level. For the central coast the curve is quite similar to that for July, but the phases are not so great in amplitude and the slopes of the max. are not so steep. At the south there is a well rounded max. which rises for half the day above a nearly level min. period; the whole curve exhibiting phases about half way between those for January and July.

At the high station of Cape Mendocino the curve for the first half of the day resembles that for July, while for the latter half of the day it is similar to that for January for the same hours; but the amplitudes are less than for those months. Having mentioned the Coast regions, we now pass to the various Inland sections of the U. S.

Inland: Northeastern U. S. (Curve 9).-For January there is a well marked maximum rising, for about ten hours of the day, above an irregular min. period, which has in some cases faint secondary phases. The crest of the max. is fairly well rounded, but flattens out (showing less rapid changes) with progress southward.

For July the maximum period is above the nearly flat min. period for about fourteen hours; in most cases the crest is well rounded but it becomes flatter toward the south.

For the Year the minimum period is usually still more flattened out than in the cases just mentioned, and the maximum portions of the curves are nearly as pronounced as for July but do not extend over quite as many hours. The curves for the Year are unusually similar to those for July, and in each very little secondary influence is noticeable.

Southeastern U. S.-For January the maximum rice is well marked, but does not cover a period of over eight to ten hours, and the minimum is rather more regularly level than is usual in mid-winter. For the relatively high exposure at Atlanta (Curve 10) there is but slight absolute variation, although there is a noticeable max. succeeded almost immediately by a rapid descent to a narrow minimum.

For July the max. covers a wider period than for January, and is more strongly marked (except at Jacksonville), while the curve of descent becomes steeper and the max. retarded until a later hour. The minimum period is hardly as level as for January. At the high exposure of Atlanta a slight secondary minimum is present.

For the Year the maximum is well developed and the curve well rounded, with a gradual ascent and a slightly sharper descent, and extends for about twelve hours above the smooth, slightly sloping, minimum period. At the high exposure of Atlanta the curve resembles that for July but has a more rounded maximum with slight secondary phases.

Lower Mississippi River Region (Curves 11 and 12).— For January the curves are very irregular and suffer but slight

absolute changes; usually the max. occurs at 14h to 16h. The irregularity is greatest along the Mississippi River, at Vicksburg and Memphis, where the maximum is so slightly developed that the secondary maximum is nearly or quite as great as the primary; this is not due to the greater development of the secondary but the lack of development of the primary. For Vicksburg, the minimum has about the same absolute value in both phases, but for Memphis the afternoon minimum (at about 19") is slightly the lower of the two. On the lower Arkansas River, with a more east-westerly river exposure, the primary maximum at 13 or 14" is well developed, but the secondary phases are hardly perceptible, and the principal minimum is hardly to be distinguished since it is but little below the long, nearly level, period of minimum wind.

For July, there appears to be but a single well defined maximum culminating about the middle of the afternoon, at about 16h, and a minimum is reached just before or about midnight. The slight traces of secondary phenomena can hardly be ascribed to other than accidental errors of exposure, etc.

The curves for the Year are very similar to those for July, but the amplitudes are not quite so great, and the minimum period is more nearly a uniform level. In all cases the descent from the culminating maximum is steeper than the ascent to it. The absolute maximum is at about 15h and the minimum at or just before midnight.

Ohio River Region (Curve 13). For January the maximum, at about 14, is well defined and with a rounded crest in some cases, and in others a pointed crest, but the slopes on either side are symmetrical. There is just a trace of a secondary maximum at some of the stations, which becomes best defined at Knoxville; while the minimum occurs for some of the stations before midnight, and for others after midnight, so that the usual times of both primary and secondary minima are represented in speaking of the principal minimum, and at Knoxville minima of about equal magnitudes occur for both primary and secondary phases.

For July the maxima are well developed, and usually with rounded, rather symmetrical crests; and the period of excess above the somewhat level period of deficiency is about fourteen or fifteen hours. The time of extreme maximum covers most of the afternoon hours, extending from 13" or 14" to 17h or 18h. There is but little more than a suspicion of a secondary maximum phase at about midnight, and although it can be noticed in some cases, yet it is entirely wanting in others. The minimum occurs a little before sunrise, and consequently just precedes the beginning of the ascent to the principal maximum.

For the Year, the maximum curves are remarkably symmetrical and have an absolute maximum at about 15h, and while the amplitude is not quite so great as for July, yet the period during which the curve is above the very nearly level. minimum is but little less than for July, and what shortening there is, is due to there being an earlier descent to the minimum level. The time of minimum is still in the early morning hours. Secondary phases are not noticeable.

Lake Region (Curve 5).--On the Lower Lakes for January, the curves are generally very irregular, but there is, in every case, a well defined, though short, principal maximum at about 14. The time of the principal minimum is variable; for Rochester and Cleveland it occurs only about four or five hours after the time of maximum, consequently at about 18h, and for Buffalo, Detroit and Toledo it occurs shortly after or at midnight. Slight, secondary maxima occur for Rochester at about midnight; for Toledo two hours earlier, and probably for Cleveland at about 23h.

For July a fairly well rounded maximum culminates at 14, but the time of minimum is by no means regular, for it occurs at about 4 hours before midnight in some cases, and 4 hours after midnight in others. In those cases in which the maximum is the best developed, the minimum occurs after midnight, both of which characteristics belong to the stations. having greater land influence.

For the Year, a moderately sized maximum with gentle, regularly sloping, sides culminates at about 14. The minimum period covers about half the twenty-four hours, and the absolute minimum follows closely the relations just given for July, but the absolute change is so much smaller than for that month that the whole minimum period varies but little from a fixed level.

For the Upper Lakes for January there is a principal maximum at times varying from 12 to 15 and the crest is the more pointed, the better the water exposure. While the principal minimum occurs in the early morning hours (4 and 8) for most of the stations, yet at Duluth a secondary minimum occurs at 17h or 18h, and this is of the same magnitude as the primary. The minimum period is very irregular and broken in all cases, and there are traces of secondary phenomena, which are well marked only in the case of Duluth, where the maximum nearly equal to the primary occurs near, or shortly before, midnight, and the minimum as just mentioned.

For July there is a well marked, rather rounded maximum at 15 in most cases, but at 14 at Alpena, with minimum at about 3 for Alpena and Chicago, but with the principal minimum just before and at midnight at Milwaukee and

Duluth. For these last mentioned places a secondary maximum occurs at 3 or 4h and a second minimum at about 6h, while for Duluth there is a slight tertiary maximum at 22h and minimum at 20. For the early hours of the descent from the maximum the slope is steeper than in the ascent.

For the Year, the maxima of the curves are nearly as well developed as for July, but the amplitude is not quite so great, and the period that the maximum is above the nearly level minimum period is only about 12h, which is less than for July. In most cases the minimum occurs at about 3 or 4h, but the whole minimum period is nearly the same, although for Milwaukee it is very irregular, and secondary phases are probably present.

Upper Mississippi River Valley Region (Curve 14).—For January there is a rather sharp-crested narrow maximum rising for about 9 above a somewhat irregular though level minimum period, and having a small amplitude. The minimum usually occurs at about sunrise, although a near approach to the absolute minimum occurs at about sunset or a little later. Traces of secondary phenomena are present about three hours after midnight for Dubuque, and three hours before midnight for Keokuk. At the high exposure of St. Louis there is a low flat-crested maximum during the time from 9h to 15h and a secondary sharp-crested maximum of nearly the same amplitude at midnight; the principal minimum is at 18h and secondary at about 6". The extreme amplitude is very slight, during the whole month.

For July there is a well developed maximum at from 14h to 16h and in most cases a secondary maximum at about midnight (sometimes a little before, and sometimes a little later). The times of minima are variable but usually at 6h and 21h, or a little later; in some cases the first is the primary and in others the secondary. The secondary phases are slight in comparison with the primary.

For the Year, there are well developed maxima at about 14h and very slight secondary maxima at St. Louis, at or just before midnight. The minima occur at or about daybreak, with very slight secondary minima at St. Louis at 20". The minimum period is quite level. Secondary phases are hardly noticeable except at St. Louis.

Great Plains (Curves 15 and 16).-For January in the north, the maximum period culminating at 13h or 14 or 15h is short (about 8h above minimum level), but well marked; and the main minimum is at from 5h to Sh. Most of the stations show secondary maxima, near midnight, and secondary minima at about 18 or 19h and of nearly equal level with the principal minima. In the south the principal maximum extends over a longer period and is more flattened at the

crest, which reaches from about noon to four hours after, but has not a greater amplitude than in the north. Secondary maxima rather better marked than usual are present, just before midnight at Abilene and Palestine; and the secondary minima for these stations at 18h and 20h are a little lower than those at about sunrise in the morning, so that the secondary phase really becomes the primary in these cases.

For July, the maxima culminating at 14" or 151 are large and well rounded, with the exceptions of Abilene and Palestine which have flat irregular crests; and the curve of the minimum is a sharply defined trough (at 5 or 6h) in some cases but rather flattened out in others, no law appearing to hold good. Secondary phases are present except in the extreme south, and in Dakota. At Palestine the usual secondary minimum is lower than the primary, and so in this case becomes the primary. Abilene and Palestine have flattened irregular crests at the principal maxima, and these show a slight recession at about noon.

For the Year the curves are quite similar to those for July except that the maximum period (culminating at 14" or 15h) is not so long or the amplitude so great. The minimum is at 5 or 6. Secondary phases are lacking or are barely perceptible, except at Palestine, where the usual secondary minimum is so low as to really become the primary. The crest of the maximum is much flattened out at the south, showing little change for several hours about noon and later.

Great Plateau (Curves 17, 18, and 19).-For January at the south, there is a single abrupt, narrow, steep sided, but slightly round crested maximum with level or slightly sloping minimum but no secondary phases; near the center there is still a principal though less marked maximum, varying from 12 to 16h, and usually secondary or even tertiary phases. While the minima are about equal, yet the customary secondary minimum phases (in the early evening) are usually slightly the lowest and thus become the primary. At the high station of Winnemucca the curve is irregular and has three sets of phases; while at the north the curves have relatively slight amplitudes and are irregular, having at least two sets of phases, and the usual secondary minimum becomes the primary in about half the cases.

For July, in the south, the maximum (at 14h or 15") becomes relatively great in Central Arizona, but is not usually large elsewhere, while a sharply defined minimum occurs at 6 to 8h; and but single phases occur where the amplitude is so excessive, while for the other stations secondary phases occur, but the minimum at about sunrise remains the primary minimum with a single exception, in which case the minimum occurs at midnight. Near the center the maximum period (culminating at