Jffpliif
SiBi'i? 1
: > ' ' ' • I ; , : ; ; - •
III!
THE SUN
THE SUN
BY
CHARLES G. ABBOT, S.M.
DIRECTOR SMITHSONIAN ASTROPHtSICAL OBSERVATORY
WITH NUMEROUS ILLUSTRATIONS
NEW YORK AND LONDON
D. APPLE TON AND COMPANY
1911
COPYRIGHT, 1911, BY D. APPLETON AND COMPANY
, 1911
Printed in the United States of America
PREFACE
WITHIN the last fifteen years we have seen the publi- cation of Rowland's great table of solar spectrum wave lengths, the establishment of the Yerkes, Ko- daikanal, Mount Wilson and other observatories largely devoted to solar researches, the photography of the spectrum of the corona and of the chromosphere at total solar eclipses, Hale's brilliant discovery of magnetic fields in sun spots, the determination of the rotation periods of the sun at different levels, as well as at all solar latitudes, Langley's bolometric investigation of the sun's infra-red spectrum, and the recent Smith- sonian determinations of the absolute intensity of the solar radiation outside our atmosphere. The great interest in such researches has been marked by the establishment of the International Solar Union, and its enthusiastic gatherings of the foremost investigators from all lands.
The time seems ripe for collecting the splendid array of new solar knowledge which such unprecedented activity has produced, and for discussing the probable nature of the sun in the light gained.
In the following pages the professional astronomer will find hitherto unpublished results of researches, and new explanatory hypotheses, illustrated by many new text figures and engravings.
v
PREFACE
Chapter II has been devoted to a description of the methods and principles employed in modern solar research, and Chapters VII to X on the relations of the sun to life upon the earth, and to the starry universe in general. Thus the book, while primarily devoted to the sun, may, I hope, serve as an introduction to the study of astrophysics for school and college use, as well as for the general reader.
In Chapters VI to IX are given many facts likely to prove of interest to the meteorologist, geologist, botanist and engineer.
Professor Young's "The Sun" is now out of print, and as it is hoped that the present work may to some extent take its place, I have been permitted to use some of his illustrations, notably in Chapter IV, and to make several quotations from his text, the longest in Chapters IV and VI. I desire also to acknowledge my obligations to many who have given suggestions, information and illustrations. Especially I offer thanks to F. E. Fowle, S. A. Mitchell, W. W. Campbell, G. E. Hale and the staff of the Mount Wilson Solar Observa- tory, E. B. Frost, H. M. Chase, A. G. Eneas, Henry Holt and Company, the Superintendent of the United States Naval Observatory, M. J. Moore of the United States Patent Office, and Messrs. Briggs and Shantz of the United States Department of Agriculture.
C. G. ABBOT.
WASHINGTON, July 10, 1911.
CONTENTS
CHAPTER I
PAGES
THE SOLAR SYSTEM. THE SUN'S DISTANCE. ITS DIMENSIONS 1-30
CHAPTER II
THE INSTRUMENTS AND METHODS USED IN SOLAR INVESTI- GATION . 31-84
The Telescope. — The Coelostat. — The Spectrum and What It Indicates. — Spectroscopes. — The Spectro- heliograph. — The Heliomicrometer. — The Comparator. —Nature of Radiation — Laws of Radiation. — Spectra of Different Sources. — Pyrheliometry. — Bolometry.
CHAPTER III
THE PHOTOSPHERE 85-127
Telescopic View. — The Photospheric Spectrum. — Row- land's Spectrum Tables. — Chemical Elements Found and not Found. — Corrections to Rowland's Wave- lengths. — Levels . — Pressures . — Convection Currents . — Limb Spectra. — The Variation of the Sun's Brightness. — Solar Temperatures. — Spectro-heliography. — The Solar Rotation.
CHAPTER IV
ECLIPSES AND THE OUTER SOLAR ENVELOPES . . 128-182
The Saros. — Eclipse Expeditions. — The Corona. — The
Chromosphere. — Jansen's and Lockyer's Discovery. —
The Spectra of the Chromosphere and Prominences. —
vii
CONTENTS
PAGES
Prominences and the Spectroheliograph.— Recent Flash Spectrum Observations. — The Heights of Different Metals in the Chromosphere. — Mitchell's Observations of 1905.— Campbell's Observations.— Chromospheric Spectra in Full Daylight.
CHAPTER V
SUN-SPOTS, FACULAE AND GRANULATION . . . 183-214 Sun-spot Periodicity —Drift —The Distribution of Sun- spots. — Their Formation and Life History. — The Sun- spot Level. — Langley's Typical Sun-spot. — Faculae. — Granulation. — Sun-spot Spectra. — Sun-spots and Mag- netism.— Radial Motion in Spot Penumbras.
CHAPTER VI
WHAT is THE SUN? 215-279
Young's Views. — Halm's Views. — Schmidt's Hypothesis. — Julius' Views. — The Author's Views.
CHAPTER VII
THE SUN AS THE EARTH'S SOURCE OF HEAT . . 280-330 Causes of Low Temperatures at High Altitudes. — Meas- urement of the Intensity of Sun Rays. — Dependence of Solar Radiation on Air Mass — The Transmission of the Atmosphere. — The "Solar Constant of Radiation." — The Light of the Sky.— The Dependence of the Earth's Temperature on Radiation. — The Fluctuation of Solar Emission. — Geological Temperatures.
CHAPTER VIII
THE SUN'S INFLUENCE ON PLANT LIFE .... 331-361 Plant Requirements. — The Assimilation of Carbon by Autotrophic Plants. — Etiolation. — Plant Geography. — Light Requirements of Plants. — Heliotropism. — Plants as Energy Accumulators.
viii
CONTENTS CHAPTER IX
PAGES
UTILIZING SOLAR ENERGY 362-390
Experiments with Burning Mirrors. — The "Hot-box" Principle. — Mouchot, Pifre and Ericsson. — Eneas' Solar Engines. — Properties of Glass. — Solar Heaters and Res- ervoirs.—Low Temperature Solar Engines.— Solar Cooking Appliances.— Solar Metallurgy.— Resume".— Quantity of Solar Energy Available. — Thermo-dynamic Efficiency. — Reflecting Powers of Mirror Surfaces.
CHAPTER X
THE SUN AMONG THE STARS 391-434
Stellar Distances. — Magnitudes. — The Sun's Magnitude and Light Emission. — The Solar Motion. — Star Groups. — Double Stars. — Stellar Masses and Distances. — Mira Ceti and the Sun. — Stellar Spectra. — The Classification of Stellar Spectra. — Radiation Distribution. — Evolution of the Solar System. — Stellar Evolution
CONCLUSION 435-437
INDEX 439-448
LIST OF PLATES AND ILLUSTRATIONS IN TEXT
LIST OF PLATES FACING
PAGE
Langley's typical sun-spot Frontispiece
PLATE I. — Smithsonian observing shelter and coelostat
(Mount Wilson) 36
PLATE II.— The Zeeman effect (King) 44
PLATE III.— Solar photograph (Ellerman) .... 86 PLATE IV. — Fig. 1. Spectrum of procyon (Adams); Fig. 2.
Spectrum of east and west limbs of the sun (St. John) . 88 PLATE V. — Calcium spectroheliogram, H2 (Ellerman) . .114
PLATE VI. — Hydrogen spectroheliogram, Ha (Ellerman) . 116
PLATE VII. — Calcium spectroheliogram, Hx (Ellerman) . 118
PLATE VIII.— Calcium spectroheliogram, Ha (Ellerman) . 120
PLATE IX.— Calcium spectroheliogram, H2 (Ellerman) . 122
PLATE X.— Hydrogen spectroheliogram, Hy (Ellerman) . 124
PLATE XL— Hydrogen spectroheliogram, Ha (Ellerman) . 126
PLATE XII.— Solar corona, 1900, May 28 .... 132
PLATE XIIL— Solar corona, 1905, August 30 ... 134 PLATE XIV. — Spectroheliograms of solar prominences (Slo-
cum) 166
PLATE XV.— Flash spectra, 1905, August 30 (S. A. Mitchell) 174
PLATE XVI. — Photographs of a portion of the sun (Jansen) 202 PLATE XVII. — Spectra of photosphere and sun-spot (Mount
Wilson Solar Observatory) 210
PLATE XVIIL— Effect of pressure on gaseous spectra (Gale) 248
PLATE XIX.— The Pleiades (Ritchey) 398
PLATE XX A.— Stellar spectra (Campbell) . . . .406
PLATE XX B.— Stellar spectra (Campbell) ... 408
PLATE XXL— Stellar spectra (Hale and Ellerman) . . 410
xi
PLATES AND ILLUSTRATIONS IN TEXT
PAGE
PLATE XXIL— The great nebula in Orion (Ritchey) . . 418
PLATE XXIIL— Nebula N. G. C. 6992 Cygni (Ritchey) . 420
PLATE XXIV.— The great nebula in Andromeda (Ritchey) 422 PLATE XXV.— Spiral Nebula M. 51 Canum Venaticorum
(Ritchey) 424
PLATE XXVI.— Fig. 1. Nebula H. V. 24 (Ritchey); Fig. 2.
Ring nebula in Lyra (Ritchey) 426
LIST OF ILLUSTRATIONS IN TEXT
FIQ- PAGE
1.— Relative distances of planets 5
2.— Geometrical parallax method 10
3.— Velocity of light (Fizeau) 19
4.— Velocity of light (Foucault) 20
5. — Determination of constant of gravitation .... 29
6. — Projecting the solar image 32
7. — Herschel's solar eye-piece 33
8. — Shade glasses 34
9. — Polarizing eye-piece 34
10. — Rapid exposing solar shutter 35
11. — Prismatic refraction of light 47
12. — Prismatic spectroscope 50
13.— Diffraction of light 51
14. — Plane grating spectroscope 53
15. — Concave grating spectroscope 54
16.— The 150-foot tower (Mount Wilson) .... 58
17. — Energy-spectra of the sun and the perfect radiator . 69
18. — Pouillet's pyrheliometer 73
19.— Abbot's silver disk pyrheliometer 76
20.— Angstrom's pyrheliometer 77
21. — Angstrom's pyrheliometer. Details 77
22. — Abbot's water flow pyrheliometer 79
23.— The bolometer 81
24. — Bolographs of the solar spectrum 83
25. — Brightness on solar disk 100
26. — Energy spectra on solar disk 109
xii
PLATES AND ILLUSTRATIONS IN TEXT
FIG. PAGE
27. — Huggin's first observation of a prominence in full sun- shine 143
28. — Double reversal of the D-lines 147
29.— Double reversal of C-line 148
30. — " Arrow-head " spectrum 148
31. — Opened slit of the spectroscope 149
32 . — Chromosphere and prominences as seen in the spectrum . 151
33. — Relative frequency of protuberances and sun-spots . 153
34-39. — Eruptive prominences 157
40-45. — Quiescent prominences .. 159
46-51. — Forms of prominences 163
52. — Motion indicated by prominence spectra . . .165
53. — Sun-spots and terrestrial magnetism 187
54. — Sun-spots and terrestrial temperatures and magnetism . 191
55. — Spoerer's curves of sun-spot latitude 194
56.— Solar diagram (Young) 216
57. — Normal and anomalous dispersion 233
58. — Michelson's hypothesis 260
59. — Displacement diagram 261
60.— March of insolation (Mount Wilson) .... 285
61. — Holographs of the solar spectrum 292
62. — Atmospheric transmission plats 294
63. — Insolation and terrestrial temperatures . . . .316 64. — Apparent variations of the sun's radiation (Smithsonian
observations) 321
65. Hypothetical temperature diagram 325
66.— Stomata (Schwendener) 338
67. — Promotion of carbon assimilation by light, and absorp- tion of light 343
68. — Eneas' solar engine 367
69. — Boiler of Eneas' solar engine 368
70. — Adams' solar cooker 380
71. — Intensity of sun rays (Mount Wilson and Washington) 384 72. — Intensity of solar radiation (Different latitudes and
elevations) 385
LIST OF TABLES
I. — Principal data of the solar system .... 3
II. — Velocity of light in different media .... 46
III. — Principal solar spectrum lines 90
IV. — Chemical elements found in the sun . . . 91
V. — Chemical elements doubtfully occurring in the sun 92
VI. — Corrections to Rowland's wave-lengths ... 96
VII. — Distribution of radiation over the sun's disk . . 107
VIII. — Energy spectrum relations over the sun's disk . 110
IX. — Daily rotation of the sun's surface .... 126
X. — Measures of ninety-two lines in the flash spectrum
near H§ . . . 178
XI. — Years of sun-spot maxima and minima and maxi- mum intensities 186
XII. — Hydrogen spectrum in sun-spots .... 206
XIII. — Spectrum lines affected in sun-spots . . . 206
XIV. — Sun-spot, hot arc and cool arc spectra . . . 209 XV. — Moissan's experiments on the vaporization of the
elements of the iron family .... 238 XVI. — Differences between observed and computed values
of atmospheric transmission : 242 XVII. — Observed intensity of solar radiation, Mount Wil- son, July 6, 1910 237
XVIII. — The intensity of radiation in different parts of the
solar spectrum . . . ... . . 288
XIX. — Transmission for total solar radiation . . . 296
XX. — Atmospheric transmission for homogeneous rays 297
XXL— Sunlight and sky light (Exner) . . . .301
XXII.— Sunlight and sky light (Roscoe) . . . .302
xv
LIST OF TABLES
TABLE
XXIII.— Sky light on a vertical surface (Exner, Schramm) 303 XXIV.— Sun and sky light. Relative brightness for dif- ferent wave-lengths (Mount Wilson) . . 305 XXV.— Average brightness of sky zones. Flint Island
and Mount Wilson . . . 395
XXVI. — Ratio of total radiations: Sky to sun . . 307
XXVII.— Yearly means and mean daily temperature de- partures 311 XXVIII. — Composition of food products
XXIX. — Economy of Helianthus Annuus . . 361
XXX.— Percentage reflecting powers of various reflectors 3 88 XXXI.— Classification of stellar spectra 410
XXXII.— Intensities in stellar spectra (Wilsing and Schei-
ner) . . . .411
XXXIII.— Spectroscopic binaries. Spectral types, periods and
eccentricities (Campbell) . . . 422
XXXIV.— Spectral types and velocities in space (Kapteyn) 430
INTRODUCTION
WE depend on the sun for life, warmth, light, and ajl mechanical and electrical powers. Its constant supply of heat is necessary to prevent the oceans and even the air from freezing. The supply of coal which we are now using is but an evidence of the sun's light in former ages. To pass to the enumeration of the comforts and luxuries and beauties we owe to the solar rays would lead us far astray, and is indeed wholly unnecessary because all men acknowledge and many worship the sun as the source of these bene- fits. It would be a gross neglect to omit the closer investigation of such unique relations as those the sun maintains to life. Yet the study of the means of increasing the usefulness of the sun has been ne- glected, and it is rather in the investigation of its curious features that solar researches have gone farthest.
The enormous brilliance and heat of the solar rays suggestive of temperatures far above any which can be produced on the earth; the marked dimness and brown shade of the edge or limb of the sun relative to the center; the fluctuating march of spots across the disk; the variable rates of rotation of the sun's sur- face in different latitudes; the brilliant markings 2 xvii
INTRODUCTION
called faculse which accompany the spots; the weird and highly beautiful phenomena of total solar eclip- ses; all these have long been the objects of minute study. In the last half century the development of the spectroscope has led to great progress in the more intimate and satisfactory knowledge of the sun; so that we now know many of the chemical elements of which it is composed; the approximate temperature of its surface ; the motion of the vapors at and near the surface; the approximate pressure under which they lie ; the magnetic character and cyclonic struc- ture of sun spots ; their relative coolness as compared with their surroundings; besides many other details hardly to be credited as known of a body situated nearly ninety-three millions of miles away.
By bonds unseen yet altogether stronger than a bar of steel thousands of miles in diameter the sun holds to itself the moving family called the solar system, comprising the earth and moon, the seven other great planets with their satellites, half a thou- sand asteroids, or minor planets, besides numerous comets and meteorites. It has required the lifelong labors of many men of exceptional genius, like New- ton, coupled with centuries of no less praiseworthy if less brilliant accumulations of accurate observations to have given us the full knowledge which we now enjoy of the distances, dimensions, masses and orbits of the solar system.
Although the distance from the sun to the orbit of Neptune is 2,800,000,000 miles, the solar system is
xviii
INTRODUCTION
but a speck in the vast universe of the stars. In the year 1901 there flared up in the constellation Perseus a new star which for a few d&ys rivaled the brightest stars of the heavens in its brilliancy, and then slowly faded away into insignificance. To dwellers on the earth this sight was new in 1901, but in reality that new star was so far away that its sudden burst of light, traveling toward us 186,000 miles a second, had been on the journey since the days of Cromwell, and the star had faded away nearly three centuries ago. At such enormous distances are the stars that, although some of them are believed to be millions of miles in diameter, they present no real disks even in the largest telescopes, so that the details of their sur- faces cannot be examined. Nevertheless, by the powerful aid of the spectroscope, much is known of the chemical constitutions of the stars; and many of them have been shown to belong to revolving sys- tems, the components of which, though separated in some instances by greater distances than is Jupiter from the sun, are separately indistinguishable to the telescope. Still we might despair of knowing much more of the physics of the stars if it were not that the spectroscope shows also that the sun is but one of them close by, and that a large class among the stars is probably in a similar condition as to temperature, and made up of the same chemical substances as the sun.
Both telescopic and spectroscopic observations have shown that the solar system is moving at a rapid
xix
INTRODUCTION
rate towards the constellation Hercules, although no change in the aspect of the heavens sensible to the naked eye would accumulate for an immense period of years. By means of the displacement of the earth in its orbit around the sun, amounting to over 180 millions of miles, semi-annually, it has been possible to obtain with fair accuracy the distances of many of the individual stars, and from these by statistical methods to go further and estimate the average dis- tances of all the stars of a given brightness. These various examples indicate the important place of the sun in stellar investigations; and indeed the study of the sun as a typical star, though quite recently developed, seems bound to throw much more light on the subject of the nature of the universe.
Considering the sun as the fountain of light and heat upon the earth, perhaps the first question which suggests itself is this: How much radiant energy reaches the earth from the sun in a given time? This utilitarian branch of solar investigation has been comparatively neglected. No more striking proof of the neglect need be cited than to say that the fore- most text-book on meteorology, published since 1900, states various determinations of the intensity of solar radiation at the earth's mean distance which range from 1.76 to 4.06 calories per square centi- meter per minute. Of these the author of the text- book prefers one which is confessedly the mean of such divergent numbers as 2.63 and 3.50, one of which numbers was thought by its originator to be
XX
INTRODUCTION
too low, and the other too high! As will be shown later, there can be little question now (1910) that the true value is about 1.95 calories; but how remark- able it is that one of the fundamental constants of Nature should have been uncertain within such wide limits, so late as the beginning of the twentieth cen- tury. Imagine, for analogy, that it had been stated in a standard work on astronomy, published in 1905, that the sun's distance (which is of no greater im- portance than the constant of radiation) might be anything between 80 millions and 200 millions of miles so far as known, and that it was generally sup- posed to be 140 millions!
Some of the more important questions connected with the sun's action as the fountain of light and heat are the following: Is the solar radiation uniform or variable? What losses does it suffer in the earth's atmosphere? Are there changes of transparency in the sun's outer layers sufficient to alter the earth's supply of radiation appreciably? How much solar radiation does the earth reflect, unused, to space? How does the earth's temperature depend on solar radiation and on the transparency of the air? If there should be variations of solar radiation, how great changes of temperature of different stations on the surface of the earth ought to follow, and how long would such responses be delayed? In short, are solar studies applicable to weather prediction? What methods, if any, can be economically used to store and employ the sun's energy for power or heat-
xxi
INTRODUCTION
ing? What influences do changes in the intensity or color of the light falling on different plants produce on their growth and fruitage? May advantageous variations of plants be promoted by the control of their radiation supply? What can be done with solar rays for the promotion of health?
In the pages which follow the sun will be considered in these three aspects : First as the controlling mem- ber of the solar system; second, as an object of in- quiry, interesting in itself, but still more so as the nearest star, and typical of a large class of stars; third, as the fountain of light and- heat, and through them of life on the earth. It is indispensable to any satisfactory understanding of the second and third branches to be familiar with the methods and princi- ples which are now being employed in solar investi- gations. For the convenience of the reader a general account of these is given in Chapter II, which there- fore has to do directly with physics, and only sec- ondarily with the sun, but which forms the ground- work of the chapters relating directly to solar phe- nomena. Illustrated descriptions of some of the instruments used in solar research will also be found in Chapter II, and appropriate references to these descriptions and to the statements of the general relations will be found in the text of subsequent chapters.
To avoid premature discussion, the various solar phenomena will be described first without much attention to their explanation, except as seems nec-
xxii
INTRODUCTION
essary to fix attention upon significant facts. Solar theories are dealt with in Chapter VI. One principal exception to this course is in the frequent noting of applications of KirchhofTs discoveries on the rela- tions of temperature, radiation, and absorption. It may be that future writers on the sun will attribute to anomalous dispersion many of the phenomena here set down as due to absorption and motion. But although anomalous dispersion hypotheses are so strongly advocated by Julius, the writer feels confi- dent that his own preference for the older views is still shared by most students of solar physics.
What, after all, is the sun, and how can we best explain the principal solar phenomena? No doubt many will find the views here advanced heretical, but for the writer the existence of the cloudy photo- sphere, so firmly believed in by most solar observers, seems so highly improbable that he has ventured to advocate the view of a purely gaseous sun. But in doing so it is not Schmidt's refraction theory to which he turns to explain the sharp solar boundary. According to Lord Rayleigh our own atmosphere, if freed from dust, would still scatter light by the action of the gases themselves. Schuster and Natan- son have computed this effect, independently, and both find that the purely gaseous scattering goes far to explain in full the observed weakening of the direct beam of the sun above Mount Wilson for rays which are not selectively absorbed. This weakening amounts to several per cent. If then the gases of
xxiii
INTRODUCTION
the atmosphere of the earth, which extend, in density sufficient to scatter light appreciably, perhaps only fifty miles in altitude, suffice to scatter several per cent, of a beam of light, it seems probable that we can see at the most not more than a very few thou- sand miles into the gaseous body of the sun, which, at the layer producing the Fraunhofer lines, seems to be under several atmospheres of pressure. Admit- ting this, how deep, measured radially, can one see near the sun's edge, where the few thousand miles above mentioned will lie along a line of sight nearly tangent to the sun? It would seem that at the sun's edge a shell of gas of only a few hundred miles in thickness must suffice to fully veil all that lies below. Viewed from the earth this would correspond to a fraction of a second of arc, so that a gaseous sun at 93,000,000 miles away would present an apparently sharp boundary. From these considerations depend various consequences adapted to the explanation of solar phenomena. To be sure there are several apparently powerful objections to this view of a purely gaseous sun, but they seem not to be in- superable.
My good friend, Prof. J. C. Kapteyn, has en- couraged me to set down several hypotheses which can be regarded as only slenderly founded. Among these are the hypotheses of the causes of some strange phenomena of geological climates, touched upon in Chapter VI, and more fully discussed in Chapter VII ; the hypotheses of the causes of some peculiar-
xxiv
INTRODUCTION
ities of stellar evolution given in Chapter X; and as some readers may be disposed to think, even the explanation of solar phenomena already mentioned which occupies a large part of Chapter VI. Pro- fessor Kapteyn is of the opinion that a bushel of chaff is worth searching by a Crusoe if it contains some grains of corn that will sprout, and so my defense for my temerity in including such speculations is that they may interest some readers to begin some more fruitful researches.
THE SUN
CHAPTER I
THE SOLAR SYSTEM. THE SUN'S DISTANCE. ITS DIMENSIONS
THE objects which appear to move among the stars, namely the sun, planets, minor planets, moons, me- teors and comets,1 compose the solar system. For- merly it was believed that all the heavenly bodies revolve about the earth. But now the theory of Copernicus is fully verified, and the earth is known to be only a planet, of much smaller size than Jupiter, Saturn, Uranus or Neptune, though larger than Mars, Venus or Mercury, and like the other planets it re- volves about the sun. Galileo was threatened with torture and forced to perjure himself because he be- lieved this, which shows how fortunate we are to live in the present age.
The moon has actually only — — as great a diameter
4UU
as the sun, although they appear to be about equal. It is the immense distance by which the sun and planets are separated from the earth in comparison with the distances we are accustomed to travel over
1 Not all the comets remain permanently attached to the solar system, but many of them do.
1
THE SUN
or even as compared with the distance of the moon, which prevents us from immediately realizing the great bulks of these distant bodies. In the following table is a summary of the approximate dimensions and principal characteristics of the larger members of the solar system. The means of determining the sun's distance, dimensions and rotation will be given later.
GRAVITATION
Accustomed as we are to regard inches and feet as ordinary, miles as considerable, and thousands of miles as very great distances, it may seem almost in- credible that there should be any bond between the the sun and Neptune, situated as they are 2,800,000,- 000 miles apart. There is, however, a bond between them so strong that it would require the strength of a bar of steel 500 miles in diameter to take its place in preventing the escape of Neptune from the sun. This bond we call gravitation. Every body in the universe is believed to attract every other body in the universe with a force proportional to the mass or quantity of matter the body contains, and inversely proportional to the square of the distance between their centers of gravity. On the one hand, this law of gravitation applies between all bodies on the earth as well as between the earth itself and any one of them; and, on the other hand, there is evidence that it holds also among the fixed stars. The weight of a stone is the measure of the attraction between it and
2
THE SOLAR SYSTEM
Dens Wate
^ 1-1 O O rH
COff-
CO O t^ O
^-' i"H O O O CO
<NOOOi-iOOOOcOO»O
Axi
rotati
Orbital revolutio
Inclination of orbit to ecliptic.
aa
"O
^i °°
a a a a 3
r~ e- CO b- O -* Jl
i^-d »o co toi-H t^
00
(M CO (M (M (
O t^
T-I s^. e^- <N
o O O t-( CO CO
an d n mi
nt an meter
§O 00 O i— i l^» O5
rco"t>^t^
O O"J
t^ CO -^ »O >O rH 00 t^ OC' «
' O *O ' CO O
i>» O^ ^O CO C4 O
111
§°9
CT> >O 00 00 •
i-l 't1 O5 • CO O
THE SUN
the earth, and with a sufficiently sensitive apparatus the attraction of two stones for each other may be clearly shown. If weighed with a sufficiently deli- cate spring balance, a weight will be found lighter at the top of a mountain than at its base, in the ratio of the square of the distances of the earth's center at the two places of observation. There are chemical balances so delicate that an object appears to weigh differently accordingly as the weights are placed side by side or one on the top of the other, and this is because of the difference in distance from the weights to the earth's center in the two cases.
The attraction of gravitation between the sun and Neptune amounts to 8 x 1016 (8 followed by sixteen ciphers) tons. If there was nothing opposed to this force, Neptune obviously would fall into the sun. It is the motion of the planet in its orbit, at right angles to the line leading towards the sun, which maintains the distance between them.
Kepler's laws of planetary motion are as follows:
I. The orbit of each planet is an ellipse, with the sun in one of its foci.
II. The radius vector of each planet describes equal areas in equal times.
III. The squares of the periods of revolution of the planets are proportional to the cubes of their mean distances from the sun.
There is but little difference between the major and minor axes of the elliptical orbits of most of the plan- ets, or, in other words, their orbits are nearly circular.
4
THE SOLAR SYSTEM
But this is not true of the orbits of Mercury and Mars, as is shown in Table 1.
It may be surprising to some readers that Kepler could have possessed so much knowledge of the dis- tances of the planets from the sun as to enable him to verify the third law, while as yet the actual distances in miles were not even roughly known. But only the ratios of the distances of the planets were re- quired by Kepler, and these could be fixed inde- pendently of the actual distances by the following method, known since the days of Hipparchus, which I quote from Young's "The Sun."
"First, observe the date when the planet comes to its opposition, i.e., when sun, earth, and planet are in line, as in the fig- ure, where the planet and earth are repre- sented by M and E. Next, after a known number of days, say one hundred, when the
planet has advanced to M'and the earth to E', observe the planet's elongation from the sun, i.e., the angle M'E'S. Now, since we know the periodic times of both the earth and planet, we shall know both the angle MSM' moved over by the planet in one hun- dred days, and also ESE', described in the same time by the earth. The difference is M'SE, often called the synodic angle. We have, therefore, in the tri- angle M'SE', the angle at E' measured, and the angle
THE SUN
M'SE' known as stated above, and hence by the ordinary processes of trigonometry we can find the relative values of its three sides. "
Thus, by means of comparatively simple astronom- ical observations, all the relative distances in the solar system can be fixed with high accuracy. It is a work of far greater difficulty, as we shall see, to meas- ure any of these distances absolutely.
Kepler's three laws were known before the year 1620, but without explanation. Sir Isaac Newton discovered, about the year 1679, that all three laws are direct consequences of the laws of motion, pro- viding it is assumed that all bodies attract one an- other with a force varying inversely as the square of the distance. This latter principle is Newton's law of gravitation.
At the present day no well-informed person ques- tions either the Copernican system or the universal sway of gravitation ; for, while not every such person has the mathematical knowledge requisite to examine all the proofs of these fundamental facts, he yet feels entire faith in the conclusions unanimously agreed upon by such masters as Kepler, Newton, Laplace, and many others of scarcely less renown, who have overcome the tremendous mathematical difficulties in which the knowledge of the motions of the solar system is involved. Every planet and satellite at- tracts every other, and perturbs its motion from the simple orbit which would exist if there were only two bodies concerned. In a large astronomical library
6
THE SOLAR SYSTEM
one may find printed in a quarto or folio volume the final equation representing the motion, for instance, of the moon. Such an equation comprises line upon line and page after page, including thousands of terms required to account for all the disturbing fac- tors. None but a master can handle such a problem.
Prof. E. W. Brown, writing in 1904 1 of his inves- tigation of the theory of the moon's motion, says:
" A few brief details about the amount of time and labour expended may not be uninteresting. From 1890 to 1895 certain classes of inequalities were cal- culated, but the work was only begun on a systematic plan, which involved a fresh computation of all the inequalities previously found, at the beginning of 1896. Mr. Sterner began work for me in the autumn of 1897 and finished it in the spring of the present year, though neither of us was by any means contin- uously engaged in calculation during that period. He spent on it, according to a carefully kept record, nearly three thousand hours, and I estimate my share as some five or six thousand hours, so that the calcu- lations have probably occupied altogether about eight or nine thousand hours. There were about 13,000 multiplications of series made, containing some 400,- 000 separate products; the whole of the work re- quired the writing of between four and five millions of digits and plus and minus signs. Although the problem now completed constitutes by far the longer
1 Royal Ast. Soc. Monthly Notices, vol. Ixv, p. 107, 1904. 3 7
THE SUN
part of the whole, much remains to be done before it is advisable to proceed to the construction of tables."
Every large scientific institution or observatory has almost daily communications from persons of very moderate attainments who presume to question, nay rather to spurn, the most well-attested facts of human knowledge. Such persons seem to prefer especially to direct their attacks on the following facts: the Copernican system; the law of universal gravitation; the first and second laws of energy; and, finally, the high temperature of the sun. No argument can refute them, because they have not the requisite learning to comprehend it, which is no disgrace, but which should make men modest enough to have faith in those who excel them immeasurably. Hence it is the policy of most scientific institutions to avoid entirely discussions of these subjects with such correspondents.
Professor Newcomb tells, in his " Reminiscences of an Astronomer," of such a critic who called upon him and announced his disbelief in Sir Isaac Newton's theory of gravitation. Professor Newcomb proposed to the skeptic that he jump out of the window and convince himself of the existence of gravitation. Being thus pressed, the visitor stated that he be- lieved that gravitation extended no further than the air, and did not go up to the moon. Professor New- comb asked him if he had ever been there to see, and when his caller answered "No," replied that, until
8
THE SUN'S DISTANCE
one of them could go to the moon and try the experi- ment, he doubted if they could ever agree!
THE SUN'S DISTANCE (1) Geometrical Methods.
Since the ratios of the distances between all the principal members of the solar system can be fixed with great accuracy by ordinary astronomical obser- vations, it suffices to measure accurately in miles or kilometers the distance from the earth to the sun or to any one of the planets, and this fixes the scale of the whole system. The great astronomical unit is the mean distance from the earth to the sun, and many determinations of it have been made in the last 250 years. Still astronomers are not quite content, although there is no doubt that we know the distance
to within part of itself. To avoid using many
J-UUU
figures, it is customary to speak of the sun's " paral- lax" instead of its distance. The parallax is the angle which the earth's equatorial radius would sub- tend if viewed from the center of the sun at the mean solar distance. This angle is nearly 8.80 seconds of arc, or about 0.000044 in circular measure. In other words, the earth's mean radius is 0.000044 of the sun's mean distance, and the latter is about 92,900,- 000 miles.
Since the solar parallax is so small, the usual method of surveyors for finding the distance of an inaccessible object is not applicable here. For, if
9
THE SUN
the earth's radius of 4,000 miles subtends only 8.8", no two stations on the earth's surface which can be seen from each other could possibly be far enough apart to serve as a suitable base line for a solar triangulation. Fortunately the observer can avail himself of the fixed stars in the investigation. These may be regarded as an infinitely great distance away, and an apparent displacement of objects in the solar
FIG. 2.
system among the stars may be observed from two stations at opposite sides of the earth, or at the same station by two observations several hours apart. The following explanation of this parallax method is quoted from Young's "The Sun."
"Fig. 2 illustrates the method of observation. Suppose two observers, situated, one near the north pole of the earth, the other near the south. Looking
10
THE SUN'S DISTANCE
at the planet, the northern observer will see it at N (in the upper figure), while the other will see it at S, farther north in the sky. If the northern observer sees it as at A (in the lower part of the figure), the southern will at the same time see it as at B; and, by measuring carefully at each station the apparent distance of the planet from several of the little stars (a, 6, c) which appear in the field of view, the amount of the displacement can be accurately ascertained. The figure is drawn to scale. The circle E being taken to represent the size of the earth as seen from Mars when nearest us, the black disk represents the apparent size of the planet on the same scale, and the distance between the points N and S, in either figure A or B, represents, on the same scale, also, the dis- placement which would be produced in the planet's position by the transference of the observer from Washington to Santiago, or vice versa."
Dr. Gill, lately the Astronomer Royal at the Cape of Good Hope, has made many highly accurate meas- urements by this method. He observed the oppo- sition of Mars in 1877 at Ascension Island, employing for his measurements a heliometer loaned by Lord Lindsay. This instrument is a telescope having its lens cut in halves, and having a micrometer screw for sliding the parts with reference to each other, thus enabling the observer to make the images of two stars formed by the two halves to coincide. This device is the most accurate one known for measuring small angular displacements between stars. Dr. Gill
11
THE SUN
determined the displacement of Mars among the stars as measured by evening and morning observa- tions, and continued the work for several weeks. From these measurements he obtained 8.780"±0.020" for the sun's parallax.
Several of the minor planets or asteroids, though at greater distances from the earth, have proved more favorable objects for these measurements than Mars. Being smaller and not so highly colored, it appears that more accurate measurements of their projections among the stars can be made. In 1889 and 1890 a concerted system of observations was made upon the asteroids Victoria, Iris, and Sappho by Dr. Gill, Dr. Elkin of the Yale College Observatory, and several German observers. Their results range from 8.796" to 8.825", and their mean is 8.807" ±0.006".
The discovery of the minor planet Eros, in 1898, furnished an object so much more favorable than any other for parallax determinations by this method that a great parallax campaign was lately carried through upon this asteroid by many of the leading observa- tories. Undoubtedly a still more thorough one will be undertaken in 1931. Eros has a very elliptical orbit; so much so that when nearest the earth in the most favorable oppositions its distance is only 13,- 500,000 miles, and its parallax then becomes as great as 60"; while at its most unfavorable oppositions its nearest distance is 74,000,000 miles and its parallax only 11". In the opposition of 1900-1, its nearest approach was 30,000,000 miles, but in 1931 it will be
12
THE SUN'S DISTANCE
within half that distance. Prof. Arthur Hinks has lately completed and published the reduction of the photographic measurements of the 1900-1 in- ternational Eros campaign, and he obtains the solar parallax as 8.807" ±0.0027".
The method just described for fixing the scale of the solar system is generally believed to be the best of all at present. But there are several other methods which deserve mention, and first, on account of, its historical interest, the method of the transit of Venus. This planet as a dark spot passes occasionally be- tween us and the sun's disk. The transits occur in pairs, about eight years apart, and the pairs occur only about once a century. Those of June, 1761 and 1769, and of December, 1874 and 1882, were all ob- served with great attention by astronomers of sev- eral nations, for they were regarded, until very recently, as yielding by far the best means of fixing the solar parallax. All the methods used depend, of course, on the displacement of the planet on the sun's disk, as viewed from opposite sides of the earth.
In Weld's " History of the Royal Society of Great Britain," may be found several quaint items relating to the transits of 1761 and 1769. We, of America, sometimes get the impression in our early school days that King George III was only a crazy old despot, and it will be a satisfaction to many of us to know that he was a very liberal patron of the best scientific enterprises. At the request of the Royal Society he ordered £1,800 to be appropriated for the observa-
13
THE SUN
tion of the transit of 1761, and in addition, the Ad- miralty directed a ship of war, the Sea-Horse, to convey the observers to Bencoolen in India. The Rev. Nevil Maskelyne, afterwards Astronomer Royal, was sent to observe at Saint Helena.
The two observers sent to Bencoolen were Mason and Dixon, the names so famous in American history on account of their survey of " Mason and Dixon's Line," which afterwards led to the popular name of " Dixie" for the South-land. Their ship, the Sea- Horse, engaged a French frigate almost at the shores of England. The stands for instruments were dam- aged by shots, and the observers could hardly be in- duced to reundertake the journey. On account of the delay thus occasioned, they observed at the Cape of Good Hope, instead of proceeding to India.
For the transit of 1769 the King, at the Memorial of the Royal Society, provided even irfore liberally. He ordered £4,000 clear of fees to be paid over, and that any balance which might be unexpended should be for the use of the Society. In addition, the ship Endeavor, under the command of Lieutenant, after- wards the famous Captain, James Cook, was ordered to the Pacific Ocean to take part in the observations. Cook observed the transit successfully at what is now called Venus Point, on the island of Tahiti. The Royal Society sent observers also to Hudson's Bay and to India on this occasion.
In 1874 and 1882 very elaborate preparations were made by the governments and private astronomers of
14
THE SUN'S DISTANCE
many countries, including our own. Observations were made all over the world and with many kinds of apparatus, including heliometers, micrometers, and photographic apparatus. Many thousands of ob- servations were made.
The results of the several transits of Venus are on the whole disappointing. A general discussion of the observations of 1761 and 1769 was made by Encke in 1822, and he found the solar parallax 8.5776". More recent recomputations have shown that the transit of 1769 may be said to indicate a parallax of between 8.7" and 8.9". From the transits of 1874 and 1882 different astronomers have computed widely different results, ranging from 8.89" down to 8.75". Newcomb adopts 8.794"0 ± .022".
(2) The Gravitation Methods.
Thus far we have considered only geometrical methods for determining the parallax, and now we may notice another class of quite a different charac- ter— the gravitational methods, so called, which de- pend on noting the perturbing effects of the different planets and satellites on one another. One of the best of them depends on observations of the motion of the moon. It was Hansen's studies of the moon's parallactic inequality which led him to announce, in 1854, the inadmissibility of Encke's value, 8.5776", for the solar parallax. The perturbation of the moon's orbit by the sun causes the interval from new moon to first quarter to be about eight minutes longer
15
THE SUN
than that from the quarter to full moon. This in- equality depends on the ratio of the radii of the orbits of the moon and the earth. Hence, as the moon's dis- tance is known, the solar parallax could be determined if the inequality could be exactly measured. New- comb gives 8.794" as the most probable result thus obtained, but Prof. E. W. Brown's recent able investi- gation of the moon's motion leads to the value 8.778".
Another gravitational method, proposed by Lev- erier, has the advantage of cumulative increase of accuracy. It depends on the secular perturbations of the orbits of the planets, especially of Venus and the minor planets, by the earth, which cause motions of their nodes and periphelia. As time goes on, the displacements thus caused are continually additive, and will eventually be so large as to be determined with very high accuracy. Leverier, indeed, thought it not worth while to observe the solar parallax by other methods, since this must eventually lead them all. Newcomb gives, as the most probable mean re- sult obtained by Leverier's method, 8.768". A fav- orable application of Leverier's method may be made in the case of Eros. G. Witt has found, thus, the ratio of the sun's mass to that of the earth and moon combined as 328,882 ±982. From this he computed a solar parallax of 8.794" ±0.009". Great improve- ment in the accuracy of this Eros result will come after the close opposition of 1931.
It is noticeable that the mean of the results obtained by the various gravitational parallax
16
THE SUN'S DISTANCE
methods falls below that obtained by the purely geometrical methods used in the minor planet cam- paigns. Most astronomers would attribute this to the lesser accuracy of the gravitational methods at pres- ent. It is certain, however, that the geometrical minor planet method tends to give too high results, owing to the difference of atmospheric refraction be- tween a minor planet, Eros, for instance, and the comparison stars; for the minor planets shine by reflected sunlight, and their light cannot be as rich relatively in the blue end of the spectrum as sunlight itself is, because of the smaller reflecting power of nearly all solid substances for blue than for red light. On the other hand, the light of most stars is rela- tively richer in the blue end of the spectrum than is that of the sun. It is probable, therefore, that, in the mean, the comparison stars for Eros, for instance, are bluer than the sun, while Eros itself is redder than the sun. Now the point of the method lies in deter- mining the apparent displacement of Eros at two stations far apart on the earth's surface, or, still better, by morning and evening observations at the same station. In the last-mentioned method, when Eros is low in the east its altitude above the horizon will be increased by atmospheric refraction, but the comparison stars, being bluer, will be raised more than Eros. Similarly in the west. The effect is to make the parallax of Eros, and hence that of the sun, too large, no matter whether the observations are made simultaneously by two distant observatories,
17
THE SUN
or by one observatory in the morning and evening. It is not yet determined how considerable this error is, but it should be investigated with care.
(3) Dependence of the Sun's Distance on Geodesy.
In all the methods thus far referred to, the solar parallax is obtained before the sun's distance, and the astronomer requires of the geodetic surveyor to tell him the dimensions of the earth, if he wishes to pass from the parallax to the actual solar distance. Accu- rate measurements of the earth require pendulum observations at many stations to fix the earth's shape, and, besides this, they require the actual measure- ments by triangulation of very long arcs of the earth's surface, and these depend finally on measurements of a base line of a few miles in length. Base lines are measured by repeatedly setting end to end, under microscopic observation and on specially leveled supports, short bars or tapes, whose temperature must be observed throughout the process. Such measurements of base lines are now made with an error of less than one part in a million. The final result from the whole net of triangulation recently completed across the United States by the Coast and Geodetic Survey is thought to be accurate to within eighty-five feet in a total distance of nearly 3,000 miles. Several determinations of the earth's dimen- sions have been made within fifty years. They give the earth's mean equatorial radius as 3963.1 miles, with a probable error less than one part in 20,000.
18
THE SUN'S DISTANCE
(4) Velocity of Light Method.
We now come to an important class of observations depending on the velocity of light, and called the " physical parallax methods," by which we may find the sun's distance directly. Several ways have been employed for measuring the velocity of light, but the two best are the toothed-wheel method of Fizeau and the revolving mirror method of Foucault. In Fiz- eau's method (see Fig. 3) a beam of light starts from a source at L and, after passing through the lens A and being reflected by the thinly silvered plane glass
D
FIG. 3.
plate B, comes to focus and passes between two teeth of a wheel at F. Thence the ray goes on to the lens C and, after traveling a great distance, is focused by the lens D upon the mirror E, which returns it on its course, so that at length it passes again between two teeth at F, and a part of it comes to the observer at H. Now imagine the wheel in rapid rotation. The light will then be cut off by every tooth which passes F, and will thus consist of a series of flashes. But, owing to the persistence of vision, the beam will still seem continuous if observed at H, though it will be weakened because it shines only intermittently.
19
THE SUN
Time is required, however, for the light to pass from F to E and back to F, and meanwhile the tooth next F has advanced, and may be in such a position as exactly to cut off the returning beam, so that the eye at H will see nothing. By gradually increasing the speed of the wheel the light is alternately cut off and transmitted. By counting these changes from light to darkness, and knowing the number of teeth and the speed of the wheel and the distance FE, the velocity of light is measured.
The method of Foucault is illustrated by Fig. 4. Light from the slit S passes through the thinly sil-
B A P
FIG. 4.
vered glass plate P, thence through the lens A, and is reflected by a plane mirror B to the concave mirror C. The radius of curvature of the mirror C is equal to BC, so that the light is returned to B in the same path that it traversed in going, and thus it again passes through the lens A, and a part is reflected by the sil- vered glass P and is observed at 0. If the mirror B is revolved slowly, the light comes out at 0 as a series of flashes. These become sensibly continu- ous to the eye as the speed increases, but when the speed becomes high the image at 0 becomes dis-
20
THE SUN'S DISTANCE
placed owing to the motion of rotation of the mirror B while the light is passing from B to C and back to B. From the amount of displacement, the speed of the mirror, and the distance BC, the velocity of light is computed.
According to the electromagnetic theory of light, the ratio between the electrostatic and the electro- magnetic systems of electrical units should also give the velocity of light. Furthermore, the electric waves used in wireless telegraphy should proceed with the velocity of light. Results depending on these last two considerations, though interesting, cannot rival in accuracy the velocity directly determined. The following are among the best results:
OBSERVER. |
Method. |
VELOCITY OF LIGHT IN VACUO |
|
Kilometers per second. |
Miles per second. |
||
Mean of Michelson, New- ) comb and others \ Mean of Cornu and Per- [ rotin f Various observers Rosa and Dorsey |
Foucault Fizeau Hertz waves . . Ratio of units . |
299,860 299,890 299,130 299,710 |
186,330 186,350 |
Accepted velocity of light . . |
|
299,860 |
186,330 |
The velocity of light just given is probably correct within one part in ten thousand.
There are three ways of employing this quantity to fix the distance of the sun. The first we will men- tion is through the aberration of light. Though light proceeds by wave motions, and not by particles, the
21
THE SUN
idea of aberration may be understood by the analogy of raindrops. If rain is falling vertically, and a man stands still, his hat screens his face. But if he move rapidly forward in any direction, the rain strikes his face, thus appearing not to come vertically but at an angle thereto. So with the light of the stars: Owing to the motion of the earth in its orbit the stars are apparently displaced when the earth is moving at right angles to the line of sight; the displacement being in one direction at one time of the year and in the opposite direction when the earth's motion is re- versed six months afterwards. Owing to aberration, stars at the poles of the ecliptic describe little circles about 41" in diameter, and those in the plane of the ecliptic merely oscillate in a straight line about 41 " long. There is an uncertainty of a few hundredths of a second as to the " constant of aberration," as astronomers call the radius of the circle of aberration. The Paris conference of 1896 adopted the value 20.47". But there is now much evidence tending to recommend a higher value. The long-continued ob- servations- of Doolittle made with instruments of different kinds seem to require us to set the constant of aberration as high as 20.51", perhaps even 20.53". The following table shows the relation of aberration and solar parallax values :
Aberration Constant. |
20.46" |
20.47" |
20.48" |
20.49" |
20.50" |
20.51" |
20.52" |
20.53" |
Solar Parallax. |
8.807" |
8.803" |
8.799" |
8.794" |
8.790" |
8.786" |
8.781" |
8.777" |
THE SUN'S DISTANCE
Another way to use the velocity of light for fixing the sun's distance is through the observations of Jupiter's satellites. Olaf Romer called attention to this method by taking the problem the other way about and computing the velocity of light from the supposed known distance of the sun. The satellites pass behind the planet and are eclipsed frequently. These eclipses occur nearly 1,000 seconds later in time when Jupiter is in conjunction than when in op- position, for there is a difference of distance amount- ing to the whole diameter of the earth's orbit for light to traverse in the two cases. Unfortunately, the eclipses are not sudden phenomena, so that it requires careful photometric work to fix the " light equation," as the time required for light to travel the radius of the earth's orbit is called. According to many years of observation at the Harvard College Obserr vatory, as reduced by Professor Samson of Durham, the light equation is 498.64 seconds. From this the sun's parallax comes out 8.799".
A third way of employing the velocity of light is through what is known . as the Doppler effect. Just as a locomotive whistle is higher in pitch when the train approaches, so the light of a star is bluer in hue when we are approaching the star in the earth's orbit. The velocity of the earth is so small compared to the velocity of light that the magnitude of the change can be measured only with a powerful spec- troscope of special design. Nevertheless, it seems possible that solar parallax determinations by this 4 23
THE SUN
method may before long compare favorably in accuracy with any others. For parallax purposes, Kiistner, and lately Halm, have photographed the spectra of bright stars in comparison with spark spectra photographed above and below on the same plate. They repeated these comparisons at inter- vals of about six months for several years, arid, after applying necessary corrections, they have determined the velocity of the earth in its orbit, and from this the sun's parallax. Their values are not far from those obtained from other methods, but are not quite accurate enough to compete with them.
It would seem more promising to determine the relative velocities of the planets and the earth, by photographing simultaneously the spectra of Venus and Mars, or of Mars and the moon at a favorable time. Suppose, for instance, that two large cce- lostats (see Chapter II) were arranged one beside the other to reflect the light of Mars and Venus simul- taneously upon a single long-focus concave mirror, and the two images were reflected together by suit- able devices so as to fall at once one above the other on the slit of a powerful spectroscope. By the use of a rotating sector one image could be made equal to the other in brightness, and the spectra of both could then be photographed absolutely simultane- ously, one above the other on the same plate. Errors such as displacements by change of tempera- ture of the spectroscope would effect both spectra alike. The ccelostats should be used alternately
24
THE SUN'S DISTANCE
on the two objects, so as to follow out Professor Turner's excellent motto of "reversing everything that can be reversed." From preliminary trials made at Mount Wilson it seemed to Mr. Adams and the writer that it would be practicable to photo- graph the two spectra on a scale ^ as large as that which Mr. Adams employed to determine the solar rotation spectroscopically. As the spectra of the planets are similar, being solar spectra slightly altered by selective reflection, there would be numerous good lines to measure. There would evidently be no necessity of introducing a terres- trial comparison spectrum. It seems probable that by this method the solar parallax could be determi- nable to about one part in 2000. However, it has not yet been tried.
(5) Summary.
Excluding parallax values not of the highest weight, we have the following mean results:
From heliometer work on minor planets 8 . 807'
From the Eros campaign 8 . 807'
From all gravitational methods 8 . 780'
From eclipses of Jupiter's satellites1 8 . 799'
From the constant of aberration of light (assumed 20 . 51") ... 8 . 786'
If we take the mean of all these results as they stand, we, in effect, give double weight to the geometrical and velocity of light methods as compared < with the
1 This determination is included, notwithstanding the large probable error Professor Samson assigns to it, because in the author's opinon, as indicated above, the constant of aberration and the minor planet parallaxes are uncertain to nearly the same degree, and the value of an independent method is very great.
25
THE SUN
gravitational method. This seems justified, and by doing so we reach the probable value of the solar parallax as
8.796".
This corresponds to a solar distance of
92,930,000 miles or 149,560,000 kilometers.
DIAMETER OF THE SUN
From the heliometer measurements of Schurr and Ambronn the sun's angular diameter, as seen from the earth at mean distance, is 1920.0//±0.03// Other determinations agree very closely with this. Hence, the sun's diameter is
865,000 miles or 1,392,000 kilometers. Poor has lately maintained that observations indi- cate that the sun's equatorial and polar diameters vary relatively as much as 0.1" during a sun-spot cycle of eleven years. According to him, the equatorial diameter is the larger at sun-spot maxi- mum, and the polar diameter the larger at minimum. Ambronn, however, denies that this is supported by the observations, and Moulton opposes so large a variation on theoretical grounds.
THE SUN'S MASS
The mass of the sun relatively to a planet which has a satellite may be obtained in several ways. One of them, as applied to the earth, is as follows: Let M be the mass of the sun, earth and moon
26
THE SUN'S DIMENSIONS
combined; and m, that of the earth and moon; let R be the mean distance between the centers of the sun and the earth; r, the mean distance between centers of the earth and moon ; let T be the number of days in a sidereal year, and t the number in a sidereal month. Then, by Kepler's law:
R3 r3 R3 r3 r3
M:m=— :-; whence M-ra : ra = — --:—.
The mass of the moon compared with the earth is known from other data to be 1/81.53. Small cor- rections to the periods T and t due to the perturba- tions are also known. Applying these corrections, the ratio of the masses of the sun and earth comes out. It is, according to Newcomb, for a parallax of 8.796":
332,800.
THE EARTH'S MASS
Astronomy must be displaced by physics if we would proceed further and obtain the mass of the sun in ordinary units; for the mass of the earth is then required. This is determined by comparing the at- traction of the earth for a body (i.e., the weight of the body) with the attraction which a known mass pro- duces when acting upon the body from a known dis- tance. During the eighteenth century attempts were made to compare the attraction of a mountain with that of the earth. The most celebrated of these was performed in 1775, under the auspices of the Royal Society, by the Astronomer Royal Maskelyne at the
27
THE SUN
mountain Schehallien in Scotland. Owing to the im- possibility of determining accurately the center of gravity and density of a mountain, this method, though suggested by Newton, and one very interest- ing to contemplate, is of little value. What is known as the method of Cavendish, though proposed by Michell, is regarded as best. In this method a pair of small balls are hung at the ends of a rod which is supported in the center by a fine wire or fiber. Two large masses are placed in a position to twist the sus- pending fiber by attracting the small balls. The force of attraction is measured by the torsion of the fiber, and this is determined by the period of vibra- tion of the system. C. V. Boys performed a notable piece of work by this method in 1894, and for it he invented the quartz fiber, without which some of the most delicate and interesting of modern physical work in other lines would be impossible. His first method of making quartz fibers was very picturesque. One bit of a quartz crystal being fastened to a little arrow, and another to a bow, the two bits are fused together by a blowpipe flame; and when the quartz is properly melted the arrow is shot out, trailing be- hind it a thread of quartz almost too fine to be seen. Such fibers are almost perfectly elastic, and as strong as steel in proportion to their size. Boys obtained for the mean density of the earth, 5.527. In other words, the earth has five and a half times as great mass as it would have if composed entirely of water. The principle of a third method, more simple to im-
28
THE SUN'S DIMENSIONS
derstand than Cavendish's, is illustrated in Fig. 5. Two equal balls, A and B, are suspended from the beam E of an equal arm balance, and two large equal balls are arranged so as to be in the positions C and D, or C' and D', at pleasure. In the first position they tend by their at- tractions to make A overbalance B, and in the second position, the opposite. Hence the mere weighing of B against A by means of a rider on the beam E is the principal require- ment in addition to knowing the masses of the balls, C and D, and their distances from A and B. By this means Richarz and Krigar-Menzel determined the earth's mean density to be 5.505. Burgess has discussed the different determina- tions, and gives the most probable value of the earth's density as 5.5247 ± 0.0013. Corresponding to this, the constant of gravitation (see the begin-
D'
FIG. 5.
ning of this chapter) is 666.07 X 10~10
cm.
- (or
gr. sec.
dynes). From these figures the mass of the earth is 1.317 X 1025 pounds, or 5.984 X 1024 kilograms, and that of the sun is 4.38 X 1030 pounds, or 1.990 X 1030 kilograms (that is 4, or 1, followed by 30 places of zeros).
29
THE SUN
THE SUN'S DENSITY
Since the volume of the sun is 1,306,000 times that of the earth, the density of the sun is only 0.255 as great as that of the earth, and is 1.41 as compared with water. A most interesting and important con- clusion follows from these figures on the density of the sun. Notwithstanding that its density is so small, we know from its spectrum that the sun has many of the heavy metals and other chemical ele- ments found upon the earth, and we presume that it includes few elements or compounds which in a liquid or solid state would be of less density than water. Water and other common liquids can not exist even as vapors on the sun, owing to the high temperature. In view of these facts, it follows that the sun is prob- ably mainly gaseous. Owing to the enormous mass of the sun, the attraction of gravitation at its surface is 27.6 times as great as that at the surface of the earth, so that a body which weighed a hundred pounds here would weigh over a ton there. Hence, the gases of the interior of the sun must be tremendously com- pressed, so that probably in their appearance they would resemble liquids, though still having the prop- erty of indefinite expansibility characteristic of gases.
CHAPTER II
THE INSTRUMENTS AND METHODS USED IN SOLAR INVESTIGATION
The Telescope. — The Coelostat. — The Spectrum and what it In- dicates.— 'Spectroscopes. — The Spectroheliograph. — The Helio- micrometer. — The Comparator. — The Nature of Radiation. — Laws of Radiation. — Spectra of Different Sources. — Pyrheliom- etry. — Bolometry.
FOR a long time the telescope and the observer's eye were the principal means of advancing solar in- vestigation, but in the last half century a number of other less familiar instruments and physical princi- ples have been employed, which require some ex- planation.
THE TELESCOPE
A few words may be said first as to the methods of employing the telescope. The sun is far too bright to view for any length of time with the naked eye, much less with the telescope, unless means are used for reducing the brightness. It is said that the Bel- gian physicist, Plateau, having looked steadily at the sun twenty seconds for the purpose of studying the after images which would be produced, lost his sight permanently in consequence.
To get a rough general view of the sun a screen is often used in the manner shown in Fig. 6. The dis-
31
THE SUN
tance of the screen from the eyepiece depends on the size of the image desired and the power of the eyepiece. By moving the eyepiece to and fro in the draw tube, a sharp image is readily obtained. It is well to put a screen on the front of the tele- scope, as shown in the figure, to cut off undesired light. For Carrington's method of determining the exact locations of sun spots on the disk, see Monthly
FIG. 6.
Notices of the Royal Astronomical Society, vol. xiv, p. 153.
Observations of the finer details of the sun cannot be made with a screen, and there are several ways of protecting the eye for direct telescopic vision. This may be done by reducing the aperture of the telescope objective with a suitable diaphragm and placing a dark glass in front of the eyepiece, but at
SOLAR INVESTIGATION
great cost of definition if the diaphragm is too small or the shade glass not perfect. A reflecting tele- scope, if its object mirror is left unsilvered, may re- quire only a shade glass for visual work with the sun; and, on the other hand, the objective lens of a re- fractor may be thinly silvered to cut down the light. But both these expedients unfit the telescope for other purposes. There are several special solar eyepieces which have been devised. Fig. 7 shows Sir John Her- schel's. The light en- tering at 0 encounters a prism of glass whose first surface is placed at an angle of 45°. More than 90 per cent of the light passes through the prism and goes out through the open end of the tube, while the reflected light goes up through the eyepiece AB. A shade glass is still necessary, but need not be very dark. It is advantageous to employ a long thin wedge of dark glass ("London Smoke," for instance) compensated by a corresponding wedge of ordinary glass, as shown in Fig. 8.
With this arrangement the image is undistorted, uncolored, and may be made of exactly the proper brightness for observing. The polarizing eyepieces on the general plan shown in Fig. 9 are more con-
33
FIG. 7.
THE SUN
venient, but also more expensive. The light is re- duced merely by rotating the upper case in its
^^^^•^•^^••Bm^l^HH bearmg m the
lower. The image is seen in its proper color and without being either inverted or reversed from right to left
FIG. 8. ^,
by the eyepiece.
When very small objects are being examined, it is sometimes advantageous to use Dawes's device of limiting the field of view by means of a minute diaphragm made by piercing a card or plate of ivory with a hot needle.
For photographic work the extreme brightness of the sun is advantageous in- stead of troublesome, because it enables the observer to employ slow plates which have much finer grain than rapid ones, and also to cut down the exposure time, which is an im-
SOLAR INVESTIGATION
portant gain, since it is favorable to making com- plete exposures during the occasional instants of exceptionally good atmospheric conditions which favor superior optical definition or "good seeing." As all solar observers know, the atmospheric effect called " boiling" is generally much worse in the day- time than in the night, and undoubtedly because the powerful heating of the surface of the ground by the solar rays causes rising air currents of un- equal density, which drift hither and thither across the line of sight. Photographs of the sun are usually ex- posed by means of a sliding shutter similar in action to the form shown in Fig. 10. B is a catch which may be released by the electromagnet, or by hand, thus allowing the spring S to draw the slide* con- taining the slit A swiftly across the opening 0, through which the rays enter the camera. The ex- posure is proportional to the width of the slit A, and is governed by the tension of the spring S. 1 1
FIG. 10.
Exposures of grknrk to 77^; of a second are required
The edges of the sun's
5000 100 according to circumstances.
TIIM SUN
disk HIT not us bright us the center, sn I hat the solar image cannot be properly exposed to show details equally well in all parts in the same picture.
It is often desirable to know the orientation of the solar image. For this purpose a cord or wire may be stretched close to the plate, in some known position, as horizontal, or parallel to the sun's drift, or vertical, and its shadow on the image will serve as a basis of computation. Sometimes in an optical system containing reflectors it is desired to know what parts of the solar image correspond to east and west in the sky. This may always be determined with ease and certainty by stopping the telescope motion and letting the sun's imago drift; for the advancing edge or "limb" of the image must always correspond to the west edge or "limb" of the sun in the sky. The data for computing the position of the sun's equator are published annually in the pamphlet called The Companion to the Observa- tory.
THE COSLOSTAT
Most modern apparatus for solar research is of a complex and necessarily bulky nature, so that it is highly inconvenient to move it. Accordingly, a fixed beam of sunlight is almost a necessity. There are several kinds of instruments for reflecting the light of heavenly bodies in a fixed direction, called heliostats or siderostats; but these all rotate the image of the heavenly object if an image is formed. This is usu- ally a great disadvantage, and, fortunately, there is
36
PLATE I
SMITHSONIAN OBSERVING SHELTER AND COXOSTAT, MT. WILSON.
SOLAR INVESTIGATION
one very simple instrument for the purpose, called the coelostat, which does not rotate the image. In its simplest form the ccelostat is a single plane mirror mounted on an axis parallel to the axis of the earth, and rotated by clockwork at the rate of one complete rotation in forty-eight hours. In this form the sun's beam is reflected in a different direction at different times of the year, according to the position of the sun north or south of the celestial equator. Even in a single day the mirror cannot be used to throw a hori- zontal beam in any desired direction, but only in two, nearly east and west respectively, the first favorable for morning hours, the other for afternoon hours. These limitations are overcome by the introduction of a second plane mirror, south of and above the level of the first, on which the beam is first reflected, and from which it can be sent in any desired direction, but preferably (for the northern hemisphere) towards the north. It is necessary to provide longitudinal and cross motions for one or the other of the mirrors, to accommodate the change in declination of the sun at different times of the year. Plate I shows the fifteen-inch coelostat of the Smithsonian Astrophysi- cal Observatory at Mount Wilson, Cal. The first or rotating mirror is provided with means of moving it on tracks both east and west, and also north and south. The second or fixed mirror reflects the beam horizontally northward to the spectroscope within the observatory.
37
THE SUN
THE SPECTRUM AND WHAT IT INDICATES
Since Kirchhoff and Bunsen's great discovery of spectrum analysis in 1859, the spectroscope has be- come more and more indispensable to progress in solar research, so that now the greater portion of our knowledge of the sun is due to this instrument. White light is not a simple but a composite thing, containing potentially all the different colors familiar to the eye, and still other rays which the eye sees not at all. As we shall describe, light can be analyzed so as to pre- sent to the eye the colors, and this presentation is usually in the form of a long ribbon of color gradation. When light is thus analyzed to show the colors which are potentially in it, the spectrum is said to be pro- duced.
If sunlight is resolved into a spectrum, under good conditions we see a ribbon of light shading gradually from dull red through brighter and brighter hues to orange, then yellow, next green, then blue, indigo and violet. If we had eyes of unlimited capacity we should see beyond the violet still other rays, and beyond the red yet others, also. We can detect such invisible rays by the heat they produce or by photography, but just as the ear cannot hear sounds above a certain pitch, or below a certain other pitch, the eye is limited as to its recognition of radiation. Rays lying beyond the violet end of the visible spec- trum are called " ultra- violet" and those beyond the red are called " infra-red." In the visible spec-
38
SOLAR INVESTIGATION
trum the shading is not perfectly continuous, for there may be seen almost innumerable vacancies of color, or dark lines crossing the colored ribbon at right angles. These dark lines are called, after the name of their discoverer, Fraunhofer lines. It is their presence, and not the beautiful colors, which has been the means of teaching us many things about the sun and stars which would have seemed to the contemporaries of the Herschels to be beyond the possibilities of future discovery.
The cause of the dark lines of the spectrum was unknown until Kirchhoff and Bunsen's researches, about 1859, showed that they correspond in position to certain bright lines which form the spectra of me- tallic vapors. For example, if metallic sodium, or any of its compounds like common table salt, is thrust into an alcohol-lamp flame, the spectrum of the flame shows two brilliant yellow lines which agree in place with two prominent dark lines in the yellow part of the solar spectrum. Not only so, but if an incandescent oxy-hydrogen calcium light, whose nat- ural spectrum shows neither bright nor dark lines in the yellow at these places, is caused to shine through an alcohol-lamp flame charged with sodium vapor and placed before the spectroscope, the two dark lines like those in the solar spectrum will appear instead of the two bright lines of the sodium-charged flame it- self. Other chemical elements, also, when heated to vaporization, emit bright spectrum lines, and the vapors of these elements, if placed in a beam of white 5 39
THE SUN
light, absorb the rays they themselves emit. If their own emission is more intense than the emission they absorb from such a transmitted beam, the effect will be brighter lines in a continuously bright spectrum. If their own emission is less intense than the emission they absorb, the resulting spectrum will be crossed by dark lines. The former effect occurs in the spec- tra of certain stars, the latter in that of the sun. As the emission of a vapor falls off rapidly as the temper- ature falls, it is natural to suppose as the cause of the sun's characteristic dark spectrum lines that the metallic vapors of the sun's outer layers, since they are free to loose heat to space, continue cooler than the sun's inner layers, and hence cannot, by their own emission, fully compensate and supply the place of the rays they absorb.
In Fig. 1 of Plate IV a part of the spectrum of the star Procyon is shown, with comparison spectra of iron above and below it. The stellar spectrum shows numerous dark lines for the reasons indicated above ; and many of these correspond closely in position and relative importance (or intensity, as it is called) to bright lines in the iron spectrum. For a reason to be explained below, the stellar lines are all shifted a little towards the violet with respect to the comparison spectrum, but it is evident that iron has left its sign in the star's spectrum as well as in that of the electric spark.
First of all, then, the dark lines of the solar and stellar spectra show what chemical elements are pres-
40
SOLAR INVESTIGATION
ent in the sun and stars. By comparing the -solar spectrum with bright line spectra of pure metals pro- duced in the laboratory, it has been shown that nearly all of the elements found in the sun are also present in the earth. In the second place the lines of the solar spectrum serve as reference marks to en- able us to recognize the effect of certain influences in the sun, such as varying degrees of temperature, of velocity, of pressure, and of magnetism.
As regards temperature: A dark spectrum line generally indicates a cooler vapor in front of a hotter source, and a bright line that no hotter source lies behind. Furthermore, many elements give in the laboratory a large number of bright lines whose relative intensities differ according to the tempera- ture of the source. Similar differences of intensity among the lines of a given element, as found in the solar spectrum, give a basis for estimating differ- ences of temperature there, as, for instance, between a sun spot and the photosphere.
As regards velocity: We have noted the Dop- pler effect already in speaking of methods of meas- uring the sun's distance. It depends on the fact that light travels by waves. Those waves which are visible to the eye, differ in length from 0.0004 millimeter (0.4 ^) to 0.0007 millimeter (0.7 p), cor- responding to violet and dull red respectively. The time period of a complete vibration of a wave of violet light is given by the ratio of its length (.0004 millimeter) to the velocity of light (300,000,000,000
41
THE SUN
millimeters per second) . Hence, 750,000,000,000,000 waves of violet light are emitted from all parts of the sun's surface each second.1 Doppler's principle is as follows: If a star is approaching the earth with a velocity v, the effect is to shorten the length of each wave of light reaching the earth by an amount vt, where t is the period of vibration of the wave. If V is the velocity of light and X the original wave- length, X = Vt. Suppose the apparent wave- length to be X,, then X, = (V- v)t. Hence, (X- X,)
If the lines of a spectrum are displaced towards the violet by amounts which are less at the violet end of the spectrum than at the red in the ratio of the wave lengths, this may be an indication that the source of light is approaching the earth. By com- paring the positions of the spectrum lines at different parts of the edge of the sun's disk, it has become pos- sible to measure the rate of rotation of the sun for all solar latitudes. Similar studies of the shifting of lines in the spectra of the stars in all parts of the heavens indicates towards which of the stars the solar system is approaching in its motion through space. In Fig. 1 of Plate IV the dark stellar iron lines show a displacement towards the violet as com- pared with the bright iron companion lines and thus we find that Procyon was approaching the earth at the time when the observation was made. In Fig.
1 For "additional remarks on this subject, see Chapter VII. 42
SOLAR INVESTIGATION
2 of Plate IV is shown a pair of superposed solar spectra from the eastern and western edges of the sun. The great oxygen band called B gives rise to most of the lines in this spectral region, and as these lines are terrestrial, not solar, they are not displaced in the two spectra. But all the solar lines show a displacement, due to the fact that one edge of the sun is approaching, the other receding.
As regards pressure: The experiments of Hum- phreys, Mohler and Jewell first showed that the spectrum lines of different elements are shifted towards the red by varying amounts, if the pressure of the surroundings of the source is increased. These pressure shifts are usually very minute, and they follow a different law from shifts due to velocity. Thus the examination of the solar spectrum can indicate the range of pressure under which its absorption lines were produced.
As regards magnetism: It was first shown by Zeeman that a powerful magnetic field may split an ordinary single spectrum line into several com- ponents, differing in the character of the polar- ization of their light waves. The polariscopic ex- amination of double, or triple, or merely widened spectrum lines may yield evidence as to the mag- netism in the sun. It is from such study that Hale has discovered the existence of magnetic fields in sun spots. In most cases the lines appear separated into two components when viewed along the lines of force of the magnetic field, and into three com-
43
THE SUN
ponents when viewed at right angles to the lines of force; but sometimes four, or six, or even more lines are seen. In the simpler cases, first mentioned, the doublet seen longitudinally consists of two circularly polarized rays, one polarized left-, the other right- handed. The triplet seen transversely consists of three plane-polarized rays, of which the central one occupies the same position as the line seen in the absence of a magnetic field, and the two side rays occupy the same positions as the two lines seen longitudinally. The plane of polarization of the central component is at right angles to that of the two side components. Hence, the central compo- nent may be extinguished by interposing a Nicol prism in a certain position, while the two side com- ponents may be extinguished and the central one again seen when the Nicol is rotated 90°. In the case of the doublet seen longitudinally, the two rays may be transformed into plane-polarized light by introducing a Fresnel rhomb. After this they may be extinguished alternately by a Nicol prism rotated 90° at a time. In Plate II, Fig. 1 shows a part of the ultra-violet spark spectrum of iron viewed trans- versely to the lines of force of a magnetic field. Figs. 2 and 3 show the effects of interposing the Nicol prism in two positions. Fig. 1 appears exactly the same as the spectrum would if seen along the lines of force without the Nicol prism, although, in fact, the polarization is not the same. Most of the lines in this spectral region are of the ordinary type, but
44
SOLAR INVESTIGATION
one is unaffected, while some are very complex. One even has twelve components; but this will probably not be discerned in the spectrum as reproduced.
Like the straws which show the way the wind blows, or the hieroglyphics which hold the history of ancient times, the lines of the spectrum yield to painstaking, minute examination a wealth of knowl- edge wholly unthought of by the careless, and, at first glance, unknowable. Hardly anything in science is more wonderful than the extent of knowl- edge of the heavenly bodies, situated millions, billions, and trillions of miles away, which has been acquired through the spectrum.
THE SPECTROSCOPE
In common practice two very different pieces of apparatus are employed to produce the spectrum, namely the prism and the grating; but a device con- taining either one of these, in suitable combina- tion with its adjuncts needed to produce a spectrum, is called a spectroscope. To understand something of the action of the prism it is needful to know that light travels with different velocities in different substances, and that in general the different-colored lights travel with different velocities in the same medium. In a vacuum all kinds of light travel with equal velocity, and in most gases, at ordinary pressures, there is little difference for the different colors, but with transparent liquids and solids the difference is very considerable, as is shown in the
45
THE SUN
following little table. The numbers give the values of the ratio
velocity of light in vacuo velocity of light in medium'
TABLE II. — Velocity of light in vacuo, and its ratios to the velocity of light in different media.
Colo |
r |
Violet |
Blue |
Green |
Yellow |
Red |
Wav |
e length1 |
4,200 |
4,800 |
5,400 |
5,900 |
6,500 |
Velo un |
city in vacu- i2 |
299,860 |
299,860 |
299,860 |
299,860 |
299,860 |
Ratio of velocity in vacuum 1 1 to velocity in: |
Air at atmos- pheric pres- sure |
1.000297 |
1.000295 |
1.000294 |
1.000293 |
1.000292 |
Water |
1.3420 |
1.3381 |
1.3354 |
1.3336 |
1.3320 |
|
Average flint glass |
1.6366 |
1.6238 |
1.6157 |
1.6108 |
1.6066 |
|
Carbon bi- sulphide . . . |
1.6835 |
1.6553 |
1.6374 |
1.6273 |
1.6188 |
|
Diamond |
2.4570 |
2.4373 |
2.4242 |
2.4170 |
2.4108 |
In consequence of the difference of velocity of light in different media, a beam of light of a single color is bent when it crosses obliquely the boundary between two media. This is illustrated in the accompanying diagram, Fig. 11. The reader must assume, what is shown by the methods of physical optics, that the direction of propagation of light at each instant is at
'The unit of wave length is the Angstrom, which ia one ten- billionth of a meter.
2 The unit of velocity is the kilometer per second. 46
SOLAR INVESTIGATION
right angles to the light front. A. beam of light, which at a certain instant presents the front ac, begins to enter a denser medium at c. In consequence of the less velocity in the denser medium the light from c moves only to cl} while that from a moves to a\t so that the lower portion of the light front is thereby turned to the direction 6^1. Suc- cessive positions of
the light front are shown at a^Ci, a262C2, et cetera. Suppose that the cross section of the denser medium is triangular in shape, so that the light from a4 begins to proceed more rapidly while that at c4 is still being retarded. This state of affairs is that of the prism of the usual form, and the light front finally emerges at a7c7, proceeding in a different direction from that which it had at first. The difference of direction de- pends on the fractional difference of velocity of the ray in the two media. Since this is greater for violet light than for red, a beam of light containing both colors will be split up by such an instrument, and the violet part will be more bent or deviated from its original direction than the red. Such action of the prism is said to be " refraction, " the difference in direction between the entering and emerging beams is the "deviation," and the difference of direction between the different colors as they emerge is called
47
THE SUN
" dispersion." The angle CcA between the entering ray and the line perpendicular to the face of the prism is called the angle of incidence, and the angle c7cB the angle of refraction. A principal law of refraction is this: The sine of the angle of incidence divided by the sine of the angle of refraction is constant for a ray of a single color entering a given substance, whatever the angle at which it enters. Calling i the angle of incidence, r the angle of refraction, and n the con- stant, or index of refraction:
sin i
n = — — . sinr
The value of the index of refraction is, therefore, a matter of much importance for calculation, and it is still more interesting because it is also the ratio of the velocities of the light in the two media concerned. For yellow light arid for ordinary telescope flint glass and air as the two media, the refractive index is about 1.61 (see Table II).
It follows mathematically from the law just stated, that for a given prism and a given color of light the deviation can never fall below a certain minimum value, whatever the angle of incidence. This small- est angle of deviation is called the angle of minimum deviation, and is secured when the angle of incidence is equal to the angle of emergence. For this position of the prism the following relation holds, if we desig- nate the angles of incidence and deviation as i and D, and the angle at the apex of the prism as A : D=2i-A. 48
SOLAR INVESTIGATION
In the use of prismatic spectroscopes it is generally preferable that the rays of light composing the beam shall be parallel to one another as they enter the prism, and that the prism shall be set for minimum deviation. The beam which emerges when white light passes the prism under these circumstances con- sists of a mixture of bundles of parallel rays of light of different colors, with the neighboring shades differing by almost imperceptible inclinations one from an- other in their paths of emergence. It is necessary to bring them to focus by means of a lens or mirror if they are to be sharply separated. If the light comes originally from a star the rays will be practically par- allel without alteration; but if they come from the sun they converge from opposite sides of the solar disk with an angle of over 30' of arc. For solar work, therefore, and often for stellar work, also, two other adjuncts to the prism are used. The first is a narrow slit between sharp metal jaws parallel to the line of intersection of the prism faces, and the second is a lens or mirror placed at such a distance as to render the rays which diverge from the slit parallel. A lens or mirror serving the latter purpose is called a collimator, and the lens or mirror which focuses the spectrum is called the objective, or image-forming piece. This arrangement is indicated in Fig. 12.
With most prisms the violet part of the spectrum is much more extended than the red, owing to the more rapid proportional variation of the velocity of light in glass at the violet end of the spectrum.
49
THE SUN
The grating, as a means of dispersing light, de- pends on the phenomenon called interference. Light, like sound, is propagated by wave motions. If a tuning fork be set in vibration and slowly rotated while held in the hand at some distance from the ear, the sound will be found to wax and wane in loudness,
FIG. 12
although the fork continues to vibrate steadily. The position of faint sound occurs because the vibration of air excited by one prong of the fork reaches the ear so much later than that excited by the other that, while one wave is in what corresponds to a crest, the other is in what corresponds to a trough. At all parts of the waves their effects are similarly opposed, and the result, if the waves are equal in strength, is silence.
With light a similar thing may occur. From two slits, di and a2 (Fig. 13), imagine light of a single color to proceed in all directions. Then at fcj, b2, etc., the waves may be supposed to arrive in opposition, and thus to produce darkness, while at cl} c2, c3, etc., there is light. From a great number of other
50
SOLAR INVESTIGATION
slits, a3 a4, etc., placed in a plane, and equally spaced, the directions of light and darkness will be the same, so that if a piece of plane glass is coated with silver, and the silver coat is scratched off in a series of parallel and equidistant lines, a bundle of rays of light passing through the slits will some of them proceed parallel to the direction AB, while the others will be deviated, or diffracted, as it is called, in various definite directions on either side of the central beam. These directions depend upon the
B
FIG. 13.
interval between successive rulings, and on the length of the wave of the given color of light. The devia- tions are less for violet light than for the red, which shows that the length of wave is less for violet rays. Such a grating, as has just been mentioned, is called a transmission grating, but it is more common to employ a reflecting grating. A carefully ground and polished surface of speculum metal is scratched with a diamond point in parallel rulings very close to- gether, not uncommonly as many as 20,000 to the
51
THE SUN
inch. One may suppose that with such close ruling, the spaces between the ruled scratches are probably like the rough ridges turned up by a plow, and, as they would reflect but weakly, they may be assumed to correspond to the opaque parts of the transmission grating, while the smooth sides of the scratches act as bright sources of light. To the late Prof. Henry A. Rowland, of Baltimore, is due the principal share of the credit for the great advance in knowledge of the solar spectrum, and of that of the spectra of vaporized substances, which has come in the last twenty-five years; for he it was who designed the perfected screw, and thereby was enabled to construct hitherto un- equaled ruling machines. Rowland gratings, having a total of as many as 60,000 lines or more, each two or three inches long, are in nearly every large labo- ratory and observatory of the world. Not only did he thus promote the work of others, but his own em- ployment of his gratings has left some branches of solar spectroscopy at the furthest forward mark they have yet reached.
Diffraction gratings may be ruled on flat surfaces and used with a collimator and objective like a prism, but many of them are ruled on concave surfaces, and are used, after a design of Rowland, without collima- tor or objective. Thus, we have the plane grating and the concave grating spectroscopes. The arrange- ment of the former is shown in Fig. 14. Frequently, however, the collimator is used also as the image- forming lens. Such an arrangement is called the
52
SOLAR INVESTIGATION
Littrow form of spectroscope. It may be employed also, for prismatic instruments if a plane mirror is in- troduced to return the beam through the prism. It is necessary to tip the grating or mirror a little so that the spectrum is formed above or below the slit. Fig. 15 shows the concave grating arrangement. In this figure, S is the slit, G the grating, and I the spectrum. C is a rigid bar which carries the grating
FIG. 14.
and the means of observing the spectrum. This bar, mounted on carriages, k and k' ', slides over the tracks, R and R'. These tracks are placed at right angles, with their intersection at S.
From a white light a grating spectroscope produces a series of spectra more and more diverging on either side of a single white band in the center. These spec- tra are called first, second and third order, et cetera, according to their divergence. Only one spectrum can be made use of at a time among all this multitude, and the greater the number of the order, the greater the dispersion of the spectrum. The higher orders overlap, so that the red of one order falls on the violet or some other color of the next higher order. When it
53
THE SUN
is necessary to separate entirely one color from the other, it is customary to interpose somewhere in the beam an absorbing screen which is opaque to the color not desired, but transparent to the other. Spec-
FIG. 15.
tra of very high orders, however, are hopelessly mixed and are practically white light. It is seldom that spectra above the fourth order are employed. The relative brightness of grating spectra depends on the
SOLAR INVESTIGATION
form of the grooves ruled. Some diamond points produce gratings very bright in one or two particular spectra, and are preferred for this reason. The selec- tion of a good diamond point is the result usually of trial rather than of microscopic examination. A spectrum may be very bright for some colors and not for others. At best, a grating seldom throws as much as one-tenth of the light into one spectrum, and, therefore, in researches where loss of light is very serious a prism is often preferred, since it may trans- mit as much as eighty-five per cent. In a prismatic spectrum the violet is greatly extended as compared with the red, while in a concave grating spectrum the dispersion is a linear function of the wave length. That is to say, equal distances along the concave- grating spectrum correspond to equal differences of wave length. Such spectra are said to be "normal." A plane-grating spectrum is nearly normal for short distances.
The wave lengths in the spectrum, as visible to the eye, range from about 0.39/*1 to 0.80/*. Beyond the violet the solar spectrum runs to a wave length of 0.29/A, where it is practically cut off, partly by the nontransparency of our own atmosphere (particu- larly the nontransparency of ozone) and perhaps imperatively by the opaqueness of the solar envelope. Beyond the red the solar spectrum extends to a wave length of about 20/*, though with several long inter-
irThe micron, or thousandth of a millimeter, is denoted by the Greek letter j*.
6 55
THE SUN
missions due to the nontransparency of the atmos- phere (especially of water vapor, carbonic acid and ozone), from which cause it practically ceases at 20fju. Ordinary glass apparatus ceases to be trans- parent at about wave length 0.35/* in the ultra-violet, and at about 2.5/* in the infra-red, but the limits differ with different kinds of glass. Quartz apparatus is transparent to rays of all wave lengths from less than 0.20/I,1 to more than 4.0u. Fluorite is transparent in the ultra-violet, and in the infra-red its transpar- ency extends to about 7.0/i. Rock-salt is also trans- parent in the ultra-violet, and as far as 17/* in the infra-red. Silvered glass mirrors reflect almost totally for all rays of the infra-red and visible spectrum, and their reflecting power remains high as far as wave length 0.33/z in the ultra-violet. Between wave lengths 0.33/x and 0.29/*, the reflecting power of silver does not reach fifteen per cent. Speculum metal, which is used for gratings, reflects much less strongly than silver in the visible spectrum, but continues to reflect forty per cent or more to beyond wave length 0.30/i.
As stated above, it is the minute study of the lines found in spectra which yields many of the most inter- esting results, and in the solar spectrum these lines become increasingly numerous towards the violet, and in the ultra-violet. Fortunately, the ordinary
1 Although solar rays of less wave length than 0.29^ are not found, terrestrial sources give rays of much shorter wave lengths, even to O.lOn
56
SOLAR INVESTIGATION
photographic plate is highly sensitive in this thickly lined violet and ultra-violet part of the spectrum, and at present most spectrum investigations are made photographically. There are special photographic plates which are sensitive in other parts of the spec- trum. By staining ordinary plates with certain dyes, they may be employed for red rays, and even a little beyond the visible limit of the red spectrum. For spectrum investigations far beyond the red, it is necessary to use sensitive heat measuring apparatus, such as will soon be described.
For some purposes it is sufficient to allow the rays of the sun to shine directly into the spectroscope, but ordinarily it is necessary to confine the obser- vations to selected areas of the sun such as a sun spot, or to the sun's edge or "limb" as distinguished from the center. To do this the slit of the spec- troscope must be placed in the focus of a lens or con- cave mirror which forms a solar image of suitable dimensions for the investigation. When the spec- troscope is large, and the work requires it to be main- tained at perfectly constant temperature for long photographic exposures, it becomes highly desirable to keep the spectroscope fixed and to employ a ccelostat to reflect light to the lens or mirror. Fig. 16 shows the new 150-foot tower telescope, with a pit 75 feet deep beneath for the spectroscope, as just being completed at the Mount Wilson Solar Obser- vatory. A smaller tower telescope has been doing good work there for a considerable time. The
57
THE SUN
coelostat is on the top of the tower 60 feet high, and reflects a beam of sunlight vertically downwards
through a lens which forms a solar image over 7 inches in diameter upon the slit of the spectroscope near the surface of the ground. The slit is in the center of a turn- table which supports, by rigid steel construction, the collimator and plane grating 30 feet below ground. Thus, the whole spectroscope can be rotated about the axis of the beam of light. The col- limating lens acts, also, as an image-forming lens (the Littrow type of spectro- scope), and the spectrum falls on the photographic plate fixed upon the surface of the turntable near the slit. Below ground the tempera- ture is very constant. At the top of the tower, the air is nearly free from the trem- ors which cause " boiling" of the image. As the beam de- scends vertically from the top of the tower it is less 58
FIG. 16.
SOLAR INVESTIGATION
likely to be distorted by " boiling" than it would be if coming obliquely, as from the sun directly. Hence, altogether, the tower plan of solar observatory is highly favorable for carrying on exact investigations with powerful apparatus. The new tower telescope of over 150 feet focus just being erected for the Mount Wilson Solar Observatory will doubtless yield very remarkable results.
THE SPECTROHELIOGRAPH
The spectroheliograph, invented by Dr. G. E. Hale, is a device for photographing the sun in the light of a single wave length. Let us suppose that the solar image is brought to focus on the slit of a spectroscope, and that the slit is longer than the diameter of the image. The spectroscope may be adjusted so that a certain Fraunhofer line, perhaps the line called C, or otherwise Ha (due to hydrogen), falls in the center of the field of view. Then, if the solar image is allowed to drift across the slit, the observer will see the masses of hydrogen on the sun which emit the light in question as their images pass in succession over the slit. But it would be practically impossible to note and remember or sketch these details. If the photographic plate is substituted for the eye, and a slit placed just in front of it, so narrow as to permit only the Ha line to pass, a photographic record would be made, but this would be a mixture of all the successive views of the hydro- gen masses, and would be useless. But by moving
59
THE SUN
the plate along at the same rate that the image of the sun drifts, there would be a new part exposed for every succeeding impression, and the result would be a photograph of the hydrogen masses which emit Ha light, as they exist over the whole sun's disk. This is one form of the spectroheliograph. In another form, which is employed for the five-foot spectroheliograph of the Snow telescope of the Mount Wilson Solar Observatory, the whole spec- troscope is floated on mercury, and moved slowly, at right angles to the beam, across the sun's image and the photographic plate, both of which remain station- ary. The solar image of the Snow telescope is about 7 inches in diameter, and, if the correspondingly long slit of the spectroscope were straight, the spectrum lines would be greatly curved, and the sun's image taken with the spectroheliograph would be distorted. This defect is avoided by using curved slits, dividing the necessary curvature between that for the spectro- scope and that in front of the plate. The curvature of these slits differs for different spectrum lines, so that as many pairs of slits are required as there are spectrum lines in which spectroheliograms are de- sired. Thus far Ha, H/3, H7, HS of hydrogen, H and K of calcium, and a few preliminary tests of other lines have been tried.
THE HELIOMICROMETER
It ordinarily requires considerable measurement and calculation to determine the positions of objects
60
SOLAR INVESTIGATION
with reference to the solar equator, seen on the solar photographs, whether direct or spectroheliographic. This labor is largely avoided by the use of a device of Mr. Hale's called the heliomicrometer. It con- sists of a sphere marked with circles of latitude and longitude, and adjusted so that its poles correspond in position with those of the sun for the date in question. A long focus concave mirror throws an image of this sphere, and of the photographic plate to be examined, simultaneously into a double-field eyepiece. Thereby, the two images are superposed, and the observer sees the solar photograph appar- ently marked with lines of latitude and longitude corresponding with those of the sun. A micrometer is provided for accurately measuring the distances of the images of the solar objects from the nearest reference lines.
THE COMPARATOR
In all photographic spectrum work, the main thing is accurate measurements of the positions of the spectrum lines with reference to each other, or with reference to certain standards of position. In many cases the slit of the spectroscope is partly covered by a diaphragm of peculiar shape, which can be moved so as to uncover different portions of the slit. Thus, successive exposures may be made to different sources of light, as, for instance, the center and limb of the sun, or the sun and the iron arc light. In the resulting photograph there
61
THE SUN
are several spectra corresponding to these different sources, all accurately aligned one above another. For measurement, the photograph is placed on the table of a measuring machine, or comparator, and this table is moved to and fro by an accurate screw with graduated head, thus bringing chosen spectrum lines to the cross hair of the observing microscope. Measurements of position to the ten-thousandth
part of a millimeter (to ^,7^ inch) are some-
times made in this manner.
The wave lengths of the solar spectrum lines and of the bright spectrum lines of the chemical elements are the fundamental data of spectroscopy. In Row- land's great table of the solar spectrum the wave lengths are given to seven places of significant figures, that is, to thousandths of an " Angstrom unit." It has lately been found that there are certain syste- matic errors of the table due to various causes, chiefly to an obscure source of error in the use of the grating for determining wave lengths, so that there are corrections of the order of one or two hundredths of an Angstrom to be applied to make Rowland's table homogeneous. To reduce to the absolute scale of the international metric system, a some- what larger correction is needed. By means of the interferometer these corrections are gradually being determined, and it is probable that within a few years we shall have a standard table of solar and terrestrial spectrum places accurate to within two or
68
SOLAR INVESTIGATION
three units in the seventh place of significant figures. It seems extraordinary enough that so small a quan- tity as the wave length of light should be measurable to such extreme precision, and still more extraor- dinary that such a decree of accuracy is at all necessary for promoting investigation. But so it is, and much of the remarkable progress of solar knowledge in recent years depends on differences of wave lengths, as in the case of pressure and velocity shifts, not larger than 0.005 of an Angstrom, or less than one-millionth part of the wave length of yellow light.
THE NATURE OF RADIATION
Not less important, perhaps, than these questions of exact wave lengths, is the measurement of the intensity of light, or rather, speaking more broadly, of radiation. All solar rays, whether visible or photographically active or not, produce heat when absorbed upon a blackened surface. Sometimes the infra-red rays are called "heat rays/' the light rays, "visible rays," and the blue, violet and ultra-violet, " actinic," or " photographic rays." But there is no distinction of kind between these things. All are regarded as transverse vibrations of the luminifer- ous ether, differing only in wave length. Just as there are sound waves too high or too low in pitch to be heard, so radiation may be too long or too short in wave length to be seen, but this implies no dif- ference in kind of vibration.
The intensity of radiation can be quantitatively 63
THE SUN
estimated only very imperfectly by the eye, or by the aid of the photographic plate, although both the eye and the plate are excessively sensitive to radia- tion of certain wave lengths. But waves of all wave lengths produce their just effects when trans- formed into heat. Though both are forms of energy, radiation is not heat, but may be transformed com- pletely into heat. We regard radiation as wave mo- tion in the ether, heat as irregular motion of the molecules of material substances. All heated sub- stances give off radiation; but the amount and quality of radiation given off at a given temperature are different for different substances. Substances at any temperature above the absolute zero ( — 273° C.) are supposed to consist of molecules in rapid mo- tion. These moving molecules may be supposed to communicate some of their energy to the unseen ether which is assumed to permeate all space, even the interstices between the molecules of solid bodies. Thereby the ether may be assumed to be set in con- fused vibration, and from this confusion is extri- cated by the prism, or grating, the orderly succession of wave lengths which we term the spectrum. The relative intensity of the several parts of such a spectrum depends on the temperature of the exciting body.
Kirchhoff introduced the notion of the perfect radiator. This is sometimes called "the absolutely black body," because a perfect radiator is a per- fect absorber of radiation, and most black substances
64
SOLAR INVESTIGATION
are also nearly perfect absorbers. The perfect radi- ator emits for a given temperature the maximum possible amount of radiation of each and every wave length; so that no other body at the same temperature can excel its emission for any wave length.1
LAWS OF RADIATION
Kirchhoff proved the following important rela- tion, now known as Kirchhoff 's law: For any given temperature and wave length the ratio of the emis- sion of a body to its absorption is a constant, and equal to the emission of a perfect radiator for the same temperature and wave length. In order to understand this law, the force of the expressions emission and absorption must be clearly grasped. By emission is meant the rate of escape of energy by radiation, and to fix ideas it may be regarded as the amount radiated from each square centimeter of sur- face in a minute of time. By absorption is meant the fraction which would be absorbed in the body if shined upon by radiation from another source. For instance, if thus shined upon, and three-fourths of the rays received are absorbed and go to warm the body, while the other fourth is reflected away, or transmitted, the absorption is said to be three- fourths. Such a body, by KirchhofTs law, would emit only three-fourths as copiously, for the wave
1 An exception must be made, perhaps, of a certain class of bodies excited to radiation by other causes than temperature, as, for in- stance, chemical action. The remarks above concern the relations of temperature and radiation alone.
65
THE SUN
length and temperature in question, as would the perfect radiator.
The importance of the conception of the perfect radiator will appear as we go on. No substance in the world answers to its requirements, but lamp- black is very nearly a perfect radiator at low tem- peratures. However, if a closed hollow chamber is formed of any substance whatever, and its walls maintained at uniform temperature, the radiation in- side the chamber will be that of the perfect radiator. If a small hole be made in the wall the radiation which escapes through the hole will be practically perfect radiation. Instruments of this form have been constructed within the last fifteen years, and careful measurements have been made of the inten- sity of their emission for a great range of wave lengths, and for temperatures from that of liquid air up to that of melting platinum. These results have been compared with the theoretical radiation formu- lae connecting temperature, wave length and radia- tion which have been proposed.
The formula of Wien, as modified by Planck, is found to express the observed results. Let e be the emission of wave length X by the perfect radiator of the temperature T, and let e be the base of the Napier- ian system of logarithms, and let c± and c2 be two con- stants determined by experiment. Then:
Cj
e = cA~5(e AT - 1 )-! (The Wieri-Planck formula) . I
As stated above, no body emitting rays by virtue of
(5(1
SOLAR INVESTIGATION
temperature can exceed the radiation determined by this formula for any wave length or temperature.
Another formula of nearly equal importance, due to Stefan, gives the measure of the sum total of radi- ation, E, of all wave lengths, for a perfect radiator of the absolute temperature, T. It is this:
E = o-T4 (Stefan's formula). II
The quantity, cr, is a constant determined by experi- ment.
A third formula, called Wien's displacement law, connecting the wave length of maximum emission, ^max. (expressed in thousandths of a millimeter, or /*) , with the absolute temperature T is as follows:
Vax. T ~ 293° (Wien's displacement formula). Ill It is from these three formulae that we are able to obtain some definite ideas of the minimum tempera- ture of the sun. Many bodies appear to approach the state of being perfect radiators at high tempera- tures, although departing greatly from it at low tem- peratures. But no body radiating by virtue of its temperature can excel, either in the sum total of its radiation, or in that of any wave length, the emission of the perfect radiator of the same temperature. Hence, if we can determine by Formula II the temper- ature which the perfect radiator would have in order that its radiation should approximate in quantity the emission of the sun, then it is sure that the solar temperature must be as high or higher.
Before giving the values of the constants in these 67
THE SUN
formulae, we must consider how energy of radiation can be measured. There are no accurate means of measuring radiant energy while it remains such. It must first be transformed into heat. The unit of measurement of heat is the calorie, or that amount of heat which is required to warm one gram of water at 15° C. through one degree. With this unit we must combine the notion of intensity. We then define the unit intensity of radiant energy as that which, if completely absorbed by a surface at right angles to the beam, will produce one calorie of heat per square centimeter per minute. We therefore meas- ure radiation in calories per square centimeter per minute.
To suit this definition, and to correspond with wave lengths expressed in microns (/*), and tem- peratures in absolute degrees of the Centigrade scale, the values of the constants of formulae I and II are as follows :
c,= 5.29 X 105; C, = 14,550; a- = 76.8 X 1Q-12.
SPECTRA OF DIFFERENT SOURCES
In Fig. 17 the curves A and B give the distribution of radiation in the spectrum of a perfect radiator at 7000° and 6200° of the absolute Centigrade tempera- ture scale, as computed from the Formula I. The curve C gives the distribution of radiation as it would be found in the average spectrum of the sun's entire disk, if it could be observed outside of our atmos- phere, according to determinations made by Smith-
68
SOLAR INVESTIGATION
a: AH3Sso
rfg'
X
at/W'XVx*
sonian expeditions on the summits of Mount Wilson and Mount Whitney. The wave lengths are given by the horizontal distances (abscissae) and are in thousandths of a millimeter, or microns, usually de-
69
THE SUN
noted by the Greek letter /*. The visible spectrum practically extends from 0.4/* to 0.7/*1, so that much of the solar radiation is invisible. The vertical heights of the curves (ordinates) are proportional to the energy of the rays of corresponding wave lengths as meas- ured by their heating effects. It will be noted that the forms of the computed and observed curves differ most in the ultra-violet, where the observed solar radiation falls off more rapidly than the computed radiation of the perfect radiator. Further remarks on the subject of the sun's temperature will be given in the next chapter.
The reader will note that the maximum ordinate of curve A occurs at a less wave length than that for curve B, and that curve A is at all points the higher of the two. The perfect radiator is supposed to emit rays of all wave lengths at all temperatures, whether high or low; but when the temperature is low the shorter wave lengths, including those which would be visible, are too weak to be detected, even by such a highly sensitive organ as the eye. As the tempera- ture increases the intensities of rays of all wave lengths increase, but the intensities of rays of shorter wave length increase most rapidly. Hence, as ex- pressed in Formula III, the wave length of the maximum emission grows less and shifts towards the violet end of the spectrum as the temperature in-
1 By special devices, the spectrum can be observed visually from 0.37 /u to 0.83 /A, but as ordinarily observed it falls within the limits above stated.
70
SOLAR INVESTIGATION
creases. Most common solids and liquids emit a continuous spectrum, which, as the temperature in- creases, grows in intensity more rapidly for short wave lengths than for long. But there are usually special regions, or bands of the spectra of solids and liquids where the radiation is stronger than that of the adjacent wave lengths. These are called regions of " selective emission," and, as follows from Kirch- hoff's law, they are also regions of " selective absorp- tion."
When gases or vapors are examined under ordinary conditions of low pressure, and with small quantity present, as when the electric arc is caused to play between metallic poles, the spectrum appears to be made up chiefly of narrow lines or bands of selective emission, without a prominent accompanying con- tinuous spectrum. Some authors hold that the con- tinuous background is totally absent in gaseous spec- tra, but it seems more likely that there is, in fact, a very slight vestige of it present, which, if the quan- tity of gas was increased, so that the observer could look towards immense thicknesses, would be increased until the emission for -all wave lengths would finally approach the intensity of a perfect radi- ator. This view is supported by the circumstance that, if the pressure upon the emitting gas is in- creased to several atmospheres, the spectrum lines widen out, till at length there is, for some distance from the lines, a perceptible continuous background. Whether or not, then, it be true that gases under 7 71
THE SUN
less than atmospheric pressure would give continu- ous spectra if in great depth, it is certainly highly probable that such gases would do so if more and more compressed with the increasing thickness. Re- gions of strong emission are regions of strong absorp- tion by Kirchhoff's law, so that in the case of a thick gas, as just proposed, it would be only the front layers which would give rise to the lines or bands of high selective absorption, while the deep-lying layers would be those which would produce the continuous spectrum. If the gas is not of uniform temperature, but grows hotter with increasing thickness, it is easy to see that the continuous spectrum might exceed the line spectrum in its intensity, so that the really bright lines would appear dark by contrast with the background. As is well known, the solar spectrum has the character of a continuous bright ground crossed by darker lines, and evidence will be pre- sented later which indicates that it is indeed to be regarded as a gaseous spectrum of the kind just described.
PYRHELIOMETRY
In the year 1838 Pouillet devised the instru- ment which he called the pyrheliometer, shown in Fig. 18, and used it for measuring the intensity of the sun's radiation. A flat silver-plated vessel a&, blackened with lampblack on its upper surface, is filled with water, and contains also the bulb of the thermometer d. The instrument is held in the
72
SOLAR INVESTIGATION
clamp c, and pointed towards the sun as indicated when the shadow of the box ab falls centrally on the plate ee. By rotating the whole apparatus in the clamp c, the water can, in effect, be stirred to equalize its temperature. To observe the intensity of the solar radiation the instrument is first shaded, and the change of temperature occurring in a certain time, as, for instance, five minutes, is noted. Then the screen is re- moved, and the observer notes the change -of temperature due to the sun's heating in the same time. Finally the shade obser- vation is repeated. Correcting the average rate of rise of tem- perature per minute during the sun exposure by the average rate of cooling, shown by the shade readings, the result gives the rise of temperature per min- ute of a mass of water and cop- per, of known heat capacity, due to the sun's rays shining at right angles and absorbed on the known area of the top of the box. A correction of about 2.5 per cent must be added on account of loss by reflection from the lampblack.
Pouillet observed the intensity of the sun's rays with this instrument at different hours of the day.
73
FIG. 18.
THE SUN
The atmosphere weakens the sun rays by the diffuse reflection of its molecules and dust particles. This effect is more and more apparent as the sun nears the horizon. The atmosphere extends upwards for a great distance, but becomes less and less dense, so that at one hundred miles elevation what re- mains above is negligible, so far as cutting off the sun's rays is concerned. Hence, we may regard the effective part of the atmosphere as a layer whose thickness is very small compared 'to the earth's ra- dius; and so, whenever the sun is 15° or more above the horizon, the length of path of its rays in air is in proportion to the length of the path when the sun is in the zenith simply as the secant of the zenith dis- tance at the time of the observation.
Bouguer and Lambert had shown independently, in the year 1760, that when a ray traverses a homo- geneous transparent medium, the intensity, E, after traversing any given thickness, t, of the medium is given by the following formula, in which E0 is the original intensity, and a is a constant which repre- sents the proportion transmitted by unit thickness:
E == E0a<.
Pouillet applied Bouguer 's formula to his observa- tions, taking unit thickness as that traversed by rays when the sun is in the zenith, so that if z is the zenith distance, the formula becomes:
E = E0asecantz.
He computed the value E0, which is the intensity of the sun's radiation outside the atmosphere, and, re-
74
SOLAR INVESTIGATION
ducing to mean distance of the sun1, obtained E0 =. 1.76 calories per square centimeter per minute. This value Radau, and, also, Langley afterwards showed must be below the true value of the " solar constant of radiation" because Pouillet made no spectrum observations, and it is necessary to do so on account of the unequal losses suffered by rays of dif- ferent wave lengths in passing through the air.
Pouillet's pyrheliometer was improved by Tyndall, who substituted an iron box containing mercury in place of the copper box containing water. In re- cent years Tyndall' s design has been improved at the Smithsonian Institution. First, a copper box rilled with mercury was employed; then a copper disk with a hole drilled radially to contain the cylin- drical bulb of a thermometer with, also, a little mer- cury surrounding -it to make good heat connection; now (1910), the Institution uses a blackened silver disk (shown in section at a in Fig. 19) with a radial hole lined by a thin steel thimble. In this is inserted in mercury a cylindrical-bulb thermometer, 6, bent at right angles so as to point towards the sun when in use. The disk is enclosed in a brass-walled, black- ened chamber, c, and this is protected from changes of temperature by a wooden wall, d, outside. The sun's rays are admitted through a tube, e (shown partly in section) , which contains diaphragms, / / /,
1 The sun's radiation varies in its intensity inversely as the square of the sun's distance. Hence the earth receives on this account nearly 7 per cent more solar radiation in January than in July.
75
THE SUN
FIQ. 19. — SILVER DISK PYRHELIOMETER.
76
SOLAR INVESTIGATION
to prevent air currents from reaching the silver disk. An equatorial mounting enables the observer to point the instrument towards the sun. Several in- struments of the type shown in Fig. 19 have been
constructed and com- pared with those at the Institution, and sent to
FIG. 20 — ANGSTROM'S PYRHE- LIOMETER.
FIG. 21.
different solar observers abroad to convey to them exactly the scale of measurements employed here. In 1896 K. Angstrom devised his electrical com- pensation pyrheliometer, which has been used very extensively. Fig. 20 gives a general view of the instrument and Fig. 21 an enlarged detail view of the interior. It consists of two thin strips of man- ganin, U U, of measured area, which are blackened on the front surfaces, and have fixed to the rear of each a thermoelectric junction for determining their tem- peratures. The binding posts, Kx K2, communicate respectively to the strips and the thermal junc- tions. A measured current of electricity is passed through one strip, while the other is exposed to the sun, and when a galvanometer connected with the
77
THE SUN
thermal junctions indicates equality of temperature it is assumed that the known amount of heat in- troduced by the electrical current is equal to that absorbed from the sun's rays. By reversing the screen, W, and the commutator, C, the two strips are heated alternately by the sun and by electricity, and the mean result is employed. After applying a correction for loss by reflection, the results are com- puted in terms of calories per square centimeter per minute. The instrument is inclosed in a dia- phragmed tube R, and is mounted on an alt-azimuth stand provided with the screws, Si S2, for following the sun. A thermometer, T, indicates the tempera- ture of the strips.
In both forms of pyrheliometers described above, if used as standard instruments, a correction must be determined and applied to allow for the radiation reflected. Besides this, there is another source of loss, arising from the fact that part of the heat produced by the absorption of solar radiation in lampblack is carried off by the air, and by re- radiation of great wrave length, and this part does not produce any effect on the thermometer or ther- moelectric junction.
To avoid these sources of error other forms of pyrheliometers have been devised in which the rays are absorbed within a hollow cylindrical blackened chamber. Such a chamber, as stated a few pages above, is practically a perfect radiator, and hence is a perfect absorber, so that no correction for rays
78
SOLAR INVESTIGATION
^reflected is needed. The rays are principally ab- sorbed at the rear end, and, as the tube is deep, the heat tending to es- cape will be absorbed somewhere on the side walls. Two means of using the hollow cham- ber have been employed, the first about 1894, by W. A. Michelson, the second, 1905 to 1910, by the writer. Michelson surrounds the chamber by melting ice and water, and determines the heat introduced by measuring the contraction of the ice as it melts.
In the form devised by the writer, as shown in Fig. 22, a measured stream of water, enter- ing at E and emerging at F, flows continually in a spiral channel round the walls of the blackened chamber, A A, carrying off the heat as fast as
THE SUN
formed. The rise of temperature of the stream of water due to the solar heating (admitted through the vestibule, B B, and the measured diaphragm, C) is de- termined by a differential electrical thermometer com- posed of four fine platinum wires wound longitudi- nally on ivory spirals. These wires are bathed by the stream of water which follows the spiral channels of the ivory. Two coils are situated at Db in the entering stream of water, and two at D2, after its passage through the walls of the chamber. The four are joined to form a Wheatstone's bridge, and their indications are read by a sensitive galvanometer. The pyrheliom- eter is protected from outside temperature changes by the Dewar vacuum flask, K K. In order to test the accuracy of the instrument two coils of man- ganin wire, G and H, are placed within the chamber near its rear, and a known quantity of heat may be produced there in either coil by the passage of a measured current of electricity. This heat is then measured just as if it were from the sun, and if all that is introduced is found, it may be supposed that the instrument is a correct recorder of solar radia- tion, especially as the coil G is very unfavorably sit- uated for giving up its heat to the walls.
Two such water-flow pyrheliometers of different dimensions were tested at Washington in 1910 and gave closely agreeing results on solar radiation, be- sides recovering almost completely the electrically developed heat used as a test. These water-flow pyrheliometers are used as standards, and the read-
80
SOLAR INVESTIGATION
ings of the silver-disk pyrheliometers are reduced to the scale they give. The water-flow pyrheliometer, when in use, is mounted equatorially and driven by clockwork to follow the sun. It is alternately shaded and exposed to solar radiation.
BOLOMETRY
For measuring the intensity of the rays in the solar spectrum, the instrument most used is the bolometer, a delicate electrical thermometer, in- vented by Langley about 1880. As now construct- ed, it comprises two exactly similar, narrow, blackened, platinum strips hardly as wide as hairs, ten times thinner than they are wide, and about half an inch long. Referring to Fig. 23, such strips, a, b, having an electrical resistance of about four ohms each, are joined, as shown, to two coils, c, d, of manganin wire, each of about 20 ohms resistance, forming with the -two strips a Wheatstone's bridge. A variable resistance, e, of several thousand ohms is shunted around one coil and serves to bring the whole to an electrical balance. Sometimes a small resistance of copper, /, is included in one arm of the Wheatstone's bridge to prevent its unbalancement as the surrounding teniperature changes. A cur- rent of about 0.1 ampere from a storage battery of
81
— L|J O-vWV-O
FIG. 23.
THE SUN
several cells, in parallel, flows constantly through the bridge, and the adjustment is observed by a highly sensitive galvanometer, g. If the radiation is caused to fall on one of the bolometer strips, its re- sistance increases, and there results a deflection -of the galvanometer proportional to the heat produced by the radiation. The record of the galvanometer is kept automatically on the photographic plate which is moved vertically by the clockwork at the same time that the spectrum is moved across the bolom- eter strip, so that rising and falling temperatures of the strip, due to changes of intensity of the spec- trum, are indicated by higher and lower parts of the curve, photographically traced by the little spot of sunlight reflected by the tiny mirror of the galvanom- eter needle. Fig. 24 gives a pair of such energy curves or holographs of the solar spectrum. Some of the principal Fraunhofer lines give great depres- sions of the curve, and are indicated on the margin of the figure. At the points marked * * a shutter was introduced in front of the slit of the spectro- scope to give the zero of radiation. At the points marked f t diaphragms were introduced to diminish the intensity of the spectrum, so that the photo- graphic trace would not run off the plate. The scale of the intensity as thus altered is indicated on the margin.
In Chapters III and VII are given the applica- tion of the bolometer for the determination of the "solar constant of radiation," the transparency of
82
SOLAR INVESTIGATION
83
THE SUN
the atmosphere for rays of different wave lengths, the investigation of the comparative brightness of dif- ferent parts of the solar image, and the determina- tion of the temperature of the sun. The astonishing sensitiveness of the bolometer may be understood when it is said that, in ordinary use, changes of
temperature of less than 77; of a degree C. are
measured, and by special installation this sensitive- ness may be increased 1000 fold. The still more astonishing sensitiveness of the eye is indicated by the fact that we receive enough light through the pupil of the eye from a star of the sixth magnitude to see it, though with the most sensitive bolometer it would require a mirror perhaps ten feet in diameter to concentrate enough rays from such a star to make its heating observable. This is the more striking because the eye is affected by only a short range of spectral colors, while the bolometer measures the total radiation of all wave lengths.
CHAPTER III
THE PHOTOSPHERE
Telescopic View. — The Photospheric Spectrum. — Rowland's Spec- trum Tables. — Chemical Elements Found and Not Found. — Corrections to Rowland's Wave Lengths. — Levels. — Pressures. — Convection Currents. — Limb Spectra. — Variation of the Sun's Brightness. — Solar Temperatures. — Spectroheliography. — Solar Rotation.
As viewed through the telescope, or photographed, the radiating surface of the sun, called the " photo- sphere, " presents a brilliant disk covered by indis- tinct mottlings sometimes spoken of as the " rice- grain-structure." Objects much less than a second of arc or 400 miles in diameter, cannot be well seen on the sun, so that these " rice-grains, " which appear according to different authors from 100 to 500 miles in diameter, are really large areas. Some authors speak of the bright areas of this mottled appearance as " granulations, " and the darker parts as "pores." Generally a few very dark patches called "sun spots" may be seen, and around them, if they hap- pen to be observed near the edge or "limb" of the sun, are found very bright areas called "faculse. " The faculse are seldom seen very much more than a quarter radius within the limb. Photography reveals at once, what the eye recognizes less easily, that the
85
THE SUN
photosphere falls off in brightness towards the sun's limb. A photograph well exposed at the center will be very weak at the limb. Plate III shows this clearly, and also exhibits the rice-grain structure, sun spots, and faculse. Sun spots march nearly reg- ularly across the sun's disk in about 13.6 l days, and appear after an equally long absence, which indicates that the sun rotates upon its axis.
THE PHOTOSPHERIC SPECTRUM
The spectrum of the sun's photosphere is a con- tinuous bright background of color crossed by dark lines and bands. Newton recognized seven colors in the spectrum, comprising violet, indigo, blue, green, yellow, orange and red, but these blend into one another by perfectly imperceptible gradations of in- numerable hues. By photography and by the bolom- eter, the solar spectrum has been followed beyond the violet end as seen by the eye (which occurs about wave length 0.38^), as far as wave length 0.29/*. Here the rays are almost wholly cut off by losses in the earth's atmosphere and in the sun's outer envel- opes. Beyond the red, which may be observed with the eye to wave length 0.80/*, Abney has photo- graphed, by the aid of specially dyed plates, to wave length l.l/*, and with the bolometer the solar spec- trum has been measured at the Smithsonian Astro- physical Observatory as far as wave length 5.3/*.
'The earth is meanwhile advancing, so that this is not the half period of the sun's sidereal rotation.
86
PLATE III
DIRECT SOLAR PHOTOGRAPH. (EllermarO 1908, April 30. G. M. T. 2 h 30 m. P. S. T. 6 h 30 m A. M.
THE PHOTOSPHERE
Probably sun rays might be recognized with the bolometer at intervals as far as 20^, but beyond this they would probably be practically all cut off by losses in the earth's atmosphere.
The dark lines and bands of the solar spectrum, named from their discoverer "Fraunhofer lines/' have two different sources. A considerable number of lines, notably in the red and infra-red regions of the spectrum, are caused by the absorption of gases and vapors in the earth's atmosphere. The chief of these terrestrial absorbents are oxygen, water vapor, and carbonic-acid gas. By far the greater number of the Fraunhofer lines, however, are formed by the absorption of solar rays by gases in and about the sun itself; notably by iron, nickel, calcium, titanium, cobalt, chromium, manganese, carbon, vanadium, sodium, magnesium, and hydrogen. The existence of these elements and many others in the sun is proved by the occurrence in the solar spectrum of dark lines, occupying the same relative positions as to wave length, and generally of nearly the same relative in- tensity, that the characteristic bright lines of these elements occupy in their spectra as produced in the laboratory. As shown by Kirchhoff and Bunsen in 1859, dark lines are produced in a bright continuous spectrum by interposing cooler vapors or gases be- tween the source of light and the spectroscope, and these lines occupy the same positions that the bright lines of the vapors or gases would occupy if the latter were themselves the sole sources of light. Conform- 8 87
THE SUN
ably to this discovery it will be shown in a later chap- ter that the spectrum of the outermost solar layer, called the " chromosphere/' when seen alone at solar eclipses, is a bright line spectrum »vhich is almost the exact reversal of the photospheric spectrum. The layer in which the dark lines have their rise is accord- ingly called "the reversing layer."
As any gases between the observer and the sun may produce dark absorption lines in this way, it is not at first apparent how to distinguish between terres- trial and solar gases. There are two ways of testing whether a given Fraunhofer line is solar or atmos- pheric. The first is by observing its intensity rela- tive to other lines at high and low elevation of the sun above the horizon. Atmospheric, or as they are called, " telluric," lines will generally be strengthened at low sun, because the layer of air traversed will then be greater. A second and better method of discrim- ination consists in forming an image of the sun and causing rays from its east and west limbs to be re- flected together simultaneously into the slit of the spectroscope, so as to give rise to two superposed solar spectra, one of light from the east limb, the other from the west. Telluric lines will occupy the same position in the two spectra, but solar lines will be shifted with reference to one another owing to the rotation of the sun, which produces a very notable Doppler effect. This is shown in Plate IV, Fig. 2, which includes the oxygen band B and some solar lines in its vicinity as photographed
88
3 g
* -a
^ o
02 g "'I
^ W
£ 5s
I o I 3
M
s I
w o
& PQ
THE PHOTOSPHERE
at Mount Wilson by St. John under remarkably fine conditions.
ROWLAND'S SPECTRUM TABLES
The solar spectrum has been photographed at great dispersion by numerous observers, but most notably by Rowland. He published about 1895, in the early volumes of the Astrophysical Journal, his great " Pre- liminary Table of Solar Spectrum Wave Lengths," which still forms the basis for solar and stellar re- searches. Rowland states in his introduction that he photographed the arc spectrum of all the then known elements except gallium in connection with the solar spectrum, but that the work of identifica- tion of lines in the solar spectrum with the arc lines would be a further labor of years. This work of identification has never yet been completed, nor has a correspondingly full comparison of the solar spec- trum with the spark spectra of the elements been attempted. In Rowland's " Preliminary Table" there are about 14,000 lines recorded. Their wave lengths are given to seven places of figures, that is, to thou- sandths of an Angstrom. For each line is given its intensity. The intensities go from 1, a line just clearly visible on Rowland's spectrum map, up to 1000 for the strong calcium lines H and K. Below 1, the intensities go down to 0000, indicating lines more and more difficult to see.
The great lines of the solar spectrum, named long ago for the letters of the alphabet, are as follows :
89
THE SUN
TABLE III. — Principal solar spectrum
A 7593.842 Oxygen3 |
a 7184.57 Waters |
B 6860.970 Oxygen3 |
C(Ha) 6562 . 835 Hydrogen |
D3 .->,SX!».!)7.~ Sodium |
i: 52(i!»..r).r)l Iron |
|
Corrected wave length1 Element |
||||||
b 5183.620 Magnesium |
F(H/3) 4861.350 Hydrogen |
G(Hy) 4340.471 Hydrogen |
H 3968.491 Calcium |
K 3933.680 Calcium |
||
Corrected wave length1 , . Element |
1 According to the table of corrections below.
2 Edge of the head of A. a Terrestrial lines.
About one-third of the 14,000 solar lines were iden- tified by Rowland and ascribed by him to various chemical elements. In a good many cases a line is attributed to several elements at once. In such cases the coincidence with them all is probably not generally exact, but only so close that, even with Rowland's very high dispersion, the several linos overlap. Investigations with much higher dispersion on bright line spectra indicate that in many cases apparently single lines of single elements are really resolvable into groups. But perhaps even such very high resolving powers would generally fail to separate the blended lines of Rowland's table, because, owing to pressure or other conditions, the several lines involved are so much widened as to overlap. For many years Lockyer maintained plausibly that the elements had common constituents which gave rise to common lines in the spectrum, but this so-called "basic line" hypothesis is not now generally held. The following summary of Rowland's identifications
90
THE PHOTOSPHERE
is taken from Young's "The Sun," with slight changes :
Chemical Elements Found and not Found in the Sun.— The first columns of the following table give the chemical elements found by Rowland to exist in the sun arranged according to the intensity of their solar lines, and with their atomic weights annexed. The last columns give them arranged in the order of the number of their solar lines, and with the numbers occasionally annexed. The sign f indicates that the element has not been identified in eclipse chro- mospheric spectra.
TABLE IV. — Chemical elements found in the sun.
1.
2.
3.
4.
5.
6.
7.
8.
9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23.
Calcium (40.09) Iron (55.85) Hydrogen (1.008) Sodium (23.00) Nickel (58.68) Magnesium (24.32) Cobalt (58.97) Silicon (28.3) Aluminum (27.1) Titanium (48.1) Chromium (52.0) Strontium (87.62) Manganese (54.93) Vanadium (51.2) Barium (137.37) Carbon (12.00) Scandium (44.1) Yttrium (89.0) Zirconium (90.6) fMolybdenum (96.0) Lanthanum (139.0) fNiobium (93.5) fPalladium (106.7)
Iron (2000 or more) Nickel Titanium Manganese Chromium Cobalt
Carbon (200 or more) Vanadium Zirconium Cerium
Calcium (75 or more) Neodymium Scandium Lanthanum Yttrium Niobium Molybdenum Palladium
Magnesium (20 or more) Sodium (11) Silicon Hydrogen Strontium 91
THE SUN
24. tNeodymium (144.3) Barium
25. tCopper (63.57) Aluminum (4)
26. Zinc (65.37) Cadmium
27. Cadmium (112.40) Rhodium
28. Cerium (140.25) Erbium
29. tGlucinum (9.1) Zinc
30. tGermanium (72.5) Copper (2)
31. tRhodium (102.9) Silver
32. Silver (107.88) Glucinum
33. Tin (119.0) Germanium
34. Lead (207.10) Tin
35. Erbium (167.4) Lead (1)
36. fPotassium (39.10) Potassium
Besides these thirty-six elements, thus arranged, it has been found that helium (4.0) and gallium (69.9) certainly show solar lines, although helium lines are hide-and-seek things, and for some reason only occa- sionally appear as dark lines in the solar spectrum. There also appear very faint dark solar lines, nearly or exactly corresponding in their position to some of the strongest arc lines of:
TABLE V. — Chemical elements doubtfully occurring in the sun. Ruthenium (101.7), Indium (114.8), Tantalum (181.0), Tungsten (184.0), Osmium (190.9), Iridium (193.1),
Platinum (195.0), Mercury (200.0), Thallium (204.0), Bismuth (208.0), Thorium (232.42), Uranium (238.5). The mean atomic weight of these elements is 186 . 95.
The lines of the important elements of the halogen group, Fluorine, Chlorine, Bromine, Iodine; those of the oxygen group, Oxygen,1 Sulphur, Selenium, and
1 Since this was written St. John has found that a triplet of faint lines attributed to oxygen occurring beyond A in the extreme1 red, shows relative displacements at the sun's limbs. Hence, we must probably admit free oxygen as giving a solar spectrum. Combined oxygen and combined nitrogen give solar band spectra.
92
THE PHOTOSPHERE
Tellurium; those of the nitrogen group, Nitrogen, Phosphorus, Arsenic, and Antimony (Bismuth doubt- ful) do not appear to have been found in the photo- spheric spectrum, or in the spectrum of the chromo- sphere. This singular omission comprises nearly all of the prominent " negative" elements, and Boron, another of them, is also absent from the solar spec- trum. Further remarks on this subject will be made later.
There is considerable interest attaching to the relations of atomic weight of the elements and the intensity of their solar lines. Taking the thirty-six ele- ments of the intensity table in order, in four groups of nine each, the average atomic weights are as follows:
Elements 1-9, 35.26; elements 10-18, 64.04; elements 19-27, 101.27; elements 28-36, 107.25.
In the last group, as thus divided, occur glucinum (9.1) and potassium (39.10). The former has two, and the latter one identified line, and, as these lines are also very weak, it is not impossible that these two elements may by future investigation fall out of their strange company.1 If so, the mean atomic weight of the remaining seven elements would be 131.00. In Group II of this arrangement appears carbon (12.00), but, judging from Kayser's " Handbuch," the solar " carbon " lines belong to carbon compounds of high molecular weights. Hardly less interesting than the
and Runge question the existence of potassium lines in the photospheric spectrum.
93
THE SUN
classification just given is the further fact that most of the elements of the platinum group, and some other elements of very high atomic weight found commonly on the earth, are only doubtfully recognized in the sun, although they give strong lines in the arc. The full significance of these relations will be further dis- cussed in Chapter VI, but it may be said here that the explanation of the decrease of intensities with in- creasing atomic weights seems to depend on the depth of these gases below the sun's surface. We may sup- pose that the interesting elements radium and ura- nium might not produce lines in the solar spectrum, even if these elements exist in the sun, because of their high atomic weights.
The element oxygen undoubtedly exists in the sun because the flutings of titanium oxide are very prom- inent in sun-spot spectra. It might be anticipated that the well-known oxygen lines themselves would be found in the photospheric spectrum if it were not that the earth's atmosphere itself contains so much oxygen as to produce such intense oxygen lines that solar effects are unrecognizable. However, photo- graphs of the spectra of the two opposite limbs of the sun show the negative, for in these spectra all solar lines are displaced by Doppler effects, but the well-known oxygen lines show none. Nitrogen, also found plentifully in the earth's atmosphere, behaves similarly. It is a peculiar feature of the solar spec- trum that very few of the so-called negative, or non- metallic, elements are recognized from it. Thus, the
94
THE PHOTOSPHERE
important halogen group of elements, which includes such common elements as chlorine and bromine, is unrecognized. So also with the important element sulphur. These omissions are very remarkable and not yet, I think, well understood.
However, it is found frequently in the laboratory that the spectrum of a mixture or compound of two elements is apt to show one of them predominatingly, or even alone. Especially does a metal thus often ex- clude a nonmetal. But yet oxygen and helium, which, although existing in the sun, are of slight effect in the solar spectrum, are very prominently in evidence in the spectra of many of the stars. Since oxygen is certainly present in sun spots as an oxide, and nitro- gen as cyanogen, though they do not give their char- acteristic lines as elements,1 the other elements just mentioned may also be present in the sun without giving their spectral lines.
Some of the " unknown" lines have now been as- signed to their appropriate elements, but more than half of Rowland's lines are still unidentified. A large number of these are, however, very weak. It is probable that within the next decade many of them will be identified, either with spark or arc spectra.
Corrections to Rowland's Wave Lengths.
It has been shown that the wave lengths assigned by Rowland must be altered. His system is based
1 Three faint lines attributed to oxygen are, however, now known to be solar. See note on preceding page.
95
THE SUN
on measurements by several observers of the wave length of the yellow sodium lines. Measurements by the interferometer in the hands of Michelson, Fabry, Perot, Buisson, and other experimenters have shown that Rowland's assumed wave-length at
D should be reduced by about . This change,
though considerable as wave lengths go, would be of little consequence, if Rowland's system was self- consistent. But it is further shown that the differ- ence from the true scale differs for different parts of the spectrum about as follows:
TABLE VI. Corrections to wave lengths in Rowland's Preliminary Table of Solar-Spectrum Wave Lengths.
Wave lengths |
3000 |
3200 |
3400 |
3600 |
3700 |
3900 |
4100 |
4300 |
4500 |
4700 |
Corrections |
-.106 |
-.124 |
-.148 |
-.155 |
-.140 |
-.144 |
-.152 |
-.161 |
-.172 |
-.179 |
Wave lengths |
4900 |
5100 |
5300 |
5400 |
5600 |
5800 |
6000 |
6200 |
6400 |
6500 |
Corrections |
-.176 |
-.170 |
-.172 |
-.212 |
-.218 |
-.209 |
-.213 |
-.212 |
-.209 |
-.210 |
These discrepancies are to be ascribed largely to certain deficiencies of the grating as a means of measuring wave lengths, and not to avoidable inac- curacy of Rowland's work, although he neglected certain small corrections not strictly negligible. An effort is now (1910) being made, with international cooperation, to establish a consistent and highly accurate system of wave lengths. The results, while not yet officially announced, can hardly differ
96
THE PHOTOSPHERE
appreciably from those indicated in the above table of corrections to Rowland's wave lengths.
An accurate table of solar wave lengths and of the wave lengths of the lines of all the chemical elements constitutes the fundamental groundwork of all mod- ern spectroscopic investigation. What the great star catalogues are to astronomy, the wave length tables are to astrophysics. On them are based in- vestigations of motion and pressure in the sun and stars, of the elements present, the magnetic fields which exist, the possibility of anomalous dispersion phenomena, and other solar and stellar conditions.
LEVELS
In the general spectrum of the solar photosphere we have an index of conditions which exist in a layer practically at the surface of the sun, for, as shown by terrestrial experiments, it takes only a little of an absorbing gas to produce a dark line in the spectrum. But it is thought that a difference of average level exists in the positions of the layers which produce lines of different elements, and even different lines of the same element. The layer of the sun which gives rise to the dark Fraunhofer lines, though thin rela- tively to the solar radius, may yet be thought of as made up of several layers of differing level. Calcium lines are thought to represent a higher level than iron lines, and hydrogen lines one still higher. Yet further, as the longer wave lengths are often more readily emitted by an element than the shorter ones,
97
THE SUN
that is, are emitted at lower temperatures, it may be that a red line of an element on the whole represents a higher level than a violet line of the same element. The continuous background of the solar spectrum represents a lower average level than any of the spec- trum lines, as, of course, follows from Kirchhoff and Bunsen's principle. However, the continuous back- ground offers less opportunities of investigation than the lines, so that less can be learned of the levels it represents than of the so-called " re versing layer" where the lines are formed. The lines them- selves are not to be regarded as dark except by con- trast. If seen against a black ground they would be dazzlingly bright, but, as they are formed in the outer and cooler layers of the sun, they are less bright than the spectrum background against which they are seen. The light of the deeper solar layers cannot get out, if it is of a wave length where great absorption occurs, as is the case in the Fraunhofer lines.
PRESSURES
The effect of pressure is two-fold. It broadens lines and shifts them in wave length. Generally the effect is the same whether a gas is compressed by a like or a foreign mass of gas. Pressure shifts can be distinguished from velocity shifts, because, while the former increase on the whole with increasing wave length, they affect different lines of the same ele- ment and of different elements with shifts of quite arbitrarily differing amounts, and some lines, indeed,
98
THE PHOTOSPHERE
are practically unaffected; velocity causes shifts which differ, it is true, in different parts of the spectrum, but which are directly proportional to the wave lengths. Several investigations have been made to determine the pressures prevailing in the reversing layer. Jewell, in 1896, by examination of grating spectra, found that for most solar lines the wave lengths are greater by a few thousands of an Angstrom than the corresponding lines in the arc spectrum at atmospheric pressure. He found, to be sure, many anomalies which tended to throw doubt on the explanation of these shifts as due to pressure, but the following estimates of the pressure in the reversing layer are given by Jewell, Mohler, and Humphreys :l
ELEMENT. |
Alumi- num. |
Cobalt. |
Silicon. |
Cal- cium. |
Chro- mium. |
Man- ganese. |
Iron, Nickel, Copper, each. |
Pressure .... |
2atm. |
4 |
4 |
6 or 31 |
5 |
6 |
7 |
1 Depending on what group of lines is observed. The H and K lines, however, are not included, nor is 4227.
In 1909 Fabry and Buisson examined numerous iron lines, mostly between wrave lengths 4,000 and 4,500 Angstroms, by interference methods, and dis- covered small shifts in the same sense as found by Jewell. They also investigated the behavior of the anomalous cases and explained them as due to un- symmetrical broadening under pressure. They con-
1 Astrophysical Journal, vol. iii, p. 139, 1896. 99
THE SUN
eluded that the solar reversing layer for iron lines lies under a pressure of 5.5 atmospheres. Evershed, however, criticises their interpretation of the behav- ior of the anomalous lines, and thinks the evidence tends to show that the pressure is less than one atmos- phere.
CONVECTION CURRENTS
Some recent measurements of Adams indicate velocities of ascent of from 0.1 to 0.3 kilometers per second in the solar layer where the metallic absorp- tion lines are formed. This seems at first sight hard to accept, because what goes up must surely come down again, so that we might suppose there would be as much of a Doppler effect of descent as of ascent. But in this connection we must consider the temper- atures of the ascending and descending currents. Adams refers to unpublished experiments of Fox which indicate the brighter areas or " granulations " of the sun's surface as yielding a spectrum strong in " enhanced" or high temperature lines, and the darker spaces or " pores" between as regions of "arc" or low temperature lines. Adams finds the "en- hanced" lines indicate maximum velocities of ascent. He argues that the spectrum would be predominat- ingly influenced by the hotter and brighter parts, and as these are shown to be ascending the whole spec- trum would hence be indicative of ascent. Evershed had advanced a similar argument in 1902 to account for peculiarities of the "flash spectrum."
It is to be supposed that vertical circulation may 100
THE PHOTOSPHERE
be active in the sun because the interior is, of course, hotter than the exterior; the latter is continually being cooled by radiation, and, being thereby made denser, would tend to fall. Velocities of 0.1 to 0.3 kilometers per second are, to be sure, greater than those of any winds we know of on the earth. On the earth, moreover, the vertical circulation and the winds are to a large extent due to the variable tem- perature conditions depending on the changes from day to night anU from summer to winter. As the sun has neither night nor day, summer nor winter, it is to be regarded rather as having approximately reached a steady state of affairs; but still, in con- sideration of the sun's enormous temperature, Mr. Adams' results give no cause for surprise.
St. John has still more recently published a beauti- fully accurate study of the displacements of the cal- cium lines H and K, and of the calcium circulation to be inferred thereby in the sun. He distinguishes three parts of each of these broad lines, which he in- dicates by the subscripts 1 , 2, 3. K3 is the narrow dark line in the center, K2 the bright lines on either edge of K3, and KX the dark, broad, diffuse edges on the out- sides of the K2 regions. Similarly for H, St. John concludes:1 "The calcium vapor producing the ab- sorption band K3 in the solar spectrum has a descend- ing motion over the general surface of the sun of 1.14 kilometers per second in the mean. . . . The cal- cium vapor to which the bright emission line K2 is
1 Contributions of the Mount Wilson Solar Observatory, No. 48. 101
THE SUN
due has an ascending motion over the general surface of the sun of 1.97 kilometers per second in the mean. . . . The wave lengths of K2 (mean of both parts of K2) and K3 reduced to the limb are 3933.667 and 3933.665 respectively. The corresponding wave length in the arc at atmospheric pressure is 3933.667. The mean pressure in the intermediate emitting layer is, therefore, approximately one atmosphere. . . . The shorter wave length of the K3 line may be inter- preted as indicating a somewhat lower pressure in the upper absorbing layer, though the smallness of the quantities involved does not permit a positive con- clusion. ... In the case of the intermediate and highest levels of calcium vapor [there is indicated an] absence of currents of appreciable velocity parallel to the solar surface. ... The widths of the H3 and K3 lines at the center [of the disk], compared with the corresponding widths in the arc, point to an extremely small quantity of the calcium vapor in the upper levels of the solar atmosphere. . . . The average appreciable height of the atmospheric calcium shown by a radial slit is about 5,000 kilometers above the photosphere. The thickness of the upper absorbing layer is approximately 1,500 kilometers. Allowing 700 kilometers for the reversing layer, the emitting layer would have a thickness of approximately 3,000 kilometers. The elevation at which the K line is appreciable is about 500 to 600 kilometers above the level at which the H line ceases to show. . . . The shift between limb and center is 0.015 Angstroms for the
102
THE PHOTOSPHERE
H3 line, and in agreement with that obtained for the K3 line."
An interesting result on the rotation of the sun as measured by the line K3 will be given below.
It is by no means to be supposed that the fact of the enormous transfer of heat from within the gase- ous body of the sun to the exterior, to supply that which is lost by radiation to space, requires us to imagine a strong vertical circulation to carry it on. At low temperatures, as for instance, between a body at boiling temperature and one at freezing, convec- tion is rather more important than radiation as a means of transferring heat ; but this is probably not the case at the temperatures prevailing within the sun. For radiation increases with the fourth power of the temperature, and convection by no means at such a tremendous rate of increase. Hence, as the material of the sun is probably transparent, we must suppose that the heat from within the sun becomes available at the surface to supply the losses of energy by radiation to space chiefly by a process of internal radiation, gradual absorption in a long path outward, and reradiation nearly counterbalancing the absorp- tion. This process is repeated as many times as nec- essary, and except for the very short time occupied by absorption and reradiation, is performed at velocities of nearly 186,000 miles a second, and produces quick communication of energy from within outward.1
'See Schwartz child, "Ueber das Gleichgewicht der Sonnenatmos- phare," Gottingen Nachr., Mathphys. KL, 1906, pp. 1-13. Prof. T. J. J. See also takes this view of the function of internal solar radiation. 9 103
THE SUN
If we admit that Adams has shown an effective velocity of ascent averaging 0.12 kilometers per second and the shifting, thereby, of average solar lines of wave length 4200 by 0.0015 Angstroms toward the violet, then a correction must be applied to Fabry and Buisson's results tending to increase by one atmosphere the supposed pressures in the re- versing layer. Adams has investigated by a purely differential method the shifting of lines between the center and limbs of the sun, and finds that, after correcting his results for this supposed velocity of ascent, there remain in the spectra of the limbs well- substantiated displacements towards the red, which are best explained by ascribing them to effects of pressure.1 Hydrogen, sodium, calcium, and mag- nesium lines show almost no displacement. Lines of titanium, vanadium, and scandium show moderate displacement, and those of iron and nickel consider- able shifts, averaging .007 Angstroms. Lines of the elements of high atomic weight show very small dis- placements, as do also lines strengthened at the limb. Enhanced lines, as a class, show maximum displacements, which apparently grow with the de- gree of enhancement of the several lines. These, at first sight highly discrepant, observations harmonize beautifully under Adams' clever discussion, which we shall reserve till we come to the chapter on solar theory. Adams confirms Fabry and Buisson's ob- servation that the violet edges of lines do not shift.
1 Contributions of the Mount Wilson Solar Observatory, No. 43. 104
THE PHOTOSPHERE
LIMB SPECTRA
The spectrum of the sun's limb is, as would be ex- pected from the general darkening of the sun towards the limb, weaker than that at the center. In violet light eight or ten times as long photographic exposure is required for the limb as for the center. This ratio is reduced to four or five for red light. But, besides this general effect, the Fraunhofer lines are much altered, especially in the violet. The stronger lines almost completely lose their side shadings or "wings" in the limb spectrum, while in sun-spot spectra, as we shall see in Chapter V, the wings have increased prominence. As against this marked difference from spot spectra, the limb spec- trum is like that of spots in having similar changes of relative intensity of lines, so that lines strengthened in spots are strengthened, though in less degree, at the limb, and vice versa. As in the spots, the so- called spark or "enhanced" lines are often weakened at the limb. The H> line of hydrogen, on the con- trary, is widened and perhaps strengthened at the limb, although narrowed and weakened in spots.
VARIATION OF THE SUN'S BRIGHTNESS
The variation of the brightness of the sun from the center to limb is much more readily determined by the bolometer than by the photographic plate. Fig. 25 shows the distribution of brightness along a
105
THE SUN
diameter of the sun's disk for rays of different wave lengths. The reader will notice how great the con- trast in brightness between center and edge is for the shorter wave lengths. This fact is also shown
\
11
FIG. 25. — BRIGHTNESS ON SOLAR DISK
by the following table, which gives the brightness at different percentages of a solar radius from the center of the solar disk, and with which the data of
Figs. 25 and 26 agree.
106
THE PHOTOSPHERE
TABLE VII. — Distribution of radiation over the sun's disk.
Fraction |
|||||||||
xvaaius |
0.00 |
0.40 |
0.55 |
0.65 |
0.75 |
0.825 |
0.875 |
0.92 |
0.95 |
Wave length |
|||||||||
i t |
|||||||||
Oft. 323 |
144 |
129 |
120 |
112 |
99 |
86 |
76 |
64 |
49 |
0.386 |
338 |
312 |
289 |
267 |
240 |
214 |
188 |
163 |
141 |
0.433 |
456 |
423 |
395 |
368 |
333 |
296 |
266 |
233 |
205 |
0.456 |
515 |
486 |
455 |
428 |
390 |
351 |
317 |
277 |
242 |
0.481 |
511 |
483 |
456 |
430 |
394 |
358 |
324 |
290 |
255 |
0.501 |
489 |
463 |
437 |
414 |
380 |
347 |
323 |
286 |
254 |
0.534 |
463 |
440 |
417 |
396 |
366 |
337 |
312 |
281 |
254 |
0.604 |
399 |
382 |
365 |
348 |
326 |
304 |
284 |
259 |
237 |
0.670 |
333 |
320 |
308 |
295 |
281 |
262 |
247 |
227 |
210 |
0.699 |
307 |
295 |
284 |
273 |
258 |
243 |
229 |
212 |
195 |
0.866 |
174 |
169 |
163 |
159 |
152 |
145 |
138 |
130 |
122 • |
1.031 |
111 |
108 |
105.5 |
103 |
99 |
94.5 |
90.5 |
86 |
81 |
1.225 |
77.6 |
75.7 |
73.8 |
72.2 |
69.8 |
67.1 |
64.7 |
61.6 |
58.7 |
1.655 |
39.5 |
38.9 |
38.2 |
37.6 |
36.7 |
35.7 |
34.7 |
33.6 |
32.3 |
2.097 |
14.0 |
13.8 |
13.6 |
13.4 |
13.1 |
12.8 |
12'. 5 |
12.2 |
11.7 |
Wave length |
|||||||||
of Max. |
Oft. 458 |
Oft. 467 |
Oft. 471 |
Oft. 474 |
Of*. 478 |
Of*. 483 |
Of*. 489 |
Oft. 496 |
Oft. 505 |
Following the lines of the table from left to right, the reader may note the decrease of brightness from the center of the sun to 95 per cent of the radius out- ward1. The results are arranged vertically in order of wave length, and the numbers have been so ad- justed that, by taking any single vertical column, as for instance, that for 75 per cent, out on the radius, the reader may find for a single zone of the sun the distribution of brightness on a uniform scale of wave lengths for the normal spectrum outside the earth's atmosphere. The data as regards distribution along the radius for wave length 0.323^ are from results
1 There is a tendency of all the data plotted in Fig. 25 to show a less rapid fall of brightness from 95 to 97 per cent out on the radius, than would be expected. This may be due to error.
107
THE SUN
of Schwartzchild and Villager, who obtained them by photographing the solar image formed by a silvered lens. The remainder of the data are from the bolo- metric results of Abbot and Fowle.
In the preceding table the maximum number in each vertical column is indicated by black-faced type. But the wave length intervals are not small enough to show accurately in this manner the amount of shifting of the wave length of maximum radiation for light coming from greater and greater distances from the center of the sun's disk. By means of plotting the values, we find that the true wave lengths of maximum intensity are as given in the lower line of the table. This shows a shifting of the maximum of radiation from 0.458/* at the center of the sun's disk to 0.505/* at 95 per cent out on the radius. We shall see that a similar shifting of the wave length of maximum radiation occurs between the photosphere and the umbra of a sun spot. The dotted curve of the accompanying Fig. 26 shows the distribution of radiation in the spectrum for light of the whole sun's disk as it would be if viewed outside the earth's atmosphere. Similar curves are given also in Fig. 26 for the center of the sun's disk and for points 55, 82.5 and 95 per cent of the radius towards the limb. No account is made in the figures of the Fraunhofer lines sepa- rately, although collectively they doubtless affect the forms of the curves, especially for the shorter wave lengths.
108
THE PHOTOSPHERE
SOLAR TEMPERATURES First Method.
These five energy curves of Fig. 26 are of inter- est as they indicate the probable temperatures in
;
OS O6 Q7 Q8 Q9 1.0
_I5 1.6
FIG. 26. — ENERGY SPECTRA ON SOLAR DISK.
the photosphere. From Wien's displacement law ( xmax. T = 2930) given in Chapter II, we may find,
by substituting the values indicated for
109
the
THE SUN
values of the absolute temperatures for which a per- fect radiator would give the same wave lengths of maximum radiation. The values are given in Table VIII.
Furthermore, as the five curves of Fig. 26 are plotted with ordinates proportional to intensities, and abscissae proportional to wave length, their in- cluded areas are proportional to the intensities of the emission of all wave lengths combined, as emitted from the selected regions of the sun's disk. If the total emission is comparable to that of a perfect radiator, then, by Stefan's law, it is proportional to the fourth power of the temperature of the emitting body. Hence, the fourth roots of the areas included by the five given curves should be in inverse ratio of the wave lengths of maximum emission. The fol- lowing table shows in its fourth and sixth lines how the matter comes out:
TABLE VIII. — Energy spectrum relations over the sun's disk.
POSITION. ~* |
Whole Disk. |
Center. |
55% |
82.5% |
95% |
Wave lengthof max- imum |
0.^468 |
O.Ai458 |
0./i471 |
0./I483 |
0.M505 |
2930 |
6260° |
6400° |
6220° |
6070° |
5800° |
Xmax. |
|||||
Ratios by maximum |
1.079 |
1.104 |
1.073 |
1.047 |
1.000 |
Ratios of Areas .... |
1.407 |
1.620 |
1.476 |
1.249 |
1.000 |
Ratios by fourth roots of areas ... |
1.090 |
1.128 |
1.102 |
1.057 |
1.000 |
'On the absolute scale of Centigrade degrees water freezes at 273° and boils at 373°.
110
THE PHOTOSPHERE
The greatest disagreement between the ratios through maximum and through the fourth roots of the areas is about 2J/£ per cent.
Second Method.
Another method of estimating the probable solar temperature is by attempting to match, as well as possible, the distribution of energy in the whole range of solar spectrum with the distribution compu- ted by the Wien-Planck formula given in Chapter II. Referring to Fig. 17, the reader will find in curves B and A the distribution according to Wien- Planck in the spectrum of a perfect radiator at 6200° and 7000° C. absolute, and also in curve C the energy spectrum for the general solar surface. No account is made in the computations of the relative values of the constant CA and the solar constant of radiation. The 6200° curve has been repeated at B7 on a larger scale of ordinates, and the observed curve also repeated at C' on a scale nearly matching that of B'. The observed curve falls below the com- puted ones in the ultra-violet, but this discrepancy is to be expected, partly because the ultra-violet solar spectrum is crowded with lines of selective absorp- tion.
On the other hand, the observed curve rises above the computed ones in the infra-red, a feature to which Professor Bigelow has repeatedly called attention. It has just been said, and it will be spoken of at greater length in Chapter VI, that the rays from the
111
THE SUN
center of the sun's disk seem to arise from a source at higher temperature than those emanating from the sun's limb. In accordance with the explanation of this phenomenon which will be advanced in Chapter VI, it would be expected, also, that solar rays of long wave lengths would appear to come from sources of higher temperature than would those of shorter wave lengths. If so, we shall thereby understand why the infra-red parts of curves C and C' (Fig. 17) rise above curves A, B and B', respectively, for curves C and C' do not represent the spectrum of a source at constant temperature. Their infra-red parts correspond to much hotter sources than do their visible and ultra-violet parts.1 It is evident, however, that the 7000° curve, except in the ultra- violet, is a better match for the observations than the 6200° curve. The large discrepancy in the ultra- violet is probably due in part to the general tendency toward lower temperature in the sun for short wave length rays, but far more to the throngs of Fraunhofer lines in that region of spectrum, which though not shown separately, very greatly affect the form of the curve.
Third Method.
Pointing seemingly to a lower solar temperature than those we have considered are the following
1 The accuracy of the observed curve for wave lengths beyond 2/i is seriously impaired by the effect of terrestrial water vapor, so that no conclusion should be drawn from the fact of the falling off of the curve in this region.
THE PHOTOSPHERE
facts. As recently done for a large number of stars by Wilsing and Scheiner, we may compute the ap- parent temperature of the sun by the formula:
where E! and E2 are the intensities of energy at two wave lengths \ and \, c2 a constant for which Wilsing and Scheiner prefer the value 14200, and T the absolute Centigrade temperature. Taking a number of values of the intensity within a given range of wave lengths, and proceeding according to the method of least squares, 1 find:
Wave length range |
0.^30- Q.ftSQ |
0./i35- 0.^50 |
0./*50- 0.^70 |
0.^80- l./*50 |
l./iOO- l./xSO |
l./dO- l./i50 |
Temperature . . . |
3932° |
5142° |
6900° |
4493° |
4006° |
3840° |
The falling off of computed temperatures for long wave length rays is due to the fact that the observed curve of Fig. 17 rises less rapidly from the infra-red towards shorter wave lengths than does the 6200° curve, and far less rapidly than the 7000° curve. But, as we have said, and in accordance with a line of explanation to be given in Chapter VI, we may assume that as the wave length decreases the effec- tive source of radiation approaches the exterior of the sun, and, therefore, is cooler. Hence, although the effective temperatures of emission for the infra- red rays are probably exceeding 7000°, the observed energy curve does not rise towards its maximum from the infra-red side as fast as does the 7000°
113
THE SUN
curve, because each successive shorter wave length is emitted from a lower average temperature than its next longer neighbor, and is, therefore, less intense than it would otherwise be. In the ultra-violet, however, we may consider the temperature of effec- tive emission not only apparently, but really far be- low 7000°, on account of the superficial region of its origin.
On the whole, the preceding review of the form of the solar energy curve inclines us to set the average temperature of the photosphere certainly above 6200°, and possibly near 7000°.
Fourth method.
It will be shown in Chapter VII that the intensity of solar radiation at the earth's mean distance from the sun is 1.95 calories per square centimeter per minute. From Stefan's law, with Kurlbaum's con- stant (see Chapter II), a perfect radiator emits radi- ant energy from each square centimeter of its surface at the rate of 76.8 X 10~12T4 calories per minute. The radius of the sun being 696,000 kilometers, and the mean radius of the earth's orbit 149,560,000 kilo- meters, we would have the following equation for a perfect radiator of uniform absolute temperature T in the sun's place :
(696,000)2 X76.8 X 10'12T4 = (149,560,000)2 X 1.95
From this, T = 5860° absolute C. As this value falls below those obtained previously, we may suppose the
1 1 I
CALCIUM SPECTROHELIOGRAM, H2. (Ellerman.) 1908, April 30. G. M. T. 12h 53m. P. S. T. 4h 43m P.M.
THE PHOTOSPHERE
sun's constant of emission is a little less than that of a perfect radiator.
An observation which may be regarded as con- firmatory of the view that the photosphere falls some- what short of perfect radiating power is stated by Jewell as follows : 1
" When some of the very best negatives of the solar spectrum are carefully examined, it is found that some of the sharp-edged, clean-cut, and unshaded lines of iron, chromium, manganese, titanium, etc., have a faint, dark shading just outside the edge of the line. It is very faint and difficult to observe (only slightly darker2 than the general background of the solar spectrum), but it is not due to contrast, as it is not always present. It is a difficult observation to make, but was observed sometime before the explan- ation forced itself upon me. The correct explanation undoubtedly is that this faint, dark shading (dark in the negative [overbright in the spectrum]) is the re- mains of an emission line, either produced at the photosphere or lower down in the solar atmosphere than the absorption line. "
This interesting observation, which has been con- firmed by Evershed, appears to indicate that the photospheric radiation in general, though undoubt- edly coming from hotter, because deeper, layers than the rays within the influence of the Fraunhofer lines, yet lacks something of the full intensity of perfect or
1 Astrophysical Journal, vol. Ill, p. 99, 1896.
2 Darker in the negative, brighter in the spectrum.
115
THE srx
" black-body" radiation. For thus it might occur that deep-lying (yet not the deepest lying) metallic vapors would give in the immediate proximity of their lines of powerful selective emission a more in- tense radiation than the deeper lying and hotter, but intrinsically less strongly emissive, layers of the photosphere.
Summary.
In all of these ways discussed of estimating the solar temperature, we have to go on the hypothesis that the sun is a perfect radiator. This is, of course, very unlikely, but if the sun's radiating power is not perfect, then its temperature must, at any rate, exceed that (5860° abs.) calculated by the fourth method from Stefan's law of radiation. It is scarcely less probable that the solar temperature exceeds that (6260° abs.) calculated by the first method through Wien's displacement law. For the influences tend- ing to distort the form of the solar spectrum energy curve seem to be of a kind to diminish the violet most, and thereby to shift the maximum of energy towards the red. Hence, we conclude that there is a high probability that the average temperature of the apparent photosphere exceeds 5860° or even 6260° of the absolute Centigrade scale, and may be as high as 7000° absolute Centigrade.
The reader may be disposed to question whether a difference of temperature probably exists between the center and edge of the apparent photospheric disk,
116
HYDROGEN (Ha) SPECTROHELIOGRAM. (Ellerman.) 1908, April 30. G. M. T. 13 h 6 m. P. S. T. 5 h 6 m P. M.
THE PHOTOSPHERE
as brought out in Table VIII, but this matter will be further discussed in Chapter VI. One highly in- teresting conclusion seems to follow from the fact of the enormously high temperature of the photosphere, taken in connection with the spectroscopic proof of moderate pressures in the reversing layer. This con- clusion is that no known substances can exist in the photosphere except as gases.1 It has generally been held that the photosphere is a cloudy layer. If so, the materials composing the clouds are not known to exist on the earth.
THE SPECTROHELIOGRAPH
When we examine the sun visually or by direct photography, the source of the light is highly com- plex. Many chemical elements, existing in a layer many hundreds, or perhaps thousands, of miles deep take part in sending the light. After tentative trials in the early days of the spectroscope and of photog- raphy, the matter of obtaining a view of the sun in the light of one element, and substantially at one level, was taken up about 1890 by Hale and by Deslandres independently, and in 1891 Hale first employed his spectroheliograph. Deslandres has long used a simi- lar principle, but with intermittent instead of con- tinuous displacement of the view over the solar sur- face, in his "spectroscope a vitesse." He has lately employed the spectroheliograph itself with great success. The spectroheliograph, as explained in
1 See also Chapter VI. 117
THE BUN
Chapter II, is in effect, a screen which cuts off all light except that of a single spectral line, and enables the observer to see how the vapor of a single element lies on the sun's surface.
We shall now examine some beautiful spectroheli- ographic results obtained on Mount Wilson by Mr. Ellerman, which Mr. Hale has kindly allowed me to reproduce here. Plate V is taken with the spectro- heliograph in the H2 line of calcium.1 Comparing it with the direct photograph of the sun taken on the same day, shown in Plate III, at the beginning of this chapter, there is seen a greater distinctness and prominence of detail. Hale has called the mo tt lings shown by the spectroheliograph "flocculi, " and dis- tinguishes between bright and dark flocculi. A pho- tograph through the Ha (C) line of hydrogen, made within a few minutes of Plate V, is given in Plate VI. The hydrogen flocculi are generally of more well- defined shapes than the calcium flocculi, and usually dark where these are bright. Bright hydrogen floc- culi, however, often appear in sun spot and active regions, and such bright flocculi frequently change in form with eruptive rapidity.
In a broad line, like the H or K lines of calcium, the slit of the spectroheliograph may be set in several positions. Hale distinguishes three such, which he terms, H!, H2, H3 or Kb K2, and K3. In an eclipse
1 The faint structure of parallel lines seen on all spectroheliographic plates is not a solar feature, but is caused by very slight irregu- larities of the motion of the instrument.
118
THE PHOTOSPHERE
photograph of the chromosphere with radial slit (see Chapter IV) the H and K lines have frequently an " arrow head" appearance. That is: The light of the center of H or K is found at a high level above the sun, and the matter which produces the light of the edges, or wings, does not extend out so far. H3 or K3 corresponds to the center of H or K (seen as a dark line in the solar spectrum). Thus, when we look at a K3 spectroheliographic plate, there is a deep layer of calcium vapor behind the regions shown, and, as it takes but a small portion of this thickness to cut off by absorption the light of this wave length, our view is of the highest levels where cal- cium occurs. The K2 and K! positions on the sides and extreme wings of K, respectively, correspond to moderate and low level calcium distribution. In the spectrum of hydrogen a similar difference of effective level in spectroheliograph observations is attained by employing lines of different wave lengths. In eclipse observations high hydrogen prominences are red, owing to the predominance in their light of rays of the Ha (C) line. Hence, pho- tographs taken through the Ha (C) line give high level phenomena, and, as might be plausibly inferred from a consideration of Wien's displacement law, the hydrogen lines of successively shorter wave- lengths would be most copiously emitted at hotter, and hence lower levels. We then regard an Ha or K3 photograph as a high-level, an H£ or K2 as a medium, and an H7 or Kj as a low-level phenomenon 10 119
THE SUN
for hydrogen and calcium respectively. However, these gases are both high-level gases on the sun, and the photographs of the sun through their lines are above the levels where most Fraunhofer lines are produced. It is to be expected that when, with in- creasingly powerful instrumental appliances, the spectroheliograph can be employed in the narrower, and, therefore, more difficult, lines of the heavier and less easily vaporized elements, the conditions at lower levels will be shown.
The following illustrations bring out the differences due to level in a striking manner. Unfortunately, it was not possible for Mr. Ellerman to furnish me a series showing all the different kinds of spectrohelio- grams above mentioned for a single day, and, indeed, it was found necessary to omit altogether an example of H3 in the calcium series. Plates VII and VIII show a spotted area of the sun's surface as it appeared July 16, 1907, in Hx and H2 calcium spectrohelio- graphic exposures. Plates IX, X, and XI illustrate a spotted region of the solar surface as it appeared September 10, 1909. They are taken in H2 of cal- cium, HY and Ha of hydrogen, respectively. In this latter series the first plate gives no hint of the pronounced vortical structure revealed by the high level hydrogen in the last plate. One is struck by the similarity of these curved structural forms to the lines-of-force diagrams given by the familiar experi- ment of shaking fine iron filings on a glass plate held horizontally over a couple of magnets. In Chapter
120
THE PHOTOSPHERE
V we shall have occasion to refer again to Plate XI when we come to deal with the magnetic character of the sun spots.
The spectroheliograph results will receive further attention in Chapter IV in connection with the study of solar prominences. These objects are great flamelike protuberances which extend for thousands, sometimes hundreds of thousands of miles above the photosphere. First observed at eclipses, the fact that they shine principally by the bright spectrum lines of calcium and hydrogen made it possible to see them at the sun's limbs at all times with the spec- troscope, and now the spectroheliograph has enabled us to recognize them frequently as dark hydrogen flocculi on the disk itself. A view of the sun through the Ha (C) line is best adapted for this purpose, and, indeed, it may well be said to reveal the sun in quite a new aspect. Direct photographs and spectrohelio- graphic results through Hx (C) and H and K all show a mottling of the solar surfaces, but in Plate XI the mottling, especially in the neighborhood of sun spots, shows a marked tendency toward curved and spiral forms, as if the hydrogen at this high solar level were definitely arranged by cyclonic mo- tions. Still there are not usually found observable motions along these curved lines, although in ex- ceptional cases series of Ha spectroheliographic plates have given evidence of definite and very rapid motion. Thus St. John observing on Mount Wilson on June 3, 1908, photographed a hydrogen flocculus,
121
THE SUN
probably a prominence, apparently moving 105,000 kilometers (60,000 miles) in 18 minutes towards a double sun spot. When near the spot the flocculus divided, and apparently each branch was sucked into a sun spot. The apparent motion in this case was almost exactly radial to the sun spot pair. A dark flocculus of a similar type, which is also probably a prominence, is seen in Plate VI.1
THE SOLAR ROTATION
The rotation of the sun has been measured by ob- serving the march of sun spots, faculae, and, of late, spectroheliographic flocculi across the disk. The classical researches of . Carrington and of Spoerer on the march of sun spots showed :
(1) That the sun rotates about an axis inclined about 7° to the plane of the ecliptic, and so that the sun's axis points midway between the polar star and Vega to a position in right ascension 18h 44m and declination 64°.
(2) At the solar equator the rotation occurs in about 25 days.
(3) The period of one rotation increases on either side of the equator about equally, and is about 27J/2 days at 45° north or south solar latitude.
(4) Individual sun spots drift in different directions on the sun's surface, so that it is only the mean re-
1 An interesting conclusion relating to the part played by eruptive prominences in the life history of sun spots is quoted in Chapter V from Bpectroheliographio observations of Fox.
122
THE PHOTOSPHERE
suit of the motions of many spots which can give accurately the solar rotation period.
(5) The daily rate of solar rotation, and the fact of different rotation periods for different solar latitudes, were both expressed by Carrington in the following formula, in which X is the daily rate of rotation, I the solar latitude :
X = 865' -- 165' sin1 I
Faye assuming on theoretical grounds that the exponent of sin I should be 2, derived from Carring- ton's observations of 1853-1861 the expression:
X = 862' - 186' sin2 1.
Spoerer, from observations of his own between 1862 and 1868, combined with those of Secchi and others, obtained:
X = 1011' - 203' sin (41° 13' + /) Tisserand from observations of 1874-1875 ob- tained :
X = 857.6' - 157.3' sin2 1.
Wilsing and later Stratonoff have determined the solar rotation from observations of faculae. As these objects can seldom be followed much more than a quarter way across the solar disk, and as their appearance is usually altered when they reappear on the other limb, the results have less weight than those obtained by sun-spot observations. Wilsing found no evidence of equatorial acceleration, but Stratonoff found from the faculae similar results to those of Carrington and Spoerer on sun spots. Very
123
THE SUN
recently Chevalier has published results of a long and excellent series of determinations of the solar rota- tion by measurements of faculae. His work con- firms that of Stratonoff.
In 1908 Hale published determinations of solar rotation from spectroheliographic plates of the hydro- gen and calcium flocculi, taken through the HS and H2 lines respectively. His results with H2 calcium flocculi are in close agreement with those obtained by Fox in 1903-4 for the same line. Their results agree, also, at all latitudes with the rates of solar rotation derived by various observers from observa- tion of sun spots. With H8 hydrogen flocculi, the rate of equatorial rotation was about the same, but there was found no retardation at higher latitudes, a fact of high interest and significance.
According to Doppler's principle the spectral lines of a source receding must be displaced towards the red with reference to those of a source approaching the observer. By forming the solar image with a telescope, and reflecting light from the two limbs si- multaneously upon the slit of a spectroscope, two spectra may be produced, one immediately above the other, which exhibit at a glance the shifting of all solar lines owing to the sun's rotation. See Plate IV. Atmospheric lines are not thus shifted.
In this way the rate of solar rotation has been de- termined with great accuracy by Duner, Halm, and lately by Adams. Their results bring out clearly the fact discovered by Carrington from the study of sun
124:
fij
THE PHOTOSPHERE
spots, namely, that the sun's angular rotation is slower at high latitudes than at the sun's equator. Since sun spots do not occur near the sun's pole, this peculiarity could not be thoroughly studied by Car- rington, but by the spectroscopic method the solar rotation period has been determined at high as well as low latitudes. There might easily be a doubt whether the spectroscopic results on solar rotation ought to agree with those obtained from observing the sun spots, the faculae, and the flocculi exposed by the spectroheliograph, for the sun, as indicated by several lines of evidence, is at so high a tempera- ture as to be probably almost wholly gaseous, and our vision may penetrate to some distance below its surface. The several objects which are sources of the light phenomena employed for the different methods of studying the solar rotation may lie at different levels, and may, therefore, move at different rates. Accordingly it is interesting to compare, in the fol- lowing table, the rotation periods indicated by the several visual methods and by the spectroscopic ob- servations of lines of different chemical elements. The table is compiled from those given by Hale l and by Adams.2
According to the results referred to in the following table, the sidereal rotation of the average solar sur- face is completed in about 24.6 days at the equator, 26.3 days at ± 30° latitude, 31.2 days at ± 60°,
1 Contributions of the Mount Wilson Solar Observatory, No. 25.
2 Ibid., No. 33.
125
THE SUN
TABLE IX. — Daily rotation of the sun's surface, Various methods of observing.
object observed. -» |
Sun spots. |
Faculse. |
Ca. Flocculi H • line. |
H. Flocruli H5 line. |
Many spec- tral linos. |
Observer. -» |
Mean of Carrington, Spoerer, Maunder. |
Mean of Stratonoff and Chevalier. |
Mean of ll:i!f and ' Fox. |
Halo. |
Adams, 100S. (Doppler effect.) |
Latitude. ; |
|||||
0° to ± 5° |
14.40° |
14 . 56° |
14.54° |
14.3° |
14.59° |
± 5 ±10 |
14.35 |
14.52 |
14.41 |
14.4 |
14.48 |
±10 ±15 |
14.25 |
14.33 |
14.30 |
14.6 |
14.33 |
±15 ±20 |
14.13 |
14.21 |
14.13 |
14.5 |
14.15 |
±20 ±25 |
13.98 |
14.19 |
13.99 |
14.7 |
13.95 |
±25 ±30 |
13.80 |
14.04 |
13.97 |
14.7 |
13.74 |
±30 ±35 |
13.60 |
13. 601 |
13.75 |
14.9 |
13 . 50 |
Stratonoff only.
Adams' spectroscopic results, including high latitudes.
Chemical elements. -» |
Many. |
'ttSy?" |
Fe, Ti, TiFe. |
Mn, Fe, Fe. |
Ca. |
H. |
Wave |
4196.699 |
4265.418 |
4257.815 |
4226 .91 |
6563 . 054 |
|
lengths. -» |
Many. |
4197.257 |
4287 .566 |
4290.542 |
(Ha or C). |
|
Latitude. ; |
4216.136 |
4288.310 |
4291 .6-30 |
|||
0.3° |
14.65° |
14.49° |
14.65° |
14.72° |
15.0° |
15.2° |
14.9 |
14.28 |
14.21 |
14.31 |
14.34 |
14.9 |
15.0 |
29.7 |
13.66 |
13.49 |
13.65 |
13.74 |
14.2 |
14.6 |
44.7 |
12.81 |
12.74 |
12.85 |
12.95 |
13.6 |
14.0 |
60.0 |
11.52 |
11.35 |
11.53 |
11.62 |
12.5 |
13.7 |
74.9 |
10.84 |
10.50 |
10.93 |
11.04 |
13.1 |
14.3 |
and 35.3 days at ± 80°. The agreement between Adams' and Duner's work, done in different years, is so exact that there seems little reason to suspect a secular variation of the retardation towards high latitudes. Adams finds his mean results and those of Duner and Halm well expressed by the following formula :
? = 10°.62 + 3°.99cos2 </>. 126
THE PHOTOSPHERE
Where £ is the angular sidereal rotation per day and <#> the solar latitude.
The highly interesting change in the observed rotation period for lines of different chemical ele- ments is regarded as indicating differences of effec- tive level of the production of the Fraunhofer lines. The results in this direction gain added value be- cause they agree with several other lines of evidence that point to the same conclusions. The whole subject will be discussed in Chapter VI.
Very recently St. John has determined the rate of daily rotation spectroscopically in the K3 line of calcium.1 He says: " The angular velocity of the high-level calcium producing the absorption line K3 is nearly constant for the latitudes of observation, being 15°.5 and 15°.4 per day at the latitudes 6°.6 and 38°. 4, respectively. The corresponding values de- duced from Adams' results are 15°. 1 and 14°.3 for hydrogen, and 14°.4 and 13°. 2 for the reversing layer. The high velocity of the calcium vapor producing the K3 line points to a higher elevation of this layer of calcium vapor than of the hydrogen effective in the production of the Ha line." It is a very singular thing that calcium occurs at such very high levels in the sun. We shall see the fact confirmed in the next chapter, but the reason for it is one of those many puzzles which whet the appetite of the student in solar research.
1 Contributions of the Mount Wilson Solar Observatory, No. 48. ^ 127
CHAPTER IV
ECLIPSES AND THE OUTER SOLAR ENVELOPES
The Saros. — Eclipse Expeditions. — The Corona. — The Chromo- sphere.— The Eclipse of 1868 and Jansen's and Lockyer's Discovery. — Spectrum of the Chromosphere and Prominences. — Prominences and the Spectroheliograph. — Recent Flash Spec- trum Observations. — The Heights of Different Metals in the Chromosphere. — Mitchell's Observations of 1905. — Campbell's Observations. — Chromospheric Spectra in Full Daylight.
WHEN the moon passes directly between the earth and the sun it sometimes completely covers the latter, and there is a total solar eclipse. At such times the brilliant glare of day ceases for a few mo- ments to illuminate our atmosphere, and in the semi- darkness we may see the objects which closely sur- round the sun. Total solar eclipses occur almost every year, but as the moon is never much greater in angular diameter than the sun, the area of the earth's surface on which the eclipse appears total at a given instant is rarely greater than 100 miles in average diameter. The rapid motion of the moon, though partly offset by the rotation of the earth, hurries the region of totality along faster than 1,000 miles an hour, making a belt seldom wider than 100 miles, but sometimes more than 5,000 miles long, on which the eclipse is total at sometime between sunrise and sun-
128
ECLIPSES AND SOLAR ENVELOPES
set. Over enormous areas on either side of the line of totality the sun is partially eclipsed, and appears for some hours as a crescent figure.
THE SAROS
The ancients discovered a cycle of eclipses called the Saros, which indicates approximately the times when solar eclipses will occur. In 223 synodic months there are almost exactly nineteen " eclipse years" of 346.62 days, the interval between the times when the sun in its apparent annual path crosses the two nodes of the moon's orbit. Hence, if we count forward 6,585 days, or 18 years 11 days, from one total eclipse, we are apt to find the occurrence of another either partial or total. A family of eclipses thus occurs, separated by intervals of about eighteen years. Such a family generally numbers about sixty- five or seventy eclipses, of which perhaps eighteen will be total, and the rest annular or partial. Many total eclipses are visible only at regions unfavorable for observation, such as oceans, the polar regions, or very cloudy localities. As totality at a given place never lasts more than eight minutes, and generally does not exceed three, there have been hardly more than a couple of hours of time employed in total solar eclipse observing in the last half century. Yet so well have the moments been utilized that a large stock of information has been gathered.
Not infrequently eclipse expeditions have led as- tronomers to experiences of hardship, disappointment
129
THE SUN
and in one instance to death. Father Perry, of Stony- hurst, who led the English eclipse expedition to Cay- enne, in 1889, was taken ill before the awaited day. He insisted on observing, supported by an attendant, and called for three cheers when the eclipse had been successfully observed, saying: "I can't cheer, but I will wave my helmet!" A few days later he died at sea.
A total eclipse having been predicted to occur at a certain place favorable for observing, astronomers journey there several weeks in advance, equipped with photographic telescopes, spectroscopes, auxil- iary apparatus, and supplies. The instruments are set up and carefully adjusted and are provided with every possible contrivance to facilitate and shorten their operation at the critical moment. Rehearsals of the eclipse begin as soon as possible. Time signals are counted off, photographic apparatus is manip- ulated, and the whole program is gone over and over again, just as if the totality were on. In this way the observers try to anticipate all possible contin- gencies, and acquire skill and rapidity in performing their parts. Mr. Langley used to say that if a pin were likely to be dropped during the eclipse the ob- server should practice dropping one and filling its place at rehearsal. The hour, minute, and second of the eclipse are predicted long in advance, so that on the appointed day all is prepared for action at a well known time. At first contact of the moon a notch begins to appear in the sun's disk, and this grows
130
ECLIPSES AND SOLAR ENVELOPES
larger and larger during the next hour and a half, until only a narrow crescent remains. This hour and a half has always seemed to the writer the saving element; for during its slow passage the unhurried march of events tends to calm the nervous agitation which comes on with the first contact, when one feete that his opportunity is now or never. As the cres- cent becomes thin the sun's light becomes noticeably weak and yellow, for only the limb now remains vis- ible, and its light, as stated already, is very much weaker, especially in the violet end of the spectrum, than the light of the center of the disk. Just before totality flickering bands, called " shadow bands," steal rapidly along the ground, and then, as the last crescent of the photosphere suddenly vanishes, a thin ring of rosy light encircles the moon, and beyond this, for perhaps one or even two diameters of the sun, blooms forth the pearly hued corona,
THE CORONA
There is a cycle of changes in the form of the corona having a period of about eleven years, sup- posed to be identical with that of sun-spot frequency, which will be noticed in Chapter V. As the corona can be observed only at total solar eclipses, the march of the cycle of changes is as yet only imperfectly known, but for the last half century it has been observed that there are long equatorial coronal streamers at the time of sun-spot minimum, while at maximum of sun spots the corona extends only
131
THE SUN
to moderate distances, but nearly uniformly in all directions from the sun. The accompanying views, Plate XII from a drawing of Mr. P. R. Calvert pre- pared from the Yerkes Observatory photographs of the 1900 eclipse, Plate XIII from a drawing of Mrs. C. G. Abbot prepared from U. S. Naval Observa- tory photographs of the 1905 eclipse, illustrate the characteristic forms of the corona at sun-spot minima and maxima respectively.
Many efforts have been made, but thus far with- out success, to devise a method of observing the corona without an eclipse. Success is unlikely, for in its brightest parts, even within TV radius of the sun's limb, the brightness of the corona is only about one-tenth as great as that of the daylight sky at 20° from the sun, if viewed from sea level. Close to the sun the daylight sky is many fold brighter still, so that the coronal brightness is insignificant in comparison with it. By ascending a very high mountain, it is true, a considerable gain might be made, for the corona wrould be a little brighter and the sky several fold less bright, but the brightness of the sky would still be far too great to permit the corona to be seen, even in its brightest parts, by any contrivance yet devised.
The corona fades rapidly with increasing distance from the sun. According to Turner, who has dis- cussed results of various eclipses, it falls off approxi- mately as the sixth power of the distance from the sun's center. L. Becker has discussed photographic
132
ECLIPSES AND SOLAR ENVELOPES
observations made at the eclipse of 1905, and gives the following formula of distribution of the intensity of the blue and violet coronal radiation at different distances, H, from the sun's limb; I is the intensity, C is a constant, and H is expressed in thousandths of a solar diameter.
I = C (H + 140) -4.
At the eclipse of January 3, 1908, the present writer, assisted by A. F. Moore, made bolometric observa- tions of the intensities of the coronal radiation at several distances from the sun's limb. These were made both with and without a screen of asphaltum varnish on glass. This screen was used to cut off the visible spectrum while still transmitting the infra-red. The following is a comparison of these results with those computed according to the for- mulae of Turner and of Becker.
H = |
45. |
121. |
364. |
Total radiation Visible radiation . . . |
100 100 |
29.9 29.8 |
0 0 |
Infra-red radiation Computed via Becker Computed via Turner |
100 100 100 |
30.1 25.2 45.7 |
0 1.8 6.3 |
The agreement between the bolometric observations and the computation by Becker's formula is pretty good, so that for a sun-spot maximum corona it seems to represent the distribution for all kinds of radiation, at least in the inner corona.
The light shows distinct radial polarization in the 133
THE SUN
outer corona, but the percentage of polarization de- creases, and at length vanishes near the limb of the sun. Polarization of the coronal light is generally interpreted as evidence of the presence of reflected photospheric rays in the coronal brightness, just as sky-light and its polarization is produced by the dif- fuse reflection of sunlight in the air. Some writers have inferred from the absence of polarization near the sun's limb that the light of that part of the corona contains almost no reflected photospheric rays. But a particle near the sun's limb must be shone upon from every direction within a hemisphere, so that the light which it reflects, being partially polarized in every plane, would show polarization in none. Hence, the absence of reflected photospheric rays from the inner coronal brightness cannot rightly be inferred from the absence of polarization.
The spectrum of the corona is more nearly con- tinuous than that of the photosphere. ,A few bright lines are found, but these are not conspicuous at most eclipses. There is a famous bright coronal line in the green at wave length 5303. This line was dis- covered by Young, in 1870, and it has been seen with more or less distinctness at many subsequent eclipses. It does not correspond in wave length to a line of any known substance, or to a photospheric line, so that it is ascribed to a hypothetical element "coro- nium." As the element helium was found in the earth after its spectrum had long been known in the sun and stars, so it may happen with "coronium."
134
PLATE XIII
SOLAR CORONA. 1905, AUGUST 30.
From Drawing by Mrs. C. G. Abbot from Photographs by the United States Naval Observatory Eclipse Expedition.
ECLIPSES AND SOLAR ENVELOPES
Several bright coronal lines have been discovered in the ultra-violet by Deslandres, Dyson, Lewis, and others. In the outer corona the Fraunhofer lines of the photospheric spectrum have been seen, and have been repeatedly photographed by Campbell, Perrine and others. Lewis found them only in the ultra-violet spectrum in the eclipse of 1908. These dark lines fade and disappear near the limb of the sun. Their presence in the outer corona is a proof of the presence in the outer coronal light of a large pro- portion of reflected photospheric rays, but Campbell infers from their absence near the sun's limb that the inner corona shines almost wholly by light of incandescence of the material there, due to its being heated by proximity to the sun. There are, however, several causes other than a great admixture of coronal light of incandescence which must contribute to diminish the distinctness of the Fraunhofer lines of the inner corona near the sun's limb. Among these are (1) atmospheric reflection of the strong bright line spectrum of the chromosphere; (2) over ex- posure of photographic spectra for the very inner- most corona, etc.
It was inferred by Bigelow and by Holden, from studies of eclipse photographs, that the corona par- ticipates in the rotation of the sun. This view is con- firmed by spectroscopic observations of Deslandres, Campbell, and Belopolsky. It has been supposed by many that, as the polar streamers of the corona ap- pear much like terrestrial auroras seen in high north- 11 135
THE SUN
ern and southern latitudes, the corona may have, like them, an electrical origin. They would regard its light, like that of the aurora, as largely of lumines- cence similar to that of a glow electrical discharge, and not true temperature radiation. The writer's bolometric observations of the inner corona at the eclipse of 1908 seemed to be incompatible with the view that it shines mainly by light of ordinary incan- descence. For by the aid of absorbing screens it was shown that the ratio of intensity of infra-red to total radiation is almost the same for the inner corona as for the photosphere. If the corona shines mainly by incandescence, and its high temperature is produced by the absorption of sunlight in its particles, then the fraction of its radiation occurring in the infra-red spectrum should be disproportionately greater than that for the photosphere, because the temperature of the corona must be much the lower. Lewis, however, at the same eclipse found the ultra-violet coronal rays disproportionately weaker than those of the photosphere, and inferred therefrom a low coronal temperature. The composition of the inner coronal light cannot yet be regarded as settled. There is undoubtedly some reflected light, some light of incandescence, and perhaps some of lumines- cence. It may be that it is the latter which is the key to the perplexing observations above re- corded. The nature of the corona will be further discussed in Chapter VI.
136
ECLIPSES AND SOLAR ENVELOPES
THE CHROMOSPHERE
Close to the limb of the sun there is seen at total solar eclipses, and by special contrivances also in full sunlight, a thin ring of rosy light called the " chro- mosphere, " from which project irregularly, some- times as much as 50,000 or even 100,000 miles, rosy forms called " prominences. " The spectrum of the chromosphere consists of bright lines on a faint con- tinuous background. These bright lines are the counterparts in position, and generally, also, in rela- tive intensity, of the dark Fraunhofer lines of the photospheric spectrum. The prominences appear to be but higher extensions of the chromosphere, yet their spectra are usually simpler. Prof. C. A. Young made a prolonged study of the prominences and of their spectra, and I cannot introduce the mat- ter better than to quote his descriptions (pages 197 to 226 of "The Sun") supplementing his story by mention of most recent work on the subject. Young's explanations of some of the phenomena differ some- what from those which the present writer would pre- fer, as Young was a believer in the cloudy photo- sphere.
The Eclipse of 1868. Janssen's and Lockyer's Dis- covery.
"Every one is more or less familiar with the story of this eclipse. Herschel, Tennant, Pogson, Rayet, and Janssen, all made substantially the same report.
137
THE SUN
They found the spectrum of the prominences to con- sist of bright lines, and conspicuous among them were the lines of hydrogen. There were some serious dis- crepancies, indeed, among their observations, not only as to the number of the bright lines seen, which is not to be wondered at, but as to their position. Thus, Rayet (who saw more lines than any one else) identified the red line observed with B instead of C; and all the observers mistook the yellow line they saw for that of sodium.
"Still, their observations, taken together, com- pletely demonstrated the fact that the prominences are enormous masses of highly heated gaseous matter, and that hydrogen is a main constituent.
"Janssen went further. The lines he saw during the eclipse were so brilliant that he felt sure he could see them again in the full sunlight. He was pre- vented by clouds from trying the experiment the same afternoon, after the close of the eclipse; but the next morning the sun rose unobscured, and, as soon as he had completed the necessary adjustments, and directed his instrument to the portion of the sun's limb where the day before the most brilliant prominence appeared, the same lines came out again, clear and bright; and now, of course, there was no difficulty in determining at leisure, and with almost absolute accuracy, their position in the spectrum. He immediately confirmed his first conclusion, that hydrogen is the most conspicuous component of the prominences, but found that the yellow line must
138
ECLIPSES AND SOLAR ENVELOPES
be referred to some other element than sodium,1 being somewhat more refrangible than the D lines.
"He found also that, by slightly moving his tele- scope and causing the image of the sun's limb to take different positions with reference to the slit of his spectroscope, he could even trace out the form and measure the dimensions of the prominences; and he remained at his station for several days, engaged in these novel and exceedingly interesting observations.
"Of course, he immediately sent home a report of his eclipse-work, and of his new discovery, but, as his station at Guntoor, in eastern India, was farther from mail communication with Europe than those upon the western coast of the peninsula, his letter did not reach France until some week or two after the ac- counts of the other observers; when it did arrive, it came to Paris, in company with a communication from Mr. Lockyer, announcing the same discovery, made independently, and even more creditably, since with Mr. Lockyer it was not suggested by anything he had seen, but was thought out from fundamental principles.'
"Nearly two years previously the idea had oc- curred to him (and, indeed, to others also, though he was the first to publish it) that, if the protuberances are gaseous, so as to give a spectrum of bright lines, those lines ought to be visible in a spectroscope of
irThis element is helium and was discovered on the earth long afterwards.
139
THE SUN
sufficient power, even in broad daylight. The prin- ciple is simply this :
" Under ordinary circumstances the protuberances are invisible, for the same reason as the stars in the daytime: they are hidden by the intense light re- flected from the particles of our own atmosphere near the sun's place in the sky, and, if we could only suf- ficiently weaken this aerial illumination, without at the same time weakening their light, the end would be gained. And the spectroscope accomplishes pre- cisely this very thing. Since the air-light is re- flected sunshine, it of course presents the same spec- trum as sunlight, a continuous band of color crossed by dark lines. Now, this sort of spectrum is greatly weakened by every increase of dispersive power, be- cause the light is spread out into a longer ribbon and made to cover a more extended area. On the other hand, a spectrum of bright- lines undergoes no such weakening by an increase in the dispersive power of the spectroscope. The bright lines are only more widely separated — not in the least diffused or shorn of their brightness. Moreover, if the gas is one which, like hydrogen, shows dark lines in the ordinary solar spectrum (and therefore in that of the air-light), the case is even better: not only is the continuous spectrum of the air-light weakened by the high dis- persion, but it has dark gaps in it just where the bright lines of the prominence spectrum will fall.
"If, then, the image of the sun, formed by a tele- scope, be examined with a spectroscope, one might
140
ECLIPSES AND SOLAR ENVELOPES
hope to see at the edge of the disk the bright lines belonging to the spectrum of the prominences, in case they are really gaseous.
"Mr. Lockyer and Mr. Huggins both tried the ex- periment as early as 1867, but without success; partly because their instruments had not sufficient power to bring out the lines conspicuously, but more because they did not know whereabouts in the spectrum to look for them, and were not even sure of their exis- tence. At any rate, as soon as the discovery was an- nounced, Mr. Huggins immediately saw the lines without difficulty, with the same instrument which had failed to show them to him before. It is a fact, too often forgotten, that to perceive a thing known to exist does not require one half the instrumental power or acuteness of sense as to discover it.
"Mr. Lockyer, immediately after his suggestion was published, had set about procuring a suitable instrument, and was assisted by a grant from the treasury of the Royal Society. After a long delay, consequent in part upon the death of the optician who had first undertaken its construction, and partly due to other causes, he received the new spectroscope just as the report of HerscheFs and Tennant's ob- servations reached England. Hastily adjusting the instrument, not yet entirely completed, he at once applied it to his telescope, and without difficulty found the lines, and verified their position. He imme- diately also discovered them to be visible around the whole circumference of the sun, and consequently
141
THE SUN
that the protuberances are mere extensions of a con- tinuous solar envelope, to which, as mentioned above, was given the name of Chromosphere. (He does not seem to have been aware of the earlier and similar conclusions of Arago, Grant, Secchi, and others.) He at once communicated his results to the Royal Society, and also to the French Academy of Sciences, and, by one of the curious coincidences which so frequently occur, his letter and Janssen's were read at the same meeting, and within a few minutes of each other.
"The discovery excited the greatest enthusiasm, and in 1872 the French Government struck a gold medal in honor of the two astronomers, bearing their united effigies.
" It immediately occurred to several observers, Janssen, Lockyer, Zollner, and others, that by giving a rapid motion of vibration or rotation to the slit of the spectroscope it would be possible to perceive the whole contour and detail of a protuberance at once, but it seems to have been reserved for Mr. Huggins to be the first to show practically that a still simpler device would answer the same purpose. With a spectroscope of sufficient dispersive power it is only necessary to widen the slit of the instrument by the proper adjusting screw. As the slit is widened, more and more of the protuberance becomes visible, and, if not too large, the whole can be seen at once : with the widening of the slit, however, the brightness of the background increases, so that the finer details of
l \->
ECLIPSES AND SOLAR ENVELOPES
the object are less clearly seen, and a limit is soon reached beyond which further widening is disadvan- tageous. The higher the dispersive power of the spectroscope the wider the slit that can be used, and the larger the protuberance that can be examined as a whole — within certain limits, however. It is not difficult with our latest spectroscopes, diffraction instruments espe- cially, to reach a dispersion so great that even the C line becomes broad and hazy, like the b lines in an ordi- nary instrument. FlG- 27.— HUGGINS'S FIRST OBSERVATION OF A PROMINENCE IN FULL SUNSHINE.
In that case each
luminous point in the prominence itself is represented in the image of the prominence, not by a point, as it should be to give clear definition, but by a streak at right angles to the spectrum lines.
Spectrum of the Chromosphere and Prominences.
" The spectra of the chromosphere and prominences are very interesting in their relations to that of the photosphere, and present many peculiarities which are not yet fully explained. At times and in places where some special disturbance is going on — fre« quently in the neighborhood of spots at the times when they are just passing around the limb of the disk — the spectrum, at the base of the chromosphere,
143
THE SUN
is very complicated, consisting of hundreds of bright lines. In the course of a few weeks of observation at Sherman in 1872, the writer made out a list of two hundred and seventy-three, and more recent observa- tions have added largely to the number — at least fifty lines within the limits of the visible spectrum, and, by photography, at least eighty in the ultra- violet. The majority of the lines, however, are seen only occasionally, for a few minutes at a tune, when the gases and vapors, which generally lie low, mainly in the interstices of the clouds which constitute the photosphere, and below its upper surface, are ele- vated for the time being by some eruptive action. For the most part, the lines which appear only at such times are simply " reversals" of the more prominent dark lines of the ordinary solar spectrum. But the se- lection of the lines seems most capricious; one is taken, and another left, though belonging to the same element, of equal intensity, and close beside the first. It is evident that the subject needs a detailed and careful study, combining solar observations with laboratory-work upon the spectra of the elements concerned, before a satisfactory account can be given of all the peculiar behavior observed.
"The lines composing the true chromosphere spec- trum, if we may call it so (that is, those which are always observable in it with suitable appliances) , are not very numerous, and we give the following list, designating them by their wave length, as given by Rowland :
144
ECLIPSES AND SOLAR ENVELOPES
1. |
7065 |
.50. |
Helium. |
||
2. |
6563 |
.05, |
C. |
Hydrogen |
(Ha). |
3. |
5875, |
,98, |
D3. (close double) . |
Helium. |
|
4. |
5316 |
.87. |
|||
5. |
4861 |
50, |
F. |
Hydrogen |
(H/B). |
6. |
4471 |
,80, |
/• |
Helium. |
|
7. |
4340. |
66, |
g (near G). |
Hydrogen |
(H7). |
8. |
4101.85, |
h. |
Hydrogen |
(US). |
|
9. |
3970 |
.20 |
(in H). |
Hydrogen |
(H«). |
10. |
3968 |
.56, |
H. |
Calcium. |
|
11. |
3933 |
.86, |
K. |
Calcium. |
"The first line is generally very difficult to see, though sometimes pretty conspicuous. It is in the led, between B and a, and has a very faint corre- sponding dark line. No. 3 has no dark line corre- sponding as a usual thing, though occasionally one appears, especially in the neighborhood of sun spots. No. 9 is quite within the broad shade of the H-line, which thus appears double in the chromosphere spec- trum.
"The eleven lines mentioned above are invariably present in the spectrum of the chromosphere; a much larger number make their appearance on very slight provocation. They are:
1'. |
6678 |
2. |
Helium. |
11'. |
5183. |
8, |
bi. |
Magnesium. |
2'. |
6431 |
1, |
Iron. |
12'. |
5172, |
9, |
6,. |
Magnesium. |
3'. |
6141 |
9. |
Barium. |
13'. |
5169 |
2, |
&•. |
Iron. |
4'. |
5896 |
2, D, |
Sodium. |
14'. |
5167. |
6, |
64. |
Magnesium. |
5'. |
5890 |
2, Da |
Sodium. |
15'. |
5018 |
(}. |
Iron. |
|
6'. |
5363 |
0. |
Iron. ? |
16'. |
5015 |
8. |
Helium. |
|
r. |
5284 |
6. |
Titanium? ? |
17'. |
4934 |
3. |
Barium. |
|
8'. |
5276 |
2. |
Chromium. ? |
18'. |
4924 |
1. |
Iron. |
|
9'. |
5234 |
7. |
Manganese. |
19'. |
4922 |
3. |
Helium. |
|
10'. |
5198 |
2. |
? ? |
20'. |
4919 |
1. |
Iron. ? |
145
THE SUN
21'. |
4900 |
3. |
Barium. |
28'. |
4236 |
1. |
Iron. |
|
22'. |
4584 |
1. |
Iron. |
29'. |
4233 |
8. |
Iron. |
|
23'. |
4501 |
4. |
Titanium. |
30'. |
4226 |
9. |
Calcium. |
|
24'. |
4491 |
5. |
Manganese. |
31'. |
4215 |
7. |
Stront-dm. |
|
25'. |
4490 |
2. |
Manganese. |
32'. |
4077.9. |
Strontium. |
||
26'. |
4469 |
5. |
Iron. |
33'. |
4026 |
0. |
Helium. |
|
27'. |
4245 |
5. |
Iron. |
34'. |
3889 |
1. |
Hydrogen |
(HO- |
"It is not intended, however, to intimate that, if one of these appears, all of them will do so, nor that they are equally conspicuous or equally common. To a certain degree, also, their selection by the writer is arbitrary, for there are nearly as many more which are seen pretty frequently, and some of them may very possibly be found hereafter to deserve a place upon the list rather than some that have been in- cluded.
"It requires careful manipulation to bring out the fainter and finer lines satisfactorily. The slit must be adjusted with extreme care to the focal plane of the rays under examination, placed tangential to the solar image, and brought exactly to the edge of the disk. A thousandth of an inch in its position will often make the whole difference between a successful operation and its failure, and even a slight unsteadi- ness of the air will diminish the number of bright lines visible by at least one half.
"As the majority of the lines are developed only by more or less unusual disturbances of the solar surface, it naturally happens that one very often finds them distorted or displaced by the motions of the gases along the line of sight (toward or from the observer) ,
146
ECLIPSES AND SOLAR ENVELOPES
as explained in a previous chapter, producing what Lockyer calls " motion-forms. " Occasionally, also, we meet with " double reversals, " so called, especially in the lines of magnesium and sodium. The (dark) lines of these substances are rather wide in the solar
in the chromosphere
I
FIG. 28. — DOUBLE REVERSAL, OF THE D-LINES. (October, 1880.)
spectrum. When reversed spectrum, the phenomenon usually consists of a thin bright line down the center of the wider dark band : in a double reversal the bright line widens and a fine dark line appears in its center, so that we have a central dark line, a bright one on each side of it, and outside of the bright lines a dark shade on both sides. Fig. 28 represents such a double reversal of the D-lines observed by the writer on several occasions in 1880. The phenomenon seems to be due to the pres- ence of an unusual quantity of the vapor at a consid- erable density, and is the precise correlative of what is sometimes seen in the spectrum of a sodium-flame. The two D-lines of sodium each becomes itself double, so that we get pairs of bright lines in place of single lines. The electric arc often shows this still more finely.
"At the base of a prominence, the C, F, H, and K lines are always thus doubly reversed. Fig. 29 is from a recent photograph of the C-line obtained at Prince-
147
THE SUN
ton, by Mr. Reed, with the large telescope and spec- troscope. The slit was tangential to the sun's limb.
Of course, an isochromatic plate and a long exposure were required to get such an impression from the "ruby light" of that part of the spec- trum. When the slit is ail- justed to cross the sun's limb radially the bright lines where they project beyond the spectrum of the photosphere assume the "arrow-headed'' form shown in Fig. 30. " Generally speaking, the spectrum of a prominence is simpler than that of the chromosphere at its base. We seldom find any lines except C, D3, F, gy h, H and K, at a considerable elevation above the photosphere, though / is sometimes met with. On rare occasions, also, the vapors of sodium and magnesium are carried into the higher regions, and once or twice
148
FIG. 29. — DOUBLE REVERSAL (Photographed.)
Fu;. .SO.
ECLIPSES AND SOLAR ENVELOPES
the writer has seen the line No. 1 of the second list (6678.2) in the upper portions of a prominence.
Observation of Prominences.
" When the spectroscope is used as a means of ren- dering visible the forms and features of the promi- nences, the only difference is that the slit is more or less widened.
"The telescope is directed so that the solar image shall fall with that portion of its limb which is to be examined just tangent to the opened slit, as in Fig. 31, which represents the slit-plate of the spectroscope, with the image of the sun in position for observation.
"If, now, a prominence ex- FIG. 31.— OPENED SLIT OP
. . . ' . . THE SPECTROSCOPE.
ists at this part of the sun s
limb (as would probably be the case, considering the proximity of the spot shown in the figure), and if the spectroscope itself is so adjusted that the C-line falls in the center of the field of view, then, on looking into the eyepiece, one will see something much like Fig. 32. The red portion of the spectrum will stretch athwart the field of view like a scarlet ribbon, with a darkish band across it, and in that band will appear the prominences, like scarlet clouds — so like our own terrestrial clouds, indeed, in form and texture, that the resemblance is quite startling : one might almost think he was looking out through a partly opened
149
THE SUN
door upon a sunset sky, except that there is no variety or contrast of color; all the cloudlets are of the same pure scarlet hue. Along the edge of the opening is seen the chromosphere, more brilliant than the clouds which rise from it or float above it, and for the most part made up of minute tongues and filaments. Usu- ally, however, the definition of the chromosphere is less distinct than that of the higher clouds. The reason is, that close to the limb of the sun, where the temperature and pressure are highest, the hydrogen is in such a state that the lines of its spectrum are widened and " winged, " something like those of mag- nesium, though to a less extent. Each point in the chromosphere, therefore, when viewed through the opened slit, appears not as a point, but as a short line, directed lengthwise in the spectrum. As the length of this line depends upon the dispersive power of the spectroscope, it is easy to see that it is possible to go too far in this respect. The lower the dispersion the more distinct the image obtained, but also the fainter as compared with the background upon which it is seen.
"Just beneath the chromosphere (at a in the cut) the appearance is as if the edge of the sun was dark, a phenomenon which for some time was very puzzling. Its explanation lies in the "double reversal" of the C-line at the base of the chromosphere, discussed and figured a few pages back.
"If the spectroscope is adjusted upon the F-line, instead of C, then a similar image of the prominences
150
ECLIPSES AND SOLAR ENVELOPES
and chromosphere is seen, only blue instead of scarlet; usually, however, since the F-line is hazier and more winged than C, this blue image is somewhat less per- fect in its details and definition, and is therefore less used for observation. Similar effects are obtained by means of the yellow line near D, and the violet line near G. With suitable precautions, using a violet shade- glass before the eye, and carefully shutting out all extraneous light, the H and K lines can also be used; but visual observa- tions in this part of the spectrum are extremely difficult and unsatisfactory.
"With photog- FIG. 32. — CHROMOSPHERE AND PROMINENCES
, , , . AS SEEN IN THE SPECTRUM.
rapny the case is
the reverse — these lines are then precisely those which can be employed most easily and conveniently. We shall recur to this a little later.
" Professor Winlock and Mr. Lockyer have at- tempted, by using an annular opening instead of the ordinary slit, to obtain a view of the whole circum- ference of the sun at once, and have succeeded. With a spectroscope of sufficient power, and adjustments delicate enough, the thing can be done ; but as yet no very satisfactory results appear to have been reached. 12 151
THE SUN
We still (in visual observations) have to examine the circumference piecemeal, so to speak, readjusting the instrument at each point, to make the slit tangen- tial to the limb.
"The number of protuberances of considerable magnitude (exceeding ten thousand miles in altitude), visible at any one time, on the circumference of the sun, is never very great, rarely reaching twenty-five or thirty. Their number, however, varies extremely with the number of sun spots: during a sun-spot minimum there are not unfrequently occasions when not a single one can be found, though even during those years the, more usual number is five or six — some of which often are of considerable size. The observations of Tacchini and Secchi have showed that their numbers closely follow the march of the sun spots though never falling quite so low.
"To Tacchini we owe our most complete record of these objects, now continuous since 1872, giving their number and distribution upon the sun, with drawings of all that were specially remarkable. Many others have cooperated in observations of this kind: the Hungarian observers, Fenyi at Kalocsa, and Von Gothard at Hereny, have given us many fine descrip- tions and delineations. Father Perry and his assis- tant Sidgreaves, at Stonyhurst, also deserve a special mention.
"Their distribution on the sun's surface is in some respects similar to that of the spots, but with impor- tant differences. The spots are confined within 40°
152
ECLIPSES AND SOLAR ENVELOPES
of the sun's equator, being most numerous at a solar latitude of about 20° on each hemisphere. Now, the protuberances are most numerous pre- cisely where the spots are most abundant, but they do not disappear at a latitude of 40°; they are found even at the poles, and from the latitude
FIG. 33. — RELATIVE FREQUENCY OF PROTUBERANCES AND SUN-SPOTS.
of 60° actually increase in number to a latitude of about 75°.
"The annexed diagram, Fig. 33, represents the relative frequency of the protuberances and spots on the different portions of the solar surface. On the left side is given the result of Carrington's observa- tion of 1,386 spots between 1853 and 1861, and on the right the result of Secchi's observations of
153
THE RFX
2,707 l protuberances in 1871 . The lengl h of each ra- dial line represents the number of spots or protuber- ances observed at each particular latitude on a scale of a quarter of an inch to the hundred ; for example, Secchi gives 228 protuberances as the number ob- served during the period of his work between 10° and 20° of south latitude, and the corresponding line drawn at 15° south, on the left-hand side of the figure is therefore made Iff or .57 of an inch long. The other lines are laid off in the same way, and thus the irregular curve drawn through their extremities rep- resents to the eye the relative frequency of these phe- nomena in the different solar latitudes. The dotted line on the right-hand side represents in the same manner and on the same scale the distribution of the larger protuberances, having an altitude of more than 1', or 27,000 miles.
"A mere inspection of the diagram shows at once that, while the prominences may, and in fact often do, have a close connection with the spots, they are yet to some extent independent phenomena.
"A careful study of the subject shows that they are much more closely related to the faculse.2 In many cases, at least, facula3, when followed to the
1 The 2,767 prominences are not all different ones. If any of the prominences observed on one day remained visible the next, they were recorded afresh; and, as a prominence near the pole would be carried but slowly out of sight by the sun's rotation, it is thus easy to see how the number of prominences recorded in the polar regions is so large.
'See page 109 [of Young].
154
ECLIPSES AND SOLAR ENVELOPES
limb of the sun, have been found to be surrounded by prominences, and there is reason to suppose that the fact is'a general one. The spots, on the other hand, when they reach the border of the sun's image, are commonly surrounded by prominences more or less completely, but seldom overlaid by them. Indeed, Respighi asserts (and the most careful observations we have been able to make confirm his statement) that as a general rule the chromosphere is consider- ably depressed immediately over a spot. Secchi, how- ever, denies this.
Magnitude and Classification of Prominences.
"The protuberances differ greatly in magnitude. The average depth of the chromosphere is not far from 10" or 12", or about 5,000 or 6,000 miles, and it is not, therefore, customary to note as a prominence any cloud with an elevation of less than 15" or 20"— 7,000 to 9,000 miles. Of the 2,767 already quoted, 1,964 attained an altitude of 40", or 18,000 miles, and it is worthy of notice that the smaller ones are so few, only about one third of the whole: 751, or nearly one fourth of the whole, reached a height of over 1', or 28,000 miles; the precise number which reached greater elevations is not mentioned, but several ex- ceeded 3', or 84,000 miles. It is only rather rarely that they reach elevations as great as 100,000 miles. The writer has in all seen, perhaps, three or four which exceeded 150,000 miles, and Secchi has re- corded one of 300,000 miles. On October 7, 1880,
155
THE SUN
the writer observed one which attained the still un- equaled height of over 13' of arc, or 350,000 miles. When first seen, on the southeast limb of the sun, about 10.30 A.M., it was a "horn" of ordinary appear- ance, some 40,000 miles in elevation, and attracted no special attention. When next seen, half an hour later, it had become very brilliant and had doubled its height : during the next hour it stretched upward until it reached the enormous altitude mentioned, breaking up into filaments which gradually faded away, until, by 12.30 P.M., there was nothing left. A telescopic examination of the sun's disk showed noth- ing to account for such an extraordinary outburst, except some small and not very brilliant faculse. While it was extending upward most rapidly a violent cyclonic motion was shown by the displacement of the spectrum lines, and H and K were reversed through its whole height.
"In their form and structure the protuberances differ as widely as in their magnitude. Two princi- pal classes are recognized by all observers — the qui- escent, cloud-formed or hydrogenous, and the eruptive or metallic. By Secchi these are each further sub- divided into several sub-classes or varieties, between which, however, it is not always easy to maintain the distinctions.
"And here perhaps is the proper place to mention that Trouvelot insists on the existence of "dark" prominences — i.e., clouds of cooler hydrogen • that absorb the light of the hydrogen behind them; but
156
ECLIPSES AND SOLAR ENVELOPES
Three figures of the same prominence, seen July 25, 1872.
FIG. 34.
AS SEEN AT 2.15 P. M.
FIG. 38. SHEAF AND VOLUTES.
FIG. 35. AS SEEN AT 2.45 P. M.
FIG. 30.
AS SEEN AT 3.30 P. M. Scale, 100,000 miles to the inch.
ERUPTIVE PROMINENCES 157
THE SUN
there is no proof, we think, that these are anything but " holes. " Tacchini, on the other hand, is disposed to assert the existence of " white" prominences, which give a continuous spectrum, and so are not reached by spectroscopic observation, though conspicuous to the eye, and on the photographic plate, at the time of a total eclipse, as in 1883 and December, 1889. But the evidence hardly warrants confident belief in the existence of such objects.
"The quiescent prominences in form and texture resemble, with almost perfect exactness, our terres- trial clouds, and differ among themselves as much and in the same manner. The familiar cirrus and stratus types are very common, the former especially, while the cumulus and cumulo-stratus are less frequent. The protuberances of this class are often of enormous magnitude, especially in their horizontal extent (but the highest elevations are attained by those of the eruptive order), and are comparatively permanent, remaining often for hours and days without serious change; near the poles they sometimes persist through a whole solar revolution of twenty- seven days. Sometimes they appear to lie upon the limb of the sun like a bank of clouds in the horizon; prob- ably because they are so far from the edge of the disk that only their upper portions are in sight.* When seen in their full extent they are ordinarily connected to the underlying chromosphere by slender columns, which are usually smallest at the base, and appear often to be made up of separate filaments closely in-
158
ECLIPSES AND SOLAR ENVELOPES
FIG. 43. DIFFUSE.
FIG. 41.
FlLAMENTAKY.
FIG. 44. STEMMED.
FIG. 42. PLUMES.
QUIESCENT PROMINENCES.
Scale, 7o,000 miles to the inch.
159
THE SUN
tertwined, and expanding upward. Sometimes the whole under surface is fringed with down-hanging filaments, which remind one of a summer shower fall- ing from a heavy thundercloud. Sometimes they float entirely free from the chromosphere ; indeed, as a general rule, the layer clouds are attended by de- tached cloudlets for the most part horizontal in their arrangement.
" The figures give an idea of some of the general ap- pearances of this class of prominences, but their del- icate, filmy beauty can be adequately rendered only by a far more elaborate style of engraving.
"Their spectrum is usually very simple, consisting of the four lines of hydrogen, and the three of helium, with H and K. Occasionally the sodium and mag- nesium lines also appear, and that even near the sum- mit of the clouds; and this phenomenon was so much more frequently observed in the clear atmosphere of Sherman as to suggest that, if the power of our spec- troscopes were sufficiently increased, it would cease to be unusual.
"The genesis of this sort of prominence is problem- atical. They have been commonly looked upon as the debris and relics of eruptions, consisting of gases which have been ejected from beneath the solar sur- face, and then abandoned to the action of the cur- rents of the sun's upper atmosphere. But near the poles of the sun distinctively eruptive prominences never appear, and there is no evidence of aerial cur- rents which would transport to those regions matters
160
ECLIPSES AND SOLAR ENVELOPES
ejected nearer the sun's equator. Indeed, the whole appearance of these objects indicates that they orig- inate where we see them. Possibly, although in the polar regions there are no violent eruptions, there yet may be a quiet outpouring of heated hydrogen sufficient to account for their production — an out- rush issuing through the smaller pores of the solar surface, which abound near the poles as well as elsewhere.
"But Secchi reports an observation which, if cor- rect, puts a very different face upon the matter.1 He has seen isolated cloudlets form and grow spon- taneously without any perceptible connection with the chromosphere or other masses of hydrogen, just as in our own atmosphere clouds form from aqueous vapor, already present in the air, but invisible until some local cooling or change of pressure causes its condensation. These prominences are, therefore, formed by some local heating or other luminous ex- citement of hydrogen already present, and not by any
1 On October 13, 1880, the writer for the first time met with the same phenomenon. A small, bright cloud appeared on that day, about 11 A. M., at an elevation of some 2^' (67,500 miles) above the limb, without any evident cause or any visible connection with the chromosphere below. It grew rapidly without any sensible rising or falling, and in an hour developed into a large stratiform cloud, irregular on the upper surface, but nearly flat beneath. From this lower surface pendent filaments grew out, and by the middle of the afternoon the v object had become one of the ordinary stemmed prominences, much like Fig. 44.
But obviously the thing is very unusual, for in more than twenty years of observation I have encountered the phenomenon only three times.
161
THE SUN
transportation and aggregation of materials from a distance. The precise nature of the action which produces this effect it would not be possible to assign at present ; but it is worthy of note that the spectro- scopic observations made during eclipses rather favor this view, by showing that hydrogen, in a feebly luminous condition, is found all around the sun, and at a very great altitude — far above the ordinary range of prominences.
" Indeed, in most cases the forms and changes of this class of prominences so closely resemble our own terrestrial clouds that one is almost forced to believe that they are surrounded by, and float in, a medium which does not greatly differ from themselves in den- sity, though it is not visible in the spectroscopic mode of observation.
Eruptive Prominences.
"The eruptive prominences are very different- much more brilliant and much more vivacious and interesting. They consist usually of brilliant spikes or jets, which change their form and brightness very rapidly. For the most part they attain altitudes of not more than 20,000 or 30,000 miles, but occasion- ally they rise far higher than even the largest of the clouds of the preceding class. Their spectrum is very complicated, especially near their base, and often filled with bright lines, those of sodium, magnesium, barium, iron, and titanium, being especially conspicu- ous, while calcium, chromium manganese, and prob-
ECLIPSES AND SOLAR ENVELOPES
FIG. 49.
PROMINENCE AS IT APPEARED AT HALF- PAST TWELVE O'CLOCK, SEPTEMBER 7, 1871.
FIG. 46.
VERTICAL FILAMENTS.
FIG. 47. CYCLONE.
FIG. 50.
As THE ABOVE APPEARED HALF AN HOUR LATER WHEN THE UP-RUSHING HYDROGEN ATTAINED A HEIGHT OF MORE THAN 200,- 000 MILES.
FIG. 51.
SPOT NEAR THE SUN'S LlMB, WITH ACCOM- PANYING JETS OF HYDROGEN, AS SEEN FLAMES. OCTOBER 5, 1871.
Scale, 75,000 miles to the inch.
163
THE SUN
ably sulphur, are by no means rare, and for this reason Secchi calls them metallic prominences.
"They usually appear in the immediate neighbor- hood of a spot, never occurring very near the solar poles. Their form and appearance change with great rapidity, so that the motion can almost be seen with the eye — an interval of fifteen or twenty minutes being often sufficient to transform, quite beyond rec- ognition, a mass of these flames fifty thousand miles high, and sometimes embracing the whole period of their complete development or disappearance. Some- times they consist of pointed rays, diverging in all directions, like hedgehog-spines. Sometimes they look like, flames ; sometimes like sheaves of grain ; sometimes like whirling waterspouts, capped with a great cloud; occasionally they present most exactly the appearance of jets of liquid fire, rising and falling in graceful parabolas; frequently they carry on their edges spirals like the volutes of an Ionic column; and continually they detach filaments which rise to a great elevation, gradually expanding and growing fainter as they ascend, until the eye loses them. Our figures present some of the more common and typical forms, and illustrate their rapidity of change, but there is no end to the number of curious and interest- ing appearances which they exhibit under varying circumstances.
"The velocity of the motions often exceeds a hun- dred miles a second, and sometimes, though very rarely, reaches two hundred miles. That we have to
164
ECLIPSES AND SOLAR ENVELOPES
do with actual motions, and not with mere change of place of a luminous form, is rendered certain by the fact that the lines of the spectrum are often displaced and distorted in a manner to indicate that some of the cloud-masses are moving either toward or from the earth (and, of course, tangential to the solar surface) with similar swiftness.
" Fig. 52 is a representation of a portion of the spec- trum of a prominence observed at Sherman on August
FIG. 52.
3, 1872, an observation to which allusion was made in the preceding chapter. The F-line, at 208 of the scale, must be imagined as blazingly brilliant, and fainter bright lines appear at 203.2, 208.8, 209.4, and 212.1 (the scale is KirchhofTs), while two bands of continuous spectrum, produced probably by the com- pression of the gas at the points of maximum dis- turbance, run the whole length of the figure. At the
165
THE SUN
upper point of disturbance F is drawn out into a point reaching to 207.4 of the scale, and indicating a veloc- ity of 230 miles a second away from us; at the lower point it extends to 208.7, and indicates a velocity of about 250 miles per second toward us. It was very noticeable that this swift motion of the hydrogen did not seem to carry with it many other substances which were at the time represented in the spectrum by their bright lines; magnesium and sodium were somewhat affected, but barium and the unknown ele- ment of the corona were not. "
An examination of the sun's limb for prominences is made on every fair observing day at many observa- tories. At the Italian observatories of Rome and Catania such observations have been continued by Secchi, Tacchini, and Ricco for about forty years. A general discussion of this highly valuable mass of observations is about to be published.
Prominences and the Spectroheliograph.
Since the introduction of the Spectroheliograph the prominences can be observed much more satisfactor- ily than before. In Plate XIV, Fig. 1 shows a large quiescent prominence photographed by Slocum with the Rumford Spectroheliograph at the Yerkes Ob- servatory. The height of the prominence as shown in the plate is 1.6 minutes of arc, or 69,000 kilometers. Slocum states that this prominence lasted probably continuously for at least fifty-five days in the spring of 1910, but Evershed traces it twenty-seven days
166
PLATE XIV.
FIG. 1. — 1910, March 17. G. M. T. oh 30m. LON. 7°. LAT. + l/° TO -18°
FIG. 2. — 1910, October 10. G. M. T. 7h 56m .8.
FIG. 3. — 1910, October 10. G. M. T. 8h 6m .4. SOLAR PROMINENCES. (Slocum.) CALCIUM (H) SPECTROHELIOGRAMS.
ECLIPSES AND SOLAR ENVELOPES
longer still. Its southern extremity remained nearly stationary near 20° south latitude, while the northern end varied greatly; from the equator on March 4, to 25° north latitude on March 18; then, retiring, reached 10° south latitude on April 28. At no time could the prominence be seen projected against the sun's disk in Slocum's calcium spectroheliographic observations. He saw it only on the sun's limbs. But Evershed and Deslandres photographed it re- peatedly; the former in H2 calcium, the latter in K3 calcium and Ha hydrogen light, appearing like a long cloud upon the sun's disk. A similar feature is shown in the Ha photograph reproduced in Plate VI from Ellerman's Mount Wilson observations of April 30, 1908.
Very beautiful eruptive prominences are occa- sionally observed with the spectroheliograph. The two lower figures in Plate XIV show an uncommon- ly fine quasi-eruptive prominence photographed Oc- tober 10, 1910, in the H line of calcium, by Slocum at the Yerkes Observatory. Although by no means as active as some eruptive prominences, this one changed rapidly, and there may be seen considerable differences in form in the two exposures, separated by a time interval of only ten minutes. The approxi- mate position was as follows: Solar latitude 24° to 39° S; longitude 225°. Height 2.5 minutes of arc, or 108,000 kilometers.
13 167
THE SUN
Recent Flash-spectrum Observations.
Following the great discovery of Janssen and Lock- yer in 1868, the next year brought the important dis- covery of helium in the sun — a chemical element not found on the earth for nearly thirty years afterwards. Young kept up the pace by the discovery of the " flash spectrum" at the total solar eclipse of 1870. Setting the slit of his spectroscope where the chromo- sphere should be, and keeping his eye prepared for what he was about to witness, he saw, as the last photospheric rays were extinguished, a bright line reversal of the photospheric spectrum flash out to view. It was not till 1896 that the flash spectrum was photographed by Shackelton with a prismatic camera.
The chromosphere appears as a very thin crescent, hence its spectrum may be photographed without slit or collimator. The appearance of such spectra may be understood from Plate XV, Fig. 1, taken by S. A. Mitchell at the eclipse of 1905. The spectrum lines are each represented by arcs of circles. Where very long arcs appear, they correspond to the great lines of hydrogen and calcium. These elements ex- tend much higher above the sun than others, and hence continue in sight longer as the moon advances.
It was feared that the astigmatism, which makes the concave grating a valuable laboratory instrument, would render it unfit for use on the flash spectrum without a slit. A slit for such work is undesirable,
168
ECLIPSES AND SOLAR ENVELOPES
on account of the loss of light. The work of Mitchell in 1898, who used for the photography of stellar spec- tra a Rowland concave grating as an objective grat- ing, without slit, paved the way for the use of grat- ings at the time of an eclipse. They were first used in flash-spectrum photography in 1900.
Successful photographic observations of the flash spectrum have been made at the eclipses of 1896, 1898, 1900, 1901, 1905, and 1908. Among the observ- ers have been Shackelton, Campbell, Evershed, Dyson, Jewell, Frost, Lord, S. A. Mitchell, Perrine, and others. The observations have shown that the flash spectrum, or spectrum of the chromosphere, is essentially the reversal of the ordinary Fraunhofer spectrum, but with some significant differences. Many of the weaker Fraunhofer lines, of course, do not appear. The lines of the two spectra in general bear different relative intensities. Taking the lines of any one chemical element by itself, however, the rel- ative intensities in the two spectra are not very dif- ferent. Lockyer, Evershed, and Dyson find in gen- eral that the so-called enhanced or spark lines are more prominent in the flash spectrum than in the photospheric spectrum.1 The cause of the dis- crepancy between the line intensities in the spectra, as a whole, seems to be that the elements of higher atomic weights are less prominent in flash spectra.
Dyson, in discussing the Greenwich observations
1 Frost and Mitchell were inclined to question that this is general, but Mitchell seems now to agree that it is.
169
THE SUN
of the eclipses of 1900, 1901, and 1905, gives the meas- ured positions of about 1200 lines, and identifica- tions of most of them with single lines, or with blends of several lines, found in Rowland's tables. The range of spectrum observed is from 3295 Ang- stroms, in the ultra-violet, to 5896 in the orange. The average deviation of the positions from the positions fixed by Rowland is 0.04 Angstroms, but as Dyson's spectra are prismatic this difference is exceeded in the green and yellow. Dyson found twenty-six strong lines of hydrogen agreeing excellently in position with the places fixed by Balmer's series formula. Helium is also a prominent element. The following are the chemical elements as they are found repre- sented by their spectral lines :
Very strong: Hydrogen, Helium, Magnesium, Calcium, Scandium, Titanium, Chromium, Stron- tium.
Strong: Manganese, Iron, Yttrium, Zirconium, Barium, Lanthanum, Cerium, Erbium, Europeum.
Not very strong: Carbon, Aluminum, Vanadium, Neodymium.
Very weak : Nickel, Cobalt, Lead.
Possibly shown: Zinc, Lanthanum, Tantalum.
Doubtful: Silicon, Gadolinium, Prsesodymium.
Absent:1 Argon, Neon, Krypton.
Not well shown within limits of spectrum : Sodium.
The arc lines of aluminum, magnesium, barium,
1 Mitchell, however, inclines to think these elements are repre- sented in the flash spectrum by weak lines.
170
ECLIPSES AND SOLAR ENVELOPES
zinc, and lead appear to be present, whereas their enhanced, or spark, lines show not at all, or faintly. In this Dyson finds these elements exceptional, for in general it is the enhanced lines which predominate in the flash.
The Heights of Different Metals in the Chromosphere. By measuring the lengths of the arcs seen as flash spectrum lines, observers have estimated the heights to which the elements rise in the chromosphere above the sun's general surface. From Sir Norman Lock- yer's report of observations of the eclipse of 1898, we have the following values. The ordinary chemical symbols for the elements are used for short :
Element |
Ca |
H |
He |
Sr |
Ca |
Mg |
Al |
Mn |
Fe |
C |
Var- ious |
Spectrum |
K |
Not |
4471 |
4078 |
4227 |
U.V. |
3944 |
Quar- |
Many |
Flut- |
Many |
lines |
given |
trip- |
tet |
ing |
lines |
||||||
let |
Includ- |
||||||||||
4027 |
4216 |
3962 |
4031 |
lines |
ing Fe |
||||||
etc. |
arc |
||||||||||
lines |
|||||||||||
Mean |
|||||||||||
height |
|||||||||||
seconds. . . |
13.3 |
10 |
7.5 |
6.0 |
4.4 |
4.4 |
3.2 |
2.4 |
3.2 |
1.05 |
1.05 |
to 1.4 |
|||||||||||
Kilometers . |
9700 |
7200 |
5400 |
4300 |
3200 |
3200 |
2300 |
1800 |
2300 |
760 |
760 |
to 1000 |
Jewell,1 from observations of the eclipses of 1900 and 1901, estimates the chromospheric heights cor- responding to separate lines of various elements. He finds, as does Lockyer (See Ca above), that different lines of the same elements yield widely different val- ues. Thus, for calcium his heights range from 15,000
lPub. U.S. Naval Observatory, 2 Series, Vol. IV, Ap. I. 171
THE SUN
down to 100 miles, and for titanium from 3,500 to 100 miles. In general his results show high levels for hydrogen, helium, parhelium, magnesium, sodium, and ytterbium; low levels for chromium, iron, cobalt, nickel, manganese, yttrium, cadmium, zinc, carbon (as cyanogen), and vanadium; contradictory levels indicated by different lines for calcium, strontium, barium, scandium, and titanium. Most lines corre- spond to heights of less than one second of arc (475 miles, 760 kilometers). Jewell regards the chromo- sphere as an atmosphere of hydrogen and a few other permanent gases, rapidly decreasing in density out- ward, and holding as temporary constituents other elements as products of eruptions from within, or meteors from without.
Frost and Mitchell, from observations of the 1900 and 1901 eclipses, respectively, have also given brief tables of the heights attained by different elements in the chromosphere, as indicated by individual spec- trum lines. Their results differ very little from those above mentioned. Mitchell states that the lengths of a great majority of the lines indicate heights not exceeding 0.5" of arc, and would set I" of arc as the average depth of the " re versing layer."
Jewell very pertinently calls attention to the mi- nute quantities of substance required to produce spec- trum lines. As some lines require less producing substance than others, this may cause part of the dis- crepancy between the heights estimated for different lines of the same element.
ECLIPSES AND SOLAR ENVELOPES
Mitchell's Observations of 1905.
My friend, Prof. S. A. Mitchell, has kindly fur- nished me, in advance of his publication, with the following description of his apparatus, and of the results he obtained as a member of the U. S. Naval Observatory expedition, at the total eclipse of August, 1905. His flash spectrum is believed to be the best which has ever been secured.
"Mitchell used two spectrographs of high disper- sion, both with gratings. The first was a six-inch Rowland plane grating of 15,000 lines to the inch, belonging to the Naval Observatory. This same grat- ing had been used by him in Sumatra, in 1901, but in 1905 a glass achromatic objective of five inches aperture was used instead of a quartz lens. With this instrument special attention was paid to the red end of the spectrum. The other instrument was a four- inch grating, ruled on a parabolic surface, instead of the ordinary spherical concave surface. This grating of 14,438 lines per inch and ten feet radius of curva- ture was very bright in the first order on one side, and in the estimation of Mr. Jewell it was one of the best of Rowland gratings, and gave spectra equal in brightness to that obtained by the ordinary six-inch grating. This grating belonged to the Rumford com- mittee, and was kindly loaned by Professor F. A. Saunders of Syracuse University.
"Such a spectrograph used for eclipse work is of the simplest form imaginable. Light from the coelo-
173
THE SUN
stat mirror, reflected horizontally, falls on the grating, where it is diffracted, and is then brought to focus on the photographic plate, five feet distant. Grating and plate holder are placed in a wooden box, and if the grating and photographic plate are perpendicular to the diffracted beam, the spectrum is "normal." As the spectrum was brought to a focus on a circle of thirty inches radius, it was impossible to bend the photographic plate, and heavy gelatine films were used. The spectra were focussed by using a col- limating apparatus consisting of a slit between two concave mirrors (which were previously adjusted by the use of a five-inch visual telescope). The spectra were focussed visually, and test photographs were made in order to check the ultra-violet focus. The excellence of the focus is shown by the flash spectra, which were photographed in the first order.
"The parabolic grating spectra extend from X 3,300 in the ultra-violet to the D lines at X 5,890 in the orange. The plane grating spectrogram continues in the red to the C line. The length of the spectrum taken with the former grating is 9.5 inches. The spec- trum is very nearly normal throughout its whole extent; the dispersion, therefore, is such that one millimeter is equal to 10.8 Angstrom units. This is a dispersion about equal to that obtained by the three- prism spectrographs attached to the great Lick or Yerkes telescopes. As the grating at the eclipse was used as an objective grating without slit, it had a dis- persion a little less than a quarter of that obtained
174
x
ECLIPSES AND SOLAR ENVELOPES
with a 21.5 foot grating in the first order on the or- dinary Rowland mounting.
"As is seen from the illustration (Plate XV, Fig. 1, where unfortunately a great amount of fine detail is lost in reproduction) the definition is excellent. The extreme ultra-violet is not in quite so good a focus as the region from K to D, where the definition is per- fect. About 4,000 lines were measured in the region from X 3,300 to X 5,900. On account of the strong con- tinuous spectrum throughout the photographed spec- trum, it was a little difficult to see the spectrum lines, especially when they were faint. The spectrum lines being curved, it was necessary to measure at the same part of each line. Moreover, since no slit was used, it was necessary to measure the position of the line at the moon's edge. Evidently the height above the sun's limb of the metallic vapor forming a given spec- trum line has much to do with its appearance on the photographic plate, and the middle of the measured line will not give its exact wave length.
"Preliminary wave lengths of the flash lines were directly obtained from the measures. These were compared with Rowland's tables. Each line from Rowland which was identified with certainty was taken as a standard to obtain adjusted values of the flash wave lengths. The smaller dispersion hi the flash spectrum caused lines in Rowland to be blended together; and in such blends it was difficult to know what exact wave length to assume. Consequently, if the flash lines were identified with single lines in
175
THE SUN
Rowland's tables, they were taken as standards. Since the scale-value assumed was only an approxi- mate one, and the spectrum was not strictly normal, a Least Squares adjustment was made in a manner suggested by Professor C. Runge, Kaiser Wilhelm Professor at Columbia University in 1909-10. As the result of this adjustment, the probable error of a single determination of a wave length throughout the spectrum is about ± 0.025 Angstrom units. The small size of this error will be appreciated when one remembers that a series of cusps were measured, and that an error in measurement of a thousandth of a millimeter, or one micron, corresponds to a discrep- ancy of 0.01 A. U.
"Such accurate wave lengths of the flash spectra lend the possibility of a close comparison with the Fraunhofer spectrum. Such a comparison shows with great certainty that the flash spectrum is but a rever- sal of the Fraunhofer lines. Almost every line in the ordinary solar spectrum with an intensity of 3 or greater on Rowland's scale is found in the flash spec- trum, in many cases two or more lines being blended into one in the photograph of smaller dispersion. Though all the strong Fraunhofer lines are found in the flash spectrum, the converse is not true; for there are many strong lines in the flash spectrum which have no equivalent in the ordinary spectrum. In addition to this fact, there is the further difference that there are remarkable inequalities in intensity between the lines of the two spectra.
176
ECLIPSES AND SOLAR ENVELOPES
"It was pointed out by Evershed that we could easily imagine two separate gases in the sun's enve- lope which would have absorption lines of the same intensities, but whose emission spectra would differ very much in intensities. A heavy gas, lying in a thin layer above the photosphere, might absorb the solar light exactly to the same extent as a less dense layer extending to greater altitudes. As the moon successively passes over layers at the time of an eclipse, the lighter gas would give lines of the greater intensities in the flash spectrum. As is well known, the helium lines appear as strong lines in the flash spectrum; they are lacking in the Fraunhofer spec- trum. Over thirty lines of the hydrogen series have been counted in Mitchell's 1905 spectra. In Plate XV, Fig. 2, a portion of the spectrum is shown greatly enlarged.
"For the purpose of a closer comparison, the re- sults of the measures of an extent of the spectrum of 62 Angstrom units to the red side of HS, i.e. from XX 4, 102-4, 164 are given in the following table.
"In the above region, where ninety-two lines in the flash were measured, there are eighty-two lines in Rowland's tables of an intensity 2 and greater. Of these eighty-two lines, but one is with certainty lack- ing from the flash spectrum, the Fe line (intensity 4) at X 4,154.976. Of the ninety- two flash lines, all have been identified with the exception of a few faint lines. The remarkable accuracy of the wave lengths of this flash spectrum, which far surpasses any results hith-
177
THE SUN
TABLE X. — Measures of ninety-two lines in the flash spectrum near H$.
Flash Spectrum |
Wave Length Rowland |
Numbe of Lines Blended |
Sub- stance |
Inten sity and Char- acter |
Remarks |
|
In- ten- sity |
Wave Length |
|||||
5o |
4102.00 |
4102.000 |
H5 |
40 N |
||
1 |
4103.10 |
4103.097 |
Si, Mn |
5 |
||
0 |
4103.65 |
4103.622 |
2 |
1 |
||
2 |
4104.27 |
4104.288 |
Fe |
5 |
||
0 |
4104.65 |
4104.623 |
Co, V, |
0 |
||
3 |
4105.21 |
4105.245 |
2 |
— , V |
3 |
|
2 |
4106.49 |
4106.502 |
2 |
Fe |
4 |
|
2 |
4107.64 |
4107.649 |
Ce-Fe-Zr |
5 |
||
0 |
4108.68 |
4108.687 |
2 |
|||
3 |
4109.371 |
4109.215) |
Fe |
3 |
||
3 |
4109. 88 j |
4 109. 609 1 |
Nd? |
1 |
||
4109 905 |
V |
2 |
||||
2 |
4110.63 |
4110.691 |
Co |
4 |
||
1 |
4111.62 |
|||||
2 |
4111.97 |
41 11. '940 |
V |
4 |
||
0 |
4112.45 |
4112.478 |
Fe |
2 |
||
0 |
4112.89 |
4112.869 |
Ti |
1 |
||
1 |
4113.24 |
4113.183 |
2 |
Fe, Mn |
4 |
|
2d |
4114.00 |
|||||
3 |
4114.73 |
4ii4.'769 |
2 |
Fe,' — |
e" |
|
3 |
4115.35 |
4115.330 |
V |
3 |
||
4116.14 |
4116.138 |
0 |
||||
2d |
4116.78 |
4116.738 |
3 |
V.'Nd? |
2 |
|
4118.02 |
4118.008 |
2 |
||||
5 |
4118.85 |
4118.852 |
3 |
Fe.'Co |
11 |
|
0 |
4119.53 |
4119.550) |
Fe |
1 |
||
0 |
4119.74 |
4119.751V |
2 |
1 |
||
0 |
4120.12 |
4120.075) |
0 |
|||
1 |
4120.35 |
4120.368 |
Fe |
4 |
||
2 |
4120.93 |
(4120.973) |
He |
Helium line at 4 120. 973 |
||
3 |
4121.46 |
4121.477 |
Cr-Co |
6d? |
||
Id |
4122.02 |
4122.049 |
2 |
Fe, Ti, Cr |
4 |
|
3 |
4122.80 |
4122.819 |
1 |
|||
5 |
4123.45 |
4123.477 |
2 |
La, Mn |
3 |
|
3 |
4123.93 |
4123.907 |
Fe |
5 |
||
2 |
4124.96 |
4124.938 |
2 |
|||
1 |
4125.93 |
4125.900 |
3 |
Fe,' — |
7 |
|
1 |
4126.35 |
4126.344 |
Fe |
4 |
||
1 |
4126.66 |
4126.673 |
Cr |
2 |
||
5 |
4127.86 |
4127.872 |
2 |
Fe |
8 |
|
5 |
4128.25 |
4128.251 |
Ce-V |
6d? |
||
0 |
4128.91 |
4128.894 |
2 |
|||
1 |
4129.41 |
4129.448 |
2 |
Ce' — |
5 |
|
5 |
4129.88 |
4129.882 |
Eu |
1 |
||
2 |
4130.83 |
4130.804 |
Ba |
2 |
||
0 |
4131.46 |
|||||
3d |
4132.16 |
ii 09 .'idol 4132.23o| |
V Fe |
2V 10 / |
||
1 |
4133.05 |
4133.062 |
Fe |
4 |
||
2d |
4133.93 |
4133.908 |
3 |
Fe, Ce |
5 |
|
2 |
4134.49 |
4134.492 |
Fe |
3 |
||
5 |
4134.84 |
4134.840 |
Fe |
5 |
||
2d? |
4135.56 |
4135.529 |
2 |
1 |
||
1 |
4136.02 |
|||||
2 |
4136.69 |
4i3e!678 |
Fe |
4 |
||
3 |
4137.26 |
4137.156) 4137. 567 j |
Fe |
6) 2J |
||
4 |
4137.79 |
4137.809 |
Fe.'Ce |
1 |
||
0 |
4138.31 |
4138.324 |
2 |
1 |
178
ECLIPSES AND SOLAR ENVELOPES
TABLE X. — Continued.
Flash Spectrum |
Wave Length Rowland |
Number of Lines Blended |
Sub- stance |
Inten- sity and Char- acter |
Remarks |
|
In- ten- sity |
Wave Length |
|||||
1 |
4139.08 |
4139.008 |
0 |
|||
0 |
4139.57 |
|||||
1 |
4140.24 |
4i40.'245 |
2 |
Fe,' — |
9 |
|
1 |
4141.81 ] |
4141.809 |
La |
0] |
||
1 |
4142.03 |
4142.025 |
Fe |
4 |
||
2 |
4142.56 J |
4142.542 |
4 |
Cr, — |
sj |
|
3 |
4143.28 |
(4143.30) |
Nd line at 4143. 30 |
|||
6 |
4144.05 |
(4143.919) ) |
He |
Helium line at 4143.919 |
||
4144.038 J |
Fe |
i5 |
||||
2 |
4144.63 |
4144.674 |
Ce |
ONd? |
||
2 |
4145.13) |
4145.152 |
Ce |
o\ |
||
0 |
4145.37] |
4145.357 |
ij |
|||
1 |
4145.84 |
4145.840 |
2 |
1 |
||
3 |
4146.23 |
4146.225 |
Fe |
3 |
||
0 |
4147.12 |
4147.145 |
2 |
|||
2 |
4147.69 |
4147.713 |
3 |
Mn, Fe |
7 |
|
4148.98 |
4148.948 |
Mn |
0 |
|||
10 |
4149.37 |
4149.360 |
Zr |
2 |
||
4149.923) |
2) |
[dentification doubtful |
||||
2 |
4150.03 |
4150. 056 / |
Ce |
00 1 |
||
0 |
4150.40 |
4150.411 |
4 |
|||
1 |
4150.68 |
|||||
3 |
4151.18 |
4isi.'i29 |
Zr.'Ti |
i |
||
6 |
4152.23 |
4152.248 |
3 |
La, Fe, Ce |
6 |
|
0 |
4152.68 |
C |
||||
0 |
4153.51 |
4i53.'542 |
Fe |
i |
||
2d |
4154.09 |
4154.112 |
2 |
Cr, Fe |
5 ' |
|
4154.65 |
4154.667 |
Fe |
4 |
|||
6 |
4156.30 |
4156.339 |
4 |
Nd, Zr |
5 |
|
3 |
4157.00 |
4156.970 |
Fe |
3d? |
||
3 |
4158.00 |
4157.948 |
Fe |
5 |
||
2d |
4159.00 |
4158.959 |
Fe |
5 |
||
0 |
4159.40 |
4159.353 |
5 |
|||
Od |
4160.57 |
4160.53 |
2 |
rlasselburg gives V |
||
4160.57 |
||||||
1 |
4161.23) |
4161.239 |
2 |
|||
5 |
4161. 65 J |
4161.682 |
Ti |
4 |
Spark line Ti |
|
2 |
4162.79 |
4162.724 |
2 |
2N |
||
10 |
4163.82 |
4163.818 |
Ti.'Cr |
Spark line Ti |
erto published makes the identification of lines a practical certainty. Hence, it must be concluded that the flash spectrum is a reversal of the Fraun- hofer spectrum, but with marked differences in the intensities in the two spectra.
" Measures of the 1905 spectra confirm Mitchell's 1901 results that hydrogen (H), helium (He), scan-
179
THE SUN
dium (Sc), titanium (Ti), strontium (Sr), vanadium (V), Zr, Y, Cr, Mn, Nd and Ce appear with a greater intensity in the flash than in the photospheric spec- trum, relative to the other elements. These eclipse results also confirm the prominence of enhanced lines."
Campbell's Observations.
Professor Campbell has invented, and used success- fully at several recent eclipses, a spectroscope with a moving plate. He begins exposures slightly before totality comes, and as the plate keeps falling, the spectra are produced in a continuous series at deter- minable times; and thus are adapted to give the whole history of the spectrum, as it changes from the photospheric spectrum (reflected by the air) to the chromospheric, or flash spectrum. It is well known that Professor Campbell and other members of the Lick expeditions have secured spectra of very high excellence with this and other apparatus in recent eclipses, of which the discussion is not yet com- pleted. Professor Campbell's full publication, and that of Mitchell, are awaited with keen interest.
CHROMOSPHERIC SPECTRA IN FULL DAYLIGHT.
Recently Adams, at the Mount Wilson Solar Ob- servatory, has obtained many photographs of the chromospheric spectrum in full sunlight. This method surpasses in accuracy of wave-length meas- urements, and may eventually rival in detail the best
180
ECLIPSES AND SOLAR ENVELOPES
eclipse " flash spectra." Adams' observations were made with the 60-foot-focus tower telescope, and the - 30-f oot-f ocus plane-grating spectroscope. Success de- pended on securing excellent definition of the sun's image, so that the spectroscope slit might be held exactly to the edge of the limb, without the light of the photosphere ''boiling over," so as to blot out the bright-line spectrum. The chromosphere is a stratum so thin that it is covered by the march of the moon at an eclipse in a very few seconds. Accordingly, there is not sufficient time during total eclipses for the ex- posure of a slit spectrograph of high dispersion, and for this reason slitless spectrographs of moderate dis- persion have usually been employed. Consequently, it is not practicable to get from eclipse "flash spectra'' such high precision of wave lengths as is necessary to decide the subtler points regarding the condition and nature of the chromosphere. Hence, the great advan- tage of supplementing eclipse work by observations at great dispersion in full sunlight, especially for the red end of the spectrum, where photography requires long exposures. Mr. Hale proposes to continue the work begun by Adams with the 60-foot tower telescope, and is making provision for an enlarged solar image. He expects greatly enriched results when the 150-foot tower telescope is available.
In the work thus far published by Hale and Adams the number of bright lines shown is far less than Mitchell obtained at the eclipse of 1905. Certain dif- ferences seem to indicate that the level of the spectra
181
THE SUN
photographed in full daylight is a little above the level best observed at eclipses. The wave length of the bright lines found seem to be practically identical with the wave lengths of corresponding dark lines in the photosphere. This circumstance, as Hale and Adams remark, does not lend support to Julius's con- tention, which will be noted in a later chapter, that the bright lines of the "flash spectrum" are due to anomalous refraction of light just outside the sun's limb. For, if this were the case, there would probably be a shifting towards the red of their apparent wave lengths from those of the dark lines of the photo- sphere. But Julius thinks the margins of discrepancy between the positions of the bright and dark lines, as given by Hale and Adams, still leave ground for his theory of anomalous dispersion. In order to permit of this interpretation, however, Julius imagines "the solar atmosphere to be honeycombed with irregular density gradients, which may be steeper than the underlying general radial gradient." Thus he finds the possibility that displacements of the chromo- spheric lines may be, now to the red, now to the vio- let, of the Fraunhofer lines. Most observers still retain the view that the chromospheric spectrum is essentially the reverse of the Fraunhofer spectrum and appears at the edge of the sun bright instead of dark because there is no such enormously brilliant spectrum background to dim, by comparison, the in- trinsic brightness of the lines themselves.
182
CHAPTER V
SUN-SPOTS, FACUL.E, AND GRANULATION
Sun-spot Periodicity. — Drift. — Distribution of Sun-spots. — For- mation and Life History. — Sun-spot Level. — Langley's Typical Sun-spot. — Faculae. — Granulation. — Sun-spot Spectra. — Cool- ness of Sun-spots. — Sun-spots and Magnetism. — Radial Motion in Spot Penumbras.
ALTHOUGH occasionally seen, and recorded much earlier without recognition of their solar origin, the history of sun-spots as solar phenomena dates from 1610, when they were independently discovered by Fabricius, Scheiner, and Galileo. The discovery fol- lowed naturally from the invention of the telescope in Holland, in 1608. There was at first some doubt (not shared by Fabricius or Galileo) whether the sun- spots were not planets. Indeed, sun-spots were for a time called in France the "Bourbonian Stars."
Viewed in a telescope, or projected on a screen, the sun-spots are plainly seen, and appear to consist of two well-marked parts; the umbra, apparently very dark, and the penumbra, a half-tone border around the umbra. Sun-spots differ greatly in size, shape, and darkness. Some large ones are TO of the sun's diameter, or five times the diameter of the earth, and sun-spot groups occasionally spread over an area of u 183
THE SUN
more than ro the sun's diameter. These great spots and spotted areas are rare.
SUN-SPOT PERIODICITY
Schwabe of Desau about 1843 discovered, as the result of systematic observing for nearly twenty years, that there is a periodicity in the occurrence of sun-spots. They are most frequent at intervals of about eleven years, and are nearly absent for a year or two in the interim. This sun-spot periodicity was exhaustively studied by Wolf of Zurich, who repre- sented the spottedness by a system now called " Wolf's sun-spot relative numbers. " These are com- puted by the formula, r = k(Wg +/), in which r is Wolf's number, g the number of groups and single spots observed, / the total number of spots which can be counted in these groups and single spots combined, and k a multiplier which depends on the conditions of observation and the telescope employed. Wolf took k as unity for himself when observing with a three- inch telescope and a power of 64. A less favored or less assiduous observer would receive k greater than unity, and one with a larger telescope and good opportunities for observing would receive a fractional value of k. Wolf's numbers seem arbitrary, but are found by photographic comparisons to be closely pro- portional to the spotted areas on the sun. One hun- dred as a sun-spot number corresponds to about TTFIT of the sun's visible disk covered by spots, including both umbras and penumbras.
184
SUN-SPOTS, FACUL^, AND GRANULATION
Wolf, by consulting all available sources, carried his sun-spot numbers back to 1610. His successor, Wolfer, has kept up the series from Wolf's death, in 1893, up to the present time. In Fig. 53 and Fig. 54 the curves show the run of spottedness during all this interval.1 It will be seen that the maxima and min- ima are not uniformly spaced ; but so that, while a mean sun-spot interval of 11.13 years is deduced by Professor Newcomb, the individual periods range between 7.3 and 17.1 years as extremes. These fea- tures are shown in the table on page 186.
Newcomb finds the average time of increasing spottedness, 4.62 years, of decreasing spottedness, 6.51 years. Having studied the total interval from 1610 to 1898 in three parts, he concludes that: " Un- derlying the periodic variations of spot activity there is a uniform cycle, unchanging from time to time, and determining the general mean of the activity."
The reader will note in the sun-spot curves that there is not only a great dissimilarity in the lengths of the individual periods, but also of their activities as measured by the maximum number of spots observed. Dr. Lockyer pointed out a relation between these phenomena which has been mentioned also by Halm and thoroughly confirmed by Wolfer. Call the time
1 In Fig. 53 the inclusion of new data leads to a modification as follows:
Year 1800 1801 1802 1803 1804 1805 1806
Mean. . .15.0 33.7 44.1 43.0 46.8 42.5 27.3 These data are corresponding to the mid-years. 185
THE SUN
TABLE XI. — Years of sun-spot maxima and minima and maximum intensities
Minima |
Difference |
Maxima |
Difference |
Maximum Wolf Number |
1610.8 |
1615.5 |
|||
1619.0 |
'8'.2 |
1626.0 |
io's |
|
1634.0 |
15.0 |
1639.5 |
13.5 |
|
1645.0 |
11.0 |
1649.0 |
9.5 |
|
1655.0 |
10.0 |
1660.0 |
11.0 |
|
1666.0 |
11.0 |
1675.0 |
15.0 |
|
1679.5 |
13.5 |
1685.0 |
10.0 |
|
1689.5 |
10.0 |
1693.0 |
8.0 |
|
1698.0 |
8.5 |
1705.5 |
12.5 |
|
1712.0 |
14.0 |
1718.2 |
12.7 |
|
1723.5 |
11.5 |
1727.5 |
9.3 |
|
1734.0 |
10.5 |
1738.7 |
11.2 |
|
1745.0 |
11.0 |
1750.3 |
11.6 |
'si |
1755.2 |
11.2 |
1761.5 |
11.2 |
80 |
1766.5 |
11.3 |
1769.7 |
8.2 |
103 |
1775.5 |
9.0 |
1778.4 |
8.7 |
151 |
1784.7 |
9.2 |
1788.1 |
9.7 |
133 |
1798.3 |
13.6 |
1805.2 |
17.1 |
47 |
1810.6 |
12.3 |
1816.4 |
11.2 |
46 |
1823.3 |
12.7 |
1829.9 |
13.5 |
67 |
1833.9 |
10.6 |
1837.2 |
7.3 |
137 |
1843.5 |
9.6 |
1848.1 |
10.9 |
125 |
1856.0 |
12.5 |
1860.1 |
12.0 |
95 |
1867.2 |
11.2 |
1870.6 |
10.5 |
132 |
1878.9 |
11.7 |
1883.9 |
13.3 |
65 |
1889.6 |
10.7 |
1894.1 |
10.2 |
84 |
1901.6, |
12,0 |
1906.4 |
12.3 |
60 |
interval from a minimum to the following maximum a, and that from the maximum to the following mini- mum b. The variations of a and of the ratio - proceed
o
in a sense contrary to the intensity of the outbreak of spottedness in each period. In other words, the more intense the outbreak of spots in any sun-spot period the shorter the time required for its development,
186
SUN-SPOTS, FACUL.E, AND GRANULATION
187
THE SUN
both actually and as compared with the time re- quired for its decay.
It is interesting to observe, also, that the interval from minimum to maximum is always much less than that from maximum to minimum. Attention will be drawn in Chapter X to the similarity between this characteristic and a certain type of stellar variation exemplified in the star Mira.
The causes which produce sun-spots being as yet doubtful, or perhaps it is better to say entirely un- known, the causes of their periodicity and of the ir- regularity of the periods are, of course, also unknown. Attempts have been made to connect the period with the times of revolution of the planets, and, indeed, the mean length of the sun-spot period is not far from the period of the revolution of Jupiter (11.86 years). No satisfactory case for a connection between these phenomena is yet made out. Schuster has recently applied a method of mathematical analysis fitted to bring out secondary periods which may underlie the average sun-spot periodicity of 1 1 . 1 3 years. He finds three well-marked periods of 11.125 years, 8.32 years, and 4.77 years. As curious mathematical coinci- dences he notes that the sum of the reciprocals of the first two periods equals the reciprocal of the third, and all three are nearly even fractions of 33| years. He finds the relative intensities of spot- tedness for his three periods variable, hence the inequality of the successive total periods produced by their combination. He inclines to attribute
188
SUN-SPOTS, FACUL.E, AND GRANULATION
sun-spots to causes outside the sun, perhaps to meteor swarms.
Several kinds of phenomena, some solar, others ter- restrial, are evidently closely associated with sun- spots and share in their periodicity. Firstly, the f acu- lae, or bright flecks on the solar surface, which are always seen most plentifully in sun-spot neighbor- hoods, naturally have the same seasons of maxima and minima. Secondly, the prominences, as stated in the preceding chapter, are most numerous at sun- spot maxima, and decrease in number, though not with so marked a change, as the number of sun-spots decreases. Thirdly, the form of the solar corona evi- dently goes through a periodic change simultaneous with the sun-spot cycle. Thus we speak of a solar corona having prolonged equatorial streamers of an arrow-head shape as "a sun-spot minimum corona," and one nearly equally developed in all directions as "a sun-spot maximum corona." Fourthly, the ter- restrial auroras (northern and southern lights) follow the sun-spot periodicity, as shown by Loomis and many others. Fifthly, changes in the earth's magnetic field occur in complete synchronism with the changes of sun-spot numbers. This connection is very close, for the agreement descends even to minute parallel- ism, as shown by the magnetic curves plotted in Fig. 53 and Fig. 54. Great sun-spots often seem to be the direct promoters of great magnetic disturbances (magnetic storms) and auroral displays. Maunder has found that the magnetic disturbances seem to
189
THE SUN
arise from restricted solar areas, not necessarily including sun-spots, and to go out in definite directions, or rather shafts of several degrees diameter, which rotate with the sun.1 When such a shaft strikes the earth a magnetic storm arises. Such lines of influence are not, he thinks, neces- sarily radial, but may follow coronal stream lines. Sixthly, the earth's surface air temperature is on the whole lower at sun-spot maximum than at sun-spot minimum. This relation is indicated at least for the United States in Fig. 54. The difference of mean temperature for the earth generally, ranging from 0°.5 to 1°.0 Centigrade for a change of 100-sun-spot num- bers, is shown by temperature statistics studied by Koppen, Nordmann, Newcomb, Abbot, Fowle, Arc- towski, and Bigelow. This will be further discussed in Chapter VII. Many other terrestrial changes, in rainfall, cloudiness, number of cyclones, panics, prices of foods, famines, growth of trees, even flights of in- sects, have been seriously compared with the sun- spot changes. In some of these phenomena there appear to be rather well-substantiated indications of a periodicity coincident with that of sun-spots, while such relations in many cases are probably purely fan- ciful.
So-called " great periods" of 33, 35, 55, and several hundred years have been proposed by various au- thors in order to explain the variations of the lengths
1 Shearman of Toronto discovered also a periodicity of auroral displays approximating the rotation period of the sun.
190
SUN-SPOTS, FACUL^E, AND GRANULATION
\ ' |
1 |
1 |
f |
890 2 4 6 6 1900 MPERATURES AND MAGNETISM. i States inland stations, 's). in diurnal range ) s, Chree. j |
||||||||||||||
f |
— - |
^ |
1 |
I |
s |
|||||||||||||
V |
\ |
V |
||||||||||||||||
« |
{* |
) |
/ |
; |
7 |
|||||||||||||
^ |
^ |
/ |
r |
y |
^ |
|||||||||||||
<, |
-f |
1 |
^ |
r |
/ |
|||||||||||||
V |
/ |
2 |
/ |
|||||||||||||||
.^r |
^ |
{ |
( |
1 |
||||||||||||||
? |
V |
\ |
I |
|||||||||||||||
^ |
x |
X |
\ |
X- |
^ |
|||||||||||||
V 1 |
»• n ^ |
5 |
A |
\ |
||||||||||||||
'-c |
2 |
1 |
1 |
J |
a"" I860 2468 FIG. 54. — SUN-SPOTS AND TERRESTRIAL Ti I. Temperature departures, Unite( II. Sun-spot relative numbers (WoH III. Magnetic declination f meg IV. Magnetic horizontal force I Elli |
|||||||||||||
^ |
ff~~> |
I |
/ |
|||||||||||||||
^ |
^^ |
1 |
j |
|||||||||||||||
'' 5 |
/ |
^ |
f |
/ |
||||||||||||||
<r |
./- |
^ |
^ |
|||||||||||||||
• |
> i> |
^ |
; |
f |
^ |
|||||||||||||
<z |
s |
\ |
c |
L |
||||||||||||||
i |
•v- |
• — |
/ |
f |
) |
|||||||||||||
\ |
^ |
\ |
\ |
S^ |
||||||||||||||
*• — |
. |
•—x |
v |
\ |
s |
v |
||||||||||||
^r |
i. — — |
> |
\ |
v |
\ |
\ |
||||||||||||
1 — |
1 r |
^r- |
1 |
7 |
||||||||||||||
~r |
-s4-* |
^ |
7 |
|||||||||||||||
D m o in o - *<3 *c> *o* '— • li + + |
o o o SO "fr CO |
0 |
! s s s |
§555 |
191
THE SUN
and intensities of the eleven year periods, and the changes of rainfall, times of harvest, and other changes of terrestrial phenomena said to be indicated by history or tradition. The tendency to groups of three in respect to the intensity of the successive sun- spot outbreaks has been mentioned by various writ- ers, and may be noted in the table. This helps to sus- tain belief in a thirty-three year period, but it will be noted that the four maxima 1830 to 1870 were un- commonly intense (perhaps excepting the intermedi- ate one of 1860). The question of the reality of "great periods" seem to require further lapse of years to de- cide it.
SUN-SPOT DRIFT
If we imagine an observer on the moon to watch the clouds on the earth's surface, they would appear to him on the whole to indicate a mean rotation period of about twenty-four hours for the earth. But he would also discover that many, and perhaps nearly all, of the cloudy areas had proper motions of their own besides, so that no single cloud would give cor- rectly the rotation period of the earth. So it is with the sun-spots, for, after allowing for the sun's average rotation period, nearly every spot has a motion of its own. Carrington found a slight tendency of spots be- tween 20° North and 20° South latitudes to approach the equator, and outside these latitudes a more de- cided tendency to approach the poles. Faye held that spots persistently describe little ellipses on the sun's surface of one or two days' period. It is said
192
SUN-SPOTS, FACUL^E, AND GRANULATION
that an actively changing spot is apt to move forward by irregular jerks. When a spot divides, the parts are apt to separate rapidly.
DISTRIBUTION OF SUN-SPOTS Sun-spots very seldom occur at higher latitudes than 40°. Within the sun-spot belt 80° wide, as thus, denned, the distribution of spots is irregular. They occur mainly in two zones on either side of the equa- tor, between latitudes 10° and 30°. As regards the northern and southern hemispheres the number oc- curring in a very long period of years is practically equal, but there is often a great inequality for several years in succession. A remarkable instance of this irregularity occurred between 1672 and 1704, when no spots were recorded in the northern hemisphere, and the appearance of a few there in 1705 was noted by the French Academy as a very extraordinary event. Newcomb draws attention in the four cycles 1856-1898 to a marked and growing preponderance of spots in the southern hemisphere. A peculiarity of sun-spot distribution likely to prove of great theo- retical significance was discovered by Spoerer, and is confirmed by Greenwich observations. There seems to be a close connection between the latitudes of great prevalence and the periodicity of sun-spots. Young states the matter as follows :
" Speaking broadly, the disturbance which pro- duces the spots of a given sun-spot period first mani- fests itself in two. belts about 30° north and south of
193
THE SUN
the sun's equator. These belts then draw in toward the equator, and the sun-spot maximum occurs when their latitude is about 16°; while the disturbance gradually and finally dies out at a latitude of 8° or 10°, some twelve or fourteen years after its first outbreak. Two or three years before this disappearance, how- ever, two new zones of disturbance show themselves. Thus, at the sun-spot minimum there are four well- marked spot-belts; two near the equator, due to the expiring disturbance, and two in high latitudes, due
18 |
55 18 |
60 18 |
65 18 |
70 18 |
75 18 |
80 |
||||||||||||||||||||
L. |
f^ |
W. |
||||||||||||||||||||||||
! |
s |
|||||||||||||||||||||||||
30° 26° |
j |
^ |
/ |
\ |
100 |
|||||||||||||||||||||
\ |
,/' |
N |
s |
^s |
1 |
-, |
||||||||||||||||||||
22° |
\ |
/ |
*" |
Si |
'\ |
|||||||||||||||||||||
18 |
50 |
|||||||||||||||||||||||||
S,^ |
-- |
-- |
s |
|||||||||||||||||||||||
14° |
/ |
X |
••v |
/ |
^> |
•^ |
v- |
's^ |
/ |
|||||||||||||||||
10° |
- |
-'' |
""" |
^. |
-»» |
^ |
'- |
0 |
FIG. 55. — SPOERER'S CURVES OF SUN-SPOT LATITUDE.
to the newly beginning outbreak ; and it appears that the true sun-spot cycle is from twelve to fourteen years long, each beginning in high latitudes before the preceding one has expired near the equator.
" Fig. 55 illustrates this, embodying Spoerer's results from 1855 to 1880. The dotted curves show Wolf's sun-spot curve for that period, the vertical column at the right of the figure, marked W at the top, giving Wolf's ' relative numbers. ' The two continuous curves, on the other hand, give the solar latitudes of the two series of spots that invaded the sun's surface in those years. The scale of latitudes is on the left hand. The
194
SUN-SPOTS, FACUL.E, AND GRANULATION
first series began in 1856 and ended in 1868; the sec- ond broke out in 1866 and lasted until 1880. During these years it happened that there was very little dif- ference between the northern and southern hemi- spheres of the sun. "
In a summary of the results of solar observations made at Greenwich from 1874 to 1902 the Astrono- mer Royal, Christie, gives data showing the prevail- ing latitudes at which they occurred in different parts of the sun-spot cycles. The maximum latitude at which spots occurred was 42°, but they could only be regarded as sporadic phenomena above latitude 33°. Preceding a time of sun-spot minimum, the prevailing spottedness occurred at low latitudes, and when the spots reappeared after minimum it was generally at high latitudes. The equatorial belt from 5° to -5° was never a center of spottedness. These facts are indicated by the following table, abridged from the data given by the Astronomer Royal.
Year |
18-* |
80 |
82 |
84 |
86 |
Q1 |
PR |
95 |
97 |
Centers of |
N |
21° |
16° |
11° |
9° |
21° |
15° |
12° |
8° |
spottedness |
S |
19° |
18° |
11° |
10° |
20° |
15° |
12° |
7° |
Wolf's numbers |
32 |
58 |
63 |
25 |
38 |
84 |
62 |
28 |
The protuberances, on the contrary, as shown by Ricco and the Lockyers, and confirmed by Mascari, have their zones of maximum frequency transferred from low toward higher latitudes as the sun-spot cycles progress.
195
THE SUN
SUN-SPOT FORMATION AND LIFE-HISTORY
As regards the formation and life history of sun- spots, Young has described the phenomena in these words :
" There is no regular process for the formation of a spot. Sometimes it is gradual, requiring days or even weeks for its full development, and sometimes a single day suffices. Generally, for some time before the ap- pearance of the spot, there is an evident disturbance of the solar surface, manifested especially by the pres- ence of numerous and brilliant faculae,1 among which, " pores" or minute black dots are scattered. These enlarge, and between them appear grayish patches, apparently caused by a dark mass lying veiled below a thin layer of luminous filaments. The veil grows gradually thinner, and vanishes, giving us at last the completed spot with its perfect penumbra. The "pores," some of them, coalesce with the principal spot, some disappear, and others constitute the at- tendant train. When the spot is once completely formed, it assumes usually an approximately circular form, and remains without striking change until its dissolution. As its end approaches, the surrounding photosphere seems to crowd in upon and cover and overwhelm the penumbra. Bridges of light, often many times brighter than the average of the solar surface, push across the umbra, the arrangement of
1 This is Secchi's view. Lockyer maintains that the spots appear before the faculae.
196
SUN-SPOTS, FACUL.E, AND GRANULATION
the penumbra filaments becomes confused, and, as Secchi expresses it, the luminous matter of the photo- sphere seems to tumble pell-mell into the chasm, which disappears and leaves a disturbed surface marked with faculae, which in their turn subside after a time. As intimated before, however, the disturb- ance is not unfrequently renewed at the same point after a few days, and a fresh spot appears just where the old one was overwhelmed.
" The spots usually appear not singly, but in groups — at least, isolated spots of any size are less common than groups. Very often a large spot is followed upon the eastern side by a train of smaller ones;, many of which, in such a case, are apt to be very imperfect in structure, sometimes showing no umbra at all, often having a penumbra only upon one side, and usually irregular in form. It is noticeable, also, that in such cases, when any considerable change of form or struc- ture shows itself in the principal spot of a group, it seems to rush forward (westward) upon the solar surface, leaving its attendants trailing behind. When a large spot divides into two or more, as often hap- pens, the parts usually seem to repel each other and fly asunder with great velocity — great, that is, if reckoned in miles per hour, though, of course, to a telescopic observer the motion is very slow, since one can only barely see upon the sun's surface a change of place amounting to two hundred miles, even with a very high magnifying power. Velocities of three or four hundred miles an hour are usual, and velocities
197
THE SUN
of one thousand miles, and even more, are by no means exceptional.
"The average life of a sun-spot may be taken as two or three months; the longest yet on record is that of a spot observed in 1840 and 1841, which lasted eighteen months. There are cases, however, where the disappearance of a spot is very soon followed by the appearance of another at the same point, and some- times this alternate disappearance and reappearance is several times repeated. While some spots are thus long-lived, others, however, endure only for a day or two, and sometimes only for a few hours. "
Carrington, Secchi, Perry, Maunder, and Sid- greaves have all noted the tendency of spots to recur in the same positions, but not in a sense indicative of permanent special eruptive places, as in the case of terrestrial volcanoes. Father Sidgreaves says: "They are indications of a more enduring state of disturb- ance than is measured by the lifetime of a single spot, for it is not improbable that a recurrence springs from the same source as its predecessor. And, if this be true, the spots must be more subject to drift than their underlying origins, for nearly always the recur- ring spot is found to the rear of the former position. "
According to the spectroheliographic investiga- tions of Fox :l i i Spot birth is always accompanied by, and generally antedated by, an eruption" (i. e., erup- tive prominence). " In the early hours of the life of a spot the eruption may partially or entirely cover the
lAstrophysical Journal, vol. xxviii, p. 255, 1908. 198
SUN-SPOTS, FACUL^E, AND GRANULATION
spot, and often may precede it, in the direction of solar rotation. An eruption is seldom seen preceding a mature single spot, but if present will be following it at the edge of the penumbra, perhaps encroaching somewhat. If the spot is actively growing, eruptions are almost certain to be found on the following edge. Eruptions accompany spots in rapid decline, being often seen at the ends of the bridges. I think the evidence of the Rumford spectroheliograms fairly conclusive in showing that the spot has its genesis in the eruption. The phenomenon of spot development following the appearance of an eruption is so general that it is possible, on the appearance of an isolated eruption, to predict with certainty the advent of a spot. When the spot is well developed it stimulates new eruptions. " The " eruptions" mentioned by Fox are, of course, seen with the spectroheliograph any- where on the sun's disk, but when close to the limb they are recognized by him to be really "the bases of the eruptive prominences. "
THE SUN-SPOT LEVEL
The level of sun-spots is a question which has been discussed for over a century, and often with consider- able vehemence. In 1769, Dr. A. Wilson of Glas- gow advocated the view that sun-spots are depres- sions of the sun's surface. He observed that when a spot first appears on the eastern edge of the sun the penumbra is well marked on the side nearest the edge of the sun, but nearly invisible on the side next the 15 199
THE SUN
sun's center, while the umbra scarcely shows at all, being as if hidden behind a bank. As the spot ad- vances towards the center, according to A. Wilson, the advancing and following sides of the penumbra become more equal, and the umbra covers an increas- ing fraction of the total width of the spot. Having passed the center, the spot naturally exhibits the op- posite succession of phenomena. This progress in ap- pearance would be conclusive evidence that spots are depressions if it were universally admitted to be real. Many spots are so unsymmetrical, even at the center of the sun, as to be unfavorable objects on which to test A. Wilson's view. Many spots alter their shape in crossing the sun's disk quite apart from any change due to the sun's spherical form. In the last twenty years several very assiduous observers have published conclusions based on very numerous observations; and, even when discussing the spots occurring in the same course of years, about as many disagree with A. Wilson's view as support him. It seems most prob- able, therefore, that the level of the sun-spot phe- nomena seen by ordinary observation differs very little, if at all, from that of the surrounding bright surface of the sun.
LANGLEY'S TYPICAL SUN-SPOT
Owing to the effect of the sun's rays in heating the surface of the earth, and thereby causing the ascent of warm currents of air which spoil the " seeing," the observer is at a disadvantage in studying the minute
200
SUN-SPOTS, FACULJ:, AND GRANULATION
features of the sun as compared with the moon or other night objects. The " seeing" on the sun is gen- erally better in the hours soon after sunrise and before sunset, when the heating of the sun's rays is dimin- ished both by passing through a thick stratum of air and by striking the earth's surface obliquely. Some- times the presence of thick haze or light uniform cloudiness appears to favor good definition, but often these conditions are connected with atmospheric dis- turbances so nearly in line with the sun as to spoil the " seeing. " Good solar " seeing" is seldom found when a clear blue sky, a brisk breeze, and high altitude of the sun occur simultaneously. With the hindrance thus occasioned by the irregularities of density in the earth's atmosphere to contend with, solar observers, as a rule, find only comparatively rare instants when really satisfactory views of the sun's surface may be obtained. By combining with extraordinary skill the impressions received in the instants of best "seeing," which were the reward of several years of assiduous observing, the late Dr. S. P. Langley produced, in 1873, his famous sketch of the "typical sun-spot," a copy of which is reproduced as the frontispiece. This is generally conceded to represent better than any photographs, and even better than anyone is likely to see for himself in the telescope, the appearance of a sun-spot and its surroundings as seen under the best purely telescopic observation.
201
THE SUN
FACUL^E
Next to sun-spots, the most prominent solar fea- tures, and closely associated with the life history of spots, are the faculae, or bright patches which are most abundantly seen near the borders of the sun's disk. Their appearance has been likened by Young to the flecks of foam which dot the water beneath a waterfall. They are very prevalent in the neigh- borhood of sun-spots, but, unlike them, they are found all over the surface of the sun, though spar- ingly near the poles. It is difficult to see them near the center of the sun's disk. As stated in Chapter III, the brilliancy of the solar surface is not uniform all over the disk, but falls off very greatly near the edges. Speaking roughly, the faculae, on the other hand, may be regarded as equally bright wherever seen on the sun's disk, and hence come out more distinct1-/ near the edges, where the background is less brilliant. The prevalence of faculae has maxima and minima synchronous with the sun-spot period.
GRANULATION
Besides the sun-spots and the faculse, there is seen under good observing conditions a general granu- lated appearance all over the sun's surface. Many years ago much controversy was waged over the exact forms of the granules, some observers compar- ing them to rice grains, others to willow leaves, and others to bits of straw. These patches of differing
202
PLATE XVI
6 h 47 m.
7 h 37m.
PHOTOGRAPHS OF A PORTION OF THE SUN. (Janssen.) Meudon, June 1, 1878. Interval, 50 minutes.
SUN-SPOTS, FACUL^E, AND GRANULATION
brilliance are really immense areas of 10,000 to 50,000 square miles, and are probably not of a regular pat- tern at all, so that little insight into solar conditions is had by the discussion of their mere forms. In Langley's sun-spot drawing they are depicted in great numbers, and with various shapes, quite as they are apt to occur. Plate XVI is a reproduction of two of Janssen's celebrated photographs of them.
SUN-SPOT SPECTRA
The spectrum of a sun-spot differs from that of the photosphere in several significant ways. (1) As meas- ured by the bolometer or other photometric methods, its energy is far weaker in the violet. This is shown in the accompanying comparisons between the intensi- ties of the spectra of sun-spots and of the photosphere near the center of the sun's disk. The data for the ultra-violet spectrum are from the work of Schwartz- child and Villiger, and the remainder from the work of the Smithsonian observers.
Wave Lengths (\ = ) |
(V.320 |
Q/i.448 |
O/x.586 |
O/i.799 |
l/t.218 |
2/i.llS |
Ratio of umbra |
0.12 |
0.377 |
0.424 |
0.53.5 |
0.610 |
0.761 |
brightness photosphere |
As different spots differ in darkness of their centers, too much reliance should not be placed on the transi- tion of relative brightness from \ = 0.320/1. to X = 0.448/4, as given above. The remainder of the data, however, all applies to the same spot observed by the
203
THE SUN
same observers, and should, therefore, be comparable. There are three ways of explaining the progressive relative weakness of the shorter wave-length rays in sun-spots. The sun-spot temperatures may be much below those of the photosphere, there may be a greater amount of absorption or scattering of the light above the spots, or, finally, the phenomenon may be due to the action of both these causes. It has lately been made practically certain that the first- mentioned cause, at least, is operative. This is proved by the work on sun-spot spectra noted below. Several observers have found that the contrast of brightness between the sun-spots and the photo- sphere decreases towards the sun's limb. Langley, and also Frost, found indications that at the very limb the total radiation of the sun-spot umbra is ac- tually stronger than that of the photosphere. W. E. Wilson observed that the ratio of the brightness of the spot umbra to that of the photosphere at the sun's center did not change from the center to ninety-five per cent out on the solar radius, whereas the ratio of brightness of the umbra to the surroundings increased from TOZT to rW- He could not confirm Frost's and Langley's result. Schwartzchild and Villiger, observ- ing at wave-length 0.32//, in the ultra-violet, found the ratio of brightness of sun-spots to the surrounding photosphere at the center ten to fourteen per cent, but close to the limb it was thirty to fifty per cent. It has already been stated that the photosphere at the limb of the sun is less bright than it is at the cen-
204
SUN-SPOTS, FACUL.E, AND GRANULATION
ter, and the exact amount of change has been given for various wave lengths in Chapter III. Accordingly it is easy to see that if, as observed by W. E. Wilson, the sun-spot umbra remains nearly unchanged in its intrinsic brightness wherever seen upon the sun, the results just mentioned would tend to follow. It seems hard to believe, however, that the radiation of the spot-umbra at the limb could actually exceed that of the surrounding photosphere, as observed by Frost and Langley, and further experiments along this line should be made.
(2) In sun-spot spectra many Fraunhofer lines are strengthened and many weakened as compared with the same lines in the photospheric spectrum.1 From Adams' summary of the subject2 1 take the following data. Calcium has sixty lines in Rowland's table be- tween \ = 0.40/4 and A, = 0.70/u, and with one possi- ble exception all are strengthened in sun-spots. The strengthening increases absolutely, and also rela- tively to the intensities of the lines affected, with in- creasing wave length. With iron there are 1 , 108 lines in the same interval of Rowland's table, of which 784 are affected in spots. Of these, 558 are due to iron
1 A spectrum absorption line is said to be strengthened when, by reason of its becoming broader without becoming less dark, or by reason of its becoming darker, or from both changes, it presents a greater contrast to the adjoining spectrum. Weakening a spectrum line implies an opposite change. In either case the term is relative, and may really mean the alteration of the adjoining spectrum, with- out change in the line itself, in such a manner that the contrast of the line is altered.
2 Contributions of the Mount Wilson Solar Observatory, No. 40.
205
THE SUN
alone, the others being blends of iron lines with very close lines of other elements. Of the 558 purely iron lines affected, 300 are strengthened and 258 weakened in sun-spots. Hydrogen has four lines in the region under discussion, and all are weakened. The case is so striking that it is worth giving in full:
TABLE XII. — Hydrogen spectrum in sun-spots
Line |
Wave Length |
Intensity |
|
Photospheric |
Sun-spot |
||
Hs |
4101.848 4340.471 4861.350 6562.835 |
40N 20N 30 40 |
1 4 10 25 |
Hv |
|||
i«.: |
|||
H£ |
|||
The following table from Adams' publication shows the behavior of the spot lines of thirteen different ele- ments :
TABLE XIII. — Spectrum lines affected in
Total |
Number of Lines Strengthened |
Number of Lines Weakened |
Percentage of Total Number |
|||||
Element |
i oT/ai Number Lines |
Com- |
Com- |
|||||
One Ele- ment |
pound Lines and |
One Ele- ment |
pound Lines and |
Strength- ened |
Weak- ened |
Affec- ted |
||
Blends |
Blends |
|||||||
Calcium |
60 |
43 |
16 |
98 |
98 |
|||
Chromium. . . |
386 |
200 |
75 |
'36 |
3! |
71 |
"\7 |
88 |
Cobalt |
118 |
26 |
25 |
17 |
14 |
43 |
26 |
69 |
Hydrogen. . . |
4 |
4 |
100 |
100 |
||||
Iron. . |
1108 |
300 |
i27 |
258 |
98 |
S9 ' |
32 |
71 |
Magnesium . . |
8 |
3 |
1 |
38 |
12 |
50 |
||
Manganese . . Nickel |
167 251 |
68 48 |
'si 24 |
15 106 |
'{) 26 |
59 29 |
14 53 |
73 82 |
Scandium . . . |
45 |
30 |
3 |
67 |
7 |
74 |
||
Silicon |
9 |
8 |
'i |
100 |
100 |
|||
Sodium |
8 |
"8 |
100 |
100 |
||||
Titanium .... |
432 |
247 |
'73 |
'46 |
28 |
74 |
"\7 |
91 |
Vanadium. . . |
176 |
114 |
37 |
9 |
5 |
86 |
8 |
94 |
206
SUN-SPOTS, FACUL.E, AND GRANULATION
COOLNESS OF SUN-SPOTS
If the layer which produces the Fraunhofer lines over the spots were of the same temperature that it is over the photosphere, the lines in spots would tend to appear weakened; because, while the emission in the lines would in that case remain really unchanged, the spectrum background against which they are seen would be weakened, and approach the brightness of the lines, as has been seen under Caption 1 . Since the reduction of the background in sun-spot spectra is greatest for short wave lengths, the violet lines would be most weakened in the case we are considering. This is, indeed, the case for hydrogen, and may be explained in that case perhaps as a consequence of high level, but in fact the majority of sun-spots lines are strengthened, and this in itself may be regarded as evidence that the " re versing layer" for most ele- ments is cooler over spots than over the photosphere. Besides this general consideration, there are several others, now to be mentioned, which point to the same conclusion.
Lines which are relatively stronger in the electric spark than in the arc, when produced as bright lines in the laboratory, are called " enhanced lines." Of 144 enhanced lines observed in spots, says Adams, 11 130 are distinctly weakened, none are strengthened, while sixteen show no marked change." This almost universal weakening of enhanced lines in sun-spots is shown as follows, to be evidence of a low tempera-
207
THE SUN
ture in the sun-spot reversing layer. By Kirchhoff's law (see Chapter II) emission and absorption are pro- portional. Hence, if it requires the conditions of the spark to produce certain emission lines strongly, it will also require the conditions of the spark to cause the operative gases to absorb strongly in these lines. But spark versus arc conditions are to be regarded as of high versus lower temperatures, a view fully con- firmed by the experiments of Hale, Adams, and Gale with strong and weak arcs, and those of King with the electric furnace at high and low temperatures. Accordingly, the weakening of the enhanced lines in the sun-spot spectrum, in opposition to the prevailing strengthening of lines in spots, is explained by assum- ing that the spot vapors are too cool to produce strong absorption of enhanced lines.
A third line of evidence showing that the reversing layer is cooler over sun-spots is furnished by a de- tailed comparison of the spectra of sun-spots and photosphere on the one hand, and of low and high temperatures in the arc or electric furnace on the other. This comparison was begun by Hale, Adams, and Gale, and continued by King. Adams gives in a long table the results of such a comparison for the lines of iron. From this table several of the most well-marked cases, typical of strengthening, weaken- ing, and neutrality, are given in the following table.
In general, within the error of measurement, lines strengthened in the cool arc are strengthened in sun- spots, those weakened in the cool arc are weakened in
208
SUN-SPOTS, FACUL.E, AND GRANULATION
TABLE XIV. — Sun-spot, hot arc and cool arc spectra
i Intensity |
Intensity |
||||||
Wave Length (Rowland's) |
Spot ratio |
Arc ratio |
Dis- crepancy |
||||
Sun |
Spot |
Hot arc |
Cool arc |
||||
4118.708 |
5 |
4 |
16 |
12 |
1.2 |
1.3 |
—0.1 |
4291.630 |
2 |
3 |
8 |
16 |
0.7 |
0.5 |
+0.2 |
4325.939 |
8 |
7 |
48 |
40 |
1.1 |
1.2 |
—0.1 |
4461.818 |
4 |
7 |
19 |
40 |
0.6 |
0.5 |
+0.1 |
4531.327 |
5 |
7 |
16 |
24 |
0.7 |
0.7 |
+0.0 |
4939.868 |
3 |
5 |
10 |
' 18 |
0.6 |
0.6 |
0.0 |
5083.518 |
4 |
6 |
12 |
22 |
0.7 |
0.5 |
+0.2 |
5202.516 |
4 |
4 |
14 |
16 |
1.0 |
0.9 |
+0.1 |
5333.089 |
4 |
7 |
7 |
16 |
0.6 |
0.4 |
+0.2 |
5405.989 |
6 |
10 |
40 |
80 |
0.6 |
0.5 |
0.1 |
6024.281 |
7 |
7 |
13 |
13 |
1.0 |
1.0 |
0.0 |
sun-spots, and those unchanged in one are unchanged in the other, and all by similar proportions. It fol- lows from this, by a similar line of argument to that just given for enhanced lines, that the reversing layer is relatively cooler over sun-spots than over the pho- tosphere*
A fourth phenomenon strongly indicating the same conclusion is the highly conspicuous presence in sun- spot spectra of flutings, or rythmic banded appear- ances, immensely numerous, and characteristic re- spectively of the spectra of titanium oxide, magnes- ium hydride, and calcium hydride. The identifica- tions of these flutings were discovered respectively by Hale, Adams, and Gale, by Fowler and by Olm- sted. These and other molecular compounds give, as Evershed has stated, very slight and not always per-
209
THE SUN
ceptible evidence of their presence in the photo- spheric spectrum. It is well known that high tem- peratures tend to produce complete dissociation of molecular compounds. The copious appearance of the lines of compounds in the spectra of sun-spots would be very strong evidence of the relatively low temperature in the reversing layer above spots, even if unsupported by the other evidences given above, and by many other minor phenomena of which space forbids the mention.
According to Father Cortie,1 steam also occurs in sun-spots, for he finds water-vapor lines among those widened in sun-spot spectra. He cites experiments, too, which indicate that the spectrum of magnesium hydride could not show in sun-spots if water vapor was not also present. Evershed, however, concludes from observations at the high and dry station of Kodaikanal that: "On the whole, it must be ad- mitted that the evidence for the strengthening of telluric lines, of whatever origin, in spot spectra is practically negligible."
An excellent photographic map of the sun-spot spectrum, contrasted with that of the photosphere, has been prepared at the Mount Wilson Solar Ob- servatory and distributed to solar observers. Plate XVII, reproduced here by the permission of the Director, shows a section of this map including the b group. Although no engraving can do full justice to the original, the reader will be able to
1Astrophysical Journal, vol. xxviii, p. 379, 1908. 210
SUN-SPOTS, FACUL.E, AND GRANULATION
note for himself some of the features mentioned
above.
SUN-SPOTS AND MAGNETISM
In the year 1908, Hale discovered the existence of a magnetic field in spots, which betrays its presence by the widening, doubling, or tripling of a great number of spectral lines. As stated in Chapter II, Zeeman discovered, about 1896, that most lines of the spec- trum are separated into two components when viewed along the lines of force of a powerful magnet, and the two components are circularly polarized in opposite directions. With less powerful fields, the lines are not clearly doubled, only widened, but their right- and left-hand edges exhibit in this case traces of op- posite circular polarization. Hale applied this test of polarization to the most widened lines of sun-spots by introducing a Fresnel rhornb to convert the sup- posed circular to plane polarization, and found the right-hand or left-hand edge of the lines could be cut off at will, according to the position of the Nicol prism used for analyzing the character of polarization of the light. Some lines are triple in spots, but these seeming discrepancies proved to be the best of evi- dence of the effect of a magnetic field. For when the same lines were examined in the laboratory they proved exceptional, and to become triple instead of double when viewed along the magnetic lines of force. Hale's .brilliant discovery has cleared up one of the most puzzling questions relating to the sun-spot spectrum.
THE SUN
By polarization studies, Hale found that sun-spot fields are not always of the same polarity. Very often a pair of sun-spots quite near together are found to be of opposite polarity. In general, the polarity of spots in the sun's southern hemisphere is opposite to that in the northern, but there are very numerous exceptions to this rule, as, of course, in the case of double spots, as just mentioned. Spots near the sun's limb, since they present their magnetic lines of force nearly at right angles to our line of sight, tend to show triple lines where doublets would be seen near the center of the disk.
The cause of the magnetic field in sun-spots is a most interesting problem. Rowland showed many years ago that static electric charges, in rotation, produce electro-magnetic effects similar to those pro- duced by electric currents in coils of wire. This seems to point the way to a solution, for, as stated in the account of the sgectroheliographic results in Chapter III, the sun when viewed through the hydrogen line Ha (C) shows curved formations (see Plate XI), which seem to indicate spiral motion in sun-spot neighborhoods. In such Ha photographs of double spots, which give opposing magnetic polarity, the curves which surround the spots seem to present the appearance not unlike those seen among iron filings on a sheet of paper acted upon by a pair of opposite magnetic poles. It seems, then, not improbable that whirling motions or vortices exist in sun-spots, and that these carry along electrically charged par-
212
SUN-SPOTS, FACUL.E, AND GRANULATION
tides which produce the observed magnetic fields. The impression was at first that these charges were the so-called ions, or bodies smaller than atoms, re- cently made known by J. J. Thomson arid others; but great difficulty was found in accounting for their isolation in sun-spots in sufficient numbers. It was suggested to Mr. Hale by the writer that the mole- cules of the compounds shown in the sun-spot spec- trum, or perhaps even the relatively cooled elemen- tary gases in spots, might very probably be regarded as sufficiently different from the surroundings to pro- duce frictional electricity, when whirled about in the spots, just as steam becomes electrified in Arm- strong's machine when, carrying water-droplets, it issues from an orifice. Further discussion of the matter will be found in Chapter VI.
« RADIAL MOTION IN SPOT PENUMBRAS
Evershed has lately observed shifting of spectral lines in the penumbras of spots situated at consider- able distance from the center of the sun's limb. This seems to indicate motion nearly radial to the center of the spots, as if material was coming to the sun's surface in the sun-spot centers, and then spreading out in all directions, like smoke from a volcano. Nevertheless, no spectroscopic evidence of motion in spots radial to the center of the sun has ever been ob- tained.1 Adams has lately sought to find evidences
1 As this is being published St. John has observed high level gases moving downwards in spots.
213
THE SUN
of increased or decreased pressure in the reversing layer over sun-spots, from shifting of lines known to be subject to large shifts when their sources are under pressure, but he was unable to discover evidences of altered pressure. The significance of these facts will be discussed in Chapter VI.
CHAPTER VI.
WHAT IS THE SUN?
Young's Views. — Halm's Views. — Schmidt's Hypothesis. — Julius' Views. — The Author's Views.
BELIEVING that the views of the late Professor Young probably are still shared by a majority of as- tronomers, even after the lapse of fifteen years since the appearance of the last revision of his work, "The Sun, " we shall begin this chapter by quoting a part of the summary which he gives in his Chapter IX. We shall then take up the solar theories of Halm, Schmidt, and Julius. In the remainder of the chapter we shall consider still another view of the matter, which the present writer inclines to adopt.
YOUNG'S VIEWS
Quoting from Young's "The Sun:"
"Fig. 56 is intended to present to the eye, more clearly than any mere description, the constitution of the sun, and the relation of the different concentric shells or envelopes as conceived by the writer.
"The picture is an ideal section through the center. The black disk represents the inner nucleus, which is not accessible to observation, its nature and constitu- tion being a mere matter of inference. The white 16 215
THE SUN
ring surrounding it is the photosphere, or shell of in- candescent cloud which forms the visible surface. The depth, or thickness, of this shell is quite un- known; it may be many times thicker than rep- resented, or possibly some- what thinner. Nor is it certain whether it is separated from the inner core by a definite sur- face, or whether, on the other hand, there is no distinct bound- ary between them.
" The outer surface of the photosphere, however, is cer- tainly pretty sharply defined, though very irregular, rising at points into faculae, and depressed at others in spots, as shown in the figure.
" Immediately above this lies the so-called 'revers- ing stratum, ' in which the Fraunhof er lines originate.
216
FIG. 56. — SOLAR DIAGRAM. (Young.)
WHAT IS THE SUN?
It is to be noted, however, that the gases which com- pose this stratum do not merely overlie the photo- sphere, but they also fill the interspaces between the photospheric clouds, forming the atmosphere in which they float, and an attempt has been made to indicate this fact in the diagram.
"Above the ' re versing stratum' lies the scarlet chromosphere, with prominences of various forms and dimensions rising high above the solar surface; and over, and embracing all, is the coronal atmosphere and the mysterious radiance of clouds, rifts, and streamers, fading gradually into the outer darkness.
"At the center of the sun the earth is represented in its true relative dimensions — or of the three inches which is taken as the scale of the sun's diameter. This scale reduces our globe to a little dot only fa of an inch across. Around it, at its proper distance, is drawn the orbit of the moon, still far within the pho- tosphere, the moon herself being fairly represented by any one of the minute points which make up the dotted line that indicates her path.
"The central nucleus is made black in the picture, simply for convenience, and not with any purpose to indicate that the matter which composes it is cooler or even less brilliantly luminous than the photosphere. It is quite probable, indeed, that this central core (which contains certainly more than nine-tenths of the whole mass of the sun) is purely gaseous, and it is of course true that, at a given temperature and pressure, a gaseous mass has a lower radiating power, and is
217
THE SUN
less luminous, than a mass of clouds, such as those which constitute the photosphere. But, on the other hand, both compression and increase of temperature rapidly raise the radiating power of a gas; and it is highly probable that, at no very considerable depth, the growing pressure and heat may more than equal- ize matters, and render the central nucleus as in- tensely bright as the surface of the sun itself.
" At the upper surface of the photosphere, however, and all through it, indeed, the uncondensed gases are dark as compared with the droplets and crystals which make up the photospheric clouds. Here the pressure and temperature are lowered, so that the vapors give out no longer a continuous but a bright- line spectrum, whenever we get a chance to see them, against a non-luminous background; and, when the in tenser light from the liquid and solid particles of the photosphere shines through these vapors, they rob it or the corresponding rays, and produce for us the familiar dark-lined spectrum of ordinary sunlight.
" Although it 'may not be possible, in the present state of science, to demonstrate that the principal por- tion of the solar mass is gaseous, this much can at least be said — that a globe of incandescent gas, under condi- tions such as have been intimated, would necessarily present just such phenomena as the sun exhibits.
"On the outer surface, exposed to the cold of space, the rapid radiation would certainly produce the con- densation and precipitation into luminous clouds of such vapors as had a boiling-point higher than that of
218
WHAT IS THE SUN?
the cooling surface. These clouds would float in an atmosphere saturated with the vapors from which they were formed, and also containing such other va- pors as were not condensed, and thus the peculiarities of the solar spectrum would result. On the other hand, the permanent gases, like hydrogen — those not subject to condensation into the liquid form under the solar conditions — would rise to higher elevations than the others, and form above the photosphere just such a chromosphere as we observe. Whether, from the mere assumption of such a constitution for the sun, one could work out, a priori, the phenomena of sun-spots and prominences, is indeed doubtful; but thus far nothing in any of them has been observed which appears to be inconsistent with this view of the subject — nothing, we say, unless it should turn out, as was once maintained, that the solar surface possesses, so to speak, ' geographical ' characteristics, evinced by the disposition to break out into sun-spots at certain fixed points — as if at those points there were volca- noes or something of the sort. Of course, the fact that the spots are distributed mainly in two belts parallel to the solar equator, involves no difficulty, for it is easy to conceive how, in more than one way, the sun's rotation might lead to such a result : but peculiarities permanently attaching to individual points on the solar surface necessarily imply rigid connections, such as are inconsistent with the theory of a gaseous or even of a fluid nucleus. But while, as has been already pointed out, there is a marked tendency in
219
THE SUN
spots to recur at or near the same points during sev- eral solar revolutions, there is no evidence which es- tablishes the existence of fixed spot-centers; and the idea is to he regarded merely as a relic of the old Herschellian theory of a solid sun. Still it is difficult to test the notion conclusively even by means of such extended observations 'as those of Carrington or Spoerer, or the auroral periods of Veeder, since the time of rotation of the solid nucleus, if it exists at all, is unknown, and this makes the discussion difficult and unsatisfactory.
"With reference to the constitution of the photo- sphere there is a general agreement among astrono- mers. A few, perhaps, still hold, as has been men- tioned, to the idea that the visible surface is a liquid sheet, while some believe that it is purely gaseous; but the whole appearance of things, the details of the granulation, the phenomena of spots and faculae, the mobility and variability of the floccules, all better accord with the theory adopted in these pages, which is a necessary consequence of the hypothesis that the sun is principally gaseous. It seems almost impos- sible to doubt that the photosphere is a shell of clouds. As to the precise constitution of this shell, however, the form and magnitude of the component cloudlets, the chemical elements involved, and the temperature and pressure, there is room for a good deal of uncer- tainty and difference of opinion. The more common view, apparently — the one, certainly, which the writer has hitherto held — is, that the clouds are
220
WHAT IS THE SUN?
formed mainly by the condensation of the substances which are most conspicuous in the solar spectrum, such as iron and the other metals. As to the form of the clouds, also, it has usually been assumed that, as a consequence of the ascending currents by which they are formed, they are columnar, their height being much greater than their other dimensions.
" Professor Hastings has proposed a somewhat dif- ferent theory, which avoids some of the difficulties of the received doctrine, though not without encounter- ing others which seem just as formidable.
"One main peculiarity is the assumption that the photospheric ' clouds ' are formed by the precipitation of either carbon, silicon, or boron (the three members of the carbon group), to the exclusion of other sub- stances which are less refractory (have lower boiling- points), and therefore escape precipitation.
" His idea that the stratum which produces the gen- eral absorption at the limb of the sun is a veil of 'smoke' — i. e., of the same minute particles which constitute the photosphere, but cooled to relative darkness — has been already alluded to in a preceding chapter. So far as we know, it is novel and valuable, clearing up a good many embarrassing difficulties. It is so obvious, on reflection, that something of the sort must accompany the photosphere, that it is surprising that the idea had not been thought of before. Of course, the particles formed by condensation must, many of them at least, be carried by the ascending currents high above the point of their formation, and
221
THE SUN
cooled so much as to become relatively dark in com- parison with the more vivid incandescence of the re- gions below, just as the ascending particles of carbon, unconsumed and cooled, constitute the smoke of a fire.
"The idea that carbon may be the main constitu- ent of the photosphere is by no means new : it was first seriously advanced, we believe, by Johnstone Stoney, of Dublin, as early as 1867, mainly on physico-chem- ical grounds, and is enthusiastically advocated by Sir Robert Ball in his recent ' Story of the Sun. '
"As regards the 'reversing stratum' very little need be added. Mr. Lockyer indeed denies its existence — that is, in the sense that there is a thin stratum, close above the surface of the photosphere, in which most of the dark lines of the solar spectrum originate. He maintains, on the contrary, in accordance with his 'dissociation theory,' that certain of the lines, due to substances the most nearly elementary, and having their molecules in the highest stage of dissociation, originate only deep down in the solar atmosphere where the heat is most intense; others, due to vapors with molecules somewhat less simple, have their birth a little higher; and others yet, due to molecules the most complex, are produced only in the most elevated regions of the solar atmosphere; each elevation thus being responsible for its own special family of spec- trum lines.
" if, however, we reject this theory as 'not proven, ' we get results not very different.
"The vapors of the photosphere and chromosphere 222
WHAT IS THE SUN?
are not to be thought of as entirely separate and dis- tinct. All the gases are found together in the inter- stices between the cloud-granules of the photosphere — the unknown substance which produces the green line in the spectrum of the corona, the hydrogen, the calcium, and helium which characterize the chromo- sphere, and the metallic vapors which give the re- versing layer its peculiar properties — these all exist together in the lower depths, unless, indeed, it may possibly be the case that at the greater elevations some compound bodies are formed which can not exist in the fiercer fires below. So far as we can distinguish between these different portions, we may define the photosphere as the shell within which precipitation is taking place; the reversing layer, as that lowest re- gion of the solar atmosphere which contains sensibly all the gases indicated by the spectroscope ; the chro- mosphere, as the region of hydrogen, calcium, and helium; and the corona, as that upper domain of the solar atmosphere which becomes observable only dur- ing solar eclipses. But the coronal gas itself is most conspicuous and abundant right in the photosphere and reversing layer, and the same is true of the hydro- gen of the prominences.
"It is well, also, to bear in mind that, if any sub- stances decomposable by heat exist upon the sun at all, we must expect to find them in the higher and cooler regions of the solar atmosphere. In and near the photosphere, or underneath it, matter must be in in its most elemental state.
223
THE SUN
"As to the mechanism of the chromosphere and prominences, if we may use the expression, much cer- tainly remains to be learned. In many cases, indeed, perhaps in most, the forms and behavior of the protu- berances are satisfactorily enough accounted for by supposing that the heated hydrogen and its associate vapors is simply forced up into cooler regions by pres- sure from below — a pressure which must result from the downward movement of the great mass of pre- cipitated matter which forms the photosphere. But evidently this is not the whole story. We must have recourse to ideas of a different order to account for the somewhat rare, but still really numerous and well- authenticated instances when the summits of prom- inences have been seen to rise in a few minutes to ele- vations of two or three hundred thousand miles, the upward motion being almost visible to the eye at the rate of a hundred miles a second or more.
"Very perplexing, also, is the indubitable fact that clouds of this prominence-matter sometimes gather and form without any apparent connection with the chromosphere below, apparently just as clouds form in our own atmosphere, by the condensation of vapor before invisible. On the whole, it looks very much as if we must regard the prominences as differing from the surrounding medium mainly, if not wholly, in their luminosity — as simply superheated portions of an immense atmosphere.
"But, then, we immediately encounter the difficul- ties so ably urged by Lane, Lockyer, and others, that
224
WHAT IS THE SUN?
the existence of hydrogen of any appreciable density, at the elevation of even a hundred thousand miles, implies a density and pressure at the surface of the photosphere so high as to be entirely inconsistent with the spectroscopic phenomena there manifested — un- less, indeed, under solar conditions, the action of gravity upon the gases of the solar atmosphere is modified by some repulsive force. That such a force is at least conceivable, is obvious from the behavior of the tails of comets; and many features in the cor- ona point in the same direction. Of its nature and origin we can not, however, assert anything as yet.
"Even more difficult than the problem of the chromosphere is that of the corona. While it is some- thing to know that the phenomenon is mainly solar, and that, therefore, it must rank in magnitude and importance with the most magnificent of natural ob- jects, we have yet to find a satisfactory explanation of many of its most obvious features. It is certainly very complex — matter meteoric and matter truly solar; orbital motion, solar attraction, atmospheric resistance, and actions thermal, electrical, and mag- netic, are probably all combined."
HALM'S VIEWS
Since the time when Young wrote, Halm has con- tributed the following theory, designed particularly to explain the periodicity of sun-spots.1 Halm calls
1 Annals Royal Observatory Edinburg, vol. i, pp. 74-151, 1902.
225
THE SUN
attention to the function of the so-called solar enve- lope, and refers to the views of Langley, Pickering, and others, that it prevents the escape of half of the solar energy. He then considers the effect of changes in its powers of restraint. He accepts Helmholtz's hypothesis that the source of the solar energy lies in the contraction of the sun, and thinks, in contradic- tion to See's views, that the sun is already gradually cooling. He then refers to Hastings' paper (cited by Young) on the nature of the solar envelope, and says : " Indeed, it seems obvious that these particles which, while ascending from the interior to the surface, are precipitated so as to form the luminous clouds of the photosphere must (quoting Hastings) 'rapidly cool on account of their great radiating power, and form a fog or smoke which settles slowly through the spaces between the granules ' and that ' it is this smoke which produces the general absorption at the limb. ' ' ' Then, to emphasize the importance of heat conservation by the solar envelope, Halm refers to Langley's early view (which apparently Halm has not noticed that Langley afterwards retracted) to the effect that the earth's temperature would fall to —200° C. if it had no atmosphere.1
He suggests that if gravitation should be temporar- ily too little to supply heat energy by contraction to balance lost energy of radiation, then the layer of
1 See "Report of the Mt, Whitney Expedition" p. 123, and "The
Temperature of the Moon," Memoirs National Academy, vol. iv, pt.2, p. 193.
226
WHAT IS THE SUN?
maximum incandescence would cool, and a lower layer would become the new layer of maximum in- candescence. The absorbing layer thereby increases in thickness, so that the new layer of maximum incan- descence dissipates less energy than the first. Thus, at length a layer is reached which dissipates energy of radiation as fast as gravitation supplies energy of heat. But, when this state occurs, the outer layers will still go on cooling, since they receive less radiation from within than formerly. Consequently they continue to grow more opaque, and the amount of energy of radi- ation dissipated to space thereby becomes less than the amount of heat energy supplied by contraction.
Halm continues: "It thus comes to pass that, while the function of the absorbing envelope is >that of reducing as much as possible the waste of energy from the photospheric layers beneath, it is, by the very nature of the process, compelled to overdo its work, and to finally preserve too much energy within the star. The outbreak of eruptions and the forma- tion of spots are the consequence of an unstable equi- librium in the photospheric layers, and take place whenever the supply of heat from the interior is so supplemented by the continuous reflection of heat from the overlying atmosphere that the photospheric layers receive more heat than is required for the main- tenance of their thermal equilibrium. .
"The function of eruptions, consisting as they do in the ejection of overheated photospheric matter, is to produce a general heating and clearing up of the
227
THE SUN
cooled absorbing layers of the solar envelope. The action of the spots consists in drawing the cooled por- tions of this atmosphere into the hotter regions of the photosphere. "
Halm then goes on with mathematical work aimed to show that the consequences of these principles lead to a periodicity of sun-spot phenomena similar, even in its details, to that actually observed, but this part of the paper does not seem to be so soundly based as to add much to the merit of his views.
There is a strong objection to Halm's mathematical analysis, which applies, also, to the several computa- tions made by Vogel, Pickering, and others to deter- mine the effectiveness of the so-called "solar enve- lope," which these astronomers regard as a layer which restrains the emission of the sun. For they treat only the losses suffered by the direct beam through scattering in this envelope, without taking account of the gains which the beam acquires from rays scattered into it by the same envelope. Their numerical results are hence of no application to the sun; for the light proceeding in a single direction from any point in the "envelope" is derived from almost a full hemisphere. Their formulae are appli- cable only to a case like that of the earth's atmos- phere, where the entering rays are practically all parallel.
SCHMIDT'S HYPOTHESIS
It was in 1891 that Schmidt published his theory of a gaseous photosphere, and he explained the appar-
228
WHAT IS THE SUN?
cntly sharp outline of the sun in a very ingenious and interesting way. It is well known that the sun and other objects are seen, after they get below the real horizon of the earth, by the refraction of the air, which curves the rays of light. The amount of curva- ture depends on the rate of change of optical density of the atmosphere from its outer limits to the surface of the earth. At sea-level the difference caused by refraction between the apparent and real positions of heavenly bodies is about one-half a degree of arc. Suppose the earth were to grow larger, but with the same atmospheric densities prevailing. There would be a certain limiting diameter, about seven times that of the earth, where the curvature of the rays would be just sufficient to cause them to bend entirely around the earth in passing from the top to the bottom of the atmosphere, so that if there were no loss of light on the circuit a man as tall as the atmosphere is thick might be imagined to stand on his head at the equa- tor, arid looking directly in front of him see his own heels all the way around the world. If the earth were supposed still larger then all the rays leaving its sur- face tangentially would be incurved, and reach the surface again at some other point, without ever suc- ceeding in escaping to space.
Schmidt conceived of the sun as a wholly gaseous body, above the limiting size just discussed. Accord- ingly, looking from the earth, there would be a cer- tain diameter at which the line of sight would curve around in an infinitely long spiral of practically con-
229
THE SUN
slant diameter. A line of sight outside of this would pass nearly straight through the outer layers of gas, and emerge on the opposite side of the sun in space. A line of sight to a smaller circumference would pass along a diminishing spiral inside the sun till it almost reached the lesser sphere, to which it would finally be tangent ; and there it would go around and around in an infinite spiral course of practically constant diam- eter. Hence, all lines of sight inside a certain limiting circumference would give brilliant effects because they would have an infinitely long path of incandes- cent gas of great density to take light from ; while all lines of sight outside this limiting circumference, hav- ing only a limited thickness or rarified gas to take light from, would give by comparison only negligibly faint effects.
According to this view the solar phenomena, sun- spots, for example, need not be regarded as superficial, but may lie at any point between the outer limiting sphere and the inner sphere to which the diminishing spiral of refraction of the line of sight at length be- comes tangent. If this is so, a sun-spot which we see near the limb may really be somewhere on the oppo- site side of the sun from the earth. This hypothesis has curious consequences if we consider the apparent rotation of the sun as measured by observing sun- spots. For the supposed inner sphere, on which the spot by hypothesis really lies, must go at a different rate of rotation from the apparent rate of the sun. Suppose the sun's equator in the plane of the ecliptic,
230
WHAT IS THE SUN?
and that a certain equatorial spot actually lay at the limb of the inner sphere, but appeared at the limb of the boundary sphere. After a synodic day's ro- tation, the light pursuing the same actual path within the sun as before would come out, it is true, at a point just as far advanced in angle on the boundary sphere as the point it started from was on the inner sphere; but coming out nearly tangent to the outer sphere, as before, it would not be directed towards the earth at all. The path of light directed towards the earth through a sun-spot, advanced apparently one synodic day's march on the boundary sphere, would pursue an entirely differently shaped spiral within the sun, and would cut our hypothetical sun-spot sphere on its equator, to be sure, but not, at the same angular de- parture from the first position as would be indicated by the appearances. But when we reach the center of the sun's disk the line of sight is straight. Hence, the total period of rotation of the supposed inner sun- spot sphere must equal that of the apparent rotation outside ; for every time the apparent sun-spot reaches the center of the disk the real one is directly behind it. Accordingly, the motion of the supposed inner sun- spot sphere must be non-uniform, which seems ab- surd. The sun-spot must, therefore, be really, as well as apparently, superficial. An interesting result of Schmidt's hypothesis appears, also, if we consider spectroscopic line-of-sight determinations of solar rotation. For the motion in the line of sight depends on how far down in the sun we consider the light as 17 231
THE SUN
arising, so that it would seem that all spectral lines should be widened, when viewed near the limb, unless the material which gives rise to them is situated close to the apparent level of the limb.
Schmidt's views have obtained considerable ac- ceptance, but not from observers of solar phenomena. The late Professor Keeler said:1 " According to this theory, the sharpness of the sun's limb and the enor- mous change of brightness at that place are not caused by corresponding abrupt changes in the con- stitution, density, or light-radiating power of the solar matter, but are the result of refraction in a non- homogeneous medium. ... In other words, the photosphere is an optical and not a material sur- face. . . . Various assumptions as to the mass, temperature, etc., are here necessary, which it is gen- erally impossible to verify, but Dr. Knopf has shown . . . that the conditions in the case of the sun are well within the bounds of probability. . . . But, however difficult it may be for present theories to account for the tenuity of the solar atmosphere, immediately above the photosphere, and however readily the same fact may be accounted for by the theory of Schmidt, it is certain that the observer who has studied the structure of the sun's surface, and particularly the aspect of the spots and other markings as they approach the limb, must feel con- vinced that these forms actually occur at practically
1 Asirophysical Journal, vol. i, p. 178, 1895. 232
WHAT IS THE SUN?
the same level, that is, that the photosphere is an actual and not an optical surface. "
JULIUS'S VIEWS
Professor W. H. Julius of Utrecht has proposed a group of solar theories composed of ingenious appli- cations of the principles of anomalous dispersion. It has been abun- dantly shown by tD03 laboratory ex- periments that
\
the dispersion of light by the L00° vapors of metals 0999 is subject to dis- continuities in the regions of 7 spectrum im- 0.9% mediately adja- cent to their lines of strong emis- sion and absorp- tion. Fig. 57 shows the anomalous two-branched dis- persion curve of sodium vapor in the neighborhood of the D lines, according to researches of R. W. Wood. For comparison, the normal dispersion of rock-salt in the same region is also given. The enormous vari- ations of dispersion of the light on the edges of the D lines by sodium vapor would cause the production of dark spectral lines under certain circumstances, not
233
FIG. 57. — NORMAL AND ANOMALOUS DISPERSION.
THE SUN
by true absorption, but by anomalous dispersion. Julius has applied this to the explanation of many of the solar phenomena, and the reader interested should consult his numerous papers and also the critical articles of Hartmann, Anderson, and others. See The Astrophysical Journal, Astronomische Nachrichten, et cetera.
We may briefly state two or three of Julius's ex- planations here, and first concerning the chromo- sphere and prominences. These objects have bright line spectra, and appear to protrude beyond the limb of the sun. Eruptive prominences often appear to shoot out as rapidly as 100 miles a second! But to Julius they are not seen by their own brilliance out- side the sun's limb, nor do they rise with such veloci- ties at all. The line of sight to the apparent summit of a prominence is really, he thinks, a greatly curved line by virtue of the anomalous dispersion caused by the non-homogeneous density of a mass of non-lum- inous gas existing there; and the true source of the principal light is in the photosphere. A slight re- arrangement of the density alters greatly the path of the rays, and causes the impression of displacement of the prominence at enormous speeds. Adjacent wave lengths of the photospheric light do not reach the observer, because not anomalously refracted. The wave lengths of prominence spectra, if unaffected by other causes, would generally be slightly greater than the wave lengths of the true absorption lines of the gases concerned, because the density must, on
234:
WHAT IS THE SUN?
the whole, diminish outside the photosphere. But Julius regards irregular density gradients in the oppo- site direction as of common occurrence, so that short wave lengths will frequently occur. The displace- ments of wave lengths themselves will, he thinks, be almost imperceptible. Whatever may be our opinion of Julius's explanation of the high prominences, we must, I think, admit a considerable probability that anomal- ous dispersion might produce many of the phenomena of the chromosphere. But, on the other hand, if the chromospheric gases are self-luminous, the anomalous dispersion effects may be almost entirely masked.
Fraunhofer lines Julius regards as " absorption lines enveloped in dispersion bands/' the latter caused by a honeycomb of irregular density gradients in the photosphere, and showing themselves chiefly as the " wings" which occur with many lines. Rever- sals of chromospheric lines he regards as evidence of local condensations of gas, in which density gradients in both directions occur, thus bringing both longer and shorter wave-length dispersion bands to the eye.
Even sun-spots he attributes to refraction, but not anomalous refraction, at least as regards their major phenomena. He imagines local strong condensations or rarefactions in the photosphere, and shows how these might produce regions of diminished radiation, on account of the re-distribution of rays, and the return of some to the sun.1
1 See further "Regular Consequences of Irregular Refraction in the Sun," by W. H. Julius, Proc. Roy. Acad. of Amsterdam, Meet-
236
THE SUN
Astronomers generally admit that in the sun there may be conditions which favor the production of phenomena of anomalous dispersion, especially in the chromosphere. With few exceptions, however, they believe anomalous effects negligible, and the observed facts to be more simply and satisfactorily accounted for on the basis of ordinary views of selective emis- sion and absorption, such as have been given in pre- ceding chapters. The test between the two methods of explanation often involves the precise measure- ment of wave-lengths, and such criteria have not thus far been applied with such rigor as to exclude entirely the explanations advanced by Julius. It is not im- possible that writers on solar phenomena ten years hence will devote much space to the discussion of anomalous dispersion.
THE AUTHOR'S VIEWS
We must leave the reader to supplement by his reading of the original papers these inadequate sum- maries of the views of various investigators, and we will pass on to the solar theory which seems to the writer most probable. In its most general aspect this is similar to the views stated by Secchi in 1877 for Newcomb's " Popular Astronomy." Also an important paper by Schuster entitled "Radiation
ing of Sept. 25, 1909; "On the Origin of the Chromospheric Light," Mooting of same Academy, Nov. 27, 1909. "Anomalous Refraction Phenomena Investigated with the Speotroholiograph," by W. H. Julius, Astropkysical Journal, Dec., 1<M)S, cl cctmi.
236
WHAT IS THE SUN?
Through a Foggy Atmosphere" 1 has some things in common with it. Still more in touch is the paper of Schwartzchild already mentioned.2
It will be assumed: A. The sun, excepting perhaps in sun-spots, is wholly gaseous or vaporous. Except in sun-spots the photosphere is too hot to contain solids or liquids. B. The density of the gases rapidly diminishes, and their temperature rapidly falls from within outwards across the apparent boundary of the sun.
The view that the sun's photosphere is too hot to contain other than gaseous constituents has been strongly combated by J. F. Hermann Schulz,3 who even argues that the sun is mainly liquid. He sets the average temperature of the photosphere at 5,400° C. (5,673° Abs.) Although admitting that the late H. Moissan placed the temperature of his electric furnace at 3,500° C., and stated that all known ele- ments volatilize at that temperature, Schulz argues that the temperature of the electric furnace is to be set higher, even probably as high as the sun's temper- ature, and the volatilization is not to be regarded as complete in the furnace. His argument is that the enormous energy of the electric current used (see table below) had no adequate escape by conduction or radiation, and must have raised the temperature
1 Astrophysical Journal, vol. xxi, pp. 1-22, 1905. 2"Ueber das Gleichgewicht der Sonnenatmosphare," Gottingen Nachr., Math-phys. Kl. 1906, pp. 1-13.
3 Astrophysical Journal, vol. xxix, pp. 33-39, 1909.
237
THE SUN
of the furnace till checked by the melting and evapo- ration of the limestone of which it was constructed.
He continues: " Now Moissan has shown that, even at the enormous temperature attained in his electric furnace, we have not yet reached the point at which all terrestrial elements are truly boiling. In this re- spect the following table is very instructive, which Moissan gave in Comptes Rendus of February 19, 1906 (142, 430)."
TABLE XV. — Moissan's experiments on the vaporization of metals of the iron family
Metal |
Weights Grams |
Time Minutes |
Amperes |
Volts |
Metal Distilled Grams |
Nickel |
150 |
5 |
500 |
110 |
56 |
200 |
9 |
500 |
100 |
200 |
|
Iron |
150 825 800 |
5 10 20 |
500 1000 1000 |
110 55 110 |
14 150 400 |
M anganese |
150 |
3 |
500 |
110 |
38 |
150 |
5 |
500 |
110 |
80 |
|
Chromium |
150 |
5 |
500 |
110 |
38 |
Molybdenum |
150 150 |
10 20 |
700 700 |
110 110 |
0 56 |
Tungsten |
150 |
20 |
800 |
110 |
25 |
Uranium |
150 150 200 |
5 5 9 |
500 700 900 |
110 110 110 |
0 15 200 |
" Moissan further adds the following remarks: 'Molybdenum. The 150 grams were not fused by a current of 500 amperes and 110 volts. After applying
238
WHAT IS THE SUN?
700 amperes and 110 volts for seven minutes, the metal was fused but nothing evaporated. After twenty minutes 56 grams were distilled. Tungsten. After applying 500 amperes and 110 volts for five minutes, the metal was not yet fused. After apply- ing 800 amperes and 110 volts for twenty minutes, boiling commenced, but only 25 grams distilled. '
" Another highly interesting paper of Moissan is: 'Sur la distillation des corps simples. ' 1 Here we find the following statement:
" ' Gold commences to evaporate in vacuo at 1,070°. It boils in vacuo at 1,800°, and should boil at 760 mm. pressure at 2,530°, ' thus showing how much depends upon the pressure under which boiling takes place. Now all Moissan 's experiments, tabulated above, are made at ordinary atmospheric pressure, and we are entirely at loss to say how much the evaporation of the various metals would have been retarded under increased pressure, such as we might expect at the very base of the solar atmosphere, close to the liquid nucleus.
"Moissan tried, also, the metalloid titanium in his electric furnace. Five hundred grams were heated by a current of 500 amperes and 110 volts; after four minutes vapor appeared, but after five minutes the stuff was fused only on the surface, and carbide of titanium had formed. Then 300 grams were treated with 1,000 amperefe and 55 volts for seven minutes; 110 grams were distilled; the stuff itself, however,
1 Annales de chemie et de physique, (8) 8, 145-181, 1900. 239
THE SUN
had been only viscous, the surface had not become horizontal. In his book 'Der electrische Of en/ p. 238, he says that even with a current of 2,200 am- peres and 60 volts, the stuff in the crucible is not com- pletely fused."
The behavior of titanium is not unparalleled. Some substances go over from solids to gases without melt- ing at all. In the second experiment nearly half of the material was distilled, although the melting was not complete. No substances are cited which failed to become largely gaseous under a few minutes heat- ing at atmospheric pressure with the electric oven. True, an increase of pressure to five or ten atmos- pheres, which may prevail in layers we can see in the sun, would certainly have hindered the evaporation. But if the electric oven is above 3,500° C., even 4,000° C., it is still far beneath the photospheric tempera- ture. For if the solar constant is 1.95 calories, as will be shown in Chapter VII, the photosphere cannot be at a lower temperature than 5,860° Absolute Centi- grade, and may be much higher if its intrinsic radiat- ing capacity is considerably less than that of the per- fect radiator. Indeed, it seems most probable that the photospheric temperature should be set not lower than 6,500° Absolute. At such a temperature, pre- vailing not minutes but milleniums, one can most easily believe all elements are entirely gaseous.
As for the sun being mainly liquid, as argued by Schulz, the sun's low specific gravity has led even those who prefer to believe in a cloudy photosphere
VI n
WHAT IS THE SUN?
to regard the interior as almost wholly gaseous. We return to our discussion.
It is required to explain : 1. Why the sun presents a sharp boundarj^. 2. Why the enormous radiation of the photosphere does not so far cool its surface as to precipitate clouds. 3. Why a more or less definite structure appears on the sun. 4. Why the spectrum of the sun is mainly continuous. 5. Why, towards the limb, there is a gradual decrease in brightness, and an alteration in spectral distribution. 6. Why the solar spectrum has dark lines.
Besides these principal requirements, there are a thousand details of fact not necessary here to re- hearse, which must not be hopelessly inconsistent with any satisfactory solar theory. Finally there are the great problems of the periodicity of sun-spots, faculse, et cetera, the variations of solar rotation with latitude, and the supply of the sun's energy.
(1 ) Why the Sun Presents a Sharp Boundary.
In Lord Rayleigh's celebrated mathematical in- vestigations of the light of the sky he has shown that, whether proceeding on the hypothesis of the elastic solid theory of light, or on the electromagnetic theory, the extinguishing effect on a beam of light of the molecules of a gas, or of a collection of particles which are small compared with the wave length of light,
may be expressed by the relation : k = ^ — ;
in which k is the coefficient of extinction, //, is the
241
THE SUN
index of refraction, and N is the number of particles, or molecules, per cubic centimeter. Schuster has proved the relation to be independent of theory if /A is approximately unity. This is true for all gases. He has applied this quantitative theory of extinction to the atmosphere.1 For N he uses Rutherford and Geiger's value, 2.72 x 1019 molecules per cubic cen- timeter. If h is the height of the homogeneous at- mosphere, that is, the height to which the atmosphere would extend if entirely at standard temperature and pressure, then e~kh is the fraction of light which would reach the observer if none were lost in any other way than by molecular scattering. From these data Schuster calculates the extinction above sea- level, and above 1,800 meters elevation, and compares
TABLE XVI. — Difference between observed and computed values of atmospheric transmission
Wave Length |
Washington Observed |
Computed -"Clear" |
Ml. Wilson Observed |
Computed -"Clear" |
||
Mean |
Clear |
Mean |
Clear |
|||
o's 0.6 0.7 0.8 1.0 |
0.55 0.70 0.76 0.84 0.87 0.90 |
0.72 0.84 0.87 0.90 0.94 0.96 |
—0.01 + 0.03 0.07 0.06 0.04 0.03 |
0.73 0.85 0.89 0.94 0.96 0.97 |
0.76 0.89 0.92 0.96 0.99 0.99 |
0.00 0.00 0.03 0.01 —0.01 0.00 |
the computed values with the transmission observed on days of mean and maximum transparency at Washington and Mount Wilson, respectively, by Smithsonian observers.
\ttlnrc, vol. Ixxxi, p. 91 242
1 <>()<).
WHAT IS THE RUN?
Schuster concludes that on a clear day on Mount Wilson molecular scattering practically accounts for the atmospheric extinction. Even at Washington he thinks the major part of the losses in the atmos- phere may be thus accounted for; although on the average day something must be attributed to re- flection and absorption of grosser dust particles.
Professor Natanson has treated the matter from the standpoint of the electron theory. He differs in some respects from Ray lei gh and Schuster, although deriving a practically similar formula for scattering, for he introduces not the number of molecules but the number of electrons per cubic centimeter. He also has compared theory with the observations of Smith- sonian observers at Washington and Mount Wilson, and finds an approximate agreement. He does not state the conclusion in so many words, but his results indicate that the extinction of light above Mount Wil- son on the best days may reasonably be accounted for by scattering of the gas itself without consideration of dust particles.
All this has apparently a very important bearing on our views of the sun. The temperature of the layers from which we get the most light, as already stated, seems to be certainly in excess of 6,000° Ab- solute Centigrade. There are no substances, so far as known, which can exist except as vapors in these con- ditions. Hence, it seems reasonable to suppose that the sun contains no solids or liquids, unless perhaps in sun-spots, and that its substance, as we see it, and
243
THE srx
within the layers we see, is altogether gaseous. But if this is so, how, it will be asked, can the sun present a sharp boundary?
According to the theory of Schmidt, which has been alluded to, this is caused by the effect of refraction. But if Rayleigh and Schuster and Natanson are right in attributing a substantial light scattering effect to gases, Schmidt's theory needs hardly to be invoked, nor, indeed, can it really be of much application. For if, as computed by Schuster, the quantity of gas in the vertical column of atmosphere above Mount Wil- son is sufficient to scatter from the direct beam of yellow sunlight six per cent of its light, a column con- taining seventy-five times as much will suffice to scat- ter ninety-nine per cent.
Several observers have found that the pressure in the reversing layer for iron is about five atmospheres. Assuming the average absolute temperature of the photosphere to be 6,500°, and that of the air 250°, the quantity of gas per cubic centimeter in the reversing layer would be about ^ as great as in air at atmos- pheric pressure. As the homogeneous atmosphere above Mount Wilson is less than ten miles high, seventy-five times the quantity of gas above Mount Wilson would be found probably within 4,500 miles of the top of the sun's reversing layer. This estimate assumes the line of sight radial within the sun, and regards five atmospheres as the average pressure. If, as Evershed maintains, the pressure of the reversing layer is only of the order of one atmosphere, still we
WHAT IS THE SUN?
must admit that the pressure increases rapidly with the depth, so that still the estimate seems to be ample.
Hence, it seems probable that gaseous scattering alone prevents us from seeing towards the center of the sun, when looking directly at the middle of the solar disk, to more than 5,000 miles below the re- versing layer.
At the limb of the sun, the direct line of sight to a position at the same distance radially below the re- versing layer would traverse fully 60,000 miles of gas. Accordingly, to obtain our column containing the requisite quantity of gas for practical extinction of yellow light, at the limb we should penetrate a layer which, measured along the radius, would be very much thinner than that required at the center of the disk. For, even to a radial depth of only 500 miles, the direct line of sight is almost 20,000 miles.
These considerations seem to point to a reasonable explanation of the sharp boundary of the sun. For at the edge of the disk, owing to the oblique line of sight, gaseous scattering will probably extinguish almost all yellow light starting from more than 500 miles below the chromosphere, while, an even less thickness suffices for blue or violet light. It is plain that an indistinctness of outline corresponding to a layer of this depth would not be readily recognized on the solar image, since it corresponds to only about one second of arc. Furthermore, the direct line of sight takes in not only the nearer, but the further solar hemisphere as well. A still thinner stratum
245
THE SUN
than 500 miles would, therefore, suffice to contribute all the light that can be contributed to the beam di- rectly along the line of sight. We therefore con- clude that within a small part of a second of arc below the reversing layer the sun would appear as a solid body, even though entirely gaseous.1
(2) Why Is there No Cloudy Photosphere?
But it is said by Young and many others that a cloudy photosphere must certainly exist as the result of the juxtaposition of the hot gases of the sun with the cold of space. Without falling back on the strong reply that the apparent temperature of the so-called photosphere exceeds 6,000° Absolute Centigrade, and that no known substances can exist except as vapors at that temperature, it may be asked whether the ab- sence of a cloud immediately above the smoke-stack of a locomotive in winter does not show that such a juxtaposition of hot gases and cold surroundings with- out forming a cloud is entirely possible. There is no cloud formed immediate^ above the smoke-stack because the steam there is superheated above the boiling point. It may be urged that a little time is, of course, required to form the cloud, and that, owing to the rapid motion of the steam, it is carried a little above the smoke-stack during this interval. But this is really admitting that while the steam remains superheated it will not form a cloud, so that all that is
1 For practical purposes of seeing, it is not the depth of the layer which scatters ninty-nine per cent., but a much less fraction that is in question.
246
WHAT IS THE SUN?
necessary to prevent a cloud is to supply heat to the steam as fast as heat escapes from it, and thus to keep it superheated.
Such a state of affairs seems to exist in the sun. Heating is communicated from the interior to the sur- face layers fast enough to maintain the latter above 6,000°, notwithstanding their radiation to space, and at this temperature no cloud forms. The convey- ance of heat from within is probably almost wholly by repeated radiation, rather than by vertical convection currents.1
(3) What, then, Is the Cause of the So-called ' ' Rice-grain Structure'' on the Sun, if there Are No Clouds f It is not to be supposed that the communication of heat from within outwards is perfectly uniform at all parts, for, as evidenced by the sun-spots, the prom- inences, and the corona, there are marked defects of homogeneity in the sun. Hence, it may readily be supposed that some regions of the gas are a little hot- ter than others, and that these differences of tempera- ture will give rise to differences of brightness. By the radiation laws, the increase of brightness is far more rapid than the corresponding increase of temperature. Professor J. Scheiner published, in 1895, a theory of the solar granulation which seems very reasonable; and which, if we consider the effects produced to be merely regions of local cooling without actual con- densations, would suit the theory of the altogether-
1 This is the view of Schwartzchild and also of See. 18 247
THE SUN
gaseous sun as well as it does the theory of the cloudy photosphere.
Professor Scheiner says (quoting from a translation in the Astrophysical Journal) : "According to the the- ory of Helmholtz, air waves are produced when two layers of air, differing in temperature (i. e. in density), glide past each other, just as waves are produced by the gliding of air over water. If the lower layer is nearly saturated with aqueous vapor, condensations will take place in the wave crests on account of the diminution of pressure. Under these circumstances the elevations or -wave crests appear as clouds, the depressions or troughs as transparent interspaces, and thus a more or less regular series of cirrus clouds is produced. If the impulses resulting in wave forma- tion act in two different directions the waves cross, and we have the cloud effect known as a mackerel sky. The great similarity in appearance between the solar photosphere and terrestrial cirrus has long been rec- ognized, and there is no doubt that the necessary con- ditions for the application of Helmholtz 's theory to the solar atmosphere — the existence of layers of dif- ferent temperature, the over-saturated state of con- densable gases (in the photosphere), and variously directed currents in the different layers — are found in the sun. I therefore regard the bright grains of the photosphere as wave crests, rendered visible by con- densation, or at least an increase of condensation, of two crossing series of waves. "
We may adopt Scheiner 's view in the present dis- 248
WHAT IS THE SUN?
cussion, only not admitting actual condensation. Hence, his bright grains would be our dark ones, be- cause the cooler regions would radiate least. The reader will see that this amendment to Schemer's in- terpretation is rendered at least plausible by the fact that spectroheliograms show bright and dark hydro- gen flocculi, and of course no such a thing as a con- densed hydrogen cloud can be thought of at solar tem- peratures.
(4) Why Is the Sun's Spectrum Mainly Continuous? Gases are noted for giving only line spectra, while
the solar spectrum is, on the contrary, chiefly a con- tinuous spectrum crossed by absorption lines. In reply to this objection it may be said that gases under pressure give more and more continuous spectrum along with the bright lines, even in layers of small thickness, like those operated on in the laboratory. (See Plate X VIII . ) Think, then, if layers many miles thick, and under pressures of at least several atmos- pheres, may not give a fully continuous spectrum.
(5) Why Does the Limb Fall Off in Brightness and Grow Redder f
As stated above, the light received from near the edge of the solar disk comes, on the whole, from more superficial layers than that received from the center of the disk; because at the edge we look obliquely, and hence by a longer path, into the sun, and the scat- tering of the molecules cuts off the view before the deeper layers seen at the center are reached. At the
249
THE SUN
edge, the layers which are emitting light to us, being more superficial, and hence cooler, will in consequence give less intense light than those at the center.
Referring to Tables 7 and 8, Chapter III, it is pos- sible to compute, either by Stefan's law or by Wien's law, the change in effective temperature required to account for the decrease of brightness towards the sun's limb. As shown in Table 8, the two methods of computation are in close accord. Extending the result somewhat, we have the following differences of temperature, assuming the central disk temperature 6,400° Absolute Centigrade. These may be compared with the corresponding differences of elevation of
the lowest observable layer, assuming a depth of — -
100
radius, or 7,000 kilometers, as the limit of visibility at the center of the disk.
Fraction of radius |
||||||
from center of |
||||||
sun's disk |
0.0 |
0.1 |
0.2 |
0.3 |
0.4 |
0.5 |
Decrease of temper- |
||||||
ature |
0° |
20° |
45° |
80° |
115° |
160° |
Increase of elevation |
||||||
of farthest visible |
||||||
layer |
Okm. |
66km. |
140km. |
315km. |
545km. |
930km. |
The small temperature gradient of the order of 1° C. per kilometer of change of level1 required for this line of explanation seems no greater than we should expect to exist in the sun's outer layers.
As scattering is greater for violet than for red rays, the violet rays will come, on the average, from more
1 Mean radiating level, not lowest visible level. 250
WHAT IS THE SUN?
superficial layers than the red, both at the center and edge. Accordingly, the diameter of the sun should be greater if measured in violet light than if measured in red, so far as this consideration goes. But the differ- ence of diameter due to this cause is probably too little to be measured. It is obscured by " boiling" of the sun's image, diffraction, and scattered light in the earth's atmosphere, any one of which alone probably produces a greater effect at the limbs than that we are considering. According to Planck's formula, the change of intensity of radiation accompanying change of temperature of the radiating source is greater pro- portionally for short wave lengths than for longer ones. Hence, it follows that the violet should be weaker with respect to the red at the limb than at the center of the sun. This is in accord with observation. Whether this effect would be augmented or dimin- ished in consequence of the fact that the effective radiating layer for violet radiation is nearer the sur- face than that for red at both center and edge, de- pends on the relative change of temperature due to this shifting of depth at the two regions. It seems impossible as yet to determine how this would be.
(6) Why Has the Solar Spectrum Dark Lines f
All the Fraunhofer lines would really be bright if seen against a dark background.1 They are dark only relatively to the brighter continuous spectrum. In
1 Different persons estimate their brightness as from one-fifth to one- tenth that of the continuous spectrum background.
251
THE SUN
these lines the selective absorption of radiation is very powerful, and cuts off all transmission within a short distance, so that, as. compared with the continuous spectrum, they are emitted very near the surface of the sun. This superficial layer in which they arise is cooler than that which lies behind, hence its emission is less intense, and hence the comparative darkness of the Fraunhofer lines. As between the center and the limb of the sun we should expect little change in the absolute brightness of the Fraunhofer lines, because, owing to the powerful selective absorption within them, they are very markedly superficial phenomena both at center and limb. Thus, but little change in the effective depth and temperature from which they are emitted occurs, no matter from what angle the surface of the sun is viewed. It is not so with the process of weakening by scattering, which requires great thickness of gas; and hence, as we have seen, the continuous spectrum is brighter at the center of the sun than at the limb. Consequently the contrast or " intensity" of Fraunhofer lines falls off towards the limb, because they change little, while the back- ground against which they are seen, falls off in bright- ness.
Why Are Not all Chemical Elements Impartially Represented by the Intensities of Their Solar Lines f
It is not to be inferred from what has been said under (6) that there is no thickness to the " reversing layer," or no change of its effective thickness from
WHAT IS THE SUN?
the center to the edge of the sun, but only that, relatively to the effective thickness of the layer which furnishes the continuous spectrum at the center of the sun's disk, the reversing layer for any one element is thin. Hence, we may dis- • tinguish between high level and low level spectrum lines. It would be expected a priori that elements (a) of high atomic weight, (6) of high vaporizing temperature would be found at lower levels, and (c) that of the spectrum lines of a single element the longer wave lengths would, so far as depending on the relations of temperature and emission, represent higher levels.1 It might perhaps be expected that the reversing layer for a heavy element could lie wholly, below that of a light one. For very low lying ele- ments it might conceivably occur, through scattering, that their entire spectra would disappear at the edge of the sun, although appearing at the center. In general, low lying elements would give weak solar spectra, because the temperature of the emission of their lines would more nearly approach the temperature of the emission of the continuous spectrum background.
Referring to Chapter III, the reader will recall the marked connection between atomic weight and in- tensity of solar spectra. On the whole, the elements^ of less atomic weight give the strongest solar spectra. The platinum group, of very high atomic weight, on the other hand, is only partly represented in the solar
1 The effect of scattering would tend in the other direction, how- ever.
253
THE SUN
spectrum. Rowland and Tatnall say,1 speaking of these elements: "The heavier lines have been exam- ined as to the probability of their occurrence in the solar spectrum, and investigation has confirmed the existence of rhodium and palladium in the sun. Ruth- enium is doubtful" (afterwards confirmed) "and it is most probable that there are no solar lines of appre- ciable intensity belonging to platinum or osmium in this region of the spectrum" (X3,000 to 4,000). "The most intense lines of the arc spectra of rhodium and palladium correspond to extremely weak solar lines." This failure of solar lines is not for lack of strong lines in the arc, for Rowland and Tatnall give many platinum •arc lines of intensities 5 to 15, to which there are cer- tainly no corresponding solar lines above intensity 00.
The comparison of intensities and atomic weights given in Chapter III has some glaring discrepancies. Carbon is found near lanthanum, although its atomic weight is but 12. It is now believed that solar lines attributed by Rowland to carbon are really due to carbon compounds of considerable molecular weight, notably to cyanogen. Glucinum and potassium fall in strange company. But they have only one or two lines each identified by Rowland, and these may lead us into error. Indeed, Kayser and Runge question the existence of potassium lines in the photospheric spectrum.
In the flash spectrum at eclipses we have another indication of differences of level. There again, as
1 Astrophysical Journal, vol. ii, p. 184, 1895, 254
WHAT IS THE SUN?
shown by Evershed, Lockyer, Jewell, Mitchell, and others from measurements of the lengths of flash spectrum arcs, the order of level agrees on the whole with that which we have just considered.1 Again, in Adams' work on the solar rotation, if we grant (as we must when we recall the relative rotational velocities observed through red hydrogen (Ha), calcium (\4,227), and iron lines) that lower levels correspond to slower velocities, then we find (CN2) and lan- thanum falling in at lower levels than iron and titan- ium, just as they appear to do from considerations of Rowland's intensities.
The absence of lines of helium, the halogens and other negative elements in the photospheric spectrum is probably due to the extinguishing effect which the metals appear to produce on the lines of such elements when the metallic and other gases are mixed. Thus, according to E. Wiedemarm, nitrogen and hydrogen lines first begin to appear in a vacuum tube showing mercury lines when the concentration of these gases is thirty per cent. Also, common salt in a flame shows the spectrum of sodium alone, not of chlorine.
What Causes the Differences of Character and Wave Length for Fraunhofer Lines between the Center and Edge of the Sun ?
The reader will recall that, after allowing for the rotation of the sun and for an apparent general rise
1 The extremely high level of calcium H and K lines is an anomaly not well understood.
255
THE SUN
of the brighter material toward the solar surface, Adams l confirmed Halm's and Buisson and Fabry's results that there is a general displacement toward the red of the centers of most solar lines as seen near the limbs and compared with the center. This displace- ment is inappreciable for the more prominent lines of hydrogen, calcium, sodium, and magnesium, and small for the other lines of these elements. Also for elements of high atomic weight the shifts are very small. Iron and nickel lines show larger shifts than those of titanium, vanadium, and scandium. En- hanced lines as a class show larger shifts than arc lines do. Lines strengthened at the limb show small shifts. The displacements are greater for long wave lengths than for short. The character of lines at the limb is also altered. Some strong lines of the elements hydrogen, sodium, calcium, silicon, magnesium, aluminum, iron, chromium, titanium, and manganese lose partially or wholly the winged appearance which they have at the center. Many lines of all kinds of elements are slightly widened. The enhanced lines and lines of elements of high atomic weight are gen- erally much weakened.
The weakening of high temperature and low lying element lines may be attributed to scattering. At the center of the sun we look straight down upon the lower reversing layers and get their rays under more favorable angles of scattering than at the limb, where they, in order to contribute to the line of sight, must
1 Contributions of the Mount Wilson Solar Observatory, No. V.\. 256
WHAT IS THE SUN?
scatter by one or more reflections through a right angle, or nearly so. Hence the continuous spectrum at the limb encroaches upon their lines, diluting them with stray light. Furthermore, what is at least equally important, the sun's continuous spectrum is weaker at the limb, for reasons already considered, and would give the lines less contrast, even without the effect we have just considered. This weakening of the continuous spectrum at the limb contributes- powerfully to reduce the visibility of the wings of lines, also, because the wings are seen against a back ground which, towards the limbs, approaches more and more their own strength of emission. The widen- ing of lines seems possibly a promiscuous Dopplei effect due to their being contributed to by different levels rotating at different velocities.
Adams explains the superior displacements of en- hanced lines by suggesting that at the center of the sun, the higher temperature gases are rising, the cooler ones falling, giving for the spectrum lines in general a rising effect; because most of the light comes from the brighter emitting matter which is rising. But for the lines which are high temperature, or enhanced, lines a maximum rate of rise (greater than that of average lines) is observed, because the descending cooler vapors do not emit or absorb en- hanced lines, so that for these lines there is a displace- ment of central spectra towards the violet which ap- pears as an increased displacement of edge spectra towards the red. Of course, at the limb these mo-
257
THE SUN
tions of rise and fall are at right angles to the line of sight, and, therefore, produce no Doppler effects.
Having thus cleared the ground of Doppler effects of rotation and rise, Adams attributes the remaining displacements to pressure depending on level. High level lines are not displaced because emitted under slight pressure both at the edge and center. Low level lines are not displaced because they arise only from thin strata at the very bottom of the layer which is visible to us, and which must be at nearly equal, though high pressures at both center and limbs. Scattering does not permit us to see much beyond the outer boundaries of such strata at center, and at the limb we see them only faintly, and after the rays have been one or more times reflected, hence such spectrum lines are weak at the limb. Lines of intermediate levels are under higher effective pres- sures at the limb than at the center, according to Halm's view, as adopted by Adams, because any line of sight drawn just inside the limb has a longer rela- tive path in the lower layers it cuts than a line of sight drawn near the center of the disk has in the corresponding layers. Hence, lower layers contrib- ute proportionately more to the spectra of inter- mediately lying elements at the limb than at the center.
The writer must confess that he feels a little hesi- tancy about adopting this last argument, because he thinks that it would be necessary to consider for these layers quite as much the rays scattered into the beam
258
WHAT IS THE SUN?
from all sides as to consider merely the line of sight. But until the proportions contributed to a beam by scattering from different distances, and at different angles, and the rate of change of density along the sun's radius, are better known than now, it seems idle to press this objection.
Julius has explained the displacements of lines towards the red as a simple consequence of anomalous dispersion. But Adams shows that the lines apt to be most powerfully affected by anomalous dispersion show no shifts at all, and that a comparison of all the known data as to the strength of anomalous disper- sion for the several lines with their observed shifts at the limb yields nothing to recommend this explana- tion of Julius.
Why Do the Prominences and the Chromosphere Give
Bright Line Spectra?
According to the line of explanation we are pursu- ing, the gases of these appendages of the sun are in a condition of extremely low pressure and density, and do not contain sufficiently many molecules contrib- uting radiation to the line of sight to emit a strong continuous spectrum. But for the spectrum lines of powerful selective emission, their radiation is suffi- ciently considerable to reveal their forms.
What of the Characteristic Forms and Occasional Immense Velocities of the Prominences f Although not shared by all, there has always been a
hesitancy among many of those who regard prom-
259
THE SUN
inences as real protruding masses of bright gas, and not, after Julius, as mirage effects, to trust their ob- servations, both direct and spectroscopic, that these gaseous masses are bodily projected at such rates as a hundred miles a second. It is hard to imagine on purely mechanical grounds how such velocities could arise. In spectroscopic determinations the motion observed in prominences is apparently tangential to the sun. W. A. Michelson of Russia has suggested that in this case we may really have moderate motion
across the line of sight,
as illustrated in the accom- FlG 58 panying Fig. 58. Let as be
a line of sight, and let a
mass of gas whose front is AA, rise to positions BB and CC. Then the source of light moves effectively from a to b to c, giving an apparently enormous motion in the line of sight, which is really a much smaller motion across the line of sight. Whatever may be thought of this explanation, it, of course, has reference only to apparent enormous tangential ve- locities.
As for apparent enormous radial motions, we all frequently see wisps of cirrus clouds stretch across the sky in a twinkling, as it were. This does not, of course, indicate motion of translation from one end of the wisp to the other, but rather the rise of a trough of cooling, which causes a precipitation of cloud almost simultaneously along its whole length. Young and others state that detached prominences
260
WHAT IS THE SUN?
sometimes form without preliminary attachment to the sun's chromosphere. Perhaps eruptive promi- nences are formed somewhat as cirrus clouds are, by the rise from the sun of some disturbance fitted to arouse emission almost simultaneously within large masses of previously non-luminous hydrogen and calcium gases, which lie where a prominence is about to appear. Perhaps an electrical excitation would be most reasonable. The apparent tremendous velocity of the outbreak could then be explained by reference to the accompanying dia- gram, Fig. 59. Let the line of sight be in the line of the arrow I, AB the photo- sphere, and ab the trough which suddenly arouses emission beginning at a and proceeding almost im- mediately to 6. The lower end, a, is lost in the glare of the photosphere, and the prominence appears to rise from c to b in the very brief time needed to extend to b the influence of the trough. If the line of sight had been in the direction of the arrow II the promi- nence would have appeared detached. The writer does not venture to recommend this suggestion very strongly.
Professor E. Pringsheim1 has suggested an explan- ation of these enormous observed prominence veloci- ties that appears very reasonable. He refers to ex- periments of J. Stark, who has found Doppler dis- placements of the order of magnitude which occur in
1 E. Pringsheim, " Physik der Sonne," Leipzig, 1910, pp. 225-228. 261
THE SUN
prominences, when observing the so-called " canal rays," which are curious electrical discharges ob- tained through rare gases by special contrivance. This indicates that the positively charged atoms, which give the light, may travel with velocities like those which appear in the prominences, when forced by ordinary differences of electrical potential in very rare gases. Pringsheim fchen draws attention to the fact that the comet of 1843 passed at perihelion within 3' or 4' of the photosphere (TJT the solar diameter) without being affected by the resistance of the material encountered. This proves the existence of a sufficient, indeed, of an extraordinary degree of vacuity there. It is not known that sufficient variations of electrical potential exist in the sun's neighborhood, but those which exist in the earth's atmosphere are abundantly sufficient to drive elec- trons with prominence-like velocities in vacuum, ac- cording to Pringsheim's computations. The exist- ence of similar potential gradients near the sun seems not improbable.
Julius's explanation of prominences through anom- alous dispersion we have already noted, but it requires us to admit the propagation of disturbances to the apparent tops of the prominences, and to believe that the gases at such enormous heights are dense enough to produce appreciable anomalous re- fraction.
262
WHAT IS THE SUN?
What Is the Corona?
Comets pass through the corona without sensible retardation. Hence, its matter, whether it be purely gaseous or partly meteoric dust, must be very rare. Since the corona gives the Fraunhofer spectrum in its outer part, it must contain reflected photospheric light. Its substance in its inner parts may well be hot enough, by virtue of proximity to the sun, to give light of incandescence. Its form suggests the auroral streamers, and inclines one to think that, as the au- roral light is of electrical luminescence, so may a part of the coronal light be. As the aurora gives bright lines in its spectrum, so, also, does the corona. The proportions of the mixture of these three varieties in coronal radiation is unknown, but, according to bolometric work, the mixture gives almost the same spectral distribution in the inner corona as photo- spheric radiation. This suggests that light of lumi- nescence and of reflection together predominate over incandescence.
On the other hand, the results obtained by repre- sentatives of the Lick Observatory at several eclipses have led Campbell, Perrine, and Lewis to express the opinion very definitely that the inner corona shines mainly by ordinary incandescence, due to the heating of its particles on account of the absorption by them of photospheric radiation. Neglecting the bolometric results, this conclusion would be perfectly reasonable and it may yet prove that there is some error in these 19 263
THE SUN
latter results which may explain the discrepancy. Yet the conditions of the bolometric work at Flint Island were so satisfactory, and the observations so concordant, that this seems rather improbable. Per- haps some new line of explanation may suit all parties.
The electrical explanation of the coronal form and brightness is receiving much attention. Professor Pringsheim devotes much space to it in his new work, "Physik der Sonne." In a recent article, however, Professor R. W. Wood1 has sought to explain both the polarization and light emission of the corona as the effects of the passage of the powerful sun-rays through comparatively cool metallic vapors, thereby exciting fluorescent light in them. By laboratory experiments with light so excited in vapors of sodium, potassium, and iodine, he finds the percentage of po- larization similar to that in the corona. He states that the spectrum of mixed vapors would be contin- uous, at least for low dispersion. The fluorescent spectrum is, in fact, made up of thousands of fine lines arranged in groups and bands, and gives no re- semblance to the bright line spectra of the same ele- ments. These lines lie so closely packed as probably to escape detection with low dispersion spectroscopes. Any color of fluorescence may occur, according to the kind of vapors mixed, and their proportions. Wood thinks it quite possible that the coronal green line is
1 Astrophysical Journal, vol. xxviii, p. 75, 1908.
WHAT IS THE SUN?
not a bright line of some unknown substance, but rather a yet unrecognized fluorescent line from some well-known element.
Schaberle has long maintained a mechanical erup- tion theory of the coronal form.1 He traces back the probable courses of the streamers, and locates them in centers of eruption on the sun's disk. His views agree well with the hypothesis that the coronal bright- ness is mainly of incandescence.
The cause of the change of form of the corona with the sun-spot cycle is unknown.
Importance of Temperature.
It will be noted that in the solar hypotheses we are recommending the temperature plays a most promi- nent part. First of all, the existence of a cloudy pho- tosphere is denied because the temperature of the photosphere is shown probably to reach 6,500°, for it is highly improbable that solids or liquids can exist in such conditions. Secondly, the presence of the so- called " granulations " is regarded as evidence of dif- ferences of temperature in the radiating gas — dif- ferences which would naturally be expected in an im- mense globe of gas giving off tremendous amounts of radiation from its surface, and known to present irreg- ularities of rotation and cyclonic motions in addition. Thirdly, the darkening towards the limb is regarded primarily as a temperature effect, secondarily due to
1 See Lick Observatory Contributions, vol. iv, 1893. 265
THE SUN
scattering. Owing to scattering, the effective radi- ating layer must necessarily be nearer the surface, and hence cooler, at the limb than at the center of the disk. We say it must be nearer the surface: For, travelling obliquely, a ray must become extinguished by scattering in the gas at the limb, before it reaches the same radial depth that it does if travelling radi- ally at the center. Fourthly, the darkening at the limb would naturally be greater for violet than for red rays, firstly, because with all incandescent bodies a fall of temperature causes more decrease of radiation for short rays than for long; and, secondly, because molecular scattering is greater for violet rays than for red, and hence at the sun's edge the effective radiating layer for the violet will be more near the surface than will that for the red. Fifthly, the Fraunhofer lines are regarded, not as dark, but as very bright, intrinsi- cally. They only appear dark because, owing to pow- erful selective absorption of the gases which give rise to them, they cut off completely the light from be- hind, and the observer sees only a relatively thin and superficial layer of the sun, when viewing it by the light of the Fraunhofer lines. The reversing layer is hence colder, and its radiation less intense than that of the continuous spectrum background which comes from deeper layers of the sun. Sixthly, the contrast of the Fraunhofer lines with the background of spec- trum decreases as our view approaches the edge of the sun's disk, because the Fraunhofer line region is so thin and superficial that its temperature is nearly the
266
WHAT IS THE SUN?
same at the edge as at the center; whereas for the continuous spectrum background, the effective radi- ating layer rapidly approaches the sun's surface as we look nearer the limb, and hence its radiation de- creases, owing to the fall of temperature.
What of Sun-spots ?
We now trace the importance of temperature in the explanation of sun-spot phenomena. In accordance with the Mount Wilson observations of Hale, Eller- man, and St. John, we may regard sun-spots as vor- tices, and, as indicated by the spectrum work of Ever- shed, we must conclude that in the Fraunhofer line region the motion along the spiral is from within out- ward. We may imagine that these vortices are sim- ilar in form to water-spouts seen at sea, with the trumpet-shaped part at the top, and the whirl carry- ing matter from below outward. In such circum- stances there would be a great cooling of the gases, owing to their rapid expansion as they approach the limb. This cooling (as appears from the discovery of lines due to the copious presence of calcium and mag- nesium hydrides, arid also of titanium oxide in sun- spot spectra) carries the temperature down to per- haps 3,500°, which is low enough for the formation of liquids, and perhaps some solids.
These dissimilar substances, by their friction (per- haps even by their very formation) we suppose may give rise to charges of electricity, which, being carried round rapidly in the stem of the vortex, produce the
267
THE SUN
effect of currents of electricity as shown by Rowland. Hence they give 'rise to the magnetic field, which Hale finds is a feature of a sun-spot. The top of the vortex, we assume, corresponds in level nearly with the upper Fraunhofer line region. There the cooled matter spreads out some distance in spirals which grow in radius so rapidly as to be almost radial to the umbra. As there is no longer further rise and expan- sion, at length the matter becomes warmed, by con- tact, to the temperature of the surroundings. The stem of the vortex is the umbra of a sun-spot, the spreading top is the penumbra.
The peculiarities of the sun-spot spectrum and the causes of these peculiarities have been dealt with at considerable length in Chapter V. We may summar- ize them as the peculiarities attending, (a) diminished temperature as compared with the photosphere, (6) the action of strong magnetic fields. The sun-spot spectrum has been shown by Hale and Adams to be the type of the spectrum of the red stars. Since we now know that, as regards the characteristics for which this comparison holds, the sun-spot spectrum results from the mere cooling of the photospheric material, this relation is very significant, and indi- cates distinctly one step in the process of stellar evo- lution. We shall recur to this in Chapter X.
We have noted particularly in Chapter V the re- markable behavior of the hydrogen lines in sun-spots. They are all weakened, and the shorter wave-length lines most weakened, as cohipared with the photo-
268
WHAT IS THE SUN?
spheric hydrogen spectrum. The lines of other ele- ments are generally strengthened in spots. This anomaly seems explainable as due to the high level of hydrogen. This gas is relatively unaffected in posi- tion, and in fact, as St. John observed, is sucked in- ward and downward rather than whirled outward and upward by the cyclonic motion in sun-spots. Hence, its temperature and (in consequence) its radi- ation is rather increased than lowered by the presence of spots. The continuous spectrum background against which we see the hydrogen lines is, however, weakened in spots, and thereby the contrast of the hydrogen lines is diminished. In other words, they are weakened. Owing to the lower temperature, the energy spectrum, that is, the continuous spectrum background, in sun-spots as at the sun's limb, is weaker in the violet as compared with the red than is the ordinary solar spectrum. Thus, in spots, the radi- ation in the violet hydrogen lines approaches more nearly the brightness of the spectrum background .than that in the red lines. Hence, the comparatively greater weakening of the shorter wave-length hydro- gen sun-spot lines follows.
In the center of the sun-spot vortex there is a ten- dency to form a vacuum. Into this partial void is sucked the superincumbent matter, which is the high- level hydrogen of the chromosphere and prominences. Hence occurs the inwardly directed radial motion of this gas shown by the Ha spectroheliograms at Mount Wilson. Between the Ha level gas, which is going in-
269
THE SUN
ward, and the Fraunhofer line gases, which are going outward, there must exist a quiescent region. Hence, the lower level hydrogen and the HI and H2 cal- cium spectroheliograms show little or no evidence of stream lines or other phenomena of stream motion. The failure of Adams to discover differences in the pressure of the reversing layer over sun-spots may be regarded as confirmatory of the superficial char- acter of the reversing layer, and of the absence of either elevation or depression in the general sun-spot level.
As for the cause of the formation of sun-spots, that is all conjecture. They are generally preceded by faculae and, according to Fox, by eruptive promi- nences. Perhaps the faculie, which on our tempera- ture hypothesis we regard merely as regions of su- perior temperature, may be formed first, owing to the presence above them of prominence or coronal mat- ter. Such formations above would impede radiation, and hence would cause the regions below to be over- heated. Being overheated, they would tend to ex- pand, and by expansion would cause the rise of ma- terial from below, owing to reduced pressure above. In this outflow a rotation would usually be set up, just as in the escape of water from a spout, and thus the sun-spot would be formed. Once formed its vortical motion would tend to continue, and would naturally remain for considerable time. Hale has noticed that the vortices of most sun-spots of the southern hemisphere go in one direction and those in
270
WHAT IS THE SUN?
the northern in the opposite. This indeed is what would be expected in consequence of the different rates of rotation of the sun at different latitudes. But it would also be expected that accidental local circumstances, irrespective of this general cause, might sometimes determine the rotation in opposite senses. This also is in line with observation.
As to the general cause of the periodical changes (1) of the form of the corona, (2) of the areas of faculaB and (3) of the sun-spot numbers, these also, are things as yet altogether uncertain. We have already noted Halm's theory of sun-spot periodicity as a conse- quence of internal conditions. Schuster and others have suggested exterior influences as the operative causes of the periodicity. For instance, the periodic returns of swarms of meteorites, and the periodic returns of certain planetary configurations have been mentioned.
As for the variable rates of solar rotation at dif- ferent latitudes and depths, these have been regarded by Wilsing, Sampson, Wilczynski, and Moulton as vestiges of some ancient actions in which the sun fig- ured with outside celestial bodies.
What Supplies the Solar Energy f
Lastly comes the greatest problem of all: What maintains the solar temperature despite the sun's enormous losses by radiation? These losses stagger expression in figures. At 90,000,000 miles (145,000,- 000 kilometers) the average radiation is about 'two
271.
THE SUN
calories per square centimeter per minute. Hence the total emission of the sun is about
22 2X4X y X (14,500,000,000,000)2 calories per minute!
Ordinary fires of coal are kept up by the combination of carbon with oxygen. Except in sun-spots no com- binations are going on in the sun. It is so hot there that most compounds would separate into their ele- ments, instead of elements uniting with the evolution of heat. If the sun had no continuous supply of heat, but, like a piece of metal lying on the blacksmith's anvil, had been cooling off, there would have been a marked decrease of the earth's temperature within historical times. Geologists show that the earth has not varied more than a few tens of degrees from the present temperatures for probably 50,000,000 years. Indeed, in that remote past the earth's temperature appears to have been a little higher than it is now. Assuming that the sun emitted its present quota of radiation during all that interval, the problem of its source of supply has been, at least until very recently, insoluble. Since the discovery of the breaking up of radio-active materials to produce elements of lower atomic weight with the evolution of heat, as for in- stance in the production of helium from radium, per- haps no such difficulty ought to be regarded as in- superable. It is objected that radium and uranium lines are not found in the solar spectrum. We have seen, however, that the lines of the elements grow more and more feeble as the atomic weight increases,
WHAT IS THE SUN?
and that this seems to be due to the fact that the heavy elements lie at low levels in the sun. Hence it is not surprising that uranium (238.5) and radium (226.4) should not show spectrum lines even if these elements are present in the sun. Dr. G. F. Becker, however, thinks radium and uranium to be elements which form only at temperatures much below those of the sun. At all events, it is more satisfactory, if possible, to account for the solar heat by known causes, rather than to invoke radio-activity of un- discovered materials.
There are certain circumstances of geology which may indicate a diminished radiation of the sun in an- cient times. Although palms used to flourish in the arctic zones, it does not appear that the tropics were then much hotter if any than now. As Manson in- sists, this uniformity of climate from the poles to the equator seems hard to reconcile with the present zonal distribution of temperature, if the sun were then as now the principal source of heat, and its effects then, as now, zonally distributed. On the other hand, there is accumulating evidence that glaciation has occurred more than once over great regions of the tropics, and most notably in the Permo-Carbonifer- ous period. In that remote period, far antedating the so-called " glacial" or Pleistocene period of com- paratively recent times, glaciation prevailed in Aus- tralia, Southern Africa, Hindustan, and perhaps in other tropical regions. It was no mere sporadic mountain-top affair, but probably a phenomenon of
873
THE SUN
more imposing extent than even the glaciation of the Pleistocene Period.1 As will be shown in the next chapter it seems very difficult to see how such a sub- tropical glaciation as this could have come about if at that time the sun's output was substantially as great as now. It does not help us to suppose that the poles of the earth were then shifted so that these countries were sub-arctic. The area involved is so vast that the glaciation would still extend further from the sup- posed pole than did that of the Pleistocene period. Besides, this would bring one pole in the vicinity of Mexico, and the Permian deposits of Texas do not justify the inference of a polar climate there.
It seems worth considering if the Permian and the still earlier tropical glaciations which geologists are
1 According to Chamberlin and Salisbury (" Geology," volume ii, pages 636, 634. Henry Holt & Co., 1906): "The known Permo- Carboniforous glaciation of Australia, India, and Africa is found in two zones, the one north and the other south of the equator. In neither zone have the limits of glaciation been accurately deter- mined, but in the former it is known to have extended from latitude 18° to about 35° and probably still further north, while in the latter it is known to have extended from latitude 21° to 35°. In an equatorial zone about 40° in width glaciation has not been dis- covered. The glaciation of these various countries has a range of about 130° in longitude. Glacial conditions must therefore have prevailed over an area, or at least about the borders of an area many times as large as that covered by ice in the northern hemisphere during the Pleistocene glacial period." Speaking of the Australian glaciation they say: "It is not to be understood that the phenom- ena here described are restricted to high altitudes; rather they are known chiefly at low levels, descending in some places nearly to the sea. The altitude of this region is not only low now, but it was probably low during the glaciation as shown by the relation of the glacial deposits to the marine beds."
2*3 I
WHAT IS THE SUN?
now recognizing and also the generally prevailing similarity of polar and equatorial climates in early epochs do not all point to one of the following hypotheses :
(A) Perhaps the sun in those early times was not so nearly exclusively as now the earth's source of heat, and the earth itself still retained so much heat that its life was practically independent of the sun except for light. In a later chapter we shall see that under other favoring conditions by no means all of our pres- ent light supply is necessary to promote maximum plant growth, and that the red end of the spectrum, which would suffer least reduction by a decrease in the solar temperature, is highly efficient for plant growth. Perhaps, then, the sun has been gradually growing in temperature and emission, and in the Per- mian times had not then become the practically ex- clusive source of heat to the earth's surface. We may, then, briefly consider if Permian glaciation was perhaps due, as Manson has suggested, to a very mod- erate elevation of land areas within a region of a still prevailing low-lying cloud mantle, with accompany- ing snowy precipitation. The great, and it seems to me insuperable, difficulty which this hypothesis en- counters is to explain in any reasonable way how the earth's temperature could be maintained for millions of years without depending so completely on the sun that the explanation of the uniformity of climates fails. Furthermore the aridity of climate indicated by the great Permian deposits of salt and gypsum
275
THE SUN
does not speak for the existence of a thick cloud man- tle. These difficulties will be further discussed in the next chapter.
(B) Perhaps the sun in very ancient times had not yet altogether condensed to a pronounced nucleus, but still existed as a nebula of very considerable size, so that the earth was illuminated and warmed from all directions, or (if no part of the nebula inclosed the earth) at least from nearly a hemisphere.1 This of course would promote uniformity of temperatures from the equator to the poles. If thus receiving ra- diation from a very large solid angle, the intensity of the radiation need have been only very slight indeed to maintain the earth's temperature. Such radiation might be furnished by a cloud of small particles (not gases) comprising the nebula. Even if they gave no considerable radiation of their own, they would re- flect that of the hotter solar nucleus. On either hy- pothesis (A) or (B) the radiation of the ancient sun, or solar nebula, to outside space may have been con- siderably smaller than the total of the present solar radiation.
If either of these views or a combination of both is acceptable, it relieves the problem of solar radiation of much of its difficulty. We may then suppose that 50,000,000 years ago the total emission of solar radi-
1 This suggestion was made by Chambcrlin about twelve years ago. In Plate XXVI, Fig. 1, is shown a spiral nebula on edge. The bright region at its center is seen to extend out of the plane of the spiral so as to fill a large sphere.
276
WHAT IS THE SUN?
ation was considerably less than the present emis- sion, and that it increased slowly for ages, reaching approximately the present output at about the Pleis- tocene period, which was perhaps not over 100,000 years ago.
Helmholtz in 1853 proposed a source of solar en- ergy supply which is everywhere recognized as cer- tainly very considerable. He pointed out that the shrinking together of the sun converts potential en- ergy of position into heat, just as the falling of a stone converts its potential energy of position finally into heat. Several authors have made computations of the quantity of energy which would be available from this source. Their results have generally been based on the assumption that the sun was originally a nebula filling a sphere whose diameter was the orbit of Neptune. It appears that the condensation of such a nebula having the mass of the sun would have furnished thus far about 25,000,000 times as much energy as the sun now loses each year. (This estimate is based on a "Solar Constant" of 2.0 calories per square centimeter per minute.)
According to Helmholtz 's view, a contraction of about 250 feet per year in the sun's diameter would suffice to sustain the present solar radiation. At this rate it would require about 10,000 years to reduce the apparent diameter of the sun by one second of arc, so that, so far as telescopic observation is con- cerned, the contraction theory is tenable, for a change of -T0- second in the solar diameter is unrecognizable.
277
THE SIX
From calculations of Newcomb the sun will require to have shrunk to one half its present size if it maintains its present rate of radiation for about 7,000,000 years longer. As shrinking cannot go on indefinitely, nor can the supply of heat from this cause have been in- finitely great, we must, from this point of view, re- gard the duration of life depending on the sun's rays as having had a beginning in the remote past, and as tending towards an end at some remote time in the future.
There has been much question in recent years whether Helmholtz's hypothesis of the sun's energy supply is adequate to account for the duration of life upon the earth revealed by the geological record. Joly has estimated from the volume and salt contents of the ocean, compared with the rates of discharge and salinity of the rivers, that the earth's geological age is about 80,000,000 years. G. F. Becker has re- cently revised the calculations, with allowance for a more rapid discharge of salt in earlier periods, and finds about 50,000,000 years. On the other hand, many geologists think the thickness of the earth's deposited strata requires us to admit more than 100,- 000,000 years. The duration of the sun's radiation at present rate of output apparently cannot have been supplied by shrinking alone for more than 25,000,000 years. But, as has been said, it seems plausible that the solar radiation was formerly less considerable than now. If so, we may lengthen several fold the dura- tion of the supply by contraction of sufficient solar
278
WHAT IS THE SUN?
radiation for purposes of supporting life on the earth, leaving the question of the earth's temperature main- tenance under the supposed circumstances to be dis- cussed in the following chapter. On these grounds we may regard Helmholtz's contraction hypothesis as adequate to satisfy the requirements of geology and physics in regard to the source of the sun's energy. Whether or not radio-active processes are, or have been, considerable sources of solar energy is not yet determined.
CHAPTER VII
THE SUN AS THE EARTH'S SOURCE OF HEAT
Causes of Low Temperature at High Altitudes. — Measurement of the Intensity of Sun Rays. — Dependence of Solar Radiation on Air Mass. — The Transmission of the Atmosphere. — The "Solar Constant of Radiation. "—The Light of the Sky.— The De- pendence of the Earth's Temperature on Radiation. — Fluctuation of Solar Emission. — Geological Temperatures.
NEARLY all of the heat of the earth's surface comes directly from the sun's rays. The heat of coal and wood and the energy of water power and wind, from which heat may be derived, are indirectly the effect of solar rays either of present or past times. Occasion- ally a person is met with whose mind works so curi- ously as to lead him to deny that the sun is hot. Such an one almost invariably calls attention to the fact that as we ascend a mountain, or are carried up by a balloon, the temperature falls. Thus, although we may be actually approaching the sun, the heating effects of the solar rays become less obvious. Of course the elevation possible for man to attain is in- significant compared with the radius of the earth's orbit, so that no change of solar radiation ought to be appreciable from the change of distance to the sun involved in climbing a mountain. But a consider- able increase in the intensity of the sun's rays attends
280
THE SUN AS THE EARTH'S SOURCE OF HEAT
the mere ascending above the lower dusty part of the atmosphere. Hence there is some excuse for surprise at the decrease of temperature? observed at high alti- tudes, which occurs notwithstanding the increase in the direct solar radiation.
One secret of this paradox lies in the fact that the sun's rays heat only objects which absorb them. Highly transparent objects like glass, or the air, de- rive little heat by being shined upon; for the rays pass through them almost unchanged. Absorbing sub- stances like lamp-black, on the other hand, almost entirety destroy the rays and convert their energy of vibration into heat. Upon the surface of the earth the air is in contact with such an absorbing substance, namely the ground, and is warmed by contact with it. At high altitudes the free air has contact with no ab- sorbing substance to warm it, and as it transmits sun rays with great freedom it derives only a little heat from them directly. It contains, moreover, ozone, carbon dioxide, and water vapor which all radiate freely long-wave rays and thus dissipate to space the heat gained. Consequently the high air is cold, and cools whatever it blows upon. Its cooling action on the surfaces of mountains is greater on account of the high winds which prevail.
Rising currents warm the upper air less than they would do but for the decrease of atmospheric density which occurs with increasing altitudes. For the air currents which rise from the heated surface of the earth expand in rising, and by expansion are some-
281
THE SUN
what cooled. A factor of considerable influence tend- ing to cause lower temperatures on elevated inland table lands, like the plateau of Thibet, is the compar- ative lack of water vapor in the air above. The water evaporated from the Indian Ocean can hardly reach the plateau of Thibet because in rising through the free air to such a great height it is so much cooled as to be mostly precipitated. Water vapor, while nearly transparent to light, and indeed to perhaps eighty-five per cent of all .the rays which the sun sends, is on the other hand a powerful absorber of the rays of great wave length which are emitted by a comparatively cool body like the earth. Hence at low altitudes where water vapor is plentiful in the air, it is a considerable hindrance to the escape of earth rays to space. In the comparative lack of water vapor at high altitudes of interior regions of large continents, the cooling of the ground by radiation to space is much more rapid than at sea level, and hence lower temperatures prevail. In the case of steep and rough mountains the configuration of the ground is con- ducive to low temperatures because it diminishes the radiation per unit area received from the sun, while increasing the area affected by the cooling winds.
We may therefore attribute the coolness of the free upper air to its transparency, its considerable radi- ating capacity, and its expansion ; the coolness of the rugged mountains to their contours, and to the con- tact of the cool winds; the coolness of the elevated inland plateaus to the dryness of the air above them ;
THE SUN AS THE EARTH'S SOURCE OF HEAT
at the same time recognizing that they all three are receiving more intense rays from the sun than is the earth's surface in general.
MEASUREMENT OF THE INTENSITY OF SUN-RAYS
That which the sun sends to the earth in such abun- dance used to be considered as three distinct things, namely: Actinic or chemical rays; light or visible rays; heat or invisible rays. These distinctions are now known to be misleading, for the rays which affect modern photographic plates extend in the spectrum from far beyond the farthest violet to far beyond the farthest red, and the rays which can produce heat in- clude all these, and many more, still further beyond the red. All rays may be totally transformed to pro- duce heat, however they may differ in their effects upon the eye, or on different chemical substances. All these rays travel with equal velocity in free space, and this velocity is about 300,000 kilometers (186,- 000 miles) per second. That which so travels is not a material substance, but waves, similar in some re- spects to the waves which travel on water, or on a stretched rope. That which distinguishes red light from blue light is the length of the wave, or the num- ber of complete waves executed per second. The wave lengths of visible light vary from about 0.0004 millimeter in the violet to 0.0007 millimeter in the red; and the corresponding numbers of vibrations per second from 750 to 430 millions of millions. But there have been recognized by means of photography
283
THE SUN
rays of wave length only 0.0001 millimeter and wave frequency 3,000,000,000,000,000. By delicate heat measuring apparatus rays of wave length 0.06 milli- meter and frequency 5,000,000,000,000 have been recognized. All this, and perhaps a wider range of spectrum, is probably included in the sun beams as they leave the sun, but our atmosphere prevents some of the shortest and longest of them from reach- ing the surface of the earth.
Since it is upon the supply of these sun rays that heat, light, power, and the growth of all living things upon the earth depends, the measurement of the in- tensity of the total supply, and the determination of the different varieties which compose it, are of first- rate interest and importance.
We measure the intensity of solar radiation by the heat which it will produce when completely absorbed on a surface at right angles to the rays. A conven- ient unit for measuring solar heating is the calory per square centimeter per minute (see Chapter II). The maximum intensity of solar radiation as measured near sea level at Washington when the sun is not more than 45° from the zenith usually ranges from 1.15 to 1.45 calories per square centimeter per minute on cloudless days, depending on the clearness and dry- ness of the air. At Mount Wilson in California, over one mile above sea level, the values observed range from 1.45 to U>2 calorics; and on Mount Whitney in California, nearly throe miles in altitude, the ob- served values reach 1.75 calories.
284
THE SUN AS THE EARTH'S SOURCE OF HEAT
Fig. 60 shows the march of intensity of sun rays during the forenoon of July 6, 1910, on Mount Wilson. The horizontal scale gives zenith distances, the verti^
1.4 1.2 1.0 OR |
' |
|
||||||||||||||||
" - |
>• |
^ |
^ |
|||||||||||||||
^ |
X |
|||||||||||||||||
\ |
\ |
|||||||||||||||||
\ |
\ |
|||||||||||||||||
\ |
||||||||||||||||||
\ |
||||||||||||||||||
\ |
||||||||||||||||||
\ |
||||||||||||||||||
\ |
||||||||||||||||||
CALORIES. P ? |
\ j |
|||||||||||||||||
j |
||||||||||||||||||
)° 10" 20" 30° 40° 5O* 60° 70° 80* 90 ZENITH DISTANCES. |
FIG. 60. — MABCH OF INSOLATION (Mount Wilson).
cal scale calories per square centimeter per minute. The decrease of intensity at the smallest zenith dis- tance observed is caused by increased humidity, due to the springing up of a sea breeze about eleven
285
THE SUN
o'clock. The individual observations are given in Table XVII, page 287.
Formerly radiation was regarded as three distinct entities, namely : actinic or chemical rays ; visible or light rays; obscure or heat rays. As already stated this view is an error now generally abandoned, and all radiation comprised in these three categories is rec- ognized as of the same fundamental kind, differing only as to wave length. The reader will therefore recognize that Table XVIII is not intended to re- vive this ancient classification, but only to fix our ideas of the amount of solar radiation found in the regions (1) where ordinary photographic plates are most sensitive, (2) where the eye is the most sensitive, and (3) in the infra-red spectrum. These facts are given for the beam outside the earth's atmosphere, and as it reaches Mount Wilson and Washington under different angles of zenith distance. The num- bers express the radiation, within the stated regions of wave length, in calories per square centimeter per minute. The zenith distances selected are 0°, 60°, 70° 32' and 75° 32', for which the "air masses"1 are 1,2, 3, and 4.
DEPENDENCE OF SOLAR RADIATION ON AIR-MASS
It is not possible to express satisfactorily the de- crease of intensity of the direct solar beam, depending
1 The "air-mass" is the ratio of tho length of the path of the sun's rays in the atmosphere to the corresponding length if the sun were vertically overhead. It is closely expressed by the secant of the zenith distance for zenith distances less than 75°.
286
THE SUN AS THE EARTH'S SOURCE OF HEAT
3
s
287
THE SUN
'O 00 CO O
<N CO iM 00
d d d d
o o o o
o o
00 £
T-l 00
<£> O
d i-L
o o o ^
<N O
T-l IO
d d
i-H IO CO ^
d d d ^H
o ^H
o p p
(N CO
o d
o o
000^
id 4.3 8 «
3332
Q »O O O
4.0 id id id
288
THE SUN AS THE EARTH'S SOURCE OF HEAT
on the decreasing elevation of the observing station. For the lower layers of the air contain a load of dust and water vapor which changes in quantity, quality, and distribution from hour to hour and day to day. In short, the variation of the solar beam in the lower atmosphere does not proceed according to any fixed law of relation to the barometric pressure.
As for the change of the intensity of the direct solar beam for different zenith distances of the sun, that may be well expressed by Bouguer's exponential for- mula, e = e0a se ' z, as explained in Chapter II, pro- vided we deal with homogeneous rays (rays which are practically of one wave length), and observe them at a single station on a clear day. If we imagine the atmosphere to be made up of a great number of shells concentric with the earth, and the shells of such thick- ness as to contribute equal amounts to the barometric pressure, each of the upper shells will transmit to the shell next below practically an unchanging fraction of the intensity the shell receives of a homogeneous ray. But when the ray reaches a layer within one or two miles of sea level the fraction transmitted continually decreases from shell to shell owing to the increasing load of dust carried by the lower layers.
The total thickness of the atmosphere necessary to be considered as affecting solar radiation is less than one hundred miles, and is so small compared with the earth's radius that the shells may be regarded as prac- tically parallel planes, except when we deal with rays entering the atmosphere at very great zenith dis-
289
THE SUN
tances. Atmospheric refraction, too, may be neg- lected in these computations for rays whose zenith distance is not above 75°. Hence we may assume that for zenith distances less than 75° the ratio of the length of the path of the ray in each shell to the thick- ness of the shell is constant, and equal to the secant of the zenith distance. Under these restrictions (as shown in Chapter II) the exponential formula of Bouguer serves to determine the intensity, e, of mono- chromatic rays at different zenith distances, even though we do not know the change of transmission from layer to layer. For as the sun rises higher and higher the thickness in every layer changes in the same proportion. In thought we may go even further, and, with the sun in the zenith, imagine that the thickness in every layer should be reduced simultane- ously in equal proportions until no air remains. In other words, we can, after the secant reaches its min- imum value, unity, substitute another function of the quantity of air in each shell, which we imagine to be decreased in equal proportion in all layers until no more atmosphere is left. Thus we may determine the intensity, e0, which our monochromatic ray would have outside the earth's atmosphere.
The quantity, a, which appears in the formula, is the fraction of the intensity outside the earth's at- mosphere which remains in the beam as it reaches the observer at the earth's surface. This quantity is called the atmospheric transmission coefficient. It differs with the altitude of the observer and the clear-
290
THE SUN AS THE EARTH'S SOURCE OF HEAT
ness of his sky. It differs also for rays of different colors; increasing, generally, as we pass from short wave lengths to longer ones. There are, however, cer- tain rays which suffer powerful selective absorption in the gases and vapors of the earth's atmosphere, and for such rays the transmission coefficients are very small. Absorption bands play a very great part in the red and infra-red spectrum, where the bands of oxygen, water vapor, and carbon dioxide are principal- ly found. This is made clear in the accompanying illustration, Fig. 61, which shows two successive ob- servations made on Mount Wilson by the bolometer of the relative intensity of the rays in the solar spec- trum of a 60° flint-glass prism. At places marked * the sun rays were cut off so as to give the base line, or line of zero radiation. At places marked | the sun rays were altered in intensity so as to keep the curve within the bounds of the plate. The heights above the base line are proportional to the energy of the spectrum rays. The length is proportional to the prismatic de- viation. Fraunhofer lines show as depressions of the curve. Prominent Fraunhofer lines are indicated by their letters. These energy curves, or holographs, were made on Mount Wilson as a part of a series of six such curves obtained at different solar zenith dis- tances in a single forenoon. They were made to de- termine the transmission of the atmosphere at all parts of the spectrum. From such observations the distribution of solar radiation as it would be outside of our atmosphere is computed. We have studied in
291
THE SUN
THE SUN AS THE EARTH'S SOURCE OF HEAT
Chapter III the significance of such work in regard to the sun's temperature.
We could not determine the intensity outside the atmosphere if the transparency of the air varied much during the several hours required to complete the series of holographs. Fortunately there is the follow- ing criterion for the excellence of any given day in this respect : In the course of the usual reductions, logarithms of the heights above the base-line (corre- sponding to intensities of radiation at given wave lengths are plotted against zenith distances. The results should show straight lines. Fig. 62, p. 294, shows how well this test is met by the Mount Wilson conditions. The tangent of the inclination of such lines gives the logarithm of the transmission at ver- tical sun, which we have called a. Values of a for a given wave length are of course greater for Mount Wilson than for Washington. By dividing the aver- age Washington values by those for Mount Wilson we obtain the average transmission of the mile of air nearest sea level, as it is above Washington. We shall see in the following tables that the loss in passing through this last mile of the air is almost the same as the entire loss above Mount Wilson.
Bouguer's formula is not exactly applicable to pyr- heliometric measurements of the total radiation (or summation of all rays of all wave lengths) of the sun. It fails because rays suffer unequal extinction in the atmosphere, some being almost completely extin- guished in the upper air owing to the action of water
293
THE SUN
vapor and other selective absorbents. Hence for these rays the intensity at the earth's surface does
0 |
^2.538 |
40 49 25 17 762 SO |
|||||||||
\ |
\ |
||||||||||
2 6 |
^ |
||||||||||
\ \ |
|||||||||||
$s* |
\ |
||||||||||
\ |
\ |
\ |
|||||||||
6 2 6 |
£779 |
*\ |
\ |
||||||||
32.928 |
X |
\ |
\ |
||||||||
» |
xx |
NNX |
\ |
\ |
|||||||
xx |
t> |
\ |
\ |
,0 |
|||||||
2n |
^ |
k |
. |
5 |
\ |
||||||
o LOGARITHM WTEc |
^ |
K |
\ |
X |
|||||||
:3.\28 |
O |
N |
K^ |
\ |
\ |
||||||
S3.I20 |
^"^ |
— o — ^0^- |
£ |
o |
•& |
^x |
sN |
||||
"---<c |
"~^ |
1 — — |
N^.'£ 'V.TC, |
||||||||
>-OSECANTZI2 24 3.6 FIG. 62. — ATMOSPHERIC TRANSMISSION PLATS. |
not alter much with the zenith distance. Neverthe- less the exponential formula holds approximately
294
THE SUN AS THE EARTH'S SOURCE OF HEAT
even for total radiation, except that the logarithmic plats like those given in Fig. 62 are slightly curved, and if continued on in straight lines to zero atmos- pheric thickness they fall below the real intensity of total solar radiation outside the earth's atmosphere, as obtained from spectrum observations. Pouillet, however, determined transmission coefficients for the total solar radiation, and was thereby led to his cele- brated value 1.76 calories per square centimeter per minute for the solar constant of radiation. Radau, and later Langley, showed clearly that, on account of the differences of transmission for rays of different wave lengths, we must observe the transmission of each color by itself, and determine what the intensity of each separate color would be outside the atmos- phere. Langley first applied this procedure experi- mentally. Following his method we may sum up the area included under the solar spectrum energy curve outside the earth's atmosphere, and compare it with the area for the corresponding curve at zero zenith distance of the sun. Thus we may find the actual vertical transmission of the atmosphere for the total radiation. Knowing by measurements of the pyr- heliometer the intensity of total radiation for any ob- served zenith distance we can determine how many heat units the area of the corresponding spectrum- energy curve represents. Summing up in similar terms the area as it would be outside the earth's at- mosphere we may obtain the true " solar constant."
21 295
THE SUN
TRANSMISSION OF THE ATMOSPHERE
In the following table there are given the mean results for the vertical transmission of total solar radiation, according to observations of the Astro- physical Observatory of the Smithsonian Institution.
TABLE XIX. — Transmission for total solar radiation
PLACE |
Washington |
Mount Wilson |
Mount Whitney |
True transmission |
0.699 |
0.817 |
0.896 |
Apparent transmission |
0.787 |
0.894 |
0.960 |
Table XX, opposite, gives the atmospheric trans- mission for vertical rays, and for the zenith distances whose secants are two and three, respectively, for rays of various wave lengths.
THE SOLAR CONSTANT OF RADIATION
From the mean results of the Washington observa- tions of 1902 to 1907, the Mount Wilson observations of 1905 to 1910, and Mount Whitney observations of 1909, 1910, all corrected to the absolute scale of heat,1 the total intensity of solar radiation outside the earth's atmosphere at the earth's mean distance from the sun (called the " solar constant" of radiation),
1 Values published in Volume II of the Annals of Astrophysical Observatory of the Smithsonian Institution were given on a pro- visional scale of pyrheliometry differing about five per cent, from the true one, and are here given as reduced to true calories.
296
THE SUN AS THE EARTH'S SOURCE OF HEAT
dddd
tOl^ O CO l^co co **
OiOiOiOi
dddd
§o^§§ dddd
OiOi CO CO
dddd
TjH <M GO O
tO to CO Oi Oi Oi GO GO
dddd
CO CO Oi F^-
<M Oi >O •* Oi GO 1> GO
dddd
GO i
OOOO
=oc5«^«^ ocOqq
dodo
§co
>>
S a
0000
O^ O^
oooo
:28
' CO CO
(M
O CO <N
(N CO CO t^
»O <M CO O T-H CO Oi iO
GOr^ ** co
CO *O C^l iO
Ss
fl «s §
Illl
CO Oil> GO
dddd
^ CO ^ Oi
•^ to CO Oi Oi Oi 1> 1>
dddd
CO "* (M tO
dddd
O
t^ CO CO ^O
dddd
CO<M i-i t^"
CD -M
2 .
.agd.t5
"SF1 §3 S^'-S S
§ 'C t*
297
THE SUN
as expressed in calories per square centimeter per minute is as follows :
PLACE |
Washington |
Mount Wilson |
Mount Whitney |
|||||
Date |
1902-1907 |
1905 |
1906 |
1908 |
1909 |
1910 |
1909 |
1910 |
Observations. |
44 |
59 |
62 |
113 |
95 |
28 |
1 |
3 |
Mean solar constant.. . |
1.960 |
1.925 |
1.921 |
1.929 |
1.896 |
1.914 |
1.959 |
1.956 |
In 1909 and 1910 observations were made simul- taneously on Mount Wilson (elevation one mile) and Mount Whitney (elevation nearly three miles) by Smithsonian observers, with the following results:
DATE |
1909 Sept. 3 |
1910 Aug. 12 |
Aug. 13 |
Aug. 14 |
Mount Wilson |
1.943 |
1.943 |
1.924 |
1.904 |
Mount Whitney |
1.959 |
1.979 |
1.933 |
1.956 |
We see that notwithstanding the differences in al- titude of the observing stations, and the differences of atmospheric transmission above them, there is good agreement between the computed values of the " solar constant" of radiation. Prior to 1905 this quantity was in great doubt, as numbers ranging from 1.76 to 4.10 had been given for it, and the ac- cepted value then was 3.0. The opinion now seems to prevail that no considerable change from the Smith- sonian result of about 1.95 calories per square centi-
298
THE SUN AS THE EARTH'S SOURCE OF HEAT
meter per minute will come from future experiment- ing.
Expressed in another way, the measurements indi- cate that if the sun's rays could be completely em- ployed to melt ice exposed continuously to them at right angles, they would suffice to melt a layer 426 feet thick in a year.1 Such a layer at the earth's mean distance, if it entirely surrounded the sun, would weigh 4 X 1025 (4 followed by 25 ciphers) tons, and the complete melting of it each year would represent as many heat units as the burning of 4 X 1023 tons of anthracite coal. This, then, is a measure of the sun's yearly output of radiation.
THE LIGHT OF THE SKY
It must not be inferred from the tables given on a preceding page that only 81.7 per cent of the sun's radiation reaches the Mount Wilson level at vertical sun. That, to be sure, is the average result for the direct solar beam, but the sky supplies an appreciable addition of indirect rays even on Mount Wilson. At sea level the sky light is a still more considerable por- tion of the total radiation, but as yet not very ex- actly measured. The relative brightness of the sun and sky differs greatly according to the manner in which the rays are received. Owing to the great ex- tent of the sky, it is not possible, when receiving rays
1 As the earth has four times the area of its cross-section, we may say that the sun's rays are capable of melting an ice shell covering the earth to an average thickness of 106 . 5 feet annually.
299
THE SUN
simultaneously from its whole extent, to have them all fall at right angles to the absorbing surface. Hence the sky light is at a disadvantage with respect to sunlight, unless we observe the brightness from every part of the sky by itself and then sum up the results. From bolometric measurements of 1905 and 1906, made by the Smithsonian observers, and reduced in this manner, it appears that the total sky radiation on Mount Wilson computed at normal in- cidence, and including all wave lengths, is from eleven to twenty per cent of the total direct sun radiation. Both sun and sky rays are in this estimate supposed to be received at right angles to the absorbing sur- face, and the sun to be not over 50° from the zenith. The percentages depend on the clearness of the sky, increasing with the haziness. If we make the as- sumption that the sky shines on a horizontal surface, and the sun upon n surface normal to the beam, these percentages become 5.2 and 7.7. If both sun and sky rays are supposed to shine on a horizontal surface, the ratio varies of .course greatly from hour to hour. Professor Exner has derived formulae for the rela- tive brightness of the sun and sky on the hypothesis that the sky light is all due to scattering from parti- cles which are small as compared with the wave length of the rays.1 He has found it necessary to make some rather rough simplifying assumptions. Never- theless his computations fall in pretty well with such
1 Sitzungsbericht, d. K. Aknd.d. Wissen., Wien., M. N. Klassr. CXVIII, Ha, 1909.
300
THE SUN AS THE EARTH'S SOURCE OF HEAT
observations as are available. In the following tables taken from Exner's publication, z is the zenith distance of the sun, H the intensity of sky light and S that of sunlight, both being measured on a horizontal surface. At normal incidence outside the atmos- phere the intensity of sunlight is taken as unity. The quantity p is the transmission coefficient of the atmosphere above the observer for a vertical ray. From Smithsonian observations we see that p = 0.6 would correspond to wave length 0.43//, (violet) at Washington 0.35/z. (ultra-violet) at Mount Wilson. Correspondingly, for p = 0.75 we have 0.59/*, (yellow) at Washington and 0.41//, (violet) at Mount Wilson.
TABLE XXI. — Sunlight and sky light. (Exner.)
P = 0.6 |
p = 0.75 |
|||||||
7, |
H |
S |
S + H |
S H |
H |
S |
S + H |
S H |
80° |
0.241 |
0.009 |
0.250 |
0.04 |
0.136 |
0.032 |
0.168 |
0.24 |
70° |
0.245 |
0.077 |
0.322 |
0.31 |
0.138 |
0.147 |
0.285 |
1.06 |
60° |
0.252 |
0.180 |
0.432 |
0.72 |
0.141 |
0.282 |
0.423 |
2.00 |
50° |
0.259 |
0.289 |
0.548 |
1.12 |
0.146 |
0.408 |
0.554 |
2.79 |
40° |
0.268 |
0.394 |
0.662 |
1.47 |
0.151 |
0.528 |
0.679 |
3.50 |
30° |
0.276 |
0.484 |
0.760 |
1.75 |
0.155 |
0.625 |
0.780 |
4.03 |
20° |
0.281 |
0.547 |
0.828 |
1.95 |
0.158 |
0.693 |
0.851 |
4.38 |
10° |
0.285 |
0.582 |
0.867 |
2.04 |
0.160 |
0.731 |
0.891 |
4.57 |
0° |
0.288 |
0.600 |
0.888 |
2.08 |
0.162 |
0.750 |
0.912 |
4.63 |
The change of H with the zenith distance of the sun is not as great in these tables as it should be. This appears from the following measurements of Roscoe for which we may assume p =0.6.
The units employed by Roscoe are not the same as 301
THE SUN TABLE XXII.— Sunlight and sky light. (Roscoe.)
z |
80° 9' |
70° 19' |
58° 46' |
47° 47' |
36° 51' |
25° 46' |
H |
0.038 |
0.062 |
0.100 |
0.115 |
0.126 |
0.138 |
S |
0.000 |
0.023 |
0.052 |
0.100 |
0.136 |
0.221 |
those employed by Exner, so that for easier compari- son the results of Roscoe may be multiplied by 2 or 2.5. As there is much difference for different days and for different stations in results of this kind, Ex- ner's computations seem to be near enough at least for giving a general idea of the state of affairs. Indeed the following summary, which I translate from Wies- ner's description of his photographic observations of light received on horizontal Surfaces,1 fits Exner 's results for short wave lengths very well :
"The direct sunlight, which is sometimes twice the intensity of diffused light, may also sink to zero. — For solar altitudes less than 19° (z = 71°) the chem- ical intensity of the sunlight as compared with dif- fused daylight is negligible. With increasing solar altitude the intensity of the direct sunlight gains in comparison with the diffused light. The solar alti- tude for which S = H seems not to be constant even for apparently clear sky, and for one and the same station. For cloudless sun the equality of direct and diffused light occurs generally when the solar alti- tude is about 57° (z = 33°), yet with clear sky it was
'Vienna Academy. Denkschriften, Bd. 64, 1807.
303
THE SUN AS THE EARTH'S SOURCE OF HEAT
observed once at 33° (z = 57°). Since the intensity of the direct beam may reach twice that of the dif- fused, the total combined chemical effect may be threefold that of the diffused light."
Exner also gives computations of the relative amounts of combined direct and diffused light receiv- able on vertical surfaces facing respectively South, North, West, East, compared with the amounts re- ceived on a horizontal surface. The southern expos- ure only of the vertical surface is supposed to receive some direct sunlight, designated by 2. V8, VN, Vw, VE designate the diffused illumination of the vertical surface toward the South, North, West, and East. S and H have their former meanings.
TABLE XXIII. — Sky light on a vertical surface. (Exner, Schramm.)
Computed, p=0.8 |
Observed by W. Schramm |
||||||
z |
2 + Vs |
VN |
Vw VE |
2 + V8 |
VN |
Vw |
VE |
S + H |
S + H |
S + H S + H |
S + H |
S + H |
S + H |
S + H |
|
85° |
1.43 |
0.537 |
0.461 |
2.73 |
0.560 |
0.542 |
0.604 |
75 |
2.21 |
0.263 |
0.241 |
3.41 |
0.268 |
0.397 |
0.386 |
65 |
1.69 |
0.148 |
0.146 |
1.81 |
0.258 |
0.331 |
0.351 |
55 |
1.25 |
0.099 |
0.104 |
1.32 |
0.147 |
0.223 |
0.204 |
45 |
0.92 |
0.075 |
0.083 |
0.976 |
0.118 |
0.195 |
0.175 |
35 |
0.68 |
0.062 |
0.071 |
0.749 |
0.091 |
0.131 |
0.139 |
It is not probable that the influence of the direct sunlight was wholly absent in Schramm's VE and Vw observations. Apart from these there is a pretty good agreement of observed and computed results, as the following summary also indicates,
303
THE SUN
Ratios of average vertical illumination to that of the North.
2 + Vs |
VN |
Vw |
YE |
Vmean |
|
Observed |
7 64 |
1 00 |
1 26 |
1 29 |
2 80 |
Computed |
6 91 |
1 00 |
0 94 |
0 94 |
2 45 |
Exner computes from p = 0.6 and z = 40°, the following :
S + H |
2 + Vs |
VN |
Vmean |
S + H |
S + VB |
|
v w— VE |
Vmean |
VN |
||||
0.662 |
0.480 |
0.106 |
0.120 |
0.207 |
3.2 |
4.5 |
As we have stated at some length in Chapter VI, Schuster has employed Lord Rayleigh's theory of the scattering of light by particles small compared with the wave lengths, to compute the transmission of the direct beam of sunlight. He assumes that the loss in the atmosphere is wholly from the scattering caused by the molecules of the air. He finds close agreement between the computed and observed re- sults for excellently clear days at Mount Wilson and Washington. This seems to indicate that the dust load of the atmosphere plays a subordinate part in affecting the solar radiation on the best days, and that under such conditions as are found ordinarily at Mount Wilson, and occasionally at Washington, nearly all the light of the sky is due to the diffuse
THE SUN AS THE EARTH'S SOURCE OF HEAT
reflection or scattering of sun rays by the molecules of air. Somewhat similar conclusions have been reached still more recently by Natanson, except that he considers the matter from the point of view of the electron theory.
The light of the sky is very much richer in the vio- let rays than that of the direct solar beam. From Smithsonian experiments the following values are taken for different wave lengths, assuming about equal intensities in the extreme red for a sunbeam and a skybeam, and giving the approximate normal spectral distribution for both as observed at the sur- face of Mount Wilson.
TABLE XXIV. — Sun and sky light. Relative brightness for different wave lengths on Mount Wilson.
WAVE LENGTH -*• |
0.422/w, |
0.457M |
0.491ft |
0 . 556/u. |
0.614/* |
0 . 660M |
Sunlight; z = 50° |
186 |
232 |
227 |
211 |
191 |
166 |
Skylight Ratio |
1,194 6 92 |
986 4 25 |
701 3 09 |
395 1 87 |
231 1 21 |
174 1 05 |
At the sea-level, especially in cities and other dusty localities, the proportion of blue in sky light is usu- ally much less than that given above; for particles large as compared with the wave length of light, such as occur in dust, do not act in the same way as small particles and molecules. Large particles, by reflect- ing sunlight, tend rather to diminish than to increase the relative proportion of the intensity due to rays of short wave length.
Skylight is brightest near the sun and near the hori- 305
THE SUN
I |
i— i d |
d |
B i |
8 |
.!! |
|
°o |
00 CO |
CD |
COCO |
iCt- |
||
\ |
i— i |
(N |
CO |
r*» 01 |
O l^- |
|
g |
o |
d |
d |
d t^ |
i-i <N' |
cod |
i |
» |
I> l-H |
•^ |
|||
i |
0 |
00 1C |
* |
coc^ |
01 Cl COO |
|
^ |
0 |
d |
d |
CO ui |
dei |
|
|2 ~ |
CO |
CO i-H |
1-1 |
|||
1 |
0 |
co |
||||
R Q |
^f |
T-l |
t£ |
•^ |
r^[^ |
O5 O |
| |
% |
d |
d |
rH (M |
00 i-H |
i~ |
>-< |
0 |
CO |
||||
•1 |
Q| |
t3 |
Oi |
2§ |
CO® |
|
^ |
d |
d |
<N d >C i— i |
o>~ |
co'^ |
|
1 |
h |
s |
01 |
00 |
28§ |
%% |
1 |
o 1C |
d |
d |
5 2 |
o^ |
*~ |
^ |
1C |
CO |
||||
1 £ |
i-H o |
d |
01 d |
CO |
:1 |
!o |
s |
||||||
1 |
d |
: g |
a |
|||
1 |
1 |
"S J |
II |
|l |
||
\ > |
£ |
I |
[Vj [^ |
|||
X |
>> |
^ — s |
||||
X 3 |
"8 cu |
1 |
^ G ? 5 o |
§ |
I-H ! |
|
5 |
1 |
Q |
i G $ |
rH 0 |
H? u |
-D : |
E |
O |
1 1 |
1 § |
-+J o 3 |
1^ |
|
i— i |
>— i |
6 |
o |
• |
£ 1— 1 |
300
THE SUN AS THE EARTH'S SOURCE OF HEAT
zon. The results shown in Table XXV on the total radiation are from bolometric measurements at Mount Wilson, and at Flint Island, a coral island in the Pacific, near the equator, lying 400 miles northwest of Tahiti.
The measurements are reduced as if the sun were in the zenith.
Summing up columns* (V) and (VI) we find the average brightness of a portion of the sky equal in angular area to the sun as compared with the bright- ness of the sun: First as received on a surface at right angles to the beam in both cases; second with the skylight received on a horizontal surface and the sunlight received normally. The measurements in- clude all the rays transmissible by a glass plate three millimeters thick. The results are as follows:
TABLE XXVI. — Ratio of total radiations: Sky to sun
STATION |
FOR EQUAL AREAS |
FOR WHOLE SKY |
||
Normal incidence |
Sky on horizontal |
Normal incidence |
Sky on horizontal |
|
Flint Island |
636X10-8 |
302X10-8 |
0.67 |
0.32 |
Mount Wilson |
176X10-8 |
69X10-8 |
0.18 |
0.072 |
Thus according to these measurements (which how- ever are not sufficiently numerous or exact) at sea level the sky furnishes to a horizontal surface thirty- two per cent as much radiation as the direct high sun. At 1,800 meters elevation only 7.2 per cent.
307
THE SUN
THE DEPENDENCE OF THE EARTH'S TEMPERATURE ON RADIATION
The temperature of the earth seems to be main- tained at present almost wholly by the absorption of solar radiation. It is thought by some that the earth's temperature is rising slowly. If so, this would indicate that the sum total of the earth's sup- plies of heat exceeds its losses. But this change of temperature, if real, is so exceedingly slow that we may practically say that the earth's heat income and heat outgo balance. The outgo, neglecting the rela- tively trifling effects of vegetable and other storage processes, is made up wholly of the earth's radiation of long-wave rays to space. It has been shown that the absorptive effects of atmospheric water vapor, carbonic acid and ozone combined prevent nearly or quite nine-tenths of the rays which are emitted at the earth's surface from escaping directly to space.
Hence the earth's effective radiating layer may be regarded as situated in the atmosphere, and as being chiefly the water vapor layers at several miles ele- vation, whose average temperature is about — 10° C. The still higher situated effective radiating layers of carbonic acid and ozone gases, whose average temper- atures reach as low as — 60° C, also radiate freely in a few limited regions of spectrum. We shall not be far astray, therefore, if we regard the average tem- perature of the earth's radiating layer as not above
308
THE SUN AS THE EARTH'S SOURCE OF HEAT
260° of the absolute centigrade scale.1 The constitu- ents of this radiating layer, namely, water vapor, car- bonic acid and ozone, owing to their powerful selec- tively absorbing properties for rays proper to this temperature, must also be nearly perfect radiators for the same temperature. Admitting that their radiating power is perfect, the emission of the as- sumed radiating layer at 260° C. absolute is, by Stefan's law (see Chapter II), about 0.34 calories per square centimeter per minute.
In order to sustain this average rate of loss of heat over the whole surface of the earth, solar radiation, shining effectively over only the area of the earth's cross-section, must be absorbed at four times this rate, or 1.36 calories per square centimeter per min- ute. Of the energy represented by the solar con- stant (1.95 calories) about thirty-five per cent is re- flected away according to the Smithsonian determi- nation of the earth's " albedo. " The remainder is 1.27 calories, and nearly suffices to furnish the heat above computed as lost from the earth. The difference (0.09 calories) may possibly mean that an appreciable quantity of heat is furnished by terrestrial sources, such as radio-active processes. However it seems quite reasonable to suppose that the difference may be accounted for (1) by assuming that the earth's effective radiating layer is not a perfect radiator, so that its radiation falls short of the 0.34 calories per square centimeter per minute which a perfect radi-
1 Water freezes at 273° of this scale, and boils at 373°. 309
THE SUN
ator at 260° C. absolute would emit; or (2) that the effective radiating temperature is below 260° C. absolute.
The surface temperature of the earth reaches 310° absolute C. in the tropics, and at the poles falls as low as 220°, and its effective mean temperature is 287.2° absolute C. or + 14.2 C. It exceeds the tem- perature of the radiating layer by over 25°. This is in large measure for the same reason that the gar- dener's hot beds, or the steamfitter's asbestos-covered pipes, exceed the temperature of their surroundings. For the sun's rays shine through the atmospheric vapors readily, and warm the earth's surface. The escape of its heat, as we have seen, is hindered by the atmosphere. Hence the earth's surface temperature rises sufficiently to force a flow of heat out to the effective radiating layer. If it was not for the blank- eting effect of the water vapor of the atmosphere, the earth's mean surface temperature would probably be nearly 20° C. below freezing, providing the reflecting power of the earth was not changed. But if there was no water vapor in our air, the sun's rays would reach the earth's surface with at least ten per cent greater intensity on cloudless days than they do now. Since clouds would then be absent there would be about 1.75 calories instead of 1.27 as now available to warm the earth. Consequently, the earth's mean temperature, if water was absent, would be about 277° absolute or + 4° C. But there would then be a much greater range of temperature between night
310
THE SUN AS THE EARTH'S SOURCE OF HEAT
and day and between summer and winter than there is now.
Upon the moon there is no atmosphere, and by the observations of Lor*d Rosse, of Langley and of Very, the moon's sunlit surface falls from about the tem- perature of boiling water nearly to that of liquid air within the short duration of a total lunar eclipse. Quite otherwise is the state of affairs on the earth. In the following table is given the yearly average of the daily range of temperatures which occurs at sev- eral stations on the earth.
TABLE XXVII. — Yearly means and mean daily temperature depar- tures. (Centigrade.)
HOUR -» |
Lati- tude / |
Mid- night |
2 |
4 |
6 |
8 |
10 |
12 |
Tirnbuctu Port au Prince . . . |
16°49'N 18°34'N |
-4M -2°. 6 |
-5°. 6 -3°. 2 |
-6°. 8 -3°. 7 |
-7°. 7 -3°. 8 |
-2°. 8 -0°.6 |
+3°. 2 +2°. 9 |
+6°. 9 +4°. 7 |
HOUR -* |
Noon |
2 |
4 |
6 |
8 |
10 |
12 |
Mean |
Timbuctu Port au Prince . . . |
+6°. 9 +4°. 7 |
+8°. 5 +4°. 5 |
+7°. 4 +3°.l |
+3°. 4 +l°.l |
-0°.l -0°.8 |
-2°. 4 -1°.3 |
-4°.l -2°. 6 |
29°. 2 25°. 9 |
Even a polar night of five months' duration in which the sun is continuously below the horizon pro- duces no such range of temperature on the earth as a total lunar eclipse of a few hours' duration does upon the moon. Witness the following mean temperatures :
Fort Conger. Latitude 81° 44'. Temperatures Centigrade.
Jan. |
Feb. |
Mar. |
April |
May |
June |
July |
Aug. |
Sept. |
Oct. |
Nov. |
Dec. |
-39°.0 |
-40°. 1 |
-33°.5 |
-25°.3 |
-10°.0 |
+0°.4 |
+2°.8 |
+1°.0 |
-9°.0 |
-22°.7 |
-30°.9 |
-33°.4 |
22
311
THE SUN
These examples indicate how slowly the temperature of the earth falls towards the absolute zero when the solar radiation is utterly cut off. The delay cannot be attributed to the influence of the inner heat of the earth. From the rise of temperature at increasing depths in the earth (about 1° C. in 28 meters) taken in connection with the observed conductivity of rock (about 0.0042 calories per second per centimeter cube) it is calculated that the heat supplied to the surface from within is but 0.00010 calories per square centi- meter per minute, which would suffice to keep a per- fect radiator at only 34° absolute (Centigrade) tem- perature and could not be expected to keep the earth's surface above 40° absolute, or -233° C. Even this is far above what the moon and all the stars combined could do to supply the place of the sun.
The following table1 of the mean monthly temper- atures (Centigrade) gives some idea of the yearly ranges of temperatures on the earth at various sta- tions in the Northern Hemisphere. Several pairs of sta- tions at nearly the same latitude, but one inland, the other oceanic, are contrasted to show the influence of the oceans in reducing fluctuations of temperatures.
The reader will notice how much smaller are the yearly ranges of temperatures for oceanic stations than those for the inland stations. It is also appar- ent that the yearly range increases with the latitude. This is due in part to the growing disparity of the
1 The data are taken mainly from various publications of J. v. Hann.
312
THE SUN AS THE EARTH'S SOURCE OF HEAT
STATION |
Ver- khoy- ansk |
Fort Conger |
St. Louis (U.S.A.) |
P. Delgada (Azores) |
Tim- buctu |
Port au Prince |
Bogota |
Jaluit (Mar- shall Islands) |
Latitude N |
67°34'° |
81°44'° |
38°38' |
37°45' |
16°49'° |
18°34'° |
4°31' |
5°55' |
Elevation (meters) |
173 |
20 |
250 |
36 |
2660 |
3 |
||
January. . . February. . March. . April . . May. . . . June. . . July. . . . August. September. October . . . November. December . |
-51°.0 -45°.3 -32°.5 -13°.7 + 2°.0 + 12°.3 +15°.o +10°.l + 2°.5 -15°.0 -37°.8 -47°.0 |
-39°.0 -40°. 1 -33°.5 -25°. 3 -10°.0 + 0°.4 + 2°.8 + 1°.0 - 9°.0 -22°.7 -30°.9 -33°.4 |
- 0°.8 + 1°.7 + 6°.2 +13°.4 +18°.8 +24°.0 +26°.0 +24°.9 +20°.8 +14°.2 + 6°.4 + 2°.0 |
+14°.l +13°.9 +14°.l +15°.4 +1G°.G +18°.9 +21°.3 +22°.0 +20°.9 + 18°.9 +16°.9 + 15°.l |
+21°.8 +23°.8 +28°. 1 +32°.5 +35°.0 +34°.2 +32°.7 +31°.l +31°.8 +31°.0 +26°.8 +21°.4 |
+24°. 1 +24°.G +25°. 1 +25°.9 +26°.0 +27°.l +27°.G +27°. 3 +26°.7 +26°.3 +25°.6 +24°.4 |
14°.2 14°.4 14°.S 14°.7 14°.8 14°.5 14°.l 13°.9 13°.9 14°.4 14°.7 14°.5 |
27°. 1 27°.2 27°.0 26°.9 26°.9 26 .8 26°.8 26°. 9 26°.9 27°. 1 27°. 1 27°.0 |
Yearly Range |
66°.5 |
42°.9 |
26°.8 |
8°.l |
13°.6 |
3°.o |
0°.9 |
0°.4 |
Yearly Mear^ |
-16°.7 |
-20°.0 |
+ 13°. 1 |
+ 17°.3 |
+29°.2 |
+25°.9 |
14°.4 |
27°.0 |
longest and shortest days at higher latitudes, and in part to the more rapid change in the intensity of illumination with change of zenith distance of the sun at high latitudes. At the equator the days and nights are always equal, and the secant of the zenith distance of the noonday sun varies only from 1 to 0.917. At latitude 45° N. the length of day varies from eight hours, thirty-four minutes to fifteen hours, twenty-six minutes, and the secant of the noonday zenith distance from 0.930 to 0.366. The value of the secant of the zenith distance influences the result in two ways, first, as it measures the length of the path of the rays in the air, second, as it measures their weakening on a horizontal surface in conse- quence of obliquity.
313
THE SUN
A minor influence which affects the yearly march of the solar radiation is the change of the sun's dis- tance from the earth. This causes an increase of nearly seven per cent in the earth's heat supply from July to January; and combined with the sun's march in declination it produces two maxima and two minima of radiation in the tropics. At Bogota temperature maxima occur March to May and Oc- tober to December — minima August to September, and January, all occurring a little after the corres- ponding maxima and minima of radiation.
By taking the three factors of solar distance, obliquity and the daily duration of sunlight into account, formulae have been devised for comput- ing the " effective insolation" as it is sometimes called. This is the intensity of the uniform beam, which if received continuously at normal incidence would yield an equal supply of radiation to that which is really effective on a horizontal surface. In such computations atmospheric losses are usu- ally neglected, but on the other hand diffused sky radiation is also neglected. We may also imagine an hypothetical earth equal in size and similar in motions to the real earth, but a perfect absorber and radiator; thin as an egg shell; perfectly conducting of heat from east to west, but perfectly non-conduct- ing from north to south. The temperature of such a structure can be computed for all times and latitudes by Stefan's law (see Chapter II). When such com- puted temperatures are compared with those actu-
314
THE SUN AS THE EARTH'S SOURCE OF HEAT
ally observed on the earth, it is found that no real stations show as great yearly ranges of temperature as the corresponding hypothetical ones. For Tim- buctu and some other desert stations the observed range is over half the computed. For average inland stations the ratio is about three-tenths; for average coast stations, one-fifth; for average island stations, one-twelfth; at Apia, in Samoa, only one twenty- fifth. For the hypothetical earth the percentage change of absolute temperature is everywhere one- fourth of the percentage change in solar radiation which would cause it.
The accompanying illustration, Fig. 63, gives the march of the " effective insolation" at the north lat- itudes 17° 40' and 5° 10', and also the yearly change of temperature at Timbuctu (16°. 49' N.), Port au Prince (18° 34' N.), Bogota (4° 31') and Jaluit (5° 55'). The curves show how much the effect of solar change may be modified by local conditions, and especially how considerable are the delays which occur at oceanic stations between solar causes and their ter- restrial temperature effects. Thus, while at Tim- buctu, an inland station, the maximum and minimum temperatures attend closely the minima of effective insolation, they are so far delayed that the maximum temperature occurs three months after the insolation is at its maximum at St. Louis, Senegambia, a coast station nearly west of Timbuctu. Such facts should be taken into consideration when seeking by studying temperature statistics to determine if fluctuations of
315
THE SUN
34
32'
JAN. APR. JULY OCT.
FK;. <;:<. INSOLATION AND TERRESTRIAL TEMPERATURES.
316
THE SUN AS THE EARTH'S SOURCE OF HEAT
solar radiation probably occur. Long solar periods like the eleven-year sun-spot period may be expected to affect the temperatures of all terrestrial stations nearly simultaneously. Not so temporary solar changes of a few days or months. These could be expected to show their influence only at inland sta- tions, and preferably cloudless ones. The lag of tem- perature minima behind the inducing solar radiation minima is about twenty days for average continental stations, only ten days for particularly favorable ones, but reaches two months or more in the case of many island ones.
FLUCTUATION OF SOLAR EMISSION
, Numerous attempts have been made to see if ter- restrial temperatures, by their departures from the normals, indicate fluctuations of the sun's emission of radiation. Koppen concluded from such investi- gations, published in 1873, 1880, and 1881, that the earth's temperature is higher at sun-spot minimum than at sun-spot maximum. This conclusion is con- firmed by Stone, Gould, Nordmann, Newcomb, Abbot and Fowle, Arctowski, Bigelow, and others. Taking a general view of their results with those of Koppen, we may conclude that for a change of 100 sun-spot numbers of Wolf's scale, which is about the average range of sun-spot activity, there is a change of the mean temperature of the earth of about 0.7° C . The cause of this cannot be in the mere darkening of the disk of the sun by the areas covered by the spots,
317
THE SUN
for if they were perfectly black all over the change of solar radiation corresponding to 100 sun-spot num- bers would be only — — , which is not J of what is nec- 500
essary to produce the observed change of tempera- ture. There must therefore be other changes of the sun (or possibly in the earth's atmosphere or in the intervening space), as yet not understood, which at- tend the increase of sun spots and which exceed in their effective reduction of solar radiation the direct influence of the darkness of the spots themselves.
It is only within the last five years that there have been direct measurements of the solar radiation suf- ficiently complete and accurate to show whether there are frequent changes of the sun's emission suf- ficiently large to affect the earth's temperature notice- ably. The reader might be inclined to suppose that a mere analysis of the deviations from normal tem- peratures at numerous stations would be sufficient to practically verify or disprove frequent variability of the sun. Indeed there are found numerous in- stances of departures from normal temperatures which seem to indicate something of the kind, but yet the evidence of different localities is so contradictory and confusing that careful meteorologists reserve an opinion on the matter. Recently the idea has gained some adherents that a few per cent of increase of solar radiation during a period of several months need not necessarily affect the temperatures of all stations on the earth in the same direction, but might make
318
THE SUN AS THE EARTH'S SOURCE OF HEAT
some warmer and others cooler temporarily, so that the effect may thus be obscured. This contradiction of effects, as these observers think, may be due to changes of cloudiness and of the circulation of the atmosphere. Statistical studies of temperatures by Arctowski, Bigelow, and others support the view that certain large regions are on the whole balanced in their temperature changes, so that the same dis- turbance which warms one region, cools another, as if there were waves of effect superposed upon the earth. Apparently the meteorological condition of the earth is so complicated by the relative configura- tion of land and sea, cloudy and clear areas and hot and cold regions that we cannot expect to determine solar changes with certainty by climatic investiga- tions, and must rather, turning the matter about, first determine the solar changes by direct observa- tions, and then search out their terrestrial effects.
Up to 1905 the measurements of solar radiation made in different countries by different investigators were so much at variance as to make it seem highly unlikely that sufficiently accurate knowledge of the solar emission could be obtained to lead to the dis- covery of variability of the sun. But the Smithson- ian observations at Washington, Mount Wilson, and Mount Whitney as given on a previous page are so highly concordant, and seem to be so probably com- petent to fix the intensity of the solar radiation out- side the atmosphere within about one per cent, that there now seems to be a good prospect of discovering
319
THE SUN
fluctuations of solar radiation if they exceed one per cent from the mean value. Heretofore the observa- tions have not been kept up continuously, but ef- forts are being made to remedy this defect by the establishment of one or more additional " solar con- stant" observing stations.
From measurements of the Smithsonian observers on Mount Wilson during six months of each of the years of 1905, 1906, 1908, arid 1909, as given in Fig. 64, it appears probable that fluctuations of the sun at irregular intervals of several days, and sometimes of several months, are not uncommon. Apparently the amplitude of such changes sometimes reaches ten per cent, and seems frequently to reach from three to five per cent. But notwithstanding that the reality of' these changes is attested by various evidences, such as the continuity of a change during several days of consecutive observing, which the reader can see for himself in the observations of 1908 and 1909, Fig. 64, and the fact that the fluctuation of radiation depending on the yearly change of the solar distance can be easily recognized,1 though it amounts to but three per cent during the six months covered by the Mount Wilson work, yet the supposed solar variabil- ity can hardly be said to be conclusively shown until another station as well equipped as Mount Wilson supports the conclusion by simultaneous measure- ments.
If occasional variations of ten, or even five, per
1 The values platted in Fig. 64 are reduced to mean solar distance. 320
THE SUN AS THE EARTH'S SOURCE OF HEAT
321
THE SUN
cent, shall be found to exist in the solar radiation, they may well be expected to produce noticeable effects on the earth's climate. We have seen that it is difficult, if not impossible, to seek the cause from the observed climatic effects, owing to their great complexity. But when the variations of the sun become accurately observed, and thus the action of the cause is known, the tracing of the climatic effects will be a matter of great interest and importance. Indeed, some comparisons of temperature statistics with the " solar constant" values have been made al- ready, and indicate a probable connection between the two. But much more work is needed for certainty. A solar change of five per cent continued for six months might well alter the mean temperature of inland stations by 2° C., or 3.6° F., and this would make the difference between an unusually hot and an unusually cold season. Its influence in temperate zones on the length of season favorable to vegetable growth would be very noticeable, as will be more clearly shown in the following chapter.
GEOLOGICAL TEMPERATURES
It is generally believed that from the Cambrian to the Pleiocene a genial climate usually prevailed at the poles, and, moreover, without evidences of ex- traordinarily high temperatures at the equator. This state of affairs seems to be inconsistent with the view that the sun controlled the temperatures then in the same manner that it does at present.
322
THE SUN AS THE EARTH'S SOURCE OF HEAT
On the other hand, the sub-tropical glaciation L which interrupted this condition of uniform temper- atures during the Permian- period seems inconsistent with the present value of the solar constant of radia- tion. If the constant is two calories per square cen- timeter per minute, the average 2 insolation per square centimeter per minute in latitude 28° is 0.55 calories. The largest possible part of this supply would be lost for purposes of heating if the earth was completely cloudy. According to observations of Smithsonian observers this maximum possible loss, at this tropical latitude of high sun, would be less than forty-two per cent, leaving at least 0.32 calories ab- sorbed. This remainder would maintain a perfect radiator at 254.3° absolute Centigrade, or —18.7 C. If the radiating layer were not perfectly " black" it would have a higher temperature than this, and any contribution of heat from the earth's interior would also tend to raise its temperature. Under the per- fectly cloudy condition we are considering, the upper part of the cloud layer would absorb most of the ab- sorbable radiation from the sun, but its own outward radiation would be restrained, then as now, by the water vapor, carbon dioxide and ozone lying higher. Accordingly there would be a "blanket" or " hot- house" effect similar to that which now "exists, and which now raises the surface temperature of the earth nearly 30° C. above the temperature of the
1 See description quoted near the end of Chapter VI.
2 For night and day for the whole year.
323
THE SUN
radiating layer. Hence we may suppose that the upper layer of the clouds would also have been nearly 30° C. above the temperature ( — .18.7) which has just been computed for a perfectly radiating layer, mak- ing the cloud temperature about + 10° C. Accord- ing to Kirchhoff's law the earth within such a mantle would also be at the same temperature, or higher if warmed at all by internal heating. Thus it seems unlikely that a perfectly cloudy earth could have been glaciated at latitude 28° while a solar constant of 2.0 calories prevailed. If the earth was not per- fectly cloudy the conditions would have been less favorable for glaciation and more like -those of the present time.1 The matter of Permian tropical glaci- ation is still more perplexing when we consider that there was no glaciation simultaneously in temperate and polar regions.
Reverting now to the hypothesis called (A) in Chapter VI : If in Permian times and still earlier the solar radiation alone was far too little to maintain the surface of the earth or its clouds above freezing, then glacial conditions would have been produced in those times by a very moderate rise of land level. For sup- posing conduction from within and radio-activity to have been considerable sources of earth heat, and the outer layers of the clouds not greatly warmed by the sun, the thickness of the water vapor bearing stratum
1 The preceding argument does not tend to show that high tropical mountains might not be glaciated. The reasons for cold tempera- tures on high mountains have already been explained.
324
THE SUN AS THE EARTH'S SOURCE OF HEAT
would have been much less then than now, since the vertical temperature gradient of the atmosphere would have been far more rapid. Accordingly a much less degree of elevation than now would suffice to reach levels of comparatively free radiation to space, and air currents cold enough to cause snow.
On this hypothesis we may represent the contribu- tory influences which maintained the terrestrial cli-
TIME.
FIG. 65. — HYPOTHETICAL TEMPERATURE DIAGRAM.
mate in geologic time by the accompanying Fig. 65. The upper curved line indicates the earth's tempera- ture, the lower curved line what it would have been if the sun had been the sole contributor of heat to the earth. No attempt is made to draw the figure to scale either in time or temperature, but only to illus- trate the idea proposed.
The grave difficulty with our hypothesis of a low intensity of solar radiation in early geological periods is of course the question how the earth's surface tem-
325
THE SUN
perature was generally maintained above freezing. We do not, to be sure, argue that no heat at all came from the sun, but only that, while increasing, it had not in Permian times reached perhaps three-fourths its present intensity. No difficulty arises in sup- posing that in the very earliest geological times the earth's own heat sufficed completely to maintain its surface temperature. The difficulty, lies in supposing that the earth could have still contributed appreci- ably after the enormous lapse of ages, estimated roughly at 50,000,000 years, to Permian times. This difficulty appears to me insuperable.
We will turn now to hypothesis (B) stated in Chap- ter VI.1 According to Laplace's theory of the origin of the solar system we are to suppose that when the earth was formed the sun was expanded so as almost to fill the orbit of the earth. Other nebular hypothe- ses recognize the probable existence of much meteoric nebulosity in the solar system at that epoch. It is to be supposed that at that stage the sun itself was a combined structure of nebulosity and condensation, such perhaps as we see in the Pleiades stars (see Plate XIX), and was not spherical, but still its polar diameter was very much greater than now. Under such circumstances the sun, with its outlying appendages, as viewed from the earth would sub- tend a great part of a hemisphere, so that its rays would be nearly equally diffused all over the earth's surface. Such a state of affairs would have pro-
1 This hypothesis was suggested by Chamberlin about 1898. 326
THE SUN AS THE EARTH'S SOURCE OF HEAT
moted the uniformity of temperatures we have already noted.
Besides this, the 'intensity of the sun's radiation must have been very small; for otherwise, coming from so large an angle, it would have melted the earth. If the rays came to the earth from a hemi- sphere, and the sun then filled the earth's orbit, the total output of the sun to space need not have ex- ceeded half its present amount. Thus our hypothe- sis may release us from the difficulty of understanding how the sun could have radiated so long without the temperature of the earth showing marked change. For in early times we imagine the solar radiation of very slight intensity, but the angle subtended by the sun very large. With increasing solar density the angle diminished, but the output of solar radiation increased. We may even suppose considerable neb- ulosity existed all about the earth, and that this neb- ulosity, by reflecting solar rays, and by sending some long wave rays of its own, helped to diminish the rigor of the demands which geology induces us to make on the sun in ancient times. Hence our hypo- thesis (B) relieves us of the difficulty of the problem of the supply of solar energy during the enormous lapse of geological time.
As regards the possibility of tropical glaciation: We may suppose that the full maintenance of ordi- nary temperatures required formerly, as it does now, the cooperation of the blanketing effect of the water vapor of the earth's atmosphere; and that in addition 23 327
THE SUN
to this the earth's internal sources of heat were then of some appreciable importance in maintaining its surface temperature. The earlier the period we con- sider, the greater we may suppose the contribution of the earth's own heat, and the less the requirement of the sun. But we may assume that all three fac- tors, solar radiation, terrestrial conduction and the blanketing effect of the earth's atmosphere were re- quired to maintain genial temperatures in the Per- mian period. As we have assumed that solar radia- tion was nearly uniformly distributed over the earth's surface, because of the large angle subtended by the sun and of the reflection of radiation by still existing outlying nebulosity about the earth, glaciation at the tropics was accordingly no more difficult to bring on then than glaciation elsewhere. Hence, a regional ele- vation of land areas, or any other means of bringing about a reduction of the efficiency of the atmosphere as a blanket, in any locality, would have produced local glaciation. Snow and ice once formed, would help to perpetuate themselves by their high reflecting power.
There are several ways in which the efficiency of the atmosphere as a blanket may be altered. One of these is by a considerable reduction of the atmos- pheric humidity, and this, though somewhat unfav- orable to great rainfall, would still be in line with the known pronounced aridity of the Permian period. But decreased humidity generally brings with it de- creased cloudiness, which permits more solar radi-
328
THE SUN AS THE EARTH'S SOURCE OF HEAT
ation to be received, and thereby tends to raise in- stead of depress temperatures. Hence it may be that we are to look for the supposed regional altera- tion of temperature rather in a considerable increase of cloudiness, due to change in the relative arrange- ment of land and oceans.
But, in whichever of the several ways suggested, or in still others, the local decrease of temperature might have been brought about, the hypothesis (B) is evidently highly favorable to the explanation of tropical glaciation, since it makes it just as easy to produce ancient glaciations in the tropics as in polar regions. When the Pleistocene period arrived we" assume that the sun had so far shrunk that its influ- ence was then, as now, zonal. We may further sup- pose, if we choose, that the sun's radiation was less then than now, and that this combined with other causes to produce the Pleistocene glaciation. It is well known that one of these causes was a consider- able elevation of the glaciated areas.
Our hypothesis (B) seems to relieve us at once of three formidable difficulties, and enables us to under- stand: 1. How the sun has continued to suffice for terrestrial needs throughout geological time. 2. Why earlier geological periods were characterized by uni- formity of climate irrespective of latitude. (3) How it was possible to have tropical glaciation at all dur- ing the Permian and earlier epochs; but especially without evidence of simultaneous overwhelming gla- ciation over all the temperate and polar zones of the
329
THE SUN
earth. Unfortunately the most favoring foundation of hypothesis (B), namely, the Laplacian nebular hypothesis, is now strongly attacked on dynamical grounds, as we shall see in Chapter X. The substi- tute proposed by Chamberlin and Moulton seems less adapted to the line of explanation just given. For it leaves little opportunity for the development of solar heat by contraction, and besides does not permit us to assume so widely spread sources of the earth's supply of radiation in ancient times. Never- theless Chamberlin himself put forth the rudiments of hypothesis (B) about twelve years ago. We need not yet despair that a nebular hypothesis may be proposed as suitable to our purpose as to other re- quirements.1
1 A reference to See's views is nuule in ( 'lutpler X.
CHAPTER VIII
Plant Requirements. — The Assimilation of Carbon by Autotrophic Plants. — Etiolation. — Plant Geography. — Light Requirements of Plants. — Heliotropism. — Plants as Energy Accumulators.
THE vegetable kingdom varies so widely in forms, habits, and every characteristic of its members, that the reader must not expect in this chapter a discus- sion of all the sun's functions with respect to all plants. But the higher plants, such as everybody sees in the forests and fields, and which provide not only food for man and beast, but countless materials for building and the arts, are directly and indirectly dependent in many interesting ways on the sun's radiation. The subject of plant growth is so full of cases of extraordinary adaptations that it is hard to avoid digressing from the story of purely solar influ- ences to speak of some of these ; and perhaps readers may pardon a few such excursions from the main highway of our subject.
PLANT REQUIREMENTS
The higher plants require carbon, oxygen, hydro- gen, nitrogen, sulphur, phosphorus, potassium, cal- cium, magnesium, and iron. Living vegetation con-
331
THE SUN
tains a very high percentage of water; but both of its constituents, oxygen and hydrogen, also enter largely into more complex compounds with carbon. As re- gards their methods of obtaining carbon, plants are classified in three groups: (1) Autotrophic, or the self-nourishing, which obtain it through their leaves, under the influence of light, from the carbonic acid gas of the air. (2) Saprophytes, or scavengers, which take it, in part at least, through their roots from de- caying vegetable and animal organisms. (3) Para- sitic plants, which take nearly all their nourishment from living vegetation on which they fasten them- selves. We shall practically confine our attention to the first class, and when we use the term plant for short in what follows, we shall mean generally auto- trophic plants.
All plants are largely composed of water and most of them employ it profusely in their vital actions. A large birch tree, according to Von Hohnel's figures, may send into the air through its leaves in one day eighty pounds of water, which it has gathered mainly from the soil by its roots. If 200 such trees grew on an acre their water output in a season would perhaps reach 1,500 tons. While not all trees and plants are proportionally as free in using water as this, or indeed can be, still they all require it, and depend on the sun not only to keep water in the liquid form, but also to promote the atmospheric circulation which promotes the rainfall. These two functions, first maintaining a proper temperature, second inducing a sufficient
332
THE SUN'S INFLUENCE ON PLANT LIFE
rainfall, are in this age almost wholly solar. Formerly it may possibly have been otherwise.
From determinations of Konig we learn the follow- ing percentage compositions of some of the common vegetables.
TABLE XXVIII. — Composition of food products
Non-nitrog- |
||||||
Water |
Fat Ether Extract |
Nitrogenous Material |
enous (as Carbohy- |
Wood Fibre |
Ash |
|
drates) |
||||||
Wheat (grain) . . |
13.65 |
1.75 |
12.35 |
67.91 |
2.53 |
1.81 |
Potato tubers . . |
75.48 |
0.15 |
1.95 |
20.69 |
0.75 |
0.98 |
Beetroot |
87.61 |
0.11 |
1.09 |
9.26 |
0.98 |
0 95 |
Lettuce (leaves) |
94.33 |
0.31 |
1.41 |
2.19 |
0.73 |
1.03 |
The various chemical substances mentioned above as plant requirements, and also some others, occur in weak solution in the water which plants so plenti- fully absorb through their roots. We cannot enter into that profoundly interesting and difficult question how this fluid rises to the tops of such immense trees as the Sequoia and the Eucalyptus, which sometimes reach heights of 500 feet, and in which the action of gravity would tend to produce outward pressures within the roots of fifteen atmospheres. Suffice it to say that in some manner the fluids obtained from the ground do reach all parts of the plant, and the water copiously passing through the leaves evaporates. This is called transpiration. The carbonic acid of the air entering the leaves during light action in a manner to be described later, is altered and combined with the various elements transported from the roots. Com-
333
THE SUN
plex nourishing compounds produced in the leaves descend to all living cells of the stem and roots, and after undergoing further transformations even re- ascend in spring to start the new growth of shoots, leaves, and buds.
The various elements are not available to the plant in all their chemical combinations, and in some com- binations may even be poisonous. Without going into details, it will be interesting to note the case of nitrogen. This element, found uncombined as a gas in air, is rather inert chemically, and none of the higher plants seem able to make use of it in its free state. Ammonia, too, though prevalent as a product of decay in the soil, and existing also in the air, is not nourishing to most of the higher plants. Nitrites are said to be poisonous in moderate concentration, al- though in very dilute solution perhaps useful. Ni- trates, then, are to be regarded as the principal nitro- gen sources for autotrophic plants. In agriculture the removal of crops withdraws available nitrogen from the soil faster than ordinary processes can pro- duce it, hence the use of fertilizers containing salt- peter. But the leguminous plants, such as peas, beans, clover, alfalfa, etc., are said to be able to use free nitrogen, and it is customary to plow under green crops of such nature to improve the soil. Careful researches have shown that in reality certain micro- organisms, often present in the soil, cause the forma- tion of nodules on the roots of these leguminous plants, and that atmospheric nitrogen is only assimi-
334
THE SUN'S INFLUENCE ON PLANT LIFE
lated when these micro-organisms inhabit the nod- ules. Different Leguminosce require different species of micro-organisms for a successful partnership of this kind. The micro-organisms combine the free nitrogen into forms useful to the plant, and the plants supply other materials, perhaps carbohydrates, for the micro-organisms. This is but one of many in- stances in which the higher plant forms are proved to depend upon the activities of the lower, quite as much as the lower on the higher. Quite recently it has be- come possible to purchase at large seed stores cultures of the proper micro-organisms, with instructions for multiplying them, so that when sowing a field with clover or alfalfa, for instance, the cultures may be mixed with the seed so that it may be certain that the crop will not suffer from lack of nitrogen.
THE ASSIMILATION OF CARBON BY AUTOTROPHIC PLANTS
Many plants, among which are corn and others of the most valuable food plants, will thrive in water cultures as well as in the soil, although the supply of carbon through their roots is made impossible. Hence the source of the carbon which is a fundamental ele- ment of all organic life must be, in such cases, the air. If grown in darkness, although in all other respects the conditions are retained identical, no considerable gain of carbon occurs and the plants remain white, for no chlorophyll is formed. We find, then, that carbon dioxide of the air is taken in under the influ-
335
THE SUN
ence of light, and is acted upon to form complex com- pounds with hydrogen and oxygen, such as hexose, sugar, starch, and also the nitrogenous carbon com- pounds which go to compose vegetation. This ac- tion is found to require the green chlorophyll bodies of the living plant cells, and chlorophyll, as we have said, is not produced without light. Oxygen is given off in the chemical transformations and escapes from the leaves. The process of absorbing carbonic acid and transforming it with evolution of oxygen, as just described, is called assimilation.
The evolution of oxygen may be demonstrated in a simple way by cutting a branch of the water plant Elodea Canadensis and placing it in a tube of water charged with carbonic acid gas. If retained in very dim light for some time nothing easily noteworthy occurs, but when well illuminated a stream of bub- bles escapes from the cut ends. By inverting a test tube, previously filled with water, the gas may be col- lected, and will be found by testing it with a glowing coal to consist largely of oxygen. Quantitative experi- ments have been made by counting the bubbles given off in this manner, and it has thus been shown that their number is usually nearly proportional to the in- tensity of the light. In darkness no oxygen is given off, but carbon dioxide is slowly evolved instead. This reverse process is called respiration.
As already stated, an essential condition for carbon assimilation is the presence in the plant of the green substance of chlorophyll. This is found in most
336
THE SUN'S INFLUENCE ON PLANT LIFE
plants almost solely in the leaves, so that these are the great organs of assimilation. Chlorophyll in alco- holic solution has a fine fluorescence. It appears green by transmitted light and red by reflected light. The spectrum of crude chlorophyll in alcoholic solu- tion is characterized by six absorption bands. Three are in the violet and merge together in strong chloro- phyll solutions. The other three occur in the green, yellow, and red respectively. By treating the alco- holic solution with benzine the crude chlorophyll, which is a mixture, will yield in benzine solution a blue green dye which seems to be the more important component. This itself is complex, and contains among other constituents one which is called phyllo- porphyrin, and differs only a very little chemically from the hsematoporphyrin of blood. But however curious and interesting chlorophyll may be, its spe- cial function, the promotion of carbon assimilation, does not go on except the chlorophyll be in the living plant cells. Artificial chlorophyll bearing cells will not answer.
It has been shown that for every volume of carbon dioxide operated on by the plant, an equal volume of oxygen is liberated. Among the principal products of the reaction is glucose or starch. Such facts may imply some such actions as are expressed in the fol- lowing symbolic manner.
6CO2 + 6H2 O = Cn H12 O6 (Glucose) + 602;
or 6CO2 + 5H2 O = C6 H10 O6 (Starch) + 602.
337
THE SUN
Starch is readily demonstrated as being produced in many plants during light action, but plants of dif- ferent families vary greatly in the quantity of it they produce. Indeed, as we shall see many times, the different plants behave so differently under given conditions that hardly a single general fact can be stated, in regard to which some kinds do not exhibit exceptions. As one person is repelled by coaxing and
moved by argu- ed ment, while another goes only as senti- ment dictates, so the plants seem to have their diverse characters, and two kinds may react oppositely to the same stimuli.
The organs of carbon assimilation are the leaves, and in these the portals of access are the little open- ings called stomata. These exist in most plants most plentifully on the under surfaces of the leaves, al- though found in some only on the upper surfaces and in still others on both. They are very minute slit- like orifices, so small that a needle prick is a huge hole in comparison with one of them. A single leaf of a sunflower may have no less than 13,000,000 stomata. Fig. 66 (after Schwendener) gives a general idea of
B
FIG. 66.— STOMATA. (Schwericlener.) A, Cross-section on tn, n. B, Plan view of half-stoma omitting parts outside a, 6. C and D, Closed and open stomata.
THE SUN'S INFLUENCE ON PLANT LIFE
the form and surroundings of these minute but nec- essary organs in the Amaryllis. There are special con- trivances called guard cells adapted for opening and closing the stomata. These guard cells, when dis- tended by containing much liquid, or when shone upon by strong light, cause the stomata to open wide, thus promoting the assimilation of carbon dioxide, and also the transpiration of water vapor, of which we shall speak later.
Although very numerous, the combined area of the stomata, even when wide open, is hardly more than one per cent of the area of the leaves, so that it was long a mystery how so much carbon dioxide could pass through them. This question was solved by Brown and Escombe (1900). They found that when carbon dioxide is admitted through an orifice to a medium capable of absorbing the gas as fast as re- ceived, the amount which diffuses through the open- ing decreases with the diameter, not with the area of the opening. This seeming paradox is explained by supposing the velocity of flow to increase as the open- ing decreases, so that a smaller hole accommodates the diffusion not merely from directly above, but also from the side areas which were before served by the larger hole. The observers found that their strange new law held for numerous openings as well as for single ones, provided the openings are separated by distances as great as eight or ten times the diameter of the holes. From this it follows that a surface pierced, like a leaf, with extremely numerous but very
339
THE SUN
small openings can admit the passage of a gas by dif- fusion at almost as rapid a rate as if the whole area of the surface were one hole. This extraordinary dis- covery raises our admiration of this excellent con- trivance of Nature, whereby the whole area of the leaves of a plant is as if available to transmit nour- ishment from the air, and to permit the escape of water vapor, although in reality nearly all of this area is actually closed, to protect the delicate cells within.
The rate of assimilation of carbon, or what is al- most proportional to it, the rate of gain in dry weight of a plant, depends on various factors. Of these we may mention first the concentration of the carbon dioxide in the air. Although, according to Eber- meyer's estimates, a square mile of forest uses up over 500 tons of carbon dioxide in a year, so that the demands of the plant life of the world are really enormous, there is nevertheless an almost steady, and everywhere nearly uniform, percentage of car- bon-dioxide in the air — about three parts in 10,000. The steady drain of plant life is to be set over against the production of carbon dioxide by the res- piration of animals, the burning of wood and coal and other sources of supply, but it is surprising that the atmospheric proportion remains so nearly uni- form as it does. Geologists are by no means of the opinion that this proportion has always been the same as at present. It is therefore of interest to inquire how the assimilation varies with the concentration of car-
340
THE SUN'S INFLUENCE ON PLANT LIFE
bon dioxide. There seems to be some disagreement between investigators as to the precise optimum con- centration, but all are agreed that the rate of assim- ilation of carbon increases steadily with the concen- tration of carbon dioxide up to a concentration of at least more than ten fold the present. Under such con- centrations the rate of assimilation may reach more than twice its usual value. According to some, the increase of assimilation is even directly proportional to the increased concentration of CO2, within these limits. While, of course, this change has no practical interest while the carbon dioxide of the air remains constant, yet it may have been of considerable prac- tical importance to vegetation in past geological epochs when the air was more highly charged.
Temperature is a still more important agency in regulating the growth of plants. Assimilation may be recognized with some plants at temperatures of several degrees below freezing, but practically speaking all growing plants of the higher forms must be maintained at temperatures between 0° and 50° C. The increase of the rate of assimilation for most plants is very rapid from 0° up to a temperature of about 35°, and at higher temperatures than this there is a still more rapid decrease in the rate. It is an interesting question whether the principal forms of vegetation could flourish on any planet if the mean temperature lay below 0°, or above 50°. Although we cannot answer this question absolutely, still it seems probable that the answer must be in the negative.
341
THE SUN
At all events, we see not only how entirely our own lives now depend on the sun, but even on that nice balance between the receipt of solar radiation and the emission of terrestrial radiation, in which even the amounts of water vapor and cloudiness are important, as stated in the preceding chapter.
We now take up the dependence of carbon assimi- lation on light, deferring the consideration of other effects of light on growth. Plants raised in darkness do not become green. The formation of chlorophyll and the assimilation of carbon require radiation be- tween wave lengths 0.39/t and 0.77/4. Experiments on the relative effectiveness of rays of different wave lengths are not altogether satisfactory. They have been confined to a few kinds of plants, and great dif- ficulty is found here, as in physics and astronomical work, in separating a sufficiently intense nearly monochromatic beam of light, and in measuring its intensity. Investigations were made about thirty years ago on the relative efficiency of the different rays by Reinke and by Engelmann. They agree in fixing the wave length of maximum effect in the red at about 0.65//, to 0.70/A, but Engelmann found a secon- dary maximum in the blue at 0.48/-6, not found by Reinke. Engelmann's observations distinguish be- tween the assimilation of the upper and lower sides of a leaf capable of such action, and he finds the posi- tion of maximum effect shifted distinctly towards shorter wave lengths for the surface which receives its illumination through the leaf. This result de-
342
THE SUN'S INFLUENCE ON PLANT LIFE
pends, it is thought, on the strong absorption by chlor- ophyll for red rays, for thereby the light which pene- trates the leaf is greatly weakened in its longer waves. Undoubtedly the relative activity of different wave lengths of light in promoting the assimilation of car- bon is closely associated with the absorption spectrum of chlorophyll, as indeed would appear from Engel-
042
6*50
$66
FIG. 67. — PROMOTION OF CARBON ASSIMILATION BY LIGHT (full curve) AND ABSORPTION OF LIGHT (dotted curve) IN GREEN LEAVES. (Engelmann.)
mann's results, given in Fig. 67. There is needed much more research in this difficult field. It would be greatly promoted by the introduction of means for obtaining nearly monochromatic light of well deter- mined and adequate intensity, covering considerable areas suitable for plant growth under otherwise nat- ural conditions.
24 343
THE SUN
Kniep and Minder l have recently made observa- tions by the bubble method with Elodea Canadensis on the assimilation of carbon dioxide in lights of dif- ferent colors. They used sunlight filtered by colored solutions so as to select red light (wave length 0.62/4 to some point in the infra-red not determined), green light (wave length 0.512/t to 0.524/*), or blue light (wave length 0.35/1, to 0.50//,) at pleasure. In each case the light could be reduced to a fixed intensity as meas- ured by a Rubens thermopile, so that they could in- vestigate the rates of assimilation under equal inten- sities of total radiation for each of the three colors. They found the green light of no effect in producing assimilation. It was like no light at all. The red and the blue they found equally effective. Hence their experiments tend to support Engelmann's, as they indicate two wave-length regions efficient to pro- duce assimilation. We must wait for more elaborate experiments before we shall know just how the effi- ciency varies with the wave length, and whether all plants are best promoted by the same rays. It is clear, however, that as the red end of the spectrum predominates in direct sunlight at the earth's surface, whereas the violet end greatly predominates in sky light, a plant may be made to assimilate carbon pre- dominatingly by red or blue light according to whether it grows in direct sunshine or not. This may offer a method of evolving new plant forms as we shall see in the next section.
1 Zeitschrift fur Botanik, vol. i, pp. 619-650, 1909. 344
THE SUN'S INFLUENCE ON PLANT LIFE
Experiments have been made on the dependence assimilation on the total intensity of light irrespective of wave length. Results very naturally differ for plants of light-loving and shade-loving habits. In general, the rate of carbon assimilation increases nearly directly proportionally to the intensity of the light, but this ratio of course cannot persist for ex- tremely high intensities, because, first, of injury to the plant, and, second, of deficiency of other promot- ing elements, especially carbon dioxide.
ETIOLATION, OR EFFECTS OF DEFICIENCY OF LIGHT
Plants grown in darkness or weak light tend to have long stem internodes and leaf stems, and small white or yellow leaves. These and other effects of deficiency of light on plant growth are termed etiola- tion. As already stated, the higher plants do not in- crease much in dry weight unless exposed to light, so that experiments on the effects of complete darkness on growth are mainly restricted to such species as have large food stores in their seeds or tubers. The object served by natural etiolation is at length to bring the leaves of the plant to suitable illumination, as is seen by the tall tree stems and climbing vines in closely growing forests. In experimental work this result cannot, of course, be reached, but nevertheless the tendency is plainly shown.
Although, as stated above, the effect of- darkness or very weak illumination is to restrict the leaf area, leaves grown in moderate light are larger and thinner
345
THE SUN
than those grown in full sun. This form of etiolation is of importance to the tobacco industry, since large thin leaves are preferred and command much higher prices. Accordingly, in Connecticut and Florida large fields of tobacco are now grown shaded by tents of open-meshed cloth to promote this improvement. Other desirable results obtained by this device con- sist in the greater and more uniform humidity of the air and soil and the prevention of disastrous winds and hail.
With many kinds of plants the buds will not de- velop if the light is too weak, and there are besides many other effects embraced by the general term etiolation. Curiously enough red light, which, as we have seen, is highly effective for promoting carbon assimilation, in many cases behaves like darkness in respect to etiolation. It is thus possible to grow plants under conditions favorable to their adequate nourishment, and at the same time to greatly alter their forms by etiolation effects. This interesting feature perhaps offers opportunities for promoting the evolution of desired forms in useful plants.
PLANT GEOGRAPHY
In natural surroundings there is a very great range of light intensity, and with it a great range in temper- ature and moisture. These circumstances produce very marked effects on plant life. In the tropics abound regions of great rainfall, from 100 to 500 cen- timeters (40 to 200 inches) annually, with the aver-
346
THE SUN'S INFLUENCE ON PLANT LIFE
age cloudiness as high as fifty to sixty per cent. The mean temperature being also high, 25° to 28° C., there results a high atmospheric humidity. Such re- gions are the home of the tropical rain forest, which, viewed from a ship presents a noticeably irregular skyline filled with every shade of green, but tending toward the more sombre hues. Flowering trees are occasionally conspicuous. The interior of such a forest teems with a varied mass of vegetation from the ground to the top of the highest trees. Vines and rugged ferns abound, so that the traveler's way is almost impassable. Fruits and flowers are plentiful. Parasitic and saprophytic plants revel among the luxurious surroundings. On viewing a tall tree one can scarcely distinguish which is its own foliage and which that of the dependent vines and parasites that load its trunk and limbs almost to breaking.
Sharply contrasting with such scenes as this are the sub-tropical deserts like the African Sahara. Here also the temperature is high, but variable, rang- ing perhaps 20° C. in a single day. Rainfall may be as slight as 5 centimeters annually, but more often reaches 10 to 20 centimeters. The scanty vegetation is provided with extraordinary contrivances to re- duce as far as possible the loss of water by transpira- tion. Leaves are small, thick, glossy and waxy, their stomata protected heavily. Thorns abound. The roots run very deep so that even at one or two meters in depth they have hardly diminished at all in size. As a rule only small plants and shrubs are found.
347
THE SUN
Some varieties have special reservoirs for the storage of water.
The periods of rest are not conspicuous in the trop- ical vegetation, for they are not governed by temper- ature, though perhaps often by rainfall. Tropical rest periods are frequently localized to single trees or parts of trees; but in the temperate and arctic zones there occurs in winter manifestly a general cessation of growth. Not all temperate and arctic trees cast their leaves, but they generally rest from the growth of shoots during the cold months. Askenasy has in- vestigated these matters at Heidelberg for the gean tree, which may serve as a type for other broad- leaved trees. The season of activity lasts from about mid-April to mid-October. It comprises the period of growth of foliage, April-May, during which next season's foliage buds appear; then, May-September, follows the period of assimilation during which stems and roots enlarge and the next season's flower buds are formed; then comes the period of decline ending in the fall of leaves. During the summer the growth of next season's buds is slow, and ceases altogether from October to early February. Then a growth begins and becomes more and more rapid. Although a warm March greatly accelerates the development, a warm October cannot start growth. From the end of November, development may be forced by hot- house conditions. During the rest period chemical changes of the reserve material go on, and it is indeed transported between different organs of the tree,
348
THE SUN'S INFLUENCE ON PLANT LIFE
Temperate forests contrast with the tropical rain forest described above in the relative absence of vines, parasitic vegetation, and undergrowth, although in moist regions herbs and shrubs are not lacking. The evergreen conifers are more and more in evidence at higher latitudes, but these become dwarfed toward the arctic zones. The growth period of arctic flora is limited to about two months, but is favored by the fact that the sun is then continually above the hori- zon. All varieties start their growing almost simul- taneously, and reach their flowering stage almost together, within a couple of weeks. Although the mean temperature of the air during the growing pe- riod may be 5° C. or more, the soil is frozen almost to the surface.
Wiesner has made extensive photographic re- searches to. determine the light requirements of plants. He employs a modification of the method of Bunsen and Roscoe. A normal photographic paper is prepared by soaking in three per cent common salt solution, drying in darkness, soaking five minutes in twelve per cent silver nitrate solution, and drying again in darkress. A normal tone or grade of dark- ening is prepared by coating a paper with a mixture of one part lampblack in 1,000 parts zinc oxide. When the photographic paper reaches this shade by expos- ure to light for one second, the light is said to be of unit intensity of the Bunsen-Roscoe scale. Those au- thors showed that for equal blackening of the photo- graphic paper the intensity of the light, between wide
349
THE SUN
limits, is inversely as the time required. Hence if n seconds are required to produce normal tone, the light intensity is 1 / n Bunsen-Roscoe unit. To avoid inconveniently long exposures in deeply shaded places, and to allow sufficient time for accurate re- sults in strong light, Wiesner introduced a gradation of shades, forming a kind of tone scale, which he standardized in terms of the normal tone.
By such procedures Wiesner has measured the light action due to direct sunlight, and to diffused light at Buitenzorg (Java), Cairo (Egypt), Vienna (Austria), several stations in Norway, and Advent Bay (Spitzbergen) . His measures were made on days varying in brightness from cloudlessness to rain and snow. The Vienna measurements extend for several years. He has made observations in the open, in leafy tree crowns, and under the shadow of thick for- ests. It is not possible to give here any adequate summary of this very extensive work, but the reader may consult the original articles of Wiesner.1
Some of Wiesner's results are as follows: The max- imum total illumination at Vienna was 1.50 B-R units; at Buitenzorg, 1.61. At Vienna the mean midday value ranges from 0.1 B-R unit in January to 0.96 in July. At Buitenzorg in December and Janu- ary the midday values range from 0.65 to 0.85. Rain or snow diminishes the light total to one-tenth or less
1 Especially in Sitzungsberichte Wien. Akad. Math. Naturw. KL, Bd. 102, I, 1893; 104, I, 1895; 109, I, 1900; 113, I, 1904. Also Denkschriften of the same Academy, Bd. 64, 1897; 67, 1899.
350
THE SUN'S INFLUENCE ON PLANT LIFE
of its normal value. At Vienna the ratio of direct sunlight to diffused skylight action is very variable, but an average value for several hours near midday is about unity. On half cloudy days the total light action is almost as strong as on cloudless days. On completely cloudy, but not stormy, days the total light action is reduced from three- to five-fold.
Taking the total direct and diffused light action in the open as the basis of reckoning, Wiesner compares with it the light action found in the crowns of trees and elsewhere. Calling the first value I, the second i, he calls the ratio (i/\ = L) the relative photic ration. When the leaves are beginning to form in spring, be- fore they get large enough to cast deep shadows, the values of L within tree crowns and under trees are not greatly less than unity. But later in the summer, when leaves are full grown, and next season's leaf buds are forming, these ratios become much smaller. For instance for the white birch (Betula alba), Wiesner finds:
Date |
Observed values of I |
L 1 Da-v>s 1 **t Minimum J |
|
Total daylight |
In tree crown |
||
April 16 |
0.834 |
0.333 |
1 2.5 |
May 1 |
0.875 |
0.219 |
1 4 |
May 14 |
1.122 |
0.142 |
1 8 |
May 29 |
1.200 |
0.109 |
11 |
351
THE SUN
This rapid increase of the shading action of forest trees, as they develop their leaves, determines the nature and habit of the underbrush. Generally the leaves of the underbrush present a scattered, or flat array, so as not to shade one another. Often the undergrowth has the habit of rapid development of leaves and blossoms in early spring, before the over- growth is fully leaved.
In arctic regions the vegetation, almost without exception, requires practically all the available light. This depends, no doubt, on the coldness and short- ness of the season of growth. Values of L much below unity seem to be insufficient for arctic plants. This may explain the absence of tree forms there. Whereas in the tropics, and even in temperate zones, most plants have means for reducing the light action on their leaves, no such contrivances are common in the frigid zones.
The range of light requirements is indicated by the following values of the relative photic ration and total light action within the crowns of trees in full leaf. The terms (Max.) and (Min.) refer to the max- imum and minimum daily values of the quantities concerned.
Among underbrush growing in a shade so deep that
L = — he found beeches, maples, and other well
Oo
growing saplings. Grasses in the temperate zones
were found, although not blooming, when L = .
oy
352
THE SUN'S INFLUENCE ON PLANT LIFE
Common 'name |
L (Min.) |
I (Max.) |
Remarks |
Boxwood |
I |
0.012 |
• |
Beech |
108 1 |
0 015 |
Isolated tree |
Maple |
85 1 |
0.030 |
Isolated tree |
White poplar |
43 1 |
0.086 |
Isolated tree |
Pine White birch |
15 1 11 1 |
0.118 0 144 |
Enclosed tree Enclosed tree |
Ash |
9 1 |
0 224 |
Enclosed tree |
Larch |
5.8 1 |
0.260 |
Isolated tree |
Blackthorn |
5 1 1.3 |
0.722 |
Blooming but not leaved |
Some tropical grasses survive L =
1
100'
Lichens
were found in the tropics which had the photic ration only L = — -. Many forms of tropical or-
chids, epiphytes, and other shade-loving plants, were
found to thrive under photic rations from — to — .
10 ou
We cannot dwell longer on the interesting work of Wiesner. From it we see how unnecessary it is, for many forms of vegetation, that the light should be of the full intensity which is now available in the open. Indeed, Wiesner remarks that in experiments made by rotating plants, so as to get equalized illumination, the buds will develop and leaves be fully grown under illuminations far below the minima observed under
353
THE SUN
natural conditions. In the natural state the well- illuminated buds grow at the expense of their less favored neighbors, and as their leaves expand they tend still further to suppress the undeveloped buds. In view of all this, and in view of the hypothesis (B) advanced in Chapters VI and VII, which treated of a more uniform illumination assumed to be formerly prevailing, it is interesting to speculate whether the great vegetation of the Carboniferous era was not produced under a far more feeble illumination than that which now prevails.
Considering the present lack of exact experiments on the efficiency of different wave lengths of light to promote plant growth, the photographic experiments of Wiesner are perhaps all that are yet demanded. But we can easily see the advantage which would re- sult to plant physiology if such an instrument as the the spectrobolometer could be employed in skilled hands to determine the relations of wave length and intensity of light to carbon assimilation and etiola- tion, for numerous plant forms.
HELIOTROPISM
It is well known that different plants vary as re- gards the angles which their organs present to the direction of strongest light. A nasturtium, for in- stance, if principally illuminated from one direction, will expose almost every leaf and flower with its face broadside toward the light. Plants within a room bend toward the window. Some species which live
354
THE SUN'S INFLUENCE ON PLANT LIFE
in dry and cloudless regions present their leaves edge- wise to the strongest illumination. Such adaptations as those we have mentioned, and others, are em- braced under the term heliotropisrn. Different plant organs differ in respect to the matter, so that botan- ists distinguish organs which are orthotropic, and those which are plagiotropic, according as they tend to lie in the direction of the principal light or at some other angle with respect to it. Also orthotropic organs may grow in a positive heliotropic manner, i. e. towards the source of light, or the contrary. Roots are usually negatively and stems positively orthotropic, while leaves may be regarded as plagio- tropic.
It was supposed by De Candolle (1832) that helio- tropisrn was a simple consequence of different rates of growth between strongly and weakly illuminated parts of an organ. It had been found (as already stated under etiolation) that stems grown in darkness exceeded in length those grown in light. Further- more, it has been shown that plants increase in stat- ure faster by night than by day. See, for instance, the following measurements by Kraus on the growth of a species of bamboo at Buitenzorg, Java, in twelve-hour intervals :
Date |
Dec. 4 |
Dec. 5 |
Dec. 6 |
Dec. 7 |
Dec. 8 |
|
Growth by day .... |
10.5 cm. |
4.5 cm. |
8cm. |
8.5 cm. |
12cm. |
|
Growth by night . . . |
16cm. |
15 cm. |
16 cm. |
12.5cm. |
355
THE SUN
On such grounds De Candolle assumed that helio- tropic curvature was simply the effect of the retard- ing influence of light on the growth of that side of the stem most strongly illuminated. This simple ex- planation may have some justification, but it is not adequate to explain the facts. For plant organs which curve away from the light also grow faster in the dark. Furthermore, the same organ may react either positively or negatively or not at all according to the intensity of the light, as shown by the experi- ments of Oltmans. This author is of the opinion that the best intensity of illumination for the general welfare of the organism is that at which it exhibits no heliotropic curvature. Direct sunlight is too bright to induce heliotropic curvature in most plants, hence they do not as a rule turn their leaves from east to west with the progress of the sun, although in the case of the sunflower this occurs with the blossoms.
It seems that illumination acts rather as a stimulus than as a force in producing heliotropism, for the effect may be produced by brief light action and the actual curvature take place in the appropriate direc- tion after the light has been withdrawn. Further- more, the reaction does not necessarily occur where the light is applied, but the stimulus may be trans- mitted some distance from the sensitive recipient organ to the position where curvature takes place, although the part of the organ where it becomes curved is shielded entirely from the action of the light.
356
THE SUN'S INFLUENCE ON PLANT LIFE
Heliotropism is without doubt of great value to plants in enabling them to adjust their leaves most advantageously to increase or reduce the illumina- tion in which they find themselves. It is especially valuable to many compound leaved plants subjected to the powerful heating effects of the direct rays of the unclouded sun. They open out their leaves in the early morning or during cloudy weather, but tilt them up edgewise in the hot sun, thus reducing the effective area for heating. Such plants, though of large leaf area, may thrive in the driest regions. Other plants have their leaves permanently set at such an angle as to receive a minimum of direct sun- light.
On the other hand many plants growing in com- paratively weak light, and some sun-loving plants in the open, turn the broad sides of their leaves toward the strongest light. The negative heliotropism of the roots of plants is of advantage, for by it they may be saved from growing out of the soil.
PLANTS AS ENERGY ACCUMULATORS
The energy now available in coal and oil was for- tunately preserved for our use in the decomposed veg- etation of former ages. Extraordinary luxuriance of vegetation is thought to have prevailed in those ancient times, and we now use the accumulated en- ergy of solar rays emitted long before the existence of man. Attempts to employ solar energy by arti- ficial engines will be referred to in the next chapter,
357
THE SUN
but none of them can as yet be compared in econom- ical success with the natural process of storage always occurring in vegetable growth. Artificial processes are for the moment far more efficient, but not in pro- portion to their great cost, and none of them, like the natural process, stores permanently the energy re- ceived. Most solar engines transform solar radia- tion immediately into heat, and this is gradually lost. Growth transforms solar radiation immediately into chemical energy, and this may be indefinitely pre- served.
Various estimates have been made of the efficiency of plants as transformers of energy. Pfeffer (1871) computed from Boussingault's work that a square meter of Nerium leaf surface formed starch at the rate of 0.000535 grams per second. Assuming the product formed to have a heat of combustion of 4,100 calories per gram, he found 2.2 calories per square meter per second to be the amount of energy eon- served. The amount of energy received from the sun would depend on the time of day, inclinations of the leaves, moisture of the air, etc., but might be esti- mated at about 150 calories per square meter per sec- ond in ordinary conditions near sea-level. This would give an efficiency of about 1.5 per cent.
Brown, in his Bakerian Lecture (see Nature, vol. Ixxi, p. 522), summarized some careful experiments on the efficiency of the sunflower. He made estimates of the temperatures of leaf surfaces and of their thermal emissivity. The latter in still air was about 0.015
358
THE SUN'S INFLUENCE ON PLANT LIFE
calories per square centimeter of leaf surface for 1° C. difference of temperature from the surroundings. Leaves evolve carbon dioxide in darkness in their or- dinary process of respiration. For a sunflower leaf respiring 0.7 cubic centimeters of carbon dioxide per 100 square centimeters per hour, the respiration causes a rise of temperature of the leaf in still air of 0°.019 C. above its surroundings. This effect is therefore practically negligible. Not so the effect of transpiration and evaporation of water, especially in windy surroundings, for this may alter the temper- ature of leaves by several degrees. The absorption coefficients of leaves of various plants in ordinary sunlight were determined. These range from sixty- five to seventy-eight per cent, and for the sunflower leaf was found to be 68.6 per cent. Such values would probably differ according to the quality of the light. The rate of absorption of carbon dioxide by the plants was measured. Air was drawn through the glazed case containing the leaf specimen and the carbon dioxide contents of the air after passage was compared with that of air unaffected by the plant. Various concentrations of carbon dioxide were experi- mented upon, and it appeared that up to concentra- tions six times the normal, the rate of assimilation was proportional to the concentration of carbon diox- ide in the air. The material formed by the plant was assumed to be hexose, whose heat of combustion is 3,760 calories per gram. The rate of assimilation seemed to be independent of the intensity of the 25 359
THE SUN light, until this was reduced as low as 0.04 calories
per square centimeter per minute, or say -— of or-
Zo
dinary sunlight. Hence the efficiency of the plant appeared much higher under weak illumination. In some cases the efficiency found was as high as five per cent, but not often above 1.7 per cent.
Two numerical illustrations will show the charac- ter of the results. Both deal with the sunflower leaf (Helianthus annuus). In the first case the activity of the leaf did not suffice to expend all the solar energy it absorbed and the leaf was above the temperature of the surroundings. In the second case, owing to the high temperature, the fully opened stomata, and the low humidity prevailing, there was rapid transpira- tion and the contrary state existed. The numbers given in the first part of Table XXIX apply to the energy reaction per square centimeter of leaf surface per minute. In the latter part of the table is given the disposal of the leaf in percentages of the solar energy received, plus the heat energy received from the surroundings.
It appears from such investigations as have been made that plants may store up as chemical energy in round numbers one or two per cent of the energy of solar radiation which shines upon their leaves. This may seem a very small efficiency, but on its results accumulated through former ages have depended the great manufacturing achievements and the comfort- able winter warmth of our dwellings for many years.
360
THE SUN'S INFLUENCE ON PLANT LIFE
TABLE XXIX. — Economy of Helianthus annuus
Case A |
Case B |
|
Total solar radiation received Amount absorbed Amount of energy used vaporizing water |
0.2569cal. 0.1762 0 1243 |
0.2746cal. 0.1884 0 3668 |
Amount of energy used in photo-syn- thesis |
0 0017 |
0 0033 |
Amount of energy lost by cooling Velocity of wind in meters per minute. . Temperature of leaf above surround- ings |
+0.0502 428 +0°.43C. |
-0.1817 200 -1°.84C. |
Energy used in photo-synthesis Energy used in transpiration |
per cent. 0.66 48.39 |
per cent. 0.72 80.38 |
Solar energy transmitted by leaf |
31 40 |
18 90 |
Heat energy lost to surroundings |
19.55 |
In the combination of water power and electricity we seem now to be passing in a measure away from the dependence on coal and steam, but there is little ques- tion that both coal and oil will long remain in exten- sive use to remind us of our dependence on the growth of ancient vegetation and its transformation of solar radiation into chemical energy.
CHAPTER IX
UTILIZING SOLAR ENERGY
Experiments with Burning Mirrors. — The "Hot-box" Principle. — Mouchot, Pifre, and Ericsson. — Eneas' Solar Engines. — Proper- ties of Glass. — Solar Heaters and Reservoirs. — Low Temperature Solar Engines. — Solar Cooking Appliances. — Solar Metallurgy. — Resume*. — • Quantity of Solar Energy Available. — Thermo- dynamic Efficiency. — Reflecting Powers of Mirror Surfaces.
AT present the manufacturing and commerce of the world is mainly carried on by aid of coal or internal combustion engines, which derive their fuel from the decomposed products of prehistoric masses of vege- tation in which were stored a small fraction of the solar energy of those bygone times. The modern great development of water power, electrically util- ized, also depends on the sun; for by solar heating water is evaporated from oceans, lakes, rivers, and the soil, is transmitted inland and precipitated by the atmospheric circulation which the sun's heat main- tains, and comes in use when it flows down in the rivers. Another immense source of water power, not as yet much utilized, resides in the ocean waves and tides, which also depend in a high degree upon the sun. It is not necessary to discuss further these well- known sources of power, and we shall pass to the vari- ous means which have been proposed for using the energy of the solar rays more directly.
UTILIZING SOLAR ENERGY
EXPERIMENTS WITH BURNING MIRRORS
It is said that during a siege of Syracuse in the year 214 B. c. the renowned philosopher Archimedes burned or scattered the Roman fleet under Marcellus by concentrating sun rays upon the ships by means of mirrors erected on the shore. Whatever may be the truth of this story, which has been doubted, such means of warfare are not likely to be revived in our day.
Buffon, the French naturalist (1707-1788), tested the possibility of the circumstance just described. In 1747 he made many experiments with a burning mir- ror constructed by mounting 360 plane glass mirrors, each 16 X 22 centimeters, on a frame in such a man- ner that each could be adjusted separately, so that all could concentrate their reflected rays to a focus at any desired distance. Corresponding to the angular diameter of the sun, the focus was about 44 centi- meters in diameter at 50 meters, and proportionately less at shorter focal distances. He found it possible to set fire to wood at 68 meters. With 45 mirrors he melted 3 kilograms of tin in a pot, at 6.5 meters, and with 117 mirrors melted silver at the same distance. By these experiments he showed the possibility of the feat of war attributed to Archimedes.
In 1755 Hoesen, a mechanician of Dresden, began to build up mirrors of paraboloidal curvature. One of these was over 3 meters in diameter, and so well made as to concentrate the sun's rays to a focus 1.3 centimeters in diameter. With one of Hoesen's mir-
363
THE SUN
rors of half this diameter Wolf reduced many metallic ores, and melted coins almost instantly.
THE " HOT-BOX" PRINCIPLE De Saussure (1740-1799), the Swiss naturalist, made five half cubes of glass of such sizes as to go one within the other with some air space between. These rested inverted on a blackened table non-conductive of heat. Thermometers were placed between the vessels and in the air outside. The one between the fourth and fifth vessels showed the highest tempera- ture, 87.5° C. In later experiments with glass-covered vessels he protected the sides and back of the vessel from cooling by wrapping it with non-conducting material. When the vessel was exposed to the sun perpendicularly he observed on one occasion a tem- perature of 110° C. within. In one experiment he heated the surrounding medium, keeping its temper- ature just below the inside temperature, and thereby practically prevented loss of heat, except through the front. In this manner he obtained a temperature within of 160° C. His experiments convinced him that two, or at most three, sheets of glass over such a hot box are better than more. He made some essays at cooking with such devices.
Sir John Herschel describes the following experi- ments made during his sojourn at the Cape of Good Hope, 1834-1838.1
1 "Results of Astronomical Observations ... at the Cape of Good Hope," etc., by Sir John F. W. Hersrhol, Bart,, published 1847. Appendix C.
364
UTILIZING SOLAR ENERGY
" (439) When, the heat communicated from the sun is confined and prevented from escape, and so forced to accumulate, very high temperatures are attained. Thus, in a small mahogany box blackened inside, covered with window glass fitted to size, but without putty, and simply exposed perpendicularly to the sun's rays, an inclosed thermometer marked, on Nov. 23, 1837, 149° F. ; on Nov. 24, 146°, 150°, 152°, etc., etc. When sand was heaped round the box, to cut off the contact of the cold air, the temperature rose on Dec. 3, 1837, to 177°. And when the same box, with its enclosed thermometer, was established under an external frame of wood well sanded up at the sides, and protected by a sheet of window glass (in addition to that of the box within), the temperatures attained on Dec. 3, 1837, were at Ih 30m P.M. (Appar. T.) 207.0°; at Ih 50m, 217.5°; and at 2h 44 m, 218°, and that with a steady breeze sweeping over the point of exposure. Again on Dec. 5, under a similar form of exposure, temperatures were observed at Oh 19m, of 224°; Oh 29m, 230°; at Ih 15m, 239°; at Ih 57m, 248°; and at 2h 57m, 240.5°. As those temperatures far surpass that of boiling water, some amusing ex- periments were made by exposing eggs, fruit, meat, etc., in the same manner (Dec. 21, 1837, et seq.), all of which, after a moderate length of exposure, were found perfectly cooked — -the eggs being ren- dered hard and powdery to the center; and on one occasion a very respectable stew of meat and vege- tables was prepared, and eaten with no small relish
365
THE SUN
by the entertained bystanders. I doubt not, that multiplying the inclosing vessels, constructing them of blackened copper inside, insulating them from con- tact with each other by charcoal supports, surround- ing the exterior one with cotton, and burying it so surrounded in dry sand, a temperature approaching ignition might readily be commanded without the use of lenses."
MOUCHOT, PlFRE, AND ERICSSON
August Mouchot, of Tours, France, was the great- est pioneer in the utilization of solar heat. He began his experiments prior to 1860 and continued them for about twenty years with aid from the French gov- ernment. He constructed solar cooking appliances, and later large machines for pumping water which he installed in Algeria. Mouchot published in 1869 a work entitled "La Chaleur Solaire et ses Applica- tions Industrielles. " A second edition appeared in 1879. He gives a history of the art, describes many applications of solar heat, and summarizes his own work, including illustrated descriptions of his great solar engines, and a report of his mission to Algeria to install for the Government solar pumping plants in the desert regions.
Solar heaters after the general form of Mouchot's, that is to say, with a conical or paraboloidal reflector, and glass-encased tubular boiler, were also con- structed after the designs of M. Pifre. One of these was exhibited at the Tuileries Garden in Paris in
366
UTILIZING SOLAR ENERGY
1882, in combination with a steam engine and print- ing press, and many copies of a paper called the "Soleil Journal" were printed by solar power.
In America Captain John Ericsson, the inventor of the famous " Monitor" type of naval vessels, devised several solar engines, 1868 to 1886. He used a cylin- dric mirror of parabolic cross section to concentrate the rays upon a tube. A two-and-a-half horse-power engine actuated by one of his solar heaters was ex- hibited in New York at the American Institute Fairs for several years.
ENEAS SOLAR ENGINES
Fig. 68 shows the solar machine of A. G. Eneas (U. S. Patents No. 670,916 and 670,917 of March 26,
FIG. 68. — ENEAS' SOLAR ENGINE.
1901). One of his solar generators was in use for a time at the Cawston Ostrich Farm near Pasadena,
367
THE SUN
California, and others in Arizona, for pumping water. The mirror is composed of facets of silvered glass ar- ranged upon the inner surface of a hollow truncated cone, whose sides make an angle of 45° to the axis. The larger diameter of the cone is stated as preferably as large as thirty-two feet, and in several in- stances was actually thirty-six feet. Decided advantage is claimed in leaving the lower end of the mirror open, as it greatly diminishes the wind pressure, and the part of the cone omitted is not very useful for gathering heat. The mounting shown in the first patent is neither equa- torial nor alt-azimuth, but this fea- ture was improved in the second by substitution of the equatorial form. A canvas shield was provided to protect the instrument from rain. An interesting feature is the form of construction of the boiler shown in Fig. 69. The solar rays are focussed upon the tube F, and the enlarged parts,/1 and/2, are respectively above and below the focal region. The up- per enlargement is a steam and water drum, the lower a settling chamber for extracting foreign matter from the water. Two concentric copper tubes,/ and/8,
368
FIG. 69. — BOILER OF ENEAS' EN- GINE.
UTILIZING SOLAR ENERGY
connect the two enlarged chambers, so that the water falls in / and rises in /8, the latter being of course the hotter. The tube /8 is enclosed by one or more glass tubes, fn, f12, whose purpose is to retard the escape of heat from /8 while admitting the solar rays.
Mr. Eneas has been good enough to furnish me the following details as to the construction of his ma- chines and their efficiency in actual operation :
"Feb. 14, 1901. Pasadena, California, llh 30m to 12h 30 m. Cross-sectional area of sunshine inter- cepted 642 square feet. Air temperature 61° F. Steam pressure per square inch 145-151 pounds. Steam condensed 123 pounds.
"Oct. 3, 1903. Mesa, Arizona. Cross sectional area of sunshine intercepted 700 square feet. Air temperature 74° F. Average steam pressure per square inch 141 pounds. Steam condensed per hour 133 pounds. Time about midday.
"Oct. 9, 1904. Wilcox, Arizona. Time 11 A.M. to 12 M. Cross sectional area of sunshine intercepted 700 square feet. Steam pressure per square inch 148-156 pounds. Steam condensed 144.5 pounds.
"The engines used were of the fore and aft com- pound condensing marine type, complete with direct connected air and feed pump. Size 224" X 6" X 41/2" and operated at 460 to 520 revolutions per min- ute, with about ~ cut off and 25" to 26" vacuum. ID
The steam used in the engine was superheated about 40° F. in the later machines.
369
THE SUN
"I find 3.71 British Thermal Units per square foot per minute given as the greatest amount of heat ob- tainable during the trial runs. The machines re- sembled in design Patent No. 670,917 with equatorial mounting. In the 1904 model the greatest and least diameters of the mirror were 36 and 19 feet, and the angle of inclination between its axis and its sides 45°. The mirrors were made of white glass similar to what
Chance Brothers of London make, and were about —
inch thick, 18 inches wide, and 24 inches long, and were sprung to the curvature of the frame. White glass was used to reduce the loss from absorption. The area of sunshine intercepted is the net area after deducting for shadows caused by the tension rods and frame work. In the later machines built, the mirrors were set so as to concentrate the reflected rays on two parts of the boiler instead of its entire length as in the Pasadena machine. This change gave better results" (perhaps because of the better protection of the remaining parts of the boiler by non-conducting wrapping instead of glass tubes). "The total cost of the machine complete with engine and pump was $2,160.
"An average day's run at Wilcox gave results about as shown in table on following page. Date October 14, 1904.
370
UTILIZING SOLAR ENERGY
TIME HOURS 7 A.M. FOCUSSED |
Steam Pressure in Pounds |
Inches on Water Gauge |
8 |
120 |
14 |
9 |
125 |
18 |
10 |
136 |
21 |
11 |
140 |
26^ |
12 |
152 |
30 + |
1 P.M. |
146 |
30+ |
2 |
141 |
30 + |
3 |
126 |
28 |
4 |
83 |
23 |
5 |
51 |
10 |
" Vacuum 23. Gallons of water pumped, 146,780. Total lift plus friction 39.4 feet. "
(This test would indicate an average horse power for the whole day of about 2-g-. From data to be given below it has been computed that this means the transformation of about four per cent of the solar radiation intercepted by the mirror into mechanical work. From coal the best engines transform from twelve to fifteen per cent of the heat of combustion into mechanical work, but probably not in so small a plant as this. The result of course depends on the efficiency of the steam engine used, as well as on that of the boiler.)
Mr. Eneas continues:
" As a result of my experience with about nine dif- ferent types of large reflectors, I believe: (1) That with similar mirrors perfected in details about 3.90 British Thermal Units per square foot per minute would be the greatest amount of heat obtainable at noontime in Arizona and other cloudless regions of
371
THE SUN
similar latitude. (2) That better progress in utiliz- ing solar heat commercially for power can be made along lines described in the Engineering News of May 13, 1909. But the actual obtaining of any great amount of power from solar rays is still an unsolved problem. "
If we take the number of calories per square centi- meter per minute available as 1.4, we find from Mr. Eneas, figure of 3.71 British Thermal Units per square foot per minute as the " greatest amount of heat ob- tainable during the trial runs" that about seventy- two per cent of the solar radiation was turned into heat in steam. His estimated maximum possible number (3.90 B. T. U.) corresponds to seventy-six per cent. This is really a very satisfactory result. The maximum steam pressures recorded correspond to a temperature of about 185° C.
PROPERTIES OF GLASS
The use of one or more glass casings as an adjunct to the boiler of the Eneas solar engine is quite analo- gous to the use of glass by de Saussure, Herschel, and Mouchot, and also to its common use by gardeners over their hotbeds. Glass transmits radiation very freely between wave lengths 0.37//, in the ultra-violet and 2.5/4 in the infra-red. This range, as indicated by Fig. 26, includes nearly all the solar radiation. The interposition of a single thin glass plate in a beam of sunlight diminishes the intensity about fif- teen per cent. This decrease is owing principally to
UTILIZING SOLAR ENERGY
reflection. The rays emitted by the outside of the boiler, if we estimate its temperature at 500° abso- lute Centigrade, would have their wave length of maximum intensity at about Q/JU and would be almost wholly prevented from directly escaping as radiation by the glass. A large fraction would suffer " metallic reflection" by the glass back to the boiler tube, and the remainder, being absorbed in the glass itself, would tend to raise its temperature and that of the air space, and so to diminish convection from the boiler to the glass. Furthermore, the glass also pre- vents wind from blowing on the boiler, and cuts off all direct convection of heat to the outside air, which is fully as valuable a function as the restraint of out- ward radiation. Thus, the employment of the glass greatly promotes the efficiency of the device, for it raises decidedly the temperature of the boiler. We shall notice below the connection between tempera- ture and the possible thermodynamic efficiency of the engine.
We have already given the interesting story told by Sir John Herschel of the dinner he cooked under glass by solar heat. The late Secretary S. P. Lang- ley was greatly interested by this story and had more than one "hot box" constructed on similar princi- ples. The writer designed one of them. It con- sisted of two round shallow wooden boxes, the inner one 50 centimeters in diameter, the outer 60 centi- meters, placed concentrically one within the other and each covered by a tightly fitting glass plate. The
373
THE SUN
boxes were further protected by a layer of feathers about 10 centimeters thick all around the sides and back of the outer box. The inner one had a black- ened metal sheet near its bottom, and suspended a little above this a blackened thermometer. The whole device was mounted equatorially and kept toward the sun. On November 4, 1897, at Washing- ton, operating with three glass plates, the thermom- eter reached 118° C. while the outside temperature was 16° C.
The question might be asked whether much higher temperatures are not practicable to attain in such a manner without mirrors or lenses to concentrate the heat. Perhaps with better construction it might be possible even to reach 200° C. with such contrivances. The limiting temperature is reached when the solar heat introduced is balanced by the escape of heat by conduction through the glasses and through the in- sulating material at the back. The effective losses diminish with increasing thickness of insulating ma- terial, increasing area of the "hot box, " and increas- ing numbers of glass plates. But, unfortunately, the increase of the number of glass plates diminishes the quantity of solar radiation reaching the inner cham- ber, so that, as found by de Saussure, two or three glasses give best results. The writer has tested with the following results the effect of introducing in a beam of 'sunlight at normal and also 45° incidence successive plates of the common glass 1.5 to 2.0 milli- meters thick used for 8X 10 photographic plates, and
UTILIZING SOLAR ENERGY
of plates 8 to 10 millimeters thick used for instrument covers :
Percentage transmission of glass plates.
Normal |
45° |
|||||||
NUMBER OF PLATES |
1 |
2 |
3 |
4 |
1 |
2 |
3 |
4 |
Thin glasses |
86.5 |
74.5 |
63.5 |
53.3 |
85.0 |
71.8 |
60.0 |
49.0 |
Thick glasses |
79. |
34. |
50. |
39. |
SOLAR HEATERS AND RESERVOIRS
U. S, Patent No. 230,323 of July 20, 1880, was is- sued to Messrs. Molera and Cebrian, who proposed to omit the costly and intricate optical devices for con- centrating the solar heat as used by Mouchot, Erics- son, and others, and even the mechanical devices for presenting the heater broadside toward the sun. They proposed a horizontal boiler composed either of a large number of blackened tubes laid side by side, or a pair of plates enclosing a thin stratum of liquid, and communicating in either case with a suitable engine designed for working at low temperatures. These inventors make no mention of a glass cover for their boiler, but its introduction would undoubtedly have increased the efficiency of their apparatus very greatly.
The erection upon the roof of a building of a series of water tanks protected by a non-conducting mate- rial at the back, and by a glass cover above, and com- municating with the water system of the bath, is 26 375
THE SUN
much used in Southern California, and doubtless elsewhere, as a means of providing warm water. Such devices ordinarily furnish a considerable supply of •water too hot to bear the naked hand in, and save the discomfort of fire in warm weather.
In all countries the sun is obscured more or less of the time by clouds and during the night hours, so that various inventors have proposed the combina- tion of a device for gathering solar heat and a large heat reservoir, which usually takes the form of a tank of water having non-conducting walls, and is situated above the level of the heater, to which it communi- cates by pipes. U. S. Patent No. 784,005 of Feb. 28, 1905, to E. C. Ketchum recognizes such features in combination with a vaporizing chamber situated within the reservoir, and containing some vaporizable liquid suitable for running a low temperature engine. In the event of a very prolonged cloudy spell the in- ventor proposes also a furnace for heating the reser- voir independently of the sun.
Low TEMPERATURE SOLAR ENGINES
Within the last ten years, at least two serious at- tempts have been made to devise commercially eco- nomical means of employing the hot-box principle for power. Both series of experiments are described in the Engineering News for May 13, 1909, referred to by Mr. Eneas. The inventors are Mr. F. Shuman, of Philadelphia, and Messrs. H. E. Willsie and J. Boyle, Jr., of Cranford, N. J. The Shuman heat absorber is a
376
UTILIZING SOLAR ENERGY
level hard rolled plot of ground rendered waterproof by covering it with asphaltum and enclosed by plank partitions rising a few inches above the bottom. In this tank water is filled to a level of about three inches and over it a thin layer of paraffin, which of course melts in the sun, and hinders evaporation and radia- tion from the water surface, while transmitting the solar rays to the water and asphaltum. The whole tank is tightly covered with a single layer of glass set in oiled cotton packing. Wind screens are erected to protect the tank from convection losses. The cost of such construction is said not to exceed twenty-five cents per square foot, and it is expected to produce a horse power for each 160 square feet (it is not stated if this is the average of all conditions or only the re- sult in the most favorable hours, but almost certainly it is the latter) . The water flows from the heater to a steam turbine operated in connection with a vacuum pump. Assuming an initial temperature of 202° F., the vacuum causes the explosion of perhaps ten per cent of water into steam and the reduction of the tem- perature of the mixed steam and water to about 102° F.
As the maximum possible thermodynamic effi- ciency under such conditions is fifteen per cent, it is unlikely that as much as five per cent of the sun's heat can be converted into mechanical work. A large storage reservoir, built below ground and well insulated, is connected with the apparatus in such a manner that the excess of hot water during the hot-
377
THE SUN
test part of the day goes in at the top of the reservoir, while water from the bottom of the reservoir is with- drawn to supply that withdrawn from the heating tank. During the morning and evening hours, or under cloudy conditions, the motor can be run from the reservoir. The plant is still in the experimental stage, but appears to be well planned, and promises considerable success.
The apparatus of Messrs. Willsie and Boyle has been more thoroughly tried, so that Mr. Willsie gives actual figures as to cost and efficiency. They prefer to build an entirely wooden basin coated with asphalt, for they find the sand even of the desert to contain moisture which injures its quality for a non-conductor of heat. In order to promote a more rapid circula- tion of the water, and its consequent higher efficiency to absorb heat, they incline the basin. In their latest construction the water runs from a first basin with one glass cover to a second with two, and from this it drips over a row of pipes containing sulphur dioxide gas. They employ a low pressure sulphur dioxide engine of the type developed 'n Germany by Pro- fessor Josse. In their experiments they run between temperatures approximating 200° F. and 100° F., but at midday their heater sometimes reaches near!y 260° F. They also combine their apparatus with a large reservoir for use at night or in cloudy weather. Four installations have been erected by Willsie and Boyle, the first at the St. Louis Exposition, the others at Needles, Arizona, a place that all travelers who have
378
UTILIZING SOLAR ENERGY
been there will agree is well qualified for experiments with solar heat! Mr. Willsie estimates the cost of installing a sun-power plant at $164 per horse power, and the cost of operating 400-horse-power steam- electric and solar-electric plants in desert regions at 2.08 and 0.61 cents per electric horse-power hour, respectively.
SOLAR COOKING APPLIANCES
Experiments in solar cooking which attracted con- siderable public attention were made in 1878 by W. Adams of Bombay, India. Fig. 70 shows the very simple apparatus employed by him for cooking pur- poses. The eight-sided conical concentrator was made of wood lined with silvered glass. It was hinged upon a board and adjusted by a wedge and by rotating the board so as to face the sun. The posi- tion of the apparatus required to be changed about once each half hour. The cooking vessel of copper was enclosed in a glass case and fixed to the back of the concentrator. Mr. Adams wrote to the Scientific American,1 that the rations of seven soldiers, consist- ing of meat and vegetables, were thoroughly cooked by it in a couple of hours, in January, the coldest month of the year in Bombay ; and that the men de- clared the food to be cooked much better than in the ordinary manner. It was also tried with success by several people in Bombay and in the Deccan. The
1 June 5, 1878. 379
THE SUN
380
UTILIZING SOLAR ENERGY
dish is stewed or baked, according as the steam is re- tained or allowed to escape. Adams' reflector was two feet four inches in diameter.
SOLAR METALLURGY
Besides devices for producing power, for cooking, and for warming water for domestic purposes, by solar heat, we may note its proposed application for metallurgy. U. S. Patent No. 277,884 of May 22, 1883, was granted to John Clark of England for a " Method of Reducing Metals from their Ores. " The inventor proposes to use a concave mirror built of segments of silvered glass or burnished metal, mounted in a manner convenient to face the sun, and adapted to focus the solar rays upon a stick of ore, for instance of the oxide or chloride of aluminum or magnesium, formed into a convenient shape by com- pression from the powdered substance. He proposes either to mix solid reducing agents with the ore or else, when the ore is heated to a suitable temperature, to convey a gaseous reducing agent, as hydrogen or carbon monoxide, to the incandescent material. The excess of the reducing agent is supposed to prevent reoxidation of the reduced metal, but this may be further guarded against by enclosing the whole ap- paratus in a glass roofed chamber filled with a neu- tral or reducing gas. The advantage claimed for the proposed use of solar rather than other sources of heat is the fact that a very high temperature can thus be readily obtained.
381
THE SUN
RESUME
In the preceding pages we have noted various devices which, singly or in combination, have been employed by numerous inventors for the utilization of solar heat. They comprise first a large surface for receiving the sun's rays. This may be fixed in a horizontal or other preferred position, or progress- ively inclined by suitable mechanism to suit the posi- tion of the sun. In the former case the surface is blackened to promote absorption, and the heat thus derived is communicated to some liquid for domestic use or for the running of a low temperature heat en- gine. More commonly mirrors (or sometimes lenses or prisms) are provided for concentrating the rays to an approximate focus. Usually the mirror is com- posed of a large number of facets of plane silvered glass or burnished metal arranged upon a frame of suitable general curvature. The form of the reflect- ing combination may be a paraboloid, or cone of rev- olution, or an arc of a cylinder of parabolic cross sec- tion. At the approximate center of concentration of the rays is located a heater for the ore to be reduced or the liquid to be vaporized. Advantage is gained in this case, and also in the fixed forms of solar heater, by encasing the heated part with glass in the direc- tion from which come the solar rays, and protecting it by non-conductors of heat in other directions. The means of presenting apparatus to the sun usually em- ployed by astronomers, such as the English type of
382
UTILIZING SOLAR ENERGY
open fork equatorial mounting, which would seem to be excellently adapted for the purpose, do not appear to have appealed to the solar engine inventors, as a rule. They have generally devised more complex mechanical movements for their purposes, including circular tracks, slotted hinged uprights, intermediate types between the alt-azimuth and equatorial forms of mounting, etc. Excepting the solar heaters for bath purposes commonly installed in the roofs of houses, it does not appear that appliances for utilizing solar heat are yet introduced with economical success in practice, for although much work has been done in this line for centuries, we hardly ever see any of the machines.
We shall conclude this chapter by a consideration of some of the data to be used in the design of solar heat apparatus.
QUANTITY OF SOLAR ENERGY AVAILABLE
We may first inquire how much solar radiation is available. The following data are computed from the Smithsonian pyrheliometric observations at Washington and Mount Wilson. Sun rays may be received on a surface at right angles to the beam (" normal incidence"), in which case the surface must be moved by suitable mechanism to follow the ap- parent motion of the sun in the heavens. On the other hand, the rays may be received on a fixed hori- zontal surface, in which case their intensity will diminish as the cosine of the sun's zenith distance.
383
THE SUN
In either case there is the decrease of the intensity of the rays depending on the length of path in the at-
f.5 1.0 |
K |
71 gives the mean intensity of direct sun- shine in calo- ries per square j centimeter per minute for Mount Wilson and Washing- 1 ton. Horizon- tal distances give ' ' air mass- es, " or, in other words, secants of the zenith 111 distances of the sun1. Vertical distances are calories. One pair of curves, III and IV, is for the receiv- s, I and II, for |
|||||||
"\ |
X |
||||||||
1 |
\ |
^ |
|||||||
V |
^ |
\ |
|||||||
G |
X |
^ |
^^ |
||||||
\\ |
'X. |
x^ |
"^ |
||||||
\ |
\ |
X, |
|||||||
\ |
\ |
^ |
^^ |
||||||
\\ |
^ |
||||||||
\ |
\ |
||||||||
\\ |
|||||||||
5* § CALORIES. £ |
\ |
X |
|||||||
X |
X \ |
s^ |
|||||||
^ |
^ |
^^ |
^i |
||||||
^ |
|||||||||
SEC.Z 234 FIG. 71. — INTENSITY OF SUN RAYS. (Mount AVilson and Washington.) I, II. Normal incidence. Ill, IV. On horizontal surface. g surface horizontal, the other |
1 The secant, of (ho zenith distance ceases to represent closely the "air mass" for zenith distances above 78^9° where sec. 7 - 5. From some measurements made at very low sun the data given below are extended to sun rising and setting.
UTILIZING SOLAR ENERGY
" normal incidence." The curves I and III are for Mount Wilson.
In Fig. 72 (upper half) is shown for sea, level and 6000 feet elevation, both for horizontal and normal incidence, the march of the sun's direct radiation
\\
\\
\
L/T.:
\
//
30"
FIG. 72. — INTENSITY OF SOLAR RADIATION.
Sea-level and 6,000-feet elevation. Normal incidence and on hori- zontal surface.
from noon to sunset on December 22, February 17 (and October 25), March 21 (and September 23), April 22 (and August 22), June 22, at which times the sun's declination is— 23J^°— 12°, and 0°, + 12° -h 233/2°, respectively. The data are computed for latitude 38° N. Horizontal distances give the hours,
385
THE SUN
and vertical distances the calories per square centi- meter per minute. Similar computations have been made for latitudes 20° N., 30° N., and 45° N., but are not shown in Fig. 72. From these results come the data represented in the lower half of Fig. 72. The curves show the number of calories of solar heating per square centimeter per day falling in cloudless weather on surfaces at horizontal and normal inci- dence, at sea-level and 6000 feet altitude respectively for the given latitudes. In each group the two upper curves are for normal incidence; the highest for 6000 feet elevation. In the following table is a summary of the whole matter expressed in calories per square centimeter per year, and also in square feet required on the average per horse-power assum- ing complete absorption and transformation, and the sun to shine 261,000 minutes per year.
NORMAL INCIDKNCK |
HORIZONTAL SURFACK |
||||
Sea-level |
6,000 feet |
Sea-level |
<;,<><)() feet |
||
20° |
292,000 |
362,000 |
185,000 |
226,000 |
Calories per sq. |
30° |
287,000 |
355,000 |
170,000 |
203,000 |
cm. per year |
38° |
271,000 |
342,000 |
152,000 |
185,000 |
|
45° |
270,000 |
340,000 |
137,000 |
169,000 |
|
20° |
10.5 |
8.5 |
16.6 |
13.6 |
Average sq. feet |
30° |
10.7 |
8.8 |
18.1 |
15.1 |
per horse- |
38° |
11.3 |
9.0 |
20.2 |
16.6 |
power |
45° |
11.4 |
9.1 |
22.4 |
18.2 |
It is not difficult to absorb ninety-five per cent of the solar radiation falling upon a surface. Lamp-
386
UTILIZING SOLAR ENERGY
black is employed as an absorber if the temperature is low, and platinum black electrolytically deposited if the temperature is so high as to burn off lampblack. There are many regions of the earth where the days are seventy-five to ninety per cent cloudless, or even more. Hence we may conclude that there are many regions where for the average daylight hours it is practicable to absorb on a surface of from one to two square yards solar heat mechanically equivalent to a horse-power. But in the production of mechanical power from solar heat only a small percentage is ac- tually utilized.
THERMODYNAMIC EFFICIENCY
It is shown in works on Thermodynamics that a perfect engine taking in heat at the absolute temper- ature TI, and rejecting it at T2, can transform only
rp rp
the fraction - l * of the heat into mechanical ^i
work. For illustration, suppose the engine taking in heat at the boiling point of water, 373° C. absolute, and rejecting it at the freezing point, 273°, the maxi-
100
mum efficiency possible will then be — = 26.8 per
o7o
cent. This thermodynamic law gives the efficiency of a perfect engine, and it does not matter what its nature, if its actuating energy is heat. A thermo- electrical engine or a steam engine are both heat en- gines, and their efficiency cannot exceed that calcu- lated by the above rule. In fact, however, no heat
387
THE SUN
engine is perfect, and the best triple expansion con- densing steam engines, with the best constructed boilers, hardly ever convert as much as fifteen per cent of the heat of combustion of the coal they con- sume into work. If a heat engine works from a very
high temperature to a low one, the fraction
•M
may approach nearly to unity. For instance, sup-
rr^ rr\
pose T! = 1000° and T2 = 300° then —L^~ 2 = 70
per cent. This accounts in part for the high effi- ciency of internal explosion engines, which develop high temperatures in their cylinders, and often con- vert twenty-five per cent of the heat of combustion of their fuel into work. On the other hand, the losses of heat by conduction, convection, and radiation in- crease rapidly with rising temperatures, so that if engines, are used at very high temperatures the ther- modynamic gain may be counterbalanced by a prac- tical loss.
REFLECTING POWER OF MIRROR SURFACES
TABLE XXX. — Percentage reflecting power of various surfaces
WAVE LENGTH |
0.35M |
0.40M |
0.45,* |
0. 50M |
0.60/u |
0.70ft. |
0.80M |
LOOM |
LOOM |
Glass1, -silvered on back. |
|||||||||
Sample A |
07 |
82 |
90 |
93 |
94 |
94 |
95 |
95 |
|
Glass, silvered on back, |
|||||||||
Sample B |
68 |
80 |
86 |
84 |
76 |
65 |
56 |
65 |
|
Glass, mercurv back .... |
73 |
71 |
70 |
73 |
|||||
Silvered on glass (Chem. |
|||||||||
Dept.) |
74 |
83 |
90 |
91 |
93 |
95 |
96 |
97 |
98 |
Nickel (Electrolyt. Dep.) |
48 |
68 |
59 |
61 |
06 |
69 |
70 |
72 |
79 |
Speculum metal |
51 |
55 |
60 |
63 |
64 |
67 |
68 |
70 |
75 |
1 The reflecting power of mirrors coated on the back differs greatly with the character of the glass used. Sample A is ordinary optical
388
UTILIZING SOLAR ENERGY
Taking all things into consideration, glass plates silvered on the back are probably the best materials for constructing the mirrors of solar heaters.
One may well question whether the solar engine of the future will have mirrors and driving mechanism. The "hot box" of de Saussure and Sir John Herschel as applied by Willsie and Boyle and Shuman is so cheap that the low efficiency inseparable from its low working temperature seems not to bar its use com- mercially. Some gain in efficiency may be made by installing the fixed heating surface parallel to the earth's axis instead of horizontal, but perhaps the increased cost may offset this gain. The efficiency of the apparatus depends on the excellence of the glass protection in front. If one could, in addition, make a vacuum under the glass economically, the efficiency would be much higher. This device de- serves much attention.
It seems highly probable that solar cooking uten- sils, combined with water heaters and heat reservoirs, and embodying the "fireless cooker" principle, will come into extensive use. For it is not hard to see that very inexpensive apparatus may be designed for combining these utilities, and that housekeepers will welcome a relief from the hot kitchen conditions of summer.
flint glass, 12 millimeters thick, silvered on the back by chemical deposition. Sample B is ordinary commercial plate glass of a greenish tinge, about 8 millimeters thick, similarly silvered. The glass of sample B perhaps has an absorption band in the upper infra- red spectrum.
389
THE SUN
Good designing, avoiding costly and complicated construction and devices likely to require frequent attention, combined with a fuller knowledge of the properties of materials available, and cleverness in adapting means to promote efficient results, — these if supported by a moderate outlay of money for experi- mental work may perhaps soon make the utilization of solar energy very extensive.
CHAPTER X
THE SUN AMONG THE STARS
Stellar Distances. — Magnitudes. — Sun's Magnitude and Light Emission. — Solar Motion. — Star Groups. — Double Stars. — • Stel- lar Masses and Densities. — Mira Ceti and the Sun. — Stellar Spectra. — The Classification of Stellar Spectra. — Radiation Distribution. — Evolution of the Solar System. — Stellar Evolu- tion.
AT first glance the stars appear to be about as much like the sun as the fireflies of a summer night. It is only prolonged investigation which has proved that the sun is merely a star, and by no means the largest of them; and that if the sun should be removed a great distance it would appear like one of the stars. The Copernican view that the sun is the center about which the earth and planets revolve seemed satisfactory enough as far as concerned the solar system, but was for centuries hard of belief as regards the stars. For it required the assumption that these were all so distant that the enormous dis- placement of the earth in space between summer and winter produced no measureable changes in their ap- parent relative positions. If the reader will walk a hundred paces in any direction within a forest, he will instantly see that the trees change their relative direc- tions from him, and if he rides in the cars he perceives 27 391
THE SUN
that the foreground of the .landscape appears to re- volve. Objects at greater and greater distances from him are seen to be less and less apparently affected among themselves by his motion. Accordingly, if the astronomer of a century ago agreed with Copernicus he had to believe that since the stars do not sensibly change in their relative positions during the year, they are so distant from the earth that a displacement of between one and two hundred million miles in the position of the earth as viewed from the nearest star subtended an angle too small for him to measure. In other words, as he could observe changes as small as a couple of seconds of arc, he had to believe that the stars were at any rate all more than 100,000,000 X 100,000 miles away. In our day we know that this is so, for the distances of some of them have actually been measured, but a century ago the astronomers took it on faith, merely because they accepted the Copernican system.
The first successful measurements of stellar pa al- laxes (a star's annual parallax is the angle the radius of the earth's orbit subtends viewed from the star) were made by Struve, at Dorpat, on Vega, 1835 to 1838, and by Bessel on the star 61 Cygni, 1837 to 1840. The latter faint star was selected on account of its large proper motion. Bessel's result was 0."35, and Struve's about one-quarter second. The former is nearly correct, the latter about twice too large. It was a great feat to measure such small angles as these. In modern practice the efforts to measure parallaxes
THE SUN AMONG THE STARS
absolutely has practically been discontinued, and in- stead relative parallaxes are determined. That is, instead of measuring, for instance, the apparent ab- solute change of polar distance of a certain star due to the earth's revolution around the sun, astronomers now for the most part determine how much a given star appears to shift among very faint neighboring stars owing to the same cause. For it is now as- sumed that the very faint stars on the average are so very distant that they have no sensible parallaxes, or at most a very minute and approximately known average parallax, which can be applied as a correc- tion. Stars are generally selected for individual par- allax measurements because they have relatively large " proper motions," or progressive apparent displacement among the stars. This is usually a safe criterion of comparative nearness, as ap- pears from our illustration of the railway train above. In a survey of ninety-two stars pub- lished a few years ago by Chase of Yale, there were found the following numbers of stars between given limits of parallax:
Number of Stars |
Limits of Parallax |
2 |
0".25to 0".20 |
6 |
0".20to 0".15 |
11 |
0".15to O'MO |
24 |
O'MO to 0".05 |
34 |
0".05to 0".00 |
8 |
0". 00 to -0". 05 |
5 |
-0". 05 to -O'MO |
2 |
-O'MO to -0". 15 |
393
THE SUN
The negative parallaxes are of course illusory, and hence we may suppose that some positive ones are also erroneous, so that of this lot no more than two- thirds have measurable parallaxes, and of the whole lot three-fourths are more distant than 100,000,000 X 2,000,000 miles. It is customary to express such enormous distances in terms of " light years. " Light travels in one year about 6,000,000,000,000 miles. Accordingly, the number given above is about thirty light years, which represents approximately the dis- tance of a star whose parallax is one-tenth second. The vast majority of stars are many times as distant as this, and the nearest yet found is d Centauri, whose distance is about four light years.
STELLAR MAGNITUDES. THE SUN'S MAGNITUDE
The relative brightness of the stars is expressed in " Magnitudes, " a star of the first magnitude giving about 2.5 times the light of one of the second. On this scale Polaris is nearly of the second, Aldebaran nearly of the first, Vega nearly of zero, Sirius — 1.4, and the sun — 26.5. A change of five magnitudes makes a change of one hundredfold in the light, so that the sun gives the earth over 90,000,000,000 times the light of Aldebaran. If removed to the distance of Aldebaran, whose annual parallax is O."ll, the sun would become a star of the fifth magnitude, and ap- pear only about as bright as the fainter stars among the six easily seen in the Pleiades. Accordingly, Al- debaran emits about forty-five times as much light
,",!) I
THE SUN AMONG THE STARS
as the sun. There are some stars, among them Rigel, Canopus, and Deneb, which are sensibly of zero par- allax and yet of first magnitude, or brighter, so that they must emit many thousands, perhaps hundreds of thousands of times as much light as the sun. On the other hand, there are many stars whose light emission is very much less than the sun's, among them the rapidly moving star whose parallax was measured by Bessel, sixty-one Cygni. Its light emis- sion is one-tenth that of the sun.
SOLAR MOTION AMONG THE STARS
As in a forest walk the trees in front seem to sepa- rate as we approach, and those behind to crowd to- gether as we recede, so the stars exhibit a tendency to move from the approximate direction of the constel- lation Hercules towards the constellation Argo in the Southern Hemisphere. In consequence of the great distance of the stars these displacements, called proper motions, are very slow, not often exceeding 100" per century, and usually very much less. Nevertheless, the observations of star places are so exact that the foci of the motions have been determined with an un- certainty of only a few degrees. That in the North- ern Hemisphere lies in R'ght Ascension 270°, Declin- ation + 30°, in the constellation Hercules about 10° southwest of the bright star Vega. The cause of the phenomenon is the motion of the solar system, rela- tively to the stars in general, toward the position just defined, called the solar apex. The rate of motion is
395
THE SUN
determined from the apparent proper motions of the stars of known parallaxes, or from the general spectro- scopic survey of the stellar motions in the line of sight.
Professor Campbell has been good enough to give me the following summary of various determinations :
"Sir William Herschel in 1783 deduced from the proper motions of thirteen stars (all then available) that the solar system was travelling approximately toward the star Lambda Herculis in Right Ascension 262°, Declination -f 26° x.
" Many determinations of the goal of the sun's way were made in the latter half of the nineteenth century, as the proper motions of stars became known in greater numbers. Of those based upon the most ex- tensive lists of proper motions we mention the follow- ing:
" Newcomb's coordinates for the apex of the sun's way, deduced from about 3,100 Bradley stars, are Right Ascension 275°, Declination + 31°L>.
" From 2,640 Bradley stars,, Kapteyn deduced the position of the apex Right Ascension 274°, Declina- ation + 29°.53.
" From the proper motions of 5,413 stars Boss has computed the apex to be at Right Ascension 270°. 5, Declination + 34°. 3; and his estimate of the velocity of the solar motion is 24 km. per second.4
1 Philosophical Transactions, vol. xv, page 405, 1783.
2 The Stars, page 91, 1901.
* Astronomische Xticttn'chtcn, vol. rlvi, page 17, 1901. 4 Astronomical Journal, No. 614, 1910. 396
THE SUN AMONG THE STARS
" Several solutions for the elements of the solar motion have been based upon the observed radial velocities of stars.
" In 1900 Campbell, from 280 stellar radial veloci- ties in the northern three-fifths of the sky, obtained for the position of the apex Right Ascension 277°. 5, Declination + 20°; and for the speed 19.9 km. per second.
"*[n 1909 Hough and Halm based a solution upon about 500 stellar radial velocities. They obtained for the position of the apex Right Ascension 271°, Declination + 25°.6; and for the velocity 20.85 km. per second.
" In 1910 Campbell deduced the elements of the solar motion from the observed radial velocities of 1034 stars and thirteen nebulae. His position of the apex was deduced as Right Ascension 272°, Decli- nation + 27°. 5; and the velocity as 17.8 km. per second.
" It should be held in mind that the motion of the solar system is a purely relative term, and in every case refers to the particular group of stars used as a basis for the solution. The computer's aim should always be to have his observational material as ho- mogeneous and as representative of the entire sidereal system as possible.
" It appears that an uncertainty of several degrees exists as to the direction of the solar motion with reference to the entire sidereal system, and perhaps of several kilometers as to the speed of this motion.
397
THE SUN
Perhaps the following values are as probable as any that we can at present assign :
" Apex at Right Ascension 270°, Decimation + 30°. Velocity 18 km. per second.
" It should be said, however, that some astrono- mers would consider the following values as more probable :
" Right Ascension 272°, Declination +33°;
" Velocity 20 km. per second.
" Kapteyn and Frost have obtained indications that the speed of the solar system, with reference to stars of spectral type B, is considerably greater than with reference to the system as a whole, but the number of B-type stars employed in the discussion is perhaps too small to yield results entirely trustworthy.
" Campbell has found that the velocity of the solar motion, with reference to stars of spectral types B, A, and F to F4, inclusive, is in essential agreement with the velocity deduced from stars of spectral types F5 to G, inclusive, K and M.
" It does not clearly appear that the direction and speed of the solar motion are functions of the dis- tances of the stars used as a basis for the solutions."
STAR GROUPS
Whether the sun has companion stars in its course is not known certainly, but there are known groups of stars which seem to form well-defined systems moving with a common trend. Such a group is. the Pleiades, including, besides the six stars easily visible, a much
398
PLATE XIX
THE PLEIADES. (G. W. Ritchey.)
Photographed with the 2-foot reflector of the Yerkes Observatory, 1901, October 19. Exposure 3^ hours. Cramer Crown plate.
THE SUN AMONG THE STARS
larger number of telescopic stars. Their connection to form a common system is shown by at least three kinds of evidence. First, it is highly improbable that so many stars of that brightness would fall in so small a region of the sky, if the distribution was purely at random. Second, excluding a few stars not regarded as belonging to the system, the stars mentioned have equal proper motions in the same direction. The common proper motion is about 6" per century. Third, there is a nebula, of filmy cloud patch in the sky, visibly connected with the several stars of the group and evidently confirming their common con- nection (see Plate XIX). The Pleiades group, in- cluding small stars partaking of the common motion, measures nearly 100' of arc in average diameter. The parallaxes of the stars are not certainly measurable, but their distance has been estimated with some plausibility to be not less, at any rate, than 200 light years. Hence, the rad us of the system is not less than three light years, or 18,000,000,000,000 miles, which is 6,000 times the radius of Neptune's orbit. If, indeed, the group is actually as small as this, it would mean perhaps a hundred good sized stars nearer together than the sun is to its nearest stellar neighbor.
The curious connection of nebulosity with the Pleiades is not without its counterpart in many other regions of the sky, and even our own solar system seems not to be devoid of it. There is observable on dark nights, nearly in the plane of the ecliptic, a light
399
THE SUN
not to be regarded as incipient twilight, called the Zodiacal light when viewed towards the sun, and called the Gegenschein in the opposite direction. See- liger has estimated, on reasonable assumptions, that it is the matter contained in this ring of nebulosity which causes the outstanding perturbation in the orbit of Mercury, not to be accounted for by the at- traction of known planetary masses. It has been supposed that stars have their origin in nebulae, and if so the Pleiades stars would seem to be less advanced in their course of evolution than the sun, but we shall recur to this.
Whether it is gravitation which controls the motion of the sun among the stars, and whether such a vast system as the Pleiades is, like the planetary systems, in orderly gravitational movement, are questions which as yet there are no means of fully solving, but the affirmative is generally believed It has been computed by Newcomb, however, that there is not enough matter in the universe to control the motion of such runaway stars as 1830 Groombridge and Arc- turus.
DOUBLE STARS
That gravitation is an universal property seems to be proved by the existence of well-observed elliptical orbits in the cases of many pairs of double stars. Since there are less than ten thousand stars to the sixth magnitude in the whole heavens, the chances are almost infinitely small that two of them should be found within 5" of one another on a random distribu-
THE SUN AMONG THE STARS
tion, for there are over 20,000,000,000 squares of 5" in the whole sky. But in fact there are many pairs closer than this among the visible stars, so that a physical connection in most such cases is practically certain. In some cases, as in that of the very bright pair of 0.4 and 1.9 magnitude composing a Centauri, stars separated by a much greater interval (in this case averaging 17". 1) are proved to be physically connected because they are observed to go through a periodic change of position with respect to one an- other. The orbit of a Centauri is completed in eighty- one years. By spectroscopic observations of motion in the line of sight a great number of stars not tel- escopically resolvable are proved to be physically connected doubles because of the variable velocity observed. In some cases of spectroscopic binaries the companions are indicated by doubling of the spectrum lines, but quite often one of the objects is too faint to give a spectrum, and its existence is noted only because the periodically variable positions of the lines in the observed spectrum indicate that the star observed is affected by orbital motion.
STELLAR MASSES AND DENSITIES
The spectroscopic method gives the projection on the line of sight of the linear velocity of one or of both components in their orbits. The telescopic method gives the projection at right angles to the line of sight of the angular motion of the components. Both methods give the period of the revolution. When the
401
THE SUN
parallax of the object is known, as in the case of a Centauri, the projected linear dimensions of the orbits are easily found. It is possible in the case of accurately observed telescopic binaries of known par- allax, or of pairs whose motions have also been ob- served spectroscopically, to determine the actual linear dimensions of the two orbits, and (assuming the law of gravitation) the relative masses of the two stars. The combined mass, as compared with the combined mass of the earth and sun, follows easily from Kepler's third law. For if we regard the mass of the earth and sun combined as unit mass, the rad- ius of the earth's orbit as unit distance, and the year as unit time; then calling the period, total mass, and mean radius vector of the binary, P, M, and R, re- spectively, we have, if matter has the same constant of gravitation everywhere:
iwr R3 M = pi-
Since R and P are both known for a well-determined orbit, we thus find M, the ratio of the combined mass of the binary to the combined mass of earth and sun. In the case of a Centauri the total mass is twice that of the sun, and the components being approximately of equal mass, they are singly about of the same mass as the sun. Their mean distance apart is 23.6 times the radius of the earth's orbit.
By such processes the combined masses of various binary stellar systems have been determined. The resulting masses are sometimes less, sometimes a few
403
THE SUN AMONG THE STARS
fold greater than the sun's. Had they come out of another order of magnitude entirely, it would have seemed doubtful if the constant of gravitation has the same value in other systems than ours. But as things are, we seem justified in supposing that gravi- tation is an universal unchanging property of matter. There is a method proposed by Pickering for find- ing a relation between the surface brilliancy and den- sity of the average star of a binary system whose period and magnitude, on the scale of brightness, are known. Without going into an explanation of the matter, which may be found in works on the stars, we shall be interested in the conclusion, which is that stars in general give much more light in proportion to their masses than does the sun. Astronomers gener- ally incline to believe that the discrepancy indicates for the stars generally a smaller density than that of the sun. In a few instances, another line of argument regarding star densities is possible. There are some binary systems whose orbits are of such small dimen- sions, and lie so nearly in a plane with the earth, that the components regularly eclipse one another, and the quantity of light of the binary thereby suffers periodic variability. In such a system the duration of the eclipse compared with the period of the orbit, gives a measure of the relative diameters of the stars relatively to the diameter of their orbits. Proceeding in this fashion it was shown by Roberts that the aver- age densities of these variables (called " Algol vari- ables" after the name of the famous spectroscopic
403
THE SUN
binary star which is the type of the class) is no more than one-eighth that of the sun. This general con- clusion was independently confirmed at the same time by Russell.
We have spoken incidentally of the Algol type of eclipsing variable stars. Without going far into the discussion of stellar variability, it will be of deep in- terest, in view of the somewhat irregular and very slight variability of the sun, to speak of another kind of variable stars of which Omicron Ceti (or Mira Ceti) is the type. This star is sometimes as bright as the second magnitude, and sometimes as faint as the ninth or fainter. Accordingly its range is several thousandfold in brightness. It goes through its cycle in an average period of about 331.6 days, but is sometimes thirty or forty days early or late in coming to a maximum. Its maxima and minima are not uniformly bright, for sometimes it attains only the fifth magnitude at maximum, and sometimes it falls only to the eighth magnitude at minimum. The time required to rise from minimum to maximum brightness is only about two-thirds the time required to fall to a minimum. The shape of the light curve is variable too, as the maxima continue longer at some recurrences than at others.
The spectrum of Mira is of the third type, to which Antares belongs,1 distinguished by the fluted spectra found to some extent in sun spots. The spectrum varies as the star's brightness varies, becoming
1 See below.
101
THE SUN AMONG THE STARS
stronger in the violet, and especially in its bright violet hydrogen lines, in the maximum phases. The spectrum indicates a high velocity of recession from the sun (66 kilometers per second), but there is no evidence from it that Mira has companions.
In many respects Mira's variability suggests the solar variability associated with sun spots. True, the fractional change of solar radiation is perhaps not
more than 773773:7: as great as that of Mira, but in the 100,000
existence of a fixed average period subject to large individual departures, an unequal intensity of max- ima, an unsymmetrical and variable light curve, there is a strong similarity to what the sun spot curve suggests for the sun. In one respect there is a di- vergence. Mira increases in brightness faster than it decreases. The change of temperature of the earth seems to indicate that the sun's radiation is at a maximum when sunspots are fewest. But the sun spots decrease to a minimum slower than they in- crease to a maximum. Still, with so many features of similarity there can be little doubt that the dis- covery of the cause and accompanying phenomena of the sun spot periodicity will indicate the secret of the Mira type of variables.
STELLAR SPECTRA
Having taken some note of the distances, motions, brightnesses, masses and densities of the stars as compared with that of the sun, and having seen that
405
THE SUN
we owe this information to knowledge gained of the sun itself, and of the solar system, we may now turn our attention to the spectra of the stars, and see wherein and how far the sun is a type in that respect. We have noted that in the solar spectrum dark lines of the metals are the prevailing feature. Calcium and hydrogen lines sometimes give bright reversals in their centers. Helium seldom produces a dark pho- tospheric line,but in the spectrum of the chromosphere its bright lines are conspicuous along with those of hydrogen and calcium. In the spectra of sun spots the dark lines of metals are still conspicuous, but are nearly overshadowed in importance by banded spec- tra of various compounds, and the violet end of the spectrum is very weak in them compared with the red, or with the violet of the ordinary photospheric spectrum.
These various peculiarities of the solar spectra find counterparts in the stars. There is a large class of stars whose spectra are hardly to be distinguished, line for line, from that of the sun. Among the most exact duplicates is the spectrum of the principal star of the brilliant binary system Capella. From this solar type we can pass either way; in the one direc- tion to stars on which the red predominates, and banded spectra overshadow the metallic lines, or in the other direction to blue stars in which lines of hydrogen or helium are almost the sole features aside from the continuous spectrum.
By the kindness of Director Campbell of the Lick 406
THE SUN AMONG THE STARS
Observatory and Director Frost of the Yerkes Ob- servatory, I give here in Plates XXA and B and XXI a series of spectra illustrative of the gradation from the so-called helium or Orion stars to the so- called carbon stars, which lie at opposite ends of the scale. While considering these, let us note more closely the diversities of stellar spectra.
The Classification of Stellar Spectra.
Father Secchi in 1867 divided the spectra of stars into four great classes. Class I comprises the blue and white stars. In their spectra dark lines of metal are few and feeble, but the dark hydrogen lines are well marked. This class is the most numerous, and includes among other prominent stars, Sirius, Vega, and Procyon. Class II comprises the yellow stars whose spectra are filled with metallic lines. This class includes the sun, also Capella, Arcturus, and Aldebaran. Class III comprises orange and red stars whose spectra show, besides many dark metallic lines found in the stars of the second type, also nu- merous dark bands or flutings. These consist, like the terrestrial oxygen bands, of series starting with well-marked heads and shading off from these to- wards the red. These flutings are now recognized to be caused by oxides of titanium and other metals, and by hydrides as, for example, of calcium. This class includes Antares and Betelgeuse. Class IV comprises some deep red stars, whose spectra also contain bands or flutings, but with the shadings
28 407
THE SUN
toward the violet. These flu tings are attributed to carbon or its compounds. The stars of Class IV are all faint. The two brightest are nineteen Piscium (5.3 mag.) and 152 Schjellerup (5.5 mag.).
The stellar classification of Secchi is still much used in general descriptions, although more detailed systems of classification have been lately adopted. The accompanying Plates XXA and B and XXI illustrate some of the differences between Secchi's types. It is however, practically impossible, with- out having had personal handling of direct spectrum photographs, to note at a glance the significant vari- ations in spectra. The spectral types of Secchi merge, of course, gradually together, so that in some cases' one would be doubtful in which of two classes to assign a star.
There are two principal modifications to be made to Secchi's classification. First, and most impor- tant/among the blue or white stars occur many whose spectra are distinguished by the absorption lines of helium, more than by those of hydrogen. Lines of oxygen and silicon also sometimes occur in these helium star spectra, but most metallic lines are extremely faint or invisible. Helium stars are nu- merous in the constellation Orion and in the Milky Way. Secchi's Class I may then be divided into two principal sub-classes, the helium or Orion stars, and the hydrogen or Sirian stars. The helium stars not infrequently show some bright emission lines in their spectra besides the dark or absorption Jines.
408
a sa
a
,>«j
1
a -813133
& ||||
i iiii
I 1111
ejfi
S3.itf
ssss
THE SUN AMONG THE STARS
Such bright lines are generally of hydrogen, but the helium line D3 is also bright. Vogel's classification includes a third division of Class I for such bright line stars.
But there is a class of stars for which Pickering at one time proposed to add a Class V to Secchi's sys- tem, whose spectra have as their main characteristics bright lines or bands in the yellow and blue, some due to hydrogen, others of unknown origin. The bright line, or so-called Wolf-Rayet stars, are situated mostly in the Milky Way or the Magellanic Clouds, and, except 7 Vela, are faint stars. Some of the ultra-violet lines bright in the spectra of Wolf-Rayet stars are also bright lines in the spectra of certain nebulae.
There has been adopted at the Harvard College Observatory a more detailed system of stellar classi- fication than either Secchi's or Vogel's, and which includes numbered gradations of the lettered main divisions, so that a very large number of varieties of spectra may be indicated. A spectrum marked B3A, or more briefly B3, is one which is estimated to be three-tenths the way from a typical B star to a typical A star, and similarly for other combinations. The following table gives parallel designations of typical stars under the classifications of Secchi, Vogel, and Harvard College Observatory. In the Harvard classification the type Q is reserved for the "new stars" which have passed their paroxysm of bright- ness. Oa is the designation of Wolf-Rayet stars.
409
THE SUN
Class 0 has other subscripts b, c, d, and class M has subscripts b, c, not included in the table.
TABLE XXXI. — Classification of stellar spectra
STAB |
Harvard |
Vogel |
Sccchi |
|
(q Argus, or Carina) (7 Argus, or Vela) (29 Canis Majoris) |
oQ. O0 |
Ib Ib |
- |
|
Alcyone Sirius Altair Canopus Procyon The Sun Arcturus AMebaran |
(\ Orionis) (8 Orionis) fo Tauri) (a Canis Majoris) (a Aquilae) (a Argus, or Carina) (a Canis Minoris) Also Capella (a Auriga?) (K Geminorum) (a Bootis) (a Tauri) |
Oe5B B B5A A ASF F F5G G G5K K Kf>M |
Ib Ib Ib I.S-II. I.S-H. Ia3-Ha 11. Ila-HIa H.-IH. Ha-IIIa |
I I I I I I I II II II IT |
Betelgeuse Mira |
(a Orionis) (o Ceti) (19 Piseium) |
M. Md N |
III. IHa Hlb |
III III IV |
In some stars the spectrum of hydrogen assumes form which was, to be sure, predicted from the nu- merical spectrum series relations, but which has never been experimentally produced in the laboratory. We cannot yet tell, therefore, what conditions such stars typify. The striking analogy between the third type spectra and those of sun spots, taken in connection with the proved relatively low temperature of sun spots noted in Chapter IV, indicate clearly a pro- gression of temperature from stars of type II to those of type III as well as of spectrum.
SPECTRAL DISTRIBUTION or RADIATION Wilsing and Scheiner have lately made a long series of spectral photometric observations on stars
410
THE SUN AMONG THE STARS
of Vogel's various spectral types, to see how the dis- tribution of energy in their spectra compares ; and with a view to estimating the temperatures prevailing in the stars by comparison with the distribution computed for " black-body" spectra. Their conclusion is that the temperatures vary regularly with the type among the stars investigated, from upwards of 10,000° of the absolute centigrade scale, to below 3,000°. As there is some question as to the validity of the tem- peratures deduced, I give here what seems a more direct and quite as interesting a summary of their results, namely, the mean spectral distributions for four average stars of each of the seven spectral classes investigated. The spectra have been put equal at wave length 0.448/4.
TABLE XXXII. — Intensities in stellar spectra. (Wilsing and Scheiner.)
TYPE |
STARS |
INTENSITY |
|||
\0?44S |
A0?480 |
A0?584 |
A0f638 |
||
»' ( |
/3 Can. Min., 12 Can. Yen. aDelphini, aPegasi |
jiooo |
836 |
579 |
505 |
Ia2 { |
a Androm. , y Coronae y Ophiuchi, y Lyrae |
} 1000 |
796 |
625 |
525 |
Ia3-IIa.. j |
a Trianguli, £ Geminorum fi Leonis, 8 Aquilae |
} 1000 |
948 |
902 |
845 |
„,. ..{ |
y Pegasi, TJ Leonis p Leonis, £ Pegasi |
} 1000 |
887 |
578 |
530 |
Ha [ |
TJ Bootis, ft Virginis ju. Herculis, y Cygni |
} 1000 |
998 |
993 |
1005 |
Ila-IIIa. { |
a Arietis, o Tauri & Cancri, /3 Ophiuchi |
} 1000 |
1205 |
1766 |
1897 |
Ilia | |
a Orionis, 5 Virginis \ Serpentis, & Sagittae |
} 1000 |
1368 |
3296 |
4406 |
According to the energy spectrum data given in Chapter III the sun's spectrum would fall in their
411
THE SUN
class (Ia3-IIa). The results given indicate that the order of the series of spectra, given by Vogel from an inspection of the character of the Fraunhofer lines, has a strong support from the distribution of inten- sities of the continuous spectra as well. Furthermore, the order given is the order proper to a series of spectra from soifrces of successively lower and lower temperatures.
EVOLUTION OF THE SOLAR SYSTEM
The inquiring mind is ever stimulated by the query: What means the order of the heavens, and can we not draw from it a reasonable view of the evolution of the universe, including the solar system? The famous Laplace, in 1796, crystallized and ampli- fied the conceptions which earlier philosophers had foreshadowed into his famous Nebular Hypoth- esis of the formation of the solar system. As mod- ified a half century later by the discovery of the con- servation of energy, it presumes a gaseous nebula, larger than Neptune's orbit, in primitive rotation. By virtue of its immense extent it contains potential energy of position which is transformed into heat as the nebula condenses, and thus is supplied the energy of radiation. The gravitation of the nebula in con- junction with the occasional collisions of its mole- cules tended, it is supposed, to produce condensation. At certain critical times the revolving mass separated rings, and these by condensation produced the plan- ets. The planets in condensation likewise threw off
412
THE SUN AMONG THE STARS
rings which formed the moons. In Saturn's case rings still persist. The view accounts for the pre- vailing tendency of the planets, their satellites, and the sun, to rotate in the same direction, and for the approximately common plane of their orbits and ro- tations. The exceptions of retrograde motion were not known in 1796, nor were they discussed, so far as is known, by Laplace, in his later revisions of his theory.
According to Chamberlin and Moulton the La- placian hypothesis, even as modified and clarified by the work of Helmholtz, Roche, Darwin, and others, fails conspicuously to account for a number of things. Principal among these are: A. The considerable eccentricities of some of the planetary orbits and the inclinations of their planes among themselves, and with respect to the sun's equator. B. The neg- ative rotation of some of the satellites, and the small periods of revolution of some of them as compared with the periods of rotation of their primaries. C. The difficulty of understanding how rings could be left off in the shrinking of the nebula, whether it were gaseous or meteoric in structure, and the still greater difficulty of understanding how a ring, if left off, could condense into a planet. D. The difficulty of accounting for the enormous discrepancy between the present moment of momentum of the system and that which must apparently have formerly pre- vailed.
Chamberlin and Moulton have proposed the 413
THE SUN
"Planetesimal Hypothesis" of the evolution of the solar system. They might start with a spiral nebula. Since there are millions of these objects in the sky this basis is justified. But the authors have gone even further, and suggested that in the course of ages two stars may approach so near together that they will mutually raise enormous tides. Tides occur in pairs at opposite ends of a diameter. Such tremendous disturbances as thus supposed, together with the eruptive tendencies due to intense heat, would perhaps combine to cause many masses of matter, varying greatly as to quantity, to be pro- jected from each tidal region. The relative motion and gravitation of the two stars would tend to change the motion of projection of the masses into motion of revolution in orbits about the primaries. When the action first occurred, the disturbing star being far off, and the attraction of the erupting star acting preponderatingly, the orbits of the erupted masses would be small, and their periods of rotation short. At closest approach of the disturbing star the con- trary would prevail. The outcome would be a two- branched spiral (see Plate XXV), containing many masses of all sizes revolving in orbits about the parent star (our sun) . As the inner orbits are of less period than the outer, the spiral form will become more and more coiled, and at length cease to present a spiral appearance.
Mutual attraction and collisions among the numer- ous masses would lead to the concentration of the
414
THE SUN AMONG THE STARS
lesser masses and particles on the larger ones, or their revolution about them as satellites. The au- thors show that collisions tend on the whole to de- crease the ellipticity of the supposed orbits, so that the larger planets, on which most collisions have occurred, would have the most nearly circular orbits. The lighter gases would be early lost as atmospheres from the smaller planets and satellites owing to the consequences of the kinetic theory of gases. But with increasing size caused by the accretion of par- ticles in collision, the occluded gases would be forced out of the interior by growing pressure, and so after a time atmospheres would be supplied again to plan- ets of medium size. The large planets would retain the gases from the start as atmospheres.
The numerous fragments called the asteroids re- mained almost unaltered from lack of large masses in their neighborhood to capture them. Their eccen- tric orbits and high inclinations are evidence of the comparative rarity of collisions among them. The retrograde motions and relatively high velocities occurring among the satellites seem to present no difficulty in the view of the authors.
The plane of the sun's rotation they believe to have been modified by the falling back of much ejected material not forced into clear orbits. Prob- ably the original plane of rotation was at consider- able angle to the present, but has been brought nearer the average plane of the planetary orbits by such collisions,
415
THE SUN
Perhaps the greatest difficulty of the hypothesis • is to account for the supply of solar radiation during the immense period of time that the earth, as shown by geological evidence, has retained practically its present dimensions and form, and its present tem- perature. Moulton assumes, as the only explanation available, and one which he thinks is also required by the Laplacian theory, that probably the contrac- tion theory of the sun's heat accounts for only a small part of the solar energy. It would seem as if the Laplacian theory had a great advantage here, for it presupposes the general extension of the nebula beyond the orbit of the earth, when the earth began to form. Hence there was an immense store of en- ergy to be gained by contraction. On the contrary, the spiral nebula of Chamberlin and Moulton appar- ently had no such general extension, but retained nearly all of its matter from and after the catastrophe in the center of things. Furthermore, the general extension of the nebula of Laplace enables us to sup- pose that the earth was for a very long time receiving radiation from a large portion of a hemisphere, or even (by reflection within vestiges of the nebula) from a sphere, so that we need not suppose that the intensity of this radiation was great, and therefore we can assign a very much longer life to the contrac- tion source of energy than we could if we were obliged to think of the solar radiation as always requiring to be at its present intensity, during geologic time, in order to maintain terrestrial temperature.
in;
THE SUN AMONG THE STARS
Prof. T. J. J. See has just published (after the above resume of nebular hypotheses was written) his vol- ume on this subject. In his view the sufficiently close approach of two stars to form a spiral nebula, as assumed by Chamberlin and Moulton, is too in- frequent to deserve consideration. He would assume the spiral nebulae to be formed by the close approach of two nebulous streams, and the curling of them together by mutual gravitation, or by the curling up of a single nebulous stream owing to its own grav- itation, but he does not show that such phenomena are apt to happen more frequently than that sug- gested by Chamberlin.
Such a spiral nebula is, according to him, the parent of the solar system, but unlike Chamberlin and Moulton, his nebula would not have its central con- densation, the sun, mainly formed before the planets began to form, but all would be forming at the same time, by capture of particles by larger masses in the exercise of mutual gravitation, and in the vicissitudes of mutual encounter between the larger and smaller bodies of the nebula. There seems to be much in common between this " capture theory" and Moulton and Chamberlin's accretion theories. See finds that the orbits of the planets will be rounded up by the resistance (that is, the continual encounter with par- ticles) which they find in the nebulous medium. Here he is in close accord with Chamberlin and Moulton, who have found, as stated above, "that collisions tend on the whole to decrease the ellipticity of the supposed
417
THE SUN
orbits, so that the larger planets, on which most col- lisions have occurred, would have the most nearly circular orbits." But See believes further that the present orbits of the planets are very far within the orbits they had when they were principally formed. It would seem on the whole, that, excepting in the method of forming his nebula, Professor See's views follow the general lines laid down by Chamberliri and Moulton, but with, this difference that they allow the sun to be forming at the same tune as the planets, and in a very extended space, so that the problem of supplying energy by contraction for con- tinuing the solar radiation in ample measure through- out geological time is easier for See than for Chamber- lin and Moulton. See's conception also permits us to suppose the solar part of the nebula was so much expanded as to shine upon the earth from a large angle in the earlier geological epochs, as was required for the foundation of what we have termed " Hypo- thesis (B)" in Chapters VI and VII.
STELLAR EVOLUTION
We will now consider a little more closely the gen- eral view that nebulae are stars in the making, and that the stars progress through a series of tempera- tures, and at length, like the earth and moon, reach a cold final condition. Plates XXII to XXVI give a series1 of nebulous forms ranging from the chaotic
1 It is very questionable if we should interpret this scries of forms as implying a series in order of development. I am greatly indebted to my friend, Mr. G. W. Ritchey, for this fine group of photographs.
418
PLATE XXII.
THE GREAT NEBULA IN ORION. (G. W. Ritchey.)
Photographed with the 2-foot reflector of the Yerkes' Observatory, 1901, October 19. Exposure 1 hour. Cramer Crown plate.
THE SUN AMONG THE STARS
nebulae in Orion and Cygnus to trhe well-developed spiral and ring forms of Andromeda and Lyra. The number of nebulae observable with the Mount Wilson reflector probably reaches into the millions. In Plate XIX we saw that the Pleiades stars are plainly wrapped in nebulosity, and seem as if still in process of condensation. This peculiarity is shared by other star groups, notably by some in Orion. The spectra of the Orion stars have a simplicity far more in common with the simple spectra of gaseous nebulae than with the lined and banded spectra of the solar and Antarian stars. Stars of the Orion type have in many instances nebulous appendages, and besides seem to be of extremely small density, according to the tests we have noted above. Hence, it is supposed that the first evolutionary step is the passage from a nebula to a helium star. Nevertheless, it is found that the great spiral Andromeda nebula gives at its center an essentially solar type of spec- trum.
But even admitting the connection of cloudlike nebulae and helium stars, why should we believe that the nebula is the first and not the last end of the chain, in point of time, or that the other types of spectrum have the same order in their secular de- velopment as they do in our arrangement of them according to their physical appearance? As to the first branch of the question, we know that gravita- tion tends to condense matter, whether by capture as of the meteors by the earth, by the opportunities
419
THE SUN
offered in molecular collisions, as required by La- place's Nebular Hypothesis, or by the collision of meteors in orbits, as proposed by Chamberlin and by See. In any of these cases the centrally directed force of gravitation inevitably seizes its opportunity to draw in the retarded particle. Excepting the dis- ruptive tendency of the close approach of two stars invoked by Chamberlin and Moulton, and the escape of gases by molecular activity according to John- stone Stoney, there is not known any cause for the separation of the constituents of a star into a nebula. This would require an enormous expenditure of en- ergy, whose possible source, except as just indicated, it is hard to conceive. The probability of the close approach of two stars would seem at first sight to be very small, for Newcomb has computed that on the average a sphere of radius 412,500 times the radius of the earth's orbit contains but one visible star. On the other hand, there may be enormous numbers of invisible bodies, and even the number of stars in space is so large that such near collisions may actually occur rather frequently, measuring time by centuries. We shall recur to the question of the order of events in stellar evolution.
Admitting the view that nebulae generally tend to condense, not to expand, their rise of temperature with condensation, if gaseous, was proved by Lane in 1876. If we adopt the usual view that yellow stars are more advanced than the blue ones, how are we to explain the circumstance that the blue stars, which
12Q
PLATE XXIII.
NEBULA N. G. C. 6992 CYGNI. (G. W. Ritchey.) Photographed with the 2-foot reflector of the Yerkes' Observatory,' 1901, October 5. Exposure 3 hours. Cramer Crown plate.
THE SUN AMONG THE STARS
seem to be nearest the nebulae in type, are by Wilsing and Schemer's observations, and by an appeal to ordinary experience, apparently hotter than the yellow ones? As a reply to this objection we must note that whether with most astronomers we accept a photosphere, or assume a purely gaseous sun as discussed in Chapter VI, the inner parts of the sun, or of a star, are not visible to the observer. The inner parts may in fact be hotter for yellow than for blue stars, without in any way altering the succession of apparent surface temperatures found by Wilsing and Scheiner, for different type stars.
But it is by no means clear that a yellow star is necessarily older than a blue star in actual time, and indeed it does not seem necessary to admit that every star of the helium or hydrogen type of spec- trum will necessarily, with lapse of time, become a solar or Antarian star. The similarity of spectrum lines proves that certain elements found in the earth exist in the sun and in the stars. When stars fail to exhibit any of the spectral lines of an element we can- not know that this element exists in those particular stars, for we are not fully justified in supposing that it does so on the assumption that conditions do not favor the production of its spectrum. It may pos- sibly be, then, that Sirius, for instance, will never show a solar type of spectrum, however cold it may grow superficially.
I develop this line of thought for consideration in connection with the discussion and catalogue of
421
THE SUN
spectroscopic binaries recently published by Prof. W. W. Campbell.1 The observed spectroscopic binary stars range in period from less than a day to more than a year, and visual binaries carry the range of periodic times up to thousands of years at least. Campbell, in summarizing the existing observations of spectroscopic binaries, draws attention to certain relations between the periods of orbital revolution, eccentricities of orbits, and types of spectrum. The following table shows these results:
TABLE XXXIII. — Spectroscopic binaries. Spectral types, periods, and
eccentricities*
Periods •* |
" short " 8 short |
Od-6d 16 2*4 (10)0.04 |
5d-10d 10 6'|9 (5)0.10 |
10d 14 73d2 (11)0.34 |
yean 1 1.9 (1)0.0 |
"long" long |
Oand BTyjkM Mean Period |
||||||
Mean Eccentricity |
||||||
A Types Mean Period. |
4 short |
10 2(.1r>5 |
1 0*2 |
12 42d2 |
2 26.45 |
|
Mean Eccentricity |
(5)0.04 |
(1)0.50 |
(S) 0. f).-, |
(1)0.59 |
||
F Types Mean Period |
0 |
G 3d! ( l) o.o.-) |
2 5dO (1)0.01 |
4 145dl (3)0.15 |
8 11.1 (3)0.44 |
1 long |
Mean Eccentricity |
||||||
G to M Types Mean Period |
0 |
0 |
0 |
3 104d8 |
9 24.3 |
13 long |
Moan Eccentricity |
(2)0.06 |
(8)0.38 |
||||
Total Mean Period |
12 short |
31 2d-.!» |
13 6d90 |
33 7 3d.-> |
15 20.5 |
14 long |
Mean Eccentricity |
(19)0.04 |
(7)0.14 |
(24) 0 . 30 |
(13)0.38 |
* From Lick Observatory Bulletin No. 181.
This summary shows clearly that the "earlier" types of spectra are associated in spectroscopic binary
Observatory Bulletin No. 181. Also "Pub. Astr. Soc. Pacific," April, 1910.
422
PLATE XXIV.
THE GREAT NEBULA IN ANDROMEDA. (G. W. Ritchey.)
Photographed with the 2-foot reflector of the Yerkes' Observatory. 1901,
September 18. Exposure 4 hours. Cramer Crown plate.
THE SUN AMONG THE STARS
stars, as a rule, with shorter periods arid smaller ec- centricities than are the " later" types of spectra. Campbell also gives a table of fifty telescopic binaries arranged in order of their periods in five groups of ten each. The periods range from 5.7 years to 194.0 years, and not one of the stars named (which gener- ally is the principal star of the pair) has a spectrum of the 0 or B type, while many specimens of types A, F, and G, and some of K, are found. As for the eccentricities, these are all large, averaging 0.461, 0.453, 0.495, 0.531, and 0.483 in the five groups. The general average period is seventy-two years, and average eccentricity of orbit 0.49.
In summary for the telescopic binaries:
SPECTRAL TYPE |
No. of Stars |
No. of Stars |
Mean period |
Mean eccentricity |
|
O-B. . |
0 9 |
Short periods. . . . |
25 |
32.8 |
0.48 |
F G-K.. |
18 14 |
Long periods .... |
25 |
108.1 |
0.51 |
M-N Unknown. . . |
0 9 |
In the words of Campbell: " Visual double stars clearly abhor the O and B types, and visual double stars of relatively short periods clearly abhor M and N types.
"What, " says Campbell, "is the significance of these facts? Let us recall that Darwin and Poincare studied the origin of binary stars from theoretical considerations, and came to the conclusion that a
29 423
THE SUN
condensing nebulous mass, rotating on its axis con- stantly faster and faster, to keep pace with loss of heat by radiation, should eventually separate into two nebulous masses revolving around their mutual center of mass. These two masses would, in the beginning, be revolving in contact in orbits essen- tially circular. With advancing time, tidal disturb- ances should cause the two bodies to draw apart rap- idly at first, and less rapidly later. In the spectro- scopic binary systems described — have we not a tol- erably complete sequence of orbits illustrative of the Darwin-Poincare hypothesis? The short-period or- bits should be circular or nearly so, and should appertain preferentially to stars of early spectral types ; the longer periods should, in general, attach to the more eccentric orbits and the older spectral types; and these are the facts. established by actual observation of binary systems. ... It will be noted that in these widely separated (telescopic binary) sys- tems there is not a single 0 or B type, representing the early stages of binary existence. There are a few A types, but the major number are of the advanced F type and G and K types. I suspect the K, M, and N types are not more fully represented for the reason that in these old-age systems the two components are in general so far apart that the periods of revolu- tion are many hundreds or thousands of years."
Campbell considers also the relative masses of the two components in the binary systems for which this is known. In seventeen cases where the components
m
PLATE XXV
SPIRAL NEBULA M. 51 CANUM VENATICORUM. (G. W. Ritchey.)
Photographed with the 5-foot reflector of the Mount Wilson Solar
Observatory. Exposure 10% hours. Seed 23 plate.
THE SUN AMONG THE STARS
are of unequal masses he finds that with one excep- tion the lesser component is of an " earlier type of spectrum, or bluer, than the more massive one." Here, at first sight, is a most surprising thing. Of two stars admittedly of equal age, the one of greater mass is in general the more advanced. Campbell says: "An hypothesis of Huggins, suggested at first rather casually, and later discussed more seriously, appears to me to be of great merit, especially when Schuster's extension of the hypothesis is applied. Huggins' suggestion is as follows: 'Another way of looking at the problem is perhaps possible. May it be that the effect of the great mass on surface density, together with the working of Lane's law, by which the temperature of a condensing gaseous mass so long as it is subject to the laws of a purely gaseous body will continue to rise, will favor in such stars the coming in of a solar type of spectrum at a somewhat rela- tively earlier time?' Schuster's extension suggests in effect that the lighter gases — hydrogen, helium, and so on — which surround a star in its early age, will be pulled down on a star of small mass but lightly, and a long period will be required for the absorption of these gases. Such a star would remain effectively young, as judged by its spectral type, longer than its more massive primary. In the latter, the greater gravitational power would lead to more rapid absorp- tion of the lighter surrounding gases, and the pre- dominant influence of the metallic absorption would enter earlier. It seems reasonable to suppose that
425
THE SUN
the greater internal gravitation of the more massive primary will generate heat more rapidly, and cause it to live its life more rapidly than in the case of the less massive secondary."
Should not this explanation of the prevailing ten- dency to earlier types of spectra for the lesser com- ponents of binary stars take more explicitly into con- sideration a difficulty suggested by ordinary experi- ence of cooling bodies, Planck's law of spectral energy distribution, and Wilsing and Schemer's results on the relative temperatures of the stars? For it is the blue stars which we should suppose to be, and which Wilsing and Scheiner find to be, superficially hottest, and as the bluer stars are generally supposed to be also of less density than the yellow ones, their sur- faces are also greater in proportion to their masses. Hence, if their radiating coefficients are equal to those of yellow stars, they should radiate more rapidly and advance more rapidly in spectral type thereby, if, as is often assumed, advance in spectral type is a mere function of radiation and consequent condensation.
I venture to suggest that if the view of Campbell as to relative masses and types of spectra is well founded,1 its significance in this respect may be the following. As we do not know that the two compon- ents of a binary are of similar constitution, may it not be that the bluer component has a smaller coeffi-
1 Not all aslronomsrs are agreed that it is the general rule for the smaller component of a binary to be less advanced in spectral type, but Campbell's review of the evidence seems very convincing.
420
PLATE XXVI.
FIG. 1. NEBULA H. V. 24. (G. W. Ritchey.)
FIG. 2. RING NEBULA IN LYRA. (G. W. Ritchey.)
Photographed with 5-foot reflector of the Mount Wilson Solar Ob- servatory. Exposures: Fig. 1, 5 hours, Seed 23 plate. 1910, March 6. Fig. 2, 45m Seed process plate. 1910, July 1.
THE SUN AMONG THE STARS
dent of radiation than the other? By this I mean that if the two objects were of equal size and temperature, the blue star would emit less radiation. We have no evidence whether this state of affairs occurs for bodies of stellar temperatures, but decided differences of radiating power were found by Paschen and others at moderately high temperatures for various solid substances, some of which might even have been expected to be approximately "black bodies." As- suming this explanation, the blue stars might radiate slower, even though of decidedly higher temperatures and larger surfaces in proportion to their masses than their yellower neighbors.
As to the assumed difference of constitution of the two components, Campbell has suggested that they were originally one object, which separated owing to too rapid rotation. In such a case might not the smaller object usually carry with it a preponderance of the lighter elements which composed the original star or nebula? We have seen in Chapter VI that the lightest elements lie furthest out in the sun, and it seems reasonable to suppose that the same holds in the case of a just separating binary, so that per- haps they might tend to accumulate in the bulging- out component of smaller mass. If this is so, then the presence of a chromospheric type rather than a photospheric type of spectrum or, in other words, the assuming of the spectrum of early stellar type, should naturally be associated with the lesser component, because it has preponderatingly the light elements,
427
THE SUN
hydrogen, helium, etc., rather than the heavy metals whose lines throng the solar spectrum. I do not mean by this to say that the lesser star has none of the heavy elements, and the greater star all of them. Rather that the lesser star has so large a supply of the lighter elements that they effectually screen by their scattering of light l the radiation of the heavier ones lower down, just as it is probable that the elements of the platinum group are obscured in the sun as ex- plained in Chapter VI. In the greater star the ele- ments hydrogen, helium, etc., while present, I sup- pose to be less plentiful, so that the heavier metals occupy practically a surface position, and, hence, give their typical spectra.
But it will be urged that this view implies too much, and does not take into account the progressive change of spectral type shown to occur with increas- ing age of binaries. In other words, that it would imply that, once a blue star, never a solar star. Before answering this objection let us examine Table I of this book, which shows that the four outer, and according to the Laplacian hypothesis probably old- est, planets of the solar system are all of low density, even lower (notwithstanding their probably low tem- peratures) than that of the enormously hot sun, and four times as low as the densities of the four inner planets. May not some support of the view just advanced be gained from this circumstance? Were not these planets constructed from the solar nebula,
1 See Chapter VI. 428
THE SUN AMONG THE STARS
accumulating with them a preponderance of the lighter solar constituents? If so, did not their forma- tion tend to advance the type of the solar spectrum? May not other stars, binary stars not excluded, whether primaries or secondaries, also give rise to relatively small planets and satellites, thereby losing their lighter surface materials, and hence advancing themselves in spectral type? Such planets, if of small masses relatively to their primaries, could not be observed, and hence may, for all we know, exist.
Dr. Johnstone Stoney showed many years ago that the lighter elements gradually escape from atmos- pheres according to the kinetic theory of gases, and the more rapidly the higher the prevailing tempera- tures. He explained by this means the absence of water vapor from Mars, -of all gases from the moon, and of helium and hydrogen in marked quantity from the earth. May not this process, favored as it must be by the high temperatures of the blue stars, aid in course of ages to divest them of hydrogen, helium, etc., and thereby tend to advance their type?
Without wholly accepting until we have fuller evidence the relation pointed out by Campbell in regard to spectral types and masses of binaries, the suggestion just made as to a possible path of stellar evolution is, of course, not limited in its application to the cases of binary stars. It may be that all the blue stars are of early spectral type because their elements of high atomic weight are obscured by the lighter gases, hydrogen, helium, etc., and that with the es-
429
THE SUN
cape of these gases to space or the formation of satel- lites, the spectral type will be advanced to the solar stage, and from that by cooling to the Antarian.1
Returning from these perhaps too presumptuous digressions, I shall finally call attention to some data noted by Prof. Kapteyn as perhaps " valuable in the classification of the stars in the order of their evolu- tion. "2I He remarks first the progressive increase of " peculiar" stellar velocities3 for stars of the advan- cing spectral types. In the following little table he sums up the results thus far available.
TABLE XXXIV. — Spectral types and velocities in space
Type of spectrum or object |
Peculiar radial velocity per second |
Number |
B to B 9 |
km 6 5 |
64 |
A to A 5 |
12.6(11.2) |
18 |
Fto F.. |
14 5 |
17 |
G to G 5 |
12 6 |
26 |
KtoKS |
15.4 |
55 |
Ma |
19.3 |
6 |
Planetary nebula |
''6 S |
13 |
Orion nebula |
0 1 |
1 |
N |
13 1 |
8 |
L |
:') 7 |
9 |
1 Consult, in this connection, T. J. J. See's "Researches," vol. ii, p. 589.
1 Contributions, Mount Wilson Solar Observatory, No. 45.
* By this is meant "the velocity freed from that part which is due to the motion of the solar system through space."
430
THE SUN AMONG THE STARS
He then discusses the " peculiar" proper motions of over 2,000 stars and finds the ratio of their average magnitudes to the solar motion as follows :
Ratio Number
Type I (B, A) 1 .02 1144
Typell(FtoK) 1.46 1093
Type unknown 1 . 45 381
He concludes that the ratio
Average linear velocity of the F, G, K stars
Average linear velocity of the A stars cannot be smaller than 1.3.
It is greatly to be regretted that there are not more radial velocities for nebulae of the Orion type known as yet, but the evidence seem's to indicate that such a nebula is to be regarded as practically stationary, and that the stars of advancing spectral types are affected by progressively greater and greater motions in space, but that the planetary nebulae are not to be classed with the nebulae of Orion type, but at the other end of the chain.
Kapteyn says:1 "The phenomenon of the increase of velocity with the evolutional stage of the stars must give rise to speculation as to its cause. The observational results contained in our table naturally lead us to conclude that the matter from which the stars originate must have little or no velocity. How is this possible under the influence of the combined attraction of the rest of the system? Is it not as if2
1 Contributions of the Mount Wilson Solar Observatory No. 45.
2 " We need not necessarily make the hypothesis that really primor- dial matter is not subject to gravitation. If, for instance, as was suggested to me by a friend, the tenuity of this matter were such that
431
THK SUN
gravitation had no effect on the cosmical matter in its primordial state? If this be so — as soon as matter changes from this state to another in which gravity begins to act, or to act freely, motion will arise, and it is evident that, as a rule, the motion must be acceler- ated, at least during immense periods, so that the longer the period elapsed since the birth of the stars the greater must be their average velocity."
After calling attention to the argument from binary systems, which we have already considered, Kapteyn mentions the two great star streams which, according to his researches, embrace the stars, and notes that, as shown by Dyson, the stars of type I " diverge less from the general drift of the two streams than the other stars. " Such a result harmonizes with the view that the Orion stars are relatively young. But, says Kapteyn: "Not only this. Observation shows fur- ther that for the Orion stars the stream velocity is small ... as compared with . . . the rest of the stars. Apart from the advantages that we may derive from this result for the classification of the stars in the order of their evolution, it has, I think, a great im- portance in its bearing upon the question of the gen- it were very materially hindered in its motion by the matter which we must assume as filling the universe in order to explain the phenom- enon of selective absorption of light recently found, the velocity of this matter could not exceed the value for which the resistance is equal to the total attraction. . . . Other suppositions may probably be made of forces which, in the primordial state of matter, counter- art gravity. But it is evident that in such cases where gravity is just counterbalanced by another force, things happen r/.s /'/ there wen- no force at all." (May not light pressure be such a force?)
432
THE SUN AMONG THE STARS
eration of the star streams themselves. For it proves that the streaming motion, too, is not an initial motion, but one generated at an epoch which, for the stars of any one type, must be placed at a time rel- atively but little preceding the time when they passed through the Orion-type stage."
The results of Kapteyn aid greatly to convince us that the progress of evolution is from the Orion type of nebula at the beginning, to the fourth type star at the end, and not the opposite, in the lapse of time. For we shall see from them that there is a real prog- ress from one stage to another, marked by the gradual march of velocities. It only remains to show that the march is in the supposed direction and not its opposite, and for this purpose the study of one part of the course is as good as another. Now we know that the stars of the second type resemble the sun's photosphere, and those of the third type the sun spots in their spectra, and that this difference is brought about in the sun by the mere reduction of temperature. A reduction of temperature, however, must finally occur when a star exhausts its sources of energy. Hence, the third type stars must probably be a later stage of evolution than the second, and the progress of evolution is therefore from the Orion nebula to the fourth type star. This conclusion is supported also by Campbell's discussion of binary stars. »
Various considerations, then, recommend the view that the stars are formed from nebulae, take first the
433
THE SUN
Orion type, and pass on with age to the solar class, and thence, with cooling, to the Antarian stage anal- ogous with sun spots in spectrum. We may suppose that a still more advanced, and usually final stage, is the cold one of which the earth and moon are types. It is speculating far from the sure ground of observa- tion to say it, but do not these conclusions, and es- pecially Kapteyn's discussion of the velocities of nebulae and stars, indicate that the entire stellar system arose from a sort of formless, relatively motionless, chaos, and will at length reach a dark and unknown end?
CONCLUSION
In our study of the sun and its relations with the earth and stars, the discoveries of two types of inves- tigators have come prominently before us. To one class belong those geniuses whose roving minds in- cline them to try this and that new thing, and whose acute perceptions enable them to turn even their most random observations into glorious discoveries. In- vestigation to them is like happy exciting play to a child. These have their place and their ever-present reward. To another class belong the patient ob- servers and philosophers who, from a love of science and a sense of duty to their age and to posterity, have gradually enlarged by tedious observation and laborious analysis that precious store of exact knowl- edge whose value time cannot impair but can only enhance. The men of both classes are deserving of admiration, the former for their brilliancy, the latter for their perseverance. As those of the former class are continually receiving their meed of praise from their contemporaries, it will not be amiss to offer our tribute to the others, and recall to mind the work of Newton, whose immortal "Principia" he suffered to remain unknown until by the importunity and finan- cial means of his friend Halley it came at last to pub- lication; Laplace, Gauss, Hansen, Newcomb, and
435
THE SUN
many more who erected the wonderful edifice of mathematical astronomy on the foundation of New- ton's law of gravitation; the long series of observers from Galileo down, whose sun spot records were com- bined by Wolf with such rare skill, after the patient work of twenty years by Schwabe had indicated the sun-spot cycle; Carrington, Spoerer, and the others whose numerous observations revealed the law of rotation of the sun; Kirchhoff, A. Angstrom, and our own wonderful Rowland, whose spectrum researches are the foundation of solar physics; the unremem- bered army of meteorological observers whose plod- ding records are sometimes scoffed at by the more brilliant, but which nevertheless share in the en- hancement of value produced by generous Time; Bradley, the father of exact stellar observation, whose thousands of accurate star places are priceless to modern astronomy; Argelander, whose enormous work, the " Durchmusterung " of the northern stars is yet in the prime of its usefulness; Huggins, whose pioneer, yet long sustained, investigations in as- tronomical spectroscopy laid the foundation of the study of stellar evolution.
These men and many more who were actuated by the same motives have passed on, but their work still lives. There still remains, and ever will remain in solar and stellar investigation, room for such work; and on the thorough doing of it in our time the won- derful flowers of future discovery, whose beauty our eyes cannot see, or our imaginations picture, must
436
CONCLUSION
largely depend. If we now had such long, unbroken, and accurate series of meteorological records of nu- merous stations in all parts of the world, on land and sea and in the air, as posterity must depend on us to supply; if we now had those long-kept, numerous, and accurate observations of stellar parallaxes, brightness, forms of spectra, velocities, and other data which Kapteyn longs for, but in vain; if we now had accurate measurements of the solar constant of radiation going back centuries ; in short, if we could, as we find the need of it, consult the records of the Past to verify the surmises of the Present, then solar and stellar knowledge would advance with such leaps and bounds that we could soon see the great pano- rama of the universal evolution unroll before us.
The child is said to long to grasp the moon. Who, in his maturer years, has never wished that he might stand upon the moon, and watch the earth at full, a glorious planet of the night, four times as far from rim to rim, and twice as bright in every part as is the moon herself! Who, thinking more gravely, has not wished sometimes he had been born in later years, when he could share the fuller understanding yet to come? Shall we not live in hope that if we worthily contribute to that happy end, we, too, may join with that great company whose patient and sound labors have given us what we know, and in a future life with them may see unrolled the wider view which here we long to see in vain?
INDEX
Abbot, C. G., 108, 190. pyrheliometer, 79. solar theory, 236-279. Mrs. C. G., 132. Aberration of light, 22. Absorption, and radiation, G4.
atmospheric, 291, 308. Adams, W., 379.
W. S., 25, 100-104, 124-127, 180-182, 205-209, 213, 255-258, 268. Air mass, 286-290. Anderson, 233. Angstrom, A, 436.
K., pyrheliometer, 77. Apex of solar motion, 395. Arago, 142. Archimedes, 363. Arctowski, 190, 317. Argelander, 436. Armstrong electrical machine,
213. Assimilation of carbon by
plants, 335-344. Atmosphere, earth's, extinction,
74, 294-297. height, 74. spectrum, 88.
Atomic weights and spectrum intensities, 91, 93, 169, 253, 317. Aurora and sun spots, 189, 220.
Ball, R., 222.
Balmer series, 170, 173.
Becker, G. F., 273, 278.
L., 132, 133. Belopolsky, 135. Bessel, 392.
Bigelow, 111, 135, 190, 317. Body, "black," 64, 65. Bolograph, 83, 292. Bolometer, 81, 84, 86. Boss, 396.
Bouguer formula, 74, 289, 293. Boussingault, 358. Boyle, Willsie and, 376-379. Boys, C. V., 28. Bradley, 396, 436. Brown, 358.
and Escombe, 339.
E. W., on Moon's motion,
7, 16.
Buffon, 363.
Buisson, 96, 99, 104, 256. Bunsen, 38, 39, 87, 98, 349. Burning mirrors, 363.
Calcium, circulation in sun, 101.
hydride, 209.
level, 97, 119, 126, 127.
lines, wave length, 102
spectroheliograms, 118. Calorie, 68.
30
439
INDEX
Calvert, 132.
Campbell, 135, 169, 180, 203,
395-398, 406, 422-429. Carbon, in sun, 93.
dioxide, absorption and radi- ation, 281, 291. and plants, 335-344. Carrington, 32, 122-126, 153,
192, 198, 436. Cavendish, 28. Cebrian, Molera and, 375. Chamberlin, 274, 276, 330, 413-
418.
Chase, 393. Chemistry, of stars, 405.
of sun, 87, 91. Chevalier, 124, 126. Christie, 195. Chromosphere, flash spectrum,
168, 177-179. heights, 171, 172. in daylight, 137, 142, 149. spectrum, 88, 143, 145, 180,
233.
in daylight, 180, 181. Clark, 381.
Coast survey, U. S., 18. Coelostat, 36.
Comparator measurements, 61. Convection currents in sun, 100,
267.
versus radiation, 103. Cook, Captain, 14. Copernicus, 1, 392. Cornu, 21. Corona, solar, 131-136, 263-
265.
Coronium, 134. Cortie, 210. Cranks, 8. Cyanogen in the sun, 94.
Darwin, G., 413, 424.
I)e Candolle, .'555.
De Saussure, 364, 374.
Dcslandres, 117, 135, 167.
Diffraction, 51, 52.
Dispersion, anomalous and reg- ular, 233.
Dixon, Mason and, 14.
Doppler effect, 23, 41, 42, 88, 100, 124, 257, 397, 401, 422, 430.
Dorsey, 21.
Double reversal, 147, 148.
Duner, 124, 126.
Dyson, 135, 169-171, 432.
Earth, dimensions and mass,
18, 27. temperature and radiation,
308-316.
and sun's variability, 317. Ebermeyer, 340. Eclipse, solar, 128-131. of 1868, 137. of 1900, 132. of 1905, 132, 173-179. of 1908, 133, 135. Efficiency, thermodynamic, 387. Elements, chemical, boiling
points, 238. in sun, 87, 89-91, 252. represented in flash-spec- trum, 169, 170, 177, 179.
Elkin, 12.
Ellerman, 118, 120, 167. Emission, selective, 71. Encke, 15.
Eneas, solar machine, 367-372. Energy, available quantity, so- lar, 383-387.
440
INDEX
Energy, spectrum, solar, 68, 82,
292.
over sun's disk, 107, 109, Engelmann, 342. Enhanced lines, 100, 104, 105,
207.
Ericsson, 366. Eros, 12, 16, 17. Escombe, Brown and, 339. Etiolation, 345. Evershed, 99, 115, 166-169, 177,
209, 210, 213, 244, 255. Exner, 300-304. Eye, as optical instrument,
84. Eye-pieces, solar, 33, 34.
Fabricius, 183. Fabry, 96, 99, 104, 256. Faculae, position and appear- ance, 85, 202. Faye, 123. Fenyi, 152. Fizeau, 19. Flacculi, 118, 121.
motion, 122. Flash-spectrum, 168, 170.
of 1905, 173-180. Fluorescent spectrum, 264. Fluorite, transparency, 56. Foucault, 20. Fowle, 108, 190, 317. Fowler, 209. Fox, 100, 122, 124, 126, 198,
270. Fraunhofer, 39.
lines, 82, 87, 98.
principal solar, 90. Fresnel rhomb, 44, 211. Frost, 169, 172, 204, 205, 398, 407.
Gale, 208, 209.
Galileo, 1, 183, 436.
Gas, levels in stars, 426-429.
spectrum, 71, 249. Gauss, 435. Gegenschein, 400. Geography, plant, 346-354. George III., 13. Gill, 11.
Glacial periods, 273-277, 322. Glass, transparency, 56, 375. Gould, 317. Grant, 142.
Granulation, solar, 202. Grating, action, 51.
concave, 54.
plane, 55.
ruling, 52. Gravitation, 2.
constant of, 28, 29.
is universal, 400.
Hale, 59, 61, 117, 118, 124-126, 181, 182, 208, 209, 211, 212, 268. Halm, 24, 124, 126, 185, 256,
271, 397.
solar theory, 225-228. Hann, 312. Hansen, 15, 435. Hartmann, 234.
Harvard, classification of stel- lar spectra, 409. college observatory, 23. Hastings, 221, 226. Heat and radiation, 64. Heliomicrometer, 60. Heliotropism, 354. Helium, in chromosphere, 170, prominences, 139. stars, 409.
441
INDEX
Helmholtz, 226, 248, 277, 278,
413.
Hercules, constellation, xx, 296. Herschel, 137, 141.
J., 33, 364.
W., 396. Hinks, 13. Hipparchus, 5. Hoesen, 363. Holden, 135.
Hot-box, 364, 376-379, 389. Hough, 397.
Huggins, 141, 142, 425, 436. Humphreys, 43, 99. Hydrogen, in stars, 409.
in sun-spots, 206.
level in sun, 97, 119, 127.
spectroheliograms, 118.
Insolation, march of, 285. Intensities, spectrum, and at- omic weights, 89, 91, 93. Interferometer, 96. Iron, level in sun, 97.
Janssen, 137, 138, 142, 168, 203. Jewell, 43, 99, 115, 169, 171-
173, 255. Josse, 278.
Julius, xxiii, 182, 259, 260, 262. Jupiter and sun-spots, 188. satellites and sun's distance,
23.
Kapteyn, xxiv, 396, 430-433,
437.
Kayser, 93, 254. Keeler, 232. Kepler, 4-6, 27, 402. Ketchum, 376-378. ,', 208.
Kirchoff, xxiii, 38, 39, 64, 87,
98, 4:;ti.
law, (if,, 71, 208. Knicp, and Minder, 344. Knopf, 232. Koppen, 190, 317. Krigar-Menzel, 29. Kustner, 24.
Lambert, formula, 74. Lampblack absorption, 386. Lane, law, 224, 255. Langley, 81, 130, 200-206, 226,
295, 311, 373, Laplace, 6, 412, 413, 435. Layer, reversing, 88, 170, 178. Levels of spectrum formation,
97, 102, 118-120, 126,
127, 171. Leverrier, 16. Lewis, 135, 136, 263. Light, and plant growth, 342-
346. direction and plant growth,
354-357.
velocity, 19-21, 46. year, 294.
Limb spectrum, 105. Littrow spectroscope, 58. Lockyer, 90, 137, 139, 142,
147, 168-171, 185, 195,
196, 222, 224, 255. Lord, 169.
Magnesium hydride, 209. Magnetism, effect on spectrum,
43.
in sun-spots, 211-213. terrestrial and sun-spots,
187-191. Manson, 273.
442
INDEX
Mars, and solar parallax, 11. Mascari, 195. Maskelyne, 14, 27. Mason and Dixon, 14. Maunder, 126, 189, 198. Measurement, spectrum, 61. Michell, 28. Michelson, A. A., 21, 96.
W. A., 79, 260. Minder, Kniep and, 344. Mira Ceti, variability, 404. Mirrors, burning, 363.
reflecting power, 388. Mitchell, S. A., 168-172, 180, 181, 255.
flash-spectrum of 1905, 173-
180.
Mohler, 43, 99. Moissan, 237, 239. Molera and Cebrian, 375. Moore, 133. Motion iu line of sight, 40, 127,
213, 397, 422. Mouchot, 366.
Moulton, 271, 330, 413-418. Mount Whitney, 69, 296-298. Mount Wilson, xxiii, 57, 69, 89, 118, 121, 210, 296-298.
Solar Observatory, 57, 60.
Natanson, xxiii, 305, 343. Naval Observatory, U. S., 132,
173.
Nebulae, and stars, 418-421. solar, 276, 326, 412-418. velocities, 430. Neptune, distance, xviii, 3.
sun's attraction, 2, 3. Newcomb, 8, 15, 16, 21, 27, 185, 190, 193, 236, 278, 317, 396, 400, 420, 435.
Newton, xviii, 6, 28, 86, 435.
Nicol prism, 44, 211.
Nitrogen required by plants,
334. Nordmann, 190, 317.
Olmsted, 209.
Oltmans, 356.
Ore reduction, 381.
Orion stars, 408, 419, 430-432.
Oxygen in sun, 92.
Parallax, solar, 9, 12-16, 22-25.
stellar, 392. Paschen, 427. Permian glaciation, 273-277,
322-329. Perot, 96.
Perrine, 135, 169, 263. Perrotin, 21. Perry, 130, 152, 198. Perseus, new star in, xix. Pfeffer, 358. Photography of sun, 34, 86.
spectrum, 57.
Photosphere, spectrum, ele- ments, 91. extent, 86. sun's, 85, 217, 220. Pickering, E. C., 226, 228, 403,
409.
Pifre, 366.
Planck, law, 66, 251, 426. Planetesimal hypothesis, 414. Planets, minor, 12, 13, 17.
principal data, 3. Plants and light direction, 354-
357.
and the sun, 331-361. as energy accumulators, 357- 361.
443
INDEX
Plants, chemical requirements, 331-335.
geography, 346-354.
light requirements, 349-354.
rest periods, 348. Plateau, 31.
Platinum black absorption, 387. Pleiades, 398. Pogson, 137. Poincare", 423. Polarization of corona, 134.
of light, 44. Potassium in sun, 92. Pouillet, 73, 74. Pressure, effect on spectrum, 43, 104.
in sun, 98, 213. Pringoheim, 261-264. Prism action, 47-49. Procyon, 40.
Prominences and sun-spots, 149, 153, 155.
classification, 156.
detached, 161.
eruptive, 157, 162, 163. rapid change, 164, 165. spectrum, 162.
in full daylight, 138-142, 149.
magnitude, 155.
number and distribution, 152, 153.
quiescent, 159-161.
solar, 121, 213.
specj;roheliograms, 166.
spectrum, 138-143. Pyrheliometer, errors, 78.
Michelson's, W. A., 79.
Pouillet's, 73.
silver disk, 75, 76.
waterflow, 79.
Quartz, transparency, 56.
Radiation and plant growth, 342-346.
and temperature, 41, 64, 70, 308-316.
measurement, 283, 284.
nature, 63, 286.
over sun's disk, 107, 110.
solar constant of, xx, 75, 295, 298.
unit, 68.
versus convection in sun, 103. Radiator, perfect, 64-66. Radium in sun, 94, 272. Rayet, 137, 138. Rayleigh, xxiii, 241, 304. Reed, 148.
Reflecting powers, 388. Refraction, anomalous and reg- ular, 233.
circular, 229.
indices, 46.
law, 48. Reinke, 342. Respighi, 155. Reversal, double, 147, 148. Reversing layer, 88.
pressures, 99.
thickness, 171. Ricco, 166, 195. Rice-grains, 85. Richarz, 29. Ritchey, 418. Roberts, 403. Roche, 413.
Rock-salt, transparency, 56. Romer, 23. Rosa, 21. Roscoe, 302, 349. Rosse, 311.
444
INDEX
Rotation, solar, 122-127. Rowland, 52, 62, 89, 90, 169,
175, 212, 254, 268, 436. " Preliminary Table," 89. wave lengths, corrections, 95,
96.
Runge, 93, 176, 254. Russell, 404.
St. John, 89, 101, 121, 127,
213, 269. Salisbury, 274. Salt, rock-, dispersion, 233.
transparency, 56. Sainson, 23, 271. Saros, 129.
Scattering of light by gases, 241. Schaberle, 265. Schehallien, 28.
Scheiner, discovery of sun- • spots, 183. J., 113, 247, 410, 421, 426. Schmidt, xxiii, 244. solar theory, 228. Schramm, 303. Schultz, 237. Schuster, xxiii, 188, 236, 242,
271, 301.
Schwabe, 184, 436. Schwartzchild, 103, 108, 203,
204, 237, 247. Schwendener, 338. Secchi, 142, 152, 156, 161, 166,
196, 198, 236, 407-409. See, T. J. J., 103, 226, 247, 417,
430.
Shackelton, 168, 169. Shearman, 190. Shuman, 376. Sidgreaves, 152, 198. Silver, reflecting power, 56.
Sky, light of, 299-307, 349-354. Slocum, 166.
Smithsonian Institution obser- vations, 69, 86, 203, 284, 298, 321, 384, 387. Smithsonian Observatory, Mt. Wilson, 37.
pyrheliometer, 75. Smoke, absorption, 386. Snow telescope, 60. Sodium, anomalous dispersion,
233.
Spectroheliograms, 118-120. Spectroheliograph, 59, 117.
and prominences, 166. Spectroscope, 45.
d vitesse, 117.
grating, 53, 54.
prismatic, 50. Spectrum, analysis, 39.
appearance, 38.
corona, 134.
limb, 105.
lines, broadening in sun- spots, 211. telluric, 43.
meaning, 39-41.
measurement, 61.
negative elements, 95.
prominence, 138, 145-147.
solar, extent, 86.
sun-spot, 203-210. Spoerer, 122, 123, 126, 193,
194, 220, 436. Stark, 261.
Stars and nebulae, 399, 418- 421.
density, 403.
distance, xix, 392.
double, 400-404.
evolution, 418-434.
445
INDEX
Stars, fission of, 429-433. groups, 398. magnitudes, 394. mass, 402, 424. Orion, 408, 419. solar, 408, 419. solar and Antarian, 433. spectra, 405-412. and velocities, 430. classification, 407-409, 422. energy, distribution, 410. spectroscopic, binary, 401,
422-429. streams, 432. Stefan saw, 67, 110, 250. Sterner, 7. Stomata, 338. Stone, 317. Stoney, 222, 420. Stralonoff, 123, 126. Struve, 392.
Sun, among the stars, 391-434. and plants, 331-361. axis, 122. brightness, distribution, 105-
107.
compared to Mira Ceti, 404. cooking, 379. corona, 263. density, 30.
dimensions and mass, 26, 27. distance, geometrical meth- ods, 9, 12, 13, 15. gravitational methods, 15. summary, 25, 26. velocity of light methods,
22, 23. energy supply, 271, 277, 383-
387.
gaseous, xxiii, 30, 217, 229, 237.
Sun, general features, xviii. heaters and reservoirs, 375-
380.
machines, 366-372, 376-379. motion among stars, 395.
in space, xx. nature of, Abbot, 236-279.
Halm, 225-228.
Julius, 233-236.
Schmidt, 228-233.
Young, 215-225. ore reduction, 381. origin, 412-418. photosphere, edge darkening, 249.
levels, 250.
nature, 246.
" rice-grains," 247.
spectrum, 249.
temperature variation, 250.
thickness, 244. prominence spectrum, 259.
velocities, 259. radiation, constant, 75.
dependence on air mass, 286.
measurement, 283, 284. rotation, 122-127. sharp boundary, 241. spectrum and chemical ele- ments, 252.
bright line, 259.
center and edge, 255.
dark line, 251. stellar magnitude, 394. temperature, 70, 109-116,
265. utilization of energy, 362-
390.
variability, 317. Sunflower, economy, 361,
446
INDEX
Sun-spots and associated phe- Temperature, attained in " hot-
noinena, 189, 190. and llocculi, 121, 122. and magnetism, 187, 191,
211. and prominences, 149, 153,
155.
and solar rotation, 122. and terrestrial temperature,
190, 191, 317. coolness of, 207-210. darkness of, 203, 204. discovery, 183. distribution on sun, 193-19G. drift, 192. formation and life history,
19G-199.
length of period, 185. level, 199.
motion within, 213. nature of, 2G7. periodicity, 184-188, 192, 225. pressure within, 213. relative numbers, 184. size, 183.
spectrum, 203-210. titanium oxide in, 94. System, solar, dimensions, 1, 3. evolution, 412-418. relative distances, 5.
Tacchini, 152, 158, 166.
Tatnall, 254.
Telescope for solar work, 31,
58.
snow, 60. tower, 58. Temperature and plant growth,
341.
and radiation, 41, 64, 70, 308- 316.
box," 305, 374. earth's, at high altitude,
280-283.
geologic, 273-277, 322. importance in sun, 265. of stars, 411.
progressive, 429, 433. of sun, 109-116. terrestrial, 311-316. Tennant, 137, 141. Thermodynamic efficiency, 387. Thomson, 213. Tisserand, 123. Titanium oxide in sun, 94,
209. Transmission, atmospheric, 294-
297. Transparency of optical media,
56.
Trouvelot, 156. Turner, 25, 132, 133. Tyndall, pyrheliometer, 75.
Uranium in sun, 94.
Veeder, 220.
Venus, transits, 13, 15.
Very, 311.
Villager, 108, 203, 204.
Vogel, 228, 409-412.
Von Gothard, 152.
Vortices, solar, 120.
Water, required by plants, 332.
vapor, radiation and absorp- tion, 281, 291. Wave lengths, accuracy, 62.
and levels, 98.
and plant growth, 342.
and pressures, 99.
447
INDEX
Wave lengths, principal solar,
90.
range, H. 2s:?. Rowland's 89.
corrected, 02, 9(>. \\ hcatstone, bridge, 81. Wiedemann, E., 255. Wien, displacement law, 07, 70,
100, 119, 250.
Wien-Planck law, 00, 111, 251. Wiesner, 302, 349-354. Wilczynski, 271. Willsie and Boyle, 370-379. Wilsing, 113, 123, 271, 410,421,
420. Wilson, A., 199, 200.
W. E., 204, 205. Witt, Eros parallax, 10.
Wolf, 184, 185, 430.
sun-spot numbers, 1M4, ISC,
194, 1 !>:.. \\oltVr, 185. Wood, R. W., 233, 204.
Yerkes Observatory, 132, 107,
407. Young, 5, 10, 134, 137, 144,
100, 101, 105, 108, 193,
190, 200. views of sun's nature, 215-
225.
Zeeman, 43, 211. Zodaical light, 400. Zollner, 142.
(1)
THE END
UNIVERSITY OF TORONTO LIBRARY
Do not
re move
the card
from this
Acme Library Card Pocket
Under Pat. " Ref. Index File." Made by LIBRARY BUREAU