766 CHARLES PAUL ALEXANDER 



The dominance of the Cylindrotominae (genus Cyttaromyia) in the 

 Eocene and Miocene of the North American fauna has already been 

 mentioned. No records of this group from the European Oligocene 

 are available. The recent species of the subfamily are practically all 

 forms belonging to cold, temperate regions, the few Oriental species of 

 Stibadocera coming from mountains at considerable altitudes. 



The Tipulinae have been found as far back as the Mesozoic, but the 

 records are not entirely satisfactory. In the lowermost Tertiaries, how- 

 ever, undoubted tipuline forms occur. Species occur in the Green River 

 shales of Colorado (Eocene). The group was well represented in the 

 Oligocene (Baltic amber, Tulameen beds of British Columbia, Krottensee, 

 and Gurnet Bay), and was very common in the Miocene (Radoboj, and 

 especially in the Florissant of Colorado, where some twenty-five species 

 of Tipula and closely allied genera or subgenera have been described by 

 Scudder and Cockerell). 



Tipulidae of the Pleistocene are not numerous, only a few having been 

 made known from the refuse of lake dwellings in England (Dicaera, 

 apparently related to Ctenophora), and in the African copal, including 

 such genera as Styringomyia, Elephantomyia, and Toxorhina. 



COMPARATIVE MORPHOLOGY 



The morphology of the various stages of crane-flies has been detailed 

 elsewhere in this paper and need not be repeated here. 



PHYLOGENETIC CONSIDERATIONS 



The eucephalous familes of crane-flies are undoubtedly lower, phylo- 

 genetically, than the Tipulidae, and the latter have been derived from 

 the former. The generalized type recurs in all three subfamilies of the 

 Tipulidae, and it is uncertain which of these three should be placed 

 lowermost. Presumably all three groups arose from an immediate com- 

 mon ancestor, or the Tipulinae and the Limnobiinae arose from one point of 

 the tree, the Cylindrotominae developing from the limnobiine stem at a 

 somewhat later period. The accompanying phylogenetic tree (Plate XII , 4) 

 graphically illustrates this apparent evolution of the group. The Limno- 

 biini show but little deviation from the fundamental type. From the 

 level of the lowermost Hexatomini (Ula and Epiphragma), in close 

 proximity to the Limnobiini, the remaining groups of crane-flies can be 



