172 PROCEEDINGS ENTOMOLOGICAL SOCIETY 



tween the submentum and the maxilla is indistinctly sepaiated 

 from the submentum. 



The thoracic segments (fig. 2) are nearly twice as broad as the 

 middle abdominal segments. The chitinized upper side is rounded 

 and protects the soft ventral part as a shield. The stout curved 

 legs (figs. 4, 8) can be drawn in under the edges of this shield. 

 The coxa (ex) is large and mobile, and the femur (fe) and tibia (ti) 

 are approximately of equal size, and only half as thick and long as 

 the coxa. The tarsus (ta) is large and unguliform, without claws. 



The abdomen (figs. 2, 4) is ten-jointed; the chitinized portion 

 of each abdominal segment forming a continuous closed ring with- 

 out any soft lateral zone. The soft-skinned portion between the 

 rings is broad and the segments can be telescoped to a consider- 

 able extent. On the eighth abdominal segment is a pair of cerci- 

 like filaments (fi, figs. 4, 7) but true cerci, which always belong- 

 to the ninth abdominal segment, are entirely absent. The fila- 

 ments are two-jointed, and consist of a cylindrical, well chitinized 

 base with a few long setae, and an apical thin-skinned, cylindrical 

 piece which is entirely filled by an enlargement of the trachea. 

 These filaments are undoubtedly organs of respiration. A pair of 

 similar, but smaller formations are found on the posterior edge of 

 the prothorax (figs. 2, 3, 4), and another pair on the posterior 

 edge of the first abdominal segment. I believe that they corre- 

 spond morphologically to the filaments on the larva of Bcro^n.^ 

 which arc not direct formations from the spiracle, but develop- 

 ments from the integument near the spiracle, provided with very 

 strong tracheae. In Berosus the spiracles are small but easily seen, 

 while in Hydroscapha they are so reduced that even the spiracle 

 opening cannot be seen, and only some of the very fine tracheae 

 of the spiracles which reach out to the surface are visible. The 

 filaments of Hydroscapha resemble greatly the peculiar flattened 

 hairs at the edges of the segments (fig. 3), but whether or not 

 they really are to be regarded as modified hairs may be disputable. 

 The tenth abdominal segment is directed downwards and carries 

 two pencil-shaped appendices (pap fig. 7). Ventrally it is devel- 

 oped as a sucker (sw). 



The systematic position of Hydroscapha has been much discussed 

 but the larva has hitherto not been known, and, its discovery by 

 Messrs. Schwarz and Barber is therefore very important. 



The entire larval structure shows that Hydroscapha must be 

 referred to the Hydrophilidae as a new subfamily, the Hydrosca- 

 phinae, closely allied to the small, remarkable subfamily, Limnebii- 

 nse, which has recently been defined by A. d'Orchymont. 2 



1 Described by Schiodte in "de metamorphosi eleutheratorum obscr- 

 vationes," Naturh. Tidskr. 3 Ser., vol. i, 1862, p. 213, pi. v, vn. 



2 Contribution a 1'etudc des larves Hydrophilides, "Annales de biologic 

 acustre. Tom. vi. 1913, p. 181." 



