chance to feed upon the various cultivated plants grown in bulk by the farmer ur gardener; many of these insects gradually desert their native host plants and to a greater or less degree change their habits, including in their fare the more succulent and easily found food.

HISTORY

The existence of the western grass-stem sawfly was first made known in 1890 when Mr. Albert Koebele reared adults from larvæ that were mining in the stems of native grasses growing in the vicinity of Alameda, Calif. During the next year, 1891, the species was described under the name of Cephus occidentalis by Messrs. Riley and Marlatt, from a series of individuals reared by Mr. Koebele and also

FIG. 1.-Western grass-stem sawfly (Cephus cinctus): Adult female. Much enlarged.

from cotypes that had in the meantime been collected in Nevada and Montana. In connection with this description the prophetic suggestion was made that: "The economic importance of this species. arises from the fact that it may be expected at any time to abandon its natural food-plant in favor of the small grains, on which it can doubtless successfully develop."

Nothing more was heard of this sawfly until 1895, when the late Dr. James Fletcher, Entomologist to the Dominion of Canada, swept adults at Indian Head, Northwest Territories, on July 5. He believed it to belong to the European species, Cephus pygmaeus L., and under

1 Koebele, A. Notes. In U. S. Dept. Agr. Div. Ent. Insect Life, v. 3, p. 71, 1890.

2 Riley, C. V., and Marlatt, C. L. Wheat and Grass Saw-Flies. In U. S. Dept. Agr. Div. Ent. Insect Life, v. 4, p. 168-179, 1891. (See p. 177-178.)

is name it was mentioned in his report for 18961 with the further atement that wheat straws containing Cephus larvæ had been sent by Mr. John Wenman of Souris, Manitoba, who stated that the indone by them was very slight. Nevertheless the prophecy of five ars before had been fulfilled, since these grass feeders actually had tacked small grain.

ry

In 1902 Dr. Fletcher reported, in a private letter, that he had found e larvæ numerous in grasses in the Northwest. In 1905 and 1906 r. G. I. Reeves, an agent of the Bureau of Entomology, noted the work the larvæ in various grasses, chiefly Agropyron sp., in Wyoming id the Dakotas, and in 1906 the same observer found the larvæ tacking wheat sparingly near Kulm, N. Dak.

FIG. 2.-Distribution of the western grass-stem sawfly in the United States.

August 31, 1907, Mr. E. O. G. Kelly, then an agent of this bureau, oted a few wheat straws near Minot, N. Dak., that had been burowed by the larvæ of Cephus.

In 1908 Messrs. F. M. Webster and G. I. Reeves found the larvæ of Cephus working in grasses in the Willamette Valley in Oregon. In the ame year Dr. Fletcher again called attention to this insect, stating hat in the previous autumn it had appeared in central Manitoba and n the southeastern part of Saskatchewan in much more serious numers than ever before, and that the quantity of broken straws in the ields was causing the farmers some alarm. Mr. Norman Criddle of Aweme, Manitoba, a close observer and practical farmer, wrote to Dr. Fletcher that this fly had increased considerably during the ast year or two, and was turning its attention to wheat and rye.

1 Fletcher, J. Report of the Entomologist and Botanist, 1896. Can. Dept. Agr. Exp. Farm, 1897. (See .229-230)

August 20, 1909, Prof. H. B. Penhallow reported from Sherwood, N. Dak., that he had examined about a hundred fields from Minot, N. Dak., north to the boundary line and several miles into Canada and had found larvæ present in every field but one. He estimated the damage in these fields as ranging from 5 to 25 per cent of the crop, but spoke of one field about 27 miles east of Sherwood where the damage was said to have exceeded 66 per cent. R. W. Sharpe reported similar damage in the Red River Valley, near Fargo, N. Dak.

During 1911 and 1912 the writer found the species occurring freely in the native grasses in various parts of Utah, and as occasion offered

FIG. 3.-Plants of Elymus condensatus growing along the railroad right of way. The natural habitat of the western grass-stem sawfly in Utah.

the life history of Cephus was learned. Most of the facts in this paper are the result of this study. (Fig. 3.)

During the years 1913, 1914, and 1915 the writer has found this sawfly almost universally distributed over the Dakotas, Minnesota, Iowa, and Nebraska, feeding in Elymus, timothy, and Agropyron at Elk Point, S. Dak., in Agropyron tenerum near Chamberlain, S. Dak., in timothy at Edgeley, N. Dak., in Bromus inermis near Merricourt, N. Dak., in Elymus canadensis at Shakopee, Minn., in practically all these grasses near Sioux City, Iowa, and in wheat, timothy, and Elymus near Minot, N. Dak. It seems to have little choice in the various native grasses and is ready to attack any of the cultivated

sorts provided the stem is sufficiently large for the larval gallery. As a rule, the larger, more robust stems are chosen for attack, especially in cultivated grasses such as timothy and Bromus. Blue grass and similar slender-stemmed species appear to be immune. It is a little surprising that a minute examination of Stipa viridula from New Mexico developed the fact that none of the stems of this robust grass were infested. This Stipa was gathered in northern New Mexico, growing in almost the same latitude as the Elymus condensatus near Pinto, Utah, where the fly abounds.

August 25, 1916, the writer, then at Pierre, S. Dak., received instructions from the Bureau of Entomology to visit Bottineau County

FIG. 4.-Wheat field of Thomas Yeam, near Souris, N. Dak., showing heavy damage done by the western grass-stem sawfly in 1916.

in North Dakota and investigate injury to wheat. It was believed locally that the Hessian fly was responsible for the damage that was being done. A very superficial examination of the injured fields proved beyond a doubt that the Cephus was present in large numbers and was doing an immense amount of mischief. Every field was infested, not only in Bottineau County, but in the adjoining counties of Benson, Pierce, McHenry, and Rolette. Near Souris, a few miles south of the Canada line, a large field of wheat on the farm of Thomas Yeam was fairly carpeted with the "straw-fallen" grain. (See figs. 4 and 5.) The loss from Cephus injury in this field was estimated at 60 per cent or more. Six feet of drill row here were taken at random and examined plant by plant. Forty-eight infested stubs were found,

an average of eight to each foot of drill row. This would mean 150 to the square yard or about 726,000 larvæ to the acre. Higher counts were made later in this same field, so the average may be larger than stated. During April, 1917, Mr. Yeam's field was again visited and a random square yard marked out and counted. Two hundred and sixty-nine infested stubs were taken from this yard, which would mean more than 1,300,000 larvæ to the acre. Fifty of these stubs were opened and 47 of the imprisoned larvæ that had spent the winter within the straw were found to be normal and very much alive. The proportion of living individuals among the hibernating larvæ seldom falls below this ratio.

FIG. 5.-Bird's-eye view of wheat in Thomas Yeam's field, Souris, N. Dak. Ninety per cent of these fallen wheat stems have been mined by the western grass-stem sawfly.

The spring of 1917 witnessed a peculiar condition of things in Bottineau and the adjoining counties of North Dakota. The dry weather hindered the growth of both grasses and grains, so that when the adult Cephus began to appear in June there was almost no opporportunity for oviposition. Stems of Bromus from chance sods growing among wheat and on waste ground were filled with eggs. Young plants of spring wheat that had barely begun to joint were attacked and often contained as many as three and four eggs placed in the stem close to the ground. With a few strokes of the net 136 adults. were swept from young wheat, so numerous were the flies at that time. In spite of the unfavorable oviposition conditions of that spring, the