140 
In 1 89 1, when 1,874 specimens were taken from lights and 836 
from trees, the dominant species at Hghts was inversa (76 per cent.), 
hirficula, tristis, and fusca following with ratios of 10 per cent., 7 
per cent., and 5 per cent, respectively. The dominant species from 
trees, on the other hand, were hirticnla (42 per cent.), inversa (32 
per cent.), and fusca (23 per cent.), the only other species being 
gibbosa (2 per cent.). The results for 1906, when 142 specimens 
were taken at lights and 3,484 at trees, were equally discordant. 
The leading species at lights this year was inversa (54 per cent.), 
rngosa and iiuplicita following with 24 per cent, and 15 per cent, 
respectively; while the leading species in trees was implicita {y2 
per cent.), followed by ilicis and rngosa, 10 per cent, and 6 per cent, 
respectively. 
Variation of Nnnibers in Different Localities and Years. — The 
numbers of the several species vary greatly from year to year in the 
same locality, and in different localities during the same season. It 
consequently happens that the dominant species in a locality may be 
different in successive years, and that the dominant species in one 
locality may be dift'erent from that in another, within the same year. 
Collections have not been made on a large enough scale or in suf- 
ficiently continuous series to enable us to exhibit these differences in 
any detail, but the following may serve as illustrations : — 
Collections made at a street-lamp in Maywood, near Chicago, by 
O. S. Westcott,* on seventeen nights from May 9 to June 14, 1887, 
contained 798 specimens of fusca and 313 of gibbosa — fusca pre- 
dominating in a ratio of more than 25^. to i ; and collections made 
the following year at the same place by the same person, on seven- 
teen nights between June 2 and July 2, gave 73 specimens of fusca 
and 1,836 of gibbosa — gibbosa now predominating in a ratio of 
25 to I. The difference in the collection period of the two years 
was due to the difference in the weather of the spring, which was 
backward and stormy in 1888. If we compare the collections of the 
same periods for these two years — June 9 to 14 in 1887, and June 9 
to 13 in 1888 — we have 96 specimens of fusca to 82 of gibbosa in 
1887, and 29 of fusca to 1,020 of gibbosa in 1888. 
A similar comparison may be made between the contents of fre- 
quent collections from trees made at Urbana through the whole 
period of activity of the May-beetles in 1891 and again in 1906 — 836 
specimens in the former year and 3,484 in the latter. (See table 
on p. 139.) In the collections of 189T the dominant species were 
hirticnla (42 per cent.), inversa (32 per cent.), and fusca (23 per 
*"Entomologica Americana, Nov., 1888, Vol. IV., p. 155. 
