NO. 2136. REVISION OF THE 0ENU8 APHYCUS—TIMBERLAKE. 575 



slightly rounded. Abdomen short, subrotund and flat in life, but 

 after death the dorsum becomes more or less concave, and the sides 

 compressed ; the fifth ventral sclerite reaches to the apex of abdomen 

 and incloses the ovipositor; the latter usually shortly protruded. 

 Legs moderate in length, rather slender, but the hind femora may 

 be slightly or moderately flattened. Wings moderate or small in 

 size; the marginal vein punctiform, the postmarginal but little longer ; 

 stigmal moderately developed, running nearly parallel with the ante- 

 rior margin of the wing, and enlarged at apex. Hind wings narrow, 

 the costal cell very narrow throughout its length but reaching to the 

 booklets. Sculpture of head and thorax alutaceous, punctures if 

 present microscopic, pubescence very fine and inconspicuous but 

 sometimes abundant. Coloration entirely yellow of different shades, 

 the antennae and legs concolorous. 



Male. — Similar to the female in all characters. 



Type of genus. — Acerophagus coccois Smith. 



The species here included in AceropJiagus are all parasitic in Dac- 

 tylopiine hosts, so far as definitely known, ])elonging to the genera 

 Pseud ococcus, Phenacoccus, and Erium. 



Metallon Walker differs in being metallic and in having rudimen- 

 tary wings; Rhopus Foerster has the body much flattened and the 

 antennae dissimilar in the male sex; Psylledontus Crawford is sepa- 

 rated by metallic coloration, the much smaller antennal club and by 

 the noncompression of the abdomen after death. In these three gen- 

 era the structure of the fifth ventral sclerite is not described, but prob- 

 ably in all three it reaches only to the middle of the venter and leaves 

 the ovipositor free. Rliopoideus Howard can not be distinguished 

 from what is here considered to be AceropJiagus Smith. Howard was 

 misled in stating that the mandibles of Eliopoideus citrinus are eden- 

 tate, as there are three teeth as described for the other species here 

 considered. Rhopoideus has no affinity with ArrhenopJiagus Aurivil- 

 lius, with which it was placed by Ashmead. 



ANALYTICAL KEY OF SPECIES. 



1. Ocelli in a right or obtuse-angled triangle, ovipositor protruded 2. 



Ocelli in a rather acute-angled triangle, ovipositor not protruded. 1. coccois Smith. 



2. Front and vertex as wide as long or nearly so 3. 



Front and vertex about one-half longer again than wide. 



Anterior ocellus in center of front and vertex, ocellar angle slightly less than 90°. 



2. texanus (Howard). 



Anterior ocellus considerably behind the center of front and vertex, ocellar angle 



fully 90° 3. gutierreziae, new species. 



3. Front and vertex fully as long as wide 4. erii, new s])ecie3. 



Front and vertex wider than long, wings small, extending but shortly beyond apex 



of abdomen 5. citrinus (Howard). 



