taxonID	type	description	language	source
03FE1B364748D449FF25FE00FEB9B593.taxon	description	There are three specimens in the Macleay Museum collection (MAMU F. 433), which measure 36.0, 40.4 and 50.2 mm SL (TL not determinable for any of the specimens owing to severe damage of the caudal fins). The old external label for the jar says “ PSEUDOAMBASSIS MACLEAYI, Casteln. NORMAN RIVER ”; an internal card similarly says “ Pseudoambassis Macleayi, Casteln. Norman River, N. Australia ”. The specimens match an index card, which states: “ Pseudoambassis macleayi, Cast. […] 2 sp. 1 ½ ″ – 2 ½ ″ Norman River ”. The miscount of two rather than three specimens is probably a result of the counting having been made with the specimens still in the jar, with one being overlooked. There are similar miscounts for other MAMU index cards. Jenny Anderson’s entry in the catalogue of the Macleay Museum fish collection lists three specimens, which she mistakenly listed as mere ‘ topotypes’. They were therefore not listed in Stanbury’s (1969) type catalogue, and accordingly were excluded from the 1970 transfer of fish types to AMS. The specimens in MAMU F. 433 are fragile and appear to have been desiccated at some stage. Nonetheless, it was possible to ascertain the following characters, which in combination are diagnostic for Allen & Burgess’s (1990) concept of Ambassis macleayi: single supraorbital spine; no nasal spine; two transverse scale rows on cheek; 10 segmented rays each in dorsal and anal fins; lateral line incomplete; pectoral fin with 15 rays (checked in lectotype only). We therefore regard the three specimens in MAMU F. 433 as the lectotype and paralectotypes of A. macleayi.	en	Gill, Anthony C., Russell, Barry C., Nelson, Gary (2018): F. L. de Castelnau’s Norman River fishes housed in the Macleay Museum, University of Sydney. Zootaxa 4459 (3): 565-574, DOI: 10.11646/zootaxa.4459.3.9
03FE1B364749D44EFF25FBD4FBD9B763.taxon	description	There are two specimens in the Macleay Museum (MAMU F. 431), which agree well with Whitley’s illustrations of the lectotype and paralectotype. The old external label for the specimens says “ PSEUDOAMBASSIS ELONGATUS, Casteln. NORMAN RIVER ”; an internal card in the jar repeats the same information. The index card for the specimens says “ Pseudoambassis elongatus, Cast. […] 2 sp. 1 ″ Norman River ”. As with the P. macleayi types, Jenny Anderson’s entry in the catalogue of the Macleay Museum fish collection mistakenly lists the specimens only as ‘ topotypes’. They too were not listed in Stanbury’s (1969) type catalogue, and accordingly were excluded from the 1970 transfer of fish types to the AMS. The specimens are in very poor condition. The larger of the two (the lectotype) is in particularly poor condition, with the head bent to the left and partially disarticulated from the body (Figure 3). The lectotype measures approximately 29 mm SL, and the paralectotype 22.5 mm SL. Details discernible from the specimens are as follows: single supraorbital spine; nasal spine absent; two transverse scale rows on cheek; dorsal and anal fins each with seven segmented rays; lateral line incomplete; predorsal scales nine; vertical scale rows from upper edge of gill opening to caudal-fin base 25; horizontal scale rows from anal-fin origin to dorsal-fin base 11. This combination of characters is unique among Australian ambassids (Allen & Burgess 1990), and we regard the two MAMU specimens in F 431 as the lectotype and paraalectotype of Ambassis elongatus.	en	Gill, Anthony C., Russell, Barry C., Nelson, Gary (2018): F. L. de Castelnau’s Norman River fishes housed in the Macleay Museum, University of Sydney. Zootaxa 4459 (3): 565-574, DOI: 10.11646/zootaxa.4459.3.9
03FE1B36474FD44FFF25FB28FF6AB198.taxon	description	Therapon terrae-reginae; Castelnau 1878 b: 46 (Norman River). This is the only non-type specimen of Castelnau’s Norman River fishes in the Macleay Museum, which we include primarily as evidence that the MAMU Norman River specimens match with Castelnau’s collection. Castelnau (1878 a) described T. terraereginae on the basis of a single specimen from northern Queensland, possibly from the Fitzroy River. He further noted that he had seen a six-inch Fitzroy River specimen of the same species in the Brisbane Museum (now Queensland Museum). He reported on an unspecified number of specimens from the Norman River in his 1878 b paper. However, he introduced confusion by noting “ The specimen on which I formed this species is not in my possession having been returned to the Brisbane Museum, so I cannot compare the specimens that I have from the Norman River with the type. ” As can be discerned from his original description, the Queensland Museum specimen is not the type. According to Bauchot & Desoutter (1987), the holotype is instead deposited in the MNHN (MNHN A- 703). Vari (1978) included T. terraereginae in the synonymy of Amniataba percoides (Günther 1864). MAMU F. 569 A includes a single 66.5 mm SL specimen (TL not determined, owing to caudal-fin damage). The old label says “ THERAPON TERRAE-REGINAE NORMAN RIVER ”. The index card for the specimen also says “ Therapon terrae-reginae […] 1 sp. 3 ″ Norman River. ” We believe this specimen is Castelnau’s Norman River specimen. It agrees well with Vari’s (1978) description of Amniataba percoides. We note, however, that it does not key to that species using Vari’s key, owing to an error in the first couplet (A versus AA): in contrast to the key, A. percoides has an exposed, serrated posttemporal (versus covered with skin and not serrate according to the key).	en	Gill, Anthony C., Russell, Barry C., Nelson, Gary (2018): F. L. de Castelnau’s Norman River fishes housed in the Macleay Museum, University of Sydney. Zootaxa 4459 (3): 565-574, DOI: 10.11646/zootaxa.4459.3.9
03FE1B36474CD44DFF25FB3BFF61B5DA.taxon	materials_examined	Identification of the species has been problematic. Macleay (1879) considered it to be a valid species of Engraulis, stating (p. 367): “ This species is described by Count Castelnau […] from one adult specimen, 7 inches long, sent to him from the Norman River, Gulf of Carpentaria. Its special distinguishing character seems to be a strong longitudinal ridge along the top of the head. ” Macleay’s mention of a single specimen is curious, given that there are two specimens in MAMU (see below). Ogilby (1910) considered E. nasutus to be a valid species of Anchovia Jordan & Evermann in Jordan (1895), and compared it with his new species A. aesturia. It is unlikely that Ogilby saw type material of E. nasutus, as all of his comparative data for the species is identical to that in Castelnau’s original description. McCulloch (1929 a) regarded E. nasutus as a valid species of Anchoviella Fowler (1911), but did not indicate whether he had examined the syntypes. Whitley (1964) considered both E. nasutus and A. aestuaria to be valid species of Thrissina Jordan & Seale (1925). Whitehead et al. (1988) tentatively placed E. nasutus in the synonymy of Thryssa hamiltoni (Gray 1835), which was followed also by Wongratana et al. (1999). Paxton et al. (2006) listed it as incertae sedis in the Engraulidae, and as a possible synonym of Thryssa hamiltoni. They noted (p. 317): “ syntypes whereabouts unknown. ” There are two specimens in the Macleay Museum (MAMU F. 1194; Figures 6 – 7), measuring 99.5 and 104.5 mm SL (TL not determinable owing to caudal-fin damage). They correspond to an index card stating “ Engraulis nasutus, Cast. […] 2 sp. 6 ″ Norman R., N. Australia ”. The specimens have the following characters (where two counts are presented, the first is from the 99.5 mm SL specimen): predorsal scutes 1; abdominal scutes sharply keeled, 14 prepelvic + 9 postpelvic = 23 total; maxilla relatively short, reaching to posterior border of preopercle; anal-fin rays 34 (iv, 30; anterior ray tips damaged in 104.5 mm SL specimen); total dorsal-fin rays 13; uppermost pectoral-fin ray not filamentous; teeth in jaws fine and conical, not canine-like; lower gill rakers 25 (checked in 104.5 mm SL specimen only). This combination of characters is unique among engraulids to the species currently called Thryssa aestuaria (Ogilby 1910). According to Paxton et al. (2006), there are three other species of Thryssa known from the Gulf of Carpentaria: T. hamiltoni, T. scratchleyi (Ramsay & Ogilby 1886) and T. setirostris (Broussonet 1782). The MAMU specimens differ from T. hamiltoni in having fewer prepelvic scutes (14 vs 15 – 20) and more lower gill rakers (25 vs 11 – 15); from T. scratchleyi in having fewer abdominal scutes (14 prepelvic + 9 postpelvic vs 19 + 12) and more lower gill rakers (18 – 23 in T. scratchleyi); and from T. setirostris in having a much shorter maxilla (reaching to preopercle edge vs to at least tip of pectoral fins), fewer abdominal scutes (16 – 18 + 9 – 10 = 25 – 28 in T. setirostris) and more lower gill rakers (10 – 12 in T. setirostris). We regard the specimens in MAMU F. 1194 as syntypes of Engraulis nasutus Castelnau, 1878, and identical to Anchovia aestuaria Ogilby, 1910. Engraulis nasutus thus is a subjective senior synonym of A. aestuaria. We here follow Eschmeyer et al. (2018, and references therein) in recognising Thryssa Cuvier (1829) rather than Thrissina Jordan & Seale (1925) (Kottelat 2013) as the correct generic name for Thryssa nasutus. Pusey et al. (2017) recorded only a single engraulid from freshwaters of northern Australia, Thryssa scratchleyi. Thryssa nasuta is known only from estuarine and marine areas (Whitehead et al. 1988, Wongratana et al. 1999); presumably the syntypes of E. nasutus were collected from the lower, estuarine reaches of the Norman River.	en	Gill, Anthony C., Russell, Barry C., Nelson, Gary (2018): F. L. de Castelnau’s Norman River fishes housed in the Macleay Museum, University of Sydney. Zootaxa 4459 (3): 565-574, DOI: 10.11646/zootaxa.4459.3.9
