136 BULLETIN 38, UNITED STATES NATIONAL MUSEUM. 



express in words that a reference to the figures of the plates, where 

 all important types are figured, is deemed sufhcieut. 



The species are not numerous, but some of them are closely allied, 

 and not a little confusion is found in collections. In separating the 

 species of the series with dusky secondaries I am not at all satisfied 

 that I have hit the best solution of the problem. The present scheme 

 worked well when prepared, and does still in the vast majority of in- 

 stances, but material since seen indicates new species which will proba- 

 bly render a re-arrangement necessary. A character not appreciated 

 when the synopsis was prepared is found in the s. t. line and the upper 

 two of the pale outward dentations. Without a complete re-study of the 

 material this can not now be made use of except as indicated in the de- 

 scriptions. I may have given some erroneous determinations in this 

 group by reason of the overlooking of this feature, and placing too much 

 stress on the form of the orbicular — a character not so satisfactory in 

 point of constancy as I had supposed. 



The group may be divided into two, nearly equal series, by the color 

 of the secondaries and the ground color of the primaries. In the first 

 series the secondaries are white, and the primaries show yellow or yel- 

 lowish gray as the predominating color. In the second series the sec- 

 ondaries are yellow fuscous or smoky, and the primaries darker gray 

 or fuscous. 



Among the first series recula may be distinguished by its small size, 

 clear, bright yellow ordinary spots, and other ornamentation, the con- 

 trasting dark shadings bringing this maculation into prominent relief. 

 Cicatricosa and qitadridentata are larger, with similar markings but less 

 intense color. These species are hopelessly confonnded in collections 

 even by Mr. Grote. Both species vary in brightness of color and some- 

 what in maculation. Cicatricosa may be distinguished by the open, V- 

 shaped orbicular and the constricted reuiform ; quadridentata has the 

 orbicular closed, somewhat variable in size and shape, and the reniforni 

 lunate or kidney-shaped. The types of all of the preceding sjiecies have 

 been examined by me. Niveilinea, which has been labeled quadriden- 

 tata in some collections, is easily distinguished by the white line cross- 

 ing the thorax back of the <;ollar. 



Tnsertans and brevipennis differ in the oblong, elongate orbicular. In 

 the former the ordinary spots are connected or fused, the orbicular not 

 oblique, the form not unusually stout. Brevipennis is an unusually stout 

 species with oblique orbicular, not fused or connected with reniforni. 

 The thorax is very heavy and large, the anterior tuft most prominent, 

 the abdomen very short and scarcely exceeding secondaries. 



Flagifjera is a well-marked species of the second series, nearly allied 

 to brevipennis. The ordinary spots are narrow, well marked, outlined 

 in black, then annulate with white, else concolorous ; the orbicularis 

 oblong, oblique, open superiorly; the t. p. line is very indefinite, 

 usually punctiform and somewhat irregular. The pale dents on veins 

 6 and 7 do not cross the s. t. line. 



