"The tussock moth caterpillar is generally supposed to be too hairy for birds, but this is another strained assumption. When they are common in Washington, D. C., nearly every robin seen carrying food to its young shows a telltale white fluff at the end of its bill. . . . Mr. Alan G. Dustan . . . in Canada . . . found that birds and ants are responsible for holding the insect at par in forests. When he exposed larvae to birds, the supply disappeared regularly and he credits birds with destroying half of the larvae hatching in forests. He further says that 'practically every egg mass laid above the snow line (and over go percent of them are) had been either partially or wholly destroyed by birds.' Cases of local extermination of tussock moths are recorded for the English sparrows in Massachusetts and the hairy woodpecker in Ohio."

McAtee goes on to report a case in which "starlings had locally extirpated a mixed infestation of brown-tail and gypsy moth larvae, and when E. H. Forbush was in charge of the gypsy-moth campaign for the State of Massachusetts, birds were observed to so hold the gypsy moth in check at one locality for several years that work by the State force was suspended. . . . It was almost impossible to complete certain experiments with larvae protected by netting bags because so many caterpillars were taken from the nets by birds. Sixty percent of the gypsy moth larvae used in these experiments were destroyed by birds."

The appleworm, larva of the codling moth, has also come in for attention by numerous birds. Even before the turn of the century, M. V. Slingerland at the Cornell Agricultural Experiment Station asserted that "by far the most effective aids to man in controlling the codling moth are the birds." This conclusion was reached by reason of the scarcity of intact hibernating cocoons and by the abundance of empty ones which apparently had been attacked by birds.

In New Hampshire, E. D. Sanderson reported : "Only 5 to 20 percent of the larvae survived the winter. An examination of seven trees . . . showed but 5 percent alive in the spring, 87 percent having been killed by birds, 4 percent by disease and 3 percent by cold. . . . It is quite evident that the birds, particularly the downy woodpeckers and nuthatches, are the most important enemies of the codling moth in New England.

And so the story continues. There are on record instances of commendable work by birds in the suppression of many other species of caterpillars, flies, beetles, ants, true bugs, plant-lice, and scale insects. Outstanding as these accomplishments are, they still may not represent the most important contribution by birds to man's battle against destructive insects. The cases I have cited, from the very nature of things, are conspicuous examples of the utility of birds; they are the high lights that have attracted attention. Their recital has been used to punctuate a story which may have its greatest significance, not in the spectacular, but in the day-by-day pressure exerted by birds. This effect is difficult if not impossible of measurement, yet nevertheless certain to be there.

Another consideration that has raised doubts in the minds of some who attempt to interpret the utility of birds is the realization that they feed not exclusively on insects injurious to man but (within certain limits) rather indiscriminately on whatever insects may be present and available to capture. Thus, both injurious and beneficial insects may be reduced.

An answer to that puzzling situation was well phrased by the late F. E. L. Beal in an article in the Yearbook of Agriculture for 1908: "Whoever expects to find in birds beneficent organisms working with a sole view to the benefit of the human race will be doomed to disappointment. Birds eat food to sustain life, and in their selection are guided entirely by considerations of their own. If all species of insectivorous birds be considered as a whole, it is found that they eat insects of the various species in about the proportions in which these species exist in nature. . . . It would appear that the true function of insectivorous birds is not so much to destroy this or that insect pest as it is to lessen the numbers of the insect tribe as a whole to reduce to a lower level the great flood tide of insect life."

To that statement I add that flexibility of food habits and a tendency to prey on what is most abundant and easiest to capture make the bird world a highly mobile and responsive force for the reduction of any insect that may be inordinately abundant significantly, the destructive insects are as a rule the most abundant ones.

One encounters fewer records of insect destruction by mammals than by birds a reflection, no doubt, of conditions as they exist. As a group, mammals do not exert the pressure on insect life that birds do. That is true notwithstanding the fact that North American bats are largely if not exclusively insectivorous; that moles, shrews, and certain small rodents, particularly grasshopper mice, skunks, and the armadillo, feed extensively on insects; and that many other species partake of insects frequently. Availability and abundance play an important part in determining the extent of insect destruction by the casual feeders on insects among mammals. Those same considerations, however, often determine the abundance or even survival of bats, shrews, moles, and the armadillo, which are highly dependent on arthropod food.

Relatively little is known statistically of the over-all or even local effect of mammalian predation on insects. The feeding of highly insectivorous bats is essentially indiscriminate in character. That I must stress despite the frequently proclaimed (yet unproved) prowess of these winged mammals in mosquito control. No doubt many a mosquito falls as prey to these nocturnal aviators, but a few moments spent in observing their flight maneuvers will convince one that moths, beetles, ephemerids, and other high-flying forms are more likely to be caught than the low-flying mosquitoes. Stomach examination likewise has demonstrated this fact.

Shrews and moles feed to a large extent on subterranean invertebrates, among which are the larval and pupal forms of numerous destructive beetles and lepidopterans. Earthworms, because of their abundance, also are a staple item of food. Mice of various kinds, particularly grasshopper mice and deer mice, eat many insects. They were conspicuous in their destruction of the range caterpillars in New Mexico in 1913. That insect appeared in nearly half of 56 stomachs of deer mice collected on open. range lands and, in bulk, they formed nearly a fifth of the food. Grasshopper mice collected under the same conditions indicated an even better performance, for, besides the consumption of an equal portion of range caterpillars, they had consumed even larger quantities of grasshoppers; the only vegetable food they had eaten were the seeds of Russian-thistle.

Skunks also rendered yeoman service against the range caterpillar in New Mexico at that time. On the basis of examined droppings, fully 85 percent of their food was comprised of the pupae of this insect. Late in the pupal season, the localities that showed signs of the presence of skunks would be largely free of pupae. Frequently areas Of 4 to 5 acres would have two-thirds of the silken cocoon webs empty. In a section near Maxwell it was reported that only 5 percent of the pupae remained undamaged. This, no doubt, was the result of attacks by mammals, including several species of mice, skunks, badgers, and even coyotes.

Without doubt the nine-banded armadillo present in considerable numbers in Texas, Louisiana, and Florida is our most insectivorous medium-sized mammal. Stomach examination has revealed that more than 92 percent of its food is insects and other invertebrates, a performance that places it closely behind the bats in its relation to insects. In volume of food consumed, it greatly exceeds the latter; in diversity of items eaten, the armadillo probably has no peer among mammals. One specimen, found near Ingram, Tex., had ingested at least 87 different food items (mainly insects) aggregating more than 3000 individuals.

Among the armadillo's insect food are numerous outstanding agricultural pests. Nearly 28 percent of the diet consists of the adults and larvae (white grubs) of scarab beetles. Termites, ants, and caterpillars (cutworms) constitute appreciable portions, and earthworms, millipedes, and crawfishes round out a regimen that is distinctly subterranean in origin. The location of the armadillo's food; beneath the surface tends to offset somewhat the benefits derived from its consumption. In its energetic search for subsurface food, the 'dillo pays little concern for the welfare of young plants. The result is that sprouting corn may be destroyed immediately by the armadillo in its removal of wireworms, which may kill the plant at a later time. In general, however, the character of the armadillo's food indicates an influence for good.

IN APPROACHING the subject of birds and mammals in relation to insects, one naturally thinks in terms of direct predation, the effect wrought on insect populations by reason of the food habits of the predators. That process of reasoning has prevailed ever since serious consideration has been given to the three-cornered relationship between man and injurious insects at opposing points and those natural factors that tend to lessen the intensity of this struggle. Research aimed at demonstrating and recording the effect of such predation has characterized the sciences of economic ornithology and mammalogy in this country and in Europe for more than a century.