102
authors showed that a deposit of pumice coincided with the
end of volcanic activity in the south of the lake Pavin (Bout
1969 ; Brousse et al 1969 ; Leyrit et al, chapter 6; Juvigné et
Miallier, chapter 8). But Bout and Brousse mistakenly attrib-
uted the pumice deposit to Montcineyre.
Finally, this story ends with the work by Camus et al.
( 1973 ) who reconciled all the existing hypotheses by estab-
lishing a consistent tephrostratigraphy of the eruptions of the
Pavin area, mapping the extent of the deposits of the Pavin
eruption and neighbouring volcanoes, describing the petrogra-
phy of the juvenile magma of the eruption (“ an amphibole-
bearing leucotrachyte ”), and by showing that the magma
produced either pumice or typical dense blocks owing to a
dynamism affected by the arrival of superfi cial water. By iden-
tifying Pavin with a maar, these authors eventually agreed with
Scrope and Dufresnoy. Meanwhile the model of a maar forma-
tion, which implies violent explosions to enlarge the crater by
concentric collapses, was clarifi ed and its demonstration has
been long documented. The results of the most recent research
on the eruption of the lake Pavin which constrain this model,
are presented in the next three following chapters.
5.5 Summary
At the time of the discovery of the Auvergne volcanoes in
1751, the model which was used as the existing paradigm
was a volcano with crater, built with red scoriae – ‘ fi r e
print ’ – and which had emitted lava fl ows. Pavin, with its
large lake occupying a deep circular depression and without
red ejecta or lava fl ow, would not be considered as a vol-
cano – in the absence of any better hypothesis – until 20–30
years later. As soon as this basic premise was accepted, the
mechanism of its formation was immediately the object of
controversies between those who advocated for gaseous
explosions without deposit, and those who defended a large-
scale collapse. Very quickly, the link with the morphology of
the maars of Eifel was established, but this did not offer the
solution to the problem because the formation of the German
maars was not better understood at this time.
No observation allowed for militating in favor of one
hypothesis or another until 1835, the year when deposits of
heterogeneous breccia were described. These deposits were
interpreted as eruptive deposits but in the absence of petro-
graphic analysis no juvenile magma was underlined.
Shortly after, the fi ndings of the action of the glaciers in the
region raised a new hypothesis, namely the damming of
glacial valleys by lava fl ows. Breccias that were previously
described were subsequently likened to moraines. The sys-
tematic geologic mapping of France at the scale 1:80,000
with its rigorous methodology allowed geologists to dis-
card the dammed glacial valley hypothesis and retain only
that of an explosive activity with associated ejecta. But at
that time nobody knew as yet how to describe and highlight
the juvenile magma in this deposit that reworked the old
formations with pumice and trachytes of Monts-Dore. After
a brief, rather confi dential return to the collapse hypothesis
coinciding with the theory of the “collapse calderas”, the
general framework of the modern interpretation was pub-
lished in 1973 with the identifi cation of juvenile magma in
the deposits, mapping of their extent and pinpointing of
their position in the local tephrochronology. The juvenileFig. 5.6 Filed note #2300 of P. Glangeaud , pp 2347–2348 (Bibliothèque
Clermont Université 2011 ). Transcription of the fi rst lines: “ Le cratère-
lac Pavin est bien [illisible] d’un cratère d’explosion car il entame les
formations suivantes : gneiss du substrat, coulées de trachyte, cinérite
à blocs, andésites (coulées) pliocènes monts Dore, moraines glaciaires
et basalte compacte labradoritique. Puis la coulée quaternaire de
Montchal. Bords découpés à l’emporte pièce et dont on trouve des
débris projetés aux alentours en blocs de taille assez considérable,
jusqu’à 1 m 3 , principalement dans les pentes N. / The crater-lake Pavin
is [illegible] of a crater of explosion because it is cut down into the fol-
lowing formations: gneiss of the basement, trachytic lava fl ow, cinerite
with blocks, Pliocene andesites (lava fl ow) Monts-Dore, glacial
moraines and massive labradoric basalt. Then the Quaternary lava
fl ow of Montchal. Edges are die-cut and one fi nds their fragments
thrown in the surroundings as blocks of rather considerable size up to
1 m 3 mainly on north slopes ”P. Boivin and S. Jouhannel