SUBFAMILIES STENOSMYI.INAE AND KALOSMYLINAE 165 A REVISION OF THE OSMYLID SUBFAMILIES STENOSMYLINAE AND KALOSMYLINAE (NEUROPTERA) By D. E. KIMMINS, Department of Entomology, British Museum (Natural History). (With eight plates and twenty-eight text-figures.) THE Osmylidae, as a whole, have attracted the attention of many neurop-terists, notably Hagen, McLachlan and Kriiger. Many of them are large-winged insects, with well-marked pattern and densely reticulated wings. These features have led to the description of many forms, but unfortunately little attention has been paid to genital structure in either sex, and it is probable that many synonyms have been created, especially in the Osmylinae and Spilos-mylinae. Kriiger, in his revision of the family in 1913-1914, based his classifi-cation almost entirely on venation and placed much reliance on rather minute venational characters. In some species at least, further material has shown these characters to be variable, even differing on opposite sides of the same insect. In spite of this, Kriiger's work forms a very useful basis for our classifi-cation. Nevertheless in the Osmylidae the venation, being complicated, is by no means stable, and should be used with some reserve as a basis for determination. The recent acquisition by the British Museum of the historic collection of Neuroptera built up by McLachlan, and also the types and many of the specimens of Neuroptera from the Tillyard collection, has enabled me to make a more careful study of the members of the two subfamilies Stenosmylinae and Kalos-mylinae. These subfamilies form a convenient unit for study, inasmuch as the genera (with the exception of Isostenosmylus) are restricted in their distribution to Australia (excluding N. Queensland), Tasmania, New Zealand and Chile. The exception is also South American, but it has a more northerly range, covering Ecuador, Southern Brazil and Peru. Further collecting in Chile and Patagonia may well extend the range of Isostenosmylus southward and will probably discover further species of these subfamilies. The major part of the Australian Osmylidae belongs to these two subfamilies, the exceptions being Eidoporismus pulchellus Esben-Petersen, Porismus strigatus (Burm.), and one or two Spilosmyline species of Papuan origin which have reached North Queensland. The structure of the male genitalia confirms the opinion that the Osmylidae are among the more primitive members of the true Neuroptera. In most Neuroptera tin-tenth sternite in the male has been retracted within the abdomen and has become much modified from the normal U-shaped sternite to a saddle or arch-like structure. In these subfamilies are found examples of a tenth sternite which is still external and covered with hairs (Stenosmylus, Oedosmylus), as well as intermediate forms leading to the more or less internal structure of Isostenos-mylus. In many Osmylidae there is found in the male a pair of eversible, finger-like scent-glands, opening on the dorsum between the eighth and ninth abdominal tergites. These are present in the Kalosmylinae, but appear to be absent in the Stenosmylinae, in which the eighth and ninth tergites are fused. novit. zool., 42, 1 14
BioStor
