405 



33. Antenna with 6 joints (PI. XXX, Fig. 4) 



(renus inco'tus D (p. 533) . 



— Antenna with 5 joints Chirononius flavus (p. 474). 



34. The four eenti-al teeth considerably paler than the lateral teeth, 



rounded apieally, the outline of the four together forming a con- 

 vex line, first lateral tooth beyond these distinctly longer and 

 broader than the centi'al pair 35 



— Central pair of teeth much stronger than any other pair, or labial 



teeth not as stated above 36 



35. Third tooth from median line (first dark tooth) very distinctly pro- 



jecting beyond the anterior transverse line of the second tooth. . . 

 Metriocnemus knahi. 



— Tooth mentioned above not projecting farther forward than the an- 



terior transverse line of the second tooth (PI. XXIX, Fig. 20) . . . 

 Orthocladius sp. C (p. 531). 



36. Central pair of teeth with a distinct shoulder, the second tooth 



fused with first (PI. XXIX, Fig. 16) 



Ortlwcladius nivoriundus (p. 525) . 



— Central pair of teeth without a distinct shoulder 37 



37. Central pair of teeth nearly twice as broad as the next pair, the 



latter longer than third pair. . .Metriocnemus lundbecki (p. 498). 



— Central pair of teeth not twice as broad as next pair, the latter 



not longer than third pair CMronomus nigricans (p. 434). 



The foregoing key is framed to include the previously described 

 North American larvae with the exception of those described but un- 

 identified, and is not intended to serve as a guide to the separation of 

 the species in the decerns group. There are several very closely al- 

 lied species in this group which it will be necessary to rear in consider- 

 able numbers, and any careful student with time to devote to the 

 work should find some interesting problems in differentiating the spe- 

 cies in the larval and pupal stages. The species of this group all have 

 red larvcT with long respiratory organs on the sides of the eleventh seg- 

 ment, and probably there are in all more than half a dozen closely 

 allied species which are much more readily separated in the imaginal 

 stage than in either the larval or pupal stages. But few reared speci- 

 mens of this group are available for study here, and therefore I make 

 at present no attempt to associate the species in their different stages. 



The form of the labial plate has been used as a convenient means 

 of separating the species and is generally very constant in form in indi- 

 viduals of the same species. Occasionally, however, aberrant ex- 

 amples occur, possiblv due to injurv, and two of these are figured 

 herewith (PI. XXIX." Fig. ii. and PI. XXXVIII, Fig. lo). 



