NORTH AMERICAN HYMENOPTERA. 209 



The Bforth American CHRYSIDIDJB. 



• • BY S. FRANK AARON. 



The materials upon which this monograph is based are the almost per- 

 fect collections of the American Entomological Society, containing the 

 types of Messrs. Cresson and Norton, together with a large number of 

 undescribed forms, and the collection of M. Provancher, of Canada, 

 kindly loaned me for study, containing all his type specimens described 

 in the " Naturaliste Canadien," with other rare and interesting species. 

 Only four species treated of in the following pages are wanting in the 

 material at hand, and furthermore I had the advantage of having Mr. 

 Norton's determinations of many species in his paper now in the Society's 

 collection. More than four times the material has since then been added 

 to the collection. 



I believe that there is generally a strong separation between the fauna 

 of the United States and that of Mexico and the Antilles, and I have 

 described only those forms found in America North of Mexico. An 

 examination of a number of Cuban forms belonging to several genera 

 convinced me that there is a slight, but constant difference, in that the 

 Cuban specimens have the thorax rich, light emerald-green, with brassy 

 reflections and without blue, while the abdomen is more or less deep blue 

 and purple with some green reflections. Color is of very slight im- 

 portance in the separation of species, but the absolute contrast of color 

 upon different parts of the same insect, together with its being constant 

 in individuals, alters the case in favor of specific distinction. For this 

 reason I have left certain Cuban species heretofore considered equal to 

 a few of our species out of the synonymy. I have also erected four 

 sub-families much after the manner of Dahlbom's families. 



It has been my purpose not only to give diagnostic descriptions of 

 genera and species, but in the descriptions of each form to mention the 

 inconstant or gradational variations of individuals in order that students 

 may readily appreciate absolute and constant differentiation. This has 

 been rendered the more certain by the large material at hand. I have 

 endeavored to avoid mentioning characters common to every individual 

 in the family, and have for the most part noticed these characters in the 

 following description. 



TRANS. AMKB. BNT. SOC. XII. (31) NOVEMBER, 1885. 



