522 MONTHLY WEATHER REVIEW NOVEMBER, 1904 zontally by the wind. In the four experiments a t St. Louis in the month of September all descended within fifty miles of St. Louis. I n ten ascents during November and Decem- ber the balloon traveled much farther, reaching 200 miles from St. Louis. I n general it would seem that even in midsummer, a t altitudes above four miles, there are perpetual freezing tem- peratures, and an upper air current blowing steadily from the west. The Blue Hill Observatory is again to be congratulated on the energy it shows in the matter of upper air observa- tions, ancl it is to be hoped that Mr. Rotch will be able to con- tinue the balloon work a t St. Louis. hhny such stations will be needed in order to perfect our knowledge of the atmosphere over the United States-C. 8. TRAILS OF METEORS. A newspaper clipping froin Portland, Oreg., says : Residents of the Knob Hill district going home last night (Dwenlber 21, 1904), shortly after 6 o’clock were startled by the appearance nut i )f the cloudy air of a meteor nith a trail of fire behind it, flying c .1 1 ,~ i)ver the city, apparently not O \P ~ 151) feet above thenl. It wab glowing 16v1 and threw out sparks behind like the tail of a ci~~iiet. The quotation relates to a plienomenon that woiilcl give u s much information about the upper atmosphere if only the ob- servers woulcl give us a more exact account of what they saw. The smokelike trail left behind by a meteor is undoubtedly higher up in the air than any altitude attained by balloons, and the changes that i t undergoes must be snpposecl to depeiid, at least in part, on the atmospheric conclitions in its neighlwr- hood. What the meteorologist wants is a sketch, as exact as possible, of the location of the trail among the stars. Seyeral successive sketches, stating the moments of time, and the ap- parent angular altitude, ani1 the bearing or azimuth, woulcl give 11s the basis for some calculation as to atmoqheric influences. When several observers happen to see the same meteor from different points of view, their separate sketches of the trail would give u s still more interehting results. A large collec- tion of data of this kind has been made by RIr. IT. P. Trow- bridge of Columbia University, New Tork, N. Y., and elery additional sketch woulcl be liiglily valued by liim.--C‘. -4. FLOOD ON THE SOUTH CANADIAN RIVER, IN OKLA- HOMA AND INDIAN TERRITORY, OCTOBER 1-4,1804. The Canadian River rises in eastern New Mexico and flows across the panhandle of Texas into Oklahoma and Indian Territory, where i t joins the Arkansas. I n the preceding number of the MONTHLY WEATHER REVIEW, page 466, we pub- lished an account of the floods in New Mexico due to heavy rains in that Territory from September 26-30. The resulting flood during the first clays of October on the Grand River, below the point where it enters Oklahoma, is described in detail by h h . C. hl. Strong, Section Director, in the October and November reports of the Oklahonia ancl Indian Territory section of the Climate and Crop Service of the Weather Bureau. That the flood originated in the niountains of New RIexico is shown 1)y the fact that only in that State were there any heavy rains on the Canadian River basin between September 25 a i d October 4. The flow1 I )f the South c’anadian Iliverof October 1 to4, 1904, will stand iiieniorable a+ the m o 4 ilestructil e one in the history of this setation since its srttlt.uient. To thororighly nuderstancl itsdebtructive effectsoue must ha\ e a k n o ~ ledge of tlie topographical features of that stream. Finaliup its source in tlia iiiouutain rauges of New Mesici), it thence iileauders wutlienstwaiil at’ross the high uplanilb of the Texas panhandle to the fertile 1ilains o f C)hlahoma, ani1 on tlirougli the Indian Territory to its wnnec+ion with the Arkansab River Thrnughuut i t s couise its Ijed lies inclt el1 in a wide valley, whose bot- tom levpls art? I j u t sliylrtly a h )\ e the surface o f the stream. Its waters are usually hhallow, nf sinal1 width, and to a large extent are under- groimil throughout tile year, tlir ubual width of the surface stream varying froni GO to 200 yards. Previoob to the flooil the river valley was cnvered with fertile farms from its entrance into Oklahi iiua to i t s mouth in the Indian Territory, rich with fields of cotton, coni, and wheat. Conceive in your mind thiz beautiful and fertile valley, rich in the finest products of the earth, and tlieii turn to the results following the Bond. The feeble stream. winilinp its way in a shallow bed, became a vabt flow1 that rolle(1 IL will o f water eighteen to twenty feet deep. in places spreading from hill to hill, with width varying from nile to two miles, sweeping everything from its path ani1 (.overing the valley with sand from one-half to fclrir feet in depth, completely ol)literating everything in the form of vegetation. The force of the water waq sn treiiieuilous that nnthing could stand in it+ course: crous. bridges-110th irou and wooden-trees, aud houses werH ah. Strong states: DARKNESS AT MEMPHIS. From a newspaper clipping we learn that on December 2, about 10 a. m., a dark pall covered MemphiR for about fifteen minutes. The sunlight was wholly cut off by a dense cloud ancl the clarkness of night prevailed. Of course artificial light was used and business temporarily paralyzed, but the light returned as suddenly as it went. We are told that in some quarters a panic prevailed, and that some were shouting and praying, imagining that the end of the world had come. A similar darkness frequently attends a heavy local cloud, and is a common attendant upon prairie and forest, fires and upon 701- canic eruptions; everyone is familiar wit,h the account,s of the great clarkness attending the fogs that prevail in London during the winter season, ancl with less intensity in New York, Philadelphia, Chicago, and other large cities where soft coal is used. Every intelligent person understands that such darkness gives no occasion for superstitious fears. However, we inust recognize the fact that there are some in every class, ignorant or educated, American or foreign, rich or poor, who are liable to loose self control when anything very unespecte(1 OCCU~S. A panic is almost as easily precipitated among men, wonien, and children as i t is among her& of animals. For the Weather Bureau men it should be a proud record that they haTe stayed many panics by words of assiirance and a cool bearing. We ought never to forget that every natural phenomenon has a natural cause, and requires to be considered and treated calmly. So matter what the iriipentling disaster, the first requisite is to preserve one’tJ self-possession and help others to do the same.-C. A. swept away liGe straif the best European iueteorologisth, and I may especially refer )oii to the accompanying letter of Piuf. .J. N. Pernter, reprinted from the ~IONTHLY WEITHER REVIEW for May, 1904. You may be interested also in reading the en~ln+.erl 11age5 from the advance proof of my Annual Eept)rt. i n \\ hich I hare refriled to long-range forecazth. Rrqvttiug that I ~‘i t i i not cncortragc~)t)ur pul~lic competition. but with tlie l i e d \\ ishl)h foi thv prosperity of the Brlgiau d:,troinical Society, I an1, Very re%pci.tfully, (Slglled J WILLIS 1 4 . M~JORE, Chi+ f‘ U. S . Tfecrther Bureciu. CORRIGENDA. RIONTHLY IvE.4THER REVIEW for August, 1904, 11. 373, COlUmII 2, line 10, “ involve ” read “ evolve.” RI~NTHLY WEATHER REVIEW for October, 1904, p. 458, column 2, Table 1, number of clays with thunderstorms in May, 1895, for ‘‘ 6 ’ I read r ‘ 5 ”; 1). 459, column 1, Table 4, average duration of thunderstorms in November, lS90, for “. ..... read ‘‘ 0 ”; p. 465, column 1, line 11, for ‘‘ Helmholz ” read ‘‘ Helmholtz,” liue 16, for ‘.Nenclioff ’’ read “ Neulioff,” line 18, for “Eck- holm ’’ read “ Ekholm.” THE WEATHER OF THE MONTH. By Mr. {VU. B. STOCKMIN, Chief, Division of BIeteorological Records. PRESSURE. southern Plateau regions. The greatest negative departures atluosl)~eric pressure is grallhicallg ranged from - .10 t,o - .15 inch, and occurred over New Eng- shown on Chart VI11 and the average values and departures lancl, ~11~1 the extreme eastern portion of New York. from normal are shown in Tables I and VI. The mean pressure decreased from that of October, 1904, in highest over the Plateau regions, New Euglancl, Miclclle dtlantic States, northern portion of the with the crest over western ~v ~~~i ~~. lowest over South Atlantic States, Ohio Valley ancl Tennessee, except the eastern New Engliznil. western portion, the Lake region, and the north Pacific dis- baronleter Was above the norrllal in the west Gulf trict. States, the Mississippi and Missouri valleys, slope ancl Plateau The inasiiuiim increase ranged from + .10 to + .16 inch, regions, the south Pacific region, except the estrenie south- ancl occurred over the middle and southern slope and Plateau western portion, and the micldle Pacific region, except tlie es- regions, the wuthern portion of the northern slope region, treme northwestern portion. I n all other districts it was be- aud southwestern North Dakota. The maximum decrease low the normal. ranged from -.lo to -.17 inch, ancl o c c u r r e d over New The greatest positive departures from the normal ranged England, and the northeastern portion of the Micldle Atlantic from + .10 to + .18 inch, and occurred in the middle and States. ~h~ distribntion of ~h~ lIarometer It ~h~ I n all the remaining districts i t increased.