426 PROCEEDINGS OF THE NATIONAL MUSEUM v 



Climacia, Rophalis, and Neurorthus — without designating trilj 

 Navas (1935, p. 16) divided the Sisyridae into three tribes, "Neuij 

 thinos" {Neurorthus) , "Sishinos" (Sisyra, Sisyrella), and his propoij 

 "Climacinos" (Climacia). Sisyrina Banks would appear to belong 

 the "Sisirinos" of Nav^s; however, the tribes will not be differential 

 in this revision. 



The Sisyridae are represented in Tertiary deposits of Baltic ambeii 

 the Lower Oligocene (Rophalis, Sisyra'?) (see Killington, 1936,i 

 5). Handlirsch (1908, p. 1252) concluded from paleontologi; 

 evidence that the Sisjo-idae arose from the stem which also gj 

 rise to the present-day families — Coniopterygidae, Hemerobiid 

 Berothidae, and Dilaridae. Handlirsch also believed that the ot 

 families, including the Osmylidae, arose on six other branches fr 

 the main stem. Tillyard's earlier work (1916, p. 272), based pi 

 cipally on the wing venation of fossil and recent groups, sugges' 

 that a reduced remnant of the osmyloid stem had thrown off 

 aquatic Sisyridae and reached the extreme limits of reduction n 

 the Coniopterygidae with theh reduced venation. Later (Hi, 

 p. 312) he divided the Planipennia into five superfamilies, the Sisyri ,e 

 being included with the Dilaridae, Psychopsidae, Berothidae, Hemo)- 

 biidae, Chrysopidae, Mantispidae, Polystoechotidae, and Osmylile 

 in the superfamily Hemerobioidea. Carpenter (1940, p. 194) f- 

 lowed this classification in his revision of the Nearctic Hemerobiidp, 

 Berothidae, Sisja-idae, Polystoechotidae, and Dilaridae, althoughlie 

 recognized that the families do not form a natural group by thii- 

 selves. Eventually Tillyard (1932, p. 29) considered the Sisyriie 

 to be a specialized side-branch distinct from both the Dilaridae : d 

 Hemerobiidae, with the Ithonidae, Dilaridae, and Hemerobiii.e 

 representing one line of evolution, the Prohemerobiidae and le 

 Psychopsidae specialized side-branches, the Berothidae the old^t 

 existing family of Planipennia, and the position of the Coniopterygiie 

 "still as much a mystery as ever, there being no close connectijs 

 between these highly reduced forms and other Planipennia, eiW 

 fossil or recent." Withycombe (1925, p. 402), basing his conclusiis 

 on the study of the immatiu-e stages, divided the Neuroptera iio 

 five somewhat different superfamilies and included the Sisyridae id 

 Osmylidae on the same offshoot, with the Dilaridae, Berothidae id 

 Mantispidae, and Myiodactylidae on three later offshoots of the si le 

 stem in the Osmyloidea. Although there is a considerable dift'ertse 

 in appearance in the adults of the Osmylidae and Sisyridae, Wilf- 

 combe (ibid., p. 400) beheved the Sisyridae to have had osmyid 

 ancestors, which went into deeper water as larvae and then tool to 

 preying upon fresh-water sponges, at the same time developing loi er 

 jaws and abdominal tracheal gills and losing the labial palpi in 

 addition, there was reduction of the two claws into a single c]ff, 



