06 April 2021

AperTO - Archivio Istituzionale Open Access dell'Università di Torino

Original Citation:

The magnetic remanence of hematite-bearing murals

Published version:

DOI:10.1029/2009GL041198

Terms of use:

Open Access

(Article begins on next page)

Anyone can freely access the full text of works made available as "Open Access". Works made available under a
Creative Commons license can be used according to the terms and conditions of said license. Use of all other works
requires consent of the right holder (author or publisher) if not exempted from copyright protection by the applicable law.

Availability:

This is the author's manuscript

This version is available http://hdl.handle.net/2318/92886 since



Magnetic remanence of hematite-bearing murals

R. Lanza,
1
E. Zanella,

1
and S. Saudino

1,2

Received 6 October 2009; revised 12 November 2009; accepted 19 November 2009; published 16 December 2009.

[1] We report on a series of experiments designed to test
the ability of hematite-bearing colors to record the direction
of the ambient magnetic field. Plasterboards accurately
oriented with respect to the Earth’s magnetic field were
painted with red tempera colors prepared with hematite
pigments. Magnetic measurements indicate that the color
film retains a remanent magnetization and acquires a well
developed magnetic fabric. The remanence direction is
close to, yet slightly deviated from the Earth’s magnetic
field. The deviation is interpreted to result from preferential
alignment of the pigment grains parallel to the plasterboard
surface and depends on both its orientation with respect to
magnetic north and the degree of magnetic anisotropy of the
color film, which in turn varies according to the pigment
used. Investigation of the magnetic remanence of murals may
complement archaeomagnetic information derived from
traditional materials such as baked and fired structures.
Citation: Lanza, R., E. Zanella, and S. Saudino (2009), Magnetic

remanence of hematite-bearing murals, Geophys. Res. Lett., 36,

L24302, doi:10.1029/2009GL041198.

1. Introduction

[2] Red colored, hematite-bearing mural paintings have
been shown to carry a remanent magnetization close to the
geomagnetic field direction at the time they were painted,
known either from direct historical measurements [Chiari
and Lanza, 1997, 1999] or archaeomagnetic data [Zanella et
al., 2000; Goguitchaichvili et al., 2004].
[3] The basic model for acquisition of a pictorial remanent

magnetization (PiRM) supposes that when the color is applied
to a wall, the hematite grains behave as tiny magnets free to
move and align their moment parallel to the Earth’s field. Once
the color dries, the grains are locked and magnetization is
preserved over time. This model was tested in laboratory
controlled conditions: colors were prepared with high-quality
pigments and used to paint oriented plasterboards. The natural
remanent magnetization was measured and its direction
checked against that of the laboratory field. A deeper under-
standing of the remanence characteristics was obtained from
measurement of the anisotropy of isothermal remanent mag-
netization (AIRM). Measurements were done at the ALP
laboratory (Peveragno, Italy) using a JR-6 spinner magnetom-
eter, a 2-G Enterprises degausser, a Molspin AF demagnetizer,
a Bussi pulse magnet and a Schonstedt thermal demagnetizer.

2. Samples and Remanence Measurements

[4] Two red colors were prepared using a 4 to 1 mixture
of water and egg yolk as binder (90% in volume) and

commercial hematite pigments (10%), namely Morellone
and Rosso di Marte (Zecchi, Fine art and restoration
materials – Firenze), which are equivalent to the pigments
with the same names used by Italian Renaissance painters.
Scanning Electron Microscope (SEM) images (Figure 1)
demonstrate that the grain size of both pigments is less than
0.5 mm and that the Morellone grains are mostly flake-
shaped, and that the Rosso di Marte ones are acicular. The
colors were applied to plasterboards placed in grassland
opposite the laboratory, in order to avoid any unevenness of
the magnetic field inside the building. The plasterboards
were accurately oriented with respect to the Earth’s mag-
netic field using bubble levels, a compass and a 3-axes
fluxgate magnetometer. Orientation accuracy was estimated
as ±1�. Samples were taken with the flexible plastic disk
(8 = 18 mm) technique [Chiari and Lanza, 1999;
Goguitchaichvili et al., 2004]. In order to make possible
thermal demagnetization of the color film, one basal face
of oriented standard cylindrical paleomagnetic specimens
(8 = 25.4 mm, h = 23 mm) was also painted. To avoid any
bias in measuring the PiRM, diamagnetic limestones were
used and the possible remanence carried by incidental
ferrimagnetic grains was checked before painting. The
measured signal was of the order of 10

�11
Am

2
, similar

to the nominal sensitivity (2.6 � 10�11 Am2) of the JR-6
spinner magnetometer.
[5] Magnetometers used in paleomagnetism are designed

to measure specimens usually 8 to 11 cm
3
in volume,

whereas the disk-shaped specimens of color film are some
tens of microns thick and a few cubic millimeters in
volume. The result of a measurement might therefore
depend on the position of the specimen relative to the
pick-up coils of the spinner magnetometer. To check this
possible effect, two groups of ten specimens were measured
in two positions. In the first one, the diameter of the disk
coincided with the spinner axis, while in the second the disk
was shifted 10 mm sideways. The group mean directions are
statistically indistinguishable (Table 1), which suggests that
the small, 1� to 3�, differences between the directions of
individual specimens reflect random errors probably due to
the positioning of specimens within the spinner specimen
holder.
[6] Six vertical plasterboards with different orientations

with respect to magnetic north were painted. The azimuths,
measured clockwise, were: 0�, 30�, 45�, 90�, 315� and
330�. Various portions of individual plasterboards were
painted with different brushstroke orientation: up-down,
side to side, random. Ten specimens for each color were
collected from each plasterboard. They were measured and
the mean direction of each group was calculated using
Fisher [1953] statistics. The painting always acquired a
remanent magnetization with magnetic moment in the
order of 3–6 � 10�9 Am2 and directions tightly grouped.
The angular dispersion was usually smaller for Morellone

GEOPHYSICAL RESEARCH LETTERS, VOL. 36, L24302, doi:10.1029/2009GL041198, 2009
Click
Here

for

Full
Article

1
Dipartimento di Scienze della Terra, Università di Torino, Turin, Italy.

2
Now at Dipartimento di Ivrea, ARPA Piemonte, Ivrea, Italy.

Copyright 2009 by the American Geophysical Union.
0094-8276/09/2009GL041198$05.00

L24302 1 of 5

http://dx.doi.org/10.1029/2009GL041198


(0.8� � a95 � 2.7�) than for Rosso di Marte (2.4� � a95 �
5.1�). The directions were not affected by the brushstroke
orientation and close to that of the local ambient field
(D = 0�, I = 59�), although small differences (0� to 9�)
in declination were detected (Figure 2). The differences
look systematic because the value of the deviation angle is
a function of the azimuth of the plasterboard and the
deviation is opposite in sense to the plasterboard orienta-
tion: it is clockwise when the plasterboard azimuth falls in
the NE-SW quadrant, and counterclockwise when in the
NW-SE quadrant. One horizontal plasterboard was painted
and the inclination of the painting remanence was 6� to

12� shallower than that of the ambient field. The possible
origin of the observed deviations is discussed below.

3. Characteristics of the Acquired Remanence

[7] The PiRM stability is good, which is typical of
remanences carried by powdered hematite. The PiRM
direction does not change throughout stepwise thermal
and AF demagnetization. More than 50% of the initial
remanence is still present after heating at 580�C, and some
60% remains after AF demagnetization at a 280 mT peak-
field. Similar observations were made for some murals at
the Bibliotheca Apostolica in Rome [Chiari and Lanza,
1999]. This demonstrates that hematite-bearing mural paint-
ings carry a stable record of the Earth’s field at the time they
were painted and may be used in archaeomagnetic studies,
as in the case of Pompeii [Zanella et al., 2000].
[8] The observed deviation of the remanence direction

from the ambient field needs further investigation. The
inclination shallowing in the case of the horizontal plaster-
board might be regarded as analogous to that typical of a
depositional remanent magnetization (DRM). Preferential
orientation of non-equant detrital grains is caused by gravity
in the DRM case, whereas in the PiRM one it could be
caused by other phenomena acting on the pigment grains,
such as surface tension in the color films, which parallels
the painted surface, or grain anisotropy in the hematite
pigment.
[9] Vertical plasterboards give a clue to understand the

deviation. Let b be the angle between magnetic North and
the azimuth of the plasterboard (Figure 3) and d the angle
between the plasterboard and the PiRM horizontal compo-
nent. Should the PiRM direction coincide with magnetic
North, d = b. The fact that d � b means that the PiRM
direction is systematically deviated toward the painted
surface and that it does not coincide with magnetic North.
The systematic deviation may be accounted for by prefer-
ential orientation of the pigment grains, which would tend
to orient their long dimension parallel to the color film,
namely to the painted surface. The easy magnetization
direction of hematite lies in the lattice basal plane, whereas
the hard direction parallels the axis of symmetry. The long
dimension of hematite grains usually corresponds to the
basal plane so that their preferential orientation would bias
the remanence direction. This hypothesis can be tested by
investigating the magnetic fabric of the color film by means
of remanence anisotropy measurements. AIRM was thus

Figure 2. Equal-area projection of the PiRM directions
from plasterboards painted with Morellone (circles) and
Rosso di Marte (squares) pigments. Full/open symbols:
vertical/horizontal plasterboards; star = laboratory ambient
field.

Figure 1. SEM images of (a) Morellone and (b) Rosso di
Marte pigments.

Table 1. Mean PiRM Direction and Intensity for Morellone

Specimens From Plasterboards With Different Orientations and

Measured in Different Positions Within the JR-6 Coils
a

Azimuth Position n D, I (deg) M (Am
2
) k a95 (deg)

30� Central 10 2.7, 57.3 4.3 10�9 318 2.7
30� Offset 10 4.5, 57.0 4.3 10�9 544 2.1
330� Central 10 358.3, 58.4 3.3 10�9 118 4.5
330� Offset 10 359.7, 59.2 3.3 10�9 148 4.0

a
Key: azimuth: plasterboard orientation; position: position of the specimen

within the measuring coils; n = number of specimens; D, I = declination,
inclination; M = magnetic moment; k, a95 = precision, semi-angle of
confidence from Fisher [1953].

L24302 LANZA ET AL.: MURAL REMANENCE L24302

2 of 5



measured on specimens from vertical and horizontal plaster-
boards. The specimens were subjected to tumbling AF
demagnetization in the maximum available peak-field of
100 mT and were then given a direct field of 80 mT for the
Morellone and 60 mT for the Rosso di Marte specimens.

IRM acquisition curves showed that these values were
enough to give the specimens a magnetic moment in the
order of 1 � 10�7 Am2. The procedure was repeated twelve
times: the direct field was applied in six different positions
according to the scheme of Jelinek [1996], and in two
opposite directions for each position, in order to cancel
the NRM component harder than 100 mT. The magnetic
fabric is similar in all specimens and is always well
developed. The minimum axes I3 are well grouped and
are orthogonal to the plasterboard plane (Figure 4), which
coincides with the magnetic foliation; the maximum I1 and
intermediate I2 axes are more or less dispersed within the
foliation plane. The anisotropy degree, P = I1/I3, is higher
for the Rosso di Marte specimens (P = 1.55) than for the
Morellone (P = 1.18) specimens.
[10] According to Uyeda et al. [1963] the relationship

between the actual direction of the remanence and that of
the external magnetizing field is (using our symbols from
Figure 3) tan d = 1/P tan b, where P is the degree of
anisotropy. Experimental values of tan d vs. tan b (Figure 5a)
have good linear correlation; interpretation of the Morellone
curve using the above equation gives a calculated value
P = 1.14, which is consistent with the experimental value of
P = 1.18, whereas the agreement is less good for the Rosso di
Marte specimens, whose calculated and experimental values
are P = 1.27 and P = 1.55, respectively. The equation of
Uyeda et al. [1963] may thus be used to calculate the

Figure 4. Equal-area projection of AIRM principal directions: (a, c) vertical plasterboards; (b, d) horizontal plasterboards.
Symbols: squares = maximum axis, I1; triangles = intermediate axis, I2; solid circles = minimum axis, I3; thick great
circles = magnetic foliation.

Figure 3. Geometrical sketch of a hypothetical PiRM
vector and plasterboard orientation. Symbols: Nm =
magnetic North; J = PiRM vector; H, Z = PiRM horizontal
and vertical components; b = plasterboard azimuth; d =
angle between plasterboard and H; b � d = D, declination
error.

L24302 LANZA ET AL.: MURAL REMANENCE L24302

3 of 5



PiRM declination error D = b � d = b � tan�1 (1/P tan b)
(Figure 5b), i.e. the expected deviation of the PiRM from the
geomagnetic field.
[11] Since the deviation is clockwise or counterclockwise

according the wall azimuth is in the NE or NW quadrant, its
effect on the mean PiRM direction derived from different
walls in the same building is strongly reduced. Paleomag-
netic directions for murals sampled from the four walls of a
room at Palazzo Venturi-Gallerani (Siena, Italy), which
were painted in 1794, are slightly different (Figure 6), yet
their mean (D = 341�, I = 65�) is indistinguishable from the

coeval Earth’s magnetic field direction (D = 343�, I = 64�)
from direct historical measurements [Cafarella et al., 1992].
[12] The equation of Uyeda et al. [1963], written in the

form tan IDRM = 1/P tan IField, is often used to estimate
inclination shallowing in sedimentary rocks,. In the case of
horizontal plasterboards, the equation works well in the case
of the PiRM carried by Rosso di Marte, and gives the same
shallowing (12�) as the observed value. The calculated
shallowing in the Morellone (5�) case is close to or half
of the observed values (6� and 10� respectively).

4. Discussion and Conclusions

[13] On the basis of the experimental results, we conclude
that the PiRM direction acquired by a mural painting is a
stable remanence whose direction is slightly deviated from
that of the Earth’s magnetic field as a function of the wall
orientation and the degree of magnetic anisotropy of the
color film. In terms of archaeomagnetic investigations, the
deviation of the PiRM direction relative to the ambient field
does not imply that murals are unreliable sources of
information on secular variation (SV). On the one hand,
the deviation is zero for walls oriented N-S or E-W. On the
other, the expected deviation may be evaluated from AIRM
measurements. Moreover, since the error depends on the
wall orientation, it may be cancelled by sampling different
walls in the same room or building and calculating the mean
PiRM direction. In conclusion, our simple model for PiRM

Figure 5. (a) Tangent of PiRM declination (d) vs. tangent of plasterboard azimuth (b) measured from magnetic North (see
Figure 3 and text for further explanation). Symbols: solid circles = Morellone (r = 0.997) specimens, and squares = Rosso di
Marte (r = 0.976) specimens. (b) Calculated PiRM declination error D = (b � d) as a function of wall orientation (b).

Figure 6. Equal-area projection of the PiRM directions
from murals painted on the four walls of a room at Palazzo
Venturi-Gallerani (Siena, Italy). Symbols: solid circles =
individual mural mean PiRM direction; star = Earth’s
magnetic field direction from historical measurements in
1794.

L24302 LANZA ET AL.: MURAL REMANENCE L24302

4 of 5



acquisition appears to work well enough for archaeomag-
netic purposes. Investigation of new types of materials, such
as mural paintings and plasters [Soler-Arechalde et al.,
2006], extends the use of archaeomagnetism beyond the
traditional baked and fired structures and provides new
opportunities to date archaeological finds.

[14] Acknowledgment. The authors are indebted to G. Chiari for
providing the samples from the Palazzo Venturi-Gallerani murals.

References
Cafarella, L., A. De Santis, and A. Meloni (1992), Secular variation from
historical geomagnetic field measurements, Phys. Earth Planet Inter., 73,
206–221, doi:10.1016/0031-9201(92)90091-9.

Chiari, G., and R. Lanza (1997), Pictorial remanent magnetization as an
indicator of secular variation of the Earth’s magnetic field, Phys. Earth
Planet Inter., 101, 79–83, doi:10.1016/S0031-9201(96)03222-0.

Chiari, G., and R. Lanza (1999), Remanent magnetization of mural
paintings from the Bibliotheca Apostolica (Vatican, Rome), J. Appl.
Geophys., 41, 137–143, doi:10.1016/S0926-9851(98)00038-X.

Fisher, R. A. (1953), Dispersion on a sphere, Proc. R. Soc. A, 217,
295–305, doi:10.1098/rspa.1953.0064.

Goguitchaichvili, A., A. M. Soler, E. Zanella, G. Chiari, R. Lanza,
J. Urrutia-Fucugauchi, and T. Gonzalez (2004), Pre-Columbian mural
paintings from Mesoamerica as geomagnetic field recorders, Geophys.
Res. Lett., 31, L12607, doi:10.1029/2004GL020065.

Jelinek, V. (1996), Theory and measurement of the anisotropy of isothermal
remanent magnetization of rocks, Travaux Geophys., 37, 124–134.

Soler-Arechalde, A. M., F. Sánchez, M. Rodriguez, C. Caballero-Miranda,
A. Goguitchaichvili, J. Urrutia-Fucugauchi, L. Manzanilla, and
D. H. Tarling (2006), Archaeomagnetic investigation of oriented pre-
Columbian lime-plasters from Teotihuacan, Mesoamerica, Earth Planets
Space, 58, 1433–1439.

Uyeda, S., M. D. Fuller, J. C. Belshé, and R. W. Girdler (1963), Anisotropy
of magnetic susceptibility of rocks and minerals, J. Geophys. Res., 68,
279–291, doi:10.1029/JZ068i001p00279.

Zanella, E., L. Gurioli, G. Chiari, A. Ciarallo, R. Cioni, E. De Carolis, and
R. Lanza (2000), Archaeomagnetic results from mural paintings and
pyroclastic rocks in Pompeii and Herculaneum, Phys. Earth Planet
Inter., 118, 227–240, doi:10.1016/S0031-9201(99)00146-6.

�����������������������
R. Lanza and E. Zanella, Dipartimento di Scienze della Terra, Università

di Torino, Torino, I-10125, Italy. (roberto.lanza@unito.it)
S. Saudino, Dipartimento di Ivrea, ARPA Piemonte, Ivrea, I-10015, Italy.

L24302 LANZA ET AL.: MURAL REMANENCE L24302

5 of 5