The Total Radiation Received on a Horizontal Surface

Home , August 1914, Diffuse sky radiation, February 1911, February 1914

474 MONTHLY WEATHER REVIEW. AUGUST,1914

SECTION I.-AEROLOGY.

TEE TOTAL RADIATION RECEIVEI? ON A HORIZONTAL phra,om contained an opening 24 inches square, and the SURFACE FROX THE SUN AND SKY AT innx diaphragm an op3ning only slightly in excess of the XOUNT WEATHER, VA. arm of the grids constitutino the thermometers, or about 2t By HERBERTH. KIMBALL. inchs square. The angurar opening between two ar- del sides of the outer diaphragm as seen from the tier- pntd Mount Woathw, Vs., AM. 18,1914.p mometers was therzfore about go. This is slightly greater I. STANDARDIZATIONOF CALLENDAR RECORDINQ PYRHELIO- than the angle subtended b the circular opening in the METERS. (1) outer diaphragm of the drvin pyrheliometer as seen from the thermal elempnt. Description qf pyrh.eliometer.-The receiving part of this The glass cover of Callendar pyrheliometer No. 9860 pyrhpliometer 1s made up of four platinum grids wound has boen accidentlv broken and removed. Previous to m series in two pairs, each air constituting an elpctrical this, however, in February, 1911, both No. 98GO and No. resistance thermometer. &e grids are wound on a hori- 9861 were compared with Smithsonian silver disk pyrhe- zontal mica plate fixed in a seal d and exhausted glass liometer No. 1, and were found to read 10 per cent and bulb, the top of which is hemi3pherical. The air of grids 9 er cent low, respectively. conatitutin one of the thermometers i3 coate a with black &he two Callendar pyrheliometers thus shielded from enamel, an% the relative positions of the four grids are sky radiation were equatoriallv mounted on a common shown in figure 1. M'hm exposed to rsdhtion the black- support, and kept pointed to the sun by clockwork, with ened thermometcr is raised to a highqr tempwatura than their receiving surfaces normal to the direction of the the other, the sires of which are naked. Awn to incident solar rays. They were esposcd on a shelf in a Callendar the diff .renc.e in temperaturz of the two tf er- third story window of the Phvsical Laboratory (see A, mometers is ve nearly proportional to the intensity of fig. 31, beside a Marvin pyrhaliometer. This latter was the radiation fa 'ng upon thorn. An automatic rp standardized bv comparison with Smithsonian silver diak tion of this temperaturex dill rence i3 cff.cted by pPa- acing pvrheliometer No. 1, and comparative readin between the two thermometers in two arms of a self-adjustin,o the two are frequentlv made. Smithsonian%o. 1 has Wheatstone brid e, which has a pen so arranged that it also been recompared from time to time with pyrhelio- traces on a recor% sheet-wound on 8 revolving drum the meters in use at the Astrophvsical Observatory of the varying difference in resistance of the two thermometers Smithsonian Institution. By means of these precautions due to thrir dii€.wnce in tem mature. it is believed that a e are able to el press the readings of Comparison of ddaivin an% Callendar pyrheliometers, the Marvin pyrheliometer in heat units of the Smithmian when er osed to solar radiation. only, at normat incidence.- revised scale o pyrheliomety (2). Most of the compari- A ccrt' cate is furni3h.d with each Calladar pyrh.50- sons of the Callendar< pvrheliometers have boen made meter, xwhich giles a factor by means of which recorded with Marvin pyrheliometer No. 3, the comparisons of difhrences in r.-sistance of the two thcmometcrs may be which with the Smithsonian instrument have given very reduced to gram-calories of heat received by them each constan t ratios. minute per square centimeter of area. This factor neces- In obtaining comparisons between the Marvin and the sarily assumes that the recordt-d difference in resistance Callendar pyi helioiiioters the latter have usually been of the two thermometers is strictly proportional to the allowed to make a continuous record for a half-day period, intensity of the radiation to which they are esposed, the record for the a hole day bein interru ted at noon without re ard to the magnitude of this intensity, the in order to obtain thc zero line on t%a o recor sheet. The wave lengt% of the radiant energy, or the temperature of Marvin pyrheliometer has been read at fre uent intemals the resistance grids. throughout the day, eight readings in ei t consecutive Both direct solar radiation and diffuse sky radiation minutes usually constituting a series, anP these readin vary in intensity and in quality from day to day, and from have been compared with the Cullendar record ma de hour to hour of the same day. They also dXcr from each siniult aneously. other in both quality and intensity. It seemed desirable, Between August 15 and 30, 1912, 29 comparisons were therefore, to corn are records obtained from a Cttllendnr obtained betu een a Marvin pyrheliometer and Callendar yrheliometcr witg the readings of some other instrument No. 9860. The results are surmnarized in Table 1. Be- !,fore undertaking to ex ress in absolute units records tueen July 27, 1912, and March 19, 1913, about 130 obtained from the Callencf ar instrument when es osed to comparisons were obtained between a Marvin pyrhelio- both sun and sky radiation. For this purpose 2allendrtr nieter and Callendar No. 98G1, and the results are sum- pyrheliometers No. 9860 and No. 9S61 were fittrd with marized in Table 2. tubes about 3) inches in diameter and 20 inches long, The final results as tabulated were obtained by plotting blackened inside, and containing two diaphragms, one the individual observations, with air masses (a proxi- just above the case containing the thermometers, the mately the secant of the sun's zenith distance) as aI! sciasas other at the outer end of the tube. The distance between Callendar these diaphragms was about 15 inches. The outer die and the ratio Mar,.in as ordinates. In the cnse of No. 9860, without a glass cover, there is no apparent differ- 1 In hrt 11 ohrerntlonel date have bem InrliiM to apt.80 1914, the date on whlch ndLUon maaummmis were dbcontlnued aL MwtW ather, ba ence betm een the values of these ratios from morning and

Unauthenticated | Downloaded 10/01/21 02:25 PM UTC 1)‘ TO face p. *N. ‘ - M. W. R., August, 1914.

FIG.3.-Physical laboratory, Mount Weather, Va. A indicates the afternoon window exposure forjthe Marvin pyrheliometer; B, the exposure for the Callendar pyrheliometer.

%. %.

ll F

I_-

_+ -”

FIG.4.-Exposure of redistion instruments on roof of the physical laboratory, Mount Weather, Va.

Unauthenticated | Downloaded 10/01/21 02:25 PM UTC AUQU~T,1914. MONTHLY WEATHER REVIEW. 476 afternoon observations. In the case of No. 9861, with a made near midday and were generally both preceded and lass cover, the ratios from morning observatjons are followed by unshaded comparisons of the mstrumepts. ifecidedly higher than the ratios from afternoon obser- The results, reduced to bring the two instruments mto vations, for zenith distances of the sun less than 78.7'. accord when unshaded, are summarized in Table 3. From Table 1 it is seen that Callendar No. 9860 reads With the sun at zenith distance 78.7' (air mass 5) its slightly lower than the Marvin instrument with the sun radiation intensity at normal incidence is reduced by less than 50' from the zenith, and higher than the Mar- atmospheric absorption from its zenithal intensity b vin instrument with the sun more than 60' from the approxiniatel 50 per cent, the diminution be' mucK zenith. greater for sE ort waves than for long waves. Table 2 From Table 2 it is seen that Callendar No. 98Gl reads shows that this decreased mtensitv is accomDanied bv an higher than the Marvin instrument with the sun more 20 than 79' from the zenith, and with a hi her sun reads increase in the ratio, Marvin oflmore thi per ient, lower than the Martin, especially in the af ternoon. while Table 3 shows an increase of only 5 or 6 per cent. Since No. 9860 and No. 9861 read practically alike in This greater increase in Table 2 as compared with that February, 1911, before the glass case was removed from in Table 3 is probably due to the decrease in the pro- No. 9860, the results in Tables 1 and 2 indicate that the portion of short wave radiation in the solar s ectrum removal of its case increased the readings of No. 9860 by with increase in the sun's zenith distance. golished at least 10 per cent. platinum has a much higher coefficient of absorption for Compa&wn of Marvin and Callendar pyrhelionzeters radiation of short wave lengths than for radiation of when exposed to radincim of varying intenszty.--In order lo wave lengths (3). In consequence the bright grids to determine the effect of radiation intensity upon the of "%t e Callendar pyrheliometer absorb relatively more of indications of Callendar pyrheliometers, Marvin No. 3 the solar radiation at noon than when the sun is at a was mounted in the place of No. 98G0, beside No. 9861, greater zenith distance, thereby diminishing the value of and on the same support. The diameter of the dis the ratio, as shown. even more marked phraped tube of No. 98Gl was enlarged to 43 inches so as Marvin ' An to a mit of ventilation at the bottom, and was shortened decrease in this ratio is found in the second part of Table so that there was only about 12 inches between the two 4 11s compared with the first part, the comparisons being diaphragms, leaving the outer one only about 13 inches between total solar radiation and the radiation trans- from the receiving surface of the thermometers. The mitted by a ray filter, to be described later. o ening in the outer diaphragm was made 23 inches square, Dijerence in results of a. m. and p. m. compa&m.s.- w\ ich gave an angular operung of 12' between its parallel An attempt has been made to ascertain the cause of the sides, and 17' between diagonally opposite corners, as seen difference in the results of a. m. and p. m. comparisons from the surface of the receiving rids. A disk with seg- between Callendar pyrheliometer No. 9861 and a Marvin ments that could be opened 3ifFerent amounts was yrheliometer. These differences are not peculiar to mounted above the two pyrheliometers, with its driving go. 9861, as they were found when corn aring the Callen- shaft between them. By rotating this disk it was os- dar instrument in use at Washington, 8. C., with Smith- sible to shade both instruments the same amount an $ to sonian No. 1, and also when comparing the Callendar any desired extent. in use at Madison, Wis., with the Marvin, except that in this latter case the Cdlendar read relatively too high TABLEl.4umy of unnpmisonS of Calkndar PyrlrcliOmCter No. in the afternoon. IPatterson __ - (4) also obtained higher 9860 (wrrnou? g (UUI envelope) ~neloredin a diapluagmcd tube, with Mmnpyphclwmetcr No. 6. values of the ratio in the afternoon than in the - c*2 Bun'sztmithdistsnce.._.... I WO'I 48.BaI 60.0°1 BB.6'1 70.7'1 73.6'1 76.7' morning. In order to eliminate the effects of excessively heating the Callendar instrument, tho fiber tube used previous 1.10 to June, 1913, which fitted t.ightly upon the case of the p rheliometer, was replaced by the larger aluminum tube, Jeady referred to, leavin an annular space about Q inch wide between the tube anf the case, for purposes of ventilation. This had no effect upon the results, however. An awning was hung over the window in which the pyrheliometers were ex osed, so as to cut off any ossible reflection from the si Be of the building into t\ e diaphragmed tube; and, finally, the Cdlendar yrholiometer was exposed on a platform above the roof oP the building; but there was no diminution in the difference between morning and afternoon ratios. However, in the course of these investi ations the Callendar gyrheliometer with its diaphragme% tube was turned 90 about its vertical axis, and the higher values of these ratios were then obtained in the afternoon instead of in the morning, as heretofore. Figure 1 shows the orientation of the black and bright grids of the pyrheliometer with reference to the meridian, N-S. The 'First position" represents the orientation Between June 18 and Au st 20, 1913, 45 comparisons when the comparisons summarized in Table 2 were were obtained between t ese two instrumpts when obtained, and the ''Second position" repregents the shaded from 50 to 90 per Fcent. The compansons were orientation after the instrument had been turned 90'

Unauthenticated | Downloaded 10/01/21 02:25 PM UTC 178 lld0NTHLY ‘WEATRER REVIEW. ci- Amm, 1914

(~8doscribed above. In the htpart of Table 4 are red radiation intensities in shli ht and sunlight, re- summanzed 29 sets pf comparative readings between s ectively, as derived from the stufics of Nichols (S), and Callendar No. 9861 and Marvin No. 3, obtained between A)bbot and Fowle (6). The filter was desiolled to reduce July 4 and July 8, 1913, with the Callendar oriented as sunlight to approximately the blueness of skylivht, but shown in the “First position,” ure 1, and 34 sets cd its transmission coefficients for green and vellow E ht are comparative readings obtained etween June 9 and probably too high. Its transnlission coefficient Bor the July 16, 1913, with the CallendarTi oriented as shown in total solar radiation is about 1/5, while for diffuse sky the “Second position,” figure 1. radiation it is about 1/3. The difference between a. m. and p. m. values of the ratio appears to be

Wars Isnprh. WQ.2-Transm~bn eoe5clent.s of ray fflter. slightly greater for solar radiation that has been de leted by passing through this ray filter than for totsP solar radiation. Wiener (7), and Abbot and Fowle (8),have shown that the sky below the sun is warmer, and the sky above the sun is cooler, than the sky on either side the sun. The four ids of the Callendar yrheliorneter, when inclosed in a riaphragmed tube as aP ready described, are exposed to different portions of the sky with reference to the sun as a center. Thus, referring to figure 1, and recalling that the angular opening between dia onally opposite corners 1. TOTAL SOLIP XADIATIOX. of the outer diaphragm as view ecf from the grids is 17O, it I I will be seen that with the pyrheliometer oriented as in the 0 . 0 “First osition” and the tube accurately pointed to the 76.S 1.016 ...... 76.1 0.M fl ...... 78.4 0.830 sun, in t e morning the black grid a is above the &xis of the M.4 1.m 0.5 0.W 71.0 0.944 tube, and must receive most of its sky radiation from 87.0 0.058 m.0 0.m ee.8 0.806 W.2 0.964 m.4 0.W 68.0 0.881 oints below the sun. The black grid e, on the other 48.9 0.917 47.2 0.890 48.8 0.884 31.0 aow as am n.o am {and, is below the axis of the tube, and must receive most of its sky radiation from points above the sun, while the 2 8OLAB PADIATIOX TIWIIIYITTED TILTEL bright grids b and d, lying to the left and right of the axis THBOUOH EAI of the tube, will receive most of their sky radiation from I I paints to the right and left of the sun. In the afternoon these conditions are reversed, so that b and d receive most of their sky radiation from points below and above the sun, res ectively, and a and c from points to the left and In the second art of Table 4 are summarized 33 sets of right oP the sun. With the pyrheliometer oriented as comparisons witE No. 9861 in the “FiIst position” and shown in the “Second position,” figure 1, the above a. m. 14 sets with it in the “Second position,” obtained between and p. m. conditions relative to the esposure of the grids Februa 26 and April 17,1914 inclusive, with a ray filter to the sky are reversed. placed ’9a ternately m front of kwt the Callendar and then the Marvin pyrheliometer. The readings of the Marvin Since the value of the ratio c::zF depends upon the p rheliometer were plotted-air mass against lo arithms extent to which the air of black grids is heated above the oP instrumental measurements- to facilitate t f e inter- temperature of the E right ,orids, reference to Table 4 will olation of measurements synchronous with the record show that the pair of grids that lies more nearly in the iythe Callendar pyrheliometer sun’s rertical, and in consequence receives most of its sky The ray filter was made by thi Vnited States Bureau of radiation from oints above and below the sun, is the Standards, and consists of the following parts: more effectivelyR eated. How ever, the data andable do 1. A glass cell 21 mm. deep containing a solution of not indicate that we are ‘ustified in attributin this to the W.NO,),. escess of sky radiation 1rom these sections of the sky as 2. A gelatino film dyed by uranine. con ared with the radiation from sections to the right 3. A gelatine film dyed by alizarine red. and 7eft of the sun. 4. A gelatine film dyed by victoria blue. Comparison of measurements of the vertical component The transmission coefficients of this filter were specified of solar radiationby Marvin and Callendar pyrhe1iometms.- by the author, and are show n in figure 2. They re resent The attempt to standardize Callendar pyrheliometers by approximately the relation between the ratio of 1lue to mounting them in a diaphragmed tube proved unsatis-

Unauthenticated | Downloaded 10/01/21 02:25 PM UTC Amm, 1914. MONTHLY WEATHER REVIEW. factory. This method was therefore abandoned for one an artificial eoli se of the sun for short periods, by inter- that is about to he described. 03ing a mova le circular screen 4 inches in diameter Cdlendar receivers No. 9861 and No. 13129 were ex- getween the sunE and the Callendar receiving grids, and posed on the capstone of the central ventilating flue of about 38 inches from the latter, as shown at D in figure 4. the Physical Laborator (see B, figure 3) with their At the receiving grids this screen subtends an thermometers horkontaf Their horlzon was ractically about 10'. The +inch arm su porting the screenT s adesOf unobstructed in every direction. In figure 4, go. 13129, a strip of sky extending from tYl e sun to the horizon and is seen in the central foreground at A; No. 9861 was +O to 2O wide, the width increasing with approach to exposed just beyond it at B. The instrument shown in the horizon. The screen has an equatorial mounting, figure 4 at B is a thermopile for mensuring nocturnil but it is necessary to turn it by hand on its axis every radiation. Beyond this at 0,projecting through a black two or three minutes in order to keep the pyrheliometer cloth screen, is the receiving plate of a Sharpe-Millar in the center of the shadow. photometer. The cloth screen covers the top of the Figure 5 shows a record made with Callendar pyrhelb photometer shelter when the roof is slid back as shown in meter No. 9861 on May 8, 1913. A cap was put over the figure. The wind vane is pointing nearly due north. the receiver at 9:C9 a. m., screenin it from all radiation Receiver No. 0861 recorded on a CPllendnr self-adiust- and fixing the zero line on the s% eet, as shown.. The __ ing bridge, and No. 13,129 on a self-adjusting bridge other depressions in the record were caused by inter-

constructed by Leeds and Northrup. The bridge wires posing the screen above described between the receiver of both these recordem have a resistance of 1 ohm for 200 of the pyrheliometer and the sun. mm. or 7.874 inches of length. Careful calibration with That part of the record (fig. 5) between fhe line A A and 0.1 ohm ste s showed no variations in either wire greater the zero or base line represents the heating effect of sky than could ie attributed to imperfect registration. The radiation upon the pyrheliometer. That part of the Leeds and Northrup register is shunted so that a differ- record between the line A A and the trace B B, made by ence of 0.1 ohm in the thermometers causes a movement the register pen, represents the heating effect upon the of the pen of 1.016 inches on the record sheet. On the Cal- pyrheliometer of the vertical component of solar radia- lendar register 8 like difference in resistance causes A, tion. This latter may be compared with the vertical movement of 0.787 inch. The record sheets are ruled to componcnt of radiation intensities measured b the tenths of inches. The time scale is 1 hour = 0.8 inch, and Marvin pyrheliometer, H=Q sin h., where h is the Eeight lines are ruled for 10 minute intervals. of the sun above the horizon. Tho values of h have been Exposed as above described, the pyrheliometers give corn uted for each observation from Ball's Altitude a continuous record of the difference in resistance of the TnbP es (9). In figure 5 the values of R am.plotted aa black and bright grids due to the heating effect of the circles in the curve C Cfrom A A as a base he. radiation received from the sun and sky. In order to Table 5 summarizes comparisons between Marvin pyr- compare this record with the readings of the Marvin heliometer No. 3 and Callcndar pyrheliometer No. 9861, p heliometer it is necessary to deduct from the total made as described above between March 28 and May 8, tr e effect due to sky radiation or, more strictly s eaking, 1913. The vertical component of the radiation intensity the effect of solar radiation reflected diffusely Prom the measured by the Marvin instrument has been divided by sky. This latter has been determined by producing the number of tenths of inches on the record htbe-

Unauthenticated | Downloaded 10/01/21 02:25 PM UTC 478 MONTHLY WEATmR REVIEW. AuGueT, 1914 tween a line similar to A A, figure 5, and the trace made evening, as the sun approached the horizon, too great a with the readings of difference in resistance was recorded. This niakes the F, is the radia- record B B quite unsymmetrical about the noon line. The line A A, which represents the heating cff sct of sky radiation, does not show this lack of syrunietry. The vertical components of Marvin pyrhelioiiieter reitdings have 5 have been obtained b lotting the niean values of F, been plotted as circlos in the curve C C with A A as a base. given in the &st part oFE t e table, against the air masses In searching for the cause of this anomaly in the Cctllen- corresponding to the sun's zenith distance. The ratios, dar record the p rhelionieter was turned on its support about its vertica r axis approsirnately 120'. The record ':&?, in the last part of the table, have been obtamed made in the early morning then showed too grent a diff 2r- b dividing the value of the reduction factor derived from ence in resistance between the black and bright grids, tg e Callendar certificate (0.0251 for pyrheliometer No. while the record made a short time before sunset shomcd 9861), by these smoothed values of F. It will be observed almost no difference. that these ratios have a higher value than the ratios in TABLE6.-Summaiy of compakons of CalkndaT pyrheliometer No. Table 2, and that the increase with increased zenith dis- 15129, exposed horizontally, With the vertical component of Marvin tance of the sun is also greater. Table 3 would lead us pyrhdiometer readings. to anticipate this result, as the vertical corn onent of solar radiation intensities is not onl less than t B e intensity at normal incidence, but it also c9ecreases more rapidly with Number of compnri- increased zenith distance of the sun. song ....._.._.____.. TARLE&--summary of comparisons of Calkndur pyrheliometer 1vo. Sun's zenith dlstance. 9861, exposed horizontally, with the vertical component of Hawin p!yrlielionietm readinp. - - _- I MWOf A. Y. WLUparisODS. I Man of P. M. comparisons. Number of compari- Number of compari- song ______..._...... ~17i12~14~9~8~7~6~8~4~~ sons .._.__._____._..16 21 34 24 11 8 4 4 I I 3 Sun's zenith distance. 30.5" 48.3' 58.5" 66.8' 70.6' 73.4' 75.7" 77.4' 78.7' 80.0' Jun's?mIthdlstance.. 30.0' 47.2' 58.6' 66.2' 71.2' 73.8' F-Marvh 0.0288 0.0265 0.0253 0 0253 0 0258 0.0256 0.0247 0.0256 0. ~~ 0.0237 Marvin (calories) 0.0375 0.0284 0.0264 0.0255 0.0258 Callendar (scale)' F- cilI-li) 0.0283 I' I' - .. _____ I Smoothed comparisons.

Sun's renlth distance...... ~25.00~48.30~60.00~66.50~70.70~73.60~75.70~77.4'~78.7O179.8&0.7'

F.. .. . ~-...... -...... ~~~~~~~~~~~ Ratlo, 0.911 .925 .936 ,950 ,061 .972 .984j ,996 1.01 1. a 1.05 Marvln F Sun's dthdistance.. 30.5" 45.L). 59.2' 66.5" 70.6' 73.6' 75.9' 77.6' 78.5' 80.0" It is believed that thrse tinonialies in thc record iire clue p-Marvin (calories) 0.0272 0.0255 0.0244 0.0236 0.0229 0.(rJ310.03210.0208 0.0190 0.021 1 to imperfections in the glass cover of the instrument, Callendar (scale) which is not perfectly hcmisphericnl and shows ine uali- 1ii -. ties in the character md thickness of the glass. heae defects might cause a concentration of the solar ra s on one of tho gids, especially when the sui is low, anztliiis Sun's zenith distance. destroy the symmetry of the recon! 8.5 noted above. Coin arison of measurements ?f short-waw rtidiation b!y Calleni ar and Marwin pyrheliometers.-A.; has already heen shown, sk light is much richer in blue light than is solar ra iation (5, 6). Hence, it is not to be expected that thex factors, F,of Tables 5 and ti, cnn be om loyetl to reduce the Callendnr record of 4iy ruciin tioii In Table 6 are summarized similar compwisons be- to Pieat miits. We may, however, make iiw of the ratio.; tween Callendar pyrheliometer No. 13129 and Marvin in the second part of Tahle 4 to obtain such :I f:ictor. pro- No. 3. There is so little diff 2rence between the a. m. and vided the ray filter has reducrtl sunlight to tlic blueness p. m. values of F that they huve been coiiibined to obtain of sliyliqht, as has been as.;uriic.(l. Vnfortmintel thr difference between the a. 111. niid the p. 111. ratios, w$:ead-y the smoothed values and the ratio, '9,the numerator discussed, introduces some uiicertninty into the results, of which is derived froin the Cttllend:ir certificate for this but the value of the fiictor nppears to lie lwtwccn 0.039 pyrheliometer. and 0.031, nncl 0.030 has bceii adopted. These ratios are only slightly higher than the a. m. Computation of the total radiation received oil n horizontal ratios for C'illendar No. 9561. given in Table 2. surfiice.--In order to obtnui factors for reclucing records In the summary in Table 6' cknpttrisons for zenith dis- of the total radintion to heat units, iiieasureiiiciits of sky tances of the sun greater than 66.5' obtained between riidin tion ni:ide by Cnllendor pyrheliometer Xo. 13129 on October 16, 1913, and February 2, 1914, have been -kO different days, fairly well cktributed throughout) the omitted, as the Ctillendar instrument did not seem to be period February 13, 1913-Februnr 24, 1914, have been roper1 recording direct solar radiation. The record for grouped according to the zenith istance of the sun at kovemser 6, 1913, reproduced in figure 6, shows the the time of the measurements. Thex results are summa- nature of the defect. After sunrise there was a consid- rized in Table 7. If the measurements nrc divided into erable period during which no difbrence in resistance two groups, the fht covering the period February, 1913- between the black and bright grids was recorded. In the August, 1913, and the second the period September, 1913- Unauthenticated | Downloaded 10/01/21 02:25 PM UTC AUGUST, 1914. MONTHLY WEATHER REVTEW. 479 February, 1914, the earlier period shows considerably In the last part of Table 7 are given tha roportions of more sk radiation than does the latter. This is in ac- average solar and sky rndiation to the totnf with the sun cord witt other observations, which show that there was at different zenith distances. By means of these propor- a gradual diminution in the haziness of the atmosphere tions, in connection with the factor 0.030 for sky radia- throughout the year 1913 (10). tion and the fuctor.3 in Table G for solar radiation, have

Fro. B.-Records of solar and sky radiation at Mount Weather, Va., on November 6, 1913. l'AnLE 7.-Solur urid crky riidintion. as recorded by Callendar pyrhelio- been obtained the first line of factors in Table 8 for re- meto- hTo.13199. ducing records of the total radiation with a cloudless sky to grmi cnlories per minute per square centimeter 1 Mean of A. M. measurementa. of horizontnl rurface. These apply to records ob- I t:iitiecl with C&ntlnr No. 131'29, which has been used Number of measurementa ...... 14 I 28 I 38 1 25 i 22 I 20 1 14 I 16 I 8 i 7 in obtaining daily records at Mount Weather since May, 1 1912. S ,ectro-photometric measurements by Ives (11) and Nic flols (12) show considerable variations in the character of the radiation from nn overcast sky. Compari- Mean of P. M. measurements. sons between plio tonietric and pyrheliometric mensure- ments made at Mount Weather during the past year Number of mure show a marked siniilarity between radiation from an overcast sky and direct solar radiation at midday, and Sun'szenlthdistance..l 30.5" 48.7' 58.5' 68.8" 70.8' 73.4' 75.6' 77.5' 78.V 80.1' marked differences between the radiation from a clear akyrac'lation....._... I 8.51 6.21 5.21 4.61 4.21 3.61 3.21 3.11 2.81 2.6 ancl an overcast sky, and between direct solar radiation with hiuh and low sun. In obtaining factors for reducing the Cdendar records of solar ancl sky radiation to heat - units, when the sky is partly overcast, the factor 0.030 Sun's zenith distance... ._.25.0' 60.0°166.50 io. 7' 3.6' i5.7'177.4O Sky radiation, Feb.-Aug., has been given a weight proportional to the amount of 1913.__._.._.__._._...._. 8.6 6.1 5.3 4.7 4.2 3.8 3.6 Sk radiation,Sept.,1913- blue sky prevniling at the time. For reducing the radia- &b. 1914 ...... ____...._._____5.2 4.4 3.9 3.5 3.2 3.0 2.8 2.6 2.4 2.3 tion from an overcast sky to heat units the factors for Sk rahiation, Feh., 191% Peb.1914.__..___.._..._. 8.: 6.3 5.0 4.3 4.0 3.7 3.4 3.1 2.9 2.7 ?.6 reducing midday solar rnclintion have been employed, Bolar&diation..__.._._._. 39.2 29.5 20.1 14.6 10.5 8.0 6.4 5.5 4.9 4.G 1 5 with a (liiiiinution of about G per cent as the sun ap- Totalradiation___....._.__ 4i.9 35.8 25.1 18.9 14.5 11.5 9.8 8.6 i.8 73'7:l proaches the horizon, corresponding to the variations m Ratio, & __...... _. 0.83 0.82 0.80 0.77 0.72 0.68 0.65 0.64 0.631 O:G3 0.63 the rntios of Table 3 for n decrease in radiation intcnsities Ratio,- __.__...... 0.18 0.18 0.20 0.23 0.281 0.32 0.351 0.36 0.371 0.31 0.37 Total of 50 per cent. These factors are given in the last line of Table 8.

Unauthenticated | Downloaded 10/01/21 02:25 PM UTC MONTHLY WEATHER REVIEW. Ao~ae~,1914 solar s ectrum, and also in the proportion of difEuse clear- sky ra $iation to solar radiation, have not been taken into account. It is believed, however,. that these variations tend to compensate one another in their effects on the value of F. A very dear SQ and a resulting increase in the values Sm'a zenith dlatsnoe. of F, is accom amed by a decrease in the quantity of sky radiation, wit l! a resultin decrease in the augmentation of F, from this source. dn the other hand, a hazy sky and a resulting decrease in the value of F,, is accompanied 0...... 10 ...... by an increase in the quantity of sky radiation, with a 20 ...... resulting increase in the au mentation of F, from this 30 ...... 40 ...... source. The records repro %uced in this REVIEWfor W ...... March, 1914, 42:140, figure 1, illustrate this point. On '10 ...... January 23, 1914, the sky was filbd with a peculiar white 80 ...... 90 ...... haze that greatly decreased the intensit of dirpct solar 100 ...... radiation, and at the same time increase the difhse slg radiation. On January 29, 1914, there wasB a dense haze While minor individual corrections are re uired for at the surface, but the sky above was a deep blue and its each Callendar pyrheliometer, the studies pf qN0. 9861 high degree of polarization indicated the absence of haze. and No. 13129 Indicate that eneral equations may he The intensity of direct solar radiation was nearly nor developed for the reduction o9 Callendar pyrheliometer mal, and the Muse sky radiation was less than the aver records to heat untits, as follows: age. With the sun at zenith distance 60.0' the factor F,, Let R = the proportion of total radiation that is received determined from direct comparisons with the Marvin dlffusely from a clear sky (see Table 7) ; pyrheliometer, was 0.0236 on the morning of the 23d F, =factor for reducing direct solar radiation to heat and 0.0252 on the morning of the 29th, and the propor- units, tions of recorded solar and sky radiation to the total = Callendar factor divided by the ratios at the end were 0.64 and 0.36, respectively, on the 23d, and 0.875 of Table 6; and 0.125, respective1 on the 29th. From these we F,=factor for reducing clear sky radiation to heat obtain for the values oP k',, 0.0259 on the 23d and 0.0257 Units, on the 29th, or 4 er cent and 5 p& cent lower, respec- =Cdendar factor divided by 0.81 ; tively, than the v Bpues derived from Table 8. F,=factor for reducing cloudy slq radiation to heat In -re 7 are reproduced the Callendar remds for May units, 20,1914,a very hazy or smo day, and June 30,1914,a =F, for solar zenith distance 26', diminishing very clear day except that li?t e sky was from 0.1 to 0.2 slightly with increased zenith distance of the covered with cumulus clouds during the afternoon. Just sun; before noon on May 20, with the sun at zenith distance 8-the proportion of the sky that is unobscured by 21.2' and its hour angle from the meridian 42 minutes, clouds. the factor F,, determined from direct comparisons with With the sun shining in a cloudless sky the reductioii the Marvin p rheliometer was 0.0253. The proportions factor for the total radiation becomes of solar and sz y radiation to the total were 0.76 and 0.24, F, (1 .O - R)Fs R Fl. (1) respectively, and the resultin value of F, is 0.0264, + whmh is 4 per cent lower than t% e value given in Table 8. When clouds are resent, the factor for reducing the Just before noon on June 30, with the sun at zenith dis- total record obtainea while the sun is shining beconies tance 18.9', and its hour angle from the meridian 48 min- F', = (1.0 - RS)F', + RS Fl. (2) utes, the factor F,, deterrmned from direct comparisons with the Marvin p rheliometer, was 0.0274. The pro- If we wish to reduce the total record obtained during portions of solar anK sky radiation to the total were 0.90 an hour with sunshine only a part of the time, we must and 0.10, res ectively, and the resulting value of F, weight F', and F, according to the roportional part of is 0.0277, whici is in accord with the value given in Table the total radiation to which each appp1 'es, as follows: 8. With the sun at zenith distance 60°, the values of F,, computed as above, are 0.0273 and 0.0271 for the F', F'$3I+ Fo(l.O-&Ic - SI+ (3) forenoon and afternoon of May 20, and 0.0270 and 0.279 (l.O-S)I, for the forenoon and afternoon of June 30, respectively. where Z and I, are the relative intensities of the total These are d closely in accord with the value 0.0271, radiation when the sun is shining and when it is obscured, derived from Table 8, with the exception of the one for res ectively. the afternoon of June 30, when the presence of clouds is & Table 8 are given the values of F, (percentage of shown by the Callendar record to have been a disturbing cloudiness= 0, and 9 - 1 .O) ; F, (percentage of cloudi- factor. ness = 100, and S = 0) ; and F', (percentage of cloudiness On Ma 8, 1913 (see record, fig. 5), the values of Fo, from 10 to 90, and S from 0.9 to 0.1). The values computezas above for Callendar No. 9E61, ranged be- adopted for Z and I, are 3 and 1,2 and 1 , 3 and 2, and 4 tween 0.0270 and 0.0281, for solar zenith distances less and 3, for solar zenith distances 25O to 63O, 64O to 74O, than 60.0'. 75O to 79', and 80° to 81°, respectively. For zenith dis- The probable error of F, for a single observation aver tances of the sun greater than 81' the factors in the last ages f0.0008, or f 3 per cent, for zenith distances of the . column of Table 8 have been employed. sun not greater than 75.7'. With lower sun the probable In the above equations, and also in Table 8 variations error increases rapidly, amountin to about f 6 per cent from day to day in the proportion of blue kght in the for solar zenith distance 78.7'. diththe sun 80 low, how-

Unauthenticated | Downloaded 10/01/21 02:25 PM UTC Au~m;1914. MONTHLY WEA.THER REVIEW. 481 ever the radiation measured is but a small part of the hour lines, have been integrated by counting in each of tod for the day. these areas the number of squares boundzd b the tenth- The probable errom of F,, F’,, and F’, may exceed inch lines and the 10-minute time lines. In c{ ear weather slightly that for F,. Also, the values of F, and P, durin the months October to February, inclusive, the given m Table 8 ap ear to be somewhat too high in hazy recopf s have been corrected by plotting on the record weather and somew!i at too low when the sky is free from sheet the vertical com onent of the Marvin p heliometer haze. readings, as explainel on page 477, and astrated in

II. MEABWEMENTB OF THE TOTAL RADIATION RECEIVED figure 6. No corrections have been made during these ON A HOBIZONTAL SURFACE. months when the sky was overcast with clouds. 2. The proportion of the sky covered with clouds dur- Callendtzr records reduced to heat unik-In Table 9 ing each hour has been determined partly from an inspec- are summarized ths dscads averages of ths total radi- tion of the Callendar records, and partl from the eye esti- ation record3d at Mount Weathsr, Va., since May 17, mate of cloudiness recorded for each oJ d hour throughout 1912, by Callendar pyrhsliomster No. 13129. The ths da in th3 “Daily Meteorological Record” for the sta- records have been reduced to heat units by the following tion. %he latter has been closely followed except when the method: clouds were some form of cirrus, in which case the propor- 1. The meas on the record shset bstween the base line, tion of the skg actually obscured by clouds may be materi- the tram made by the register pen, and the successive ally less than the proportion occupied by the cloud-sheet.

Unauthenticated | Downloaded 10/01/21 02:25 PM UTC 4sa MONTHLY WEAT3ER REVIEW . AUGU~T.1914

TABLE0.40ka and sky radiatbn. czprmaed in gram-cabrirS per sqwre CcntinrStcZ of hozisontul swfa.d Mount W&. Vi . [Ut. 39' 4' N .. long .77.53' W .. altitude. MO meters.) . .. . . Decade average during hours'mding (apparent the)- t Sa s Decade. A . M . P .M . :i ...... __ .. . . i! A 22 5 6 I 8 9 10 11 1 2 3 4 5 6 7 8 g 2 Em ...... P 1912. h .m C7r.S Or.+ Ur.i Gr.1 Gr.- Gr.i Gr.i Or., 05.- 0r.i Ur.1 f3r.T Gr.- Or.+ Gr.-c Or.+ GI.- % aCep 17-27 ...... +19.7 708 .... 4.0 17.5 30.7 54.I 61.0 62.6 68.E 73. I 6s .1 47 .e 37 .I 29 .I 13.1 2.7 0.2 571 77 n 3 21-31 ...... 21.2 7 13 0.2 4.2 16.7 31. m 43.I 52.8 64.2 74 .E 72.4 68.0 63 .a 42.E 25 .I 16.6 4.5 0.4 586 75 ea b June 1-10 ...... 22.6 7 19 0. 2 4.9 16.3 32. a 49 .I 61.9 m. 3 89 .E 76. t 67.7 61.4 46 .I 31.t 17.3 4.0 0.4 80( 79 Ba 5 11-a0 ...... 23. 3 722 0.2 3.9 11.7 24.5 29.4 41). 9 49.1 47 .I 47 .i 45.7 40.5 36 .L 23 . 10.3 2.9 0.3 41: 54 34 8 21-39 ...... 23.3 722 0.3 5.1 14.4 28.7 39 .i 45.9 53.1 56.5 49.1 45.6 35.2 28.3 m .I 10.7 3.4 0.3 431 58 35 8 July 1-10 ...... 22.7 7 1s 0.5 8.2 a. 6 33.3 44 .I 80.4 66.1 70.5 69 . 63.0 51.5 44. I 32. ( 19.2 4.9 0.5 58i 79 a 4 11-20 ...... 21.4 7 14 0.2 5.3 16.1 31.9 45. L 54.0 80 .5 62.6 58. , 47.3 40.2 28 .7 18. i 11.3 4.7 0.5 iK)i 67 52 7 21-31 ...... 19. 4 708 0.1 2. 0 11.4 19.3 31 .I 4s .0 53.1 59.9 67.I 60.6 4s .0 38 .I 27.1 12.4 3.8 0.2 48! 88 47 5 dug . 1-10 ...... 16.8 658 .... 2.1 9.4 2a. m 33 . 46.3 55.9 59.7 55 .I 47 .6 45.3 36 .B 21 .t 12.7 3.0 0.1 45! 68 39 6 11-20 ...... 13.9 648 .... 1.9 9.1 19.a 34 .E. 48.0 51.4 51.1 48. i 45.4 43.2 31.2 23 .t 9.0 1.6 .... 4li e5 33 7 n-31 ...... 10.4 631 .... 0.1 7.7 18. m 28 .% 44.1 49 .a 57.E 6.5. I 57.1 47. 0 34.4 au .( 9.6 1.3 .... 444 73 38 4 sep t. 1-10 ...... 6.6 6!U ..... 0.4 6.0 16. i 27 .E 36.3 46.2 56 .I 60.4 58.1 44.7 34.3 20.4 7.5 0. 6 ..... 41: 73 58 4 11-20 ...... + 2.8 608 ..... 0.1 3.9 15.7 21 .E 31.8 46.2 54.1 53.E 52.0 40.3 28 .B I6. E 5.4 o. a ..... 3a 88 u 5 2153...... - 1.1 556 ...... 1.1 0. 0 12. 9 zu . 1 a6.2 324 35.3 34.8 29 .n 21.5 11.4 3 .0 ...... m 48 3d 7 oct . 1-10 ...... 5.0 544 ...... 2.7 13.9 29 .1 42.8 52.3 5s. 9 80 .4 53.6 44.4 30.9 14 .S 3.9 ...... m 91 81 2 11-10 ...... 8.7 5 32 ...... 1.3 10.1 22.0 33.0 44.0 45 .a 44 .E 42.1 33.7 m .z s. c 1.3 ...... 3x 75 50 5 21-31 ...... 12. 5 5 19 ...... 0.5 6.0 18.3 30.4 37.9 37 .g 35 .I 33.1 27.5 17 .3 5. i 0.5 ...... 251 67 60 5 NOV. 1-10 ...... 15.9 5 07 ...... 0.1 4.5 14.1 22.6 29.8 31.4 31.1 29.6 24.6 14.7 5.E 0.4 ...... m 62 55 6 1140...... 18. 6 455 ....I ...... 2.7 10 .I 22.5 31.9 36.3 33.E 29.9 25.5 15.7 4.E 0.1 ...... 213 m 61 6 21-30 ...... ao .8 448 ....I ...... 2.1 10.4 21.8 32.0 39 .a 36.E 38.7 22.9 13.1 3 .E ...... 211 75 67 4 DUX 1-10 ...... 22.4 4 42 ...... 1.4 7.7 14.5 21.5 24.7 24 .I 21.2 14.7 a. 5 1.E ...... l& 53 30 6 11-20 ...... 23.3 4 38 ...... I .... 1.4 8. m 17.5 27.7 29.2 27 .€ 25.6 18.8 9.3 2.I ...... 1Q 88 57 6 21-31 ...... 23.4 438 .... I ...... 1.1 8.3 17.2 27.5 32. il 32.5 27 .4 19.3 9.4 l.€ ...... 17t 70 Bo 5 1913. Jan. 1-10 ...... P. 6 4 41 ...... 1.5 7.8 16.3 2a. 0 25.7 27.3 2s .6 18. 7 10.3 2.5 ...... 158 Bo 40 6 11-ao ...... 21.0 4 47 ...... 1.2 8 .r) 10.0 22.5 29.2 30.4 25.2 20 .2 11.0 2.7 ...... 168 57 33 7 21-31 ...... 18.7 458 ...... 1.9 9.2 17.3 25.3 31.7 31.5 27.1 ao .o 12.3 3.1 ...... 178 63 47 6 Beb . 1-10 ...... 15.8 5 07 ...... 0.3 5.3 16.3 28.7 38.6 42.3 45.4 42.4 33.4 19.5 7.1 0.4 ...... 280 73 75 4 114...... 12.5 5 18 ...... 0.4 6.7 18.a 29.3 38.9 43.4 43.3 39.9 27.0 17.6 6.5 u. 6 ...... 272 63 61 5 21- a...... 9.3 5 30 ...... 0.5 4.4 14.4 26.5 31.7 39.3 36.2 28.6 30.8 10.8 6. B 1.0 ...... 241 50 32 7 Mar. 1-10 ...... 5.9 5 41 ...... 1.3 9.7 21.1 35.5 43.6 43.4 47 .a 44.0 32.9 18 126 2.0 ...... 320 Bo 51 6 11-20 ...... - 2.0 6 54 ...... 1.7 9.5 ia 7 29.7 35.1 39.2 48a 44.4 36.7 25.8 13.5 4.1 ...... 308 53 48 6 21-31 ...... + 22 607 ..... ai 3.4 125 23.0 33.2 40.5 44.9 53.0 44.9 37.2 28.3 16.4 4.6 0.1 ..... 342 55 44 6 Ap. 1-10 ...... 6.2 620 ..... a4 6.8 22.3 37.1 62.8 61.9 61.4 5.5.0 4a 7 39.6 33.8 22.6 7.8 0.6 ..... 451 68 62 5 11-20 ...... 9.9 633 ..... 0.7 5.5 14.3 24 .a 2LL6 33.3 39.2 4a a 38.6 34.2 n .6 17.9 7.2 0.8 ..... 312 45 38 6 2l-30 ...... 13.3 6 44 ..... 29 12.4 27.3 42.7 54.4 6R8 58.0 67. Q 57.4 60.1 38.1 25.8 10.5 1.3 ..... 495 69 do 4 xay 1-10 ...... 16.4 655 ..... 3.7 15.5 31.5 44.2 60.7 69.6 72.8 73. a 64.2 57.4 44.4 n. a 1a 1 1.9 ..... 575 78 60 4 11-20 ...... ia 9 704 ..... 3.3 120 223 30.9 39.6 52.6 56.9 56.0 53.1 49.2 35.6 24.9 11.0 27 ..... 451 do (0 6 21-31 ...... 21.2 7 13 ..... 3.8 11.3 23.5 31.5 40.4 45.3 48. 2 50.5 45.3 34.8 30.3 19.7 9.3 22 ai 390 62 34 7 Iune i-10 ...... 22.6 7 19 a4 7.2 ia e 3L 3 44.2 59.4 70.2 75.8 73.6 89.6 58.8 a o 31.0 17.3 4.8 0.6 611 . 80 70 3 11-20 ...... 23. 3 722 as 7.7 19.2 38.6 51.6 65.2 72.5 8a1 74.7 69.6 Bo .7 48.1 30.5 16.5 4.0 a3 63E 84 89 3 21-30 ...... 23.4 7P a3 4.5 11.9 19.2 25.4 34.0 38.7 45.1 51 .8 44.4 37.2 B1.4 21.4 11.3 3.3 a3 883 51 37 7 July 1-10 ...... 22.7 720 a4 6. 2 ia 7 33.4 45.1 59.8 65.8 65.6 61.0 58.5 588 44.3 281 13.0 3.2 0.4 583 7% 64 4 11-20 ...... 21.4 7 15 0.3 4.4 13.1 26.5 38.3 46.2 58.8 60.4 58.4 53.9 48 6 35.2 23.4 12.6 4.8 0.5 484 67 m 6 9-31...... 19.4 707 0.1 6.2 16.6 31.2 45.8 58.8 64 .6 65.0 w. 5 53.9 40.8 243.5 22. Q 11. 2 3.1 0.2 w(I 72 64 6 Aw . 1-10 ...... 16.9 6 57 ..... 3.4 1a 3 23.5 31.4 44.2 53.0 59.2 54.1 47.7 43 .2 30 .3 21.8 9.7 1.7 ..... 431 64 48 11-20 ...... 13. 9 6 46 ..... 1.2 6.6 15.3 25.1 33.4 40.5 45.7 49.4 45. 6 36.4 29.3 ia 5 7.4 1.1 ..... 350 56 40 5 21-31 ...... 10.4 6 34 ..... 1.5 10.8 23.8 39.6 54.2 58.1 65.0 89.3 57.1 45.3 38.0 23.0 8.2 0.9 ..... 450 74 67 5 Sept. 1-10 ...... 6.7 6 22 ..... a4 5.8 17.5 31.4 44.8 51.0 55.6 60.1 53. 6 15.2 34.0 2a4 7. 2 0.7 ..... 421 75 68 4 11-20 ...... + 29 609 ..... 0.2 4.1 14.0 26.6 35.5 40.6 42.7 420 3a 8 31.5 22.4 11.4 3.3 a2 ..... 314 69 34 8 21-30 ...... - 1.0 558 ...... 2.3 12. 4 28.2 41.0 51.3 52.8 50 .a 45.9 38.8 25.6 13.5 3.3 0.1 ..... aea 74 58 6 oat. 1-10 ...... 4.9 5 44 ...... 1.2 7.5 17.1 24.6 29.5 34.4 37.0 29.9 n .5 a0. 9 9.2 1.8 ...... 211 64 36 7 11-20 ...... a6 5 32 ...... 0.8 7.3 17.8 m. 5 31.8 34.9 38.2 36. 4 26.0 16.5 7.3 1.1 ...... 245 60 40 6 21-31...... 124 5 19 ...... a4 6.5 15.2 27.3 35.1 37.1 38.0 3li 4 26 .1 ia 2 7.6 0.7 ...... 241 e5 46 6 Nov . 1-10 ...... 15.8 507 ...... 0.2 4.7 15.9 27.1 35.1 39.0 386 31.7 27.5 17.8 5.8 0.3 ...... 247 73 89 5 11-20 ...... la 6 4 55 ...... I 2.6 8.8 lb. 0 26.2 32.8 32.5 287 !a. 1 12.6 3.9 ...., ...... 186 61 44 8 21-30 ...... 20.8 4 48 ...... I 28 9.9 18.2 22.9 23.7 22.9 a5 14.9 as 2.3 ...., ...... 147 5a 42 7 Dec . 1-10 ...... 22.4 4 42 ...... , 2.3 1u 2 19.6 24.0 25.4 24.8 21.1 17.2 9.6 1.9 ...... 150 a 28 8 11-20 ...... 23.3 438 ...... I 25 11.8 23.6 34.8 33.9 3s. 3 31.5 21.8 11.0 3.a ...... 218 85 811 3 21-31 ...... 23.4 438 ...... I 1.8 9.6 18.5 u. 0 m .7 24.9 21.0 13.5 7.0 1.5 ...... 14Q 69 53 7 1914. Ja 1-10 ...... -22 6 4 41 ...... 1.0 8.4 15.0 m .e 24.6 22 1 17.1 11.1 6. 1 1.5 ...... 128 48 33 8 11-10 ...... 21.1 4 47 ...... 23 11.0 la 1 29.8 31.4 320 3R6 18.9 ia 9 28 ...... 189 e5 50 6 21-31 ...... 18. 8 456 ...... ai 3.3 11.6 21.8 29.0 34.4 33.3 32. 2 24.3 12.2 3.6 a2 ...... 2(w 61 55 Feb 1-10 15.8 5 07 a4 6.4 ...... 16.3 27.8 36.5 41.2 40.9 35.8 27.6 16.0 5.5 a4 I ...... 255 66 61 ;I 11-20 ...... 12. 6 ...... a7 7. 6 1e. 9 24 .6 35.4 39.3 36.8 37.2 SO 17.3 6.1 0.6 ...... 248 57 50 7 2l-as ...... 9.4 2 ...... 2. 0 128 24.7 40.3 52.2 56.6 57 .6 45 0 36.2 20.2 IO. 0 1.5 , ...... 3% 74 63 5 Mar . 1-10 ...... 6. 0 5 40 ...... 1.8 11.7 22.4 33.4 42.8 51.7 47 .6 44.2 30.6 I9. 3 10.0 2. 3 ...... 318 80 62 6 11-20 ...... - 2.1 553 ...... 2. 5 12.7 28.9 43 .6 50.0 56.3 58.6 47.5 37.8 15.4 12.9 3.4 ...... 376 e5 Bo 7 21-31 ...... + 2.1 6 07 ..... a2 6.2 17.6 27.3 3a 3 46.0 53.3 49 . 1 43.7 33.0 24.6 13.8 4.9 0.3 ..... 358 57 56 7 Apr . 1-10 ...... 6. 1 620 ..... 0.7 7.8 14.6 23.7 35.9 39.0 18.2 44.7 46 .6 39.4 a8 LO .8 i .5 1.0 ..... 354 53 38 7 11-20 ...... 9.8 6 33 ..... 1.0 10.2 20.2 32.4 41.9 4s0 52.0 53.0 45.7 39.3 P 4 17.8 a9 1.4 ..... 401 68 51 6 21-30 ...... 13.2 6 44 ..... 22 13.0 ae. 7 36.9 54.5 58.2 59.9 57.4 522 43.0 $1.3 21.3 11.2 3.0 ..... 473 88 61 8 May 1-10 ...... 16.3 656 ..... 2. 6 13.4 25.7 37.1 14.0 55.6 53.7 80.6 56.6 50.7 15.2 25.0 14 .6 4.9 ..... 480 e5 53 7 11-20 ...... 18.9 704 0.2 5.1 18.3 34.3 50.3 80 .5 io. 4 73.8 T3.2 64.4 54 .s 11.2 27 .9 15.8 4.9 0.3 594 79 80 3 21-31 ...... 21.1 7u 0.3 5.0 14.6 31.8 44.0 M. 0 67.4 66.1 67.3 83.4 54.7 38.5 B.8 14.9 5.1 a4 556 73 65 6 JUne 1-10 ...... 22.6 7 19 a4 tr .4 ia 5 328 13.8 53.7 64.4 66.6 37.7 59.0 53.2 19.5 25.2 14.3 5.8 0. 6 552 72 88 a 11- 20 ...... 23.3 7 22 0.5 5.8 17.3 81.4 14.6 58.6 67.0 71.4 5b. 8 63.5 5.3. I 12 0 29.5 15. 6 7.9 1. I 576 75 67 7 21-30 ...... 23.4 732 0.4 5.9 14.7 29.8 14.7 56.2 57.7 64.5 33.2 40.8 40.0 17.0 23.5 13.7 7.6 1.2 510 67 53 6 July 1-10 ...... 22.8 720 a3 5.0 11.7 21 .1 10.0 311.2 45.4 4s. 8 52.4 55.8 44.4 19 .2 29.7 13.9 6. 0 a9 443 Bo 46 8 11-20 ...... 21.5 7 15 a2 3.6 124 24.2 34.4 44.4 58.3 57.8 52.7 60.2 4a 5 13.0 29 .8 !M.4 .0. 4 1.1 511 70 60 6 2l-31 ...... +19.5 7 07 0.2 3.7 15.9 .w .9 13.0 50.9 1.3 ea8 94 .1 b2.2 47.9 14.2 27.3 17.3 6.2 0.5 577 82 74 7 AN. 1-10 ...... 16.9 6 57 0.1 2. 6 11.5 22.2 14.9 45.4 51.9 49.9 54.8 51.9 44.0 P5 25.4 16.4 5.2 a4 450 67 51 8 11-20 ...... 14.0 6 47 ..... 2.4 14.2 26.8 41.8 57.3 56.2 56.1 58.2 53.9 42.4 8.4 23.3 12.9 3.9 ..... 4113 75 65 7 21-31 ...... 1a 5 635 ..... 0.8 8.4 1s1 28.1 35.3 38.8 43.4 44.3 381 31.6 &l 17.1 9.0 2.5 ..... 377 62 47 7 Sept. 1-10 ...... as 1 22 ..... a8 9.3 220 36.9 46.3 54.8 Ea1 53 .6 61.9 4s. 4 17.1 25.6 14.4 2.1 ..... 4So 84 75 5 11-20 ...... ao 6 10 ..... ai 85 14.7 27.0 37 .6 41.6 44.3 47 .6 47 .4 40.5 12.5 16.9 9. 2 1.0 ..... ass 68 Bo 6 21-30 ...... - 0.9 5 57 ...... 6.3 19.2 31.9 45. 9 54.8 57.9 57.0 5al 40.0 15.5 15.3 6.0 a3 ..... 410 83 78 3 ------I

Unauthenticated | Downloaded 10/01/21 02:25 PM UTC Amwr, 1914. MONTELY WEATHER REVIEW. aa 3. A table has been repared from which the zenith distance of the sun can i!)e readily determined at the time in each hour.when half the radiation for that hour. has lormal UrIaCe. Horirontal surfam. been recorded. 4. Entering: Tltble 8 with the zenith distance of the sun Solar and the percenta e of cloudiness as arguments, the factor dlation. Solar and sky radiation. F‘, is obtained, t7l e product of which into the number of - squares in each hourly area of the Calendar record mul- Maxi- Msxi- uaxi- Maxi- Mini- tiplied by gives the number of gram-calories of heat lum per urn per mum per um per 10 i‘ 1inute. EK low. day. day. received during the successive hours per square. centi- a 4 6 6 1 meter of horizontal surface. These results have been 1 I’ - tabulated, and the sums for the successive houw of each 1912. .? 9r.cal. Ir.-Clll. fr.-cal. day taken, to obtain the total radiation received during mag 17-20 ...... 19 22 ...... 1.63 3m the day. The decade averages for the successive hours May 21-31 ...... 17 57 1. n 1.57 312 June 1-10 ...... 16 28 1.41 1.63 427 and for the day are iven in Table 9. June 11-20 ...... lb15 1.32 1.01 91 June 21-30...... E44 1.25 1.63 122 Daily extremes and avewgcs of radiation.-- In columns July 1-10...... 16 21 1.21 1.56 82.8 I 683 377 6 and 7 of Table 10 are ken the masinium and the Jdg 11-20...... 17 39 1.21 1. m 91.9 I 759 217 July 21-31...... 19 40 1.19 1.60 86.7 i B&1 75 minimum daily amounts oP radiation, respectively, that Aug. 1-10 ...... 22 16 1. OB 1.69 90.2 gP3 1% hate been recorded in each decade. The mininium daily Aw. 11-20 ...... I 25 12 1.13 1.56 76.4 497 308 Aug. 21-31...... a8 42 1. I8 1.49 78.1 650 315 amounts have been recorded on foggy days with rain, &pt. 1-10 ...... 32 29 1.18 1.50 71.2 525 250 Sept. 11-20 ...... 36 16 1.17 1. I 73.4 4sB 130 and the maximum dail amounts have ususllv been Sept. 21-30...... 4008 1.19 1.33 w.5 480 50 recorded on esceptiona 1s clear days. The absolute Oet. 1-10...... 4402 1.25 1.25 68.1 450 309 T Oct. 11-20...... 47 48 1.34 1.17 61.6 416 masima thus far recorded in each decade have been Oet. 21-31.. I 61 33 1.17 1.10 3rd % ...... NOV.1-10...... 64% 1.25 1.08 49.0 303 36 plotted in Trace I, fi re S. This trace, therefore, rep- NOV.11-20...... 57 41 1.22 0.94 48.8 282 94 resents graphicallv t e daily amounts of radiation that Nor. 2l-30 ...... 6965 1.22 0.87 1.5 249 138 f Dw. 1-10...... 61 30 1.31 0.75 43.8 a50 niight be ex ecteh if the sky were continuously clear, or Dec. 11-20...... 62 21 1.24 0.89 43.0 247 l:; Dsc. %a1 ...... 6324 1.30 0.81 41.4 !&?I3 56 the possible Bail37 radiation at Mount Weather, throughout the year. The “Percenta e of possible radiation” in 19n. Jan. 1-10 ...... 61 31 1.78 0.74 43.6 4G Table 9 has been obtained y ditidin the “Daily aver- J~~11-20...... 6006 ...... 0.78 45.1 49 Jan. 21-31 ...... 67 47 1.30 0.88 47.0 arc ageJJgiven in that table by%% the possi le daily radiation Feb. 1-10...... 64 51 1.30 n. rn 56.7 32 for the decade, obtained from Trace 1, figure 8. Feb. 11-20 ...... 51 35 1.33 1.11 59.5 Fob. Zl-ls...... 4822 1.18 1.30 61.0 % In Table 9 are also given the “Percentage of possible Mar. 1-10...... i 4468 1.23 1.32 73.4 91 Mar. 11-20 41 03 1.35 1. %I 70.7 52 sunsliine” derived from the records of the Marvin sun- ...... 91 mar. n-31...... 3664 1.40 1.47 78.9 shine recorder, and the “ blean daily cloudiness” de Apr. 1-10...... 32 51 1.37 1.59 84.7 Apr. 11-20...... 29 10 1.41 1. M 88.3 2z rived from the eye estimates entered in the “Daily A r.21-30 ...... ma 1.30 1.62 91.1 96 dy 1-10 ...... aa 43 1.30 1.51 90.3 Meteorological Record ” ahead referred to. May 11-20 ...... 2008 1.30 1.61 Qo. 2 c In figure 8, traces I1 (+I, I1 (O),and IV (01, are may 21-31 ...... 17 I 1.28 1. e 91.9 75 I June 1-10 ...... 16 30 1.42 1. a 98.5 Jso the plotted decade averages of daily radiat,ion for the June 11-20...... 15 44 1.36 1.63 92.9 June 21-30 ...... 15 42 1. n 1.62 s.n 9 years 1913, 1913, and 1914, respectively. There are July 1-10 ...... 10 20 1.28 1. Q 850 marked variatioiis in these tlverages from decade to July 11-20 ...... 17 31 1.31 1. RE 89.7 % July 2141...... 19 31 1.23 1.51 86.3 394 decade, principdly on account of variations in the cloudi- Aug. 1-10 ...... M 11 1.30 1.44 82.8 Am.11-20 ...... =OR 1.32 1.33 78.7 c ness; but there is no marked difference in the successive 21-31...... 28 37 1.41 1.41 81.4 years. Tlie depression in the intensity of direct solar :a. 1-10 ...... 32 24 1.38 1.43 81.4 g swt. 11-20 ...... 36 11 1.35 1.41 73.7 radiation that culminated in August, 1912, is masked *t. 21-30 ...... 4001 1.42 1.31 71. a 73 Oct. 1-10...... 4358 1.37 1.33 85. a 62 by the cloud efk’ects in these plotted records of the total oet. 11-20 ...... 47 41 1. I 1.2: 81.9 47 radiation. The ercentage of possible radiation during Oct. 21-31 ...... 61 2e 1.z 1.2: 57.6 18 Nov. 1-10...... 54 51 1.38 0.91 51.7 35 the six months b:pril to September, inclusive, lias aver- NOV.11-20...... 57 38 1.82 0.H 47.7 32 Nov. 21-30 ...... 69 51 1.21 0.M 48.5 8 aged GS per cent, wlde for the remaining months it lias Dw. 1-10...... 61 1.80 0.w 45.1 29 averaged 63 per cent. The departures from these aver- Dw. 11-20...... 6a 2E 1.10 0.K 44.1 Dw. 21-81 ...... 6324 ...... 0. !3! 49. l: ages have not esceeded 1 per cent in any half-year period. 1914. The percentage of possible sunsliine during the above JSn. 1-10 ...... 61 S 1.21 1.6 46.1 12 periods has averaged 55 per cent and 53 per cent, respec- Js~.11-20 ...... 60Q 1.29 1. a 50.1 93 Jan. 21-31 ...... 67 5l 1.24 1.6 50. I 79 tively. Feb. 1-10...... M5! i.4a 1.11 65. I 24 Feb. 11-20...... 51 I( 1.30 1.2: ear 71 In a previous paper (13) it has been shown that during Feb. 21-ls...... 48% 1.4 1.31 81. 85 t,he second half of 1912 tlie total radiation on clear days Mar. 1-10...... 45501 1.35 1.6 77. i 73 Mar. 11-20 ...... 41 01 1.48 1.6 sa. f 90 averaged from 3 to 5 per cent less than on clear days bhr. 21-31 ...... 37 a 1.29 1.5 76. i 101 Apr. 1-10...... 32 51 1.s 1.51 m. ( 50 during corresponding months in 1913. ADr. 11-20...... 29 11 1. u 1.41 84. f 60 Tlie absolute mesimuin dail amounts of radiation for A- .2Ia...... 26 51 1.43 1.3 86.3 m 1%ay 1-10 ...... aa 4; 1. r 1.61 88. t 169 the first and tlie second decnc9 es of June and the second May 11-80...... 20 11 1.31 1.5: 85. * 178 mymi...... 18 0: 1.35 1.51 88. I 356 decade of July were not recorded on clear days, but on June 1-10 ...... 16 31 1.35 1.6 88. i m days when there was some cloudiness which, wlde ob- June 11-20 ...... 1s 4 1. 1. n rn! 93 June 2l-30 ...... 15 4 1.41 1.5 88. i 216 scuring tmliesun but little, materially increased the radia- July 1-10 ...... 16 21 1.s 1. fi 87. I 135 July 11-20 ...... 18 b 1. u 1.8 88. : 2i9 tion from the sky. Trace I, figure S, for clear sky radia- . July 21-81...... 19 3 1.s 1.6 a.t 3M tion, haa therefore been drawn below the maxima for 15~.1-10 ...... mol ...... 1.6 77. i 334 Aug. 11-20 ...... 250 1.3i 1. s 80. : a2a these decades. This trace shows that tlie total radiation Am. 21-81 ...... m3 1.31 1. I 77. : 74 Bept. 1-10 ...... 32 1 1.4! 1.6 78. 283 per day per square centimeter of horizontal surface, with eept. 11-20...... 360 1. I: 1.4 73. : 48 the clearest sky, varies from 250 calories on December 20 mpt. Pa...... 886 1. M 1.2 75: 185 ------. to 765 calories on June 10. On March 20 it is 605 calories

Unauthenticated | Downloaded 10/01/21 02:25 PM UTC 484 MONTHLY WELITHER REVIEW. AUGUST, 1914 snd 515 caloriea on Se tember 20. In general, the radia- In Qure 9, Trace I(+),represents the absolute max- tion received on clear gays during the half year December ima of solar radiation at normal incidence, based on the 21 to June 20 exceeds that received on clear days during pyrheliometric measurements made at Mount Weather the half pear June 21 to December 30, by about 8 per since September 21, 1907. Trace II(O), represents the cent. This is principally because of the increased water absolute maxima of radiation on a horizontal surface, va or content of the atmosphere during the latter period. based on the data hi column 4 of Table 10. Trace I11 %aximum radiation per minute and per hour.-In (a), represents the extreme hoilrly maxima of column column 3 of Table 10 is recorded the maximum solar radia- 5, Table 10, reduced to minute rates. It therefore gives tion measured by the Marvin pyrheliometer in each decade, the absolute maxima of radiation on a horizontal surface extra olated to the noon hour unless there is evidence when the sky is free from clouds. The actual maximum that tR e sun was obscured by clouds at noon on the day measurements for each decade are indicated by crosses when the highest readings were obtained. Usually the or circles in connection with the respective traces.

extrapolation was for a few minutes of time only, and Trace I for maximum radiation at noma1 incidence for not more than 0.01 or 0.02 gram-calorie of radiation. shows but little annual variation. A rate of 1.40 calories The maximum radiation per minute recorded in column per minute per square centimeter has been recorded in 4 of Table 10 has been taken from the Callendar recorda. ever month except January, and a rate of 1.45 calories, In nearly every case the record was made when clouds or ag ove, in every month from February to October, surrounded the sun but did not obscure it. Reproduc- inclusive, with an extreme of 1.50 calories in Mav and tions of records obtained under such conditions will be September. The maximum for January is 1.37 calories. found in the Bulletin of the Mount Weather Observa- The smallness of this range is accounted for by the fact vol. 5, pages 169-171. In reducing these records that in winter, when the sun's zenith distance at noon is "Tto eat units equation (2), page 480, for FL, has been a maximum, its distance from the earth, and the dust and ernnloyed. water vapor content of the earth's atmosphere, are at a The maximum radiation per hour given in column 5 minimum. of Table 10 has been recorded in most cases during the Trace I11 shows that when the sky is free from clouds hour just preceding or following noon on a day with an the maximum radiation per square centimeter of hori- -=aY~akY* zontal surface rangas from $0.77 ealorie par minute in Unauthenticated | Downloaded 10/01/21 02:25 PM UTC AWQUST,1914. MONTHLY WEATHER REVIEW. 486

December to 1.55 calories in June. It is twice a9 great At South Kensington, England, he luwimum radis in June as in December, as we would expect, since in tion per minute recorded by a Callendar pyrheliometer December, with a zenith distance of the sun of about 62O, in 1911 was 1.33 calories per s uare centimeter (16). the vertical component of the solar radiation is on1 Difwe 8ky md&zCon.-In 4able 11 are summarized er cent of its intensity at normal incidence, whie9 47in measurements of diffuse radiation from the sky, expressed iune with a zenith distance of the sun at noon of about in gam-calories per minute per s uare centimeter of 16O it is nearly 97 per cent. horizontal surface in the &st part o? the table, and as a Trace I1 shows that cloud reflection increases these ercentage of the direct solar radiation received on a intensities on a horizontal surface by about 0.15 calorie. Rorimntd surface in the second art of the table. The Thus, in December the maximum is about 0.93 calorie, data in the first line for Mount keather have been ob- and in June it is about 1.65 calories, with one observation tained from Table 7, and represent average clear-sky showing 1.69 calories. Both traces I1 and 111 show that conditions. The data for May 20 and dune 30, 1914, the rates during the half year December 21 to June 20 have been taken from the record sheets reproduced in

FIO. B.-MaxImnm redIatkm per minute. Curve I (+), llopmal inddence: I1 (0).hdzontd surface., with clouds ne8r the sun but not obscnrlog it; III (e),hdmntnl surface, wlth cloudles, !&y. avera e about S er cent higher than during the half fi ure 7. Ma 20 was a cloudless day, but unusually hazy. year f une 21 to 6ecember 80, as is the case with the Tie sky on Jyune 30 was unusually clear, escept that it maximum daily rates plotted in figure 8, trace I. was about two-tenths covered with cumulus clouds in It is of interest to compare the maximum rates of the afternoon. On May 20,32 per cent of the total radia- radiation given above with those obtained at the Trans- tion received during the day on a horizontal surface was vaal Observator , Johannesburg, latitude 26O 11’ south, diffuse sky radiation, and on June 30, 13 per cent. Or, altitude 5,925 Peet (14). The Callendar pyrheliometer on May 20 the diffuse sk radiation was 47 per cent of records give a daily maximum of 748 calories per s uare direct solar radiation, an on June 30 it was 15 per cent. centimeter of horizontal surface, in December and anu- For purposes of comparisonB there have been added to ary, and a maximum rate per minute in November of Table 11 measurements made 1.71 calories. The maximum rate per minute as measured Cal., with a bolometer (17), by the Angstriim pyrheliometer is 1.61 calories per square Canada, with a Callendar centimeter of normal surface, and it occurred in January. line for the latter The statement is made that the A triim instrument on three days with has been employed to standardize theT allendar pyrhelio- sents the results of meter, but the method of standardization is not given (15). The smallness of the excess of the masimum rates of radiation obtained from the Callendar records for Johannesburg over those for Mount Weather is probably z&%%eather measurements. due to the difference in the methods by which the instru- The Mount Wilson measurements are higher than we ments in use at the two observatories have been stand- would expect from the results obtained at Mount Weather lmlkid. . and Toronto. We may, however, compare these with

Unauthenticated | Downloaded 10/01/21 02:25 PM UTC 486 M0N"HLY WEATRER REVIEW. AUGTJBT,1914 earlier measurements by Abbot (IS), summarized in pyrheliometer, and has done most of the computing Table 12, most of which were made near midday. connected with the reduction of the records; and to the mechanician, Mr. A. J. Weed, who has introduced modi- TABLBIl.--Meaawements of difuae sky mdiation at diJerent stations fications and improvements in a paratus employed. (Czprased in grum-calories pm minute per s uare centinicter of hori- Also to the United States Bureau of tandards for experi- zontal swjace, ad also ns a percentage oj &e direct solar radiution 8 receised a horizontd sur/ace). mental work in determining the character of color screen on to br rmployed and for the loan of the screen.

! Altitude of sun. SUMMABY. By using, in connection with the Callendar pyrhelie meter, a screen that eclipses the sun at mtervals throughout the day, t.he record of the total radiation received upon a horizontal surface has been divided into ...... two components. One of these represents direct solar ...... radiation, and by comparing it with the vertical component of solar radiation intensity at nomrtl incidence, ...... as measured by the Marvin pyrheliometer, the factor to MOUNT WILSON, CAL be employed in reducing this solar radiation component Sept.1913 ...... 0.056 I 0.110 I 0.162 I 0.189 I 0.205 I 0.228! 0.240 of the record to heat umts is at once obtained. TORONTO, CANADA. The other coin onent represents diffuse sky radiation, July and Aug.. 1911...... 0.115 which is much ricR t'r in violet and ultraviolet light than Aug. 1, 1911 ...... is direct solar radiation. To obtain the factor for reducing ...I ... .. this part of the Callcndar record to heat, units the solar radiation transmitted b a ray filter that absorbs most f Ratio. skylsolar. cxpessed as a pmcentrrgc. I of the red and infra-reg radiation has been alternately measured by a Callendar and a Marvin pyrheliometer. I MOUNT WEATHEB, VA. Radiation reflected directly from clouds, or reflected % Feb., 1913-Feb., 1914...... diflusel through a layer of clouds, has been regarded Mwao. lg14, a. m ...... as of t e same general character as solar radiation st .m...... K lgll, a. m ...... 11 midday. b.m ...... For reducing records obtained when the sky is partly YOVNT WILSON, GAL overcast with clouds, the factors for reducing direct Bapt., 1913 ...... im I 17 I 31 I 24 I a0 . 17 . 16 solar radiation, diffuse sky radiation, and radiation from a cloud sky have been combined, giving each its TOEONTO, CANADA. pro er weight. &e probable error in the reduction of the component due to direct solar radiation is about 3 per cent. It is believed that the probable error in the reduction of the total radiation does not greatly exceed this. The total radiation per square centimeter of horizontal surface, with the clearest sky, varies, at Mount Weather, from 250 calories er day on December 20 to 765 calories sky miintion. on June 10. On E€arch 20 it is GO5 calories and on Sep- I tember 20, 515 calories. In eneral, the radiation statim. Date.. Gr.eal. Percentage received on clcar days during tf e half-year December I mic./cm:. 1 of solar. 21 to June 20 averages about 8 per cent more than that received during the half-year June 21 to December 20. Vug.1S. 1m ...... 0.081 1 8 The average daily radiation in summer is about 68 per Mount Wilson, Cal. (1731~m.) ...... Sept. and Oct., 1808.. . 0.108 12 Mount Whihiey &I. (4420 m.) ...... ~ug.17. lain...... 0.135 8 cent of that received on clear days, and in winter it is Fllnt Island. Sohh Paciac ...... Dec. 3.1907 ...... O.%5S 28 Baraour, Algeria (1160 m.)...... Sept.. 1912 ...... 0.245 23 about 63 per cent. - 1 1 . The maximum solar radiation at normal incidence varies from 1.37 calories per minute per square centimeter Taking into account the respective altitudes, the bad in January to 1.50 calories in May and September. The sky conditions that revailed at Flint Island, and the total radiation on a horizontal surface with a clear sky hazy condition of tE e sky at Bassour in Sc tember, varies from 0.77 calorie per minute in December to 1.55 1912, it appears that the measurements of d'zf use sky calories in .June. When clouds are near the sun, but do radiation made at Mount Weather fall well within the not obscure it, the momentary maxinium rates are limits of measurements made at other stations with increased by about 0.15 calorie. both similar and with very diffzrent ap aratus. The diffuse sky radiation received on a horizontal I wish to acknowledge my indebtc (Pness to my asso- surface at noon averages about 25 per cent of that received oiates for many valuable suggestions durin the rogress on thc same surface from the sun. On a clear summer of this investigation, and especially to &-of. 6m. R. day it amounts to from one-eighth to one-third of the Blair, with whom originated the scheme eniplo Pd for total radiation. reducing the intensity of solar radiation by ifferent Espressed in units of work, the receipt of one calorie amounts; to my assistant, Mr. Irving F. Hand, Bwho has of heat per minute per square centimeter of area repre- made nearly all of the observations with the Marvin sents 697 watts per square meter, or 0.78 horsepower

Unauthenticated | Downloaded 10/01/21 02:25 PM UTC AO~UST,1914. MONTHLY WEATHER REVIEW. 487 per square yard; and 1.50 calories per minute, or 90 tion of the solar spectrum in the ultra-violet. If the calories per hour, which is not an uncommon rate at light from the sun is examined bv means of a quartz spec- midday 111 summer with a clear sky, represents over I troscope, the spectrum is found‘ to terminate rather ab- kilowatt per square meter, or nearly 1.2 horsepower per ruptly near wave-length 3,000 b. U.,I while if the s ec- s uare yard. The radiation received on a square meter trum of the iron spark or arc is examined, it is founf; to o9 horizontal surface during a clear day in midsummer estend to near wave-length 2,000. That the ‘difference is equivalent to 5 kilowatt-hours; and a daily total per square centimeter of GOO calories, which is equalled or exceeded at Mount Weather on clcm days from the mid- dle of March to the end of August, is equivalent to 4 kilowatbhours er square meter. The daily average at this season of t5l e year, including all kinds of weatrhw. is about two-t,hirds of that for clew weather. REFERENCES. (1 Callendar givea a brief description of this instrument in Proc. roya BOC.,London, 1906, 77A:6. (2) Abbot, C. G., & Aldrich, L. B. Smithsoxian pyrheliometry empirical formula- revised. Smithsmian misc. coll., Washington, 1913,80, No. IS. (3) Coblentz, W. W., Radiation Constants of Metals. Bull. U. S. Bur. Standards, Washington. 1909, 6: sin 8 = 0.49 x ,-o.oss~(~-soo) (4) Patterson, J. Compariscin Callendar sunshine recorder. Ottawa, 1912. M. S. 50.) which eshibits the limiting wave-length 1 in its relation (5) Nichols, ddward L. Theory of the color of the sky. Phymcal to the sun’s altitude 8. NOW,as the dtitude decreases, Review, 1909, 26:498. the sun’s rays must pass through eater and greater thicknesses of the atmosphere; the o served phenomena, therefore, have been taken to prove rthat it is the atmosphere whose absorption determines the liniit of the spec- 1909, 91, Nr. 2. trum. In order to take the next step, however, and (8 Astrophysical observatory of the Smithsonian Institution, Annals, account for the exact form of the empirical expression, it Ws&rgtrm, 1Y13,3:147. 9) Ball, Fredrick. Altitude tables. Ixlndon, 1907. is not only necessary to espress the thickness of the 10) Kimball, E. E. A returu to normal atmospheric transparency. absorption layer hi terms of the sun’s altitude, a thing Jour. Wnshirgton acad. sci., Jan. 19, 1914,4: 17-25. which can be simply done if the layer is considered (11) Ives, Eerbert E. Color measurements of illuminants. Trans. but also the relation connectiii the absorption coe cient 1lluminrtir.geng. soc., 5, 1910, p. 196. (12) Nichols, Edward L. Daylight and artificial light. Trans. of the absorbing gas with Be wavelen th must be Illunlinatifig eng. soc., 3, 1908, p. 30.1. known. This considerably extends the fiek d of inqui , (13) Bulletin of the Mount Weather Observatory, Washington, 1914, for it becomes necessary to analyze the action of tY e 8209, T: ble 4. atmosphere and determine which of its constituents are (14) Transvaal observator . Annual report of the meteorological de .rtment. Years endipg fune 30, 1908, 1909,1910. responsible for the observed effects. r5) Transvaal observat AnnuPl report of the meteorological A study of the subject has made it clear that it is suffi- de rtment. Year eitding%ne 30, 1908, p. 5. cient to fis the attention upon oxygen and ozone since &) Corless, R. On the radiation records obtained in 1911 at South nitrogen and the other constituents do not appear to be n, together with a comparison between them and the cor- major factors, at least in this region. Going a step ZFnZ abdute observations of radistion made at ICew Observa- Meteorological Office, London, Geophyhydcs 1 memoirs, No. 4. farther, it seems that the part layed by oxygen is inter- 17) Abbot, C. G. Astron. jour. Mar. 14, 19!4, N?. 16, 28:!29.. esting, and may be important, \ ut as numerical data are 7’18) Astro ysical observatory of the Smithsomin Inshtutmn, lacking, it will be well to consider first the action of ozone Annals, War>on, 1908,2:146-153, and 1913,8:141-149. for whch the necessar figures are at hand. Hartley(1) suggestei some time ago that as ozone was known to possess a strong absorption band in that part THE ABSORhON OF THE ATHOSPRERE FOR ULTRA- of the ultra-violet where the sun’s spectrum ends, the VIOUT LIGHT. action of the atmosphere mi ht be ascribed to the ozone By Prof. THEODORELYMAN. which it contains. Recenty,f Fabry (2) and Buisson [Dated: Jderm Physlcal Laboratory, HardUnlverslty, Sept. 2l, 1014.1 The abso tion of the air, to which this article is de- voted, mani? ests itself in two regions of the ultra-violet and is of particular interest to two classes of observers. The first region lies at the less refrangible end of the ultra- violet where glass is no longer transparent; the second region is the extreme ultra-violet where even quartz loses ita transparency and where the ordinary photographic plate is no longer sensitive. It is the first region that is of interest to the meteorologist and to the student of cosmical phenomena, since it is here that atmos heric absorption influences the li lit we receive from ceP estial bodies. It is the second w% ich is chiefly of importance~ to the. pure phpicist. The absorption of the atmosphere in the first region 1 The hgstr6m unlt d wave length Is 10-u meters and k uc.mlly abbreviated 1. U.- finds ita most striking illustration in the abrupt termina- IC. A., Jr.1

Unauthenticated | Downloaded 10/01/21 02:25 PM UTC