36 ILLINOIS BIOLOGICAL MONOGRAPHS [354 



of the ultimate body-segment were used in separating the families and 

 subfamilies. The number and location of abdominal legs, the food-plants, 

 types of cocoon, and coloration furnished the basis for the separation of 

 tribes, genera, and species. Middleton (1915, 1917) has characterized the 

 larvae of the genus Dimorphopteryx and of the family Cephidae. It 

 must be noted that these writers were concerned only in the preparation 

 of recognition tables for the separation of the particular species they had 

 in hand, and, with the exception of Middleton, no one has attempted 

 to construct a synopsis of families, genera, and species as such. 



In the following pages, the author has attempted to define and describe, 

 as far as possible with the materials at hand, families, subfamilies, genera, 

 and species by the use of larval characters. With a few exceptions, no 

 attempt has been made to incorporate data from previous writers for the 

 reason that the characters recorded by them were found in most cases of 

 little or no value for the present purpose — not because they were inaccu- 

 rate, altho that was true in many cases, but chiefly because they were not 

 of specific significance. For example, Dyar's descriptions of species are 

 usually very accurate and dependable but most of the characters noted 

 excepting coloration often proved to be only of family or subfamily sig- 

 nificance. The definitions given here are correct for the materials actually 

 studied, but it is not surprising if they do not hold good in many cases 

 when more materials become available for examination. It is obviously 

 impossible to attain perfection in the face of so many missing links in the 

 series of genera and species. These missing links will be filled in as rapidly 

 as accurately identified materials become available, but it must be remem- 

 bered that absolutely correct identification is only possible, in the majority 

 of cases, after carrying individual larvae of the species thru to the adult 

 stage, exuviae being saved for each instar. 



In this study the classification of the Tenthredinoidea proposed by 

 MacGillivray in 1906, with later additions, has been adopted in the main 

 in arranging and restricting the families, subfamilies, genera, and species. 

 For generic synonymy, Rohwer's "Genotype of the Sawflies and Wood- 

 wasps" (1911) has been followed. In this section all references to the 

 bibliography of the different divisions and subdivisions have been omitted. 



