504 ANNUAL REPORT SMITHSONIAN INSTITUTION, 1914. 
light. Or possibly the birds feel safer on the water at night and are 
in the habit of remaining there in a flock during a major portion of 
the night when uneasy. 
During the night of July 5 and 6, 1907, I camped on Partridge 
Island, in Lake Superior, about 1 mile from Gull Island, where a 
colony of gulls were breeding. In the latter part of the night, just 
before dawn, I heard the cries of gulls ftying overhead. The night 
had been very dark as the sky was clouded. 
Some observations on the nocturnal activities of herring gulls 
have been reported by Schuster.1. He noted these birds feeding on 
the river Mersey at Liverpool. Large quantities of food, thrown 
into the river at night from ships, are stated to be responsible for this 
nocturnal feeding. The gulls are described as flying and feeding 
silently. 
Various writers speak of gulls ‘‘roosting” at night, and my captive 
culls apparently sleep during the night, asa rule. It seems probable 
that gulls usually rest during the night, except during the breeding 
season or when food is especially available at night. It is also prob- 
able that gulls are not active when the darkness is intense. 
XII. VARIABILITY AND MODIFIABILITY IN BEHAVIOR. 
According to Herrick,’ whose conclusions are in general supported 
by my own observations: 
The life of birds is one of instinct irradiated by gleams of intelligence. Their 
mental faculties exhibit a wide range of gradation from excessive stupidity to a fair 
degree of intelligence with strong associative powers of things with ideas. 
In my study of the herring gull I have been especially interested in 
attempts at determining the extent of the ‘‘gleams,” a fascinating 
but very elusive topic. The resourcefulness which animals show in 
new situations and the extent to which their behavior may be modi- 
fied by new conditions may be considered fair criteria of their intelli- 
gence. Variability in behavior, however, has some bearing on the 
problem of intelhgence. We must recognize perhaps two types of 
variability in behavior which do not indicate intelligence as it is 
commonly understood. They may even tend toward confusion in 
our analyses of behavior, as acts which may seem to indicate resource- 
fulness or adaptiveness may be only variations in stereotyped be- 
havior. 
It is to be expected that so-called pure lines or strains may be 
found among the behavior characteristics of a species as well as in 
other characters. Some of these strains may possibly be the result 
of or be accentuated by such segregation as is afforded by separate 
1 Schuster, W., Mowen als Nachtvégel. Zool. Beob. Frankfurt a. M. Jahrg. 47, No. 3, S. 79, 1906. 
2 The home life of wild birds. Revised ed., 1905, p. 212. G. P. Putnam’s Sons. 
