Tests and results. This summary results from extensive tests, including from five to ten plats during three of the four years, with seedings at week intervals from March 15 to May 18. The plats were 100 feet long and one or more rows were sown the entire length of each plat for each seeding. The screened beds were carefully covered with the cheesecloth (from 20 to 30 meshes to the inch) and the bottoms of the side and end boards were carefully banked up with earth to keep out the maggot flies and flea-beetles. The differences between the screened and unscreened beds have usually been apparent almost from the time of sowing the seed, as the slight cover given by the thin cloth retains moisture and increases the temperature of the screened bed. Without such protection the surface dries out and, on heavy soil particularly, forms a crust that interferes with the little plants. As a rule the seedlings in the open beds were one or two days behind the screened plants in the time of their appearance above the ground, and the radishes under the cloth grew more rapidly and seemed more vigorous thruout the growing season. The roots grown under protection were usually more succulent and tender and more satisfactory for market, while the leaves, also, were larger, smoother and less leathery. Screening has usually kept out practically all of the maggot flies and many, though not all, of the flea-beetles. The work of either insect has been negligible on the screened beds; but on the open beds both have worked destructively. The flea-beetles were usually first to appear, and ate small round holes through the cotyledons of the tiny radish plants. Young seedlings frequently died from the first attacks of these insects; and others were so stunted by their continued work that they never produced roots of marketable size. Very soon after the maggot flies emerge from the ground, which may be during quite a period extending from early May to midJune, the females lay their eggs in some crevice or crack on the stem of the radish plant at or a little below the surface of the ground. The eggs hatch in four or five days and the larvæ maggots attack the roots. They make tunnels in the tissue or grooves on the surface which injure the radish or deform it. Their work may often be followed by decay of the injured tissue. Other pests under screens. During the past two seasons, two other pests have worked on radishes; and have apparently been worse under the screens. These are slugs," or shell-less snails, of three species, and millipedes or “thousand legs " of two species. These are not insects, but their work is similar in many ways to that of radish maggots. The snails rasp and eat the leaves, and both snails and millipedes eat holes in the radishes so that the white tissue shows through the red skin over large or small areas. Both of the creatures prefer cool, moist weather, and found the spring and early summer of 1916 and 1917 much to their liking, so that both, but particularly the snails, were more abundant and more. harmful than ever before. As the screened beds retain moisture better than open beds and the large radish leaves developed there shade the ground and make it cool, both pests seemed more abundant on such beds than in the open. The snails were destructive to many other plants, especially in gardens, such as lima beans, corn, dahlia, potato, chicory, dandelion and lettuce. They also mutilate ripening strawberries. They work mainly at night when they may be found by using a lantern or flash light and be destroyed. They may also be trapped by using slices of raw potato, which they seem to prefer to almost any other food. Attempts to poison them have not been successful as they seem very resistant to arsenicals altho they eat the poisoned baits freely. They seem to find the copper sulphate of bordeaux mixture distasteful, however, and will avoid plants coated with this spray. The remedies that may be used with fair promise of success are applying dry air-slaked lime about the plants, collecting the snails at night by lantern light, trapping them with sliced cull potatoes, and spraying with bordeaux mixture as a deterrent. The millipedes do comparatively little harm, fortunately, since no methods of control for them are known. USING THE MICROSCOPE IN MILK GRADING. * A rapid, F. H. HALL. The number of bacteria in milk has been used for more than a decade by certain milk distributing companies as a means of judging the commercial quality of the milk they received. The abundance of bacteria does not alone determine the quality of milk, by any means, but does furnish a guide by which we may judge of the keeping quality of the milk and make an approximate estimate of the time limit of its usefulness in fresh condition. Until recently this application of the bacterial count to the milk industry has been made by a method which could not give the counts until after the milk had been used or had soured; that is, the agar plate method reported conditions which were already several days past. It was useful in showing what kind of milk, from the bacterial standpoint, a producer had brought to market some time before; but it could not give notice, in time to be of much use, of any transitory change in the quality of the milk of a herd or any particular cow in it. By this method the bacteria in milk diluted with germ-free water were allowed to grow for from two to five days in a shallow glass dish containing a "culture medium," when the separate bacteria, or collected groups of them, developed "colonies" or masses large enough to be seen and counted. By the microscopic method, whose practical application we are to discuss, a small portion of the milk is spread on a small rectangular piece of glass and dried, then prepared by immersion in liquids and staining solutions in such a way as to make the bacteria visible and the whole suitable for examination under the high magnification of a compound microscope. This slide, which forms a permanent record of the bacterial condition of the milk, can be prepared in ten to fifteen minutes and the preparation examined at once, or it can be kept until some convenient * Reprint of Popular Edition of Bulletin No. 443, December, 1917; for Bulletin see p. 244. [698] time. The rapidity and simplicity of the method are sufficient to permit examination of the bacterial quality to be made before the milk is consumed or placed on sale. By this method the individual bacteria are seen and counted, whether each is separated from all others or many are united in a clump. These counts are therefore greater than those obtained by the ordinary methods in which clumps of bacteria are counted as if they were single bacteria. All of these facts give the microscopic method many advantages over the generally used plating method for judging the quality of milk produced on farms. Preliminary tests of this method made in 1913 gave Practical very promising results and it was decided to try tests made. it out under commercial conditions and in comparison with the plating method. This test was made at Hobart, N. Y., in 1914 and proved the new method at least as valuable and as reliable, if not more so, than the older one; so it was used alone for more than two years to grade the milk of the City of Geneva. For two full years of the time it was accepted by both dairymen and milk distributors as a guide to the payment of premiums for good quality milk. Tests at The preliminary tests were made in a laboratory operated by the Lederle Laboratories for one of the largest milk distributing companies of New York City. This Laboratory tests the milk received at Hobart for butter making, and that brought to three pasteurizing plants at Bloomville, South Cortright and Cobleskill for shipment to the city. At the shipping stations premiums are paid for milk with low bacteria counts, but at Hobart no such stimulus to the production of milk with a good bacterial quality was offered, and only 50 per ct. of the milk brought in there graded A, while the supply at the shipping stations averaged 80 per ct. Grade A. The grading of the milk was based on the agar-plate counts made by an assistant of the Lederle Laboratories, and the microscopic slides were made at the same time, from the same milk samples, by an assistant_bacteriologist of the Station. The gradings were made independently. During February, March and April, 1504 samples of milk from the three Grade A pasteurizing stations were examined by both methods. The ratings made by the microscopic method were found to agree with those made by the plate method on 1339 samples, over 89 per ct. When the two methods disagreed the microscopic method gave too low counts in a little more than half the cases, and the plate count gave too low counts in the others. In the following July, 426 additional samples of milk from the same stations were graded by both methods, with an agreement in 85 per ct. of the comparisons. Many of the discrepancies in both winter and summer work were due to slight variations above or below the number fixed as the limit for Grade A milk; so that most of the error in grading really falls within limits that would be considered reasonable in any experimental work; but there were occasional wide variations in both directions. These variations, small or great, so offset each other that the totals of samples placed in the two grades by either method agreed very closely. Thus out of 1504 samples examined in cold weather, 1280 were graded A by the microscopic method and 1270 by the agar plate method; and of the summer samples 212 were graded A by the microscopic method and 234 by the plate method. This is a surprisingly close agreement and is probably as good as could have been secured if duplicate analyses of the same samples had been made by the agar plate method. method So satisfactory was the agreement between the two Microscopic methods as used at Hobart, it was thought there could be no unfairness to anyone in using the microscopic method alone in grading the milk received at two milk stations in Geneva, especially as slight changes in the details of the method. promised even better results than were secured in the comparison tests. After a preliminary period of two months, the two milk distributing companies of Geneva and the dairymen who sold the milk agreed to accept the grades as determined by the new method. These grades were made to conform as nearly as possible to the three grades established by the State Sanitary Code designated by the letters A, B and C. To prevent confusion with these, the ratings given by the microscopic examination were called "good," "medium' and “ poor." Because of the fact that the individual bacteria are seen under the microscope, the numbers used as the limits between the grades given microscopically are necessarily much higher than those used in the Sanitary Code for milk of similar quality. Thus the milk with less than 1,000,000 bacteria per c. c. was graded as "good," that with 1,000,000 to 10,000,000 "medium" and that with more than 10,000,000 as poor." Actual counts were not made unless the sample appeared to be close to the limits between the grades. Usually an examination of 30 "fields" of the microscope was sufficient to place the milk in its proper grade, without counting. The samples were collected at least once weekly from each can of milk delivered at the milk stations, temperatures taken, and records. kept to show which samples were taken from night milk and which morning milk. As a check upon the accuracy of such weekly sampling, each can of milk brought in was tested every day during two periods of one week each. Reports showing the grading of each dairyman's milk were sent after each examination, and if anything of special interest developed in the milk of any patron at any time, he was informed of it by telephone. Many of the |