CIHM Microfiche Series (■Monographs) ICMH Collection de microfiches (monographies) C«n«ll«» Iiwtituto for Hlrtorical Mlcror»productloi» / hwtltin e«i^^ Technical and Bibliographic Notes / Notes techniques et bibliographiques The Institute has attempted to obtain the best original copy available for filming. Features of this copy which may be bibliographically unique, which may alter any of the Images in the reproduction, or which may significantly change the usual method of filming ar« checked below. Coloured covers / Couverture de couleur Covers damaged / Couverture endommag^e □ Covers restored and/or laminated / Couverture restaur^e et/ou pelliculie Cover title missing / Le litre de couverture manque Coloured maps / Cartes g^ographiques en couleur Coloured ink (i.e. other than blue or black) C Encre de couleur (i.e. autre que bleue ou noire) D I — I Cotoured plates and/or illustrattons / D n n Planches et/ou illustrations en couleur Bound with other material / Reli6 avec d'autres documents Only ed'rtton available / Seule iditton disponible Tight binding may cause shadows or distortion atong interior margin / La reiiure serrie peut causer de I'ombre ou de la distorston le long de la marge intirieure. Blank leaves added during restorations may appear within the text. Whenever possible, these have been omitted from filming / Use peut que certalnes pages blanches ajout^es lors d'une restauration apparalssent dans le texte. mais. lorsque cela MA possible, ces pages n'ont pas M filmies. L'Institut a microfilm^ le meilleur exemplaire qu'il lui a M possible de se procurer. Les details de cet exem- plaire qui sont peut-ttre unk^ues du point de vue bibli- ographique, qui peuvent modifier une image reproduite, ou qui peuvent exiger une modification dans la m^tho- de normale de filmage sont indiqu^s cl-dessous. I I Cotoured pages/ Pages de couleur fj/l Pages damaged / Pages endommag^es □ Pages restored and/or laminated / Pages restaur^es et/ou pellicul^es Pages discotoured. stained cr ^ .' Pages dteolor^es, tachet^e; . 'i-'u>. ■ s I I Pages detached/ Pages d^tachoc: I I Showthrough/ Transparence [■/] Quality of print varies / D D D Quality in^gaie de I'impresston Includes supplementary material / Comprend du nutiriel suppl^mentaire Pages wholly or partially obscured by errata slips, tissues, etc.. have been refilmed to ensure the best possible image / Les pages totalement ou pariiellement obscurcies par un feuiilet d'errata. une pelure. etc., ont M filmies ft nouveau de fa(on k obtenir la meDleure image possible. Opposing pages with varying colouration or discok>uratk>ns are filmed twice to ensure the best possible image / Les pages s'opposant ayant des colorations variables ou des decolorations sont film^es deux fois afin d'obtenir la meilleure image possible. rri Addittonal comments / Commentaires suppiimentaires: Pagination is as follows: p. 177-189, 198-199. La pagination est comme suit: p. 177-189, 198-199. Th!« htm is fi»m«d »t th« reduction rtlte chtektd btlow / €■ decwmtnt t*t film< au uux d« rMuetlen Indl^ui ei-dttiou«. 10x 14x 18X 22x 26x 30x 1 / 1 1?¥ Ifx 20x 24x 2IX 32x Th« copy filmed h«r« has b««n raproduead thanks to tha ganarosity of: L'axamplaira film4 f ut raproduit grAca i la ginireait* da: University of Toronto Gerstein Sciences Information Centre Tha imagas appaaring hara ara tha baat quality poMibIa eonsidaring tha condition and lagibility of tha original copy and in Itaaping with tha filming contract spacificationa. University of Toronto Gerstein Sciences Information Centre Laa imagaa tuivantat ont iti raproduitas avac la plus grand soin, eompta tanu da la condition at da la nattat* da Taxamplaira film*, at •n conformity avac laa conditiona du contrat da fiimaga. Original capias in printad papar covars ara filmad beginning with tha front cover and ending on the last page with a printad or illuatratad imprea- sion, or tha back cover when eppropriata. All other original copiea ara filmad beginning on tha first page with a printed or illuatratad imprea- sion, sfWf ending on the last page with e printed or illuetrated impression. Laa asempleires originauK dont la eouverture en pepier ect Imprimte sent filmte en commencant par la premier plat at an terminant soit psr Is derni^re page qui eomporte une empreinte d'impression ou d'illustration, soit par la second plat, selon le cas. Tous lee autres exemplaires originaux sent fllmOs en commencant psr la premiere pege qui eomporte une empreinte dlmpreasion ou d'illustration at an terminant par la demiire pege qui eomporte une telle empreinte. The last rscorded frsme on each microfiche shsll contain the symbol -^ (meening "CON- TINUED"), or the symbol ▼ Imeening "END"). whichever oppiies. Maps, plates, charts, stc. may be filmed at different reduction ratios. Those too large to be entirely included in one exposure ere filmed beginning in the upper left hend comer, left to right and top to bonom, as many frames as required. The following diagrams illustrate the method: Un dee symbolee suivonts spparaftra sur is darniire image do cheque microfiche, sslon le cas: le symbols -^ signifie "A SUIVRE", le symboie ▼ signifie "FIN". Les cartas, planchaa. tableaux, etc.. peuvent *tre filmto A des taux da rMuction diffSrents. Lorsque le document est trop grsnd pour turn reproduit en un soul clichA, il est film* A psrtir do I'sngle supArieur gauche, de gauche A droite. et de haut an bas. en prenant le nombre d'imeges nAcsssaire. Les diegrsmmes suivsnts illustrent la mAthode. 1 2 3 1 2 3 4 5 6 f-nf l>r* ^ i^d .^---/''' THE EFFECT OF HAZE ON SPECTROSCOPIC MEASURES OF THE SOLAR ROTATION ? NOTE ON A SUPPOSED VARIATION IN THE SOLAR ROTATION "cr ' J- .r- KM. I'll I 1>I I I KV Ki I'liii. .1 r.ir |iri\,il' u. iiIjimmi Ir. 'h PPWPiiP mfm% 4* THK KFFi:CT OF HAZE ()\ SPKCTRi iSCOI'IC MKASIKKS OF THK SOLAR RO'IATIO.N Uy KAI.I'H K. in. I.l'KV Spectroscopic «leterminiitions of the rati- of the sun's rotatimi by diflerent observers at various times present ntnarkahle and puzzling dilTerences. Certain sources of error havi- l)een prove cation which seems to clear up much of the uncertainty attached t* the problem of the solar rotation. DIFFKRENCES IN MEASURES OF SnI.AR RdTATION Notable observed differences are: (i) The measurements of the solar rotation made by diferent observers exhibit a lar^e range of values. -Wn example, the values of the equatorial solar velocity, derived from about twenty groups of determinations, range from 2 . 1 1 to i . 86 km jwr sec. Further- more, measurements by the same obsen^er of a series of plates taken over a short internal of time frequently show a considerable range in their values. (2) Some observers have found a difference in velocity for differ- ent spectral lines, while others have not.— The determinations by Adams and Miss Lasby at Mount Wilson in iqo6 1908 show such differences," and this is confirmed' in 191 5 by St. John, Adams, and Miss Ware, and also in Ottawa by the writer in 191 5, a summary of these measurements being given later in this paper. On the other hand, all other observers (the writer included) find no serious difference for different lines in the interval 1909 i i^ Hence ■ .\dam?, Ml. Wilson Conlr., \os. 20, 24, 29; Astrophyskal Journal, j6, 20.?, 1907; il, 13, 1908; 39, no, 1909; .Adams and L.isby, Puhlicalion So. ,?*(, Carnegie Inslilulio). of Washinglon. • St. John, .\dams, and Ware, Popular .islronomy, ij, (141, ig'S- >77 V 178 RALPH E. Db LVRY the conclusion : The dijference in vrlocity for dijjerent spectral lines is a variable, being present in some ohsenations and apparently absent from others. {j,) One observer found that the northern and southern hemi- spheres of the sun rotated at different rates. -The observations of Hubrecht at Cambridge' alone give information on this jioint. In 191 1 the writer suRgested the method of using simultaneous ex|v«ures from the center of the solar disk and from the limi)s ft)r determining the rates of rotation in the two hemispheres in«le- pendentiy,' and since the apparatus was received in 1913 he .as been making such observations; in 1915 a similar method was started at Mount Wilson,^ so that more evidence on this jKiint will soon be at hand. (4) Some observations show a value of the rati- of rotation pro- gressively increasing with wave-length over the small range of wave- lengths covered by a plate; a great many more obsemUions do not exhibit this eject. — The 1906- 1907 series^ of Mount Wilst)n meas- urements show this effect, while the 1908 series' does not. Some measurements by Schlesinger in 1909. and those by Hubrecht* in 19:1 show the effect. All other observations appear to be free from this ctTcct. MECHANICAL EXi'LANATION OF DIFFERENCES The following explanations of the foregoing results have been or may be offered: {a) Instrumental errors. — Such instrumental errors as would be caused by uneven illumination of the prism or grating, com- bined with observations of the spectrum out of focus, may account for part of results (i), (3), and (4). (b) Observational errors. — Small inaccuracies in determining the points observed are possible, but it is unlikely that these could ever equal 0.5 per cent. ■ Hubrecht, Monthly Xolices, 73, 5, igij. ' T>e1,nTy, Report of the Chief Astronvmer, Ottawa, i9ii,p. 390. > St. John, Adams, and Ware, loc. cit. » Ibid. * Six- footnote i, p. 177. 'Hubrecht, loc. cil. V EFFECT OF UA/.E «70 (f) Errors of measurrmrnl. In iqio the writiT sukj'i'!**^"' ••>-i< errors of tm-.isurement miRht iucount for (i) and (i). He tested this explanation of (2) by meehanically intnuiucinK displaienienls of the sjK'Ctral lines the siime for all lints and of lonUnurations and magnitudes of displacement similar to actual oJ)>ervation>; a slight tendency to systematic ditTerence for dilTerenl lines was found in a series of twelvr- of these "imitation" rotation plates, as well as a systematic difference deiH-ndinj; on the direction of the plate.' These plates were taken in the region of X 4250. where Adams and Miss Lasby founed hy systematic error of measurement in their iqo6-iqo8 determinations remained unsettled. This ex])lanation of (i) was tested by having various observers measure the same lines on the same plates. J. S. Plaskett kindly offered to co-ojwrate with the wriur in meas- uring the above-mentioned twelve plates of the mechanical shifts, with the result that a systematic ditTerence between the two measures of ab^ut 2 per cent was discovered. This difference persisted throu^ out the measurements of the solar rotation in 1910-1Q13. These suggestions of the writer followed by the com- parative measurements led to these recommendations made at the meeting of the International Solar Uniim held at Bonn in igi.r "It is highly desirable to trace to their source the ;ystematic differences that are found in the values of the solar rotation by ditlerent observers Investigation should also be made into the personal differences that are found in measures of the same plates by different observers." (In this connection it would seem advisable to have a series of plates, say one or two from each observer, measured by the automicrophotometer at Mount Wilson, and then passed around among the various observers for measure- ment.) » De Lur>-, op. cit., p. 264; Journal of ihe Royal Astronomical Socitty of Canada, 5,384, 1911. • See footnote i, p. 177. I, So RM.Pll E. Df URY I'lIYSICAI. EXIM.WATIONS OF DIFKKRKNCKS Tlir fi>riv!?iiinK ex]>l:inations are hased U|xin the jxissihility of in-itrumeiital, observational, or jiersonal errors; those whieh follow are jja-ied on i>hysieal eonsiderations: (i/l Connilion in the solar atmos fthtrr . Ijk;\\ convertion currents undoubtedly account for some of the dilTereiues obtained by the same observer under apjiarently similar conditions (i), and it is <|uite iM)ssible that in small series of obser\ations the mean may Ik; con-iderably distorted by this cause. Adams found instances of such local motions in the neighborhood of sjM)ts.' The writer ffclru!>- II:' i nutid tlir iHi>>,il)ilily oi trmr i.iiim«1 by the sky >iHTtruni I !v..(linn with tlu-
  • i»lait'iH»truiii <>i tin- linili. ;iii(i ohscrvirs have U>t the fnt part hiin ( lul t<> mIiU the clearest days for uh-i rvation. Ilnwever. thi.r -Mini- to l>e error due to this s«»uree in -.onie of the oI>m r\ations. I ht '.vriitr made, in i()ii, some tests of the elTet t of >ky s|Hitriini in lissininj; the rotation displacement, with the result that for tln' \cr\ ilearot ciays there seenuil to he little error from this Mturn ' rhi-» work led to the consideration of ihe general (jueslio'i of hK lulid spectra, and in 1912 measurements of hlemls of >i)ectra of I'mh and center were nuule which showed a striking though predic rel,ition>hip between measured \\'m^ to the fact that the dilTerence in intensity for a lir;, at !imo at:d at »er.ler increases in general with ilecrease in intci i y oi the line.' Tht^e results led to the following , 'lan.ilion V though mentioned previously, presented only n ■ because . lent results in the measureiii- .U of the solar rotation at Ottawa by the writer, and at Mount Wilstm^ by St. John, .\dams. and Mi>s Ware are strikingly well exjjlained by it. (/j) Spectrum of haze. It has been shown* (see also later) that a vrariable haze, between the observer and the sun. cau>iiig to be blended on the spectrum of the limb a sin-ctrum of variable intensity and of character somewhat similar to that of the center of the solar tiisk in regard to intensity and wave length of the spectral lines, causes: (i) the spectroscopic determinations of the solar rotation to vary, and (ii) the vc >cities of rotation from the ditTerent lines to decrease in general with decrease in intensity of the lines, the amount of the decrease in velocity for a given line dejicnding on the strength of the continuous spectrum due to the haze relative to the continuous spectrum of the limb and on the ratio of the intensities of the line in the spectrum of the haze and in the spectrum of the limb. Observations already made make it seem jirobable that the variable terrestrial atmosphere and its ' I)c I.ury, Rcporl of the Chief Aslnyitomir, Ollaua, lyi i, p. rSi ■ l)e Lur>-, Journal of the Ro\,tl Ailronomical ^ i ■(■<>• of Cintulu, to, .'oi, lyid, • St. John, .\(Iams, and Ware, lotr. cil. » SfO fiH)tnote 2 . l82 RALPH E. De LURY clouds ;ind hazes arc sufficient to account for differences (i) and (2), after eliminating the systematic and accidental errors men- tioned alx)ve. But if in any series of observations the sjiectrum of terrestrial haze can be proved of insufficient strength, then we may introduce the idea of haze existing between the earth and the sun, near the sun, or even in the solar atmosphere (such as produced by matter falling in variable amounts into the sun and requiring an interval of time before being swept along in the general rotation). Such a variable haze possibly could account for the differences in the solar radiation observed by AblxJt and others; it would be interesting to make simultaneous observations of solar rotation and radiation to sec whether the changes in their values sj-n- chronize. That explanation (//) is the true explanation of the residual differences in (i) and (2) above, after due allowance has been made for the other known sources of error, seems established from the similarity of the following three series of results, dealing with measurements of blended spectra, measurements of the solar rotation at Ottawa on plates made through different amounts of haze, and measurements of the solar rotation made at Mount Wilson. MEASUREMENTS OF BLENDED SPECTRA In the paper cited, it has been shown that the measured rota- tional displacements of the lines from the limb when blended with the lines — undisplaced by rotation — from the center of the solar disk decrease progressivel}' with decrease in the intensity of the lines; and this was explained as due to the fact that the difference in intensity between lines in the spectra of center and limb decreases, in general, with increase in intensity of the lines. There are exceptions to this latter generalization which serve to test the vari- ous theories (see later). The accompanying summary (Table I) of the first table in the paper quoted will suffice to illustrate the general results. It is thus seen that the lessening of the displacement due to rotation in the blend with the spectrum of the center is greater progressively with decrease in intensity of the line, which, in turn, is accompanied by steadily increasing values EFFECT OF HAZE 183 of the ratio of intensity from center to limb, and decreasing values of the ratio of width at center to limb. (Thus decrea>inj; intensity at the limb seems to be accompanied by increasing width. 'J"o explain this the writer has advanced the hyiH)thesis that the TABI.K I BLE>mED SPECTKA, X 5')00 Mean displacements of equalorial limb lines l)Iendo(l with five dilTcrent Minds with center spectrum in . '.ch the ratios of the densities of (le|H)sit on the photo- graphic plate from the continuous siwctrum of ti.e limb to the total of (onliniious spectrum were o 8g, o 8j, 0.74, o 6i and 0.54; mean ratio, o ;;. Plate, L854, September 20, ic i Line-intensity, center. ..!i!2l3'4:S<* * Line-intensity, limb 011-2 35"'^ 7 Line-width, center 48 i4.i 5 4 ''8 7 8 7 ^ gjkmiiersec. Line-width, limb 6.2 ,56 s« <> <> 82 80 8 6 " " No. of lines 3 I 6 ; s ^ ' ' ' Mean velocity from I i ,, . blends : l.S53i "5731 '575 '594 i <'Ai ' ''4^ I ''5' Kqii torial velocitv not i 1 ' bUnded ' \ 1.02b 2.053 2.014 J 053 ^ °>^5 ' °^5 ' O;** " Mean | i 1 1 ' '•°4' widening and weakening of the lines at the limb arc due to convec- tions similar to those in the penumbral regions of sjiots. Other factors come in to play and account for many exceptions. Ihe question will be discussed soon in another communication.) MEASUREMENTS OF THE SOLAR ROTATION AT OTTAWA ON HAZY DAYS The results from the measurements of the solar rotation on hazy days for varying degrees of haze show a striking similarity to those from the measures of the artificial blends as shown in Tables II, III, and IV. It will be thus seen that the dilTereiices of percentage between the values for intensity i and 22 are: Table II, 1.2; Table III, 4.6; Table IV, 8.2. After the observations of Table IV were made, a photographic comparison of the intensity of the spectrum of the haze relatively to the si)ectrum of the limb was secured; however, the haze was continually varying so that only a rough approximation could be arrived at. and from this it would seem that the average ratio of intensity of the contiituous i84 RALPH E. De LURY spectrum of the haze to the continuous spectrum of limb and haze for the observations of Tabic IV was 12 ± per cent. This would involve the assumption that the haze in Table II was about 2 per cent, while the haze for observations of Table III was al)out 7 per TAUI.K II Solar Rotation, X j.'oo March ii, 1 130 p.m., 1916, very slightly hazy, (> double obscrxations, i.e., 2 striiw ot spectrum from each limb Mean Me;m Intensity j 1 2 Number of lines j 3 11 Kquatorial velocity 1 i . 956 1972 Mean S3 7 1.972 3 1.968 Mran Si I 967 TABLK III Solar Rotation', X 5200 June 16, 4: IS P.M., 1915, slightly hazy, 6 double observations I Ifcfta I Mean { Mean Mean Intensity i 2 33 2^ Number of lines j 3 11 i 7 3 Equatorial velocity ' 1.808 < 1.842 j 1.845 1883 Me 5 3 ■!4 1 843 TABLE IV Solar Rotation, X 5200 March 3, 12:55 p-m., 1916, very hazy, haze varying, 3 double observations Mean Men Mean Mean Mean Intensity I 3 I 738 2 II 1.760 Si 7 1. 814 22 3 1.887 5 3 Number of lines Equatorial velocity ^4 I 816 cent. These arc of course only rough estimates, but the\- serve to point out the necessity of very accurate measures of the relative strengths of the spectrum of the haze and the spectrum of the limb. When such are made and accurately correlated with measurements of solar rotation for groups of lines of different intensities, it will be jwssible to eliminate the effect of the spectrum of the haze from any similar series of measurements of rotation. Such being the r EFFECT OF II A /.E 185 case, it should be ix)ssil)le t<> estimate the stnn;,'th of h.i/e ]iresiiil durinj^ the Mount Wilson observations.' MEASUREMENTS OF THE SOI.AR K(>TATIi>\ AT Mdl NT WTISmN TABI.K V Solar Kotatiun, X 5:00 1914-.915 moasurcmtnls (St. John, .\iliims, ;iml Ware, I'opiiltir .1 -/mh.j.hv. ij, 64 1 , I q 1 5 1 i Mr.in Mem Mcin [ Mr.in Mrin Intensity i 2 4 g ::; t Number of lines 2 5 g s 10 Kqu-itiirial velocity i q;4 i 0}.i i q45 -' 04 i i i^o It is seen from Table V that the ditTerenci- between the \:ilues for lines of intensity i and 22 is 6. i j)cr cent. It would seem that, if this is altoi^ether ascribable to ha/e, there was an overlapping spectruiTi of the haze of about q± per cent in the.se observations. The three lines of average intensity 22 were the same as in the Ottawa observations, namely, the three stronjr Mg lines in the b group, X 5167 to X 5184, but the lines of inten>ify i could easily yield dilTerent results in the two series, ^ in the ( )ttawa oi)servations and 2 in the Mount Wilson observations. However, it seinis likely that there must have l)een a cotisiderable elTect of ha/e during the latter observatit)ns. The large ditTerence between the values of the rotation in the two series is probably accounted for by some of the other sources of error, though the Ottawa \ahiesin Table II (very slight haze) are nearly the same in the mean the strongest lines being, however, e.\cei)tionally high in the Mount Wilson measures. That is a (juestion which can best be attacked after the influence of the spectrum of light scattered from haze or optical parts has been accurately eliminated. MEASUREMENTS TO TEST THE I.EVEI. HYPOTHESIS The measurements of the solar rotation given in Tables \T and VII seem to support the haze e.x])lanation and to dis])rove the level hypothesis, unless the latter be assumed to be variable, as pointed out above. ' St, John, .Xdanvs, and Ware, loc. cil. 1 86 RALPH E. De LURY In Table VI are given the measurements of 6 lines of intensities o and I paired off with 6 lines of intensities 4-15, giving a great difference in penumbral displacements in spots, interpreted as TABLE VI SoLAH Rotation, X 4500 Ottawa, June 30- July 25, 1910, 32 observations Mean Intensity 0.7 Xumbor of lines 6 Tcnumbral displacement . . +0 028 A Kquaturial velocity 1.968*0.003 Mean Mean 8 5 6 — o 001 A 1.972*0.007 46 12 +0.015 1.970*0.004 (lines) *o. 010 (plates) Seven of the foregoing plates taken on cloudy or hazy days, yield : Equatorial velocity. 1.909 »-939 1.924 km per sec. TABLE VII Solar Rotation, X 5600 Ottawa, December 6-12, 1910, 32 observations I Mean I Mean ! Mean Intensity 1.6 6.2 ■ 3.9 Number of lines ! % : S J! 10 Equatorial velocity j 1.930*0.006 ^ 1.936*0.002 ] 1.933*0.003 (lines) ! I * 0.005 (plates) indicating range in level in the reversing layer (Evershed and St. John).' If Adams' hypothesis of increasing angular velocity for increasing elevation in the sun be true, there should be a con- siderable difference between the .elocities of rotation from these two groups of lines. There is no appreciable difference, however, and the results of Table VII show this also. We are thus forced to abandon the level hypothesis, or else to modify it by adding the idea of variability. From the seven plates of Table VI taken on days when the spectrum of the haze was stronger than for the other ■ Evershed, Kodaikunal Ohservalory Bulletin, 15, 1909. St. John, ill. Wilson Conlr., Nos. 6g, 74; Aslrophysical Journal, 37, 322, 1913; 38,341, 1913. 'f EFFECT OF HAZE 187 '( plates, it is seen that there is a difference between the (Ieterminatit>ns of velocity of the two groups of lines, of o . 0,^0 km ]>er sec. a ditTer- ence explainable by the spectrum of the haze blendin); with the spectrum of the limb. The mean value of the December determi- nations (Table VII) is smaller by 2 i)er cent than the value from the July determinations (Table V). jxissibly owin^ to the lower declination of the sun in December and to the lower nvan intensity of the lines, as well as to the probably greater relative stren-^th of the sky spectrum in December than in July. SOME GENERAL DISCUSSIONS It has been mentioned that observers during igoq-iqiT, found little difference for different lines. Can this be due to the fact that at sun-spot minimum there is less danger from the error due to haze than during sun-spot maximum, pointing either to the presence of varying quantities of matter al)out the sun or to var\ ing haziness in the terrestrial atmosphere caused by the v:_ 'atinn in its ion- ization accompanying the s]H)t-activities ? In most of these igcx^- 1913 observations the lines were not considered in groups as to difference in intensity, but rather with regard to the element pro- ducing the line — in accordance with the recommendations of the Solar Union in 19 10— the imjiortant relationship between peiium- bral displacements and intensity and level' not having been fully developed at that time. It would seem ad\isablc to investigate the published results from this point of view. This has been done in a preliminary way by the writer. Some results show no ajipre- ciable relationship of velocity with line-intensity, some show evi- dence of this, and some seem to indicate the reverse of what would be e.xi)ected from Adams' level hyjwthesis, i.e., a lower rate of rotation with increasing level, a physically iH)ssil)le and quite probable state of affairs. Some exceptions to the level hyiK)thesis are readily explained on assumption of blended spectrum of haze, e.g., X 4287.566 of intensity i at the center of the solar disk is strengthened and widened at the limb, and it has a penumbral displacement of 0.026 A; if this is inteqireted as meaning low level, it is to be expected on the level hypothesis that this line ■ IhUi. i88 RALPH E. I)E LIRY shmilil '^w :i lower rotational value than the mean. Adams and Lasln- lind' in i(>o8 that this line has an etjuatorial velocity 0.004 km per sec, above the mean; this is exi)Iainal)le by the fact that this line is strengthened, not weakened, at the limb and therefore should yield a larger value than the mean of the other lines which are for the m<itii-s(>l lini •<. say from two groups ')f lines, otu- grt-ally wi-aki'iU'd at tin- limhaiid tlu' other not wcakeind at the iinil). I'he \ 5200 region olTers the hest chance for such measurement, since the strongest lines there are quite nieasureahle, and it is iH)ssihle to elinnnatt- instrumental and other errors In usinj; when desired either iodine or chlorine comparison spectra las suji^ested by the writer' in kjio and kjii and emplinearisons as follows: July ij, H«y H. H. Plaskett's values for five lines of intensities 3 to 8, XsQOO De Lury's values: 4225.6, Fe, 3 j 4226 g, Ca, 20 i 4232.8, Fe, 2 Difference between Ca line and Fe lines km pence. 1.846 I. 712 1-794 1 .711 0.082 July ao, Bright km per >ec. 2.026 1.966 I 983 1.972 0.014 ■ H. H. Plaskett, Astrophysieal Journal, 43, 156, 19 16. 198 ML\OR COST RI BUT loss .\SD SOTES 199 These measurements show three results, which are all exi)lainahle by the blending of the spectrum of haze with that of the solar limb: (i) The values of the solar rotation are smaller on the hazy day than on the bright «lay. (2) The difference In-tween the values for weak and stmng lines is greater on the hazy day than on the bright day. (0 The values at the greater wave-length. X 5(>oo. are greater th;in those at the smaller wave-length. X 42.^0, the sjH'ctrum of haze being sti >nger relatively to the sjH'ctrum of the limb for smaller wave-lengths than for the greater wave-lengths. The values at X 42,^0 ix)int to a value for the ecjuatorial velocity of about 2.0,? or 2.04 km per second for a zero difference In-twcx-n the values for weak and for strong lines. Measurements of the X 5900 plates for gn)ups of weak and strong lines will no doubt confirm the conclusion that the variation in question is due entirely to variations in the terrestrial haze. Ralph E. De Lvrv SoLAK Physics Division DoyiNio.'j Obsekvatory, Ottawa .Xugust 1916 *