rays, 15 of which are divided. Anal short, with iii 7-12 rays. 

 Pectoral Aiith 12 to IG rays, varying from pointed and asymmetrical to rounded and 

 symmetrical. Ventrals close together, inserted well behind the pectorals. Gill- 

 membranes separate, free from the isthmus ; six branchiostegals ; pseudobranchiae 

 large ; gill-rakers in moderate number, short and stout. Air-bladder large, mesially 

 constricted. Pyloric appendages in small or moderate number. Vertebrae 12 + 13=25. 

 Ribs inserted on parapophyses. {0 parrwy, an attendant.) 



Fishes of small or moderate size from the IncUan and Western Pacific Oceans 

 and the fresh waters of AustraUa and New Guinea. 



In its commonly accepted form, the genus Therapon includes a number of very 

 dissimilar fishes, the tyi^ical forms of which are almost entirely marine, while others 

 are estuarine, and many are confined strictly to fresh waters. Their affinities have 

 been subject to considerable discussion by various authors, and they have been included 

 in the famihes Pristipomatidce by Giinther, and in the Lutjanince by Boulenger, while 

 Jordan and his colleagues have adopted the family Theraponidce to accommodate 

 them. Regan, having examined the skeletal characters, regards Therapon as a Serranid. 



Their subdivision into natural groups has been attempted by several writers 

 with more or less success. Cuvier and Valenciennes distinguished Datnia and Pelates 

 in 1829, which, however, have been reunited with Therapon by most authors.^ 



The latter genus is readily separable by the junction of the gill-membranes with 

 the isthmus and by its dentition, but we are unable to define the former, though its 

 general appearance is very difierent to that of the typical Therapon. Eutherapon has 

 been founded on T. theraps by Fowler which has larger scales than the typical T. servus. 

 The relative lengths and strength of the dorsal and ana*l spines, and the presence or 

 absence of strong bony ridges on the cranium, suggest characters which might be used 

 for subdivision purposes, as also does the varying structure of the scales. The extremes 

 of all these apparently merge one into another, however, and any one character is not 

 always constant in obviously closely related species. An investigation of the skeletal 

 characters may reveal a ready means of subdivision, but we have been unable to under- 

 take this phase of the work. 



The freshwater species of Australia include such widety diverse representatives 

 as the small-scaled T. bidyana in which the suprascapular bone is exposed, and 

 T. percoides in which it is liidden and the scales are large. They are united, hoAvever, 

 by a chain of intermediate forms which appear to defy subdivision, so we reluctantly 

 accept the genus Therapon in its broad sense, distinguishing only Pelates as a well- 

 defined group. 



^ The genus Helotes Cuvier & Valenciennes has also been included in the synonymy of 

 Therapon by som',' authors, but the trilobate character of its teeth is apparently constant, and 

 readily disinguishes it from its allies. 



