showers, the whole of the following day, March 21, to an amount estimated by Mr. Titus at more than thirty gallons per tree.

The temperature of the twentieth was thirty-four degrees at seven a. m. and fifty-seven degrees at noon; that of the twenty-first was forty-four at seven a. m. and fifty-two at noon, the wind from the southeast both days. Observations on this experiment continued only until the twentyfifth, but counts of the scales were made daily up to that time-three thousand scales for the four experimental trees, and two thousand and fifty for the two checks.

In this small experiment no differences of any significance were made out in the action of the insecticides, the total general effect being the destruction of approximately ninety-five per cent of the scales, and variations from this average in the individual trees being too slight to take into account. So far as any conclusion can be drawn from an experiment on so small a scale, we can only infer that a rainfall such as described, occurring at the time of the insecticide treatment, would have no appreciable effect on the action of either of the washes. The apparent extraordinary efficiency of the washes on these trees is plausibly explained by Mr. Titus, as connected with the poor condition of the trees and the probable consequent low vitality of such of the scale insects as remained alive.

EFFECTS OF RAIN AND WATER SPRAYS IN WASHING OFF DEAD SCALES.

Noticing that many scales were loosened and washed away after insecticide treatment of the trees, Mr. Titus made some careful counts from day to day of selected lots of scales on the experimental trees to determine the circumstances and the ratio of their diminution in numbers. Selecting, for example, a definite part of a branch, counting a hundred scales on it when the insecticide was applied and marking the area occupied by them, he counted them each day thereafter for several days and thus arrived at an exact conclusion as to the effect of the fluid applications and the incidental rains. Thus, on No. 1, three hundred scales counted March 3 were reduced to one hundred and eighty-eight by March 15-a loss of thirty-seven per cent. On No. 11, four hundred scales were reduced in the same time to two hundred and twenty-three-a loss of twenty-two per cent. Both these trees, it will be remembered, were sprayed with the insecticide March 3, and daily thereafter for one week with fifteen gallons of water. On No. 3, one hundred scales were reduced in eight days to seventy-two-a loss of twenty-eight per cent, this tree having been three times sprayed, with fifteen gallons of water each time. On No. 6, sprayed once with thirty gallons, the loss was twenty-five per cent in eight days; and on No. 14, receiving the same treatment except that the insecticide used was the Oregon instead of the California wash, the loss for the same period was eleven per cent. No. 21, sprayed also but once, with fifteen gallons of water, lost twenty per cent of its scales in seven days; No. 42, exposed to rains for a day and a night, lost in

five days eleven per cent of its scales; and No. 43, similiarly exposed, lost fifteen per cent.

The check trees 40 and 41, on the other hand, kept without treatment of any kind, lost within five days but four scales out of five hundred counted.

It was further apparent from observations made in the field that a brief but hard and dashing rain would detach many more scales than a light rain longer continued, and that a fine misty rain did not loosen the scales at all.

PRACTICAL CONCLUSIONS.

The foregoing described observations and experiments go to show that the Oregon wash of lime, sulphur, and blue vitriol, is a valuable insecticide for winter use in the climate of Illinois for the destruction of the San Jose scale; that its full effect will be produced in about a week; and that frequent short rains will not noticeably diminish or delay its action, even when they come within the first five days after the insecticide treatment. It is entirely harmless to any leafless tree, and hence may be freely used in winter (but in winter only) for all kinds of trees, shrubs, and vines.

The California wash of lime, sulphur, and salt, prepared as described previously, is a little less effective than the Oregon wash as a scale destroyer, and is considerably more likely to deteriorate after application if exposed to rains within the first few days. Otherwise its effects and characteristics are very similar to those of the Oregon wash.

It should be generally known that both these washes corrode brass and copper rapidly, and that consequently an iron pump may be used to better advantage in spraying them than one made in part of brass.

With respect to the comparative effectiveness of these washes and the better known whale-oil soap and coal-oil mixtures we have as yet no accurate knowledge. There is nothing to indicate, however, that the former are less effective here than on the Pacific Coast, where they seem to have been found very satisfactory in the hands of the ordinary orchardist for the destruction of the San Jose scale. Over both the latter applications they have the very great advantage that they are harmless to the tree, and that they do not endanger the crop of the following year. They are also decidedly cheaper than either the whale-oil soap or the kerosene emulsion. In our winter's work the cost of the materials for these various mixtures has been $1.12 per hundred gallons of the Oregon and the California washes; $2.80 per hundred gallons of kerosene emulsion, diluted to contain twenty per cent of kerosene; and $6.50 for the same quantity of the whale-oil soap solution, at the usual strength of two pounds to the gallon of water.

ADDITIONAL INSECTICIDE EXPERIMENTS FOR THE SAN JOSE

SCALE.

BY PROF. S. A. FORBES, STATE ENTOMOLOGIST, Urbana, Ills.

The last observations reported in bulletin 71 of the Illinois Agricultural Experiment Station were made March 25, at which time it now appears that the insecticide effect of the experimental applications made to trees infested by the San Jose scale was not yet complete, or at least not yet fully manifest. At the above mentioned date there were found on trees which had been treated with the Oregon and California washes twenty and twenty-two days previously, living scales in numbers varying from six to thirty-one per cent of those alive in the beginning for trees treated with the California wash and from one to seventeen per cent for those treated with the Oregon wash (see tables in bulletin 71).

A careful examination of these experimental trees, made by Mr. E. S. G. Titus May 12, and a systematic count of dead and living scales showed that by that time extremely few scales remained alive on any of these trees. Five hundred young scales of the preceding year were critically examined on each of eighteen hundred trees-nine thousand scales in all -care being taken to choose lots from all parts of the tree up to the terminal twigs. Of these nine thousand scales, only thirty-five were living, the ratio of living to dead varying from none at all to a maximum of one per cent-as near complete destruction as any field operation is likely to accomplish.

One half the trees on which these counts were made had been sprayed with the California wash and the other half with the Oregon wash; and sixteen of them were chosen in pairs such that the only difference between the treatment of the trees of each pair was the difference in the insecticide applied. Comparison shows that at the time these counts were made all the difference of effect between the two insecticides had disappeared, one proving finally as efficient as the other.

These eighteen trees were so selected as to represent ten variations of treatment with water, ranging from daily spraying for seven days in succession beginning with the next day after the insecticide treatment, to simple applications of water after an interval so long as to have practically no effect. On a comparison of the reports concerning the different lots of these trees, I find no evidence that variations with the water treatment made any difference with the final effect of the insecticide. The destruction of the insects was retarded in some cases by frequent and early wettings, as shown in bulletin 71, but was practically complete in every instance before May 12.

By those who have read bulletin 71 in will be remembered that large percentages of the young scales of the preceding year were dead at the beginning of the experiment in consequence apparently of the character of the weather of the preceding summer. If allowance is made for this fact and the ratio of living to dead scales is reckoned with refer

ence to those alive when the experiment began, we find that the ratio of scales still living on these eighteen trees on the 12th of May varies from none at all for five of the trees to two and seven tenths per cent for one tree, the average ratio of living scales for the eighteen trees being a little less than eight tenths per cent.

In preparing bulletin 71 I had no data in hand for a comparison of the insecticide effects of the lime and sulphur washes with whale-oil soap or petroleum, but Mr. Titus's visit of May 12 enables me to make good this deficiency. All the infested trees at this place not used in the experiment had been sprayed with whale-oil soap in March by one of my regular field parties, and an examination of five thousand scales on eighteen of the trees so treated gave a total of thirty-nine scales, a little more than twice the ratio of living to dead scales found on those which had been treated with the lime and sulphur washes.

From the foregoing it appears that certain of the statements made in bulletin 71 must be modified with reference to these later and more complete results. It may be safely said that the washes applied were extremely efficient insecticides. Even such failures to kill the scales as occurred were probably due to imperfect distribution of the spray. The action of the washes is more prolonged than I have been previously led to infer, extending evidently beyond three weeks, and although at first considerable differences were noticeable between the Oregon and California washes, we must conclude from the above report that these are differences in promptness and rapidity of action rather than in the final effect. A similar statement may be made with reference to the effects of rain as represented by the application of water to our experimental trees. This clearly has the effect to delay, but not to prevent, a complete destruction of the scale, and it is apparent that these western washes, costing $1.12 per hundred gallons, are at least as destructive to insect life as the solution of whale-oil soap, costing $6.50 for the same quantity.

The following table will give some details of observation not reported in the text. The numbers used for the trees are the same as those in the tables of bulletin 71 and the experimental history of each tree can also be found in that bulletin. It is to be understood that five hundred scales were counted for each tree May 12. The first tree of each pair was sprayed with the California wash, and the second with the Oregon wash.

OSWEGO STRAWBERRIES.

AN ACCOUNT OF EXPERIMENT WITH FERTILIZERS, AND RECORDS OF STRAWBERRY-GROWING, IN THE OSWEGO DISTRICT.

BY PROF. L. H. BAILEY, Cornell Agricultural Experiment Station.

I. SUMMARY OF FIELD RESULTS WITH THE USE OF FERTILIZER ON STRAWBERRIES.

In the spring of 1897, at the request of the Oswego County Fruitgrowers' Association, the horticultural department of Cornell University planned, and now has under way, a series of tests with different fertilizers for the purpose of determining, if possible, the one best suited to the needs of the strawberry when grown as a field crop. These experiments were begun by the late E. G. Lodeman, and they have been continued under the personal supervision of C. E. Hunn.

The three essential fertilizers, nitrogen, phosphoric acid and potash, were used separately and in combinations in different weights and seasons, careful notes being taken both as to growth of plant and yield of fruit. The fertilizers were applied to young plantations in spring after the first tillage and before the plants bloomed, a year in advance of the recorded crop. The materials were scattered alongside the row, within a few inches of the plants, and were cultivated in. The plats were located on a variety of soils, ranging from typical strawberry soil, i. e., gravelly loam, through meadow land to black muck.

In 1897, in co-operation with the association, six plats, in as many different localities, were selected and a careful line of experiments was planned.