Variable Stars And Observations

VARIABLE STARS ROBERTO BARTALI

ABSTRACT This project is around stars that for some reason changes its luminosity. I divide this work in two sections, the first is a general description of different types of variable stars and the second is about the observations of that kind of stars. The introduction contains, synthetically, the main characteristics of almost the most common variable star types, for best understanding, I include some representative light curves. To help readers not involved in the study of these stars, I also explain how them are classified. The observing section contains information about how to do a useful work on variable stars, explaining, briefly, which methods and which the instrumentation is needed to achieve good results. The reader can also find some work I did and the results obtained with some selected variable stars INTRODUCTION A star changing its luminosity due to some physical or mechanical phenomenon is considered a variable star. Variable stars are very important because understanding what happened with them, we can learn the formation, evolution and death of stars and the

Figure 1 Nova explosion sequence Nova Cygni 1975

Universe. We can observe stars in the Universe at different evolution stage, some very young, new born, other very hold, just dieing. Sooner or later, every star would be a Variable at certain moment in its life (Figure 1). These stars are classified depending on the type of variability observed. Sometimes the name of the group, or the variability class, is the name of the prototype star showing some kind of variation property, like “RR Lyr”; other times the name is allusive to the kind of phenomenon that produce variability like “Eruptive”. The name of a variable star is a compound of 1 or 2 capital letter and the constellation name. The letter sequence start with R until reaching Z, at this point follow RR, RS until RZ, then SS, ST until SZ. Reaching ZZ, just a V and a progressive number starting with 335. The first 1

variable star catalogue was redacted in 1934. Each variable star class is divided into many subclasses. Here is a condensed classification based on many catalogues. Each class is referenced by a letter (capital for the first) followed by 1 to 4 other letter, capital or not. Sometimes a star shows variations that are a compound of 2 or more classes, so the star can not be classified as belonging on a particular one and there is necessary to indicate each class separately by a “+” (for example E+UG). The main classes are as follow: a. Eruptive irregulars b. Eruptive nova-like or Cataclysmic c. Pulsating d. Eclipsing e. X ray and others Now I will explain briefly each class and their relative subclasses ERUPTIVE IRREGULAR Ia Irregular variables, spectral class O,B,A, typical example: BU Tau In Irregular variables, normally in diffuse nebulae like Orion nebula. Main secence and subgiant stars. Ina Irregular variables in Orion. Spectral class O,B,A. Typical star: T Ori. Ins Same as In, but the star present a rapid variation. Inb Variables in Orion, spectral class F,G,K. Typical star: AH Ori. InT T Tau stars. Spectral class F,G,K,M. Emission lines at 406 and 413 nm. Belongs to diffuse nebulae. IT Same as InT, but not in diffuse nebulae. UVn Rapid flares variables, normally in diffuse nebulae, specially in Orion, spectral class K,M. Very luminous stars. Typical star: V389 Ori. Is Fast irregular variables, magnitude difference 0.5 to 1, lasting from hours to days, not connected to diffuse nebulae. Isa Irregular variables of spectral class O,B,A. Typical star: XX Oph. Isb Fast irregular variables of spectral class F,G,K,M. Typical star: AQ Dra. UV Fast flares, the maximum is reached in seconds and the duration is several minutes. The magnitude difference can be up to 6. Typical star: UV Cet.

2

ERUPTIVE NOVA-LIKE or CATACLYSMIC N Novae, magnitude variations from 1 to 16, in one day, then, slowly return to the initial magnitude, At maximum the spectral class is similar to A,F stars with strong H and He emission lines, when reducing the luminosity the spectrum is complex, at normal minimum it is almost continuum. Na Typical nova. Fast increasing of luminosity, Figure 2 followed to 3 magnitude decreasing in about Nova va Cygni Cygni 1920, light curve 100 days. (Figure 2).Typical star: nova Per 1901. Nb Nova with slow decreasing magnitude. This type of nova, lost 3 magnitudes in more than 150 days. Don’ t care about the secondary minimum. Typical star: T Aur. Nc Very slow nova, maintains the maximum brightness for years and reach the normal after many years. Typical star: RT Ser. Nr Recurrent nova, it was observed at maximum many at least for 2 times. Typical star: T Crb. Nl This kind of stars share some characteristics of novae, like magnitude increment and spectral type, but they don’ t belongs to novae. Typical star: P Cyg. Z And Symbiotic stars, their spectra present both characteristics of red giant and blue dwarf together. Typical star: Z And. RCB Very luminous stars, spectral type F,K. The star remains at normal bright magnitude and suddenly fade Figure 3 Corona Borealis light curve up to 9 magnitudes. Remain in the minimum for several days up

3

to hundred of days. When at minimum, present metallic emission lines in the spectrum. Typical star: R Crb (Figure 3). UG Dwarf stars, normally present small magnitudes fluctuations, eventually increase the brightness up to 6 magnitudes, then fade in several days. Time between maximum is not regular, but for each star is possible to compute its own mean time. The period is proportional to the magnitude of the eruption. Many of these stars have a closer companion. The spectrum at minimum is continuous and contains H,He,Ca emission lines, they appears at maximum as absorption lines. Typical star: U Gem. Z Cam Similar to UG, but they have a static intermediate maximum and they stay there for long time. Typical period is from 10 to 40 days, the magnitude increment less than 5 units. Typical star: Z Cam. SN Like novae, supernovae increase the brightness of 20 or more magnitudes in very short time. Depending on the light curve and spectra they are classified as type I or type II. After the explosion they form a planetary nebula. Typical star: CM Tau. Other Eruptive subclasses: FU, GCas, RS, SDor, WR, Ib PULSATING Cep Cepheid stars. Very luminous stars, the magnitude change from 0.1 and 2, the period is in the range 1 to 70 days. The relationship between radial velocity and luminosity is constant. The spectral type at maximum is F and at minimum is G,K. DCep Classic Cepheid, in the Galaxy plane. The Figure 4 relationship between luminosity and period Eta Aguilae light curve is constant, so we can use this stars to measure distances (Figure 4). Located in open clusters. Typical star: Delta Cep. CW Cepheid in the galaxy halo and closer to the nucleus. Located in Globular clusters, the relationship between luminosity and period is constant but their light curve are displaced 1.5 to 2 magnitude below from the one of classical cepheids. They are less luminous. The period is between 3 and 10 days. Typical star: W Vir. L Irregular, long period variable stars. Many stars are classified in this group because to the poor knowledge of their light curves or spectral class. There are no trace of a periodicity for these stars. Lb Irregular, long period variables, spectral class K,M,S, giant stars. Typical star: CO Cyg. 4

Lc Irregular variables of spectral type K,M. supergiant stars. Typical star: TZ Cas. M Very long period variables, up to 5 or even more magnitude changes in a period from 80 to 1000 ore more days. Most are giant stars of spectral class M,S,C. Typical star: Omicron Cet. SR Semi-regular variables. Normally giant or supergiant stars. Magnitude difference up to 2 and period up to 1000 or more days. The light curve present some perturbations. SRa Semi-regular variable stars of spectral class M,C,S. More stable than SR and magnitude difference less than 2.5 units. Frequently the light curve change from one period to another. Typical star: Z Aqr. SRb Semi-regular variables, like SRa, have spectral class M,C,S, the period change frequently and sometimes present a period of stability between cycles. Typical star: RR Crb. SRc Semi-regular supergiant of spectral class K,M, located in the spiral arms of the Galaxy. Typical star: MU Cep. SRd Semi-regular supergiant and giant stars of spectral class F,G,K. Typical star: S Vul. RR Regular variabnitudesles, cepheids of short period. Giant stars of class A, like CEP, but with period from minutes to 1.2 days. Magnitude difference less than 2 units, even when the period – luminosity relationship is constant, many stars present periodic superimposed variations. Located mostly around the center of the galaxy. Typical star: RR Lyr (Figure 5).

Figure 5 RR Lyrae light curve

RRab Similar to RR, but the light curve is asymmetric, the rise time is faster then the falling time. At maximum the luminosity is constant, the amplitude variation is less than 0.5 mag. Periods are from 0.5 and 0.7 days. RRs Similar to RRab but the period is short, less than 0.21 days. They are dwarf Cepheid. Located in the Galaxy disc. The magnitude is 3 units less than other type of RR. Typical star: SX Phe. RRc Cepheid variables, the light curve is symmetric, almost sinusoidal. Mean period of 0.3 days. Typical star: SX UMa.

5

RV Supergiant stars, variation amplitude of 3 magnitudes, the light curve present two minimum, the main and the secondary. Many times they are interchanged. Period extends from 30 to 150 days. Spectral type from G to K, when at maximum they are closer to G. Typical star: RV Tau. RVa Like RV, mean luminosity constant. Typical star: AC Her. RVb Like RV, mean luminosity not constant. Typical star: R Sge. BCep Cepheids of small amplitude variation, about 0.1 magnitudes. Periods from 0.1 to 0.6 days. In these type of stars, the maximum luminosity is when they have the minimum radius. Spectral class B0 to B3. Typical star BETA Cep. DSct Pulsating variables of spectral class A7 to A9. Magnitude variation about 0.1, sometimes 0,3. Located in Open clusters, the light curve varies from one star to another, periods of 0.2 days. Typical star: DELTA Sct. CVA Magnetic variable stars. In the spectrum there are lines of Silicon, Strontium, Chromium and other Rare Hearts. The variations of the spectral lines is the same as luminosity. Periods from 1 to 25 days and magnitude variation about 0.1 units. Typical star: ALFA Cvn. Other Pulsating subclasses ACyg, BCep, BCeps, CWa, CWs, DCeps, DsSct, PV Tel, ZZ, ZZa, ZZb ECLIPSING E Very close binary system. The orbital plane of the stars are near or the sane as visual plane, so during each orbit one star is eclipsing the other. The light curve is the sum or the rest of the brightness of the 2 stras. EA Eclipsing binary. If there is no eclipse, the luminosity is constant. It is possible to generate ephemerides for minimum and maximum. Their periods depends on the radius of the orbit and the mass of the stars and are from 0.2 days to hundred of years. Typical star: BETA Per. Figure 6 EB Eclipsing binary. (Figure 6) Stars of this kind Beta Lyrae Light curve are ellipsoidal rather than spherical. The magnitude varies continuously, so it is impossible to predict minimum and maximum, because the stars are always in eclipse. Spectral class O,B,A, periods around 1 day and magnitude variations less than 2 units. Typical star: BETA Lyr.

6

EW Binary ellipsoidal stars, very close together, they are almost in contact. Periods of 1 day or less, and the variation is 0.8 magnitude or less. Due to the very close distance off the stars, the main and secondary minimum are the same. Spectral class F,G. Typical star: W UMa. Ell Binary stars, ellipsoidal, but they don’ t eclipse each other. The variations are due to the different amount of surface luminosity the star present to us as it revolve. Typical star: b Per. Other Eclipsing subclasses PN, RS, WD, AR, D, DM, DS, DW, K, KE, KW, SD X-RAY and OTHERS X, XB, XF, XI, XJ, XND, XNG, XP, XPR, XPRM, XM

Figure 7 Radio pulse of a typical pulsar

OBSERVATIONS We can observe variable stars in several ways, due to the kind of variation and the physical phenomenon that produce the variation itself. The graphical representation of the luminosity versus time is called “ Light Curve “. Each different variable class has its own kind of light curve. This graph is the fingerprint of the variable. If the star is not irregular, it is possible to predict the magnitude at will shine in the future. In the introduction the reader can observe some examples of light curves. The main goal to observe variable stars is to plot the light curve as precisely as possible in order to studying which are the processes in the interior or something else external, that is the reason for the variability. Visual observation made with naked eyes or through a telescope, is not easy because is not objective, depends Figure 8 of the eye sensitivity to light and colors, so there are UCEP at minimum (lower) various methods to track and record star luminosity, And at maximum (higher) most common is the Argelander method. Recording star June 12, 1964 brightness in a photographic plate (Figure 8) or electronically with a CCD sensor, is easy and it is objective, the sensor or the plate always seen the star as it really is, depending only on the sensitivity of the instrument. But 7

photographic plate and emulsions are more sensitive to the short wavelength (blue part of the visible spectrum) and CCD are more sensitive to the red part of the visible spectrum. If we take a picture, the magnitude of the star can differ greatly from the CCD to the photograph. The very high quantum efficiency of the CCD permit to record the light from very faint stars in a short time exposure. The advantage of using electronic instruments, like CCD or photoelectric photometer is that we can record directly the brightness of the star, instead of visual or photographic methods that needs some processing of comparison. Not all the variable stars change in the visible part of Figure 9 the spectrum, we can observe variation in the entirely Typical Finder chart electromagnetic spectrum. Radio telescope can help observing neutron stars, pulsars (Figure 7) and quasars, another type of variables can be observed through X-Ray, UV and IR sensitive telescopes or detectors. Some stars have variations in the lines of the spectrum, here is where spectroscopic observations can help. In order to achieve a complete knowledge of the star we can use all, or most of the instruments and methods available, because the star can have very different properties and behavior at different wavelength. Whichever method we use for computing the magnitude of the variable star, we have to know previously the magnitude of some stars in the field of view called “Comparison stars”. More precisely we know the magnitude of the comparison stars, more exactly we can compute the variable magnitude, and, the light curve, Figure 10 correspond to the reality. Obviously the Hipparcos and Tycho star catalog comparison stars have to be “Normal”, not Plot of UCEP field variables. From a few years ago, there is an astrographic satellite, Hipparcos, which task is the measurement of the position and the magnitude of the stars very precisely, because it is out of the aberrations introduced by the Earth atmosphere to the path of light. Comparison stars can be founded on catalogs and they are plotted in “Finder Charts” like the one in Figure 9. Just those stars labeled as Comparison can be used, (Figure 10) because we know precisely the magnitude and they are not variables for sure.

8

OBSERVATIONS OF U CEP R.A. 01h 02m 18.34s DECL. +81° 52’ 32.1” Mag. 6.85 – 9.40 Type EA/SD Period 2.49307 days Spectral Type B7 – G8 III

Figure 11 Palomar Sky survey picture of UCEP field

Figure 13 File: UCEP-5

Figure 15 File: UCEP-7

Figure 12 UCEP light curve

Figure 14 File: UCEP -6

Figure 16 File: UCEP -8

9

Figure 17 File: UCEP-9

Figure 19 File: UCEP-11

Figure 18 File: UCEP-10

Figure 20 File: UCEP-12

10

UCEP DATA MJD

FILE NAME

CCD

JD-2450000.5

.FTS

2751.342581

UCEP_1

1

2754.497928

UCEP_2

2755.282454

DATE

TYPE DD

TIME UTC

TIME

EXP TIME

MM

YY

HH

MM

SS DECIMAL

SEC

22

4

2003

6

23

19 6.388611

100

1

25

4

2003

10

7

1

10.11694

120

UCEP_3

1

26

4

2003

4

56

44 4.945556

120

2756.347975

UCEP_4

1

27

4

2003

6

31

05 6.518056

10

2757.248738

UCEP_5

1

28

4

2003

4

8

11 4.136389

5

2757.250139

UCEP_6

1

28

4

2003

4

10

12

2

2758.274630

UCEP_7

1

29

4

2003

4

45

28 4.757778

5

2758.334201

UCEP_8

1

29

4

2003

6

11

15

5

2758.356863

UCEP_9

1

29

4

2003

6

43

53 6.731389

5

2758.388148

UCEP_10

1

29

4

2003

7

28

56 7.482222

5

2759.284873

UCEP_11

1

30

4

2003

5

0

13 5.003611

5

2759.375012

UCEP_12

1

30

4

2003

7

10

1

7.166944

5

2760.667523

UCEP_13

2

1

5

2003

4

21

14 4.353889

5

2760.779271

UCEP_14

2

1

5

2003

7

2

9

7.035833

5

2761.665729

UCEP_15

2

2

5

2003

4

18

39 4.310833

5

2761.737072

UCEP_16

2

2

5

2003

6

1

23 6.023056

5

2762.693692

UCEP_17

2

3

5

2003

4

58

55 4.981944

5

2764.699699

UCEP_18

2

5

5

2003

5

7

34 5.126111

5

2764.778970

UCEP_19

2

5

5

2003

7

1

43 7.028611

5

2764.781042

UCEP_20

2

5

5

2003

7

4

42 7.078333

2

2765.730914

UCEP_23

2

6

5

2003

5

52

31 5.875278

2

2765.741655

UCEP_24

2

6

5

2003

6

7

59 6.133056

5

2765.743032

UCEP_25

2

6

5

2003

6

9

58 6.166111

2

2767.674965

UCEP_26

2

8

5

2003

4

31

57

2

2767.676366

UCEP_27

2

8

5

2003

4

33

58 4.566111

4.17

6.1875

4.5325

5

11

The above table is the file information and observational data. I take 27 pictures in total, in this work I only show 8 that represent the first part of the light curve (Figures 13 to 20). In this sequence of observations there are 2 minimum, the first on MJD = 2758.388148 and the second around MJD = 2768.3, unfortunately the sky was not clear enough to take more pictures. The following minimum was on MJD = 2768.36. The original FITS files are stored in a CD rom, and they are available on request. The 8 pictures are a Jpeg copy of some of that files. The Palomar Sky survey plate is a 15’ x 15’ field around U Cep, and for coincidence is almost the same field of view of the telescope I am using to take the CCD pictures, a Takahashi Mewlon 300 equipped with a Dream Machine 1024 CCD sensor and a Takahashi TRC300 equipped with a SBIG ST8XE sensor camera, this way it is simple to find comparison stars. Not shown in this work because it is not relevant is the table of all the star in the field of view. I am using the Viziar Catalogue services for this purpose. The following table represent the value of the CCD pixel in the picture that correspond to the star. The back ground value is the value of the dark frame in the picture. The star value is the result dividing the pixel value (star value) and the back ground value. Half of the pictures are taken with a SBIG ST8XE 1530x1020 pixel CCD, and the other half with a Dream Machine IMG1024 CCD, the different pixel size make the difference in sensitivity, so I do a kind of normalization between the two CCD back ground. The graph show the relative brightness of the star at some MJD value. To get the magnitude I have to do more analysis and compare with some comparison stars, but no one is in the field of the picture. STAR VALUE

BACK GROUND

MJD

RELATIVE BRIGHTNESS

60771 54559 55171 54235 55015 54679 54803 45675 39547 30585 55119 54039 61871 60818 61737 62176 61802 61592 62317 60176 60746 62403 59625 52854 61930

17622 17622 17652 17221 17188 17162 17178 17172 17176 17178 17175 17175 13699 13652 13701 13711 13731 13692 13693 13685 13693 13712 13705 13689 13701

2751.34258 2754.49793 2755.28245 2756.34797 2757.24874 2757.25014 2758.27463 2758.3342 2758.35686 2758.38815 2759.28487 2759.37501 2760.66752 2760.77927 2761.66573 2761.73707 2762.69369 2764.6997 2764.77897 2764.78104 2765.73091 2765.74166 2765.74303 2767.67497 2767.67637

3.449 3.096 3.125 3.149 3.201 3.186 3.190 2.660 2.302 1.780 3.209 3.146 3.577 3.528 3.569 3.592 3.565 3.563 3.605 3.483 3.514 3.605 3.446 3.058 3.580

12

From my observation and measurement data, I realize the follow plot:

The variation period of the star make almost impossible the record of 2 or 3 consecutive the minimum brightness. It is possible to see in the light curve the beginning of 2 more minimum cycles MJD=2754 and MJD=2765.

Zoom-in on the light curve of U CEP around the minimum.

13

In this plot there are clearly 1 minimum and for the others it is possible to predict when they will occurs by interpolation of the values in the table. The separation between the 2 in this plot is exactly 4 periods. OBSERVATION OF DY HER R.A. 16h 31m 18s DECL. +11° 59’ 52” MAG. 10.1 – 10.6 TYPE DELTA SCUTI PERIOD. 0.148631353 days SPECTRAL TYPE A7 III – F4 III

Figure 21 Palomar Sky Survey Field DY HER Field

14

DY HER DATA FILE NAME .FTS

CCD TYPE

2756.271076

DYHER_A

2756.298009

MJD JD-2450000.5

DATE

TIME UTC

TIME DECIMAL

EXP. TIME SEC.

DD

MM

YY

HH

MM

SS

1

27

4

2003

4

40

21

4.6725

30

DYHER_B

1

27

4

2003

5

19

8

5.318889

30

2756.330417

DYHER_C

1

27

4

2003

6

5

48

6.096667

30

2756.400289

DYHER_D

1

27

4

2003

7

46

25

7.773611

30

2760.684583

DYHER_3105_1

1

1

5

2003

4

45

48

4.763333

30

2760.710463

DYHER_3105_2

1

1

5

2003

5

23

4

5.384444

30

2760.748634

DYHER_3105_3

1

1

5

2003

6

18

2

6.300556

30

2760.770266

DYHER_3105_4

1

1

5

2003

6

49

11

6.819722

30

2760.827720

DYHER_3105_5

1

1

5

2003

8

11

55

8.198611

30

2785.773009

DYHER_1

2

26

5

2003

6

53

8

6.885556

30

The method to record data is the same as I explained above in the U CEP paragraph. The instruments are also the same. I choose this star because it has a very short period, and a deep minimum. I suppose that I have been able to “Catch” a complete minimum cycle in one night. Fortunately the sky help me and I can see 2 minimum. STAR VALUE

BACK GROUND

MJD

RELATIVE BRIGHTNESS

5724 4884 3615 7189 9261 8829 5725 3409 4113 29310

63 62 62 57 61 62 61 61 61 5823

2756.271 2756.298 2756.330 2756.400 2760.685 2760.710 2760.749 2760.770 2760.828 2785.773

90.857 78.774 58.306 126.123 151.820 142.403 93.852 55.885 67.426 95.280

15

This plot is the complete light curve of DY HER based on my observation data. The period of the minimum is exactly the published data (see the table at the beginning of the discussion about this star). The two minimum are separated 30 full periods. The ephemerid for the second minimum is (30 cycles * 0.148 days) + 2756.33 = 2760.77, exactly as it appear in the light curve.

Light curve showing the minimum on MJD = 2756.3304

Light curve showing the minimum on MJD = 2760.7703 16

Figure 22 File DY HER-a

Figure 24 File DY HER-c

Figure 23 File DY HER-b

Figure 25 File DY HER-d

This pictures (figure 22 to 25), a Jpeg, copy of the original Fits files, represent the sequence for the first minimum. For helping me to find comparison stars I compare this fields with the Palomar Sky Survey plate of the same area in the figure 21. The reader can find the reference data for this pictures in the observation data table. CONCLUSIONS The realization of this work was not an easy task, because I first start to observe on my site, but light pollution and bad sky condition, don’ t help to much. I decided, then, to rent a telescope in Arizona, USA, and operate it on line via internet (www.arnierosner.com). This way I could use a CCD ( this was my first time) and the image process work was another problem. When I observe visually I used the Argelander Method, but for CCD data reduction I use some software like IRIS, DS9 and CADET. 17

In this work I only include the observations of 2 stars, because of the result achieved. The other stars I was working with are: SS LEO, BL HER, W UMA, BB HER and AN UMA. For those stars I can not realize a satisfactory light curve, due to the times I take the pictures, and, the availability of the telescopes not all the time when I need it. Anyway I get a good result, because for DY HER the minimum of the stars in published data and in my own data are the same, With a little interpolation work I can say the same for U CEP. I am working now to transform the pixel value to a magnitude value, but I need more time and I can find almost 2 comparison stars in the field of the pictures, and I have not found anyone yet. REFERENCES Burnham R.,BURNHAM’S CELESCTIAL HANDBOOK, Dover, 1978 Bartali R., LAS ESTRELLAS VARIABLES, La Via Lactea num 2,5, 1979 Cora A., L’ OSSERVAZIONE AMATORIALE DELLE STELLE VARIABILI, Nuovo Orione num.2, 1992 Bulletins of the Unione Astrofili Italiani, Variable Stars section, 1976..1981 Dogget L et al, THE ALMANAC FOR COMPUTERS FOR THE YEAR 1978, U.S. Naval Observatory, 1978 Internet references www.aavso.org http://aladin.u-strasbg.fr/aladin.gml http://vizier.u-strasbg.fr/cgi-bin/VizieR http://www-gsss.stsci.edu/support/data_access.htm http://www.time.gov/timezone.cgi?Central/d/-6/java www.arnierosner.com http://astro.estec.esa.nl/Hipparcos/HIPcatalogueSearch.html http://www.aerith.net/misao/index.html http://www.astrosurf.com/astronosur/variables.htm http://www.astrogea.org/VARIABLE/variables.htm http://aladin.u-strasbg.fr/AladinPreview Software references DS9: http://hea-www.harvard.edu/RD/ds9/index.html CADET: http://www.terra.es/personal2/oscarcj/introeng.htm IRIS: http://www.astrosurf.com/buil/us/iris/iris.htm Image credits Figure 1,2,3,4,5,6,8,9,12: Burnham’s Celestial Handbook Figure 7: Coelum num. 5-6, 1976 Figure 11,21: Palomar Sky Survey Figure 10: VSNET Univ. of Kyoto Figure 13,14,15,16,17,18,19,20,21,22,23,24,25: Roberto Bartali

18