— 4— 



B. With second lobe entirely rudimentary. 



C. Median lobes erect, produced, no notch on inner marfijin, 

 incisions even, chitinous processes small and even. 



osliorni. 

 CC. Median lobes usually notched on both sides, interval be- 

 tween first and second lobes wide, inner process of first 



incision larger than outer amyous. 



BB. Second lobe slightly developed. 



C. Outer chitinous process of first incision largest huntfri. 



*CC. Inner chitinous process slightly the largest. 



D. Median lobes invariably notched on both outer and 

 inner margins. 

 E. Median glands always absent. 



F. Caudolateral and cephalolateral groups dis- 

 tinctly separated. Cephalolaterals. six to 

 seven; caudolaterals, three to four. . Jiowardi 

 FF. Cephalolateral and caudolateral groups merged 

 together in a line or but very slightly separ- 

 ated. Cephalolaterals and caudolaterals to- 

 gether, eleven to fifteen; caudolaterals when 



separable, about nine coloratus 



EE. Median glands, none to two; cephalolaterals, four 



to nine; caudolaterals, three to eight ttvoe. 



DD. Median lobes never notched on inner margin, median 



group of glands absent aesciili 



CCC. Inner chitinous process first incision much elongated, en- 

 larged at apex i su/isiini/is 



BBB. Second lobe with inner angle decidedly developed. 



C. Median lobes normal in length, usually more or less notched 

 on outer margin. 

 D. Inner chitinous })roc(^ss of first iucision twisted or club- 



h'haped .forbesi 



DD. Inner chitinous process of first incision not club- 

 shaped, both processes of very nearly equal size. 



E. Ventral grouped glands absent pemiciosus. 



EE. Ventral groupel glands present. 



F. Chitinous processes of first incision long and 

 about equal; caudolateral glands less than 

 five; medians exceeding six ostreaeforviis 



♦Under tbls head belongs presumably .4. t(>ii')i,se)uli Ckll. It has been Impossible 

 to obtain specimens of this species for study and the available desirlptlon, which 

 win be found word for word on a subsequent vaixe., is not sufflciently diagnostic to 

 place this species with certainty in Its proper place. According to Prof. CocUerell it 

 is closely related to uvae. 



