Weather and Weather-Related Environmental Phenomena Including Natural Hazards in Medieval Hungary I: Documentary Evidence on the 11th and 12th Centuries

5
Weather and Weather-Related Environmental Phenomena
Including Natural Hazards in Medieval Hungary I:
Documentary Evidence on the 11th and 12th Centuries
Andrea Kiss
Introduction: aims and sources – the spatial and temporal distributions
The present paper is the first part of a series, providing evidence on high
and late medieval weather events and related phenomena reported in and/or
referring to medieval Hungary. When available, other Carpathian Basin-related
weather events (like those in Slavonia and northern Bosnia) are also presented.
Thus, an important aim of these papers is to provide information, analysis and
an easy access to the presently available evidence on the subject. Moreover, the
series is also designed to act as a background to later, long-term analysis papers.
After providing a concise, systematic database of floods that occurred in
medieval Hungary1 in the form of case studies on individual events reported,
this is the first attempt to give an overview of weather-related source evidence
related to the medieval Carpathian Basin. Since no systematic investigations
have been carried out in the area so far, these papers are meant to provide useful
information for medievalists in general, and those studying military, social,
economic, environmental, or climate history in particular.2
As part of this series on medieval weather-related documentary evidence,
in the first paper the weather events and other weather-related phenomena (e.g.,
fires, floods, locust invasions, famines) concerning 11th- and 12th-century Hungary
are described. For a better insight into large-scale processes, where possible,
some parallels concerning the neighbouring areas are also provided.
1 A. Kiss, “Floods and long-term water-level changes in medieval Hungary,” Diss. (Budapest:
Central European University, 2011).
2 When collecting Hungary-related high-medieval evidence (especially those in foreign
sources), as a first step, I applied (and examined the originals of) the references available in
the following collection: A. F. Gombos, Catalogus fontium historiae Hungariae aevo
ducum et regum ex stirpe Arpad descendentium ab anno Christi DCCC usque ad annum
MCCXI, 4 volumes (Budapest: Akadémiai Kiadó, 1937-1943).
6
For the 11th and 12th centuries only sporadic reports are available, mainly
in foreign narratives; only in exceptional cases do domestic sources provide any
information. This is due to the fact that very few domestic documents have been
handed down to us. Out of these sources the information preserved in the 14th7
century chronicle composition, in which some high medieval domestic
narratives were also identified, probably contains the most useful data.
As regards the temporal distribution, at present almost the same amount of
evidence is available for the 11th century as for the 12th century. Due to some
better-documented military campaigns and battles, more is known about the
years 1044, 1074 and 1147. In western evidence, often the entire country is
mentioned, even if sometimes only the western parts may have been meant.
When specific locations and/or regions inside of Hungary are reported in the
texts, the spatial distribution of known events is relatively even between the
western, and the central and southern parts of the country (see fig. 1), whereas
practically no evidence is yet available for the eastern and northeastern parts of
the Carpathian Basin (e.g., Transylvania and most of the Great Plain).
In contrast to the late medieval period, most of the evidence for the 11th
and 12th centuries is available in foreign narratives. The references in source
evidence have an interesting (but logical) distribution: for the entire country as
well as the western parts, the authors of western narratives provide more data –
the Lower Danube area being ‘covered’ specifically by Byzantine sources3 –,
while evidence related to the central parts chiefly appear in domestic sources (in
the 14th-century chronicle composition). Many of the sources that contain
information are not contemporary. This is mainly due to the fact that almost no
contemporary domestic evidence has remained from this period. Even basic
political history data (concerning the general history of Hungary) can often only
be found in foreign and non-contemporary sources (whose datings were,
therefore, closely examined and potential misdatings extensively discussed in
the literature).
Uncertain fragments or lucky survivors? The role of 11th-century evidence
1022 (1020, 1021): two suns (and fires)
In the year 1022, besides a fire and an earthquake, both the Chronicon
Monacense4 and the 14th-century chronicle composition (Chronici Hungarici
compositio saeculi XIV)5 mention that two suns were seen in the sky (halo-
3 In the present analysis I could use some contemporary Byzantine source references. These
sources were not included in the catalogue and analysis of Ioannis Telelis [“Climatic
Fluctuations in the Eastern Mediterranean and the Middle East AD 300-1500,” Jahrbuch
der Österreichischen Byzantinistik, 58 (2008): 167-207]. However, due to their great value
and broad use in reconstructing high medieval Hungarian military history, I decided to
include these references.
4 I. Szentpéntery, Scriptores Rerum Hungaricarum, 2 vols., new ed. (Budapest: Nap Kiadó,
1999) (hereafter: SRH), vol. 2, 67 (Chronicon Monacense): 31. Anno Domini MXXII. In
multis locis grandia incendia facta sunt et terre motus.
5 SRH vol. 1, 318 (Chronici Hungarici compositio saeculi XIV): (S) Incidentia de aliquibus /
(V) Incidentia: Anno Domini M° XX° II° in multis locis incendia multa et magna facta sunt,
ingens etiam terre motus contigit. IIII° Idus Maii decima hora diei, sexta feria post
Ascensionem Domini quasi duo soles visi sunt decimo Kalendas Iulii.
8
phenomenon). The 14th-century chronicle composition, referring to the 11th and
12th centuries, is mainly built on contemporary evidence. The question of how
close in time it was as well as the application of foreign narratives raise several
further questions. In many cases, misdatings are possible. In our present case,
for example, the 14th-century chronicle composition, besides an early (later
vanished) domestic gesta dated to 1060, presumably applied the evidence
available in the Annales Altahenses maiores (or both narratives used the same
source). Subsequent parts of the chronicle composition are frequently built on
legends6 or a later-vanished early 13th-century Hungarian chronicle (dated to the
1210s) that also seems to be partly based on earlier works.7
The information concerning 1022 similarly appears in the Annales
Altahenses maiores where the event is, however, mentioned for the year 1021.8
Also, the Annales Altahenses, for the year 1020, refer to fires in many places
(without specifying any location),9 which appears in the 11th-century Hungarian
chronicle composition for the year 1021.10 The same data can be found in the
Chronicon Dubnicense (15th century, compiling earlier narratives), for the year
1022.11 As a possible parallel, it is worth mentioning that R. Glaser especially
highlights the hot summer of 1022, as part of a warm and dry phase.12
Rainy or not? German military campaign in 1043 –
and a comparison with the 1044 events
Concerning the year 1043, Herimannus Augiensis mentions a rainy
summer when the German ruler came to Hungary to fight against the Hungarian
King Aba. It was in fact so wet in the German territories that there was shortage
of food because of a bad harvest.13 At that time, in August 1043, Emperor Henry
6 SRH vol. 1, 312-44.
7 Gy. Kristó, A történeti irodalom Magyarországon a kezdetektől 1241-ig (Historical literature
in Hungary from the beginning to 1241) (Budapest: Argumentum 1994), 99 (hereafter:
Kristó 1994).
8 E. Oefele, Scriptores rerum Germanicarum in usum scholarum. Annales Altahenses Maiores,
vol. 4 (Hannover: Impensis Bibliopolii Hahniani, 1891) (hereafter: Oefele 1891), 17
(Annales Altahenses Maiores): 1021. Ingens terrae motus IIII. Idus Mai, hora X. diei, feria
sexta post ascensionis Domini, quasi duo soles visi X. Kal. Iulii. See also Kristó 1994, 100.
9 Oefele 1891, 17 (Annales Altahenses maiores): 1020. Benedictus VIII. papa, ab imperatore
Babinperc vocatus, advenit. In multis terrarum locis multa et magna incendia.
10 SRH vol. 1, 318 (Chronici Hungarici compositio saeculi XIV): III° Idus Maii decima hora
diei sexta feria post Ascensionem Domini quasi duo soles visi sunt decimo Kalendas Iulii.
11 M. Florianus, Chronicon Dubnicense, Historiae Hungaricae Fontes Domestici. Pars prima:
Scriptores, vol. 3 (Leipzig, 1884), 47: 66. (22a) Anno domini millesimo vigesimo secundo
in multis locis incendia multe et magna facta sunt, ingens eciam terre motus contigit
quarto Idus May decima a hora diei feria sexta post ascensionem domini; quasi duo soles
visi sunt decimo kalendas July.
12 R. Glaser, Klimageschichte Mitteleuropas: 1200 Jahre Wetter, Klima, Katastrophen
(Darmstadt: Primus Verlag, 2008), 61.
13 G. H. Pertz, Monumenta Germaniae Historica. Scriptores, vol. 5: Annales, chronica et
historiae aevi Salici (Hannover: Impensis Bibliopolii Hahniani, 1844) (hereafter MGH SS
9
III entered Hungary, but the swamps and inundations of the Rábca River
frustrated his progress in Northern Transdanubia (Western Hungary; see fig.
1).14 From this description, although the German troops could not cross the
swamps and the waters of the Rábca river, it does not necessarily follow that
there was a rainy summer season in Hungary in the period before or during the
military campaigns.
When Henry III entered Hungary in 1044 (perhaps in June), before
defeating King Aba at Ménfő (on 5 July), he also had to cross the Rábca River.
Similar to 1043, it was again a rather difficult task due to the swampy areas
(invenit iter stagnantibus aquis).15 This is supported by the chronicle of Henrik
Mügeln, who worked with some early sources.16 However, people of the ‘other’
Hungarian king (Peter I) showed Henry a ford where it was easy to cross the
river. No more difficulties with crossing were mentioned this time, only that
Aba’s army was waiting for them on the other side of the river.17
There is a chance that, similar to West Central Europe (with Bohemia),
some parts of the summer of 1043 were also wet in the more easterly parts of
Central Europe. Nevertheless, no such information is available from other parts
of Central Europe for (the early summer of) 1044. Yet, as was reported, the
emperor’s army was again (at least for a while) obstructed by extensive swampy
areas in Hungary. This is due to the fact that the Hanság (Wasen) wetlands are
located south of them. The waters of the Hanság were collected and led to the
vol. 5), 124 (Herimannus Augiensis Chronicon): 1043. … Aestas pluviosa frugum et
vindemiarum penuriam effecit. – Heinricus rex iterum Pannonias petens, ab Ovone, vis
impetrante pactum, satisfactionem, obsides, munera, regnique usque ad Litaha flumen
parte accipiens, discessit.
14 Oefele 1891, 33: 1043 … Veniens ergo rex ad terminum regni, ubi exercitus recenseretur,
decrevit, ut hostilis terra procul dubio invaderetur, et factis machinis contra opus, quo
fluvium Rapinisa occluserant, in crastinum impugnare disposuerant.
15 Oefele 1891, 35.
16 SRH vol. 2, 158 (Chronicon Henrici de Mügeln): Do het sich daz wasser Raba genant und
Rapcha genant ergossen, daz der keyser nicht uber daz wasser moch kumen, wann die
weld warn dick ind die pruch warn tieff. Do czoh der keyser das wasser auf die ganczen
nacht und kom dez morgens fru, do dye sunne schein, an ein furt und czoh laytlich uber dy
Rab.
17 Oefele 1891, 35-36 (Annales Altahenses maiores): 1044. … Quo die cum noster heros ad
locum condictum venisset et illum minime reperisset, volens eum insequi ultra Rabanisa
fluvium, invenit iter stagnantibus aquis et machinis more illis solito interclusum. Ducatum
autem praebentibus Ungris, quos secum habuit, tota nocte equitato sursum per ripam
crepusculo facili vado transit. …. Tercia demum die post condictum praelium, cum primi
nostrorum transirent Rhaba fluvium, ecce innumerae acies armatorum eminus apparent,
qui campum latissimum, quasi silva succrevisset, operuerunt. Besides the Annales
Altahenses maiores, the 13th-century domestic Gesta Hungarorum, written by Simon de
Keza, described the situation in a rather similar way (SRH vol. 1, 176): Concitato igitur
cursu invasit fines Hungarie intrans per Supronium et, cum vellet in Bobut pertransire,
non poterat propter aquas. Hungari ergo, qui erant cum ceasare et Petro rege, duxerunt
caesaris exercitum sursum iucta flumen Rebche et utraque flumina tota nocte equitando
orto sole facili vado transierunt.
10
Danube by the short lower section of the Rábca river. Due to the small
difference in elevation, it was rather easy to ‘manipulate’ the hydrological
conditions of the lower Rábca, and thus it was possible to make crossing
difficult or impossible for those not familiar with the environmental conditions
of the area. As such, a natural flood event was not necessarily needed to ‘produce’
flooding.
Is there any information available to suggest that Hungarians influenced
the water system or applied any sort of ‘manipulation’ of this fragile water
system? Already in the case of the 1043 campaign the Annales Altahenses
maiores mentions a certain opus, an artificial object on the Rábca river (et factis
machinis contra opus, quo fluvium Rapinisa occluserant),18 which indirectly
reappears in 1044.19 In the 14th-century chronicle composition the even more
direct word obstacula is mentioned.20 Thus, presumably the same sort of
machinae appears in the Annales Altahenses maiores, and in both cases artificial
intervention in the hydrological regime produced stagnant bodies of floodwater
in 1043 and 1044. While, however, in 1043 the emperor’s army could not cross
the river due to the extensive wetlands, in 1044 they found an easy crossing
place with local help.21
Concerning the neighbouring areas, there is a famine dated to 1043 in the
Czech Lands (by Cosmas Pragensis). Although some later authors suggested
that the famine was caused by the occupation of Bohemia, R. Brázdil and O.
Kotyza (1995), similar to Glaser, drew our attention to the fact that there was a
wet and cold spell in the summer at Lake Constance (e.g., Reichenau) and St.
Gallen.22 As we will see, for 1044 a different picture may be gathered from
18 Oefele 1891, 33 (Annales Altahenses maiores).
19 Oefele 1891, 35-36 (Annales Altahenses maiores): Quo die cum noster heros ad locum
condictum venisset et illum minime reperisset, volens eum insequi ultra Rabanisa fluvium,
invenit iter stagnantibus aquis et machinis more illis solito interclusum. Ducatum autem
preabentibus Ungris, quos secum habuit, tota nocte equitando sursu per ripam crepusculo
facili vado transit. Quorum adventum prospectantes, qui machinas custodiebant, relictis
omnibus, quae habuerant fugerunt.
20 SRH vol. 1, 329.
21 SRH vol. 1, 331 (Chronici Hungarici compositio saeculi XIV): (S) Concitato itaque cursu
invasit fines Hungarie intrans per Supronium et cum vellet per Bobuth Rabcha pertransire,
fluvium Rabcha stagnantibus aquis et densissimis nemoribus et scaturinetibus paludibus
impermeabilem transire non potuit. Hungary ergo qui erant cum cesare et Petro rege,
duxerunt exercitum tota nocte equitantes sursum iuxta fluvios Raba et Rabcha, quos
illucenscente sole facili vado transierunt.
22 R. Brázdil and O. Kotyza, History of Weather and Climate in the Czech Lands I (Period
1000-1500), Zürcher Geographische Schriften 62 (Zurich: ETH, 1995), 92, with reference
to P. Alexandre, Le climat en Europe au Moyen Age. Contribution à l’histoire des
variations climatiques de 1000 à 1425, d’après les sources narratives de l’Europe
occidentale (Paris: Ecole des hautes études en sciences sociales, 1987), 326 and 338, using
the chronicle of Herimannus Augiensis and the Annales Sangallenses maiores. See also
Glaser, Klimageschichte, 61.
11
contemporary documents: more of a dry than wet character of the early and midsummer
times is described.
Dust veil on 5 July, 1044:23 The battle of Ménfő
Referring to the battle of Ménfő in 1044, both the Annales Altachenses
Maiores24 and the 14th-century chronicle composition mention that during the
battle first a cloud arose above Aba’s army and then a sudden dust storm, which
disconcerted the army and was largely responsible for the victory of King Peter
I’s troops.25 This event was as well mentioned by Herimannus Augiensis.26
Further circumstances and background conditions of the battle were studied in
detail by L. Négyesi.27 Based on the latter’s reconstruction of the geographical
circumstances of the battle, the cloud appeared roughly towards the east, since
Aba’s army occupied this position (north-northeast from the medieval village of
Ménfő). The event happened in the middle of summer, so hot and dry weather at
that time was not unusual. Since the wind storm seemed to come from an
23 Calculating it for the Gregorian Calendar (hereafter: GC): 12 July.
24 Oefele 1891, 37 (Annales Altahenses Maiores): 1044. … Igitur eo die, cum vellent confligere,
ut supra diximus, uterque populus, tantum progressus, ut facilis invicem esset
prospectus, tenuis nubecula, id est caeleste signum, adparuit, quos apostolicus, successor
beati Petri, eo quod regem suum dehonestarant, iam pridem anathemizarat. Et ecce turbo
vehemens ex parte nostratium ortus pulverem nimium adversariorum ingessit obtutibus.
Hoc sedato, nostris Dominum invocantibus et illis econtra incondite vociferantibus, pugna
nimis atrox inchoatur et summa vi utrimque decertatur.
25 SRH vol. 1, p. 332 (Chronici Hungarici compositio saeculi XIV): (S) …. Tradunt autem
Teutonici, quod cum ad prelium appropinquassent, cum celesti signo nebula tenuis
apparuit, turboque vehemens divinitus incitatus terribiliem pulverem obtutibus ingessit
Vngarorum, quos iam pridem, ut dicitur, quia Petrum suum regem dehonestaverant,
dominus apostolicus anathemate dampna verat feriendo.
26 MGH SS vol. 5, 124-25 (Herimannus Augiensis Chronicon): 1044. Maxima pestis pecudum,
et hiems satis dura et nivosa magnam vinearum partem frigore perdidit, et frugum
sterilitas famem non modicam effecit. …. – Ovo rex cum iusiurandum pactumque
infregisset, Heinricus rex cum perparvis copiis Pannonias petiit. Quam Ovo, qui maximum
paraverat exercitum, in tantum despexit, ut eum, quasi facile occidendum seu capiendum,
provinciam ingredi sineret. Sed Heinricus cum parte copiarum Rabam fluvium festine, in
divino confisus auxilio, transvadavit, pugnam cunctis militibus passim festinantibus iniit, et
in prima congessione innumerabilem Ungariorum exercitum fugavit atque prostravit, et
paucissimis suorum amissis, ipsemet fortissime pugnans, gloriosissimam victoriam adeptus
est 3. Non. Iul.
27 L. Négyesi, “A ménfői csata” (The battle of Ménfő), Hadtörténelmi Közlemények, 107/3
(1994): 136-46. It seems likely that, as Négyesi also suggested in his article, the windstorm
prevented Aba’s traditional Hungarian archers from using their arches which provided
favourable circumstances for the western type of ‘close-fighting’ tactics. Négyesi’s paper
is one of the nice examples in Hungarian research, in which – discussing the circumstances
of the battle in great detail – the author also kept and eye on and provided a concise and
systematic analysis of the geographical/environmental background, with a certain emphasis
on weather conditions. When talking about the actual weather conditions of the battle,
Négyesi even applied the results of scientific meteorological literature (p. 141).
12
easterly direction and there was great dust, dry (and warm/hot) conditions might
have prevailed earlier for a time.
According to the Annales Altahenses maiores, after King Aba and his
army had lost the battle of Ménfő, he ran away to the east where he was struck
by lightning in a village,28 while Herimannus Augiensis claimed that after the
battle Aba was killed by his own people.29 Similar to the domestic 13th-century
Gesta Hungarorum of Simon Kézai,30 Henrik Mügeln (14th century, but applying
Hungarian narratives) suggested that Aba died from serious injuries.31
An important addition can be found in the (late-11th-century) Chronicon
Bernoldi: Right after the (otherwise copied) sentence about the Hungarian
campaign of Henry III, it is mentioned (without providing any exact location)
that a great famine occurred (Magna vis famis homines immunda animalia
comedere coedit).32 Even if no further details are available concerning the
location of this famine, it may have some connection with the famine in the time
of King Peter (who became king once again right after Aba had died) when
bones of St. Coloman were taken as a gift for him to Hungary (see the next
case). Also, as said before, famine was mentioned for Bohemia in 1043.
Drought and famine in 1044?
An early-11th-century famine might be documented in the Chronicon
Austriae Mellicense breve: When the bones of blessed Coloman were taken to
Hungary, disease and famine occurred there. The dating of this possible event is
rather uncertain. Even if the only date mentioned around in the text was 1012
(1018?), it can be more probably be dated, as the same chronicle also suggests,
to the reign of Peter I.33 Until 1038 King Stephen ruled in Hungary (1001-1038),
28 Oefele 1891, 37 (Annales Altahenses maiores): 1044. … Caeterum Obo fugiens in quendam
vicum pervenit, villaque fulmine succensa ipse pene interiit, et vix elapsus fugitando
latitabat, usque dum comprehensus communi iudicio nostrorum et suorum capitalem subit
penam.
29 MGH SS vol. 5, 125 (Herimannus Augiensis Chonicon): 1044. … Et Ovone rege vix fuga
elapso, omnes Ungarii ad dedicionem Heinrico regi catervatim concurrunt, subiectionem
serviciumque promittunt.
30 SRH vol. 1, 176 (Simonis de Keza Gesta Hungarorum): Aba vero rex fugiit versus Tizam et
in villa quadam in scrobe veteri ab Hungaris, quibus regnans nocuerat, iugulatur et iuxta
quandam ecclesiam sepelitur.
31 SRH vol. 2, 261 (Chronicon rhythmicum Henrici de Mügeln): Post annos vero plurimos /
Effoditur per Hungaros, / Qui eum invenerunt / Sanatum a vulneribus. / Quem ceteris
funeribus / In Sarus posuerunt.
32 MGH SS vol. 5, 425.
33 According to the Annales Mellicenses, Coloman died in 1012, and was buried, then reburied
in 1013 and 1014, but still not in Hungary. See G. H. Pertz, Monumenta Germaniae
Historica. Scriptores, vol. 9: Chronica et annales aevi Salici (Hannover: Impensis
Bibliopolii Hahniani, 1851) (hereafter MGH SS vol. 9), 497 (Annales Mellicenses): 1012.
Beatus Cholomannus martyrizatus est, et suspensus apud Stocchaerouwe. … / 1013. Hoc
anno beatus Cholomannus sepultus est eodem loco. / 1014. Hoc anno sanctus
Cholomannus a Megingaudo Heibstatensis episcopo in Medilccha sepultus. …
13
and Peter became the king only afterwards (1038-1041, 1044-1046).34 The noncontemporary
(15th-century) compilator, Vitus Arnperckius,35 not only
mentioned the same diseases and famine, but also the excessive lack of rain (nec
pluviam habuerint).36
Thus, the most probable dating of these events is the early-mid 1040s,
especially the period of 1044-1046 when King Peter had close ties with Henry
III.37 About the same time, according to Glaser, in 1044, there was famine in the
area of Nuremberg.38 The great famine (which occurred in the Czech Lands) was
dated to 1043 in the Czech sources.39 As Glaser suggests, wet and cool summers
prevailed in the German areas in the 1040s and 1050s; the summer of 1043 was
particularly cool and rainy, and the wine of that summer was sour.40
While mentioning famine generally in Hungary, it is unclear, whether the
western sources really referred to the whole country or only the better-known
34 G. Waitz, Monumenta Germaniae Historica. Scriptores, vol. 24: Annales aevi Suevici
(Hannover: Impensis Bibliopolii Hahniani, 1879) (hereafter MGH SS vol. 24), 71 (Breve
Chronicon Austrie Mellicense): Hic autem H. marchio duravit per annos circiter quadraginta
usque ad tempora Petri regis Ungarorum, qui sancto regi Stephano successerat.
Istius Petri [T. Peter] precibus et minis circumventus Poppo Treverorum archiepiscopus,
frater marchionis, obtinuit a fratre suo, ut mitteret in Ungariam ossa beati Cholomanni;
que tamen postea Ungari pestilentia et fame compulsi reddiderunt. H. vero marchio
decedens ipso anno, filium reliquit Adalbertum, attavum vestrum. (…) Ille cum filio suo
L.II. commissa pugna cum Ovone rege Ungarorum, quo Petro regi superpositus fuert,
maximam partem exercitus eius delevit, victor existens.
35 Lexikon des Mittelaters, vol. 1 (Munich and Zurich: Artemis-&-Winkler-Verlag, 1980), col.
1011.
36 H. Pez, Scriptores Rerum Austriacarum (Leipzig 1721) (hereafter: SRA), 1181 (Vitus
Arnperckius): Sed tempore, quo sacrum corpus [Saint Coloman] in Hungariam permansit,
Hungari multas plagas sustinuerunt, pestilenciam, famem, nec pluviam habuerint.
37 This dating is also clearly suggested in the academic literature: Poppo, the archbishop Trier,
brought with him the bones to Hungary. However, due to the political changes in Hungary
(Peter lost his crown in 1046), the bones soon reappeared in Melk. In spite of this fact, his
herma was still mentioned in 1515 as part of the treasures in the basilica of Székesfehérvár.
See: A. Borosy and J. Laszlovszky, “Magyarország, a Szentföld és a korai keresztes
hadjáratok” (Hungary, the Holy Land and the early Crusades), in J. Laszlovszky, J.
Majorossy and J. Zsengellér (eds.), Magyarország és a keresztes háborúk. Lovagrendek és
emlékeik (Hungary and the Crusades. Military orders and their recollections) (Máriabesnyő
and Gödöllő: Attraktor, 2006) (75-89), 78.
38 Glaser, Klimageschichte, 71.
39 See, e.g., B. Bretholz, Die Chronik der Böhmen des Cosmas von Prag. MGH SS Nova
series, vol. 2 (Berlin: Weidmannsche Buchhandlung, 1923), 100 (Cosmas Pragensis): XIII.
Anno domince incarnationis MXXXXIII. Tanta fames fuit in Boemia, ut tercia pars populi
interiret fame. / Anno dominice incarnationis MXXXXIIII. For a further analysis of Czech
evidence, see Brázdil and Kotyza, History of Weather I, p. 92.
It is interesting that while Cosmas Pragensis mentioned hunger for 1043, he fails to
mention the year 1044. As a parallel, Herimannus Augiensis (MGH SS vol. 5, 124) in his
chronicle mentioned that the summer of 1043 was rainy and a shortage (of grain and wine)
occcured.
40 Glaser, Klimageschichte, 61.
14
western parts. However, despite the controversies, 1044(-1046: Peter’s second
reign) seems to be the most likely date for the drought and famine.
Continuous rainfall: August 1051 – an(other) unlucky campaign of Henry III to
Hungary?
A major German military campaign for Hungary was realized in 1051.
While one part of his army proceeded along the Danube, the emperor himself
rode to the south and invaded Hungary from the ‘Carantan borders’ (per
Carentani fines), from today’s Styria, downstream along the Rába river.
According to Herimannus Augiensis, all this happened in autumn.41
At this time, however, learning from past mistakes, the emperor planned
his attack more carefully, avoiding the ‘fatal triangle’ of the Fertő-Hanság-
Rábca(-Danube) wetland area. Two armies entered Hungary: one (by ship) on
the Danube, and the other, main one downstream along the Rába river, from
Styria. As for the route of their march, the 14th-century chronicle composition
mentioned the Zala and Zelice (Szőcei-patak) Rivers.42 On the Austrian border
these two waterflows lay quite close to each other. As the Annales Altahenses
maiores interprets the events, it was a difficult campaign, largely due to the fact
that the whole summer had abundant rains (Totum namque aestivum tempus,
pluviis abundans, inmanes fecit prorumpere aquas). During this campaign,
many people and horses went beneath the floodwaters (Qua de re plures tam
hominum quam equorum summersi sunt).43 Nevertheless, the emperor took the
long way until he reached Hungary (in mid-August; which in GC is late
August). Thus, the rainy conditions might mainly refer to the conditions of the
German/Austrian areas. In addition, it is interesting to note that, according to the
14th-century chronicle composition, the Hungarians were able to successfully
41 MGH SS vol. 5, 130 (Herimannus Augiensis Chronicon): Sequenti autumno magno cum
exercitu imperator, pctum Andreae regis ut per legatos offerebatur accipere contemnens,
Pannonias petiit, Gebehardoque atisponense episcopo et Welf Brezizlaoque ducibus ad
septentrionalem Danubii partem depopulandam transmissis, ipse, devecto e navibus,
quantum in equis poterat, sumptu, longo propter fluviorum illuvionem circuitu per
Carentani fines transiens, periurumque regnum invadens, hac ac illac Ungarico exercitu
latrocinantium more fugitante et nusquam aequo Marte conferre manum audente, omnia
circumquaque, quoad sumptus militi suppetebant, vasta reddidit.
42 SRH vol. 1, 347 (Chronici Hungarici compositio saeculi XIV): Sequenti vero anno cesar
cum magna multitudine bellatorum ob easdem causas venit in Hungariam iuxta fontes
rivorum Zala et Zelice. The medieval name of the Zelice River is preserved in the name of
the village called Szőce. See G. Kiss and B. Zágorhidi Czigány, “Víznevek névalkotó
szerepéről – Vas megyei példákon” (The role of water bodies in shaping the names of
locations – examples from Vas county) Helynévtörténeti tanulmányok, 6 (2011), 111-119.
43 Oefele 1891, 47 (Annales Altahenses maiores). A rainy August was also mentioned for
Lobbes (today in South-Belgium, close to the French border). See Brázdil and Kotyza
1995, 93; Alexandre 1987, 338.
15
apply scorched-earth tactics against the emperor and his army.44 So, this
information would not by itself imply that rainy conditions prevailed in West-
Hungary during the emperor’s military campaign.
More is available in the description of Aventinus. Apart from the unpleasant
and unusual fighting tactics of the Hungarians, he also blamed the
unfavourable weather conditions for the disastrous result of the campaign stating
that the rivers and swamps of Hungary were extended due to the rainy
conditions.45 In fact, Aventinus refers to the fact that not only the waterflows and
wetland areas were extended, but there were also continuous showers (heavy
rainfall), and the military campaign had to be aborted that year mainly because
of the atrocious weather, bad floods and a shortage of food. Thus, it seems that
rainy conditions were not only typical for Bavaria that summer, but they also
most likely went with the German armies on the western border of Hungary (in
Austria; and perhaps in the Czech Lands too).46
The coronation of King Salamon and the great fire event in Pécs in April 1064
According to the Chronicon Monacense, after the coronation of King
Salamon on 11April, 1064,47 at night the entire palace and the basilica in Pécs
44 SRH vol. 1, 347 (Chronici Hungarici compositio saeculi XIV): Cumque cesar intrasset ad
Hungariam et venisset in regiones combustas, neque militibus suis invenire potuit victualia
nec equis et nesciebat ubi naves eiu nullumque subsidium ex eis poterat habere.
45 N. Cisnerus, Ioannis Aventini Annalium Boiorum (Basel: Impensis Ludouici Regis, 1615),
327: Liber quintus: Imperator petiti coactoque exercitu Ungariam invadere imperioque
subigere tentat. …Vbi illi fines hostium attigere, agrosque vrunt, aedificia, vicos incendunt,
eiusmodi tempestates sunt consecutae, vt expeditio necessario intermitteretur, &
continuatione imbrium, torrentum, fluminumque diluuiae, plerisque equorum & mortalium
absumptis, sub papilionibus milites contineri non possent: itaque cunctis praeda, igne,
ferro attritis, domum reuertuntur. Porro Imperator cum valida Francorum, Gallorum,
Polonorum, Longobardorum, Venedorum, Sueuorum, Saxonum manu, Pannoniam
Australem Vngariae partem, terrestri itinere, absque carrucis, annona equis clitellarijs
aduecta, inuadit: sed paludibus fluminumque eluuie iter sibi interclusum offendit:
longisque anfractibus alluuiones amnium, qui iugi pluuia aucti excreuerant, alueosque
egrediebantur, circuit, per Chaironum fines Vngariam intrat. Hostes in continentes syluas
ac paludes se suaque omnia contulerant, pugnandi aperto marte potestatem nequeaquam
Caesari faciunt, pabulum, aquarum fontes corruperant. ….
46 It is also worth mentioning that in the catchment areas of the Rába and Rábca Rivers,
according to 20th-century observations, the main flood period is usually spring. And there
is a secondary flood maximum on the Rába River (especially on its upper sections, in
Styria) in the June-July period. For the annual course and main hydrological characteristics
of the river (and those of its upper catchment) see, e.g., P. Ambrózy, “A Felső-Rába
vízgyűjtőjének éghajlati jellegzetességei” (Climatic conditions of the upper Rába
catchment) Vízügyi Közlemények, 79/4 (1997): (498-517) 508-509. See also: L. Goda and
V. Vasvári, “A Felső-Rába vízjárásának statisztikai jellemzése” (Statistical analysis of the
upper Rába water course). Vízügyi Közlemények, 79/4 (1997): (518-538) 530.
47 GC: 18 April.
16
were destroyed by a fire.48 The 14th-century chronicle composition (being
probably the main source of the Chronicon Monacense) provided even more
details, but no information is available in the sources concerning the possible
cause of the fire.49 This event is still worth mentioning because, among other
reasons, major fire events are often associated with previous dry (or in winter
with prolonged cold) conditions; and (as one will see in the next case) strong
winds may also exacerbate the effects of a fire.
The weather conditions of 1064 in the Czech (or Polish) lands are not
described; no data is available. For the German areas, Glaser refers to the winter
of 1063 as a hard and snowy one, with a cold March and a cold mid-April, but
also no information is available for 1064.50
Fire in strong eastern winds: autumn 1071
According to the 14th-century chronicle composition, in the third month of
the siege of Belgrade (historical Nándorfehérvár), in autumn, fire caused by the
enemy (Greeks) was fanned by strong eastern winds, and the buildings all over
Belgrade were engulfed by fire.51 According to previous historical
investigations, both the political and military background suggest that the siege
was begun in late August and ended in late November.52 For this year, no
comparative information is available in the literature for the German territories
and the Czech Lands (neither for Poland or for the closer part of the Balkan
area).
48 SRH vol. 2, 75 (Chronicon Monacense): Et de manibus ipsius Geyse rex in Qinqueecclesiis
coronatur, ‘et facta est letitita’ magna ‘in populo’. Nocte sequenti ecclesia, pallatia omnia
cum edificiis annexis repentina flamma corrumpuntur. Rex et dux attoniti ob inopinatum
incendium in bona pace convenerunt et a rege Dalmacie rogati in subsidium eius
transierunt et revertuntur.
49 SRH vol. 1, 363 (Chronici Hungarici compositio saeculi XIV): (S) 98. In eodem autem
anno, ducibus ibidem existentibus ecclesia oribiliter est combusta.
(V) 98. Nocte autem secuta prognosticum future discordie et turbationis provenit. Totam
enim ecclesiam illam et omnia pallacia ceteraque edificia ei apendentia repentia flamma
corripuit et universa castante incendio corruerunt. Accepit autem omnes timor pre nimia
concussione flammarum stridentium et terribili furore campanarum et turribus ruentium;
nex sciebat quisquam quo se divertet.
50 Glaser 2008, 72.
51 SRH vol. 1, 373 (Chronici Hungarici compositio saeculi XIV): (V) 108. In tertio vero
mense obsidionis quedam puella Hungarica olim illuc per captivitatem abducta in quarta
feria civitatem succendit. Flabar autem subsolanus vehementissime, cunctaque edificia
occupaverunt incendia. Exercitus itaque Hungarorum irruit in civitatem ea parte, qua
muri eius per tormenta iam corruerant, et Grecos, Sarracenos atque Bulgaros crudeliter
trucidarunt. … Hungari ergo capta civitate sequenti die postquam ignis resederat, ….
52 For a detailed analysis of Byzantine and Hungarian sources, as well as Hungarian and
international literature of the period and the siege itself, see, e.g.: F. Makk, A turulmadártól
a kettőskeresztig. Tanulmányok a magyarság régebbi történelméről (From the ‚turul‘ to the
double-cross. Studies from the early history of Hungarians) (Szeged: Szegedi Középkorász
Műhely, 1998), 148-52. See also Kristó 1986, pp. 63-64.
17
1074 (1): frozen solid Tisza River before and after 26 February53;
the battle of Kemej
According to the Viennese version of the 14th-century chronicle
composition, shortly before the battle of Kemej on 26 February (quarta feria
post Dominicam, que cantatur Exsurge),54 King Salamon crossed the frozen
Tisza river at Thoroyd/ Cothoyd.55 Then, after the successful battle, the king
again crossed the Tisza on the way back.56 Although the site of the battle is not
precisely known, it was probably in the neighbourhood of Szolnok. The words
glaciatam (before 26 February) and congelatam (after 26 February) are
mentioned only in the (otherwise generally more detailed) Vienna-codex version
of the Chronica Picta, whereas only the fact of crossing was mentioned in the
less-detailed Sambucus (Buda)-codex version, both containing the 14th-century
chronicle composition. Similarly, the ‘frozen-Tisza’ version appears in the
chronicle of Henrik Mügeln, presumably taken from the Vienna chronicle.57
In order to cross in late February with a whole army, it was of course
necessary that the Tisza was still frozen solidly. This means that there was a
February (much) colder than usual in the central parts of Hungary. A cold winter
month in 1074 is also mentioned by Alexandre, referring to Lower-Saxony,
Vestfalia, Frankonia and Hessen.58 According to Glaser, the winter of 1074 was
especially severe in the German areas: the rivers froze “to the ground”.59
53 GC: 5 March.
54 SRH vol. 1, 385. For the chronology of events, see, e.g.: J. Bánlaky, A magyar nemzet hadtörténelme,
vol. 3 (Budapest: Athenaeum, 1930), pp. 119-21.
55 SRH vol. 1, 383 (Chronici Hungarici compositio saeculi XIV): (V) 117. Dux vero cum
exercitu suo obviavit regi in Kemey. Principes autem ducis ymo traditores, miserunt
clamculo nuncios ad regem dicentes, quod sir ex eos in dignitatibus teneret et in gratiam in
bello relicto duce ad regem confluerent. Rex autem certificavit eos super hoc prestito
iuramento et securus tunc transivit Tysciam glaciatam super ducem. As Bánlaky (1930)
suggested (p. 119; also based on the 14th-century chronicle composition), the two armies
probably crossed the Tisza on the one hand somewhere in the area of the present-day
Cibakháza, on the other hand in the area of Egyek (close to the ferry of Dorogma; see fig.
1).
56 SRH vol. 1, p. 386 (Chronici Hungarici compositio saeculi XIV): 119. … Rex [Salamon]
victoria potitus transivit Tysciam congelatam in Thoroyd et descendit Ladizlao et Otthoni
ducibus cum valida multitudine pugnatorum ad se venientium.
57 SRH vol. 2, 183 (Chronicon Henrici de Mügeln): Do der herczog Geysa ersah, daz in die
sein liessen und der streit verlorn was, do wirt er sich dennoch, so er lengst mocht.
Darnach floch er aus dem streit an die Teysa gegen Totoyd (sic). Do er kom gen Woczen,
do begegent ym sein pruder Lasla und Otto, sein swager, mit grosser hilff. Darnach czoh
der kunig mit sige uber die Teysa, die was gefrorn, und legt sich nider in den hoff Petus.
58 Alexandre 1987, 792, 803.
59 Glaser 2008, 72.
18
1074 (2): gloomy weather on 13 March60
Hardly more than a fortnight later, another battle, this time fatal for King
Salamon, took place about 120-150 km from the previous one, at Mogyoród (in
the vicinity of Pest), on 14 March (in diluculo … sexte ferie). In fact, the battle
should have taken place on 13 March, but due to gloomy weather conditions
(very dense fog), they could not commence fighting until the following day.61
Famine in Hungary: 1074 or 1075?
The 14th-century chronicle composition mentions famine referring to the
first year of the reign of King Géza I.62 In a later chapter, a famine (probably the
same one) is mentioned during Géza I’s rule (1074-1077).63 It is the year when
the latter king gained power and stabilised his reign in March (battle of
Mogyoród) after a probably hard and late winter when armies marched through
large territories of the country. Not only the two Hungarian armies of Salamon
and Géza, but also Germans, Czechs and Moravians (each with their own
armies) were present and participated in the battles of Kemej and Mogyoród.
However, although documentary traditions indeed dated Géza’s coronation
to 1074, the latter (with a ‘royal’ crown received from the Byzantine
basileos) presumably took place in early 1075.64 Thus, it may be that the actual
famine year was not 1074 but 1075. No information is available concerning the
Czech lands for 1074, nor for 1075; just the mid-summer heat and drought are
mentioned in connection with a battle near the Unstrut river (Southwest-
Saxony).65
60 GC: 20 March.
61 SRH vol. 1, 389 (Chronici Hungarici compositio saeculi XIV): 121. Hoc dicto castra
moverunt et descenderunt in allodio quod dicitur Zymgota (Cinkota – today part of
Budapest). Mons autem quidam nomine Monorod (today Mogyoród) interiacebat exercitum,
et lucescente quinta feria volebant preliari, sed densissima obscuritas supervenientis
nebule eos impedivit. Nocte vero insequenti dextrarios suos per habenas in
manibus tenentes ortum solis prestolabantur. In diluculo autem sexte ferie rex ordinatis
agminibus suis equitavit trans verticem montis Monorod et precepit salmariis, ut in latere
eiusdem montis remanerent, quatenus hostes videntes turmas illorum, non sarcinas, sed in
presidio constitutos esse putarent. For the battle, see, e.g., Bánlaky 1930, 122-25.
62 SRH vol. 1, 403 (Chronici Hungarici compositio saeculi XIV): In primo etiam anno regni
sui (Géza I called Magnus) validissima fames regnum Hungarie afflixit.
63 SRH vol. 1, 457 (Chronici Hungarici compositio saeculi XIV): In eodem autem prelio
comes Vros cepit comitem Rapolt Teutonicum, quia iam pridem nocturnis insidiis castrum
Poson occupaverat, et Gabriel Accionem. Post hec in diebus illis fames afflixit Hungariam,
que magnam partem hominum in mortem obsorbuit.
64 Makk 1998, 155-157. Probably this is one reason why the Annales Posonienses dates the
coronation itself for 1075 while the controversies are dated there to 1074. See SRH vol. 1,
126 (Annales Posonienses): 1074. inter regem Salomonem et ducem Magnum [Géza]
gravis discordia oritur, et Salomon regno privatur. / 1075. Magnus rex coronatur.
65 Brázdil and Kotyza 1995, 93.
19
1081: The basilica ‚Crux Domini‘ in (Székes)Fehérvár
was destroyed by thunder
Both the Annales Posonienses and (partly based on it) the 14th-century
chronicle composition also reported on the lightening which struck and
destroyed the basilica Crux Domini in Székesfehérvár some time in the year of
1081.66 No information is available concerning this year in the German or Czech
(or Polish) literature for Central Europe. Nevertheless, it is interesting to
mention that, according to the Greek climate historian Ioannis Telelis, in Epiros
drought and lack of water affected the Acheron river (reported in the summer),
and in mid-July strong winds whipped up the sea at Corfu.67
Frozen Lower-Danube and dense snowfall: winter of 1087?
Similar to the (Hungarian) 14th-century chronicle composition, the 14thcentury
chronicle of Henrik Mügeln contains a description about the last years
of the former King Salamon I.68 After King Ladislaus I had allowed him to leave
the country, Salamon (with Cumans and Pechenegs by his side) attacked the
eastern parts of Hungary, and the battle ended only in 1085 when King
Ladislaus forced them out of the country and down to the Lower Danube. Then
Salamon participated in the Cuman(/Pecheneg) military campaign against
Bulgaria in 1087:69 According to the chronicle tradition, at this time in winter
(mentioned after the failed battle), there was dense snowfall, and Salamon
crossed the (Lower-)Danube when the river was frozen.70
66 SRH vol. 1, 126 (Annales Posonienses): 1081. Ladizlaus rex et Salomon frater eius pacem
fecerunt et (bis) crux Domini fulgure percussa est. SRH vol. 1, 407 (Chronici Hungarici
compositio saeculi XIV): (V) Eodem etiam anno crux Domini, que Albe constituta fuerat,
percussa est a fulgura.
67 Telelis 2008, 203.
68 SRH vol. 2, 190-91 (Chronicon Henrici de Mügeln): Das erfur der kunig und sante in
gevangen gen Nytra in den kerker. Darnach lies man in aus dem kerker, do man erhub dem
leichnam sant Stephans des kunges und sant Emreichs, seins suns. Also wart kunig
Salomon ledig und frey. Darnach zoh kung Salomon heimlich czu dem heydenischen
herczogen, der waz genant Kunkerly (sic), und globt ym sein tochter czu geben (sic) und
das lant Sibenpurgen, das er ym hulf gen dem kunig Lasla. Do liesz sich der herczog
Kumkerly uberreden und zoh mit grosser menig gen Vngern. Do daz erhort der kung Lasla,
do fur er aus mit dem her gegen den Heyden und gegen dem kunig Salomon und slug ir
manig taussent czu tode. Do floh der kunig Salomon und Kumkert (sic) aus dem streyt.
Darnach zoh der kunig Salomon gen Pulgrey und nam ein raub in dem land. Do volg in
der keyser und slug ir vil czu tode, und kunig Salomon (sic) uber die Tunaw, die was
gefrorn.
69 Dating the siege of Plovdiv to 1087 seems to be relatively well-confirmed; see: J. Bánlaky,
A magyar nemzet hadtörténelme (Military history of the Hungarian nation), vol. 4 (Budapest:
Athenaeum, 1931), 7.
70 SRH vol. 1, 409-10 (Chronici Hungarici compositio saeculi XIV): (V) 134. … At rex
Salomon cum latrunculis Cunorum invasit Bulgariam et confinia Grecie, ubi ab exercitu
imperatoris Grecie miserabiliter sunt percussi, e quibus pauci valde evaserunt. / 135. Cum
enim vidissent Cuni magnam multitudinem loricatorum, timuerunt valde ceperuntque
20
No information about the exact time (and place) of this Danube-crossing
is available, and one does not know how much of the river was frozen. But the
fact that the ice was thick enough for an army to cross suggests hard preceding
winter conditions at least for a few weeks before. However, according to Glaser,
heavy snow was reported in the Nuremnberg area for the winter of 1086 (and
not 1087).71
Falling loess wall in 1092? Some natural phenomena and the possible
environment behind
In the year 1092, an interesting set of natural phenomena, possibly related
to wet conditions and flood, was included in the description of Bernoldus
monachus, based on the reference of contemporary eyewitnesses.72 He was not
present at the events described, only heard about them from others, the actual
eyewitnesses (ut audivimus). Since the monk (who died in 1100), in those few
cases when he talked about Hungary, was otherwise quite well-informed about
the Hungarian matters (e.g., 1083: canonisations; 1096: events around the First
Crusade) and all of his reference concerning Hungarian events mentioned in his
work were correctly dated, it is worth studying his account in more detail.
But before giving full credit to the present description of Bernoldus
monachus, it is worth taking a closer look on his geographical and weatherrelated
background of these ‘possible’ events. The monk mentioned five
separate events he had heard about Hungary:73
festinare, ut transirent Danubium priusquam ab hostibus concluderentur. Qui ergo loricas
non habebant, celerius pre levitate precedebant. Salomon autem et sui atque alii loricati
armorum pondere prepediti precedentes socios tardius sequebantur, et yems erat
ningebatque densissime repleverantque nives oculos eorum et preoccupante nivium
densitate non poterant videre socios suos, qui precesserant. Errantes itaque ferebantur per
inania. Sed et hostes persequebantur eos cominus. Cumque devenissent ad quodam
castrum desertum et vacuum, intraverunt illuc et tota die et nocte ibidem manserunt.
Hostes autem circumdederunt castrum undique. Obsessi vero videntes se periculo famis
perire, elegerunt magis pugnando mori, quam fame perimi. Summo itaque diluculo
exierunt in hostes et simul cum interfectoribus suis interierunt et quamplures ex eis
occidebantur, qui eos interficiebant. / 136. Rex autem Salomon fugiens (S: evasit cum
paucis) (V) cum paucis vix evasit (S+V) et facili cursu transivit Danubium, quia glaciatus
erat.
71 Glaser 2008, 73, without reference to the original source.
72 J. R. Strayer (ed.), Dictionary of the Middle Ages, vol. 2 (New York: Charles Scribner’s
Sons, 1983), 197: „Bernold of Constance (or of St. Blasien, around 1050-1100) was canon
of Constance; monk of St. Blasien and then of Schaffhausen; chronicler, canonist, liturgist,
polemicist; and earliest and most influential disseminator of the ideas of Pope Gregory VII
in Germany. His chronicle (autograph in Munich, MS lat. 432, the Codex latinus
monacensis) is the most extensive source for the history of the Investiture Controversy.“
See also: A. Vauchez, B. Dobson and M. Lapidge, Encyclopedia of the Middle Ages
(Cambridge: James Clarke, 2001), 176.
73 MGH SS vol. 5, 454 (Bernoldi Chronicon): 1092. …. In Ungaria his temporibus multa
prodigia contigerunt, ut audivimus. Nam quidam mons se in Danubium praecipitavit: unde
21
1. A hill located along the Danube ‘slipped’ or ‘fell down’ (Nam quidam mons
se in Danubium praecipitavit) and devastated areas around it. The Danube
changed its riverbed and destroyed wide areas (unde fluvius ille alveum suum
mutare coactus, circumiacentes terras longe lateque vastavit).
2. In another place, a great lake emerged in a dry area (dryland) (Item in sicca
terra lacus magnus alibi emersit, …).
3. In another lake, a hill suddenly appeared (in alio lacu mons quidam de
repente apparuit).
4. The River Tisza(?) flowed with blood for three days (Fluvius quoque nomine
Thisaha per triduum sanguineus fluxit).
5. Immensely great thunder, which had not been heard for centuries, frightened
people, which was sensed to be crossing over the ground and then returning
under the ground (Tonitru quoque maximum et a seculis inauditum homines
ibi terruit quod supra terram transivit, et sub terra redire sentiebatur eadem
via).
First, we need to ask how he came to this information. Bernoldus at that time
(1092) lived in Schaffhausen, located in northern Switzerland on the Rhine
River, west of Lake Constance. Since Bernoldus himself died in 1100, we can
presume that he had relatively up-to-date ‘news’ about Hungary. Probably the
main sources of information were the pilgrims coming from the Holy Land via
Hungary (or travellers/merchants using this main route), crossing the town of
Schaffhausen as well. And while staying at the monastery or in the town, they
may have narrated their ‘adventures’ to Bernoldus, his fellow monks or others.
This seems more likely when we look at the location of the transcontinental
pilgrimage (and travel/trade) route through Hungary (see fig. 2). Not only each
of the five events might be analysed separately, but with the grouping of most of
them (no. 1, 2, 3, and 5) the hypothesis of one (complex) phenomenon can be
suggested (after changing the order of some events).
The environmental hazard which can only be discussed separately is
reference no. 4: namely that the Tisza(?) flowed with blood for three days. It is a
‘typical’ biblical expression, one of the scourges that affected Egypt and is
mentioned in the Old Testament. Although the exact mechanisms have not been
explained in every case, the bloody water flow on the one hand might be related
to the phenomenon of ‘blood rain’, when red Sahara dust, with hot and dry
African air masses, reached the Carpathian Basin in great quantities and affected
the colour of the rain. On the other hand, it may be explained by the process of
fluvius ille alveum suum mutare coactus, circumiacentes terras longe lateque vastavit. Item
in sicca terra lacus magnus alibi emersit, et in alio lacu mons quidam de repente apparuit.
Fluvius quoque nomine Thisaha per triduum sanguineus fluxit. Sed et terrae motus ibi
inauditus factus est, ita ut homines eius impetum stando sustinere non possent. Tonitru
quoque maximum et a seculis inauditum homines ibi terruit, quod supra terram transivit, et
sub terra redire sentiebatur eadem via. On the next page (455) Bernoldus reported an
earthquake in the same year, which occurred in the bishopric of Constance.
22
river eutrophication.74 Of course, in this particular case, one cannot arrive to any
firm conclusions concerning the reason of the “three-day blood flow”. This one
reference clearly stands out as a separate event, and cannot be related to the
other events described by the monk.
Event no. 1 contains the reference to a Danube flood, when a hill or
mountain (mons) fell into the Danube, which then caused a devastating flood
event. At first sight it sounds like a typical description of a rapid mass erosion,
which – in the case of the Middle-Danube – would most likely occur in the areas
where loess walls are located close to the Danube. Events no. 2 and 3, namely
that a great lake emerged in a dry area and in another lake a hill suddenly
appeared, are more difficult to explain. In fact, these later events sound rather
‘mysterious’, even in spite the monk Bernardus – based on his other works –
seems to be a particularly trustworthy (and well-informed) annalist of his time.
Nevertheless, he said he only heard about these events and was not present when
they took place.
Case no. 5, namely that an “immensely great thunder – which had not
been heard for centuries, frightened people, and was sensed to be crossing over
the ground and then returning under the ground” – is also difficult to interpret. It
should be mentioned that an earthquake in Hungary, exactly dated to the same
year (1092), is mentioned in the Annales Augustani.75 The author in this case
clearly referred to a major earthquake. Due to the possible similarities in
physical sensation, these two cases might have some connections to each other,
so it may be interesting to seek some later parallels.76
In the Carpathian Basin, the greatest earthquake known in detail is the one
that occurred along the Danube with an epicentre close to Komárom/ Komarno
(between Győr and Komárom) in 1763. The earthquake devastated the town of
Komárom (over 600 buildings collapsed or were seriously damaged, 63 people
died). The damage was also great in the settlements located several tens of
kilometers away from Komárom.77
74 For an explanation on the ‘blood rain’ event and its meteorological background, see, e. g., J.
R. White, R. S. Cerveny and R. C. Balling, “Seasonality in European Red Dust/‘Blood’
Rain Events,” BAMS, 2012 April: 471-76. For a possible relationship to eutrophication,
see, e. g., K. Richardson and B. B. Jørgensen, “Eutrophication, history and effects,”
Coastal and Estuarine Studies, 52 (1996): 1-19.
75 G. H. Pertz (ed.), Monumenta Germaniae Historica. Scriptores, vol. 3: Annales, chronica et
historiae aevi Saxonici (Hannover: Impensis Bibliopolii Hahniani, 1839) (hereafter: MGH
SS vol. 3), 134 (Annales Angustani): 1092. In provincia Ungarorum una die, 6. Kal. Iulii
ter terrae motus factus est urbesque submersae sunt.
76 And even if we treat case no. 5 separate from the earthquake report, due to the later
complex interpretation issues, it is worth to discuss possible parallels already at this place.
77 For an overview in Hungarian, see for example: P. Varga, “Földrengések a történelemben.
Komárom katasztrófája 1763-ban” (Earthquakes in history. The catastrophe of Komárom
in 1763) História 8 (1998); for an overview in English see, e. g., the case study in: U.
Eisinger, R. Gutdeutsch and Ch. Hammerl, “Historical Earthquake Research – an Example
of Interdisciplinary Cooperation between Geophysicists and Historians,” in R. Gutdeutsch,
23
In line with the known evidence, cases no. 4 and no. 1 can be separately
analysed and an interpretation given for each. No. 2, 3, and even no. 5 cannot be
readily explained as separate events. However, if we combine these latter three
events with the ‚easy case‘ of no. 1, a rather reliable-looking ‘story’ can be
provided, and it is relatively easy to find a modern, well-described parallel for
this complex event.
As can be seen in fig. 2, the main branch of the high medieval
transcontinental route through Hungary mostly led along the Danube. Entering
Hungary, it first went along the Danube, then turned away from the river
towards the southeast, and reached the Danube again at its middle section,
continuing along the river for hundreds of kilometres. If travellers saw or heard
about any natural phenomena related to the Danube, it is more likely to have
taken place either at the upper, or the latter middle sections (with loess walls
alongside it; see fig. 2).
In historical Hungary (except for the Devín Gate area), the Upper-Danube
section is completely flat colluvium with several islands. The Danube passes
through a hilly area in the Danube Bend. Here the hills are of volcanic origin
and hence in this area with the hard volcanic stones it is relatively difficult to
imagine that an entire hill would have just ‘slipped’ into the Danube. Along the
sections south of Budapest, however, there is a rather extensive area that seems
to be an ideal place for the phenomena apparently described by the monk.
Although this is also a lowland area – for several dozens of kilometres the southeastern
edge of the Mezőföld loess plain (plateau) descending towards the
Danube –, it is divided into a series of (locally called) ‘hills’78 following the
western banks of the river (see fig. 2). The edge of this loess plateau is
approximately 50-60 metres higher than the water level of the Danube. The
eastern edge of the plateau, facing towards the Danube, is even today a high-risk
area, and subject to such ‘slipping’ or ‘falling‘ events (which actually happened,
for example, in the extraordinary wet year of 2010).
G. Grünthal and R. Musson (eds.), Historical Earthquakes in Central Europe, vol. 1
(Vienna: Geologische Bundesanstalt, 1992), 33-50.
78 In medieval and modern Hungarian terminology, the Hungarian word ‘hegy’ (Latin mons)
was equally applied for a ‘real’ hill/mountain (in its present geographical meaning), but
also for smaller mounds rising above its surroundings. See, e. g., Gy. Kristó, Tájszemlélet
és térszervezés a középkori Magyarországon (Landscape perception and the organisation
of space in medieval Hungary) (Szeged: Szegedi Középkorász Műhely, 2003), 37.
24
Fig. 2: International route of medieval pilgrimage (11th-12th centuries) going through
Hungary,
and the high-risk edge of the loess plateau along the Danube79
Nowadays this middle section is the area where there could be such a
mass-erosion event most likely within a very short period of time (several
minutes to a few hours) and thus it is worth having a closer look on modern
parallels of mass erosion events in this area. Here the loess plain of the
Mezőföld reaches the Danube and forms an unstable set of loess walls between
Paks and Dunaszekcső (fig. 2). Underneath the Mezőföld loess plain, clay layers
form a hollow declining towards the Danube. This clay hollow collects and
79 For the Hungarian section of the medieval route of pilgrimage, see Borosy and Laszlovszky
2006, 78, 81. See also L. Gerevich (chief ed.), Budapest története az őskortól az Árpád-kor
végéig (History of Budapest from the Neolithich period until the end of the Arpadian
period) (Budapest: Budapest Főváros Tanácsa, 1973), 274. The map is based on T. Dudar
(chief ed.), Történelmi világatlasz (Atlas of world history) (Budapest: Kartográfiai Vállalat,
1991), 109.
25
leads waters towards the Danube, under the loess layer. There, with the slippery
surface between the clay and the loess, a wet preceding period as well as human
activity over the loess surface, (even minor) earthquakes and/or the existence of
some nearby water reservoir, and water-level changes in the Danube may be
chiefly responsible for the fall of a loess ‘hill’.80 Very frequently, such an event
occurs at a place where a previous falling event already obstructs the water in its
natural downflow underneath the loess layer. It should also be mentioned that
the occurrence of this certain phenomenon requires the combination of specific
conditions (e.g., wet period, water-level changes, minor earthquake/artificial
damming) listed above. This phenomenon is very typical even today, and we
know of several such cases like that in Dunaszekcső, Dunaújváros and
Dunaföldvár (for the modelling of such a phenomenon, see fig. 3). In these
towns, several buildings and streets are in constant danger of ‘being engulfed’
by a loess fall event, or being buried under a falling loess wall.81
The fall of the loess wall of Dunaföldvár called ‘Öreg hegy’ (‘Old Hill’,
or it can also mean ‘Big Hill’ in old Hungarian), shown in fig.3, that occurred in
1970, perhaps provides a good parallel with the account of Bernoldus monachus.
Here the previous, almost one-year period was wet (with a 20% precipitation
excess) in the Mezőföld area. Moreover, the summer of 1970 was very rainy.
Then the (sandy clay) layer between the clay and loess became wet and slippery.
The water level of the Danube was high in late August – early September. Also
some (minor) earthquake events had occurred previously. As a result, on 15
September the Öreghegy collapsed with an enormous cracking sound (just like
case no. 5 in Bernoldus‘ description?). The loess pushed the clay up, forming
two new rows of islands in the Danube. It also led to a sudden flood event that
devastated and inundated the surrounding areas (see fig. 3).82
In summary, a preceding wet period (over several seasons, a year or
several years) over the underground catchment area (the Mezőföld) as well as
some physical disturbance (minor earthquake) and improper outflow (abundant
rainfall, high water-level conditions of the Danube) are all necessary circumstances
that could have induced the mass-erosion event described above. Hence,
whenever such an event happens, it already suggests a previous wet period and
high water-level conditions over a well-defined, larger area.
80 M. Pécsi, “Az 1970. évi dunaföldvári földcsuszamlás” (The loess-fall event of Dunaföldvár
in 1970), Földrajzi Értesítő 20/3 (1971): 233-38.
81 M. Pécsi, Geomorfológia és domborzatminősítés (Geomorphology and morphological
evaluation) (Budapest: MTA Földrajztudományi Kutató Intézet, 1991), 171.
82 Pécsi 1971, 233; Pécsi 1991, 171.
26
Fig. 3: Modelling the fall of a loess wall called ‚Öreghegy‘ in Dunaföldvár, 1970
(after Pécsi 1971)
One can see several similarities between the description of Bernoldus
monachus and the 1970-fall of the Öreghegy at Dunaföldvár. Not only the first,
but nos. 2, 3, 5 (and probably even the earthquake ‘story’) as well might be
worth taking into account, even if some of them are described as separate
events. A separate analysis of the events nos. 2, 3 and 5, due to lack of
information, would not lead us too far. Nevertheless, when taken together they
show clear similarities to a prototype of a loess-fall event, like that described
above: a hill falling into the Danube, the river changing its bed and destroying
(flooding) broad areas, a lake emerging, a ‘hill’ appearing in the lake, an
earthquake, a great thunder-like noise frightening people, which felt like it
crossed over the ground and then returned under the ground.
It should be noted, however, that Bernoldus monachus was not an
eyewitness. He narrated accounts from Hungary that he had not seen and only
heard from others. A biblical meaning and therefore an attempt to find a set of
events related to the disasters and signs predicting the end of the world might
also have influenced his description of the events. Moreover, the present
interpretation is a likely possibility, but no firm conclusions can be drawn in
general. Therefore, further information and other possible parallels for the year
of 1092 might have great significance when trying to explain what happened
then. Based on Cosmas Pragensis, for example, a great intrusion of winter with
ice, frosts and abundant snowfall was witnessed on 1 April, 1092.83 No other
direct, weather-related information in the vicinity is currently available for this
83 B. Bretholz (ed.), Monumenta Germaniae Historica. Scriptores rervm Germanicarvm nova
series, vol. 2: Cosmae Pragensis Chronica Boemorvm (Berlin: apvd Weidmannos, 1923)
(hereafter MGH SS rer. Germ. NS vol. 2), 157: Quos dux iam mutato animo benigne
suscepit et celebravit cum eis pascha in urbe Wissegrad. Et in ipsa pasche ebdomada circa
kal. Aprilis descendit maxima nix et tantum inhorruit frigus mixtum cum glacie, quantum
raro in media contigit hyeme. For a further analysis, see Brázdil and Kotyza 1995, 94.
27
year. Nevertheless, somewhat wetter autumns were mentioned for the 1090s by
Glaser.84
Weather or not? Events documented in the 12th century
Hard winter of February 1109
Reported by Cosmas Pragensis, during the military campaign of the
Czech prince Svatopluk into Hungary in February 1109, the waters were
frozen.85 Due to the favourable winter conditions, the prince very quickly
reached the town of Nyitra (today Nitra in Slovakia). The information about his
invasion refers to the northwestern areas of medieval Hungary, what is today’s
western Slovakia.86 In the Middle Ages, a winter campaign in most cases was
made when the circumstances were appropriate, and thus it usually took place in
colder weather conditions.87 Glaser, without further details, mentioned that the
winter of 1109 was cold.88
End of 1129: (Hard?) winter conditions in Barancs/Branitzova
The Byzantine Ioannes Kinnamos blamed winterly weather conditions
and the lack of the necessary equipment for the hardships of soldiers at the
fortress of Barancs/Branitzova (today Braničevo in Serbia).89 In the Annales
Mellicenses, during the year 1129, a controversy between Greeks and
Hungarians at the Sava river is mentioned where the siege of a fortress took
place.90 The medieval fortress of Braničevo, nevertheless, was located on the
Lower Danube. It was destroyed at the beginning of the Hungarian-Byzantine
84 Glaser 2008, 88.
85 G. H. Pertz, Monumenta Germaniae Historica. Scriptores, vol. 6: Annales et chronica aevi
Salici (Hannover: Impensis Bibliopolii Hahniani, 1844) (hereafter MGH SS vol. 6), 747
(Annalista Saxo): 1109. .. Mense Februario, quia magnum frigus inhorruerat, et omnes
aque congelaverant, dux Svatopluk, iam eruti sanato vulnere oculi, coadunato iterum
exercitu, tribus diebus et tribus noctibus festinans intrat Ungariam, et omni regione illa
devastata, cum multis captivis et magna preda reversus est.
86 MGH SS Nova series vol. 2, 115 (Cosmas Pragensis): Lib. III. (14 Feb. 1109) XXVI. Anno
dominice incarnationis MCVIIII. Decima luna mensis Februarii, quia grande inhorruerat
frigus et omnes aque erant congelate valde, dux Zuatopluk iam eruti sanato vulnere oculi,
continuo iterum coadunato exercitu tribus diebus et tribus noctibus continuis festinans
intrat Ungariam et nemine eorum presciente ex inproviso applicuit cum exercitu ad urbem
Nitram; et irrupisset in eam, si non vigiles, qui semper ibi sunt custodientes, clausissent
portam.
87 See also Brázdil and Kotyza 1995, 95.
88 Glaser 2008, 72.
89 Gy. Moravcsik, Az Árpád-kori magyar történet bizánci forrásai (Byzantine sources of
Hungarian history in the Arpadian Period) (Budapest: Akadémiai Kiadó, 1988), 198: O δέ
έν Βρανιτξόβη τό δεύτερον γεγονώς άνικοδομεϊν αύτήν σπουδήν έποιεϊτο. Χρόνου γοΰν έζ τό
έργον τριβομένου ξυνέβαινε χειμώνι καί σπάνει τών άναγκαίων δυσπαθοϋν τό στρατιωτικόν
ές άγαν ταλαιπωρεϊσθαι.
90 MGH SS vol. 9, 502 (Annales Mellicenses): 1129. Greci et Ungari discordantes apud
fluvius Sowa castra metati sunt.
28
war that had begun in 1127, when in summer the Hungarians attacked
Braničevo. The Byzantine emperor took the fortress back in 1128, but the
Hungarians destroyed it again in 1129. It was already winter time when the
(Byzantine) emperor’s army reached Braničevo to rebuild it, presumably around
the time when the peace treaty was signed at the end of the year.91 Based on the
above-mentioned account, the winter of 1129(-1130) – probably late autumn or
December – seems the most likely for dating the hard winter conditions
mentioned by Ioannes Kinnamos. No information is available for this winter in
the neighbouring regions.
The famine year of 1141
According to the 14th-century chronicle composition, during the reign of
Géza II, a famine occurrred in 1141 (Post hec in diebus illis fames afflixit
Hungariam, que magnam partem hominum in morte obsorbuit).92 Following this
statement, the Chronicon Monacense also reports the same event (with the same
presumable exaggeration concerning the number of casualties).93 The text
mentions it right after the castle of Pozsony (today Bratislava in Slovakia) was
taken back from the Germans. Hence, this famine could have occurred both due
to poor harvests but also due to hostilities that were also going on in the
neighbouring regions at about the same time.
In the Czech Lands, the winter of 1140-1141 was “changeable”, while a
great flood occurred in Prague in April 1141 (also mentioned in the same work
at another place concerning Bohemia in general). The most interesting information,
however, is provided for 1142: a great famine occurred in summer. According
to Brázdil and Kotyza, it was not due to poor harvests caused by bad
weather conditions, but rather due to the war. Thus, if the datings are correct in
both cases, a famine was reported for Hungary in 1141, while the Czech Lands
had to face with the same problems in the forthcoming year, namely in 1142.94
91 Gy. Kristó, Az Árpád-kor háborúi (Wars of the Arpadian period) (Budapest: Zrínyi Kiadó,
1986), 75-76.
92 SRH vol. 1, 457 (Chronici Hungarici compositio saeculi XIV): Ibi usque adeo contritum est
robur Teutonicorum, quod nec rusticis Hungaris in confinium (sic) eorum habitantibus
aliquam iniuriam seu quodcumque gravamen inferre presumperunt. Rex autem divina
gratia felici potitus victoria, simul cum universa gente sua benedixit et gloriavit Deum. In
eodem autem prelio comes Vros coepit comitem Rapolt Teutonicum, qui iam pridem
nocturnis insidiis castrum Poson occupaverat, et Gabriel Accionem. Post lec indiebus illis
fames afflixit Hungariam, que magnam partem hominum in morte obsorbuit.
93 SRH vol. 2, 81 (Chronicon Monacense): 54. Geysa, filius Bele Ceci coronatur. Contra hunc
Henricus dux Austrie congregato exercitu de Saxonia et Bauaria, rex cum Hungaris
accelerat super eos, et commissum est relium et Hungari a furore Teutonicorum territi
sunt, ita quod a principio Teutonici prevaluerunt. Ipsi autem magna strage post prostrati
sunt, circiter VII milia, reliquis fugientibus. Post hoc fames ingens afflixit Hungariam, ut
magna pars hominum interiret.
94 Brázdil and Kotyza 1995, 98.
29
If we examine the other neighbouring areas, no famine and related
problems are known for these two or the subsequent years. In the Auctarium
Zwetlense a two-year famine is mentioned (for Austria) during the year 1149,95
whereas the Auctarium Lambacense reported on the great death of domestic
animals (from disease?) during the year 1148.96
It is also interesting to see that, according to Glaser, in the 1140s wet and
cold summers prevailed in the German areas, whereas in Flanders the summer of
1141 was especially cold. No information is provided concerning 1142, but
1143 was mentioned by Glaser as a year with an especially rainy summer. The
winter of 1141 was mild, but the winter of 1143 was long and snowy. The spring
of 1141 was markedly warm, while in the autumn of 1141 there was extensive
flooding along the coast.97
Frozen Danube: winter of 1147?
In her work entitled ‘Alexiada’ (about the life and work of her father), the
educated Byzantine Princess Anna Komnena provided a testimony of the event
that the (Lower) Danube froze in the winter of 1146-1147. The ice cover of the
Danube was thick enough to hold the weight of an army when the Hungarians
crossed it.98 According to Glaser, there was heavy snow in the winter of 1147.99
If both datings are correct, the winter of 1147 led to severe conditions in large
parts of (Central) Europe.
Rainy weather caused the inundation of swamps around the Drava river in
1147?
Documented by Odo de Deogilo who travelled with the second Crusade
through Hungary, an enlargement of local swamps occurred due to heavy
rainfall in 1147. Here the river itself is not directly mentioned to be in flood, but
rather the swamps in the vicinity, even if the river did burst its banks.100 As the
95 MGH SS vol. 9, 540 (Auctarium Zwetlense): 1149. Fames valida duravit per biennum.
96 MGH SS vol. 9, 555 (Auctarium Lambacense): „1148. … Mortalitas pecorum maxima est
facta.“
97 Glaser 2008, 62, 73, 83, 88.
98 Moravcsik 1988, 103 (Anna Komnena): Διά ταϋτα καιρόν έπιτηροϋντες, έπεί τόν Ϊστρον
άποκρυταλλωθέντα εϊδον, ώσπερ ήπείρω τούτω χρησάμενοι κεϊθεν πρός ήμάς μετανίστανται
ολον εθνoς τοϊς ήμετέροις έπιφορτισθέν όρίοις καί δεινώς έλήζοντο τάς παρακειμένας πόλεις
καί χώρας.
99 Glaser 2008, 73.
100 J.-P. Migne, Patrologia Latina vol. 185 (Paris, 1855), 21-23 (Odo de Deogilo): De
profectione Ludovici VII in Orientem (Odonis de Diogilo De Ludovici VII Francorum
regis cognomento junioris profectione in orientem Cui ipse interfuit Opus septem libellis
distinctum. Epistola Odonis ad venerandum abbatem suum Sugerium). Libellus Secundus:
…. Hungaria ex hac parte aqua lutosa cingitur; ex alia vero a Bogaria amne lucido
separatur. In medio sui fluvium habet Droam, qui stanni more unam ripam proclivem
habet, et alteram arduam, unde modica pluvia effluit, et adjutus vicinis paludibus, etiam
aliquanto remota submergit. Audivimus eum multos Alemannorum, qui nos praecesserant,
30
swamps around the river extended, crossing the Drava river caused great
difficulties; many Crusaders inadvertently went under the waters. There was
also great pestilence.101
Since on 24 June the Crusaders were already at the Hungarian-Byzantine
border,102 the time of crossing the Drava River could have been around earlymid
June. No significant problems were mentioned concerning the crossing of
the (even larger) Sava River. As such, probably May – early June (and/or the
preceding period) was rich in precipitation in the Drava catchment. As we saw
earlier, Glaser suggested snowy winter conditions for 1147, and the
contemporary Byzantine evidence (Anna Komnena) also provided information
on prolonged cold winter conditions for the same part (frozen Lower Danube). If
the winter was rich in precipitation, this could have affected the water system of
the Drava River (as well as the lowland areas) for several months (especially if
spring – early summer was not dry). It is interesting to compare this description
with the case of 1189 when Crusaders again travelled through the country and
again crossed the Drava River at that time without any notable difficulties.
1154(-1155)?: Cold winter, dense snowfall and flood
Theodoros Prodromos mentions in his work the dense snowfall and
coldness of winter, in connection with the military campaign of Hungarians.103
Since late 1150, with shorter or longer interruptions, a war was going on
between the Byzantine Empire and Hungary, until spring 1155 when a peace
treaty ended the controversy. In the course of this five-year war, the military
campaign of the Hungarians described by Ioannes Kinnamos for the end of the
year 1154, seems to be the most likely candidate for the event (and its dating).104
A probable Danube flood event was recorded by Ioannes Kinnamos,
either for late 1154 or early 1155 (during the attack of Hungarians against
subito inundasse; nos autem ubi castra eorum fuerant, vix potuimus transvadere. In hoc
parvas naves habuimus et paucas; ideo fuit opus equos natare, qui facilem ingressum et
egressum difficilem habentes, cum labore quidem, (1213B) sed tamen Deo volente sine
damno transibant. Caetera omnis aqua terrae hujus, lacus sunt, et paludes et fontes, si
tamen fontes sunt, quos paululum fossa humo, etiam in aestate faciunt transeuntes,
excepto Danubio qui hanc satis in directum praeterfluit, et multarum regionum divitias
nobili civitati Estrigin navigio invehit. ….
101 G. H. Pertz (ed.), Monumenta Germaniae Historica. vol. 17: Annales aevi Suevici
(Hannover: Impensis Bibliopolii Hahniani, 1861) (hereafter MGH SS vol. 17), 27 (Annales
ducum Bavarie Altahenses): 1147. Cuonradus rex Ungariam intrat, ac ducem eorum
Wardiz nomine bello petit, igne praedaque vastat universa, et rebellem ducem obsides
dare sibique iureiurando fidelitatem confirmare coartat. Pestilentia magna facta est.
102 Borosy and Laszlovszky 2006, 86.
103 Moravcsik 1988, 179 (Theodoros Prodomos): Επί δέ τούτοις άμελώς ό Παννονάρχης έχων
καί μόνοις τείχεσιν ύγροϊς θαρρών καί περιβόλοις άμεριμνών έκάθητο δεινόν ούδέν
έλπίζων καί τοϋ χειιμώνος τό δριμύ καί τοϋ καιροϋ τήν ώραν καί τών νεφών τήν πύκνωσιν
όρών έπεγαρία παρέβλεπε τούς πρεσβευτάς, τόν Άρην έπεζήτει, έκόμπαζεν ούκ άνεκτά,
ξυρόν τήν γλώσσαν εϊχεν, …
104 Kristó 1986, 87; Moravcsik 1988, 209-14.
31
Braničevo): “However, when they (Hungarians) arrived (at the crossing place),
the Danube was in flood (raging stormy weather), and therefore they turned
towards Belgrade.”105 Since Byzantine authors of that period often used
symbolic pictures to make fights more heroic, the evidence of a flood event
cannot be completely confirmed, even if Kinnamos on political and other
questions seems to be a reliable source. No other weather reports are currently
known concerning the areas near the Carpathian Basin.
1189: Memoires of a Crusader about Hungary
During the Third Crusade, in June 1189, while crossing Hungary along
the Danube, Ansbertus clericus noted that, in spite of all previous information
and expectations, they travelled in much more pleasant conditions through
Hungary than was usual here in the summer time, namely, without the disturbance
of the ‘usual animals’. The clericus mentioned the lack of mosquitoes,
snakes and flies.106
From the same author we know that they entered the country in the last
week of May, in the Pozsony (today Bratislava in Slovakia) area, where they
stayed for a couple of days. Then they continued their way to Esztergom, down
along the Danube and they crossed the Drava river; afterwards they went to
Belgrade. After crossing the border, at the beginning of July, they reached the
Byzantine Braničevo. Therefore, Ansbertus’ (general) description refers to (the
entire month of) June 1189 when the Crusaders crossed Hungary from Austria to
the Byzantine empire, along the Danube.107
It seems likely that the weather conditions, unfavourable for the abovementioned
animals, occurred before or during the time when the Crusaders
crossed the country along the Danube (thus, along or through its extensive
floodplains). Due to biological factors necessary for the development of these
animals (with special emphasis on mosquitoes), either constant dry conditions or
prolonged floods are less favourable. The lack of these animals altogether might
105 Moravcsik 1988, 212 (Ioannes Kinnamos): Οπως μέντοι Βρανιτζοβϊται τήν πόλιν τέως
αύτώ τηρήσωσι, γράμμα χαράξας οσον ούκ ηδη άφίξεσθαι τοΰτον μηνΰον τών στρατιωτών
τινι ένεχείρισε κελεύσας έπί τήν πόλιν αύτό διαφεναι βέλει ξυνδήσαντα. Ό μέν οΰν κατά τό
κελευσθέν έποίει. Τό δέ βέλος πορρωτέρω ή έχρήν άφεθέν Οΰννοις είς χείρας ήλθεν. Εύθύς
οΰν ταραγμός είσηλθεν αύτούς, καί δή τάς τε έλεπόλεις καί οσα πρός τειχομαχίαν αύτοϊς
παρεσκεύαστο πυρί φλέξαντες έπί τόν Ίστρου πορθμόν ήεσαν. Πλήθοντι δέ αύτω
έντετυχηκότες (χειμών γάρ έκ των άνωθεν ήπειγεν), ώς έπί Βελέγραδα πόλιν έφέροντο.
106 A. Chroust, Monumenta Germaniae Historica. Scriptores rerum Germanicarum, nova
series vol. 5: Quellen zur Geschichte des Kreuzzuges Kaiser Friedrichs I: Historia de
expeditione Friderici imperatoris (Berlin: Weidmannsche Buchhandlung, 1928) (hereafter:
MGH SS rer. Germ. N.S. vol. 5), 27 (Ansbertus clericus Austriensis): A. 1189 II
nonas Iiunii … In maxima denique tranquillitate et aeris temperie plus solito mitiore nobis
arridente in tantum, ut culices, oestra, musce, serpentes, que in Ungaria estivo tempore
iter in equis agentes graviter inquietant, non solum nos vel animalia non lederent, verum
etiam raro a nobis visa sunt, Ungariam permeavimus. VI nonas Iulii ….
107 MGH SS rer. Germ. N.S. vol. 5, 17-27.
32
rather suggest prevailing dry conditions. According to Glaser the summer of
1189 was also warm in the German areas.108
Danube floods in 1194 in Bavaria and Austria – any relationship with
Hungary?
Without mentioning the location, Radulphus Niger also provided
information on a sudden great flood of the Danube in 1194 (‘around this time’),
which caused a large number of casualties and great material damage.109
Referring to the Danube at Neuburg (Bavaria), flood in 1194 was also reported
by the Continuatio Cremifanensis, according to the author, caused by ice. This
fact would suggest prolonged cold winter conditions in Bavaria in the preceding
period. In fact, the Continuatio Cremifanensis referred to three floods, and also
mentioned that fire devastated towns and villages in the same year. All these
events destroyed the harvest, and the floods caused significant damage to houses
and settlements.110 A devastating drought and fire in Vienna, and an unusual
Danube flood are described for 1194 in the Continuatio Claustroneoburgensis.
111 Moreover, famine is also mentioned at the same time in
Teutonia.112
Since the first Danube flood event, reported for the Bavarian sections, was
most probably an ice (jam) flood, this did not necessarily appear on the lower,
Austrian sections. This would explain the fact that for Bavaria there were three
floods mentioned in the sources, while for Klosterneuburg ‘only’ two floods. It
108 Glaser 2008, 62.
109 G. H. Pertz (ed.), Monumenta Germaniae Historica. Scriptores, vol. 27: Ex rerum
Anglicarum scirptoribus saec. XII. et XIII (Hannover: Impensis Bibliopolii Hahniani,
1885) (hereafter MGH SS vol. 27), 332 (Radulphi Nigri Chronica Universali): /A. 1194/
Circa hec tempora Danubius subito irrupit et multos fines terrarum extinxit, homines et
iumenta et alia terrarum volatilia. Redeunte demum Danubio in alveum suum, exierunt de
terra culices et vastaverunt agros et vineas et arbusta et depopulaverunt universam
terram. Irrupuerunt igitur porci et alique alie bestie et irruerunt in culices, devorantes eas
et ingrassati sunt vehementer, et muti mortui sunt.
110 MGH SS vol. 9, 548 (Continuatio Cremifanensis): 1194. … Danubius in Austria glaciali
molestia versus Niwenburch ebullit, et meatum suum longius vagando relinquens, homines
cum villis mergendo adnichilavit. Eadem quoque regio mirabili incendio in civitatibus et
opidis et villis et uredine in agris longius late raa annona consumendo depauperatur.
Inundationes tribus vicibus magnae et intolerabiles atque insolitae omnes pene regiones,
villas et circumpositas domos demoliendo occupaverunt.
111 MGH SS vol. 9, 619 (Continuatio Claustroneoburgensis II): 1194. Hoc in anno propter
ariditatem terre que totam Austriam invaserat, coloni eiusdem terre cruces suas tollentes
cum virginibus diversa limina sanctorum pecierunt. ….. Idus Aug. …. Wienna incendio
flagrat. Danubius vice altera contra solitum inundans, vicina loca cum hominibus et
iumentis occupat. These events (according to the editors erroneously) are partly also
described under the year 1193.
112 MGH SS vol. 9, 619 (Continuatio Claustroneoburgensis II): 1194. Fames valida facta est
in Teutonia.
33
is possible that these two floods occurred later (e.g., spring, summer or autumn),
and affected larger sections of the Danube.
Why is this interesting from a Hungarian perspective? Since the
Klosterneuburg annals clearly referred to the flood events, the Vienna area was
most probably affected and thus, there is a good chance that these flood waves
also arrived to Hungary. However, for the autumn of the same year, a locust
invasion, usually associated with dry and hot (pre)conditions, was reported in
Hungary at the Drava–Sava interfluve (see below).
Locust invasion in 1195 and/or 1194 (1193)
In the year 1194, whose spring and summer seasons were dry in Austria113
and in large parts of Europe,114 locusts were reported for Hungary, which
occurred in the autumn. The chronicle of Albricus dated the locust invasion
(pluvia locustarum densissima) to the autumn of 1194, which, according to his
knowledge, was rather memorable and affected the Drava–Sava Interfluve.115
Locust plagues are usually assumed to be a consequence of prevailing dry
conditions, and it seems that in many parts of Europe the spring and summer of
1194 were quite dry. The shorter Austrian chronicle of Melk116 as well as the
Chronicon Belgicum magnum117 (both non-contemporary) dated the locust
invasion to the same time and place, namely to autumn 1194 for the Drava-Sava
Interfluve. Vitus Arnperckius (non-contemporary: 15th century) provided 1193,
and as locations Hungary, Styria, and Croatia.118
113 See, e. g.,: MGH SS vol. 9, 619 (Continuatio Claustroneoburgensis II): 1194. Hoc in anno
propter ariditatem terre qu totam Austriam invaserat, coloni eiusdem terre cruces suas
tollentes cum virginibus diversa limina sanctorum eligitur.
114 For a broader European overview see Alexandre 1987, 368-69; see also Brázdil and
Kotyza 1995, 100-01.
115 MGH SS vol. 23, 870 (Chronicon Albrici monanchi): 1194. Memorabile quiddam contigit
hoc anno in Hungaria, quia in autompno quadam die fuit inter Dravum et Savum pluvia
locustarum densissima.
116 G. Waitz (ed.), Monumenta Germaniae Historica. Scriptores. vol. 24: Annales aevi
Suevici. Gesta saec. XII. XIII. (Hannover: Impensis Bibliopolii Hahniani, 1879) (hereafter:
MGH SS vol. 24), 223 (Chronicon Austrie Mellicense breve): 1194. Henricus imperator
Calabriam, Apuliam et Siciliam sibi bello subiugavit. Item mirabile quiddam contigit in
Ungaria hoc anno, quia in autumno, quoddam die, fuit inter Dravum et Savum pluvia
locustarum densissima.
117 J. Pistorius, Scriptores Rerum Germanicarum, vol. 3 (Regensburg: Sumptibus Joannis
Conradi Peezii, 1731), 223 (Magnum Chronicon Belgicum): Item anno Domini MCXCIV.
Henricus Imperator Calabriam, Apuliam & Siciliam sibi bello subiugavit. Item mirabile
quiddam contigit in Ungaria hoc anno, quia in Autumno, quodam die, fuit inter Dravum &
Savum pluvia locustarum densissima. It is interesting that the author, who otherwise was
seemingly rather well-informed in ‘hot news’, did not mention such or related events for
any other parts of Europe in this year (neither in the preceding nor subsequent years).
118 SRA vol. 1, 1204-05 (Vitus Arnperckius): 1193. Ea tempestate magna turma locustarum
quatuor alas habentium et volantium per Hungariam, Croatiam et Styriam cuncta
devastantium: et cum volabant, terram abumbrabant et nebula spissa.
34
The contemporary Annales canonici Sambiensi119 as well as the
Klosterneuburg annals (Continuatio Claustroneoburgensis III), however, dated
an invasion of four-winged locusts to the autumn of 1195.120 The latter narrative
generally talked about the devastation caused by locusts in Hungary (as well as
in the Marchia and Carniola) and no longer mentioned the invasion of the
Drava–Sava Interfluve.121
It is also interesting to note that the Annales S. Stephani Frisingenses
mentioned many devastating fires in 1193, when – at the beginning of spring –
their monastery completely burnt down.122 As mentioned earlier, the Continuatio
Cremifanensis also reported on the fires in the Neuburg area for the year 1194.
Moreover, as provided by U. Bieber, an extremely devastating, prolonged fire
event (having started around All Saints) was also documented for Novgorod in
1194.123
All these narratives may be of help to us if we take into account the fact
that a hot and dry summer, followed (or preceded) by cold and snowy winters
may provide favourable conditions for locust swarms124 – conditions also
favourable for the start of such an invasion in their usual breeding areas (i.e.,
where locusts are endemic, e.g., around and between the Caspian and the Black
Seas). The winter of 1194 could have been a hard one and resulted in an ice
(jam) flood on the Danube in Bavaria. Presumably, there were other flood(s) of
119 G. H. Pertz (ed.), Monumenta Germaniae Historica. Scriptores, vol. 19: Annales aevi
Suevici (Hannover: Impensis Bibliopolii Hahniani, 1866) (hereafter MGH SS vol. 19), 698
(Annales canonici Sambiensi): 1195. Autumpnali tempore quoddam genus locustarum
quatuor pennas habens venit a desteris regionibus transiens per Ungariam et Marchiam
et Carniolam, vastans in circuitu omnia.
120 MGH SS vol. 9, 634 (Continuatio Claustroneoburgensis III): 1195. … Hoc anno
autumpnali tempore quoddam genus locustarum, 4 pennas habens, venit ab exteris
regionibus, transiens per Ungariam et Marchiam et Carniolam, vastans in circuitu omnia;
cum elevaretur a terra quasi nebula consurgebat, et pre multitudine sua lucem diei et
splendorem solis obtenebrabat.
121 See for a catalogue of medieval locust invasions in Italy: D. Camuffo and S. Enzi, “Invasioni
di cavallette e fattori climatici dal medioevo al 1800,” Bollettino Geofisico 14/2
(1991): 1-51.
122 G. Waitz (ed.), Monumenta Germaniae Historica. Scriptores, vol. 13: Supplementa
tomorum I-XII, pars I (Hannover: Impensis Bibliopolii Hahniani, 1881), 55 (Annales S.
Stephani Frisingenses): 1193. Hoc anno locus noster multis malis et infortuniis attritus
est, tam incendii devastatione quam grandinis depopulatione; nam monasterium nostrum
cum claustro totaliter combustum est 5. Idus Maii, incipiente noctis crepusculo.
123 U. Bieber, “The Reports of ‘Climate Change’ in Medieval Russian Chronicles and their
Impact on Everyday Life,” paper presented and handout provided at the International
Medieval Congress 2008 (10 July), University of Leeds, p. 1: entry No. 1 (quoting original
text in translation).
124 See e.g. A. Kiss, “A sáska járások néhány területi és tájtörténeti vonatkozása a Kárpátmedencében”
(Some regional and areal consequences of historical locust invasions in the
Carpathian Basin), in A táj változásai a Kárpát-medencében. Történeti emlékek a tájban
(Landscape changes in the Carpathian Basin. Historical monuments/memorial sites in the
landscape), ed. Gy. Füleky (Gödöllő: Szent István Egyetem, 2012), 78-87.
35
the Danube in this year in Bavaria and Austria. A dry and hot spring and
summer as well as devastating fires were also reported in Austria, Bavaria and
other parts of Europe for the same year. The locust invasion can be dated with
certainty only for 1195. However, if great devastation was also caused in 1195
in the (south-)western parts of the Carpathian Basin, the locusts most likely
appeared there earlier (in 1194 or even in 1193).
An overview and discussion of results
Which weather and weather-related environmental information could be
extracted from this period?
There is a small but well-identifiable group of evidence, which directly
holds (direct or indirect) weather-related information. Apart from this, there is a
mixed set of information which usually has some kind of connection with
preceding or actual weather conditions and is hence interesting, also concerning
their possible long term impact. The following results can be summarized:
1) Evidence of (prolonged: monthly, seasonal) weather conditions:
a. Frozen-solid rivers – hard winter conditions
Cold winters with great frost (the length of hard conditions unknown)
can be identified for February 1074, when the River Tisza was frozen
solid. Hard winter conditions with a frozen Lower Danube and dense
snowfall was presumably reported for 1087. Winter conditions were
also hard in February 1109 in the area of present-day western Slovakia.
Related to a military campaign, problems caused by winter conditions
in 1129 at Braničevo (Lower Danube area) were mentioned. A solidly
frozen Danube was again mentioned in the south in 1147. The coldness
of the winter and dense snowfall were blamed by Byzantine authors for
difficulties in military operations, probably referring to events in late
1154 or early 1155.
b. Dry conditions, heatwave and a drought
A drought might have occurred in the summer of 1044, and dry
conditions were potentially present in June 1189, when Crusaders
crossed the country along the Danube (although in this case no firm
conclusion could be drawn). Great heat and a drought prevailed in
Central Europe, and also in the neighbouring countries (i.e, Austria and
the Czech Lands) in 1194, which might have affected Hungary as well.
c. Rainy summers?
A rainy summer may have affected (Western)Hungary, but even more
probably the upper (alpine) catchment area of the main Western
Hungarian rivers in 1043. A rainy summer could also have affected the
same area in 1051, before and during the military campaigns of the
German emperor in Hungary. If we accept the hypothesis of the collapse
of a loess wall in 1092, then this event, amongst others, had to be
36
preceded by a prolonged wet period (months) and high/changing water
levels of the Danube.
2) Evidence indirectly related to weather conditions
a. Locust invasion
Probably an indirect consequence of the heat and drought of the mid-
1190s was the locust invasion clearly reported for the year of 1195, but
locusts might have already been there in 1194 (and in 1193), and not
just in the Drava–Sava interfluve, but also in other parts of the country.
b. Floods
In 1043, inundations of the Rábca are mentioned. In 1051, the flood
conditions of the Zala and maybe those of the Rába are described, even
if one admits the possibilty of human intervention. In 1092, also the
Danube was mentioned to be in flood (maybe also partly due to the fall
of a hill; see 2/c). A (Lower) Danube flood event might have also
occurred in 1154 or 1155.125
3) Individual weather events reported
a. A veil of dust (windstorm), arriving from an easterly direction, was
described in connection with the battle of Ménfő that took place in 1044,
which clearly suggests a prevailing dry spell in the area. This
phenomenon might have had an impact on the battle, because the storm
prevented Aba’s soldiers from applying the eastern fighting tactics.
b. Dense fog hindered the armies from commencing fighting in 1074 at
Mogyoród; the battle took place the next day, when weather conditions
did not obstruct the fight.
c. According to some source evidence, lightning killed King Aba after he
had fled from the battle of Ménfő in 1044. Other sources, however,
claimed that he was killed by his own people. In 1081, the destruction of
the basilica of Székesfehérvár was mostly attributed to some other
lightening event.
d. The effects of strong easterly winds were emphasised during the siege
of Belgrade, which occurred in autumn 1071.
e. An atmospheric phenomenon (halo) might have been observed; namely,
two suns were seen on the sky in 1022 (or 1020, 1021).
4) Events indirectly connected to weather conditions, which directly affected the
human population
a. Hunger and famine
125 Concerning medieval Danube floods (in the Carpathian Basin) and source analysis
problems, see also: A. Kiss, “Dunai árvizek Magyarországon a középkori írott források
tükrében: 1000-1500. Esettanulmányok, forráskritika és elemzési problémák” (Danube
floods in Hungary in medieval documentary evidence: 1000-1500. Case studies, source
critics and analysis problems). In: A. P. Kiss, F. Piti and Gy. Szabados (eds.),
Középkortörténeti tanulmányok 7 (Research in medieval studies 7) (Szeged: Középkorász
Műhely, 2012), 339-55.
37
A famine in Hungary was reported for the years 1044, 1074 (or 1075),
and 1141. All these three years coincide with periods of military
conflicts. However, there were many other years, when wars occurred
but no report of any hunger or famine survived. Moreover, since in
these cases the domestic evidence (compiled in the 14th century)
mentioned the events as well, it seems less likely that these are all cases
where foreign authors would have been inspired by the impressions of
foreign (mainly German) armies passing through (Western) Hungary. It
is also an interesting fact that both in 1044 and 1141 there were also
famines in Bohemia reported for the years before (1043) or after (1142).
In both cases, wars were blamed for a bad harvest that led to severe
food shortages. For 1043, however, it is clear that in the Czech lands
bad (wet, cool) summer conditions and a generally bad year for grain
harvest could also have caused severe food shortages.
b. The cause of the fire which occurred in Pécs in mid-April 1064 at the
night after the coronation is unknown, but one cannot exclude the
possibility of human intervention. The fire of 1064 has some similarities
with another one that occurred in 1071 when, as it was reported,
during the siege of Belgrade the fire was fanned by strong easterly
winds.
* * *
As could be shown, reports for 11th- and 12th-century weather conditions
and weather-related environmental phenomena are only available in some cases.
While sometimes rather clear and unspecific evidence is given, especially
concerning the hard character of the winter or a (partly) wet summer, many of
the other reported events describe very specific weather phenomena. In some
cases the reports refer to events that strongly affected the human population
(e.g., hunger, fires, locusts). Whereas many of the events refer to specific
locations, some cases (indirectly) hold clear large-scale weather-related signals
(e.g., river floods or locust invasions). As to be shown in the next part of the
series, in which weather events and weather-related phenomena reported for the
13th century are discussed, the number of cases will significantly increase, but
the reports will be related to phenomena similar to the ones reported in the 11th
and 12th centuries.
Acknowledgements: The author is grateful for the useful suggestions of György Szabados
(Hungarian Medieval Studies Research Group, Hungarian Academy of Sciences, Szeged,
Hungary) and Éva Révész (Department of Medieval and Early Modern Hungarian History,
University of Szeged, Hungary); to the Rachel Carson Center for Environment and Society
(LMU, Munich) for their support; and to David Curley (University of Szeged) for
proofreading of this article. The paper has been completed in the framework of the „Flood
Change“ ERC project (TU Vienna, Austria).
M E D I U M A E V U M
Q U O T I D I A N U M
66
KREMS 2013