WHO would ever suppose that such coarse, heavy stuff as rusty barrel hoops, iron filings and chips, lathe turnings, and other scraps of the machine shop would enter into the manufacture of the most tenuous and imponderable substance on earth? Of course the initiated know that hydrogen may be made by treating iron with sulphuric acid, and that this is the way that the gas with which balloons are inflated is commonly made. It is the sulphuric acid that contains the hydrogen; the iron serves merely to take up the sulphur and oxygen that form the. bulk of the acid, thus liberating the hydrogen. Since in sulphuric acid there is but one part by weight of hydrogen to forty-eight of sulphur and oxygen, a large quantity of acid is required to produce enough hydrogen for an ordinary balloon, and when it comes to such huge airships as the Akron with which Mr. Vaniman intends to attempt a transatlantic flight this fall, the amount of acid required is enormous. The gas bag of the Akron is 258 feet long, and 47 feet in diameter, with a capacity of 400.000 cubic feet. As the average street blocks of New York measure 250 feet and the width of Broadway downtown is but 40 feet from curb to curb, some idea of the huge bulk of the balloon can be obtained. To fill this huge gas bag, a large plant was built at Atlantic City. Nearly 80 tons of scrap iron was carted to the plant and 100 tons of sulphuric acid in large casks, containing, over 16,000 pounds each. These materials were capable of producing 450,000 cubic feet of hydrogen gas. An equal quantity of coal gas would supply an average town of 10,000 inhabitants for three days, and if the. gas were fed to an ordinary five-foot gas burner, it would take more than ten years to exhaust the supply. For five days and nights the plant operated continuously, generating hydrogen, which was carried directly into the huge gas bag. The bag is made of three. thicknesses of cotton cloth, gummed together with thin layers of rubber, and its total weight is considerably over two tons. This, of course, is the weight of the bag in air when not inflated. It is always confusing to speak of weights when referring to a balloon. It is difficult to conceive that the hydrogen in the balloon has any weight, and yet if the balloon were in a vacuum it would crash to earth with a tremendous impact, for not only does the envelope weigh many thousand pounds, but the gas itself would weigh over a ton. The accompanying drawings, which are somewhat diagrammatical, will serve to illustrate the process of producing the gas for the Akron One is a true plan view of the plant, and the other a distorted elevation, partly in section. Those who are acquainted with the usual methods of generating hydrogen for balloons will notice a number of innovations introduced by Mr. Vaniman, whereby there is no interruption in the generation of the gas. As indicated in the plan view, there are four generator tanks A, which are made of wood with all the iron parts well coated with pitch to prevent the sulphuric acid from attacking them. These tanks are partly filled with scrap iron. The sulphuric acid is fed into the tanks from one of two large reservoirs B. Over the reservoirs is a track C. on which the sulphuric acid casks D are supported. In order to prevent too rapid a generation of gas and choking of the tanks with ferrous sulphate, the sulphuric acid is diluted. For every part of sulphuric acid, eleven . parts of water are used. While this mixture is being prepared in one of the reservoirs, -the other is tapped, and the dilute sulphuric acid takes the course indicated by the arrows to the bottom of the generator tanks. It may flow directly into any one of these tanks or may be made to flow from one pair of tanks into the other. The latter is the more. economical method, as it insures complete action of every par-tide of the acid. The gas generated in the tanks rises to the top, where it is trapped by gasometers E. and flowing into a common chamber II, passes down to the washer G. Here it is forced to make its way upward through a series of perforated plates, while a spray of water flows downward through the same plates. Thence the gas passes through four tanks H, the first containing coke, the second potassium permanganate, the third caustic soda, and the fourth calcium chloride. The first two serve to purify the gas of such materials as arsenic, sulphur and phosphorus, which are apt to be picked up from the iron. The other two tanks serve to remove all traces of moisture from the gas. Hydrogen is an odorless, colorless gas, and it would be impossible to detect leaks in the balloon were not some means used to impart an odor to the gas. The “perfume” commonly used is muronine. This is placed in a sponge as indicated at I, and thence the gas is fed directly into the balloon through the tube J. The “perfume” used has a most penetrating sickish odor, that can readily be detected, no matter how small the perforation in the bag through which the gas is escaping. One of the advantages of the arrangement shown in the drawing is that when it is desired to charge one of the generators with fresh scrap iron it may be cut out of the system completely by turning off the. pipe conducting the sulphuric acid to it and also placing a clamp on the rubber hose connecting the gasometer of that particular generator with the- chamber F. The spent liquor flows from the generators through traps K to a trough L, which leads to a large trench or drain. Heretofore the purifying and drying tanks have been filled with coke in which the various cleansing chemicals were sprinkled, the purpose of the coke being to prevent the materials from clogging. This method has caused much trouble because the materials would slowly gravitate to the bottom of the tanks, choking the flow of gas. Whenever such a condition arose, it was necessary to shut down the entire plant and clean out the tanks. Mr. Vaniman has adopted a different method. He uses a set of trays of copper netting secured to iron straps, as indicated at M in the drawing, and on these trays the purifying and drying materials are placed. Thus the mass is kept in a porous condition, through which the gas can easily percolate, and in case of any trouble the entire set of trays can be lifted out bodily. The Trade Winds of Porto Rico TAR. O. L. FASSIG, of the United States Weather -L' Bureau, is one of America's few able climatog-raphers, and is well known for his exhaustive treatises on the climate of Maryland and of the city of Baltimore. For the past two years he has had charge of the Weather Bureau station at San Juan, Porto Rico, and he has of late published a number of useful and timely monographs on Porto Rican climate. The latest, which appears in the Monthly Weather Review, deals with the trade winds. Although the trade winds of the northern hemisphere are commonly supposed to blow from the northeast, this is hardly their prevailing direction in the West Indies. In Porto Rico their general direction, taking the year as a- whole, is between east and southeast, tnough the percentage of northeast winds increases during the winter. As to diurnal variations, the prevailing direction during the daytime (9 A. M. to ] 0 P. M.) is east, while during the night, after 10 o'clock, it is southeast. Tliis applies to the year as a whole, but during July and August the prevailing east wind extends far into the night; viz., to 1 or 2 o'clock in the morning. The average hourly velocity of the trade winds- for the year is a little under 11 miles, the monthly values ranging from 8 miles an hour in October to 13 miles in July. On an average these winds are lightest about sunrise and strongest about 2 o'clock, at San Juan. The trade winds of course never rise to the force of a gale, and • the few high winds that occur in Porto Rico are generally due to the passage of tropical cyclones. Judging from the movement :Jf clouds over San Juan, the average depth of the trade-wind current in this region probably does not exceed 10,000 feet. The upper clouds—the cirrus forms—have a nearly uniform movement from the western quadrant; that. is, they show that the upper air current is opposite in direction to the trades. The trade winds contribute greatly to the good health and comfort of the inhabitants of Porto Rico. In fact, it is the presence of the steady, moderate flow of the trades that takes this island out of the class of areas in the tropics to be avoided by white men, and makes it for more than half the year a very desirable place of residence for business as well as for pleasure. The utilization of the trade winds as a source of motive power is gradually receiving prop ;r recognition. and the number of wind-mills is increasing. The south side of the island is comparatively dry, and the rainfall is irregularly distributed through the year, necessitating extensive irrigation operations; while on the north side the drainage problem is almost equally important. In both these operations wind-driven pumps might be used to great advantage. Nature's Parachutes CAREFUL examination has been made of the heads of Canada thistle-downs, in order to determine their effectiveness as parachutes, carrying the seeds of the plant to great distances through the air. The results of this examination are quite remarkable. Calculation shows that a thistle-down starting from an elevation of 20 feet, in still air, would require two-thirds of a minute to reach the ground. With a wind blowing 20 miles an hour it would be carried, on the average, about a fifth of a mile. The total surface exposed to the air in an average thistle-down is, on account of the great number of hairlets, a little more than one-third of a square foot. Another well-known and very beautiful example of nature parachutes is furnished by the light silken threads with the aid of which the little gossamer spider makes long aerial voyages