506 ANNUAL REPORT SMITHSONIAN INSTITUTION, 1914. 
will hold together in its tree location. Of course it may be said 
that tree nesting by the herring gull may be due to a so-called latent 
instinct which appears when persecution compels the bird to seek 
a safer place for its nests. We have, however, no reason to believe 
that instinct behaves in inheritance differently than other charac- 
ters. Our knowledge of the laws of inheritance does not furnish 
any basis for thinking that an instinct for tree nesting can exist for 
long periods of time in a species that has another habit, without itself 
appearing. 
Herrick! seems to consider tree nesting for the herring gull to 
be a variation without much significance. He found a small per- 
centage of gulls nesting in trees at a height of from 6 to 10 feet. In 
his judgment, this position affords no protection to the birds. I 
have seen no nests above the ground, though trees and bushes cover 
most of the ground on all but two of the islands where I have seen 
gulls breeding. 
To me such tree nesting as Audubon? described suggests real 
resourcefulness, but we know too little about it to be warranted in 
making any generalizations concerning the intelligence it may 
involve. In some colonies tree nesting may possibly be the inherited 
habit or instinct of certain strains or ‘‘pure lines”’ of herring gulls. 
The nestling offspring of tree-nesting gulls are reported as remain- 
ing in the nest when observed, though gulls of the same age on the 
ground would never be found in the nest, but would always be hiding. 
That tree-nesting juvenals do not leave their nest until they are able 
to do so without injury is probably due to a realization of the danger 
involved in such an attempt. This remaining in the nest under 
such circumstances may possibly be regarded as intelligent behavior 
of perhaps a low order which prevails over any blind instinct to 
leave the nest to hide when intruders appear. Just how much this 
behavior is tied up with instinctive activity is of course beyond our 
knowledge. 
The promiscuous feeding of juvenal gulls at Gravel Island appeared 
to me to bea variation from the probably usual habit of parents 
feeding their own offspring. Unfortunately, we lack data for estab- 
lishing the extent of this variation. It could easily be the conse- 
quence of the congested life on the island. I have noticed that 
juvenal Wilson’s terns seek food of any adult that may happen to 
come near them with food in its beak, but all of my observations indi- 
cate that the parent tern probably feeds its own offsprmg. The 
gull must go through the somewhat complicated process of re- 
gurgitation, which seems far from voluntary. Large numbers of 
1 Herrick, F. H., Nests and nest building in birds. Part 2, Journ. Animal Behavior. July-August, 
1911. Vol. 1, No. 4, pp. 244-277. 
2 Audubon, J. J., Ornithological biography. Edinburgh, 1835., vol. 3, pp. 588-589. 
