292 NEW YORK STATE MUSEUM 
k Median pair of spines of the ninth abdominal ter- | 
gite distinctly longer than any of the other tergal | 
SPINES! 20% 6 bec) cee ee eee Erotylidae 
kk Median pair of spines of the ninth abdominal ter- 
gite not longer, all the tergal spines subequal in 
henierihy 30.0 Pees oe eee Coccinellidae 
ii Body sometimes covered with fine, translucent setae 
but never with elongate black or brownish chitinized 
spines 
j Body covered with fine clavate setae 
k Ninth abdominal tergum with a pair of short, 
stout, chitinized spines.......... Hrotylidae 
kk Ninth abdominal tergum not with a pair of stout 
Chitimized spines) .e see Endomychidae 
jj Body not covered with clavate setae 
3 k Body completely covered with a white flocculent 
mass; larvae aphidivorous...Coccinellidae 
kk Body never covered with a white flocculent mass 
1 Larvae provided with abdominal prolegs on at 
least one abdominal segment 
: Chrysomelidae 
ll Larvae not provided with abdominal prolegs 
m Thorax abruptly broader than the head and 
abdomen; living in burrows of mud or sand 
intwet places. Nee Heteroceridae 
mm ‘Thorax not distinctly broader than the abdo- 
nen, if broader, only gradually so and then 
only the prothorax markedly so 
n Tergum or sternum or both with distinct 
chitinized areas, frequently prominent, 
proleglike 
0 Body usually depressed; the distance be- 
tween the prothoracic iegs greater than 
the length of the legs 
Cerambycidae 
00 Body usually cylindrical; the distance be- 
tween the prothoracic legs not as great 
as the length of the legs 
Oedemeridae 
nu Tergum or sternum never with distinct chiti- 
nized areas 
0 Ninth abdominal tergite with two or more 
Short cuticular spines 
p Apex of the abdomen obliquely trun- 
cate; antennael with three segments 
Cioidae 
1JIn counting the segments of the antennae, it is frequently difficult to 
determine whether the enlarged globular basal portion should be counted 
4S a segment or not. Some writers have considered it Simply as a pro- 
tuberance of the head, while others have looked on it as the first antennal 
segment, In this table it is considered as a distinct segment. 
