tor has been steadily at work endeavoring to perfect his method, and it is now announced that he has succeeded in eliminating those irritating substances which were so troublesome in the early preparations. From his observations Koch concluded that the affected patient gradually acquired a degree of immunity through the absorption of certain constituents of the bacilli themselves. This immunizing, however, usually occurs too late in the disease to save the patient. In order to collect these substances, whatever they may be, he dries and triturates the dead bodies of the bacilli, and from the resultant powder makes two extracts. The first of these is opalescent, and is practically identical with his original serum. He calls it tuberculin O (TO). The remaining sediment is again dried, digested with water, and then centrifuged, and this is continued until the water is perfectly clear. These later extracts contain the essential immunizing principle which he calls tuberculin R (TR), and with which such encouraging results are said to have been obtained.
Some Notes on a Dust Storm.—Apropos of our recent article on Dust Storms, we have received the following item from Mr. W. S. Jackman: "On the 19th of February, 1896 (Ash Wednesday, by eternal fitness), a remarkable dust storm descended upon the city of Chicago. As the ground had been completely covered with snow to the depth of several inches a day or two previous, it was an easy matter to form some estimate of the quantity of dust that fell. The storm began in the evening and lasted several hours. The next morning, by selecting suitable open areas that were likely to be free from eddies and local currents, an average per square yard of 1·34 ounces of dust was obtained. This was gathered by scraping up the snow to the depth the dust had penetrated—about one inch—and, after melting it, the water was evaporated and the dust dried. The amount thus collected measured 2·45 cubic inches. At this rate the quantity of dust deposited upon a square mile would be about 129·6 tons; the volume would be on the same area 4,352 cubic feet, equal to a pile about thirty-four feet long, sixteen feet wide, and eight feet deep. On being strongly heated in a clay crucible the dust turned reddish brown and lost twelve per cent of its weight. The microscope showed it to be largely composed of irregular and rounded quartz grains. As the storm began after dark, people who were unfortunate enough to be caught in it were at a loss for a time to know just what was the matter. In many instances nothing unusual was suspected until, entering their homes, their smeared and blackened countenances called forth an unwontedly hilarious greeting. After the snow melted, in many instances the sidewalks were so covered with the slimy mud as to need scraping and washing. The storm seemed to come from the northwest, and was accompanied by a moderate gale."
The Law of Mississippi Floods.—An important point in Mr. James L. Greenleaf's study in the Engineering Magazine of the Times and Causes of Western Floods is the topographical division of the country drained by the Mississippi into several large water sheds, covering a total area of 1,259,000 square miles. Although this topographical division has been wrought by natural causes, the consideration of the water sheds must be combined with that of climatic areas for the purposes of the present study. All the rivers tributary to the Mississippi show a decided tendency to low water in the autumn, the southern waters beginning to fall from a high stage in June and the northern in July. All begin to rise from low water in winter—the southern tributaries earlier, and the northern ones later. Two freshets occur during the year in each of the large tributary basins. The coincidence of the highest stage in more than two of the large branches is extremely rare, and hence it is an exception for the main river to be subjected to the enormously congested state which would otherwise result. The varying st.ages of the Mississippi accurately reflect the fluctuations of rainfall and of temperature which occur upon its tributary basins, and these are followed in detail in the author's paper. The Mississippi being virtually created by the union near one point of three large rivers—the Ohio, the Missouri, and the upper Mississippi—these three branches, of which the Ohio is regarded as the most important, naturally stamp their characteristics upon it to a very marked de-
