No, 39] ODONATA OF CONNECTICUT 29 



The mid-dorsal line is likewise produced and sometimes orna- 

 mented with very large hooks. At the tip of the abdomen there 

 may be found either short, spine-like projections or long, leaf-like 

 gills. 



It should be remembered in studying dragonfly nymphs that 

 there is more variation in the younger stages of a species than is 

 usually found in the adult and that in species the adults of which 

 are closely allied, there is no feature known which will put the 

 nymph in its proper place. It therefore behooves the student, if 

 he desires a more accurate determination, to rear the specimen 

 considered, whenever the individual belongs to a group of closely 

 related species. Characters are now available, however, which 

 enable one to refer most species to their proper families and 

 genera with certainty, and it is probable that continued study will 

 show constant characters of sufficient importance to enable separa- 

 tion of the more closely related species, in the earlier stages. 



How TO Distinguish the Sexes 



Separate keys to the sexes are found in many places below, and 

 it is important to know the male and female in order to make 

 satisfactory determinations. The male is easily recognized by the 

 presence of accessory genital appendages on the ventral surface 

 of the second abdominal segment. This character is always pres- 

 ent and easily found. The anal or terminal appendages of the 

 male consist either of two upper and two lower, or two upper and 

 one lower of characteristic shape. The males of some groups 

 possess an anal triangle (Fig. 4, b), a secondary character. 

 Females usually have a prominent ovipositor, though this is some- 

 times modified and frequently reduced. There are no accessory 

 genitalia and the anal appendages consist of simple cerci and 

 larger ventral processes. 



Variability in the Odonata 



There is sometimes considerable variation in the Odonata, both 

 in color and size. Specimens described from Connecticut may be 

 smaller or larger than those found in other localities, but as a 

 rule this variation should be within five millimeters of the dimen- 

 sions given. Occasional examples will naturally occur that may 

 be stunted from lack of food or, having found conditions more 

 favorable, may be larger than those described. Even wing veins 

 show variation in number, especially cross-veins, which is to be 

 expected, since this part of the dragonfly has undergone rapid 

 evolution. It is well known that dragonflies change color as they 

 become older, the brighter colors being replaced by duller hues, 



