The purpose of our study was twofold. On the one hand, we observed repeatedly stars already known to have resolved magnetically split lines, to study the variation of their mean field modulus through their rotation cycle. On the other hand, we pursued a systematic search for additional Ap stars with magnetically resolved lines. We started by taking spectra of stars which for some reason we suspected to have resolved lines. But as time passed and as we obtained data for an increasing number of stars, we progressively widened the search, which should soon be complete for a magnitude limited sample.
Table 4: Mean magnetic field modulus
We report here the discovery of 16 new stars with resolved magnetically split lines: HD 965, HD 29578, HD 59435, HD 61468, HD 70331, HD 75445, HD 93507, HD 119027, HD 142070, HD 144897, HD 150562, HDE 318107, HD 166473, HD 177765, HD 208217, and HD 216018. With the 13 stars in which the detection of magnetically resolved lines had been reported in Papers I to III, a total of 29 stars having that property have been found within the framework of the project described in this paper. Thus our work raised to 41 the number of Ap stars with resolved magnetically split lines known. Even more recently, a 42nd star of this type, HD 47103, was discovered by Babel et al. (1995).
Table 2 (click here) summarizes some properties of the 42 stars with resolved
magnetically split lines known to this date. On the left page,
the first four columns contain the HD/HDE number, another identification,
the V magnitude, and the spectral type, as it appears in the
catalogue of Renson et al. (1991). The next two columns give
information about the stellar rotation period and the corresponding
reference. When the latter is not given, the period information in
Col. 5 comes from the present paper. The heliocentric
Julian date taken as origin for the computation of the rotation
phases, the particular property (e.g. a magnetic extremum)
characterizing the phase origin, and the relevant reference, appear in
Cols. 7 to 9. Again, the absence of a reference in Col. 9 means that
the corresponding information is given in the present paper.
Table 2 (click here) continues on the opposite page, where in the first column
the HD/HDE number of the star is repeated, for the sake of clarity. In
Col. 2, we give the Strömgren photometric index (b-y), retrieved
(whenever available) from Renson et al. (1991)
or from Martinez (1993). Columns 3 and
4 of the right page give respectively the average 
and the ratio 
of the
observed maximum and minimum values of the mean magnetic field
modulus. 
is not the arithmetical average of all
our measurements of the field modulus of the considered star, but rather
an estimate of the mean value of 
over the rotation cycle.
When the star has been repeatedly observed throughout this
cycle, 
characterizes well its actual mean field
strength, while 
is representative of the amplitude of its
variations. When the phase coverage achieved so far is incomplete,
these quantities only give a preliminary indication; the
corresponding entries appear in italics (
) or as a lower
limit (
) in Table 2 (click here). The use of
italics has the same meaning for the minimum (
) and the
maximum (
) of the mean longitudinal magnetic field,
which are found in Cols. 5 and 6. The references from which the
longitudinal field information has been retrieved are mentioned in the
text. Columns 7 and 8 give the date of the first observation of
magnetically resolved lines in the considered star (when known;
otherwise information about the discovery year is given) and
the reference of the paper where this discovery has been announced
(no entry in Col. 8 means that the presence of resolved split lines is
reported here for the first time). Finally, some additional remarks
appear in Col. 9 (SB stands for spectroscopic binary; SB2 for
double-lined spectroscopic binary; and
roAp denotes a rapidly oscillating Ap star - see
e.g. Kurtz 1990), with the corresponding reference in Col. 10
(if it is not the present paper).
The properties summarized in Table 2 (click here) are discussed in more detail in the rest of this paper (especially in the next section).
From the observations described in Sect. 3, we have repeatedly
measured the mean magnetic field modulus of 40 of the 42 stars with
magnetically resolved lines, inferring it from the wavelength
shift between the components of the line 
, as explained in
Sect. 2. A portion of one spectrum of each of these 40 stars,
comprising the line 
, is shown in Figs. 2 (click here) to 4 (click here).
The two stars appearing in Table 2 (click here) for which we present no
results here are HD 47103 and HD 215441. The discovery (Babel et al.
1995) of resolved magnetically split lines in the former is too
recent (let us recall that this report is limited to data acquired before
end of August 1995). A number of spectra of HD 215441 have been recorded
in the course of this programme, but its magnetic field is so strong
(34 kG) that one can no longer use Eq. (1) to interpret the splitting
of 
: the latter must be modeled accounting properly for the
physics of the partial Paschen-Back effect. This will be the subject
of a future publication.
We present an overview of the measurements of the mean magnetic field modulus of the 40 remaining stars in Table 3 (click here). The columns give, in order, the HD or HDE number of the star, the number n of measurements of its magnetic field modulus that we have performed, the average and the standard deviation of these measurements, their rms deviation about a best fit curve of their variations, and their estimated uncertainty. The data appearing in the last two columns are explained more in detail in Sect. 6. Note that in contrast with Table 2 (click here), the quantity given in Col. 3 of Table 3 (click here) is the plain arithmetic average of all our measurements of the considered star. The standard deviation about this average, which appears in Col. 4, results from the superposition of the measurement uncertainties and of the variations of the stellar field.
Table 5: Correspondence between the instrumental configuration code
numbers (see Col. 2 of Table 1 (click here)) and the symbols used in the plots
The 752 individual measurements of the mean magnetic field
modulus are presented in Table 4 (click here). Column 1 contains the
heliocentric Julian date of mid-observation. An asterisk next to this
entry identifies the spectrum of each star that is shown in Figs. 2 (click here)
to 4 (click here). In Col. 2, the stellar
rotation phase is computed (whenever possible) using the ephemeris
elements appearing in Cols. 5 and 7 of Table 2 (click here). The values of the
mean magnetic field modulus 
are given in Col. 3 of Table 4 (click here),
with in Col. 4 the code referring to the instrumental configuration used to
perform the corresponding observation (as defined in Col. 2 of
Table 1 (click here)). All the mean
field modulus measurements that we have performed before end of
August 1995 are included in Table 4 (click here). A few of them correspond to
observations that have already been presented in Papers I to III.
However, for the sake of homogeneity and completeness, these spectra
have been remeasured together with all the others discussed here: the
revised values of the field that appear in this paper supersede
the ones that had been previously published. Note that the Julian dates
of some of the observations reported in Paper II were erroneous: they
have been corrected in the present paper.
In the next section, we discuss the mean field modulus measurements star by star. We present plots of these measurements against rotation phase or against Julian date. In these plots, different symbols are used to distinguish measurements performed with different instrumental configurations. Unless indicated otherwise in the figure caption, the meaning of these symbols is given in Table 5 (click here). Column 1 contains the identification number of the configuration (from Col. 2 of Table 1 (click here)), of which a short reminder is given in Col. 2 (which should be self-explantory from the consideration of Table 1 (click here)). The description of the corresponding plot symbol appears in Col. 3.
Figure 5: Mean magnetic field modulus of HD 965 against heliocentric
Julian date. The meaning of the symbols is as given in Table 5 (click here)
