in some cases hot, in others cold, at the rate of eight gallons per square rod. The results were generally satisfactory, with solutions stronger than 3%, that is, containing one-fourth pint or more to the gallon of water. This chemical was quicker in its action than was any other under trial. In twelve hours after its application all parts of the weeds above ground were browned and apparently killed. These weeds included plantain, dandelion, chicory, rag-weed and grasses. A week later they appeared the same. Two months later there was evidence of the survival and recovery of some of the plants. The dandelion roots had not been entirely killed and some new shoots were appearing. A large proportion of the grass (quack and Kentucky blue) had likewise thrown up new shoots, and some seedling weeds were growing which had germinated since the application. It was one of the best chemicals used for killing knot-grass. These results indicate that carbolic acid is a valuable herbicide and very quick in its action, but that it does not penetrate deeply enough into the soil to destroy all of the underground parts of certain weeds and that the herbicidal action is of short duration as compared with some other chemicals. There are two objectionable features associated with the use of carbolic acid. First, its odor. This is strong and to many disagreeable, and renders its use in the immediate vicinity of a house somewhat objectionable. This odor disappears soon after the application is made. Second, the insolublility in water. It is of a heavy, oily nature and does not form a permanent mixture with water. It is necessary, therefore, to agitate it frequently while the application is being made in order to insure uniformity in the distribution. WHITE ARSENIC AND SAL SODA MIXTURE This was used in the manner suggested by Shutt,' viz: white arsenic, one pound; washing soda, two pounds; water, three to nine gallons. The weaker solution proved strong enough to destroy practically all weeds. The perennial grasses were the only plants which survived the applications and most of them were killed. It was not so quick in its action as was the carbolic acid, but its effects were more enduring. Some plots treated with it in July, 1898, were still practically free from weeds in August, 1899. There are two objectionable features associated with the use of this mixture. First, the trouble of preparing it; and, second, the fact that it is a deadly poison to animal life. Neither of these is so serious as to preclude its usefulness, but both need to be taken into consideration in estimating its practical merits as compared with other herbicides. I Canada Exp. Farms, Bul. 28, p. 9, (1897). ARSENIATE OF SODA This is a chemical which is readily soluble in water. It was chosen as offering for this reason a convenient substitute for the mixture of sal soda and white arsenic. It was used at the rate of one pound in four gallons of water, and also of one pound in eight gallons of water. The latter strength proved sufficient to destroy practically all weeds. It was a little prompter in its action than was the preceding mixture, and, on the whole, was rather more effective. Its persistence in the soil seemed about equal to that of the arsenic and sal soda mixture. The objection holds against this also that it is an active poison to animals. It is also somewhat more expensive than is the white arsenic and sal soda mixture, but for operations on a small scale its greater convenience serves to recommend it for use. COMMERCIAL WEED KILLERS In addition to these compounds, trial was made of two commercial weed killers. One of these was obtained from Peter Henderson & Company of New York, under the trade name of Herbicide. It was in the form of a liquid and put up in a one-gallon can. The cost was one dollar a can. The directions were to dilute this with twenty-five gallons of water. It was so diluted and applied like the other solutions, eight gallons to the square rod. It proved very effective, its action being similar to that of the preceding arsenical solutions. It was not as prompt in action as was the carbolic acid, but fully equal to any of the other chemicals tested. At the end of one week practically all the weeds, including the grasses, were apparently dead. At the end of two months these herbicide plots were considerably cleaner than were those treated with carbolic acid, and were in fully as good condition as were those treated with any other chemical under trial. Chemical examination1 showed this herbicide to contain much arsenic associated with an alkali. It is apparently very similar in composition to the arsenical solutions already discussed, and is of course open to the same objection as a poison. The cost of the diluted solution is four cents a gallon, which is considerably greater than that of the other arsenical solutions. This greater cost is to be weighed against its greater convenience as compared with some of the others. Another preparation tested is that known as Smith's Perfect Weed Killer, sent out by Mark Smith, Louth, England. It is put upon the market in two forms. One of these is a concentrated liquid similar to the herbicide just discussed, the other is a dry powder. This latter is said to be the same thing as the liquid in a different form, and was used in our trials diluted with water according to directions. It behaved similarly to the arsenical solutions and the herbicide already discussed. Its action was I See chemical analysis, p. 147. rather slower than these during the first few days, but ultimately the results were practically alike. COST OF THE VARIOUS CHEMICALS It is evident from these results that it is possible to kill weeds by the use of chemicals. The practical usefulness of the method depends largely upon the expense. Two elements enter into this, first, the cost of the chemicals; and, second, the frequency with which the application must be repeated. The price of the chemicals will, of course, vary. The Burlington Drug Company recently gave us the following quotations on one hundred pound lots : Crude carbolic acid, 25 cents a gallon. Sal soda, 14 cents a gallon. White arsenic, 64 cents a pound. Arseniate of soda, 11 cents a pound. The cost of these as diluted and used in the proportions and amounts indicated in our experiments becomes as follows: Crude carbolic acid.-1 pint in 4 gallons of water, or 4% solution, cent a gallon or 4 cents a square rod. Arsenic-sal soda mixture.-Weaker form, 1 cent a gallon or 8 cents a square rod. Arseniate of soda.-1 pound in 8 gallons of water, 14 cents a gallon or 11 cents a square rod. Henderson's herbicide cost us $1.00 for the can, to be diluted to 25 gallons; cost of a gallon as used 4 cents, or 32 cents a square rod. Smith's weed killer cost in England 43 cents a can (dry form) to be diluted to 25 gallons. Cost for a gallon about 2 cents, or 15 cents a square rod. The cost of the liquid form is practically the same. The duty and cost of carriage must of course be added to these prices. These additions would probably bring the cost of the english article to nearly that of the american. The frequency with which it may be necessary to repeat the application of these weed killers is not fully decided by our experiments. Apparently, one thorough application each year would be sufficient of either of the arsenical solutions, or of the two proprietary compounds. It is doubtful if one application of carbolic acid would suffice, but two would quite certainly do so. It seems probable also that the effects of many of these arsenical applications will be cumulative owing to the retention of the poison in the soil. These conclusions are borne out by the experience of Prof. Nicholson of the Royal gardens, Kew, England, who writes us that the Smith's Perfect Weed Killer is used in the walks and drives of that institution and that one application yearly suffices. GENERAL CONCLUSIONS Gravel walks, drives, tennis courts and similar places can be kept free from weeds by the use of certain chemicals. Common salt can be used for this purpose, but very heavy applications are required, and when used in such amounts it is liable to be washed into the borders of adjacent lawns. Salt should always be applied in the dry form. The weeds may be more fully suppressed without such danger from washing by certain other chemicals. These are to be applied in solutions, and at the rate of about eight gallons to the square rod. Crude carbolic acid is a very powerful and quick acting herbicide. One pint in four gallons of water is usually sufficient, cost as diluted, cent for a gallon, 4 cents to the square rod. Its effects are not as enduring, however, as are those of the arsenical solutions. Various arsenical compounds are available, including arseniate of soda, a mixture of white arsenic and sal soda, and two proprietary articles. The choice between these latter becomes largely a matter of relative expense and convenience. In general, the choice should in our judgement lie beween the crude carbolic acid and the arseniate of soda. One, or at most, two applications each season of one or another of these chemicals will, it is believed, suffice to keep down the weeds. REPORT OF THE HORTICULTURIST F. A. WAUGH The work of the horticultural department during the year now closing has followed the same lines as heretofore. For reasons elsewhere discussed principal attention has been given to the study of plums. The matters to be reported upon at this time are discussed in the following order: Geography of variation in the genus Prunus in America. 2. Descriptive, historical and nomenclatural notes. II. Culture. THE POLLINATION OF PLUMS I. GENERAL CONSIDERATIONS 'Cross-pollination is advantageous to many varieties of plums, and necessary to at least a few." "For all practical purposes all classes and varieties of native plums may be regarded as absolutely self-sterile. sometimes quite self-sterile."2 I Vt. Exp. Sta. Bul. 53, p. 50 (1896). 2 Vt. Exp. Sta. Rpt. 10, p. 88 (1897). * * * * The Japanese plums are |