BULLETIN OF THE BUREAU OF FISHERIES. 
1 66 
with coarse granules or tubercles, and the scale appears to be strictly cycloid, but with a microscope 
it is possible to see minute short teeth all along the apical margin. 
I give a key for the separation of the cichlid scales now before me : 
Apical margin apparently cycloid, the rudimentary teeth microscopical. Tilapia nilotica 
Apical margin evidently ctenoid, the ctenoid patch highly developed, the elements imbricated ... i 
1. Scales not greatly broader than long; elements of ctenoid area very distinct, relatively large, 
the submarginal ones broadened basally; basal radii 8 to lo (Lake Hyanuary, Brazil, 
M. C. Z., 5285; Cichla inonoculus Agassiz, det. Steindachner). 
Chxiobranchopsis ocellaris Bloch & Schneider 
Scales much broader than long; elements of ctenoid area smaller 2 
2. Nucleus much less than a third of total length from apex; scales large, about 8 mm. long and loj 
broad; basal radii about ii; basal margin very deeply crenate (Coary, Brazil, M. C. Z., 
15516) Geophagus jurupari Heckel 
Nucleus a third of total length from apex 3 
3. Basal radii about ii or 12 (Rio Novo, Brazil, M. C. Z., 15739). 
Geophagus brasiliensis Quoy &. Gaimard 
Basal radii about 16 (Lake Hyanuary, Brazil, M. C. Z., 15706). .Chcetobranchus flavescens Heckef 
On the label of the G. brasiliensis is a manuscript generic name proposed by Steindachner for this 
fish, but not published. As the fish was named by Steindachner, it is perhaps really G. steindachneri 
Eigenmann & Hildebrandt, which Steindachner confused with brasiliensis . 
The present distribution of the Cichlidas is much like that of the Characinidae, except that they occur 
in India and Madagascar, where the characinids are absent. Their northern origin is strongly suggested 
by the occurrence of a genus (Priscacara) in the Eocene of Wyoming. This genus has ctenoid scales of 
the normal cichlid type, and numerous basal radii, closely placed in the central part of the scale. The 
character of the ctenoid area confirms the view that the genus is cichlid, not pomacentrid. 
Suborder Pharyngogn.\Thi. Labroid fishes. 
The Pharyngognathi are variously limited by different authors. Thus Gunther (1880) includes the 
Pomacentridse, Labridae, Erabiotocidre, and Chromides. Goodrich (1909) has the same series: Pomacen- 
tridae, Cichlidse, Embiotocidae, Labridae, and Scaridae. Jordan (1905) includes Labridae, Odacidae, Sipho- 
nognathidae, Scaridae (Scarichthyidae). Jordan’s arrangement is strongly supported by scale char- 
acters (I do not know the siphonognathid scale, however); the Giinther-Goodrich series includes forms 
with extremely diverse scales, and is surely unnatural. The scales of the true Phatymgognathi (sens. Jor- 
dan) are characterized by the very thin apical margin, wholly without teeth, and especially the presence 
of apical as well as basal radii. They are thus quite unlike any other acanthopterygian scales knoivn to 
me, unless they may be associated with the Atherinidae, in which the shape and general appearance of 
the scale is quite different. Such atherinids as Chirostoma crvstallinutn Jordan & Snyder (Lake Chapala, 
Mexico; J. N. Rose) and Menidia notata Mitchill show weak but evident apical radii, in the Chirostoma 
leading to a strong lobulation of the apical margin. Outside of the acanthopterygian series the pharyn- 
gognath scale is strikingly similar to that of the cyprinoids. 
LABRID.-E. Wrasses. 
Scales more or less elongate, parallel-sided, having the form seen in the cyprinid genus Labeo (see 
Biological Bulletin, vol. xx, pi. 1, fig. 8). The lateral line canals of Labrichthys are branched anteri- 
orly, as shown by Gunther. The regular basal radii in Pinielomeiopon, Enimeekia, and Coris are in 
the manner of the acanthopterygians. The species examined are Coris venusta Sauvage, Honolulu, 
H. I. (Jordan & Evermann); Emmeekia venusta (Jenkins & Evermann), Gulf of California (Albatross); 
slippery dick, Iridio bivittatus Bloch, Key West, Fla.; fathead, Pimelometopon pulcher (Ayres), Santa 
Barbara Island (Albatross); tautog, Tautoga onitis (Linnaeus), Woods Hole, Mass.; and cunner, Tau- 
iogolabrus adspersus (Walbaum). 
