State of Wyoming, thus connecting the work of previous years in the Great Basin with the western part of the Great Plains. A large part of Wyoming was found to be from 1,000 to 3,000 feet lower than represented on the latest maps, and consequently to have a warmer summer climate and to belong to a more southern life zone than previously supposed. Thus the Wind River and Bighorn basins and the plains east of the Bighorn Mountains fall within the Upper Sonoran Zone instead of the Transition. This area is of considerable importance from an agricultural standpoint, from the fact that it marks the northern extension of the ranges of certain southern species of plants and animals. A special effort was made to determine the position of the boundary between the Upper Sonoran and Transition zones in the States of Nebraska, North and South Dakota, and Montana. Further south, field work was carried on in Kansas, Colorado, and New Mexico; and further west, in California, Nevada, and western Oregon.

The study of the faunal affinities and agricultural resources of the arid tropical and semitropical or Lower Sonoran zones on the table-land of Mexico has been continued during the year, with the result that material and data of great value have been accumulated which throw light upon problems presented in corresponding areas in the arid regions of the southwestern United States.

ECONOMIC RELATIONS.

Work under this head consists in the study of those habits of birds and mammals which have a direct bearing on agriculture, favorable or unfavorable, and in the preparation of the results for publication. Information is obtained mainly by the following methods: (1) By study of the habits of species in the field, both by the division staff and by voluntary agents; (2) by the examination in the laboratory of the con. tents of stomachs and crops, with a view to the accurate determination of the character of the food; (3) by study of the literature of the subject, which involves the collation of published notes from widely scattered sources.

Supplementary investigations and experiments frequently become necessary for the settlement of particular questions, such as the best methods of limiting or exterminating harmful species, or of protecting and encouraging those which are beneficial.

The bulletin on the common crow, announced last year as practically finished except for the report on the insect contents of the stomachs, was somewhat delayed by the receipt of several hundred additional stomachs, and still further by the fact that the entomologist's report on the insect food materially modified some of the conclusions, necessitating many changes in the text. It is now about ready for transmittal. This bulletin is based primarily on the examination of the stomachs of more than 900 crows, young and old, taken at all seasons of the year and from every section of the United States; and, in addition, contains summaries of the reports of several hundred correspondents, and a review of the published matter relating to the economic status of the crow. The percentage of each item of food has been determined and its importance estimated; and the insect food has been studied and reported on by specialists under the direction of the entomologist of the Department.

Good progress has been made on a similar bulletin relating to the crow blackbird, and more than 1,100 stomachs of that species have been examined. About 150 stomachs of other blackbirds were also exam

ined, as well as 545 stomachs of woodpeckers, about 200 of the crow, and a few of other birds.* The total number of bird stomachs examined during the year is 2,064.†

The collection now numbers 20,102 stomachs, having been increased during the year by the addition of 3,900. This is not only the largest number added during any one year in the history of the division, but the stomachs are of unusual importance, being mainly those of the species directly affecting the interests of the agriculturist. The reference collection of seeds and other samples of bird food has also been materially increased.

EXHIBIT AT THE WORLD'S COLUMBIAN EXPOSITION.

Considerable time and labor were expended by the division in the preparation and installation of an exhibit for the World's Fair. The principal object in view was to show the character of the work in which this division is engaged. This was accomplished by graphic illustrations showing the geographic distribution of the mammals and birds in the United States and the economic status of the species. In the former line, the great fact that animals and plants are distributed in broad belts or zones, the boundaries of which are fixed primarily by temperature, was shown by means of three large models. One of these was a miniature mountain slope on which mounted specimens of characteristic mammals and birds were so assembled as to bring before the eye at a glance the successive faunas of different elevations. Another was a large topographic relief model of the area covered by the Death Valley Expedition. On this model, and on an adjoining relief map of the United States, the life zones were shown in different colors. The models were accompanied by enlarged photographs of desert scenery, with characteristic animals and plants. The subject was still further illustrated by maps of the United States, colored to show the areas inhabited by individual species and genera of mammals, birds, reptiles, and plants.

Groups of mammals and birds, beneficial and harmful, each handsomely mounted and surrounded by its accustomed food supply, constituted the exhibit which was designed to illustrate the economic relations of the species.

The importance of bird life to the farmer was further shown by an exhibit of the food of various species of birds, each item of which was carefully labeled. Groups of mounted hawks and owls in the act of killing or eating some animal habitually preyed upon, illustrated the gain or loss which each is likely to bring to the farmer. A series of stuffed skins of the birds, mice, squirrels, and other animals which have been found in the stomachs of various species showed interesting facts regarding the food habits of owls. These are but a few of the many novel features of the exhibit, which it is believed was both instructive and practical and justified the time devoted to its preparation.

*One hundred and forty-five stomachs of the kingbird or bee martin were examined during the year, and the results, together with those obtained from twentyfive previous examinations, form the basis of a special paper on the food of the kingbird, an abstract of which accompanies this report.

This determination of the food of species by examination of the stomach contents forms one of the most important features in the economic work of the division. Most of the errors due to ordinary observation are eliminated, and a foundation of actual facts is obtained which, in connection with competent field work, insures

reliable conclusions.

FOOD HABITS OF THE KINGBIRD OR BEE MARTIN.

(Tyrannus tyrannus.)

By WALTER B. BARROWS.

The kingbird or bee martin is the largest common flycatcher in the United States, and is too well known to need any extended description. It is widely distributed during the breeding season, nesting abundantly in all the States east of the Rocky Mountains, and less commonly in Idaho, Utah, Washington, eastern Oregon, and northeastern California. It winter it migrates southward to the West Indies, Mexico, Central and South America, a few remaining along the Gulf coast and in southern Texas.

During its stay in the United States the kingbird feeds mainly on insects, which form at least 85 per cent of all its food from April to September.

The material available for a study of the food of this bird includes the published records by various naturalists of something less than fifty dissections made at different times during the last twenty years, to which are now added the results of the examination of 171 stomachs by specialists of this Department. These stomachs came from 19 different States, the District of Columbia, and Canada, and were collected during six months of the year. The preliminary examination was made by Prof. F. E. L. Beal, who estimated the percentage of vegetable and animal food and, with the assistance of the writer, made a provisional analysis of the insect material. Subsequently the entire insect material was submitted to the Division of Entomology, and the insects were carefully identified under the direction of Prof. C. V. Riley, mainly by Mr. Theodor Pergande and Mr. D. W. Coquillet. A careful examination shows beyond all question that the kingbird eats many more injurious insects than beneficial ones, and although it eats many predaceous insects such as dragon flies (Neuroptera) and Ichneumonid insects which are decidedly beneficial, yet these evils are far outweighed by the immense numbers of harmful insects regularly consumed.

Scarcely any trait of the kingbird is so well known, or at least so largely believed, as its fondness for honey bees, but the mere fact that the birds are seen lingering about the hives is not sufficient evidence that they are doing any harm to the bees. Many of our best naturalists have contended that the injury to beehives was too slight to deserve notice, and that the myriads of destructive insects which were eaten repaid the farmer or even the apiarist many times over. It has been noticed also that the bird appeared to prefer the drones. Among the 171 stomachs examined only 14, or less than one-twelfth of the entire number, contained any traces of the honey bee (Apis mellifica), while the total number of bees found was but 50; of these, 40 were positively identified as drones and only 4 were unquestionably workers. The remains of the other 6 were so fragmentary as to render impossible anything beyond the determination of the species. In several cases in which the birds had been shot near a hive no traces of bees were found in their stomachs. The results of these examinations show that not more than 1 kingbird in 12 catches honey bees, and that not more than 10 per cent of the insects caught are workers. Among the enemies of the bee may be mentioned certain insects, notably the so-called robber flies belonging to the family Asilide, which do far more damage than birds. These robber flies are large, stout,

long-bodied flies, often covered with stiff hairs. According to Prof. C. V. Riley, one species (Trupanea apivora) has been known to kill 141 honey bees in a single day. In the stomachs examined by this Department not less than 6 kingbirds were found to have eaten these robber flies, and in one case at least 6 flies were found in a single stomach. It is very probable, therefore, that the killing of these robber flies would compensate for the 4 working bees, which were all we know to have been killed by the 171 kingbirds.

Although practically insectivorous, the kingbird has long been known to feed on fruits of various kinds. Among many hundreds of complaints received by the Department during the last seven years regarding the destruction of fruit by various birds, the kingbird has been mentioned as a fruit-eater in only three instances, when it has been accused of injuring cherries and small fruits. Nearly 50 per cent of the 171 stomachs examined contained vegetable food in amounts varying from a mere trace to 100 per cent, but only 2 contained vegetable matter alone. Of the total quantity of food eaten about 12 per cent was vegetable matter, probably all fruit; the remainder consisted entirely of insects, spiders, and myriapods. No less than 22 species of fruits have been identified in the stomach contents. Three stomachs out of 24, collected in May, contained vegetable matter, mainly remains of raspberries or blackberries. In the stomachs collected in June and July were found remains of cherries, mulberries, blueberries, and huckleberries, the average amount for the month of June being 4 per cent and for July 8 per cent. In August more than twice as much fruit was eaten as in July, the average for the former month reaching 22 per cent, although the relative number of stomachs containing fruit was slightly smaller than in July. The relative, amount of fruit and insects found in the stomachs is shown in the following table:

REPORT OF THE BOTANIST.

SIR: I have the honor to submit herewith my report as Botanist of the U. S. Department of Agriculture for the period from March 8 to December 31, 1893.

Very respectfully,

FREDERICK V. COVILLE,

Botanist.

Secretary.

Hon. J. STERLING MORTON,

The former Botanist, Dr. George Vasey, was removed by death on the 4th day of March, 1893, after a period of twenty-one years' labor in the same official capacity. The untiring industry with which he devoted himself to his work can be appreciated best by reference to the résumés which were published in the Annual Reports of the Commissioner and later of the Secretary of Agriculture.

As partially indicative of the work accomplished by the Division of Botany before March 8, 1893, a list of its publications up to that date is here given. Prior to the year 1883 all official reports of the work of this division, like those of nearly all other branches of the Department, were published in the Annual Report of the Commissioner of Agriculture.

PUBLICATIONS PRIOR TO MARCH 8, 1893.

BULLETINS.

Bulletin No. 1. Report of an Investigation of the Grasses of the Arid Districts of Kansas, Nebraska, and Colorado. By George Vasey. 1886. 8, pp. 19, 13 plates. Bulletin No. 2. Report on the Fungous Diseases of the Grape Vine. By F. LamsonScribner. 1886. 8, pp. 136, 7 plates. Bulletin No. 3. Grasses of the South. A Report on Certain Grasses and Forage Plants for Cultivation in the South and Southwest. By George Vasey. 1887. 8, pp. 63, 26 plates.

Bulletin No. 4. Desiderata of the Herbarium for North America North of Mexico. Ranunculace to Rosacea, inclusive. By George Vasey. 1887. 8, pp. 15. Bulletin No. 5. Report on the Experiments made in 1887 in the Treatment of the Downy Mildew and the Black-rot of the Grape Vine; with a chapter on the Apparatus for Applying Remedies for these Diseases. By F. Lamson-Scribner. 1888. 8, pp. 113; illustrated by figures in the text.

Bulletin No. 6. Grasses of the Arid Districts. Report of an Investigation of the Grasses of the Arid Districts of Texas, New Mexico, Arizona, Nevada, and Utah, in 1887. By George Vasey, S. M. Tracy, and G. C. Nealley. 1888, 8, pp. 61, 30 plates.

Bulletin No. 7. Black-rot (Læstadia Bidwellii). By F. Lamson-Scribner and Pierre Viala. 1888. 8, pp. 29, 1 plate.