BY R. J. TILLYABD. 421 



Type in Mr. John Mitchell's Collection. 



Locality. — Burnt shale fi'oui the embankment of the railway at Mere- 

 wether Beach, near Newca,stle, N.S.W. 



The specific name indicates the burnt condition of the rock on which the 

 impression was found, with the addition of a diminutive in allusion to the 

 small size of the wing (Latin, itstua = burnt). 



The discovery of the two fossils described above leaves the known history 

 of the Hemiptera in late Palaeozoic times in a verj' interesting condition. True 

 Homoptera belonging to the Division Auchenorrhyncha have already been dis- 

 covered in the Tipper Permian, viz. Permoscarta, belonging to the Scythiop- 

 terickie, anil Permofulyor representing the raonotyjiic family Permofuhjorickie; 

 the former was found at Newcastle. N.S.W., the latter at Belmont, at a some- 

 what higher horizon. Tlius we now know four very distinct types of Upper 

 Permian Hemiptera from Australia, if my determinations of the systematic 

 positions of these fossils are correct. To tliese may be added the two known 

 Upper Permian genera from Russia, viz. Prosbole and Scytinoptera. We 

 thus get the following classification for the Order in TT])per Permian times: — 



Suborder Palaeohemiptera. 



A single family, Proabolidae. 



Genera Prusbole (Russia), MiteheUoneura (Australia). 



Suborder Homoptera. 

 Division Auchenorkhyncih. 

 Family Scytinopteridae. 



Genera Scytinoptera (Russia), Permoscarta (Australia). 

 Family Permofulgoridae. 



Genus Pernio fidgor (Australia). 



Division Sternorrhyncha. 



Family Lophioneuridae. 



Genus Lophioneiira (Australia). 



The Suborder Heteroptera had apparently not yet been evolved. As I had 

 previously pointed out, in dealing with the Upper Triassic family Bufistaniidae, 

 which is the oldest known type of true Heteroptera, their origin is to be looked 

 for in the Palaeohemiptera, from some type a little earlier than Prosbole itself 

 (Tillyard, 1918, p. 589). The discovery of the genus MiteheUoneura shows 

 that the ancestral type there postulated was probably present in Australia in 

 Upper Permian times, just as the ancestors of the Upper Triassic Homoptera 

 of Australia were already present there at that same period. 



The fact that the Homoptera were already sjiecialised into their two main 

 Divisions in Upper Permian times indicates a considerably earlier origin in 

 geological time for the Order Hemiptera as a whole. This indication is borne 

 ■ iut by the occurrence, in the tapper Carboniferous of Commentry, of a fossil 

 forewing, left unclassified by Handlii-sch (1908, p. 325, and Atlas, PI. xxxiv., 

 fig. 1) '.rhich Professor Lameere and myself both agree, after studying the 

 specimen itself, should stand close to the earliest beginnings of the Order, but 

 definitely witliin it. This fossil is Dictyocicada antiqua Brongniart. We are 

 thus led to envisage the Order Hemiptera as a highly specialised but exceed- 

 ingly ancient offslioot of some more generalised Upper Carboniferous Order, 

 which already possessed the stegopterous or roof-like manner of folding the 

 wing-s. This Order must surely be the Protorthoptera of Handlirsch, in which 

 so manj- diverse types are known to have occurred that they may well fonn 



