2S Inscriptions and lieconJs of Dees. 



series of papci's may be readily separated by means of the 

 following tabic : — 



AbJonion ferrii<riiiouf> red, wjih a black 



spilt on oacl) t.icle of seiomi segment ; 



fceconJ tiubuiaryiiial cell verv small . . hiniaculKhiA iSmitliU 

 Abdomen not ferruginous red 1. 



1. IScutellum covered with long dense i'lil- 



vou*-}ello\v hair, contrasting with the 

 surrounding pai ts ; hair of niesothorax 



black JJdvomaculalus, Ckll., J *. 



Not so -2. 



2. Abdomen distinctly metallic, with blue, 



green, or jiurple (very obscurely so iu 

 1'. obgcurij)eiiuis) ; size small or me- 

 dium 3. 



Abdomen not (nr liardly) metallic 

 (slightly si) in P. ohscnripennis) .... 9. 



3. Till rax green or blue 4, 



Thorax black ; colours of abdomen never 



brilliant G. 



4. Abdomen crimson-]. urple and brassy; 



flaj;ellum ferruginous beneath except 

 at base ; stigma rather large, amber- 

 colour ctiprcus semi2>urpurcuf>, 



Abdomen very dark purple ; thorax dark [Ckll., 5 . 



green ; (lagellum all dark pldtiwselhis, Ckll., J . 



Abdomen blue, with green tints; narrow 



and very strongly punctate ca-ndeoti/utiis, Ckll., J. 



Abdomen olive-greeu T). 



o. Second recurrent nervure joins third 

 submarginal cell before end ; area of 

 raetathorax with no sharp trans%erse 

 keel anutbilis (.Sni.;, var., $ . 



Second recurrent nervure joins third sub- 

 marginal cell at ei.d ; area of meta- 



thorax with a sharp trans\erse keel . . ciirinatulus, Ckll., J. 

 0. Abdomen black, with the broadly de- 

 jiressed hind margins cf the segments 

 olive-green; much black hairon vtrtex, 

 mesothorax, and apex of abdomen ; an- 

 tennae entirely black . ciridivimius, Ckll., 2 • 



Hind margins of abdomii.al segments 



not green 7. 



7. A\ ings strongly dusky, with purple tints ; 



abdomen hardly metallic o/iscurijienrnti, Ckll., J . 



Winjrs clear 8. 



* When describing P. Jlavomaculatus I omitted to call attention to its 

 close resemblance to P. cristatus (Smith), which is known only in the 

 female. The two may be sexes of one, but the type oi' Jlavomuculalns 

 belonged to Smith, and he did not label it cristciius. 1 have not seeu 

 the type of cm^rt^Ms; it is not at the British Museum, but probably at 

 Oxford. I visited the Oxford Univer-ity Museum, but was not able to 

 ascertain the whereabouts of the type bees presumed to be there. 



