748 CHARLES PAuL ALEXANDER 
Anal gills—In the Ptychopteridae and in the Tanyderidae there is 
but a single pair of anal tracheal gills. In the former these are very small 
and are cylindrical, in the latter they are large, branched, fanlike struc- 
tures. In the Tipulidae the anal gills number from four in most Lim- 
nobiinae to six or eight in the Tipulinae. The increase in number is brought 
about by a division of the original gill of each side. In generalized forms, 
as Antocha, Pedicia, and others, the anal gills are constricted into segments, 
the apical ones being more or less telescopic into the preceding ones. As 
a rule the anal gills are entirely or almost entirely retractile within the 
body. : 
In the Tipulinse the gills vary with the genera, being blunt and con- 
structed for propulsion in the wood-inhabiting species, such as those of 
Ctenophora, Tanyptera, Tipula, and other genera. In the semi-aquatic 
species of Tipula the number of gills varies from four to eight. In the 
latter case each of the four principal gills is deeply bifid and the gills are 
arranged transversely, asin 7’. zgnobilis; in species with six gills the posterior 
branches of the posterior gills are atrophied as a rule; in other species, 
which have but four gills, the four anterior branches are preserved, the 
posterior pair being usually atrophied. In Longurio and Aeshnasoma, 
the four anal gills are pinnately branched, each with about six lateral 
branches. A similar condition is found in the Tanyderidae. 
The head 
The primitive crane-fly head was undoubtedly of the eucephalous, 
non-retractile type, as found in the Ptychopteridae, the Rhyphidae, and 
other families. The retracted head capsule of the Tipulidae is a derived . 
condition. 
The massive, compact capsule is found in all of the lower groups of the 
three subfamilies of the Tipulidae. Such a head is easily derivable from 
the condition in the Rhyphidae or in the Tanyderidae, for instance. The 
dorsal median sclerite, the prefrons, is almost as large and conspicuous 
in Ula as in the eucephalous families. The lateral plates that constitute 
the capsule are shaped like a mussel in the generalized groups, with the 
posterior incisions shallow. In the more specialized forms, with the 
capsule greatly dissected, the two plates of either side are entirely sepa- 
rated —the innermost, next to the prefrons, being the znternal-lateral 
of De Meijere, and the outermost being his external-lateral. The prefrons 
