SEPTENBER-DECEMBER 1960 MONTHLY WEATHER REVIEW 353 THE WEATHER AND CIRCULATION OF SEPTEMBER 1960* Including a Discussion of Tropical Storm Activity C. F. TISDALE Extended Forecast Section, US. Weather Bureau, Washington, D. C. 1. WEATHER HIGHLIGHTS September 1960 will be remembered by residents of the Atlantic Seaboard, from Florida to Maine, as the month and year of hurricane Donna. This severe tropical storm dominated weather headlines from its det,ection in the tropical Atlantic on September 2 until it became extra- tropical in the Maritime Provinces of Canada on Septem- ber 13. Preliminary estimates of total damage mark Donna as the most destructive storm ever to strike the United States mainland. The number of fatalities and injuries from Donna were slight in comparison with those sustained from other, lesser storms in the past,, and this fortunate outcome can be attributed in major part to widespread and prompt dissemination of accurate and timely warnings. The sudden development of hurricane Ethel in the central Gulf of Mexico on the 14th was cause for consider- able concern to residents of Gulf coastal sections, especially those along the southwestern coast of Florida who had just endured the fury of Donna. The intensiby of this new tropical threat fortunately decreased quite rapidly before it reached the coast and moved inland into southern Mississippi on the 15th. September weather was also highlighted by extreme variability, specifically in respect to the temperature regimes exhibited over the country during the course of the month, as periods of marked cooling alternat'ecl with periods of summer-like warmth in n~any sections. I n spite of this variability, temperatures for the nlont'h as a whole averaged near to above normal over practically the entire country. The sharp contrast between record precipitation in tlle extreme Southeast and Northeast and near to record dryness in some interior and far western areas was an additional highlight of September weather in the United Stat>es. 2. MONTHLYMEANCIRCULATIONANDWEATHER Strong polar blocking, which had dominated the circu- L'Articles on the weather and circulation of October, November, and Dcccmhcr 1960 will appear in the Review issues for January 1961, February 1961, and March 1'361, respectively. This apparent lag is due to the change in monthly designation of the affected. Review beginning with the current issue. The calendnr date of publication is not lation ol August [I], receded southeastward during Sep- tember to it center north of Scandinavia, where 700-mb. heights were 300 ft. above normal (fig. 1). Along with this development, a gradual contraction of the circumpolar vortex took place during the month. This circulation upheaval is indicated by the changes in monthly mean 700-mb. height departures from normal from August to September (fig. 2 ). S o t e t h e significant height falls over the polar regions and tlle positive changes over the major portions of the United States, Canada, and middle latitudes of the Atlantic. The resulting mean circulation for September was predominantly one of high index from the coast of Asia eastward to t8he British Isles (fig. 1). Planetary waves were of small amplit'ude, and the axis of mean 700-lnb. zonal wind speed maxima was displaced a substantial distance north of normal over North America (fig. 3 ). Broad cyclonic curvature characterized the central Pacific trough and its downstream counterpart over t'he extreme eastern United States. The wave- length between these two systems was long by summer- time standards and appears to have been sustained by the st'ronger than normal westerly circulation. The trough normally active along the west coast of t'he United States was extremely weak, and positive height departures from normal dominated pritct'ically the entire region from t8he northwestern LTnited States east,ward to the centra1 Atlantic. The simple temperature and fairly well-defined pre- cipitation patterns for the month of September (fig. 4) can be related quite well to the monthly mean 700-mb. height, and anomaly pat'terns (fig. I), despite the changes in tml-lese patterns which took place during the month (sect'ion 3). Temperatures averaged above normal in practically all sections of the count'ry (fig. 4A) under predominantly above normal height's. The warm tem- peratures in t'he western plateau areas represented a sharp reversal fronl t8he cool weather experienced there during August [I]. This warmth was t~ consequence of upper- level ridge development over the Xort'hwest (fig. 2) and frequent anticyclonic passages in the vicinity of the Rockies (see Chart IX of [2]) which kept the region mainly in southerly flow at sea level. It was the warmest Sep- t8enlber of record Rt Ely, Nev. and the month was also notably 1varn1 in some int'erior sections of California. FIGURE l.--Pvlean i00-mb. contours (solid) and height departures from normal (dotted) (both in t,eris of feet) for September 1960. Note contracted nature of circumpolar vortex with strong zonal flow over middle and high latitudes of North America. Some warming was also experienced in central portions of the country under above normal 700-mb. heights and abundant sunshine, with South Bend, Ind. experiencing the warmest September since 1933. The cooling which occurred in the Middle Atlantic States from August to September can be attribut'ed to increased cyclonic circu- lation aloft and to stronger onshore flow from the Atlantic. Near to below normal temperatures persisted in most of the extreme Sout.heast and Gulf coastal sect'ions, where below normal heights and easterly, anornalous Pow doni- inated during the month. This was the coolest Septem- ber since 1889 a t Brownsville, Tex. Heavy precipitation from Florida northwestward to eastern Tennessee (fig. 4B) resulted mostly from tropical activity moving inland during the month. It was the wet8test September and month of record a t Miami, Fla., where rainfall totalled 24.4 inches. Well above normal rainfall in the Xortheast occurred in conjunction with the passage of Donna early in the month, coupled with scattwed showers and frontal rains later in t8he month. SEPTEMBER-DECEMBER 1960 M O N T H L Y W E A 355 FIGURE 2.-Changes in 700-mb. height departures from normal (tens of feet) from August 1960 to September 1960. Subst,antial height falls over polar regions are especially noteworthy. FIGURE 4.-(A) Departure of average surface temperature from normal (" F.) for September 1960. The pattern is noteworthy for the prcdomitlance of above normal temperatures (shaded) over practically the entire nation. (B) Percentage of normal precipitation for September 1960. (From [ll].) FIGURE 3.--?vIean 700-mb. isotachs (in meters per second) for Sep- tember 1060. Solid arrows indicatcx axis of wind s1:cc:d maxima, dashed arrows the normal for the rrlontll. The primary axis was north of normal over Korth A4mcrica. Subtropical ridge line is denoted by solid line cormecting ctntcw of slow wind spccd (S). New record precipitation amounts for September were established a t Wilmington, Del., Scranton, Pa., Albany, N.Y., and Concord, N.H. Frequent outbursts of frontally induced convective activit,y were resporlsible for twice t,he normal precipitation in southern Arkansas and sub- stantially above normal rainfall from Wisconsin south- west8ward to the p d l a n d l e areas of Texas and Oklahoma. Prcdomirmlce of eastcrly anom:rlous flow precluded the possibility of any sustained intrusions of moist, Gulf air northward into the mid-sections of the country, resulting in m a r l d dryness from c1:Lstcr.n Oklahoma northeastward to western New Yorlc. I t WIS the driest September and rnontll of record a t Akron, Ohio, and the second driest Septetnlwr of record at, Detroit, Mich. ant1 Columbia, 110. Almvr normnl heights and strong zorlal flow col- 1:Lbor:ltetl in tjhe production of lit,tle to no precipitation over most of North Utlkota :md Alontarla, whcre rainfall deficiencies a t Glasgow made this September t'he second driest of record. Scattered showers, while accounting for the heavy rainfall indicat,cd in north-ccnt~ral Nevada did lit'tle to alleviate general drought conditions which have persisted over the State for two years. That there was a marked predominance of ant,icyclorlic act'ivit'y in the month's weather, is clearly illustrated by the abundance of migratory anticyclones and the almost complete absence of migratory cyclones in the United States, except along ext,rerne rlortllerrl border areas (Charts IX and X [2 ]). This is in good agreement with what one SEPTEMBER-DECEMBER 1960 M O N T H L Y W E A 355 FIGURE 2.-Changes in 700-mb. height departures from normal (tens of feet) from August 1960 to September 1960. Subst,antial height falls over polar regions are especially noteworthy. FIGURE 4.-(A) Departure of average surface temperature from normal (" F.) for September 1960. The pattern is noteworthy for the prcdomitlance of above normal temperatures (shaded) over practically the entire nation. (B) Percentage of normal precipitation for September 1960. (From [ll].) FIGURE 3.--?vIean 700-mb. isotachs (in meters per second) for Sep- tember 1060. Solid arrows indicatcx axis of wind s1:cc:d maxima, dashed arrows the normal for the rrlontll. The primary axis was north of normal over Korth A4mcrica. Subtropical ridge line is denoted by solid line cormecting ctntcw of slow wind spccd (S). New record precipitation amounts for September were established a t Wilmington, Del., Scranton, Pa., Albany, N.Y., and Concord, N.H. Frequent outbursts of frontally induced convective activit,y were resporlsible for twice t,he normal precipitation in southern Arkansas and sub- stantially above normal rainfall from Wisconsin south- west8ward to the p d l a n d l e areas of Texas and Oklahoma. Prcdomirmlce of eastcrly anom:rlous flow precluded the possibility of any sustained intrusions of moist, Gulf air northward into the mid-sections of the country, resulting in m a r l d dryness from c1:Lstcr.n Oklahoma northeastward to western New Yorlc. I t WIS the driest September and rnontll of record a t Akron, Ohio, and the second driest Septetnlwr of record at, Detroit, Mich. ant1 Columbia, 110. Almvr normnl heights and strong zorlal flow col- 1:Lbor:ltetl in tjhe production of lit,tle to no precipitation over most of North Utlkota :md Alontarla, whcre rainfall deficiencies a t Glasgow made this September t'he second driest of record. Scattered showers, while accounting for the heavy rainfall indicat,cd in north-ccnt~ral Nevada did lit'tle to alleviate general drought conditions which have persisted over the State for two years. That there was a marked predominance of ant,icyclorlic act'ivit'y in the month's weather, is clearly illustrated by the abundance of migratory anticyclones and the almost complete absence of migratory cyclones in the United States, except along ext,rerne rlortllerrl border areas (Charts IX and X [2 ]). This is in good agreement with what one SEPTEMBER-DECEMBER 1960 MONTHLY WE:\THER REVIEW 357 would expect under a high index regime with the westerlies north of normal and 700-rnb. heights above norrnal over most of the country. Below nornlal height's and cyclonic upper flow (fig. 1) resulted in correspondingly below norrnd temperatures and above normal precipitation over most of Alaska during September. Ternperature ttnorrlalies of as much as -4 O F. and precipit,at'ion ttrnounts greater t h m twice normal were recorded in some interior areas of the St,at'e. This was t,he wet'test September since 1925 at Fairbanks, Alaska. 3. TRANSITION WITHIN THE MONTH The main features of the general circulation over North America underwent three major oscillations during t,he month of September (fig. 5 ). There was rapid progressiotl during the early part of the rnonth, followed by apparent retrogression and/or new trough development during mid- month, with gradual progression and amplificat'ion again occurring during the latter part of the rnonth. Cou- cornitant wit>h t'hese circulat'ion changes were variations in the tenlperature and precipitation regimes. These are portrayed on a weekly basis in figures 6 and 7 , while the evolution of the circulation is shown by R series of over- lapping 5-day mean maps in figure 5 . FIRST WEEK The circulation over the country during the early part of the week was tnore typical of mid-sulllnler, with troughs just off the east and west coasts and tt strong ridge over the Mississippi Valley (fig. 5A). 1Jnusu:tll)- hot, hunlid weather, which began in middle and nort~hern areas east of the Rockies in late August, continued until the 7th or 8th of September. Maximum temperatures exceeding 100' F. occurred a t several stations in the Dakot;ts. A reading of 94' F. a t Lansing, Mich. on thc 7th est:tblislled a new record there for so late in the se:tson and some loca- tions in New Eng1:land experienced their hottest weiither of the year. Rapid progression of the &day rllean wave trtiirl occurred during the latter part of the week, indirating an end to this period of abnorrrlally high temper' he nliddle-Atla~~tic coastal sections through eastern New England and the Maritime Provinces of Canada early in the week (fig. 5C). AFove- normal tides, heavy rain, and wind gusts up to 130 n1.p.h. marked the storm's passage into southern portions of New England during the, afternoon of the 12th. Tropical activity- continued during the week with the rapid developrnent~ of hurricane Ethel in the central Gulf of Mexico on the 14th (fig. 5D). I t s position and intensity 358 MONTHLY WEATHER REVIEW SFZTEMBER-DECEMBEE 1960 FIGURE 6,"Departure of average surface temperature from normal '1" F.) for weeks in September 1960 centered on the 5-da~7 mean periods sholvn in maps B, D, F, and G of figure 5 , and end- ing (A) September 11, (B) September 18, (C) September 25, and (D) October 2. (From [ll].) A'ote the many, and ofttimes marked, reversals in temperature rcgime which occurred during the month. FICIJRE 7.-Total precipitation (inches) for the same weeks and from the same source as figure 6. Heavy rains were confined to various sections in the eastern half of the nation during the month. SEPTEMBER-DECEMBER 1960 MONTHLY WEATHER REVIEW 359 were established on the basis of reports by t'he Navy MAMOS (Marine Automatic Meteorological Observ. '1 t ' Ion Station), the first detection of a hurricane by this new meteorological instrument. This initially severe storm advanced northward and moved inland with decreasing intensity near Biloxi, Miss. during the afternoon of the 15th, with hurricane force winds observed a t b u t a few scattered stations in the immediate Delta area. The storm continued to fill rapidly as it moved northward during the 16th and 17th, spreading locally heavy rains from Alabama to the central Ohio Valley (fig. 7B). The development and motion of Ethel will be discussed more fully in sections 4 and 5. The discovery of tropical storm Florence northeast of Puerto Rico early on the 17th (fig. 5E) and its subsequent westward motion along a track closely paralleling that of Donna were additional highlights during the latter part of the week. THIRD WEEK The principal circulation features underwent a second major change over the country during the third week, as a trough developed over the Rocky Mountain St'ates and a strong ridge again dominated eastern areas (fig. 5E and F). Temperatures averaged unseasonably high in the lower Lakes Region and Mississippi Valley (fig. 6C) under anti- cyclonic conditions and above-norrnal 700-mb. hcigh t,s. Large areas east of the Mississippi were as much as 9O-12' F. warmer than the previous week. Cooling in north-central sections and the northwestern interior was associated with more cyclonic flow and near to below normal 700-mb. heights. Rapid drying accompanied the ridging in the East, with little or no precipitation rerorded in most areas (fig. 7C). Heavy rain in southern Kew England occurred in conjunction with cyclonic devclop- rnent just off the coast early in the week. Moderate to heavy rain in the mid-continent was the result of frontal activity during the latter part of the week. Precipitutiorl continued to be very meager in the Far West, where the only significant amounts fell in extreme western Washing- ton as a weak cyclonic center passed across soutllcrn Vancouver Island early in the period. Tropical storm Florence maintained a general westerly course and ceased to pose a major threat to the mtlinlwntl when i t passed inland over northern Cuba late on the 20th and decreased rapidly in intensit,y (fig. 5E). The much weakened center oscillated sharply southward and t'hen northward across western Cuba during the 22d and 23d (fig. 5F). Remnants of the storm caused hettvy rain over southeastern Florida and in areas of extreme southeast,ern Alabama and northwestern Florida during the last two days of the week. FOURTH WEEK The general circulation underwent its third major oscil- lation over the country during the last week of the month. The trough over the Rocky Mountain States (fig. 5F) advanced eastward to the Lakes Region and amplified to full-latitude proportions, extending from eastern Hudson Bay southward to the western Gulf of Mexico (fig. 5G). Ridging occurred over the western Plateau, and the low- latitude trough in the Far West receded westward to the northern coast of California. Average temperatures for t'he week were well above normal in interior areas of the Far West, with departures ranging up to +So F. or more (fig. SD) under anticyclonic conditions and above normal 700-mb. heights. h'ear to below normal temperatures predominated in the East under correspondingly below normal 700-mb. heights (fig. 5G). Three cold air masses advanced eastward across northern portions of the country during the week (see Chart IX of [Z]). Frosts and freezing temperatures from the Dakotas southward to northern Kansas on the last day of the month spread eastward to upstate New York, New England, and scattered sections of Pennsylvania and West Virginia by October 2 . Little or no precipitation fell over the western half of the nation (fig. 7D) under above nornlal 700-nlb. heights and drying, northeasterly anomalous flow(fig. 5G). Heavy rain in the Southern Plains was associated with cold- frontal activity early in the week. General rains in the Southcast and along the Atlantic Seaboard occurred under st,rong southerly flow of moisture from the Gulf of Mexico, with 3 or 4 days of rather dismal weather highlighting the week's activity in these areas. 4. TROPICAL STORMS RELATED TO THE MONTHLYCIRCULATION Of the three tropical storms which formed in the Atlantic and Gulf of Mexico during this September, two developed t o full hurricane intensity. These figures correspond t'o mean frequencies of t,hree storms and two hurricanes during the past 73 Septembers, but they are slightly less than the mean Septembcr frequencies of the past 10 years. The mean 700-mb. anornaly pattern for t'he month (fig. 1) did not exhibit any striking similarity to the overall patterns found by Ballenzweig [R] as favoring develop- ment in particular segments of t'hc Atlantic hurricane areas, except for some correspondence to the pattern favor- ing development in the eastern At'lantic. This is in itself of some interest, considcring the diverse areas in which the month's tropical activity originated; Donna off the west coast of Africa, Ethel in the central Gulf of Mexico, and Florence northeast of Puerto Rico (fig. 8). All three storms developed, however, under a circulation regime considered most propitious for hurricanes; i.e., during a time of generally high index with t'he westerlies somewhat north of normal. This is in good agreement with the proposal by Xamias [4] that a northward shift of the westerlies generally is associated with frequent shearing of polar troughs, thus providing a mechanism for maintaining cold air and cyclonic vorticity in the Tropics after trans- portation from northerly latitudes. The behavior of the tropical storms (fig. 8) can be 360 MONTHLY WEATHER REVIEW SEPTEMBER-DECEMBER 1960 FIGURE 8.-Preliminary tracks of tropical storms during September 1960 in the Atlantic and eastern Pacific Oceans. Open circles and dates indicate 1200 GMT positions. FIGURE 9,"Preliminary tracks of tropical storms during September 1960 in t,he western Pacific Ocean. Ope11 circles and dates indi- cate 1300 C,MT positions. i;lt,erpreted in terms of the mont'hly mean circulation (fig. 1). Donna's track paralleled very closely the ~n c m cont,ours across the sout,tlern Atlantic from western Africa to Florida. After recurvature, Donna moved sharply northward just, east of and in t'he mcan trough along the east coast. Hurricane Ethel moved north- northeastward along the mean trough in the central Gulf of hlexico. Similar motion of hurricanes in relation t'o the mean circulation has been not'ed by Klein [5] and discussed in many articles of this series. The movement of tropical storm Florence southward across western Cuba does not fit the 30-day mean patt'ern too well. However, weak remnants of the storm did eventually move northward east of the mean trough in the Southeast. Kote also the st,ronp easterlv. anornalous flow corn~onents on the mean TOO-mb. map (fig. 1) from sout,hcrrl New England south- ward to Florida and the eastern Gulf Coast States, imply- ing a high degree of vulncrahilit'y to tropical activity in t~hcsr areas. EASTERN PACIFIC Of interest this month were two tropical storms in tlle eastern Pacific (fig. 8), both of which were of hurricane intensity. Estelle was actually a carry-over from the previous month, having formed just south of Guatemala on the 29th of August. The track of Est'cllc paralleled the 1IIcan flow (fig. 1) as t'he storm moved ~est-north\~~-cstwarrl and then rccurvcd northward and westward and dissipated in t,he low-lntitutlc 111ean trough west of Baja C'alifornia on Sept,ernber 9. Ferrlanda tlevelopcd in approximately the same area as Est'clle, just South of Guatemala, on the B d and moved ~\~est-northwcstward witti the n ~e a n Aow, dissipating west of ccrltral hlexico on the Sth. WESTERN PACIFIC Tile most striking aspect of t,he mont'h's tropica; twtivity in the westerr) Pacific was t'he complete absence of st'orrns of typhoon intensity. Four storms were in evidence during various periods of the month (fig. 9), none of which developed beyond the tropicd storm stage. In f w t , t,wo ol the storms, Hest,er :t d Gloria (a carry-over from August) were so weak and srllall that their respective circulations apparently were not discernible for days at a t,irne (dashed portions of tracks in fig. 9). This was in marked cont'rast' to the unprecedented t'yphoon activity which dominated this ares during the previous nlontll [I]. A cornparison of tlle I n e m circulation of August [I ] with t'hat of Sept'ernber (fig. 1) suggests that the whole Dattern in t'he western Pacific shifted sout'llward a n d SEPTEMBER-DECEMBER 1060 MONTHLY WEATHER REVIEW 361 eastward about loo of latitude during September. The persistent Low over Formosa in August, which served as a sink for the cyclonic vorticity of t’he many t’yphoons, might conceivably have been absorbed northward into the westerlies early in Sept’ernbcr. This could have con- tributed to the development of the mean trough a!ong the coast of eastern Siberia. The subtropical ridge was stronger than nornlttl in the west-central Pacific (+ 180 f‘t,., fig. 1) and extended st’rongly westward south of Japan to eastern China. The marked southeasterly and convergent anomalous flow south of Japan in August (see fig. 1 of [lj) was replaced by a northeast’erly and strongly divergent anoma!ous Aow during September. Tropical storms Gloria and Irma developed some dis- tance east of the Philippines and moved west-northwest,- ward with the mean flow. Both storms eventuul1~- dissi- pated in the same general area of the South (lllintL Sea (fig. 9). Hest’er and Judy were relatively weak and short lived storms which deve!oped southeast of Japan. Both storms moved northwest’ward and were absorbed into strong cold-front’al systems which had nlovetl southenst- wlrd from Japan. 5. TROPICAL STORMS RELATED TO THE5-DAYMEANCIRCULATION The relation of the tropical st’orms of the Atlantic to the &day mean circulnt’ion is deserving of special att’en- tion, especially since t’he paths of the storms are often influenced by short’-period means. The exact conditions which led t’o the inception of Donna off the western coast of Africa during late August can only be surmised due to the paucity of data in this particular segment of t h e A t l m t i c . I t is quite likely, however, that the initial disturbance was in the form of a severe West African squall line of thc t’ype described by Regula [6]. The first advisory on Donna was issued on September 2 with the storm some 700 miles east of the Lesser Antilles and already of full hurricane propor- t’ions. Donna rnoved 011 a general west~-northu~estward track with the m e m flow, arriving in the extreme eastern Bahamas on the 7th (fig. SA). Note the slowing down of the storm’s forward progress from the 5th 60 t’he 7th and t’he t’urn to a Inore northwesterly direction as it approached the mean t’rougll off the east coast. I t was unfortunate that the general circulation underwent its first major oscillation of the month at this time when conditions appeared rather propit’ious for possible north- ward recurvature of the storm well off the coast. As it was, rapid progression of the 5-day mean waves occurred, with the trough off the coast shearing and moving to the central Atlant’ic, and the strong High over the Mississippi Valley advancing eastward to the coast and effectively bridging the gap north of the storm (fig. 5B). Donna, accordingly, turned again to a more westerly course, recurving sharply northward across Florida on the 10th and up along the full length of the Atlantic Seaboard in advance of the full-latitude trough which settled in over the East (fig. 5C). A rather intensive report on hurricane Donna, including a general summary of the storm’s track, major features of t’he storm in seriously affected island and continental areas, plus some comparisons of various forecasting techniques in their application to Donna, is being prepared by Haggard and Cry of the U. S. Weather Bureau*. Dunn [lo] has indicated that trailing, stationary, or fractured portions of old polar troughs may provide the initial concentration of cyclonic vort’icity necessary for hurricane formation. This cert’ainly appeared t’o be the case in regard to the development of hurricane Ethel just east of the t’railing, southern portion of the mean polar trough in t’he central Gulf of Mexico on September 14 (fig. 5D). The st’orrn moved north-northeastward along this trough, entering t’he coast near Biloxi, Miss. on the 15th. The st’orrn continued northward with rapidly diminishing intensity, eventually dissipat’ing in the Tennessee Valley on the 17th. Tropical storm Florence developed from an easterly wave which moved through the Lesser Antilles during the l6t,h. The circulat’ion was first det’ected nort’heast of Puerto Rico on the 17t’h, and t’lle storm moved westward with the mean flow, entering land in northwestern Cuba during the 21st’ wit’llout ever attaining hurricane strength (fig. 5E). Weak remnants of t’hc storm oscillat’ed south- ward and northward itcross western Cuba to Florida from the 21st to t’he 24th near a mean t’rough in an area of very weak gradient (fig. 5F). This track may have been influenced more by prevailing surface conditions in the area t’han by mean circulation features. REFERENCES 1. L. P. Stark, “The Weather and Circuiation of August 1 9 6 G A Month Dominated by a Circulation Reversal,” Monthly Weather Review, vol. 88, No. 8, Aug. 1960, pp. 286-293. 2. U.S. Weather Bureau, Climatological Dala-National Summary, ~o l . 11, No. 9, Sept. 1960. (in press) 3. E. M . IMlenzweig, “Relation of Long-Period Circulation Anonlalies to Tropical Storm Formation and Motion,” Journal of Meteorolcgy, vol. 16, KO. 2, Apr. 1959, pp. 121-139. 4. J. Xamias, “Long-Range Factors Affecting the Genesis and Paths of Tropical Cyclones,” Proceedings o,f the UNESCO S y m p o s i u m o n T y p h o o n s , 9-19 November 1954, Tokyo, 1955, 5 . R . H. Klcin, “The Weather and Circulation of Junc 1957- Including an ilnalysis of Hurricane Audrey in Relation t o the Mean Circulation,” Monthl?] Weather Review, vol. 85, SO. 6, June 1957, pp. 208-220. 6. H. Regula, “Tornadoes and Pressure Variations on the Rest Coast of Africa,” Annalen der Hydrographie und Maritimen Mefeorologie, vol. 64, Jan. 1936, pp. 107-111. 7. Ti. Butson, “Hurricane Donna in Florida,” Weatherwise, vol. 13, 8. A. V. Hardy, “Hurricane Donna in North Carolina,” Weather- wise, vol. 13, No. 5, Oct. 19€ 0, pp. 213-214. 9. J . K. MeGuire, “Hurricane Donna in thct Xortheast,” Weather- wise, vol. 13, No. 5, Oct. 1960, pp. 215-217. 10. G. E. Dunn, “Tropical Cyclones,” Compendium of Meteorology, American Meteorological Society, Boston, 1951, pp. 887-901. 11. U.S. Weather Bureau, WeekZty W e a t h e r a n d C r o p B u l l e t i n , N a t i o n a l S u m m a r y , vol. XLVII, Kos. 37-41, Sept. 12, 19, 26, Oct.. 3 , 10, 1960. pp. 213-219. No. 5, Oct. 1960, pp. 210-212. *Shorter accrnnts of the storm’s ~.msa::e tl,Iouy!l verins parts of the Vnited States have h e m [~lblishul [i, 8, 91