MONTHLY WEATHER REVIEW VOL. 50, No. 10. W. B. No. 7S9. OCTOBER, 1922. CLOBED DECEMBER 2,1922 ISOWED JAXUARY 20, 1923 PROCSRESS IN RADIATION MEASURMENTS. Uy C. DORNO. [DNIP, Switmr13ni1, -4ugust, 1922.1 sY.yorsw. The aut.hor discueses apparatus that. has been em lo! ed at the Davos Physical-~let~oroloical 0 bservatory to ineaaure tfe total solar radia- tion received at the surface of the earth, the intensity of radistion received from restricted region# of the eolar spectrum, the intensity of irradiat.ion, or outgoinw nocturnal radiation from a hlack body. and also the d i a t .i o n rcceGed at the earth's surface from the atmosphere. Especial att.ention is given to regist.ering apparatus? priucipallp of the photographic typo. A thermopile for measuring the total radiation, potassiuni and cad- mium cells for meaauring the radiation ffm t.he. visible and the ultra- violet regions of t.he solar spectrum. and ngstrom's tulipan lor nwamr- ing nocturnal radiation find favor wit.h the author. In a brief summary of the Dams inemurcnient~ it. is shoKn that. il we take into account the albedo of snow in winter and of the ground surface in simmer, throughout the year "I-Iardly a third of the incoming radiation has cont.ributed to the heating of the air and the IncltiEg [of mow or ice and evaporation of moisture." The resu r' ts of recent medical researches are cited to B o w the value of the measurenient,s of radiation int.eiinity in the uhix-violet region of the solar spectrum.-H. H. I<. From its be inning in 1907 the Davos Physical- Meteorological bbservatory has been endeavoring to dcvelop reliable methods of regist.ering sun and sky radiation, both as to their total intensit and also the The author induced Carl Zeiss in Jena to construct for hini a permanent spectrograph for ultra-violet radiation. This instruiiient was described in 1911.' Its chief object is to fix the variations in the est.ension of the sun's ultra- violet s ectruin with the time of day and season. Quite lately t f e tables given (1. c.> have met with an unespected and keen interest, since Haussor and Vahle have proved by their very careful researches that the igment-forming power properly belongs to a very smai portion of the s ectrum with a sharply pronounced iliaximuin between t E c lines 0.302 and 0.297p, from which the effect upon the epidermis sensibility falls ra idly on b0t.h sides (the brane). In Strahlenthempie, Band SIV, Heft 1, t.he aut.hor has stated the interestin biological and svolu- interest t.0 medicine, especially to the therapy of tuber- culosis of the lungs, since years ago Rollier and others advanced the thesis that healing result and pignieiAiig power nre goportional. EVIEW : n method of reaistering local atnioa- pheric clearness has been describes; that is to sax, the pliysiolo ical action of sun and sky on the hurnnn cye is measure 2 in term of its effect on the horizoiitd plane by the photoelectric method, by means of a h~ghly evacuat.ed potassium cell, at a low potential (2 to 4 vo1t.s) under milk-glass and filter (Scliott F 5599). This is n very reliHble nnd easy mode of registering radiation for H, s ectruni region of the reatest practical importance. In intensity of the more important regions o P the spectrum. research has not been e-xtende 8 to the mucous iiiuni- tional conclusions this result inay f eacl to. It is of specia.1 In this t K e same paper the aut 5 or has given data, to prove that 1 Studfc iiha Licht uad Lu t dtr Hachgrbirgcr l'ieweg 1911. 3 Slrablmthcraph Band d111, Heft 1, Seiteil ff., Lhnn Sr Svlrwarzeiiberg. Berlin. J MO. WEATHER hEV., June, 1920,48: 348351. under conditions of hqh altitude the nocturnal effective rutlistion may be reghered with sufficient correct,ness by ineuns of hgst.riirn's tulipan,4 which is hased on t.he overdist,illation of &her. Comparisons between the records of Angst.roin's tuli Ian and t,he pyrgeomoter are being continued. The incicntions t of t.he former arc in error only if the yeatest gossible [condensation for Daros occurs; then t ley are n out 15 er cent too high; that is to say, only durina the time o actual condensa- tion, not; in the nocturnay mmn. Therefore corrections are genernlly unnecessary. Farther on, this inode of registry of radiation will be more fully t,rc?atetl. There is also mentioned a niethotl of regist.ry of t.lie nocturnal effective radiation by nicans of the yrgeometer, t.he compensation beinc. eliniinatecl, hy p R otographic. indi- cation of the oscifiation of s sensit.ire g;ilvitnometer. This methocl is not to be used, escept on perfect.ly calm nights in miclwinter, since the faintest breath of air is also recorded Tho chief aim in these endeavors must be the con- struction of a reliable pvheliograph for the purpose of reaistering the solar radiation in absolute heat units or caTories. The author has sketched in rough outline a vrheliograph which combines t,he principles on which 1Slichelson's actinometer and Angstriim's prlieliometer are based. These outlines were elaborstec in detail by Dr. ltud. Thilenius in Darmst.adt, and t,he co per body and the thermocouple wero constructed by iini. The firm of A. Pfeiffer, in Wetzlttr, undertook t.he construction of the other parts of the inst.rament.. (See frontispiece.) Registering sun intensity by inems of a lheiiiio element 5eenis very simple to an ouhider, but whoever is acquainted with the historical development of t.he pyrheliomet.er remembers thst for a whole century the most prominent scient.ific men have hceu occupied with the problem 0: how t,o keep off the radiation ii:tluence of tho surroundings, when iliaking radiation measurements, before t.hey were satisfactorily snlved. 111 t,he present case-as with Micheleon's actinometer-the result has h e n attained by putt.ing the semitire thermocouple in a small c~1nem of a mewive cylindrical copper body of 3 kg. weight., whrxrc! large cnpacity keeps the caloric variations of the surroundings sutfiriently damped, and whose great transiniasibi1it.y briugn about t.he temperature cwnpensation around the small camera very ra idly ("are ia taken that in t,he forepart of tho camera plenty of rc.!ection leads the last traces of stray rays back to the intcrcepti1:g surface. The back of the camera hss been carefully blarkoncd in order to absorb :dl the rays that may have pierced through from t.he front. The cylindrical camera is 8 mni. in diameter and 23.92 mm. long: the therniocouple is ou.ac*tly centered and stands symmetrical to the declination axis. I t is 15.18 Ium. long and consists of 1Y elemen& of constantan clopper wirh an intercepting surface of 30 nim. The latter is put slightly extrafneal (2.8 mni.) to the focal plane of a quartz lens of 20 mni. opening ahid 8ti Inm. focal distance which t.herefore show in the focal plane objects of 1' angular extension 1.5 inm. in size. Owing to five diaphragms which can be u t before the!enn about. +, 4, f. 4. I% of the incoming intenaity may ge utilized. while a biightlg poliehcd shelter prowcts the copper body from insolation. The in-radiating 11111 and sky zone may, at will, be limited to 8.4 or 12.8 or 47. 8 arc-minutes from the limb of the sun by three exchangeable diaphragms lying in the focal plane. For permanent registration the front diaphragm of 7.08 mn. diameter and 4 XoEa acta Rcgiae Societatis Srirntlarum Cpaulirnsin, Ser. IV. vul. 2, XI. I], 1910. 515 26495-2-1 ' M. W. R., October, 1922. (To face p. 515.) -- I .I FIG. 1.-Pyrheliograph. (Seep. 516.) MONTHLY WEATHE$ R.EVIEW. the largest back diaphragm are used, which lets in a sky zone of about three-fourth degree of arc. Small deviations in the day run are in- evitable, partly on account of the variability of the solar declination during the day and part.ly because the mounting of the instrument, which is attarhed on the roof of the house, does not possess nat.ronomica1 fixity; but the chief reason is t.he slight variat.ions of the clock due to the great variat.ions of temperature. With this arrangement t.he elvanom- eter readin-s remain unchanged even when the diopter image de- viates half -its diameter from it.s exact centering. The temperature developiug inside the copper part rould be controlled by means of an added thermometer. Even under est.reme conditions it did not nearly attain thc critical temperature of FJOO C. at which t.he shellac of the bhemocouple mi-ht. l a w softcned. Vnder t.he stated conditions, the thermocouple indiicates a t full insolation about 0.546 millivolt., so we were obligc4 to put. in a resistance of 2.000 Q before the Hartmann & liraun mirror galvaiiometer of I .'LRX10-9 of sensitiveness, whose resist- ances \+-a3 1,000 12. When the sun is Hhining brightly the rcgiskred wading8 vary. accord- ing to the ~olaralt.itude, between 153 and 60 nim. (1 1111ii.=O.OlI.i4 cal. ). an hour int.erval corresponds to 20 mni. width. Owing to the slight heat capacity of t.he thermocouple which, in spite of its minuteness. call not be iiidc to dieappcar eiit.irely, there exist,s a Aoniewhat great.er inertia of the gdi-aiioniet.e which aniounts to that observed with the registering 1)y nicans of knk pstriim's pyrononietcr. During the gal- variometer's period (hardly I f J secondsj B3 per cent. of the incoming intensity may be regist.ered. Thc rise progreeses in the form of a &pep exponential curve; but the last scwenth follows slowly in about 13 niinutes. 1 n parinanelis registry there result8 autonlatically a sli$it compensat.ion of the rume which howevcr, is hardly noticeable wtli the scale wed. Tho innsinid., which must show inore distinct.ly with permaiieii t registering t.lian wit.h single measurements, were nc~ert.lieless lower during t,he rcgist,ering period (October, 1931, to June, 1923) than the abso1ut.e masi- mum. For 11 years (1910-1921) since March 5, 1910. the 1att.er has been l .j i 5 (Smit.hsonian Scale Revised, 1913) and was only sur mssed May 6, 1921, after a : *now- During the registoring period the caloric sums of cloudless days have been recorded 1.S per cent higlicr than those derived from single measurements in the years 1909 and 1910. A comparison for all days between the number of reg- istered calories by the pj-rheliograph aiid the number of caloriea derived from registrations of Campbell-Stokes heliograph (sunshine duration t.imcs number of cloudi- ness) indicates & high. Froin t.he beginning of January, however, they wcre below tlic normal height.. optical dist.urbance was not recorcld-except. between May 24 and June 9, perhaps. The quest,ion nriscs wliethcr t.his optical dis- turbance, w1iic.h occurs awry year t i t t.liis epoch of tlic calendar, is not. an opt.ica1 ii:dicatioii of the approa.ch of the well-known return of cold in June. In a paper in t,lie iUttwro7o;yi.sc71~ Zeitd)~i#t, which trent.s niore fully oi t.heso results, tlie optical disturbances during and afhr the important aurora in May, 1921, snd during the t,iino of great.er solnr nctivity at the end of June, .July. and August, 1021, are enlarged upon. The frontispiece shows the instriinient from the side view in the Flare where it is set up. The polar axis resta with globe-shaped pivots in the cylindric layers of a latitude arc of +-shaped cme section. which is adjustable in the limita 35' and 65' by means of a clamp at the edge. The lower part forms a console. This carries (1) the clockwork with the minute dial and (3) the springe of copper, which are insulat,ed with slate and whose places of contact are plated with 0.2 mm. of fine gold. The rings, also, whiFh are borne by an ivory ring, laid over t.he lower end of the polar axis, are of electrolytic copper with 0.2 mni. plating of fine gold: so the thermoelectric effects remain extremely minute. The conducting wires of the column are led throuph the hollow axis pivot (visible in the figure) of the declination 8x18, and joined to the gliding ringe over copper transmkion fixed to the framcl part of the polar axu. . On the other side of the declination axis is a volute wheel with 3G0 cop and degree scale on the front Burface, int.0 which a tangent screw encroaches for the fine adjustment i n declina- tion. Ita drum scale indicatee 1'. Waterproof covered azimuth-cor- fall with 1.557 at 5 5 O so \ ar altitude. er cent higher registered values. Until t,lie end o P Decenihrr the values were ahnornially OCIWBER, 1922 rection screws whose full rotation roducea 30' azimuth change, per- mit the full adjustment in the meri8ian. A le- diopter permanently controls i n the easiest manner the correct adjustment of the axia of the tube parallel to the radiation, a number of counterweighta regulate the cornpeneation of the mmes, EO the clockwork remains uninfluenced i n all the different positions of the copper part. The copper part con- sists of a right and a left half; they are joined by a front and a back centering cover and by three stron screws w i n g acme. When them? are loosened, the small camera insiie, whici contains the thermocouple is accessible. The lens tube (to be seen on the cutj wi% screen, front and rear dkphra m ctnd quartz lens is carefully put i n the copper part by means 088 cone and is kept i n the front centering cover by hinges and pressed into the co per part to the complete contact of the surface. Doctor Thilenius wifi 8oon publish a minute description of the pyrheliopraph in the Zeiifschrijt Jiir Insbzcniettttlnl~rrirle. Owing to the skill of Doctor Thilenius a dill smaller miniature thermocouple of 13 elements of special alloys has been constructed, which, when fully exposed to the sun, gives about 4 millivolt, a power which would enable us to replwe the photographic by a mechanical registration. The instrument, wliirh has been in permanent opera- t.ion since October, 1921, i. e. now more than 10 months, has stood all t,ests. Comparisons with Bngstriim's conlpensation pyrheliometer have givcm the relation 1 nim. = 0.01174 cal., the deviations from t.his relation keep within one unit of the fourth decimal and are in nowise sys bematic with one escept,ion; when the sun is low aiid the sky not quite clear, t.ho relation always rises to 4 per cent,. The reason for this IS not t.0 be found in a different reackloii to the irradiating int,ensity, but in variations of the intensit,)-. A s has been said above, the yrhelio rtiph permits only a sky sector of Michelson silver disk instrument, however, u much larger one of about 5'. Thus tlie pyrheliograph gives the more esact (the sninller) readings. This question having. been treat,cd .elsewhere 9, we shall not enter upon it here. Of the rcsu1t.s realized w e shall only mention the following: If monochromatic filters are placed before the pyrhelio- graph, the intensity of separate part.s of t,he spcctruni may theoretically also be regist,erecl. But two mis- givings esist: (1) The avttilahle filters are not sufficiently monochromatic and es ecittlly not suficientlr examined as t,o their t,ransmissikility in the u1t.m-red spectrum; (2) owing t.o the nbsorption of the incoming radintion the filter itself becomes the source of long-wave heat radiation, which vitiates t.lie measurotl result,s. This latter danger mav be met by learing. a sufficient air layer between the filter and the opening; the former can only he avoided in t,he rad spectrum. The other- wise escellent Wrat.ton-gelatine filters must be trans- parent to u1t.m-red mgs, else they could not., contrary t.0 proved laws, record n.t sunrise smaller intensity increasc in blue-riolot than in green, and in green smallcr than in red. Red glass Scliott F-4513 has proved highlv monochromat.ic and seems to bc opnquo to ultrwred. After deducting t.he absorption of tlie 2 mni. filt,er, there resultod for t.he mean of the yFar the following t.rue red port.ions of the total i n t ~n s ~t y as a function of the air muss 6): f 0 of the ec P F ge of t ie sun t.o irradiate, the Angst,riim- __ -. - - ._ .. - .- . ' I :,, I : \\h l t C ; I ! -- --- -. ... I . I I 2.5 i 3.0 O.tii3 i 5.0 1 0.748 i 0.892 ! 0.i11 ! 3.5 ! 4.0 dbliafldlungm de6 PrrussisclrPn iUeeleorulopbehcn Iflstiluls. Band VI, 1919, p. 61. OOLY~BER, 1922. MONTHLY WEATHER REVIEW. 517 193. : ' C . ' drill. KO\,. 19 ................. I -5.3 I 0.1 I ! I Dec.55 ................. I 1 .2 ) 2 .i 8 Jon.55 ................. i -4.6i 2.38 1921. Peb.23 ................. 1 -6.6 I 1.47 Mar. 10. ................ , -2.7 I J.72 An annual variation in the red >ortion is clear1 to October till March; in A ril and May it diminishes but for the most part below the spring values, but some- times surpassing them by far. In the atnios here of from September to the October minimuni. Parallel measurements made by this met,hod at Davos (1,600 m.) and Pot.sdam (100 m.) show that in the clear month of October the red portion of the solar s ectrum a t Davos a t 20' solar altitude is 7 per cent, ami at 30' solar altitude is 4 per cent. smaller than a t Pot,sdam. In this periodical e A. Aiigst.rom and the author gave an account of the pprmameter, constructed by Angstrom, its slight imperfections, its ap ropriateness for registering with conipensation eliminate$ and the first resu1t.s real- ized at Davos in November and December, 1920. The re istering has been carried on during an entire year (govember, 1920, to October, 1921). Another slight ini- perfection of the instrument has been found: When per- manently used the white magnesium oxide loses its delicate freshness; minute quantities of the light powder detach themselves under the influence of heat and arti- ficial dryness, and disa pear v.-it,liout, visible traces when the inst.runient is move x int,o a rooni to be protected clur- ing the niuht or for other mc:tsurenients. A slight hue of green tKen lies on the white. According to the dinii- nution of the reflective power the constant of tlic instru- ment increases. It is difficult bo re-whiten t,he two strips on account of their position between the black strips, be- sides tliere exists the danger that the magnesium smoke might at thesaine time touch the thernio elements. There- fore the author did not dare to do it himself. Rose, of Up- sala, however, carried it through in May antl June, and as a result reuistrationsweresecuredfor two decades inMay md one in &ne. The constant, that has been permanent,ly controlled, as was described,' ke t satisfactorily at 12.93 reason mentioned, it rose slowly t.0 16.68 on May 6. After the return from Upsala the instrument, whose st.rips had been newly whitened, recorded the value 12.90, but in the heat of summer and being sonietimes intermittently used for other pur oses, it rose to 15.53 at t.he end of tion and t,he inconvenience of having to calculate the regis- tered readings with changing factors in calories niight be avoided, or at, least be reduced to a niinumum, if t,he instrument could be kept free from every disturbance. The readings of t,he pyranometer and t,liose of the new pyrhelio trph could to some estent be coniparedwith each other. Fhe difference of registered solar + sky radiation, which falls on the horizontal surface, and the sky rudia- tion alone, iised by single measurements, must be &out equal to the solar radiation falling on t.hc noiiii:il surface times the sine of the solar altitude. The coin arisons were satisfactory and proved the relinbility of t R e read- ings of both instruments. The radiation of II cloudless sky on a horizontal surface may increase tho direct solar radiation by 18 per cent in the brbliter spring and autumn; in t.he summer months, wit,i an atmosphere rich in water vapor, b 24 er cent (relative to the mean .of tlie dag totac). J n clear autuinn days it amounts t.0- be observed. It rises at first rapi m y, then slower P roni slightly; June, July, an CQ August show changing values, autumn, rich in water vapor, the red portion fa P 1s again from November to the middle o P February; then, for the July and at the en c f of October to 20.50. This imperfec- .- a Mo. WEATEEB REV., vol. 49, No. 3, p. 135-138. 7 LOC. C i l . 0.236 (0.177) 0.N (0. Liz) 0.179 (0.161) 0.215 (0.1651 0.215 (0. lill Solar altitude: Cal. 10 O ... ....................................... 0.020 15'. ......................................... 0.038 Yo____ ...................................... 0.053 25O .......................................... 0.075 35 O............ .............................. 0.080 40 O ............ .............................. 0.075 4 5 ' .......................................... 0.078 Bright cumuli raise it to 20 per cent,, even if they cover only one-tenth of the sky; bright stratus at solar intens- ityS2-,raise it to the 3 or 4 fold value. As highest value of sky radiation was found 0.406 cal. at 44' solar altitude, the sky being bluish white with fohn. On cloudless days the absoluto masimum of sun +sky radiation was 2,165 (in June after a snowfall at GGo solar aIt,itude). Only once, on September 5, with clear sunshine mid clouds com- bined, has it been surpassed--by the estraordinarliv hi li value 2.491, when the whole sky was covered wit6 li&t gray nimbus clouds which the sun pierced sudden- ly. Of course the general action of clouds lessens radia- tion. In the pearly mean the normal values lose 21 per cent through i t ; in the three winter months only 19 per cent, in the three summer months, however, 27 per cent. During the snnie period (November, 1930, to October, 1921 ) the registering of nocturnal radiation has been car- ried on by means of the tulipan during all the favorable niulit,s, beginning antl ending with a solar depression of Go: The very considerable material obtained enabled t.he rruthor to make the sunis of the effectim radiation 2nd of the radiation of the atniosphere derived from it,, and to esamine into the influence of cloudiness, as well to its degree as to its kind. TABLE l.-i\'ncturnd efleclire mdiution -4 nnd r a d i d o f i nf the atma- phere E. (1utegrat.ion values of the whole night from 6' after sunset to 6 O before sunrlse.1 Temper- Abs.hu- A Date. 1 ature. 1 midity. 1 8 n. 4% 0.463 0 .4 5 n. 113 0.427 0.469 0.430 n. s24 0.51s 0.54.5 0. .W{ n. 4v; E __ 0. 18.1 0.2H2 0.246 0.195 0 .2 2 0.2'73 n. 'm 0.339 0.320 0.366 0.303 n. 3111 Em -- 0.34 0.368 n. 349 n. 280 n. 351 0.307 0.319 0.390 0.3i3 0.405 0.363 0.35: Nov. I9 means night from Nov. 19 to 20. Radialion constant- 6.1~4 i: 10-1 I . A= Efferti\-e radiation. S=Computed radiation of a black surface. E- Radialion of the atrumphere. EP Radiation of the atmospher of 3oo C. The nuriibers in parantlieses are -fngstroni's parallel rahies. The tablc shows that on perfect,ly cloudless nights the effective radiation depends more on the humidity of t.lio atmosphere than on 1t.s temperature. The radiation of 518 MONTHLY WEATHER REVIEW. O ~B E R , 1922 3 819 6:W 3,872 3 W4 4’897 5 152 5:787 5’220 3’766 r:m \-ear ....................................... I 1.50,ml ~.3 6 3 the atmosphere, however, de ends more on the tempera- other in their effect on the effective raJation, the yearly variation is slight, and only the clearest months of spring and autumn show high values. The yearly run of the ra- diation of the atmosphere, however, shows, as was to be expected, great amplitude-nearly 1 : 2 (0.1S-l in November and 0.366 in Auoust) from tho cold to the warm season. Corn )ared to h e effective radiation d u e s ture. Since temperature an s humidit counteract each calculated from 8. t i ngstriim’s formula8)- E,,=K-C- I O -y P K = 0.439 c =0.158 y =0.069 p = vapor pressure in nun. those found at Davos are 23 per cent, higher, in perfect oorres )ondence to A. Angstrom’sR) proportional numbers This results from the addition of the corresponding col- iinins. This conforniity proves once more that the atmospheric conditions of t.he Swiss Alps and the Cali- forninn Sierra wliere hgst.riim has collected the nlnterial for his formula are very similar. Of 123 nights whose mean cloudiness at the be inning percentage loss could be derived for the cloudiness degrees 10 and 5, lis compared with the effective radiation result- ing under the same conditions of humidity and tempern- ture when there were no clouds: for di f+ erent altitudes (0 meter, 0.41; 1,500 meters, 0.34). and at the end has been tnken into account the fo f lowing +W1 +s388 +8’M +11’975 +12’767 +14:89j +ll,Fll +a %I +5:331 +m,m +14’Sw) ~___ I B ................... Bs ................... I ~ If we express the relation between cloudiness and effec- tive radiation with AngstrCm’s formula- Am = ilk.nd - A,. m =degree of cloudiness (in tenths of the covered sky A, = efl’ective radintion ut, this degree of cloudiness 9, =effective radiation with cloudless sky X. =constunt the following constant vnlues for t.he different, kinds of cloudiness will result: i!i.Pt. .................................. 0.031 A.Mt ..................................... 0.062 St.Cu. .................................. 0.085 The results of observation and calculation with other degrees of cloudiness a ee satisfactoril . Compared cirri is greater in the high mountains. This is hardly to be wondered at, since the I add a screen of c ual thickness Admitting that the long-wave radiatign of tlie day sky corresponds to that of the night sky, based on C. G. Abbot’s and A. Angstriim’s measurements during solar eclipses, the long-wave radiation of the day sk has been surface) NIJ ..................................... 0.089 wit,li results of the lowlan r s it seems that t Y io influence of Yl f to a relative1 thick one a f ready esistin in t e valley, and to a relative T y thin one a t the high a titude. calculated according t,o hgstriim’s formula 9 rom nbso- ~. . H Smithsonian Minec.lluaeuur cdleetions, Y O ~. 65, No. 3, 1915. * $f~trorologisthhe Zritselirijl, 1916, 11. 534. lute humidity tion times for all the night total formula -4--- erature from the three observa- to twilight, of which using the well-known For the other 195 days, of which no nocturnal observations were available, the mean a*+b+c 7 tlie fraction resulting from the formu T a A4m= (1 -h )A ,, was used to calculate the radiation of the 24-hour day. Of the radiation values to be expected with a cloudless sky under the same conditions of humidit and temperature (for the four different k values mentioned above) was tnken for every kind and amount of cloudiness. In the values of the monthly radiation totals thus received the maxima are again to be found in the months known to be the clearest; the am litude, howerer, is much greater. 1.6 from June to Marc[, and cloudiness eserts n greater influence than either temperature or humidit . Finally, the whole heat eschange by rach.tion has been derived from this material (1) for the effective prevailing conditions, (2) under admission of n cloudless +ky. TABLE 3. Month. ......... .... .... i-----i------i-- 19!20. I ?;730 I tal. 1,925 1 -1% Xovember. ................................... ........................................ 1)ecember 43i6 4,641 -285 1921. i I January.. ....................................... J 770 February. .................................... Msrrh.. ........................................ 1 l$g 1 A P ~~I . .......................................... 15: &)i h h y . .......................................... 16.671 Tune. ........................................... 13.616 July. .......................................... August.. ....................................... 1 I September.. .................................... 14,037 October.. ..................................... 11,118 THEORETICAL COMPUTED CHANGE OF IIEAT; CLOUDLESS SKT. ~--- I I I S, 079 7,991 9.590 R. 381 9,241 Sew. ...................................... ___-- ‘ 1920. N uvem ber .................................... ............................. ..........I 193. 1 ._-.. January. ........................................ Pebrtnrg.. ................................... 1 12% I Unrrh. ..................................... 15: WI .\pril.. ......................................... ‘~0.510 .\lily. .......................................... June.. .......................................... July. .......................................... ~ ~~~~~ AUClllSt.. ................. I. .................... P;501 September. ..................................... I J ,R I October.. ....................................... ./ la, ;E 1 -4,139 -2,1131 -1 3.10 +1’371 +12,Wl +I4 481 + 19’ 294 +15:7fll +14 510 +3,11 +ti:w; +7: W! The effective yearly heat csch:iii:;.r shows with 92.7 kilogram-calories persquare centimctt.r ti somewhatgreater g:iin (S per cent greater) tlinii thr thcoreticd one; the reason for i t is this : t h nutgoing mlintion is diminished 45 per cent nnd the incoming only 21 per cent by the clouds. With :t continidly cloudless sky the months oi November t.o J when the sun is lowest, would bring 11 considifyzrks of heat through radiation, but in reality we find it only in the months of November and December, when their cloudiness is not considerable and even then not in n high degree. Altomether 3s per cent of the incoming radiation was returnea and 62 per cent OCTOBER, 1922. 1CIONTI.ILT WEATHER REVIEW. 519 was retained. This calculation holds for the absolutely black surface and changes completely if we take into account the albedo of the earth's surface. If we ut for cover (December to April) the snow absorption 0.3 (albedolo in the mean 0.7) ancl during the seven months without snow corer the mean libsorption of meadow and stone 0.9 (albedoso meadow 0.06, gravel 0.13, while the long-wave eff ectire radiation remains unchanged (for according to J. Maurer's measurements and from theo- retical reasons t-lie snow radiates like a black surface, and the same may be admitted of the vegetation, humus nnd stone cover) we get: lnroming radiation in December-April ..................... 14,453 (lutgning radiation in Dercmher -April. .................... 3 .6 3 6 the incoming radiation during the five months wit K snow Oiitradiated.. ...................................... 0. IS3 91.666 33,727 -- Incoming radiation in May-Xovenibcr. .................... Outgoing radiation in May-Sovemher. ...................... Inradiated .......................................... 87,939 Incoming radiation in the year. ........................... 1OG.llS Outgoing radiation in the year.. ............................ 57.363 -- Inradiated .......................................... IS, 751; In reference to the total incoming ra.diation indicated 1 9 7 t.he black surface (150,029 according t.0 Table 2) only 32.5 per cent R ~S retained in the year. and G7.5 per cent was given out by radiation: in summer 35.6 (tr cent incomin radiat.ion besides 6 per cent more was given out. uted to the heating of the air an( the melting and evap- oration of the recipitation. Under t.he conditions of and its five-month snow cover, two-thirds of the incomin radiation has conserved radiation energy t,hrough bacf radiation, and only one-third is transformed into heat- another proof for tlic high importance of radiation for the high-a1 titude climate. It is true the nlbcdo values taken into account a.re based on photometric determinations, nnd it still remains t80 be proved, whether they also hold good for other parts of the spectrum, especially the infra-red ones. A4~~~rdinc* t.0 investigations in different parts of the spectrum ctirrie8 on a t Daws, the albedo of newly fallen snow rises slowly but syst.ernat.icallp from ultra-violet thou h blue, green, alhedo of meactow nnd stone would also increase toward the infra-red end of the s ectrum; that. would give t.he result still more point. i; owever, the difference of the albedo values is only an apparent one, produced by the heterooencity of the li ht sources acting simultaneously. The saar radiation. w % ich is rich in yellow and whose incidence is determined, is more intensively reflected from the surface of the ice crvstals than the blue-sky radiation whosc incidence is clifhsed. retained and (31.4 per cent given out; in winter t R e whole So hard F y a third of the incomi D mdiation had contrib- the highland va P leg with its thin atniospheric envelope 3 t.o yellow and even to red. It is not impro % able, that the PER3ldNEXT REGISTERING OF THE ULTRA-VIOLET SOLAR RADIATION, NOVEMBER, 1920, TO .JUNE, 1952. Permanciit rcgistration of ultra-violct radition offers to niet.eorologica1 optics the fairest prospect of con- tinuous data relat,ivt? to the degree of optical purity of tlie atmosphere; for its variations may best be noticed in the short-wwre spectmni. One may especially es- pect to obt,ain evidence on the variations of the amount of ozone in the high ntmos here--an element which surely is very important-an # on a iparallelism of the variations of the o tical transmissibi Ity wit,h t,he rotn- A parallactically mount.ed cadmium cell, supplied with a thin argon filling, mid made to follow t,he sun b glass l1 nt a distance of 112 mm. therefrom on which an image of the sun of 10 mm. diameter was projected by a quartz lens of 53 mm. diameter. The uviolglass disc having 30 mm. diameter, on1 3" sky zone round the current was conducted through a mirror galvanometer of 8.55 x lo-" a t 1-m. scale distance and was photographi- cally registered. Giintlicr & Tegetmeyer, Braunscliweig, have const-ructed the instrument w-itli their wonted precision, and it is similar to the well-knomn electric hotometer for visible light., made in the same worksho . cell and especially to protect it from too reat heat. exaggerated, for cadmium me1t.s in vacuum only a t 320' C. under a simultaneous vapor ressure of 0.001 mm., in tho former res ect the precaution mas absolutely In order to obtain su P ciently high readings an ausiliary potcntial of about 160 Folts was necessary. With it. t.he day curves always show a erfectly s.ymmetrica1 course, whose sitiveness of tlie cell, however, decreased continually during the registerin period of 1% year. The decrease of sensitiveness can %e stated by coinparison wit,h cnd- mium cell 11, which has not cliunged since 1915. The latter has been employed t,o make only single mensiire- ments, and its sensitiveness has been constantly cont.rolled by a controlling cell. During the first dags, while only esperiments were made ancl comparisons mcre unfortunately not yet carried on systematically, the decrease of sensit.ireiiess was very considerable (about two-thirds of the original sensitiveness). Later on, it,s fall was neitrly linear and continued to go slowly down, every month, so that now, after 20 months of permanent use it ymesses only about half of the original sensitiveness. he reverse of an uniform measure was operated, as has been t.olcl, by reduction to t,he re ularly collected parallel readings of cell 11. Whether tks decrease of sensit-ivcness is caused by the loading up of the glass in rosimit,y of t.he anode, of the anode, or by chawes of the surface, or perhsps o tlie gas filling, has not Keen inrestioated. A chanoe of poles, which might. have given inkrnia t,ion, has Been intentionally omitted, in order not to disturb the uni- formity of the series. Grounding of the positive pole, which was done three tiines a clay, in order to fis the zero line, did not operate to change t,he curve line of !lie zqro line, not even at noon after htilf x day's permn.nent irradis- tion. Neither was tlierc n (lecreasc of sensitiveness observed after a long period of fint> weat.her, when thc csposure could last for weeks, nor when after long interrii t,ion by a period of bad weat.her, reg.ist.ering c.o$d wain l e taken up. Itc&t.ering wns carried On ctur!ng an undisturbed dnys. TEe considerable m-at,erinl, whwh tion period and wit R single revolutions of the sun. means of a clockwork. was esposed under a mnt u\-io f - sun radinted in upon the ce .y 1. The resulting photo- $he uviolglass is only employed in order to spare t i e P In the latter respect the precaution may B ave been which, however, is below that of t rl e argon-fillcd cell; but necessary. For, a P though tlie uviolglass was used, the accuracy is absolute P y not to be doubted. The sen- lienomenon could be observed. 5 or by an invisible thin cadmium P ayer on the platin rin 520 MON"TA1- WEATHER REVIEW. OCTOBER, 1922 resulted, confirms anew the yearly run which has been fixed by single measurements since 1909, as follows: With the same solar-altitude there is a minimum in midsummer, and a maximum in midwinter; months of the second half year show greater intensity than the niont,hs of the first half year; especially is autumn intensity niuch greater than s rin intensity. The intensity rises from The transmission coefficient is 0.229 in the yearly mean in 1921, while it was 0.253 in 1916 and 0.278 in 1917. 'me increased obscrva.tiona1 material collected, and which has not been confined to the finest days, esplains, at least to some extent, the deviation toward a diminution of the transmission coefficient. The absolute numbers are somewhat lower than in the preceding ears, which, to a sniall extent, may be owing to the kss of radiation in the snidl zone surrounding the sun. which has just been mentioned. If we look at all the long series from 1915 to 1922, we have, in s ite the ultra-violet solar intensity (not taking into calcu- lation the yearly run) has, ge.nerally speaking. de- creased permanently from 1915 to 1982, from 1915 to 1917 ra idly, afterwards more slowly. Only the part of autumn 1920 and 1921 are an esception. The accuracy of t.his result may of course be doubted. There may.be objections, viz, that even the best- rotected used for single measurements only, may have diminished somewhat in its sensitiveness. However, it has from time to time been controlled by cell No. I, which until the autumn 1921 was used only for controlling purposes, and no change in the relation of sensitiveness could be ascertained. Even though we were inclined to correlate the result with the sunspot period, we should have to come to the conclusion, that just in the beginning of the increased activity (spot eruption) the emitted radia- tion was most considerable and not at the time of the masimuni size of the spots. I t will need a much greater amount of material, permanently registered and collected at different places of the earth, to substantiate. t.his result. An influence of the rotation period could not be estab- lished with certainty during this t,inie of slight solar activity. Cirro-stratus alone weakens even in its lightest, shape the ultra-violet radiation of the sun's disc 10 er cent,, while the total intensity is only decreased [alf as much. The ultreviolet nzaximuni values coin- cide in the same season with the clearest and bluest sky; they are found after snowfall and together with foehn descending into the valley. On days wit.h tendency to thunderstorms they do riot appear, as the zinc lid1 photonieter eiToneously indicates-most probably the error is due t.0 the hutiiidity of the air. Also the sudden and quick junipings, which this instrument records, do not exist. in reality and niay also be due t.o the cause just mentioned. The re ist,erinv curws run undisturbed on all cloudless days. #he ten2ency of the atmosphere to condensation shows itself wry soon and very charact,er- istically in the ultxa-violet solar radiation. When cumuli arise, radiation undergoes a sniall loss; ~1 greater one, however, when a1 to-cumuli arise. The telluric solar corona, whic.h the author measured, described and e.xplainec1 in former publicat.ions 12, lessens the ultra-violet solar radiation lit.tle, if my, and the register- ing curve runs perfectly undisturbed. The case is dif- 5 to 340, whi P i ? e t e sun is niounting from 10' to (55". of the numerous gaps, the convincing impression t K at values of t R e extraordinarily clear days in the latter standard cell of the observatory (No. 111, w K ich was - __ - -- . .. 1. A hlriiadlria~riol de* Prruwiwhm dfrtmrolngirdm Inrlilrls, Vnl. V, 11. .i. 1917. ferent with the obtrusive, dazzling white corona, which is the first si n of t.endency to condensation. Under and i t also lies somewhat below the normal level. Besides its great advantrtges, accuracy and possibility of registering the oalvanometric method prese1it.s also a considera\ile disa&wntage to measurements in ultra- violet. The instrument is not hransportable. The very slight photo-current, produced by the cadmium cell, needs an estreiiiely sensitive galvanometer, that must be most carefully protected from vibrations and all in- fluences from out.side, and whose foundation must be very firin. However, a special way of employing the cadmium cell is t.o he advocated here, which remedies this defect. K. Kiihler lS first inaugurated this niethod in Kolberg. Nine months ago it was adopted here with slight changes. This electrometric method may best be called the d~iisckmr 'ng method in op osition to t.he e7wgin.g method. F he cell, which stan B s in connection with a Wulf electromet.er, is charged up to a certain potential, wbse discharging velocity under the influence of ult.ra-violet radiation gives the measure of intensity. In order t.0 diminish the too rent intensity, suitable diaphmmns were used here. 4iey are easily switched on in tge well-known 'i Elect,ric photometer for visible light." An auxiliary capacity was not needed. There esist.ed at first strong miswivings as t.o whether the des- cribed application would %e possible, and whether the photo-current would always reniain proportional to the incident intensity, t.he cell used being neither evacuated nor in a s h t e ot saturation and t,he connection between the current and t.he potential a t constant illumination bein not given linear but by the characteristic. Innu- meru 5 le comparison measurements made with t,he regis- tered values dvanometricallp recorded, aiid also those with two cel?s used siniultaneously according to this discharging method, have proved that the method is reliable, if the following im ortant points are considered: (1) In the 1imit.s of the {ischarge t.he characteristic of the cell must haw a linear course. (8) In the limits of t.he discharge the characteristic of t8he electrometer must have a linear course. (3) The positive, not the negative, polo must be charged in connection with the electronicter (in order to avoid photo .effects on the electrometer). (4) A too nearly eshausted cell must not be used- on1 experiences with normally argon-filled cells are If these rules are obserred, this simple mode of meas- urement will prove nstonishingly efficient. Having only a bifilar electromct.er at! my disposal and being therefore unable to fix c uite exact,ly the crossing t h e of a certain following espedient: Within a large scale area, in which the charactmistics of the. cell and the gaugc curve of the pliot~ornetcr run liiieally, the clischarKe area was but approsimat.elg adhwed to (with the Daws instrument between 150 and 100 vo1t.s) and the readings were taken imniediabely before and after the exposure, while the fibers stopped. its influence t K ie curve is constantly slightly wavering, a t z and. scale division \ y means of a stop watch, I adopted the The ineasure was then t,he following: li in seconds Fa*,, and V beginning and end potent,ial with radiation rJ,,' and V I beginning and end potential without radiation for the controlling of insolation. -- Is Abhundlungcn dC.9 I'rc ussiwchcn .lfilrorolngirmctl~ch~n I n s ~i l a b . H.iud VII, Nr. 2, 19.3. OCTOBER, 1992. MONTHLY WEATHER REVIEIT‘. 52 1 I multiplied by 10,000 to get easy numbers. The inso- lation loss always kept below 1 per cent, except at quite low solar altihde, and was generally negligible. I always made three measurements, of which the mean was taken. At constant sun their variations keep within narrow limits: they mostly aniount to about 1 per cent of the mean value. Sudden jumpings like t,hose occurring with the zinc ball photometer were not seen. The time of discharge was regulated by different large diaphragms, and nearly always varied between the limits of 12 and 25 seconds. The method is especially fit to be used in fixing the- ultra violet local dearness by switching in several niat quartz plat,es. The diminution of radiation, which ensues, is fatal to the galvanometric method, but in this case it is generally desirable. Further, this inethod has been applied t.0 the comparison of the intensity of t,lie quartz lamp (so inuch used in medicine now) and t,hat of the sun within that ultra-violet spectrum section, which both kinds of radiat.ion have in common. Ten different burners partly new, artly old ones were examined. The exact. dat,a were pu B lished in “ Strahlent,herapie,” Band XIV, Heft 1 (Urban k Scliwarzenberg, Berlin). The result, which will interest most here, is: The Hanauer uartz lam supplied with a new burner (so called At 100 cm. distance.. 3.7 fold sun intensit.v. in relation to Daws At 70 cm. distance. - 6.1 fold sun intensity at the mean of sun’s :It 50 cm. distance.. 12.2 fold I altitude. Some simultaneous comparison measurements? made in Chur (557 m.), Davos (1.590 in.), Schatzalp (1,StjO ni.), furnished the result that wit,hin t,hese alt.itudes with a cloudless skv obserration and calculation from t.rans- mission coeficienb (t.he cliffcrent sea l e d s being taken into consideration) are in good agreement, escept during special weat,lier conditions. For instance, when there is a foehn revailing, decending far into the valley of the arise above the latt.er light strata, rccognized by the whitish l h e color of the sky, in the former. However, t.he air is then of the highest possible t.ransmissibi1ity- in this case the Chur values hare been considerably higher t,han those resulting from’ calculation. %un,stZ