142 Fkitz Gerbig, 



appear as broadlj- oval dark spots showing in surface view two 

 distinct regioiis : 



1. A central area consisting of an iraperforate disc of cliitin 

 occupying about one half of the total diameter. 



2. A surrounding margin formed of numerous rods of chitin 

 radiating from the central disc to the circumference of the spiracle, 

 and lying side by side so closely as to leave but very narrow slits 

 between them. These slits appear further to be crossed by nnmerous 

 transverse connections, giving the whole a lattice-like appearance. 

 Air enters between the radial bars. Seen in sections this marginal 

 lattice-work is formed of three sets of parts: 



1. Passing in a radial direction from the margin of the spiracle 

 to the central disc, but at a lower level than the outer surface 

 (and hence not seen at all in surface view), is a series of hollow 

 chitinous radial bars, irregularly oval in section, some bifurcating 

 towards the centre, while others are joined with their neighbours 

 by connecting branches. 



2. Arising from these are the series of Y-shapes upstanding 

 chitinous pillars, each of the radial bars bearing a complete series. 



3. Supported by the upper ends of contignuous Y-pillars is a 

 second series of radial bars, slightly flat-topped but wedge-shaped 

 below. These being supported by branches of neighbouring Y-pillars 

 will necessarily alternate with the lower radial bars. Further, 

 these are the bars seen in surface view, the transverse connections 

 being the Y-pillars seen from above. 



To complete the structure the Y-pillars are connected together 

 by very numerous and excessively fine chitinous threads, which 

 brauch and intercommunicate, the whole forming a close network. 



This spiracle cover appears quite incapable of closing, and the 

 arrangement seems to be a complicated form of flltering apparatus, 

 piobably also preventing the entrance of water to the spiracles 

 when submerged." 



M. Keilin (1912) hat die Arbeiten von de Meijeke und Brown 

 berücksichtigt und findet am Stigma der Larve von Trichocera 

 Memalis ähnliche Verhältnisse, wie sie von jenen Autoren auch ge- 

 schildert sind. Das Stigma, das dem von mir in Fig. 24, Taf 4 

 abgebildeten sehr ähnlich ist, besteht nach ihm aus einem verdickten 

 Stigmenmittelstück (= un bouchon cicatriciel), das durch Chitin- 

 streben (= des batonnets chitineux) mit der Tracheenwand ver- 

 bunden ist. Keilin nennt den von den Streben durchkreuzten Raum 



