10 PROCEEDINGS OF THE NATIONAL MUSEUM vol. 121 



appear to have little consequence from a taxonomic standpoint 

 (Viallanes, 1882a, 1882b, 1885; Kiinckel d'Herculais, 1879; Henneguy 

 and Binet, 1892). 



Phylogeny 



Handlirsch (1908) postulated that the Stratiomyidae arose in 

 Jurassic times from a primitive tipuloid stock that had evolved to 

 the point where it more closely resembled the Xylophagidae than 

 the Tipulidae. The families most like the Stratiomyidae, the Xylo- 

 phagidae and the Rhagionidae (Leptidae), are similar to the stra- 

 tiomyid subfamily Xylomyinae but only in the adult stage (wing 

 venation, spurs on tibiae, etc.). From the Rhagionidae arose the 

 Tabanidae and Therevidae and, in turn, this line of descent led di- 

 rectly to the more advanced Diptera (Handlirsch, 1908; Bischoff, 

 1925; and Lindner, 1937). 



The most primitive subfamily in the Stratiomyidae is the Xylomyi- 

 nae. Characters occuring in the immature stages of this subfamily 

 that I consider to be primitive are an integument that is only partly 

 shagreened, the generalized type of its mouthparts, an incompletely 

 enclosed pupa, and the type of habitat, which is usually under bark 

 or in rotten logs. Characters of the adults are equally primitive. 



From this primitive subfamily 4 lines appear to have evolved, 

 one of which represents a continuation of the basic type wdth slight 

 modifications in morphology (Xylomyinae, Beridinae, and Pachy- 

 gastrinae). The other lines represent a divergence toward different 

 ecological zones in the larval stage: one to an aquatic or semiaquatic 

 mode of life (Stratiomyinae and Nemotelinae) ; another to life in a 

 rich, soft food source such as animal excrement and decaying organic 

 matter (Sarginae and Clitellariinae) ; and another line, represented 

 by Altermetoponia rubriceps (Macquart), which lives in sod 

 (Chu'omyzinae) . 



As previously indicated, the environment in which the larvae live, 

 their feeding habits, and the type of mouthparts are closely aUied 

 and provide an insight into the evolution of the family. Table 3 

 illustrates the distribution of these characters through the subfamilies. 

 In an attempt to show degree of relationship, I have analyzed 11 

 characters and given them an arbitrary numerical value (for details 

 of this technique, see James, 1953). A list of the characters and their 

 numerical values is as follows: 



A. Habitat of larva 



1. aquatic or semiaquatic 



2. terrestrial 



3. terrestrial-arboreal (close association with trees) 



