GEOFIZIKA VOL.

30 2013

Original scientific paper

UDC 551.594.5

De Aurora Boreali (1737) – Contemporary insight

to the young Bošković’s treatise
Vladis Vujnović and Inga Lisac
University of Zagreb, Faculty of Science, Department of Geophysics, Zagreb, Croatia

Received 6 November 2012, in final form 11 December 2013

An analysis of Bošković’s early paper concerned with the polar light (au-
rora borealis) gives an insight into the validity of the mathematical method
which the young Bošković applied to significant geophysical phenomena in order
to comprehend their nature. Based on the given data, we have examined nu-
merical results about the height of atmosphere and the heights of the auroras
which were observed over Europe in 1726 and in 1737. In his text there is not
a single mathematical formula. Following his instructions we have derived a
formula which may be used to determine the distance of an aurora by observing
it from one station. For the assumed aurora model, his method is applicable. We
find that described ideas on the physical cause of aurora, conceived in a fluid
coming from the sun, have elements of modern knowledge.

Keywords: Ruđer Bošković, aurora borealis, atmosphere dimensions and den-

sity

1. Introduction

Ruđer Josip Bošković (1711–1787), Croatian scientist and philosopher, one

of the last universal scholars, was born in Dubrovnik. At the age of 14 he entered
the Collegium Romanum where studied rhetoric, philosophy, theology, physics,
mathematics and astronomy. There he became a priest without pastoral duties,
had heavy teaching duties and conducted different research – purely scientific
or applied. After 1760 Bošković worked and resided in several European places,
became a member of scientific societes, and used to serve in diplomatic missions.
Belonging to Jezuits, distinguished promotors of science and scientific institu-
tions (Udias, 2003), he had easier approach to high society of the time. Having
a thorough knowledge of mathematical physics, he was engaged in different
scientific and practical tasks, in statistics, astronomy (improved instruments,
developed a method for finding orbit of celestical object from the three close posi-
tions), geodesy and cartography (measured meridian between Rome and Rimini,
introduced a sort of geodetic stand), geophysics, meteorology, optics, civil engi-
100 V. VUJNOVIĆ and i. lisaC: De Aurora Boreali (1737) – Contemporary insight ...

neering and archeology. Under his guidance the Brera astronomical observatory
was built.
Bošković is mainly recognized as a founder of a theory which unifies all forces
in nature. His fundamental work Theory of natural philosophy reduced to the
single law of forces existing in nature, printed 1758 in Vienna, influenced works
of many famous physicists and philosophers (Whyte, 1961).
Our comments refer to Bošković’s treatise which the 27-year old Bošković
submitted to the academic public on two occasions, first in August 1738 at the
Roman Seminary (Seminario Romano), and then in September of the same year
at the Roman College (Collegio Romano) (Bošković, 1738). In the education of
the Jesuits, scientific discussion were stimulated, in which prominent students
got a chance to prepare and lead discussion and shape its conclusions in a liter-
ary form. A leader of discussions was named „academician“. Title of treatise does
not bear his name, but it is acknowledged as his work and is listed in his refer-
ences (Kutleša, 2011).
The treatise has ten pages of text and eight figures. Original text in Latin
was translated to Croatian by Martinović (2012) from the specimen found in the
Archivium Ragusa, Dubrovnik. As far as we know, this is the first translation to
Croatian. The treatise we analyse formally consists of propositions and corol­
laries, and its subject belongs to the physics of the atmosphere. The first two
propositions serve to introduce measuring units used to describe the Earth’s size
and distances on its curved surface. In the first proposition, Roman miles are
used to express the Earth’s radius. In the second proposition an improved meth-
od to find the “height of the atmosphere” was described. In this way the reader
is ready to grasp the planetary dimensions of the natural phenomenon as au-
rora borealis is.
In the original text some circumstances are not explicitly stated. There is no
mention of the shape of the Earth, probably presuming that it is spherical. May
we speculate that the Earth was considered still, non-rotating, although Newton
already explained the Earth’s oblateness by its rotation. In the case of a spheri-
cal Earth, Bošković’s notion of the “largest circle” could be applied to meridians
as well as to the equator. Anyway, flattening of the Earth’s body should not influ-
ence accuracy of the physical dimensions of the natural phenomena treated in
this paper. In the time we witness, Jesuits officially adhered to Ptolemaic cosmol-
ogy. Notion of the Earth’s rotation axis was suppressed. In the fifth proposition
Bošković describes the geometry of the aurora and writes: “… aurora borealis is
a bright circle equally far from the Earth, with the center on the equatorial axis
(in Latin, axis Æquatoris), therefore it is parallel to the equator.” Although the
term “axes of a circle” in geometry is not usual, centre of aurora’s circle is con-
veniently placed on the Earth’s rotation axis. (Quotation marks we will use when
citing the original text, translated to English.) However, rotation of the Sun was
acknowledged, since proved by motion of the sunspots, and it was the crucial
geofizika, vol. 30, no. 2, 2013, 99–118 101

proof in sixth proposition, in the explanation of the aurora as a natural phenom-

enon.
Generally, text discloses Bošković as a young imaginative natural philoso-
pher who adopts new ideas of others, elaborates his own attitudes and mathe-
matical procedures.

2. Analysis of propositions one to four

The first proposition deals with the relation between different length units
and the Earth’s size. In order to recall past times, let us mention the Roman mile
and the Paris mile with corresponding smaller units: double step, feet, palm and
line. There are relations:

1 Roman mile = 1 000 double steps = 5 000 feet = 20 000 palms.

According to different sources, the Roman mile is about 1480 m. Jakobović

(1981) gives 1 Roman mile = 1478.5 m. For further calculation we will use the
round number 1480 m.
For the ratio between the Paris and the Roman measures Bošković uses

432 : 392 = 54 : 49 = 1.1020

Bošković quoted that one degree of the “largest Earth’s circle” accommodates

343 752 Paris feet = 378 828.734 Roman feet.

It follows: 68.7504 Paris miles = 75.7657 Roman miles.

For the planetary radius Bošković quotes 4341.0545 Roman miles, which is
equal to 6,425 km. He remarks than one degree of parallel going through Athens
equals 60 roman miles, while it has about 59.22 roman miles. This shows the
precision of the contemporary measurements; mean Earth’s radius is equal to
6371 km. Bošković realized his interest in measurements of our planet’s size
later on (1750–1752), when measuring the arc of the meridian through Rome
and Rimini together with Ch. Le Maire.
For the Ludolph’s number p, the ratio between half circumference and ra-
dius, Bošković uses 355 : 113 = 3.141 592 92, which differs a little from the more
exact value equal to 3.141 592 65.
The associated Corollary gives notion of the geographical foot. Bošković
gives the ratio:

1 geographical foot = 100 000 : 79 191.51 Roman feet = 1.263 Roman feet.

This ratio of the Roman and the geographical feet follows from the ratio of
numbers of Roman and geographical miles which are needed for one degree of
102 V. VUJNOVIĆ and i. lisaC: De Aurora Boreali (1737) – Contemporary insight ...

the largest circle, which means, 60 geographical miles = 75.7657 roman miles.
(The number 60 was also mentioned when stating that one degree of the parallel
through Athens, Greece, is equal to 60 Roman miles; nowadays we find it is 60.98
Roman miles.)
For the Roman miles we will hereinafter use the abbreviation r. m.
In the second proposition Bošković presents a geometrical model for
calculating the “height of the Earth’s atmosphere” – whatever it may mean. Also,
he states that this height “is higher than the true one”. This expression will be
elaborate further.
The model starts with assumption that the first solar rays are seen when
the Sun is 18° under the observer’s horizon. This angle is obviously the result of
astronomical practice. Today we also use the solar depression of 18° as the end
of astronomical twilight, 12° as the end of nautical twilight, and 6° as the end of
the civil twilight.
Twilight arises since the solar rays are scattered by air molecules. Bošković
elaborates the procedure of his predecessors who used only one reflection. In
order to get solar ray directly from the horizon, Bošković used a model with two
reflections (Fig. 1). The first reflection is at M, and the second at H; the deflected
ray reaches an observer at A. The height of the atmosphere is then defined as
the height above the earth where the second reflection is present. By dividing
the angle of 18° into four angles equalling 4° 30’, the both reflections should be
at the same height above the Earth’s surface. In this way one gets the same
atmospheric density at the place of reflections. If the angle ACF was divided into
three equal angles, the reflections would take place at different heights.
In the described method one has to calculate the distance BH which is a
part of triangle CAH, by using the known Earth’s radius and the angle of 4° 30’.
The phraseology for the 18th century is usuall one. Bošković writes that hy­
po­tenuse of triangle CAH is the „secans of the angle 4° 30’“. This is under­stand­

Figure 1. Equal to Fig. 1 in the original text (Bošković, 1738).

geofizika, vol. 30, no. 2, 2013, 99–118 103

able if the other triangle side AC is a measuring unit. Then the author explains
that the ratio of AC and BH is equal to the ratio of 100 000 versus 309.22, e.g.
as „whole sinus 100 000 versus excess of secans 4° 30’ above the radius“, i.e.
BH sec 4.5 − 1
= where sec 4.5° = 1.003 092 2.
AC 1

The „whole sinus“ means the sinus of the right angle; “excess of secans 4o 30’
above the radius” is sec 4.5° – 1.
Let us calculate: BH = AC ⋅ 0.003 092 2 ≈ 13.4 r. m. ≈ 20 km . This height appro­
xi­mately corresponds to the height of troposphere, the atmospheric layer of highest
density. When taking only one reflection at I, the height should be 54 r. m.
Now we can understand the phrase “height which is higher than the true
one”. If the refraction measured by Tycho Brahe was taken into account, then
the ray HA (the refraction is continuous as the ray proceeds towards the earth)
would be bent from below, lowering the point H named as a place of the “last
reflection”. The model used by Bošković neglects the continuous refraction.
We have to acknowledge that scientists at the beginning of 18th century had
a good insight into the extension of the atmosphere. Although Bošković’s result
is the best estimate of the atmoshpere height for his time, it is rather the result
of happy chance than of a sound model, since the passage of light through the
atmosphere is much more complex. Sky light is the multiply scattered light of
all primary rays. Moreover, one makes a logical error by drawing a primary ray
tangentially to the Earth’s surface from S to M, and from M to H; touching the
ground, ray should stop.

Contemporary knowledge of density variation with height, owing to Newton,

satisfies our knowledge, at least to the height of 50 km, as shown in the table:
Height Relative density

r. m. km Time of Bošković Today (Hedin, 1991)

0 0 1 1
7 10 1/4 1/3
14 21 1/16 1/14
21 31 1/64 1/63
28 41 1/256 1/280
35 52 1/1024 1/1180
70 104 10–6 3·10–4
140 207 10–12 2.7·10–10
210 311 10 –18
2.5·10–11
104 V. VUJNOVIĆ and i. lisaC: De Aurora Boreali (1737) – Contemporary insight ...

Bošković took the tabulated data from Newton’s Opticks, first published in
Latin in 1706. The cognition that the air is a real medium was proved by the
famous experiment of Magdeburg’s Lord Mayor Otto van Guericke in 1657, not
much before Newton’s analysis. Newton properly described gravitational sedi-
mentation of the atmosphere and calculated variation of its density. Bošković’s
achievement was in understanding the atmospheric density structure. By his
rational mind he adhered to Newtonists.
Bošković identifies the problem in the third proposition by saying: “To find
the nearest distance of aurora borealis, which is less than the true one.” This
phrase prepares readers to grasp the approximate manner of calculations.
The first two propositions were not only a sort of exercise before the main
task, but made a didactic introduction to the extension of the atmosphere whose
integral part aurora is.
In this proposition Bošković shows how to calculate the height of the aurora
seen over Europe on October 19, 1726, whose elevations were measured in
Tusculum (or Tuscul, a city close to Rome) and in Paris. The method – which
together with the measured elevations uses corrections for the atmospheric
refraction – is explained in Figs. 2. and 3.

Figure 2. Original Fig. 2., Bošković (1738).

Figure 3. Reduced version of the original Fig. 3. from Bošković (1738).

geofizika, vol. 30, no. 2, 2013, 99–118 105

First, in Fig. 2 (the original figure reproduced) Bošković explains the influence
of the atmospheric refraction, refering to the astronomical observation of Tycho
Brahe (1546–1601). The line RF is a tangent on the ray path SIR. True star
height is given by the straight line SR. Apparent star height is given by the
tangent RF. Astronomical refraction, the angle which should be subtracted from
the apparent star height, is the angle SRF. However, the aurora is situated inside
the atmosphere on the curved ray path SIR at I, not at infinity. When one
subtracts the astronomical refraction from the observed aurora height BRF, the
correction will be greater than necessary. The error is negligible only if point I
was very far.
The phenomenon was observed by observers, in Paris, by Jacques Philippe
Maraldi (Italo-French astronomer and mathematician, 1665–1729) and in Tus-
culum (Rome) by Francesco Bianchini (Italian philosopher and naturalist). The
observers described the phenomenon in the form of an arc open towards the
Earth’s pole. They measured its highest position. It is not likely that Bošković
witnessed the aurora in year 1726 when he was still a teenager, and we may
only speculate about that. The elevation measured in Paris (P) was 37° 20’, and
in Rome (R) it was 20°.
Aurora is situated at A where the light paths of two stars, one at T and an-
other at S, have their crossing. It is clear that point A where the aurora is situ-
ated, is not very distant; it should be nearer than stars since the difference of el-
evations measured in Paris and Rome is substantially larger than the difference
of the observers’ geographical latitudes; in other words, the tangents RF and PF
are not parallel. Using the tabulated values for the astronomical refraction,
Bošković diminishes the elevation measured in Rome by 3’, obtaining 19° 57’.
Certainly, instead of the exact refraction of the light coming from aurora A (angle
FRA) he uses the refraction angle FRS. In this way there appears a new point I
on the intersection of lines SR and PF, line of sight from Paris. (Bošković leaves
refraction of the curved line PA uncorrected, probably because the refraction is
small, about 1’.). Selecting the point I whose height from the Earth |NI| should
be determined, Bošković justifies the task of determining the distance of aurora
borealis which is “less than a true one”: the true height should be the height of the
aurora spot A. Position of I is obviously lower than the position of A.
Figure 3. places the position of observers at R and P on the same meridian,
a procedure borrowed from Mairan (reduction to the same meridian), whose
important book on aurora borealis (Mairan, 1733) Bošković was obviously knew.
The reason for this procedure was the following: ”...since in both cases the auro-
ral phenomenon appeared as a circular arc, he cleverly concluded: the elevation
on which the phenomenon was observed at Tusculum should be observed on the
whole parallel going through Tusculum.”
This is an important intervention to the observational circumstances. The
elevation measured at different parallels could be set at the same meridian only
in the case if the aurora is A circular phenomenon parallel to the equator, and
106 V. VUJNOVIĆ and i. lisaC: De Aurora Boreali (1737) – Contemporary insight ...

Figure 4. Explanation of the aurora model used by Bošković (1738).

Figure 5. Angles in the geometrical analysis of aurora (Bošković, 1738).

centered on the equatorial axis, as the Fig. 4 shows. Point A is therefore at any
point on the auroral circle closest to the observer, either in R or in P.
The difference of the geographic latitudes between Paris and Rome amounts
to 6o 56’. Following the geometrical analysis, as described by Bošković, we pres-
ent the angles in Fig. 5, from where we find the aurora’s height |NI| = 779 r. m.
= 1152 km. The original gives 720 r. m. = 1054 km.
In the investigation described hereinafter, Bošković neglected the atmo-
spheric refraction. The omission of the refraction and its consequences are dis-
cussed in the first and second corollary. As written in the second corollary,
Bošković was aware of observational difficulties and concluded that the correc-
geofizika, vol. 30, no. 2, 2013, 99–118 107

tion of the refraction has a little sense. This is quite reasonable since the maxi-
mum normal refraction of the stellar light amounts to only half degree – that is,
when the rays come parallel to the horizon. The aurora as a phenomenon usu-
ally covers a very large part of the sky and is not clearly outlined.
A question remains, do observers at different locations observe the same
auroral spot? The auroral height is a pretty strong function of the measured
elevation. By substituting the elevation at Paris by 40º, instead of 38º 20’, for the
height of the aurora we obtain 708 r. m. instead of 779 r. m.
The fourth proposition is concerned with the aurora of December 16, 1737.
The task was to find its distance which is „less than a true one“. We do not know
if Bošković observed this phenomenon. However, this 1737 event was observed
in Croatia (Lisac and Marki, 1998, 2003; Penzar and Penzar, 1997). Fr. Nikola
Gojak, annalist in Franciscan monastery in Makarska (note extracted from the
Franciscan monastery annals in Makarska), writes:
„16/10/1737 Around 02 h Makarska – On the clear night and W-ward the air
was illuminated fiery-red; the light was distributed into three parts: towards
Split, towards the Island of Hvar and towards the mountains.“

Figure 6. The onset of auroras treated by Bošković (1738).

At the beginning of the 18th century the aurora was given extreme attention
since the phenomenon had not been observed for sixty years (Maunder’s minimum
of solar activity).
During 1726, as well as in 1737, solar activity was on its rising slope. It
should be noted that the solar cycles to which the auroras of 1726 and 1737
belong were quite prominent and they should have strong impression on
population. The frequency and intensity of auroras depend on the solar activity.
Auroras usually appear almost daily at south and north latitudes above 70°, but
during the maximum of the solar activity, their range may extend towards
middle and lower latitudes; so they can be observed in the whole central Europe,
including Paris and Rome .
In the year 1737, Giovanni Poleni in Padua measured the elevation of 20° of
some prominent aurora details with a quadrant; the quadrant was at that time in
108 V. VUJNOVIĆ and i. lisaC: De Aurora Boreali (1737) – Contemporary insight ...

Figure 7. Explanation of the Bošković’s procedure for calculating properties of aurora observed in
1737.

general use in astronomical observations. According to original Fig. 4, aurora was

seen in zenith at the “ultimate borders of Britain” (the term indistinctly defines
the geographical position). This time Bošković did not place both observations on
the same meridian but used the vertical circle through the “borders of Britain” and
Padua. We therefore sketched a great circle through Padua (P) and Britain (N),
Fig. 7. Such a situation is much better founded than in the previous case since only
 (or its central angle), and the elevation at Padua.
two data are needed, the arc PN
Bošković obtained the arc PN = 18° from geographical tables; the choice depends
on the notion “ultimate borders”. In the triangle CPI all angles are then known
and so is the Earth’s radius as its one side.
Bošković quoted the height of the aurora in 1737 as being 836 r. m. = 1220 km;
our calculation using the sine rule gave the same result.
From the physical point of view, aside from the purely observational indefi-
niteness, it is rather difficult to explain the heights determined in years 1726
and 1737. They can be understood as overestimated values.
The aurora may start at the level of 80–150 km, broaden and grow to 600 km,
and on some occasions to 1100 km. When observing from the Earth’s surface,
observers from higher latitudes notice the most brilliant parts, in the majority
of cases, close to 100 km. However, observations from satellites, or from the In-
ternational space station, reveal the vertical structure of aurora in the form of a
curtain, extending vertically over several hundred kilometers. In their lower
parts colors are bright, green, and in higher parts hazy and red. The colors de-
pend on the molecular atmospheric constitution and on excitation conditions.
The auroral base is visually dominant due to a higher density of the lower
atmosphere; in these layers blue and magenta-red light is emitted by nitrogen
molecules. In the upper auroral parts, formed inside the less dense atmosphere,
prevailes green light of the monoatomic oxygen, and in still higher and more
rarefied parts, prevailes the red light of the monoatomic oxygen. Given oxygen
geofizika, vol. 30, no. 2, 2013, 99–118 109

spectral lines originate at metastable levels having vastly different lifetimes; a

level having longer lifetime deexcites radiatively when atoms are found in me-
dium of lower density, otherwise it is deexcited collisionally.
Due to the Earth’s curvature, observers from central or southern Europe look
at higher parts, red and foggy. This leads to the observation of red paint aurora
seen from moderate latitudes. In Croatia the last auroras were seen around
midnight over several places, in April 2000 and in October 2003, all in diffuse
red colour.

3. Analysis of the fifth proposition

The fifth proposition deals with a very interesting mathematical problem.

Bošković conducted this task following F. C. Mayer, whose treatises on aurora
were published in St. Petersburg 1728 and 1735. Bošković corroborated it only
in principle, giving advice.
The problem consists in determining the height of the auroral oval (circular
aurora) which is centered on the geographical pole and whose plane is orthogonal
to the Earth’s „equatorial axis“ (Fig. 8a). (It is instructive to present original
Bošković’s drawing, Fig. 8b.)

Figure 8a. Observation of polar light in a special case, from only one position.

Figure 8b. Original drawing.

110 V. VUJNOVIĆ and i. lisaC: De Aurora Boreali (1737) – Contemporary insight ...

Figure 9. Situation at the observation position.

Figure 10. Geometry of polar light. The problem is solved when one finds the relation between AB
and BG .

From a single position V, the observer should measure the elevation h of the
closest aurora point and the angular width g of the auroral arc EAD seen above
the horizon (Fig. 9). The situation within the auroral circle is presented in Fig.
10 (Figs. 8, 9 and 10 were derived from the original Fig. 5 in order to explain the
original text in Corollary 1).
Let us enumerate what is known or measured: CV = 4341 r. m., h, the
elevation of the point A, and the amplitude g. One has to determine the distance
AN , the height of the prominent auroral point A above the Earth’s surface. We
followed the original text as a prescription (written in italics).
 , the complement of the geographic latitude,
(A) For a given place V, arc VF
and therefore the angle VCM is also given.
The right-angled triangle VCM is given, and since the Earth’s radius CV is
known, VM is also given.
(B) The right-angled triangle BGM will also be given, since the angle in M is
common with the given right-angled triangle CVM.
Hence angle GBM and angles at tips of triangle ABV will be given.
geofizika, vol. 30, no. 2, 2013, 99–118 111

(C) As observation gives the elevation AVB, triangle VBA will be separately
given. And lastly, as semiamplitude DVB is given, the right-angled triangle DBV
is also given.
(D) Thus, ratios AB  : BV and  BV  : BD , are given, hence the ratio composed
of them, AB  : BD.
On the circle it holds: AB  : BD = BD  : BH , hence the latter ratio, as a ratio
AB  : BH .
(E) Or, if one makes HO  equal to AB  , ratio AB   against the difference BO,
and against semidifference BG will be given.
(F) The ratio BG  : BM is already known, and hence the ratio AB  : BM will
be known, and since the ratio VB  :  BA  was given, it follows that the ratio compo-
sed of them: VB  : BM or VM  : BV will also be given.
(G) Therefore, since VM is given, VB and the second side VA of the triangle
AVB will be given.
(H) Finally, in the triangle CVA with given sides CV , VA , and the angle CVA
between them, one will know CA , and, when the radius CN is subtracted, one
will obtain the sought distance NA . Q.E.F. (Quod Errat Faciendum.)

The mathematical translation of Bošković’s prescription:

(A) VF = 90° − j =  VCM

CV = 4341 r. m.
VM = CV tan (90° − j ) = CV cot j

(B) M =   BMG =  VMC

  =j

GBM = 90° − j =  VBA

  = a ,   VAB = b = 180° − a − h,  

(C) a = 90° − j

b = 180° − a − h = 90° + j − h , h is the measured elevation

g
DBV = is the measured semiamplitude.
2
g
VBD = 90°,    VBD = 90°  −
2

AB sin h BV g AB sin h g
(D) = ,    = cot ,    = cot = tan d   
BV sin b   BD 2 BD sin b   2
112 V. VUJNOVIĆ and i. lisaC: De Aurora Boreali (1737) – Contemporary insight ...

We have added angle d which was not mentioned by Bošković. Furthermore:

AB BD
=
BD BH

2
AB BD  2
=  = tan d
BH  BH 

(E) BO = 2BG ,   OH = AB

AB AB 1
= =
BH 2BG + AB BG
2 +1
AB

AB
(F) BG = BM sin j ,    tand 2
BH
1 BM cot 2d − 1
 tan2d =  ,    =
BM AB 2sin j
2 sin j + 1
AB
AB sin h sin h
=  ,     AB = VB  
VB sin b sin b

BM sin h cot 2d − 1
=
VB sin b 2sin j

(G) VB + BM = VM = CV  cot j   
sin h cot 2d − 1
VB = CV cot j − BM = CV cot j − VB
sin b 2sin j

CV cot j
VB =
sin h cot 2d − 1
1+
sin b 2sin j

cos j
VA = VB
sin b

(H) The distance CA is obtained by using the cosine rule for the triangle
CVA, knowing sides CV and VA.
geofizika, vol. 30, no. 2, 2013, 99–118 113

Let us substitute all from the above into the cosine law:

2 2 2
CA = CV + VA − 2CV  VA cos (90° + h )          cos (90° + h ) = − sin h  
2 2
CA  VA  VA
2
=1+   + 2sin h
CV  CV  CV
−2 −1 
 cos2j   cos2j  
2
CA  cot 2d − 1   cot 2d − 1  
=1+     sin b + sin h   + 2sin h   sin b + sin h  
CV
2
 sin j   sin j   2sin j  
 2sin j  
 
1
 CA 2  2
CA = CV   ,   AN = CA − 4341 r. m.
 CV 2 

g
 is the measured semiamplitude
2
h  is the measured elevation
b = 90° + j − h
sin b g
cot d = tan
sin h 2

An example. By taking j = 56.5°, g = 58°, h = 21.5°, we get the height AN

of 865 r. m. or 1280 km.
What is the precision of an observation from only one position on the Earth’s
globe? The exact expression of relative errors, depending on the error of the
measured elevation and of the arc amplitude, would be cumbersome. Therefore,
by using the Excel spreadsheet we used the range of elevations between 10° and
25° for two amplitudes, g = 58° and 90°, and j = 56.5° for the latitude. The
evaluated height is approximately linearly dependent on the elevation h. For
amplitude g = 58°, the obtained heights AN are above 1000 km, and the error
of 1° in h generates an error of about 50 km. For amplitude g = 90° the heights
are in the range of 500 km, and the error of 1° generates an error of about 30 km.
This proves that the method is quite sound and can be used, only if the aurora
has been modeled according to the given assumptions. In reality, the auroras
inside the auroral oval are centered on the magnetic pole instead on the
geographical pole. Therefore the geographical latitude should be replaced by the
geomagnetic latitude. So much about the measurements from the Earth’s surface.
One can find another expression in order to solve the crucial step in the
calculation, e.g. to find ratio VB / BM . Therefore we have introduced an auxiliary
angle e (Fig. 10), as presented in the Appendix.
In the first corollary, the method is proposed of aurora observation from
two positions. From both positions the direction to the Earth’s pole should be
known. The second corollary mentions observations of the phenomenon from
114 V. VUJNOVIĆ and i. lisaC: De Aurora Boreali (1737) – Contemporary insight ...

many positions, since Bošković was, obviously, aware of spatially and temporally
inhomogeneous aurora pattern. The best way were, as he suggested, precise
measurements of elevations from two positions in the same “vertical” – what
means the vertical plane passing through the observation positions.

4. Physical cause of the aurora

The physical cause of the aurora is the subject of the sixth proposition,
and Bošković discusses the viewpoints which existed at the time. According to
one viewpoint, the polar light causes solar rays, refracted in increased density
(an explanation for such a claim, is missing in the original text) of the atmosphere
around the pole. Bošković opposes this opinion since reflecting body is not seen.
Another viewpoint seeks the polar light in the so-called atmospheric combustion
connected with evaporated water in the atmosphere. The difficulty in this opinion
is the height to which the evaporations, when ignited, rise to be seen from lower
geographical latitudes.
Bošković adheres to the third theory, the theory of Mairan which explains
the auroral phenomenon by invoking an extension of the solar atmosphere. The
solar rotation, documented by the regular motion of sunspots over the solar disk,
deforms the solar atmosphere which therefore may obtain a convex shape and
extend far into the ecliptic plane. In his words, the solar substance flows to the
Earth, into its atmosphere, and even further from the Earth. This is, in Bošković’s
opinion, proved by the appearance of the zodiacal light; it is found in the sky
opposite to the sun. The conical shape of the zodiacal light is pointing along the
ecliptic plane and this corresponds to the convex shape of the rotating solar
atmosphere. The another effect he envisaged, is the occasional excitation of polar
lights. “The refined solar substance”, Bošković writes, is a light source. It enters
into the high atmosphere and flows towards polar areas, heating and igniting
the Earth’s atmosphere. He does not explain the reason of attracting the solar
substance (light) above the polar areas.
We may compare this scenario with the modern scientific picture about the
solar wind – a flow of rarefied and extremely hot gas composed of atomic particles,
ions and electrons, which penetrates through the geomagnetic field, forces the
magnetosphere to react and direct very rarefied content of the magnetosphere
together with the solar wind particles into the auroral ionosphere. So the very
strong electric current is formed which excites the atmospheric constituents to
emit light.
Bošković’s convinction of solar-terrestrial interaction and his results
regarding auroral structure and dimension was acknowledged by contemporaries.
De Mairan himself comments his findings (Marković, 1950): „He understood very
well that property of my view of the aurora borealis which is the most probable.
Even more, he substantiated my view and proved it in a way honourably to him
... Bošković enlarges value of my hypothesis with conclusions derived from the
geofizika, vol. 30, no. 2, 2013, 99–118 115

phenomenon seen in 1737, and particulary with his calculations applied to the
distance to which the matter of the phenomenon was far from the Earth.“
Somehow vague, but sufficiently general visions about the genesis of the
phenomena like the zodiacal and polar light, have close resemblance with the
modern view. The source of both phenomena is found in the sun. However, the
truth is closer in the case of a glowing aurora than in the case of a feeble zodiacal
light. The zodiacal light is seen because the solar light is scattered by tiny solid
particles, grains, in the space where most planetary bodies are moving – around
the plane of ecliptic. The polar light is more specifically caused and induced by
the genuine solar medium. The interaction between the solar wind and the
Earth’s high, geomagnetically active atmosphere, manifests itself in dynamics,
shape details and colors of the polar light.
The solar activity which was fully renewed after Maunder’s minimum and
during Bošković’s life, three hundred years ago, shows ambiguously less activity
during the beginning of the present 24th solar cycle period 2010–2012, Fig. 11.
Since the solar cycle in 1755 has the number 1, the cycles of the auroras in 1726
and in 1737 appeared immediately before it.

Figure 11. End of 23rd and part of the present 24th solar cycle period.

In the near future there is a probability that we shall see the polar light from
our middle latitudes, more or less expressively. The phenomenon should be
expected during the spring and autumn equinoxes, as supported by the
observational statistic and explained by the convenient orientation of active
regions on the Sun towards the Earth.

Ackowledgements – We would like to thank Dr. I. Martinović for his manuscript of the
translation from Latin to Croatian, and for fruitful discussions. We are also indebted to the
State Meteorological and Hydrological Service, Zagreb, for the financial help in the translation
of Bošković’s treatise.
116 V. VUJNOVIĆ and i. lisaC: De Aurora Boreali (1737) – Contemporary insight ...

References
Bošković, R. J. (1738): De Aurore Borealis Dissertatio, Rome, 1738, 1–12.
(„De Aurora Boreali, Dissertatio habita in Seminario Romano ab Augustino Fanucci Academico redi-
vivo, Jo. Bapt. Amalfitano Equite Hier. Ac. Rediv. …“, 12 pages, 8 figures, in Latin, published twice:
in August 1738 printed by Komarek and in September 1738 by Antonio de Rubeis).
Hedin, A. E. (1991): Extension of the MSIS thermospheric model into the middle and lower atmosphere,
J. Geophys. Res., 96, 1159–1172.
Jakobović, Z. (1981): Leksikon mjernih jedinica (Lexicon on measure units). Školska knjiga, Zagreb,
120 pp, in Croatian.
Kutleša, S. (2011): Ruđer Josip Bošković. Technical Museum, Zagreb, 174 pp, in Croatian and English.
Lisac, I. and Marki, A. (2003): Report on polar light, Croat. Meteorol. J., 38, 103–111, in Croatian.
Lisac, I. and Marki, A. (1998): The auroral events observed from Croatia and a part of surrrounding
countries, Geofizika, 15, 53–68.
Mairan, D. de (1733): Traité physique et historique de l’aurore boréale. Suite de Mémoires de l’ Acadé-
mie Royale de Sciences, Paris.
Marković, Ž. (1950): Ruđe Bošković, Almanach Bošković, 137–191.
Marković, Ž. (1968–69): Ruđe Bošković. JAZU, Zagreb, 1144 pp.
Martinović, I. (2012): Ruđer Josip Bošković, On the Aurora Borealis, Rome, 1738. (in Croatian,
translation from Latin, part of a Sci. project sponsored by State Meteorological and Hydrological
Service, Zagreb), private communication.
Penzar, B. and Penzar, I. (1997): Notes from the 17th and 18th century about the weather in Makarska,
Croatian Met. J., 32, 59–68, in Croatian.
Udias, A. (2003): Searching the heavens and the Earth: The history of Jesuit observatories. Kluwer
Academic Publishers, 373 pp, ISBN: 978-1-4020-1189-4.
Whyte, L. L., ed. (1961): Roger Joseph Boscovich SJ FRS, 1711–1787: Studies of his life and work on
the 250th anniversary of his birth. George Allen and Unwin, London.

SAŽETAK

De Aurora Boreali (1737) – Suvremeni uvid u raspravu

mladoga Boškovića
Vladis Vujnović i Inga Lisac

Analiza Boškovićeve rane rasprave o polarnoj svjetlosti pruža uvid u snagu

matematičkog postupka koji je mladi Bošković primijenio na istaknute geofizičke pojave,
s namjerom da shvati njihovu prirodu. Na temelju danih podataka provjerili smo
numeričke rezultate o visini atmosfere i visini polarnih svjetlosti koje su u Europi viđene
1726. i 1737. godine. U njegovu tekstu nema niti jedne ispisane matematičke formule.
Slijedeći pak njegove upute, izveli smo formulu koja se može koristiti da bi se visina po-
larne svjetlosti odredila promatranjem s jednog motrišta. Primjenjivost metode, uz pret-
postavljeni model polarne svjetlosti, prikladna je. Nalazimo da izložena ideja o fizičkom
uzroku polarne svjetlosti, koji se traži u zamišljenim emanacijama Sunca, sadrži elemente
suvremenih shvaćanja.

Ključne riječi: Ruđer Bošković, polarna svjetlost, razmjeri i gustoća atmosfere

Authors’ corresponding e-mail addresses: [email protected], [email protected]

geofizika, vol. 30, no. 2, 2013, 99–118 117

Appendix

Instead of closely following Bošković’s prescription regarding the relations within

the auroral circle, we can find ratio AB / BG by introducing an auxiliary angle
e (Fig. 10):
(E’) e = 90° − 2 d

(
BG = GD sin e = AG sin e = AB + BG sin e )
since GD and AG are circle radii.

AB 1 − sin e
=
BG sin e

BG AB AB BG 1 − sin e
= sin j ,    =   =  sin j
BM BM BG BM sin e

(F’) ...and since the ratio VB  : BA, was given, the ratio composed of them: VB  : BM
or VM  : BV follows

VB VB AB sin g  1 − sin e
= =    sin j
BM AB BM sin h sin e

VB + BM = VM = CV tan (90° − j ) = CV  cot j

(G’) Therefore, since VM is given, VB and the second side VA of the triangle
AVB will be given:
−1
  sin b  1 − sin e −1
 
VB = CV  cot  j  1 +     sin j  
  sin h sin e  


−1
  sin b  1 − sin e −1
cos j cos j  
VA = VB = CV    cot  j  1 +     sin j   .
sin b sin b   sin h sin e  

Comparing the expression for VB / BM in (F) and (F’), we find an identity

2sin e
cot 2d − 1 = ,   e = 90° − 2d .
1 − sin e
118 V. VUJNOVIĆ and i. lisaC: De Aurora Boreali (1737) – Contemporary insight ...

(H’) Finally, in the triangle CVA with given sides CV , VA , and angle between
them CVA, one will know CA , and, when the radius CN is subtracted, one will
obtain sought distance NA :

CA = CV + VA − 2CV VA cos (90° + h ) , cos (90° + h ) = − sin h,

2 2 2

CA = AN + CN .

To sum up:

2 2
 AN + CN   cos2j −1   cos2j −1 

 CN


= 1 +  
 sin j sin b   
 
+ 2
 
(
 sin j sin b    sin h,     CV = CN )
−1
  sin b  1 − sin e −1
−1  
where     = 1 +     sin j  
  sin h sin e  

g = 90° + j − h
 sin h    g 
e = 90° − 2d = 90° − 2arctan  cot  .
 sin b 2