“lie instruments are sent well packed to the observer with R small spike and instructions for mount.ing it., which in our crae n.re the Iollowing: INSTRUCTIONS FOR MOUNTING THE BAROMETER. 1. >h I i vow height on tlw mall where the inst.rumeni.is to i.le p l w ~~. 2. K‘nil :he sm:LIl tipi’te S cci:timei.eiy a111ne the n i d i . so as to plwi~ the cent.w of din: :it the snme hei:,lht. I I ~ p u r e y s . ’k ANTES DE. LEER PIG. l.--Did Io: aneroid barometer we9 in the Cnl-eii XetwroiugicR1 Si.xirr Besides, in the letter to t,he future ohserver nre recom- niend to him to place his h e d in such n position t,ha.t. lie 11in.y see by reflection his eyes in tlir instmnient., avoiding as far as possible pidla.ctk errors. The instrunients are sent t,o any place witliout. know- ing t.he height, above sea level. In fact., we do not care about trliis, u.s measures will be expressed finally to cor- respond to those of niercuria.1 barometers, corrected for tmi erature, sea level, and normal gravity. twronieters for the determination of the constant of each inst,rument. We roceed in the following manner: Eeadings are t.a-en f a t 7 a. in. and a value is assigned to t.lic word sent each day from the isobar that crosses t.he plaice of the observer. In normal went.her, 20 obscrva t.ions will suffice to eliiniaate all errors of dra,wing c! isolxrs and give n very a,pproximu.te vnlue for a chosen \vorcl. For instmce, suppose the observations sent are the following: 0 F course, we have to depend on readings of mercurial [!.mistance between words- 1 mm.] ~I i :w I ...................... ..j IO.:;% ......................... 1 ifi2.5 I t o ~==i ~.6 .Ti!iw 2 ......................... togalam:. .................... iti3. I ! t&a=;B?.ti JII:;~. x.. ..................... 8 !ocl,~ ......................... toga=il;2. i .Iii!ie 1 ......................... two .......................... &. 1 [ toga-X?. 1 .~!!I P i ......................... olivau. ....................... iB1.S ! toga=iG?.8 ia. i I , -0 .Sii:n~ l i ..................... .._I !!ivaiiroId L.. ................. 761.1 I to~a=iE.li br~!:~v : .................... ._._; r a h . ..__...................I iw.0 togs=i&O The inem of 20 readings will give ti fair value for the rending “ t.c,ga’’ a t t.his specin1 place. Then all otlier ralues for tIic Y tntion are inserted nnd t.he instrument cnn t.lien reader service. Other detsrniinntions of the constniit for t,he svorcl chosen can be made from time to time. We hnvr found this to be the most economical and esiwt procedure of obtaining air pressure observations from nont,e:.linicnl observers. I t might, he mentioned t h t this device of introducinc. t,lie code directly in the instruinent could be utilize3 dso jn wjncl direction nnd velocity. We intend to do so in the future. In developing this system the writer had the valuable assistance of Mr. Mieuel Gutierrez Cenizos. observer of THUNDERSTORMS OF JULY 13, !92& IN THE DISTRICT OF COLUMBIA, MARYLAND, AND VIRGINIA. By ,~I.IWED J. I-Imiay, llcteornlogi:i?;t. [Weather Unreou, Wrshiiigtoii, 1’. P., .\u:ost l.:, 1922.1 The series of thunderstornis that occurred on July 13, 1982, as well as ot,liers which ha.ve been esperienced dur- ing the resent season, clearly lacked the usu2d evidence to another that is a characteristic of thunclerstornis in this vicinity. It has seeined rather that the storms have developed directly over the city and spread locally froiii that point as u center. The usual progression of thunder- storms in the vicinity of Washington is froni the n0rt.h- west to the sout.hea.sb. The object of trliis note is to ascertain whether t,he usud stmatistics on the occurrence of t,hunderst,orms in the ad- joining States of Mar land and Virginia would throw any light of a definitive. c. T iaracter upon the progression of the thunderstorms of Julv 13, 1923. Mekorologicd con&io.ns.-At 8 a. ni. of the clute in question a trough of low pressure stretched from the mouth of the St. Lawrence sout~liwest~ward to t,lie Caro- linas. This trough was flanked on either side by higher pressure7 thd level of the bnronieter u t the center of the respective anticyclones being 30.30 inches. The winds of a de B nite niovement from one point of the compass were south t,o west in the trough and northerly on its western intq$n. The temperature u t S a. m. at Wash- ingt.on, P,,zlt.uiiore, and in southeastern Pennsylvania was SOo or al~ove and the humidity was relatively high, both conditions baing favorable to the development of thun- derstorms liiter in t,lie dtiy. Tile A t y i t t i i i q of tke .sh~rm at Tashiqtton., D. 0.--The first thunder was heard n t 4 p. m. and rain began 5 min- utes lnt,er. The locd sky sips? however, indicated that preliniiiiag thunderstorm activity had been in progress over the city for at least an hour. At 3:05 p. m. thunder was heard to t,he east of t’he stttt,ion, but furbher develop- ment of this dist,urbance ceased. At 320 p. ni. ty ical the northwest.ern segnient, of t3ie sky mas devoid of tliunderst,orin cloucls, although shortly before t,he storni broke a confused mass of clouds R peared in that which hac r hitherto been confined to the south and east, appeared in the northwest and very quickly thereafter bliunderstorm clouds tip eared to the southwest and t s lese thickened and apparelit P p moved to the east; meanwhile of the sk . At about 4 p. ni. t P le electrical dis8;: JULY, 1922. MONTHLY WEATHER REVIEW. 361 rain began. The winds before tlie storm were froni the northwest and north, occasionally backing to northesst. Heavy rain began at 4:43 p. m. with the wind in t,he southeast; the rain slackened somewhet about 5 p. in. and the wind shifted to northwest, rain again becoming very heavy. From 4:43 to 5:43 p. m., 2.79 inches of rain fell, being the greatest amount for an hour ever recorded a t the station. The total for the stoiiii was 3.37 inches. Inquiry as. to the amount of rainfall a t near-by points in Virginia ap eared to shorn that. the intensit.$r of the storm in that J tate was much leas thn:i in the District of Columbia. At this juncture i t was decided to make an inquiry as to the dei>t;li ol r.in€ idl and tile roaressimi oi' the st.orm in the adjoining Shte.'~ of Maryfanx a i d Vir- ginia. The section director,i for those States have $11 - lied me with data on both point,s and I have fnrtuaat.e r y gee, able to secure from the sewer depsrtnient, of t.lie The least catch was but 65 per cent of the average for the District and the greatest was 139 per cent of 'the average, or a range from 39 per cent above to 35 per cent below, and t,his is probably not far from the normal es ectancj- in similar c.ases. So far as known, the only was a t Aberdeen, Md., and at Richmond, Va. The re- ot I! er point a t which escesaive rains fell in a short time direct interest to city often possible to trace across a State, and t,he thunder- st,orins of Julyv 13 were 110 osceptions to the rule. On the date in qut!stion a t least two serias of thunderstornis were observed, the first. bpginning in the sstreme western vouiitks of hIarylanc1 11s early us 11 a. m. and progressing eastward or developing irregulnrly fnrther and farther t'iist, the second apparently developd about 3 :30 or sewer construction.l FIG. 1.-Precipitatioii in Washiiigtai. 1). C. liiid!es!, July 13. XI??. District oi Columbia rainfd records for tlie s t, orm as measured at 24 points, mosth- within tlie populated area of the city of Washin ton. hiese records, in addition to two others mintaine 5 by the Wentlier Bureau, will give u distribution approximately of one rniilfnll station for each square mile of territory. Mr. J. B. Gordon, snuitnrJ- engineer, has supplied R sketch ma of Washington upon Fi . 1.) h e catch of each gauge has bccn entered directly upon t.he map in the position occupied by the gnu e. It will one in the northeastern suburbs, as a t Benning, and tho other in the southwest in t.he Rock Creek region. Be- tween the two there i3 an nren of diminished fall estcnd- ing in n NNE-SSW direct.ion with the minimum fall 1.P5 inc.hes at Congress Height.s. whch I have plotted. the rainlid f of the storm. (.See be seen at once that there were two areas of a eavy rains 4 p. m., ju3t west! of Frodrrjck County niicl progressau cctst.ward atid sout.liward rather slowly until it reached Dditwarc? Bay between T and 8 p. m. In Virginia the earliest t.ime of beginning was 3:20 p. m. a t Mount Weather, in tlie estreme northwestern ;'art, of the State. In Loudoixi County, directly east of Mount Weather, two stations report the beginnili as a t 4 p. ni. At Washington, D. C., nbout 60 miles grectly east, the beginning mas a t 4 : 05 p. in. Thunderstorms also began a t the same hour at Ciiarlottesville and Hopewell, Va., both points being in the south-central part of the State. The times of beginning in both Maryland and Virginia points is shown in Table 1 on nest page: 1 CI. The distribatiuu oi rsiiifall owr restricted 3teas. MO. WEATEEB REV., 49: .IO1 -.104. 362 MONTHLY WEATHER. REVIEW. JULY, 1022 rnentsd- ................... Northern rou Frostburg.. .................... Cumherland .................... Clefu Spring.. .................. Chewsville ........................... Keedywille.. ....................... Frederlck ....................... Emmltsburg ......................... Westminster .................... Freeland ........................ Derllngon. ..................... Fakton.. ............................ Aberdeen.. .......................... Wilmlngton, Del ................ Ceeilton. ....................... Boyds.. ........................ Great Falls. ........................ Takoma ............................ Washlugton ..................... College Park..- ................. Laurel ............................... Woodstoek Middle group: ..................... TABLE 1.-Time of beqinning of thunderstorms in Maryland and Vir- ginia, July 13, 1922. MARYLAND. G3rret.t.. ......................... 11:nn a. In. .:l!efny.. ........................ i:W 1). m. o ............................. I l? noon Mk3hington ....................... I ?:%I p. m. do ............................ .i 4:nO p. m. .do. ........................ ....I 4:W p. m. Frederick. ........................ 4% p. m. do ............................. 4:!5 p. n. Carroll. ........................... ?:oO p. m. Raltimore ......................... ?:oO p. m. do ............................. 00 n. m. do ......................... ....I %5 i). m. New Castle ........................ I 5:. 0 p. m. Cecil.. ............................ Monlgomrry ...................... .do.. ........................... .do.. ........................... lristrict d roliimhis.. ............ Prince Georges.. .................. do ............................. Baltimore i IIarford.. ........................ .I 9:30 p. m. ......................... Station. ................... ......................... .................... ................... ................ ................... ........................ Baltimore. Sudlersvilla Rock Hall-.. Bouthern group: Cheltenham. La Plats Ferry Landing .................. Solomons.. .......................... Almspolii.. .................... Ridgely.. Dover, Del ......I.. ...................... County. .... ............................ ........................... ..................... .............................. .. ............................. .................. ............................ ::m! p. rn Qucen Anue. l e n t Prince Georkw.. 4::W p. nr. Charles :,:Yo 1,. 111. Calrert. ........................... do ............................. {.&O p. m: Anue -4nmdQl.. ................... 5:lM 11. m. C‘nrolinc h.?O 11. Ill .do.. .do li:MJ p. 111. ........................... {.m. ,iTL~emlL. ....................... ~ i:w 1’. Ill. ................... 1 D. N.-Dutiqg night: D,. D..-During day. The stntiohs are arrFpd arcordim to gsogra hit? position beglnwg rn the e\trrme northwest and procwing due a t . then ret& to the w e t and proceding east as 1Jrfnre. The evidence of the above table is not conclusive either for or against the idea of nonprogressire move- ment; it, however, seems to su port the belief that thunderstorms developed east 0’; the Blue Ridge in both Maryland and Virginia at ap rosimately the same vigorous of them moved east-southeast or due east. It hour on the afternoon of the 13t K and that the more is also clear that there were at least two distinct series of storms in both States and that the second one was the more violent of the two. The concentration of the heavy rainfall in one or two localities must be explained as due to local and favorable conditions at those oints. Very heavy rain also fell in Maryland on the got!, due to a barometric depression central on that date on the coast. TORNADOES IN SOUTH DAKOTA, JULY 8, 1922. ByM. E. BLYSTONE. [Weather t :m u , Huron, 8. Ualt.1 In the evening of J u 1 ~ 5,1922, two tornadoes occurred in rather close prosimity near the southern border of South Dakota, one a t St. Charles, Gregory County, which borders the Missouri River on the west, and the other on the southern shore of Lake Andes, Charles Mis Count,y. w11ich borders the Missouri River on the east. The distmce apart was about 30 miles. That these were separat.e and distinct tornadoes seems t,o bo shown by the f w t , that no destruction occurred between the t,wo places, and by the further fact t,hat the one a t La.ke Andes is reported t.0 hhre occurred at 10 p. m. and the one at St. Charles nt 10:15 p. m., while both moved in a genemlly eastward direction. Both these tornadoes appear to have been of great violence. The one which occurred at Lake Andes came from t,he west.. It missed the t .o m of that name, but it dest,royed about, 80 summer cottages, 2 large denm halls, ttnd 5 large barns on t.lie. soutli shore of the lrtke. Fifteen persons were injured, but no one wm killed. roperty loss is est.imnt,cd nt npproshately $35,000. southenst, dipping clown again at a distance of about 13 milps. but appnrc.nt.lp causina no destruction. The debris 1n.y t,oir-ard the south, in8catin t.hat the build- ings dcstroged were on the northern si f e of t,he tornado. There wits a fnnncl-shaped cloud. The tornado :,t St. Charles was more destructive, possibly due to the fact t,hat more buildings were in its path. This tornado is believed to have originated near Brockshurm Nebr., about 15 miles southwest of St. Charles, wv%ere some barns were destroyed. However, it appears t,o have caused practically no destruction in South Dakota except at and near St. Charles. Prac- tically all the buildings in St.. Charles were destroyed or daniaged. One gra.in elevator was upended and pritc.ticnlly destroyed. Another near it was dam*ed only slightly. One lumber vard was destroyed and the 1umbe.r scattered. The totd pro erty loss 18 estimated a t $150,000. One person was tilled and nine were injured. There is no report that a funnel-shaped cloud was seen, probably due to darkness, but as the path of destruction did not estend beyond the limits of the town there seems to he no doubt but that the storm was a tornado. s was slight. After causing ths destruc- tlppwently lifted and moved toward the