STUIUKP IN ATSIKAI.I.W KISHKS Mc('UI,l,()i'H. M 



of tlie iHiiiy with a large dark blue spot, wliicli is ufLeii elongate and 

 coalescent with its neighbours. S[»inuu8 dorsal trans[)arent orange-brown, 

 with a few pale gieen spo(s on the sftines ; soft dorsal, anal, and caudal 

 orange like the l)ody, with dark blue lines and spots between the rays. 

 I'ei'toral olive-brown, with blue lines and spots. Ventral orange and pink, 

 mottled with olive. Eye light green and pale gtdd. 



Described and figured from a specimen 29 !• mm. long from the snout 

 (o the end of the middle caudal rajs. 



Viin'dtioii. — An extraordinary range of variation in the colour and 

 nuirking has been recorded by both Castelnau and h>tead. The male, 

 Avhich has been figured by Richardson, is usually daik bluish-black with 

 a rich blue streak on each lobe of the caudal and sometimes on the pectoral 

 also; the female is lighter in colour, and ornamented in the manner 

 described above. But this colour marking is not constant, and Stead 

 (1906) records a male example having all the colouration of the female. 



The posterior dorsal spines resemble the rays so closely that it is 

 difficult to determine their number, while the first two anal sjjiues are so 

 hiilden in the membrane as to be easily overlooked. The following table 

 illustrates some variation in these and other details. 



>^il"""il""J- — According to Kichardson's original desci'iption, the type 

 of O. cyanoiitelas had only 4-8 scales in a longitudinal series instead of the 

 usual 52-58. This statement, combined with the difficulty of counting 

 the dorsal and anal spines and rays, and also the extraordinarily ynriable 

 colouration, has led to the establishment of two other specific names, 

 (J. J>nijiin'i(fi, Macleay, and 0. Iiroiniii, Johnston, both of which are appar- 

 ently synonyms of 0. cijaiiomtdas. Castelnau regarded O. hrinnumfi as a 

 mere colour variation of O. ci/aiio))ieh(s, and though this synonymy was 

 doubted by Macleay, it has been accepted by all later authors; the type is 

 brown in colour without markings, but it does not differ in any structural 

 details from the specimen here described and figui-ed. The validity of 

 U. hrowuii, was even doubted by its author, and 1 tind no reason to 

 separate it from Richardson's species. 



Localities'. — Several specimens are in the Australian Museum from 

 near Sydney, one from Tasmania, and one from the Abrolhos Islands, 

 Western Australia. The species has been recorded from Sonthport, 

 Queensland, by Ogilby so that its range extends from Southern Queensland 

 to Tasmania and south-western Australia. 



