was written for readers familiar with only elementary mathematics, but it is doubtful whether the effort to simplify the treatment has made the subject intelligible to those who are not specialists in the field.

BIGELOW, F. H., Report on the International Cloud Observations, May 1, 1896, to July 1, 1897, 4°, 787 pp., many illustrations and plates. Published as volume II of the Report of the Chief of the Weather Bureau for 1898-99. Government Printing Office, Washington, 1900. This report is very technical and mathematical in parts and doubtless is the most complete treatment in English of considerable portions of the whole science of meteorology.

HANN, JULIUS, Lehrbuch der Meterologie, 4°, XI + 642 pp.; first edition, Leipzig, 1901; second edition, Leipzig, 1906.

HANN, JULIUS, Handbook of Climatology, Part I, General Climatology, translated from the second German edition by R. De C. Ward, 8°, XIV + 437 pp., 11 figures. The Macmillan Company, New York, 1903. This work covers volume 1 of the German edition, is complete in itself, and is the standard work on the subject. (Out of print.)

HANN, JULIUS, Handbuch der Klimatologie, second edition, 3 vols., 404, 384, and 576 pp., Stuttgart, 1897; third edition, Vol. 1, XIV + 394 pp., Stuttgart, 1908; third edition, Vol. 2, XII + 426 pp., Stuttgart, 1910.

CHAPTER IV.-SOURCES OF DATA AND

METHODS OF COMPILATION

METEOROLOGICAL OBSERVATIONS IN THE UNITED STATES

The earliest record of systematic daily weather observations made in America, of which we have any knowledge, is that for the years 1644-1645, made by John Companius, a Swedish clergyman, living near Wilmington, Delaware. From that time until the early part of the nineteenth century, a few men possessing scientific curiosity were attracted to the study of weather conditions and made daily observations for varying lengths of time, but rarely or never contemporaneously. These observations were too meager in scope and of too short duration to be of general utility.

In 1814 the Surgeon General of the United States Army made it the duty of each hospital surgeon and director of a department to keep a diary of the weather. No attempt was made to systematize the observations or to standardize the instruments with which they were made, and the results were necessarily vague. This service was continued until the outbreak of the Civil War. A part of the results were published in three volumes entitled Meteorological Registers; the last, published in 1851, covers the period from 1831 to 1842.

Contemporaneous with the foregoing observations were those begun in 1817 by Josiah Meigs, then Commissioner of the General Land Office. He prepared blank forms for taking meteorological data, and issued them to the local land offices scattered throughout the states. This service was later transferred to the U. S. Patent Office, and continued until 1859.

Early in the nineteenth century the State of New York established a literature fund, a part of which was devoted to collecting meteorological data at several educational institutions in the state. Albany Academy began collecting such data under this fund in 1825, and is still continuing this work. It has cooperated, first, with the Smithsonian Institution; second, with the Signal Service of the United States Army, and last, with the U. S. Weather Bureau. At the time this station was established, Joseph Henry was connected with the Academy and became familiar with the meteorological work being done there. When he was chosen as the first Secretary of the Smith

sonian Institution, upon the founding of the latter in 1846, he made plans for organizing "a system of extended meteorological observations for solving the problem of American storms."* He called to his aid James P. Espy and Elias Loomis, two scientists already well known for their studies in meteorology. On December 15, 1847, the Board of Regents of the Smithsonian Institution appropriated $1000 for inaugurating the work, but Mr. Henry realized the impossibility of establishing stations with so small an amount and used it for the purchase of observing instruments. In August of the following year, Espy was appointed Meteorologist of the Navy Department, was given an appropriation by Congress, and was directed to cooperate with the Smithsonian Institution. Henry and Espy then formulated plans for establishing a large number of observing stations scattered over the entire country. These stations were of three classes:

1. Stations without instruments, where observations were made of frost, cloudiness, the direction of the wind, and the time of beginning and ending of rain.

2. Stations supplied with thermometers, where in addition to the above the daily maximum and minimum temperatures were recorded.

3. Stations equipped with full sets of instruments, consisting of barometer, thermometers, hydrometer, wind vane, and snow and rain gage.

These instruments were all standardized.

In February, 1849, a

call for voluntary observers was sent out and 412 responded, 143 of whom had formerly worked under Espy's direction. This is the real origin of the United States Weather Bureau. The work was almost totally discontinued during the Civil War, and in 1865 many valuable records and instruments were destroyed by fire, but the importance of the service had been too clearly shown to allow the work to be discontinued.

In 1870 Congress passed a joint resolution placing the work under the direction of the Chief Signal Officer of the United States Army. Later by act of Congress, approved October 1, 1890, the work was reorganized as the Weather Bureau and placed under the U. S. Department of Agriculture. All meteorological records previously taken were transferred to the new bureau.

There are a number of observing stations in the United States which were established and conducted by private individuals prior to 1846, when the Smithsonian Institution undertook the work. The long time records of these isolated stations when supplemented by the later records are of great value in determining the frequency of storms of different degrees of intensity.

* Report of the Smithsonian Institution, 1846.

As stated above, the Smithsonian Institution records are incomplete prior to 1865, but after that year there are no serious interruptions. A large number of new stations, made possible by the more generous appropriations, were established between 1870 and 1875 by the Army Signal Service. There are now about 200 so-called Regular Weather Bureau Stations. These are the most important and completely equipped stations of the service, each one representing a territory of about 15,000 square miles. Observations are taken daily at 8 a. m. and 8 p. m., 75th meridian time, by trained observers employed by the Weather Bureau. There is a continuous recording attachment to the rain gage, by which it is possible to determine the rate of precipitation and the total for the storm at any time. By this means it is possible to compute the precipitation from midnight to midnight, and it is this figure which appears on the reports. In addition to the regular Weather Bureau stations, the service includes about 4500 cooperative stations throughout the United States, operated by voluntary observers. These stations are equipped with fewer instruments than the regular stations, but those used are tested and are of the same high grade. The rain gages do not give a continuous record of the precipitation. Observations are made once each day at about 8 a. m. or 8 p. m., 75th meridian time.

The Weather Bureau also maintains so-called River Stations, where river gage readings are taken; these generally perform the functions of the cooperative stations also. The records for many of the stations are incomplete for one or more years, due to various causes, but these interruptions of the service at individual stations are of slight importance when the records for all the stations are considered.

In 1883 cooperative stations, known as Cotton Stations, operated by voluntary observers, began to be established in the cotton growing sections; observations are made daily at 8 a. m., 75th meridian time, during the crop growing season. In 1896 similar service was extended to the corn and wheat sections. Observations are not made at these stations during the winter months from November to April, and for this reason the data taken at such stations has not been of much value in these investigations.

EXISTING WEATHER BUREAU RECORDS

The records of the Weather Bureau available at Washington, D. C., and consulted for this investigation are as follows:

I. Smithsonian Institution Records: About 400 volumes. All records prior to 1873 except military records, taken principally during the period from 1848 to 1873; bound by months, each volume containing the data for all the stations in the United States for one month. Voluntary observers.

II. Military Station Records: (1) Original monthly reports from 1843 to 1859 for all the military posts in the United States; about 32 volumes. The data for each station for one year is together, 2 volumes containing the data for all the stations for each year. (2) Copies of the original reports from 1860 to 1892; about 600 volumes. The originals were kept at the posts and copies sent to Washington from time to time. Each volume contains the complete record for one station from 1860, or from date of establishment if after 1860, to 1892.

III. Compiled Data from 1860 to Date: About 36 volumes, bound by states, contain the data for each station bound together for the period from 1860, or from the date of establishment if after 1860, to 1891. Another set of volumes contains compiled data for regular Weather Bureau, cooperative, and river stations, generally for the period subsequent to 1891. The records for one month for all the stations in a given state appear together.

IV. Regular Weather Bureau Station Records: Original monthly reports (daily observations) of the regular stations from 1872 to date, bound as follows: From 1872 to 1887, for each month there is one volume containing the daily observations for all the stations in the United States; from 1888 to date, for each year there are 10 to 12 volumes containing the daily observations for all the stations in the United States, the stations being arranged alphabetically, and the records at each station for the year appearing together.

V. Data Compiled from the Records of the More Important Regular Weather Bureau Stations: Daily precipitation figures are for periods from midnight to midnight. (1) From 1871 to 1902 the data for stations A to K is arranged alphabetically and bound in 10 volumes; data for stations L to Z was never compiled. The data for each station for the whole period is bound together. (2) From 1871 to 1891 the data is arranged alphabetically by stations and bound in 6 volumes. The data for each station for the whole period is bound together, each page containing the summarized data of one station for

one year.

VI. Voluntary Observers' Records: Original monthly report sheets, showing daily precipitation from 1874 to date, bound as follows: (1) From 1874 to 1888, about 285 volumes; the data for the whole period for any given station is together, and the data for each state is arranged alphabetically according to station name. (2) From 1889 to 1892, about 60 volumes; the data for one year for any given station is together and the stations in each state are arranged alphabetically. (3) Subsequent to 1892, about 350 volumes; the data for a period of 3 years for any given station is together and the stations in each state are arranged alphabetically. There are about 50 volumes for each 3-year period.