State, the tabulated results of which were, for many years, published annually in the Reports of the Regents. In preparing these tables, the prevailing direction of the wind was computed in the then common though imperfect manner already described, and the results were as chaotic as can well be imagined. I concluded to try the new method upon them, and the results were published in the Regents' Report for 1840, accompanied by a note from the Secretary of the Board, inviting special attention to them. They were of the most satisfactory character, and when mapped showed the course of the dominant current of air over the State, with occasional deflections, dependent upon the geographical features of the adjacent country, as clearly defined as the courses of the Hudson or the Mississippi rivers. Encouraged by this, I undertook the task of collecting observations on winds over the entire extent of the United States, which was then no easy matter, as there were no such instrumentalities, to aid in the work, as are at present accomplishing so much-the Smithsonian Institution and National Observatory not being in existence, and the only collection of observations, covering any wide extent of country, was that at the Surgeon General's Office in Washington. This had been commenced under the Surgeon General, Dr. Lovell, in the year 1822, and consisted of registers kept at different military posts, and others that had been forwarded there at the request of the late Prof. James P. Espy, who was then connected with the office. None of the latter had been published, and of the former, only those for the first nine years, and embracing only from eleven to twenty posts, the number differing in different years. The rest was all in manuscript, unpublished and unreduced. My attention was called to this collection by the late Col. J. J. Abert, Chief of the Topographical Bureau, who, in 1839, invited me to visit Washington for the purpose of inspecting it. Here I was not only allowed free access to all the manuscript material in the office, which I spent several weeks in examining and reducing, but when I left, I was permitted to take home with me all the more valuable registers of Mr. Espy's collection, indeed all that I desired, and to make the requisite computations from them there. Beyond what I thus obtained, I was dependent almost solely on private correspondence for the means of prosecuting my proposed work.

"It was while engaged in slowly collecting material that, at a meeting of the American Association of Geologists and Naturalists, held at New Haven, in 1845, I was appointed a committee to report on the present state of our knowledge of the winds of North America and the North Atlantic Ocean. This greatly enlarged my field of labor, and as I knew that I could obtain material such as I wanted from many European countries, I concluded to enlarge it still further, and make it embrace the entire northern hemisphere.

[For this purpose he availed himself of all the materials relative to meteorology found in the libraries of New York, New Haven, Philadelphia, Princeton, and Washington. As much of this material was unreduced, he was obliged to spend a considerable portion of time at each of these places in the performance of this work.]

"Observations of the winds at several places in Persia, Syria, Palestine, and at Constantinople, were kindly made at my request, for a year or two, by mission

C December, 1875.

arics residing there, and forwarded to me in manuscript. Officers of the British Hudson's Bay Company were so kind as to copy for me in manuscript the entire series of observations on winds at several of their posts in the remote parts of British America-at one of them for a period of seven years. To secure observations at sea I was aided by the late Gerard Hallock, Esq., one of the editors of the Journal of Commerce, in making arrangements with ship-owners in New York, for the loan of the logs of their different vessels. I had not, however, proceeded far in this latter line of research, when Lieut. Maury commenced his labors in the same direction at the National Observatory; and his facilities for procuring material were so superior to mine that I relinquished the field to him, and relied on his published charts for the data I needed at sea, except in the latitudes above 60°, beyond which his charts did not extend.

"It was not till three years after the date of my appointment by the Association that I was prepared to report, which was at the first meeting of the American Association for the Advancement of Science, at Philadelphia, in 1848; the body which appointed me having in the mean time changed its organization and name to that just given. The report, derived from a period of over 2000 years of observation at 550 stations, contained the announcement that between north latitude 3310 and 60° there is a general current from a little to the south of west, extending entirely around the globe; but that, as those limits are approached, it gradually loses its decided character, and at the limit, on either side, all trace of any fixed direction disappears, the current at any place being controlled entirely by local influences, as illustrated in the winds of Augusta, Georgia. After passing the limit on the south, a current from the opposite direction sets in, which, as we go south, gradually assumes a more decided character, till we come fully within the limits of the trade-winds, North of latitude 60° there are indications that a uniform current that comes down from the north, in the polar regions, veers towards the west, thus establishing a third system, which breaks up at about latitude 60°.' It was while preparing this report, and by applying the improved method of investigation to the winds in the high northern latitudes, that the interesting discovery was thus made of the system of the polar winds, entirely distinct from those which prevail south of it, the physical causes of which have since been so admirably demonstrated by Prof. Ferrel, and which is now beginning to be generally recognized as a valuable contribution to meteorology.

I may here remark that when first announced all the evidence I had of the existence of the polar system of winds was derived from observations made in the northeastern portions of the American continent, Greenland, Northern Iceland, Northern Spitzbergen, and the seas adjacent; the limit attaining so high a latitude on the eastern continent that only the extreme north of Europe and Northern Siberia fell within it, and I was not able to procure reliable data from these inhospitable regions. I have, however, since obtained an abstract of the observations of Lieut. Anschu, for nearly two years, made on the shore of the Arctic Ocean, in Siberia, and valuable material from several places in Northern Finland, Southern Spitzbergen; from Kane, Hayes [and Hall], in the Greenland Seas; and also from the vicinity of Behring Strait on both sides, contributed by parties employed

in explorations for the Russo-American telegraph line. The results of all these observations, with the exception of those of Dr. Kane, at Van Rensselaer Harbor, are in accordance with the doctrine in question. And in regard to these latter, which are utterly discordant, it is worthy of remark, that while the mean direction of the wind is almost diametrically opposite to what it is at Port Foulke, only a few leagues distant, the progressive motion in the mean direction is very small, indicating local disturbance. For I have found, as a very general rule, the world over, that wherever, from local causes, the atmospheric current is diverted from its mean course, the progressive motion is reduced. Northeastern Asia merits a more careful study, and I have long made efforts to procure observations therefrom, but without any prospect of success, until 1869, when I was so fortunate as to receive from the Meteorological Committee of the Geographical Society of Irkutsk, in Eastern Siberia, an offer of co-operation. It is still difficult to obtain the requisite observations, as the region to be studied lies north of all the settled parts of Siberia, and aid can probably be had only from missionaries of the Russian church, stationed at some of the settlements on the rivers flowing into the Arctic Ocean. In respect to these localities I acknowledge aid received through the kindness of Col. Thomas W. Knox, of New York, and George Kennan, of Norwalk, Ohio.

"In the same report, above named, I pointed out and illustrated the peculiar ~'S-shaped' curves described by the wind in its mean course for the different months or seasons of the year, on both sides of the Atlantic, though I was not then prepared to fully explain them, nor did I perceive the interesting conclusions about to be deduced from them. [Illustrations of these curves are found in Plate 26, which also exhibits the graphical method of deriving from them the monsoon influences, which determine the direction and amount of their curvature. manner of computing them is explained in the Preface.]

The

"The results reached in this report, with the data from which they were derived, forming a quarto volume of 200 pages, were subsequently published in the Smithsonian Contributions to Knowledge, constituting a part of Vol. VI. This, though as perfect as the materials known could make it, and pointing out truths of importance never before recognized, was, as was soon perceived, not what it ought to be. On sending it abroad the meager filling up of portions of the eastern continent was noticed, and persons residing there kindly lent their aid in procuring material to fill them. Among these I may mention particularly Chevalier Kahnikoff, Mr. Wesselesky and Prof. Kaemtz of Russia, and Prof. Buys Ballot, Director of the Royal Observatory of Holland, from whom collectively I received records from not less than one hundred new places; and by the exchanges and collections of the Smithsonian Institution many more were added. Subsequently additional offers of aid were received from the eminent European meteorologists, Alexander Buchan, of Scotland, Dr. Alexander J. Woeikof, of Russia, Baron Meydall, and Messrs. Aguilar and Mack. In the mean time in this country, the acquisition of California, New Mexico, and Arizona largely increased the number of military posts at which observations were taken, while by the active efforts of the Smithsonian Institution there was secured a vast number of new observers in all parts

of the country, and many of them at points very remote. Lieut. Maury was also prosecuting his work on the seas, and had covered by his published charts, the entire Atlantic and Indian Oceans, the South Pacific, and all the North Pacific except a portion of comparatively small area, between the meridians of 150° E. and 165° W. from Greenwich, the chart for which was referred to by him in his latest report as 'not yet printed';-implying that it was substantially complete in manuscript, and, if so, it would seem very desirable to have it completed and published.

"In view of all these facts, and also that my original work lacked scientific arrangement, it was thought desirable to revise and enlarge it, and the Smithsonian Institution generously made appropriations to aid in the computations, as well as put at my disposal all the material at its command. The plan proposed for the new work was that followed in the present treatise, to divide the earth into 36 zones, by parallels of latitude 5° asunder, and so extending from the north to the south pole; in each of these zones commencing at the 180th meridian from Greenwich, and proceeding easterly according as observations furnished the data, around the earth to the same meridian again. Between the parallels of latitude 60° N. and 60° S. where observations are more abundant, records have been obtained from about 2000 places in North America and the West Indies, 27 in South America, 23 at islands in the Atlantic, over 700 in Europe, 206 in Asia and the East Indies, 70 in Africa, 48 in Australia and islands of the Pacific and Indian Oceans, including the extreme southerly ones of Kerguelen's Land and Heard's Island-the most southerly points where man has remained for any considerable length of time; and for over 1000 years at sea. If this area be divided into geographical squares, by drawing meridians and parallels of latitude 5° asunder, of the 1728 squares so formed, 1402 are represented in the contents of this work. The 326 vacant squares from which no observations have been obtained are as follows:

21 in North America, mostly in British America,

40 in the interior of South America,

None in Europe,

75 in Central Asia,

66 in Africa,

15 in the interior of Australia,

108 in the North Pacific Ocean, and

1 in the South Pacific Ocean.

North and south of the parallels of 60°, it is more difficult to obtain observations, and the material is therefore less abundant. Between 60° and 65° N., results are given for 57 stations, embracing a period of 316 years, mainly in Northern Russia. Further north, about 34 stations have been obtained; so that all these 36 zones are represented in the work except three, one about the north pole and two about the south, which had never been visited by man.

I had proceeded so far with the work in the southern hemisphere that, in 1859, I read a paper at the meeting of the American Association at Springfield, Mass.,

in which I showed that observations clearly indicated, and, indeed, all but demonstrated, the existence of a system of winds about the south pole, and extending from 25° to 30° from it, analogous to that which had been proved to exist about the north pole. Although the visits of explorers to this inhospitable region had been limited to periods of a few days each-too short a time for any well-defined results yet the observations disclosed the remarkable fact that while in the contiguous zone further north, and between it and the trade-winds, the mean direction of the wind was always from some point between N. and N. W., with most wonderful uniformity, far more so than in the northern hemisphere, owing undoubtedly to the less amount of land to obstruct its passage, yet out of fifteen visits by explorers to as many different points in this southern polar zone, in none was the wind from any point in the N. W. quarter, a series of coincidences without a parallel, if merely accidental, and no such system exists.

[Next, in this monograph, occur the author's remarks on the influence of difference of velocity in modifying the mean direction of the wind, which have been placed on pages 637–639, in the introduction to the Velocity Tables. Though a longer time would be desirable, the discussion is limited to observations for a period of four years, owing to the great labor and expense of making the computations.]

"The discussion of the remaining point named as belonging to the investigation, viz., the connection between the direction of the wind and the rise or fall of the barometer, may not be prepared for appearance in my new work, though it is not inferior in point of interest and practical value to either of the others. It was commenced in its present form about the same time as that of the mean direction of the wind (1836–8), and, like that, nearly simultaneously in Europe and in this country, neither party having any knowledge of what the other was doing. Inquiries had been previously instituted as to the direction of the wind which usually attended a maximum or a minimum pressure of the atmosphere, and statements had been published in England, and in this country also (?), that the former was N. E. and the latter S. W.; but the far more important question was, "What change takes place in the barometer during the continuance of different winds?" And it was to this point that the new investigation was chiefly directed. The statements just quoted may be true, but the inference drawn by some therefrom, that winds from the former point tended specially to raise the barometer, and those from the latter to depress it, was not well founded. It was as though the astronomer should conclude that the difference between the mean and true motions of a planet is greatest about midway between the apsides of its orbit, because the equation of the centre is greatest there. If winds from the west, northwest, and north tend to raise the barometer, and those from the east, southeast, and south tend to depress it, and if the wind is prone to shift its direction in the order just named, it is obvious that when it reaches the N. E. point, the barometer must show the accumulated effects of all the winds through the preceding 180 degrees, and so of course stand high, although the N. E. wind itself were neutral in its influence. To study the question properly, we need either self-registering instruments (both barometer and wind-vane), or very frequent observations; and conse