UvA-DARE is a service provided by the library of the University of Amsterdam (https://dare.uva.nl)

UvA-DARE (Digital Academic Repository)

Optical and ultraviolet studies of X-ray binaries and magnetic stars

Hammerschlag- Hensberge, G.C.M.J.

Publication date1977

Link to publication

Citation for published version (APA):Hammerschlag- Hensberge, G. C. M. J. (1977). Optical and ultraviolet studies of X-raybinaries and magnetic stars.

General rightsIt is not permitted to download or to forward/distribute the text or part of it without the consent of the author(s)and/or copyright holder(s), other than for strictly personal, individual use, unless the work is under an opencontent license (like Creative Commons).

Disclaimer/Complaints regulationsIf you believe that digital publication of certain material infringes any of your rights or (privacy) interests, pleaselet the Library know, stating your reasons. In case of a legitimate complaint, the Library will make the materialinaccessible and/or remove it from the website. Please Ask the Library: https://uba.uva.nl/en/contact, or a letterto: Library of the University of Amsterdam, Secretariat, Singel 425, 1012 WP Amsterdam, The Netherlands. Youwill be contacted as soon as possible.

Download date:10 Aug 2021

ASTRONOMY Y Astron.. Astrophys. 54, 443—449 (1977) A N D

ASTROPHYSICS S

Photometryy of Silicon Stars

H.. Hensberge1, C. De Loore1, E. J. Zuiderwijk2 and G. Hammerschlag-Hensberge2

11 Asirophysical Institute, Vrije Universiteit Brussel, Pleinlaan 2, B-1050 Brussels, Belgium Europeann Southern Observatory, La Silla, Chile 22 Astronomicallnstitute, University of Amsterdam, Roetersstraat 15, Amsterdam, The Netherlands Europeann Southern Observatory, La Silla, Chile Receivedd July 19, 1976 Summary.. Five of the brighter silicon stars in the list off Bidelman and MacConnell (1973) i.e. HD3580, HDD 187473, HD 206653, HD 207188 and HD 212432, weree observed at the ESO, La Silla, to search for variabilityy in the wuby-photometric system. All of them turnedd out to be variable in the four channels showing obviouslyy their largest amplitude in u (>0rü'2 in the casee of HD 187473) and periods could be derived for alll of them. One of the comparison stars, i.e. HD 185183, alsoo turned out to be variable. Colour indices, amplitude, characterr and periodicity of the variations and a 122 A/mm spectrogram suggest that this star might be a broad-linedd silicon star.

Keyy words: Ap-stars — photometry — Si-stars

1.. Introduction

Inn continuation of our search for periodicities among peculiarr A stars (Hensberge et al., 1976) we present in thiss paper four-colour uvby photometry of five southern siliconn stars of the list of Bidelman and Mac Connell (1973):: HD3580, HD 187473, HD206653, HD 207188 andd HD 212432. The stars were selected because of theirr brightness. One of the comparison stars, HD 185183,, turned out to be a variable peculiar star itself.

Inn §2 we give a short description of the observations andd reduction methods. The results for each program starr separately are discussed in §3. In §4 our conclu-sionss are summarised.

2.. Observations and Reductions

Thee observations were made by E. Zuiderwijk with the 500 cm Danish telescope at La Silla between July 25 andd August 23, 1975. This telescope is equipped with a

SendSend offprint requests to: H. Hensberge

photometerr for simultaneous measurements in the Strömgrenn uvby system. The photometer is used in combinationn with a photon counting system and is describedd in detail by Gronbech et al. (1976).

Eachh observation sequence contained measure-mentss of the program star (P) and three comparison starss (CI, C2, C3) as follows: C1-P-C2-P-C3-P-C3-P-C2-P-C1.. Extinction corrections were computed in the usuall way. The observations of two comparison stars weree reduced to the magnitude of the third one by using thee mean difference in magnitude over the entire observationn run. The dispersion of these differential magnitudess around their mean was low enough to permitt this procedure. The differential magnitude for thee program star was obtained as the average of all measurementss of two consecutive sequences. The stan-dardd deviation of this differential magnitude is practi-callyy always less than OTOOS.

Standardd indices were computed by transformation too the uvby system of the catalogue of bright southern starss (Gronbech et al., 1975). The transformation formulaee appear to reproduce the catalogue indices withh errors smaller than 0^01. However, the majority off program and comparison stars have not yet been measuredd in the uvby system.

Thee period of variation is determined by a method basedd on the technique used by Lafler and Kin man (1965).. For a sequence of trial periods, the phase </>, off each measurement x, is calculated. Assuming that thee xf are arranged according to their phase, S is defined as: :

N N

II (xH+l-x„) 2

cc »= 1

II (*„-*) 2

NN is the number of observations; for xN + l the ,v, value hass to be adopted and x is the average of the vn's. Obviouslyy S is a function of the guess periods since the arrangementt according to phase depends on the adopted

157 7

444 4 H.. Hensberge et al.: Photometry of Silicon Stars

Tablee I. Summary of results. Consecutive columns denote: HO number, silicon or comparison star, number of observing nightss and number of differential magnitudes, visual brightness denoted by (V) since transformation from y to (V) is somewhatt tentative for Ap-stars because of the complicated variability in this wavelength region . Slrömgren indices hh - r. ml and el and, for the variable stars, the period of variation

HD D number r

3580 0

35X1 1 4247 7 4622 2

187473 3

185183 3

187578 8 190285 5

206653 3

205348 8 205417 7 209468 8

207188 8

205705 5 207439 9 208482 2

212432 2

210739 9 210931 1 214172 2

Type e

Si i

comp. . comp. . comp. .

Si i

comp. .

var. . comp. . comp: : Si i

comp. . comp. . comp. .

Si i

comp. . comp. . comp. .

Si i

comp. . comp. . comp. .

n{N) n{N)

11(19) )

— —

15(21) )

— —

— —

13(16) )

—— .

11(15) )

— —

9(12) )

— —

(10 0

6.64/6.63 3

7.104 4 5.233 3 5.586 6

7.36/7.23 3

6.75/6.71 1

7.666 6 7.239 9

7.24/7.18 8

6.775 5 6.226 6 7.543 3

7.65/7.60 0

7.446 6 7.559 9 6.757 7

7.52/7.49 9

6.197 7 7.401 1 7.421 1

b-y b-y

-0.065 5

0.267 7 0.226 6

-0.029 9

-0.04/0.00 0

- 0 . 0 3 /--

0.032 2 0.053 3

- 0 . 0 4 /--

-0.043 3 0.008 8 0.011 1

-0.041 1

-0.033 3 0.159 9 0.035 5

- 0 . 0 5 /--

0.091 1 0.014 4 0.058 8

0.05 5

2 2

-0.06 6

ml l

0.119 9

0.162 2 0.142 2 0.131 1

0.15/0.18 8

0.09/0.12 2

0.096 6 0.147 7

0.135 5

0.112 2 0.151 1 0.174 4

0.152 2

0.101 1 0.189 9 0.182 2

0.248 8

0.193 3 0.184 4 0.191 1

cl cl

0.53/0.50 0

0.491 1 0.554 4 0.975 5

0.73/0.55 5

0.56/0.53 3

0.874 4 1.089 9

0.63/0.50 0

0.764 4 1.133 3 1.009 9

0.54/0.58 8

0.699 9 0.769 9 1.059 9

0.69/0.63 3

0.920 0 1.005 5 0.960 0

P(days) )

1.4800 5 orr 2.94 2

— —

4.755 4

-- 1.737

— —

1.7888 5

_ _

2.677 1

— —

4.699 5

— —

triall period. For the true period of variation, S is ex-pectedd to be near its minimum value for each of the filters.. Generally, when S<0.5 for each of the filters, thee correlation is good (in the case JV«20); in excellent casess S«0.2. As a rule, only for a very small number off trial periods S does not exceed 1.

Thiss technique was used by Lafler and Kinman (1965)) for the case of Cepheids. For that special case the shapee of the light curve is well known and the number off observations ranges roughly between 30 and 50. Whenn the number of observations is smaller, i.e. less thann 20, the sum S may contain a term arising from the comparisonn of observations with rather large phase differences.. Moreover, when the light curve is es-sentiallyy a double wave curve, it may be expected that itt contains none or at least very few observations be-tweenn succeeding extrema. In this extreme case, it is exactlyy this term which may produce the major con-tributionn to S and which might lead to the rejection off the trial period (although the observations are not inn contradiction with this period).

Too avoid this difficulty we decided to add an extra parameterr Q, the phase difference for which the in-creasee in magnitude in the case of a double-wave light

curvee never exceeds the probable error on the observa-tionss by a considerable amount. Terms in S, arising fromm phase points which differ more than Q are omitted. Wee compute

NN 2 > n + 1-xn)2<5(n,H+l)

S** = n = l l

N*N* I (xn-x)2

where e

<Hn,nn + l) = l if fm+l-fm£Q

== 0 if fn+l-fn>Q

andd N* is the number of <5's differing from zero. Thiss S*y depending on Q, is assumed to be a better

indicatorr for the period of the Ap stars than S. If Q is overestimatedd double-wave periods may be overlooked (inn fact 0 = 1 in the original version). If Q is under-estimated,, the test is not selective enough. This latter factt imposes a lower limit on Q in the case of large amplitude.. 0 = 0.1 seems to be a reasonable choice in mostt cases.

158 8

H.. Hensbergc et al.: Photometry of Silicon Stars 445 5

Tablee 2. Differential magnitudes after transformation to the uvby systemm of the catalogue of bright southern stars (Grenbech et al.)

Tablee 2,

JDD = 24426000 +

Ay Ay

HDD 3580-HD 4622

19.890 0 24.883 3 34.815 5 37.732 2 37.884 4 38.741 1 38.888 8 39.695 5 39.883 3 40.697 7 40.886 6 41.688 8 41.875 5 42.718 8 42.889 9 46.702 2 46.877 7 47.673 3 47.867 7

HDD 185183-

19.578 8 19.795 5 21.573 3 23.583 3 24.554 4 24.770 0 34.608 8 34.792 2 35.711 1 37.709 9 38.716 6 39.508 8 39.751 1 40.510 0 40.747 7 41.502 2 41.734 4 42.691 1 45.551 1 46.741 1 47.703 3

HDD 187473-

19.578 8 19.795 5 21.573 3 23.5X3 3 24.554 4 24.770 0 34.608 8 34.792 2 35.711 1 37.709 9 38.716 6 39.508 8 39.751 1

1.150 0 1.149 9 1.154 4 1.153 3 1.155 5 1.146 6 1.146 6 1.153 3 1.139 9 1.153 3 1.155 5 1.145 5 1.148 8 1.146 6

.1.143 3 1.155 5 1.155 5 1.148 8 1.148 8

-- HD 190285

-0.525 5 -0.504 4 -0.504 4 -0.479 9 -0.531 1 -0.524 4 -0.521 1 -0.530 0 -0.488 8 -0.494 4 -0.525 5 -0.503 3 -0.530 0 -0.507 7 -0.498 8 -0.515 5 -0.518 8 -0.502 2 -0.523 3 -0.513 3 -0.489 9

- H DD 187578

-0.320 0 -0.310 0 -0.425 5 -0.330 0 -0.317 7 -0.315 5 -0.343 3 -0.371 1 -0.420 0 -0.338 8 -0.312 2 -0.361 1 -0.395 5

Ah Ah

1.115 5 1.114 4 1.119 9 1.117 7 1.124 4 1.110 0 1.112 2 1.116 6 1.108 8 1.120 0 1.121 1 1.107 7 1.113 3 1.113 3 1.109 9 1.120 0 1.121 1 1.111 1 1.113 3

-0.618 8 -0.593 3 -0.589 9 -0.562 2 -0.621 1 -0.619 9 -0.611 1 -0.629 9 -0.567 7 -0.582 2 -0.616 6 -0.589 9 -0.619 9 -0.605 5 -0.583 3 -0.607 7 -0.618 8 -0.570 0 -0.621 1 -0.608 8 -0.579 9

-0.383 3 -0.376 6 -0.466 6 -0.404 4 -0.379 9 -0.376 6 -0.398 8 -0.422 2 -0.461 1 -0.406 6 -0.377 7 -0.413 3 -0.442 2

Av Av

1.064 4 1.065 5 1.066 6 1.065 5 1.071 1 1.061 1 1.063 3 1.063 3 1.052 2 1.068 8 1.067 7 1.060 0 1.063 3 1.060 0 1.054 4 1.066 6 1.071 1 1.064 4 1.066 6

-0.747 7 -0.731 1 -0.718 8 -0.697 7 -0.757 7 -0.753 3 -0.743 3 -0.752 2 -0.703 3 -0.709 9 -0.749 9 -0.724 4 -0.748 8 -0.734 4 -0.712 2 -0.737 7 -0.742 2 -0.707 7 -0.752 2 -0.741 1 -0.706 6

-0.382 2 -0.382 2 -0.435 5 -0.399 9 -0.381 1 -0.382 2 -0.390 0 -0.409 9 -0.432 2 -0.399 9 -0.379 9 -0.403 3 -0.414 4

Au Au

0.563 3 0.554 4 0.569 9 0.564 4 0.571 1 0.551 1 0.559 9 0.552 2 0.532 2 0.569 9 0.572 2 0.551 1 0.560 0 0.542 2 0.532 2 0.568 8 0.571 1 0.558 8 0.561 1

-1.415 5 -1.396 6 -1.383 3 -1.364 4 -1.437 7 -1.4J9 9 -1.427 7 -1.434 4 -1.363 3 -1.378 8 -1.418 8 -1.394 4 -1.432 2 -1.393 3 -1.379 9 -1.418 8 -1.418 8 -1.364 4 -1.424 4 -1.417 7 -1.373 3

-0.557 7 -0.539 9 -0.732 2 -0.598 8 -0.541 1 -0.539 9 -0.583 3 -0.625 5 -0.728 8 -0.602 2 -0.535 5 -0.620 0 -0.673 3

JDD = 24426000 +

HDD 187473-

40.510 0 40.747 7 41.502 2 41.734 4 42.691 1 45.551 1 46.741 1 47.703 3

HDD 206653-

19.687 7 23.818 8 24.796 6 35.623 3 37.758 8 38.765 5 39.583 3 39.835 5 40.720 0 41.581 1 41.829942.740 0 43.569 9 46.807 7 47.558 8 47.812 2

HDD 207188-

19.827 7 23.850 0 24.687 7 24.862 2 34.841 1 35.676 6 37.808 8 38.642 2 38.827 7 39.723 3 40.589 9 40.821 1 41.712 2 42.567 7 42.804 4

HDD 212432

19.861 1 24.839 9 34.869 9 37.682 2 37.860 0 38.798 8 39.808 8 40.618 8 40.846 6 41.801 1 42.596 6 42.841 1

Ay Ay

HDD 187578

-0.420 0 -0.409 9 -0.345 5 -0.332 2 -0.335 5 -0.401 1 -0.334 4 -0.324 4

-HDD 205348

0.404 4 0.451 1 0.407 7 0.405 5 0.418 8 0.438 8 0.427 7 0.443 3 0.417 7 0.437 7 0.455 5 0.404 4

0.449 9 0.425 5 0.449 9 0.425 5

- H DD 205705

0.153 3 0.179 9 0.189 9 0.176 6 0.189 9 0.159 9 0.195 5 0.149 9 0.164 4 0.186 6 0.196 6 0.179 9 0.193 3 0.180 0 0.184 4

- H DD 210931

0.100 0 0.104 4 0.113 3 0.102 2 0.094 4 0.102 2 0.114 4 0.096 6 0.096 6 0.103 3 0.093 3 0.091 1

Ah Ah

-0.467 7 -0.474 4 -0.415 5 -0.408 8 -0.403 3 -0.457 7 -0.403 3 -0.393 3

0.423 3 0.454 4 0.416 6 0.421 1 0.430 0 0.442 2 0.434 4 0.449 9 0.425 5 0.443 3 0.460 0 0.418 8

0.453 3 0.439 9 0.448 8 0.435 5

0.142 2 0.170 0 0.180 0 0.167 7 0.176 6 0.152 2 0.189 9 0.144 4 0.156 6 0.178 8 0.187 7 0.173 3 0.187 7 0.173 3 0.177 7

0.030 0 0.036 6 0.053 3 0.025 5 0.025 5 0.033 3 0.054 4 0.026 6 0.026 6 0.037 7 0.024 4 0.019 9

Av Av

-0.436 6 -0.435 5 -0.414 4 -0.398 8 -0.397 7 -0.437 7 -0.407 7 -0.386 6

0.465 5 0.479 9 0.459 9 0.461 1 0.467 7 0.471 1 0.470 0 0.477 7 0.458 8 0.474 4 0.483 3 0.458 8

0.477 7 0.468 8 0.475 5 0.467 7

0.187 7 0.210 0 0.222 2 0.211 1 0.218 8 0.197 7 0.226 6 0.190 0 0.199 9 0.222 2 0.228 8 0.215 5 0.229 9 0.215 5 0.217 7

-0.107 7 -0.100 0 -0.085 5 -0.108 8 -0.111 1 -0.099 9 -0.082 2 -0.102 2 -0.103 3 -0.097 7 -0.110 0 -0.112 2

Au Au

-0.722 2 -0.688 8 -0.569 9 -0.566 6 -0.596 6 -0.677 7 -0.590 0 -0.572 2

0.243 3 0.346 6 0.248 8 0.240 0 0.271 1 0.325 5 0.287 7 0.334 4 0.273 3 0.325 5 0.365 5 0.240 0

0.354 4 0.297 7 0.345 5 0.297 7

0.068 8 0.127 7 0.118 8 0.097 7 0.127 7 0.080 0 0.136 6 0.070 0 0.085 5 0.135 5 0.129 9 0.104 4 0.118 8 0.126 6 0.125 5

-0.601 1 -0.580 0 -0.541 1 -0.597 7 -0.601 1 -0.585 5 -0.542 2 -0.576 6 -0.577 7 -0.563 3 -0.600 0 -0.615 5

159 9

446 6 H.. Hensberge et al.: Photometry of Silicon Stars

Lett us focus on HD 3580 and HD 212432. In the followingg section it is argued that the period of HD 35800 is either Pl = \d4S or P2 = 2d94. Inspection of the lightt curves for each of the two cases does not lead to aa final choice. The second period induces a double-wavee light curve. The S-test prefers P,: S(P,)%0.35 whilee S(P2)> 0.6 in all filters. The S*-test gives S*(P2)*S*(P,)*S(P,). .

Thee case of HD 212432 is more illustrative. Ac-cordingg to the S-test P1 = ld65 and P2 = 2d52 are the bestt candidates. However S>0.6 in both cases in all filters.filters. The plots of the observations for periods P, or P22 are not convincing. The S*-test prefers P, to P, : thee u-variation, which has the largest amplitude and hencee should be the best indicator of the period, gives noo satisfying result (S*%0.6). However P = 4d7 happens too be an excellent candidate with S* <0.15 in each of thee four filters. This candidate could easily be dis-cardedd in the S-test, since S(P3)«1 as a consequence off the large phase-gaps containing no observations in thiss double-wave light curve. In these S*-tests a Q value off 0.1 was adopted.

Thereforee it may be concluded that the S-test favourss the selection of single-wave light curves. How-ever,, as many Ap stars are known to possess light curvess with a secondary extremum, the S*-test is more selectivee in the case of period determination for these stars,, especially when only a restricted number of observationss is available.

11 ' ' ' I .00 2 .4 .6 .8 .0 2 _

AV V

AA a A -

AA A A A

Ö

"A"A « " A "

11 , , , , 1

-- Ab £"

AA A A

** A V ^

" A ^^ A ' r f .

11 , . , , 1

A y A A

A A AA A

nn a.

AA A

^^ A A A A

j-022 0 .2 .4 .6 .8 0 .2 I t ,, , , , i P ^ » ,

11 , , , . , .00 .2 .4 .6 .8 0 2

A U U A A

A A A A

A A

A A

AA *

AA a ^

A A

AA ö

A ii » AA "

.00 .2 .4 .6 .8 .0 .2 11 , , . , 1

Fig.. 2. Photometric variations of H D 187473. Phas andd phase zero at JD 2 442 619.578

.. A U

" " ' '

----

-- A V

----

-- b

" "

-- A X

--

> >

i i .2 2

AA A

A AA a

A f ii A

^ AA *" A

1 1

/ ' ,

A A

00 .2

1 1

.4 4

A A

A ' '

^ ^ A A A

A A

a a

4 4

' '

r r

6 6

AA A

AA g

.6 6

i i

i i

.8 8

.rA A

AA ^ A

'a a

A / 11 * A

^^ A

.8 8

| |

1 1 00 .2

--

AA A

t aa a

I I

' A ' * »

" "

A A

A A * * A A A A A

00 .2 || phase,

Fig.. 1. Photometric variations of HD3580 in the instrumental systemm of the Danish 50 cm telescope at La Silla, plotted with respect too a period of ld48 and phase zero at JD 2 442 619.890

160 0

H.. Hensberge et al.: Photometry of Silicon Stars 447 7

&&&& A

I" " iphasi i

Fig.. 3. Photometric variations of HD 206653, plotted in a period P== 1?788. Phase zero corresponds to JD 2 442 619.687

3.. Results AA summary of our results, including standard indices, amplitudee and periodicity of variations, is listed in Tablee 1. The standard indices enable us to estimate thee effective temperature of the silicon stars involved, sincee u - b is a relevant temperature indicator (Megessier, 1971).. Stars showing the silicon anomaly generally havee 1.3>u-b>0.45 corresponding to effective tem-peraturess between 11000 K and 18000 K. Our program starss have u-bxO.62 (HD 3580), 0.77 (HD 206653), 0.788 (HD 207188), 0.93 (HD 187473) and 1.05 (HD 212432).. This places HD 3580 at the same temperature ass 41 Tau and somewhat hotter than Babcock's star HDD 215441.

Inn the subsequent section we discuss the results forr each of the variable stars separately. One should noticee that the plots represent the variations in the instrumentall system. Phase zero is chosen arbitrarily att the time of the first observation. The transformed differentiall magnitudes are given in Table 2.

A.A. HD3580 Thee observations are plotted in Figure 1; phases are com-putedd with respect to a period of ld48. All light curves

Fig.. 4. Photometric variations of HD 207188. Phases were computed withh P = 2?67 and phase zero at JD 2 442 619.827

aree in phase with each other. Although the variations aree rather small, the lightcurves are well defined. The b—yb—y and ml indices are intrinsically constant.

However,, also P=2d94 represents the observations equallyy well. In this case, all curves show secondary extrema.. Unfortunately, the curves with this period aree not as well defined as with P=ld48 because there remainn three gaps of 0.2 in phase where no observa-tionss exist. This is, however, not a sufficient reason for eliminatingg this longer period.

Accordingg to the accuracy of the measurements— demonstratedd by the accuracy of differential magnitudes betweenn comparison stars—we can place the following limitss on the value of P: ld d005 or 2d d02.

B.B. HD 187473 Sincee the comparison star HD 185183 proved to be variable,, only two comparison stars were used. HD 1874733 is variable with extra-ordinary amplitude: aboutt 0T1 in y and b, 0T07 in v and more than 0T2 in u. Weakk secondary extrema show up in M; they are less pronouncedd in the other channels. Maximum brightness inn y corresponds to a maximum in b — y, which shows noo secondary extrema anymore. The variation in m 1 is

161 1

448 8 H.. Hensberge et al.: Photometry of Silicon Stars

-- AU

--

.0 0 .2 2 .4 4

i i

6 6 .8 8

t> t>

öö <&

1 1

A U U

-- Ay

„Ab b

" "

" "

_Ay y

------

T.02 2

G G

"t, "t,

| |

A A

AA A

| |

A A

| |

6 6

P P 1 1

.2 2

„» »

* * & &

^ * *

. . a & &

.2 2

*ft t

A A

& &

.4 4

* *

& &

& &

* *

.4 4

6 6

^ ^

\ \

* *

a & &

a a

.6 6 i i

i i

.8 8

* *

i i

" "

& &

.8 8

11 ' : .00 .2

»' '

/ f i i

1 1

» » * *

1 1

--

ss -

A A

1 1

A A

ftft a.*.

öö Aw

" " 00 2 || phase |

Fig.. 5. Photometric variations of HD 212432. Phases were computed withh P = 4d69 and phase zero at JD 2 442 619.861

lesss clear. All curves are in phase with the period PP = 4d d04(Fig. 2).

C.C. HD 206653

Thee observations are well represented with P= l d d0055 (Fig. 3). All curves are well defined single wavess in phase with each other; b—y varies slightly, m\m\ remains constant.

D.D. HD 207188

Figuree 4 shows the results plotted in the period P=2d67. Secondaryy extrema occur in the four channels, but less pronouncedd in w; b — y and ml are constant. Obviously observationss are lacking around primary minimum. Thee frequent observations which define the slope betweenn phases 0.7 and 0.0 permit to rely on the period upp to d01.

E.E. HD 212432

Differentiall magnitudes are plotted in Figure 5. Phases aree computed with P=4d69. The accuracy of P is

d05.. Pronounced secondary extrema should exist

Fig.. 6. HD 185183 proved to be variable in the uvby system. The observationss are plotted with respect to a period P= 1?737 and phase zeroo at JD 2 442 619.578

butt they are scarcely defined by lack of observations betweenn phases 0.25 and 0.65—except two points near phasee 0.45. Again, ml is constant and the b-y variation iss hardly detectable. Further observations are required too confirm this first set. However no alternative value off P is consistent with our observations within the accuracyy limits set by differential magnitudes between comparisonn stars.

Thiss star is rather at the cool end of stars with siliconn anomaly. Consequently, the importance of the contributionn of line blocking due to metals is probably growing,, resulting in a much larger ml index.

P HDD 185183

Thiss star was selected as comparison star of HD 187473. Thee reductions showed that this star is obviously variablee in the four channels within a range of 0T08. Thee star is classified in the Henry Draper catalogue as B9.. The effective temperature derived from the u—b valuee of 0.68 is about 14500 K; its mean colour indices andd the amplitude of the variations, as well as its T_((

aree similar to those of silicon stars. We therefore

162 2

H.. Hensberge et al.: Photometry of Silicon Stars 449 9

decidedd to search for a periodicity in the order of days. Thee best correlation was found near P=1?737. The dispersionn around the mean curves displayed in Figure 6 iss larger, however, than in other diagrams, above all inn y. This can be partially due to the fact that each comparisonn star was observed less frequently than a programm star as appears from Section 2. So one blue 122 A/mm spectrogram was taken in May 1976 by Dee Loore at the ESO. The star appears to have relatively strongg but broad silicon lines. We therefore suggest thatt this might be a broad-lined silicon star.

5.. Conclusions

Alll five silicon stars are variable. Variations in the four channelss are in phase; the largest amplitude occurs inn all cases in u. One of the stars, HD 187473, shows extra-ordinaryy large amplitudes, comparable among

siliconn stars only with those of HD 215441 (Babcock's star).. It should be noticed that at least three and possibly fourr stars show more or less pronounced secondary extrema. . Acknowledgements.Acknowledgements. H. Hensberge acknowledges support by the Nationall Foundation of Collective Fundamental Research of Belgium (FKFO)) under No. 10303.

E.. J. Zuiderwijk acknowledges support by the Netherlands Organisationn for the Advancement of Pure Research (ZWO).

References s

Bidelman,, W.P., Mac Connell.D.J.: 1973, Astron. J. 78, 687 Gr0nbech,B,, 01sen,E.H„ Strömgren,B.: 1975, private communica-

tion n Gr0nbech,B.,, 01sen,E.H., Strömgren,B.: 1976, Astron. Astrophys.,

too be published Hensberge,, H, de Loore,C, Zuiderwijk, E.J., Hammerschlag-Hens-

berge,G.:: 1976, Astron. Astrophys. 48, 383 Lafler,J.,, Kinman,T.D.: 1965, Astrophys. J. Suppl. 11, 216 Megessier,G:: 1971, Astron. Astrophys. 10, 332

163 3