WASP GENTJS CERCERIS — SGULLEN 339 



Color shades and color patterns that have been used extensively 

 by earlier describers are used by the writer to a more limited extent. 

 He has found the color markings of some parts are far more stable 

 than those on other parts within the same species. Futhermore, the 

 overall tendency for color markings to vary is different for different 

 species. One of the most extreme cases of color variation within 

 a species is C. calijornica Cresson. Since, in some cases, the color 

 changes are associated closely with geographical distribution, it seems 

 desirable to recognize subspecies. 



In the females the form of the clypeal process and clypeal border 

 continue to be the most important characters to use in separating 

 species. For the males, the relative width of the hair lobes on the 

 clypeus is important in separating groups but not individual species. 

 Denticles on the mandibles, which were seldom mentioned by earlier 

 workers, have been found of value in numerous cases. Other char- 

 acters seldom or never mentioned by former workers but found useful 

 by the writer are the form and sculpturing of the tegulae, the sculptur- 

 ing of the enclosure, the mesosternal process, and the form of the 

 pygidium. The form of the pydigium has been of special value in 

 separating females. Extensive studies have been made of the male 

 genitalia in an effort to find more dependable characters to help in 

 their identification. Noticeable differences are found when comparing 

 the males of one group with those of another, but, when species are 

 otherwise closely related, no usable differences have been found. 



In general, the writer has found it far more diflBcult to separate 

 males than females. Even in some cases wherein the males definitely 

 have been recognized by their association with known females, no 

 characters as yet have been found that will separate them. Here is 

 an important field of study in this genus that should be pursued. 



Natural subdivisions. — Structures and biological behavior show 

 certain natural divisions within the genus Cerceris. In the present 

 publication the writer recognizes five more or less distinct groups. 

 Further study may show that some or all of these groups are sufficiently 

 distinct to warrant recognition as subgenera. For the present, it 

 seems desirable simply to separate them as groups. Other natural 

 subdivisions, not herein recognized, may appear with fiu"ther studies. 

 Future structural and field biological studies may help in separating 

 distinct groups. Ungrouped species are arranged alphabetically 

 separate from the groups. The characters of each group are presented 

 under each heading. 



Material studied. — Through the kind cooperation of many 

 persons, the writer has been able to study well over ninety percent of 

 the material from north of Mexico now found in the collections of 

 this country and also many collections in Europe. Over 40,000 



