101This poorly supported – or perhaps dogmatic – explana-
tion will be defended again by Morin in 1888 and later by
Giraud in 1909 and will spread confusion among the
naturalists who were not geologists at this time (e.g. Bruyant
in 1894). This will not escape the mind of those who had
gained enough fi eld experience like Vimont ( 1874 ) who, in
his 1889 publication, strongly opposed Julien’s hypothesis
and defended the “ crater-lake ” nomenclature (Vimont 1889 ).
However this author, who neglected the observations
made by Lecoq on Pavin’s ejecta, confused them with the
trachytic tuffs of the Monts-Dore and invented his own radi-
cal explanation to explain their absence:
“ Un trait caractéristique et fréquent chez les grands cratères
comme celui de Pavin est l’absence complète ou à peu près com-
plète du bourrelet saillant, ou paroi circulaire formant relief au-
dessus du sol préexistant qui constitue le cône volcanique, et qui
est dû à l’accumulation des déblais produits par le creusement
du cratère et rejetés circulairement tout autour par la force de
projection des gaz et des vapeurs. L’absence de ces déblais est
ici d’autant plus réelle qu’on ne voit nulle part aux alentours de
fragments de gneiss, contrairement à ce qui devrait être, Pavin
s’enfonçant, comme nous venons de le démontrer, dans cette
dernière roche. Voici l’explication probable de ce fait. Les
cratères de ce genre paraissent avoir été formés par une série
d’explosions violentes et répétées qui ont successivement brisé
les roches et les ont broyées en particules assez ténues et d’assez
faible poids pour que la presque-totalité ait dû être projetée et
dispersée au loin, ou même entraînée par les vents, semblables
en cela à ces cendres volcaniques qui, lors de certaines érup-
tions, sont transportées à d’immenses distances. / A characteris-
tic and frequent feature associated to the big craters as that of
Pavin is the more or less complete absence of the prominent
ridge or circular wall forming relief above pre-existing ground
level which constitutes the volcanic cone, and which is due to the
accumulation of the debris produced by the excavation of the
crater and ejected circularly around by the strength of projec-
tion of gases and vapors. The absence of these debris is all the
more real here as nowhere do we see fragments of gneiss, con-
trary to what should be, if Pavin were sinking, as we have just
demonstrated it, in this rock. Here is the likely explanation for
this fact. This kind of crater appears to have been formed by a
series of violent and repeated explosions which successively
broke rocks and crushed them in rather tiny particles and of
rather weak weight so that it was almost entirely thrown out and
dispersed away, or even entrained by winds, in a similar way to
this volcanic ash which, during some eruptions, is transported at
great distances. ” (Vimont 1874 , p. 8).
Finally, the century ends with a note written by Boule
who, using the bathymetric sections of Delebecque later pub-
lished (Delebecque 1898 ), concluded abruptly the debate:
“ b. – Cratères d’explosion ou d’effondrement. – Ceux-ci sont les
plus profonds; leurs parois latérales sont très escarpées; le fond
est constitué par un grand espace plat. L’auteur les regarde
comme produits par un effondrement et il critique l’expression
de cratères d’explosion qu’on leur applique ordinairement. Tels
sont le lac d’Issarlès (Ardèche) dont la profondeur atteint 108
m50, le lac Pavin (92 m), le lac Chauvet (63 m), le gour de
Tazanat (66 m50), dans le Puy-de-Dôme. / b. – craters of explo-
sion or collapse. – The latter are the deepest; their side walls are
very steep; its bottom is a large fl at space. The author consid-ered them as resulting from a collapse and he criticized the
expression of craters of explosion which one usually applies to
them. Such are the lake of Issarlès (Ardèche) the depth of which
reaches 108.50 m, the lake Pavin (92 m), the lake Chauvet (63
m), the gour of Tazanat (66.50 m) in Puy-de-Dôme ”. (Boule
1896 , p. 759).5.4 The Twentieth Century: More
Scientifi c Rigor and DataThe XIXth ends as no true professional geologist studied in
detail the issue of the lake Pavin and all the explanations are
equally dismissed. The XXth sees Philippe Glangeaud’s
arrival who returned to geological mapping at scale 1:80,000
of Auvergne, in particular its volcanic areas. Thanks to its
fi eld notes, rich in maps and cross sections, we can follow
the progress of his very rigorous work (Bibliothèque
Clermont Université 2011 ). This is how we learn that at the
beginning of June 1911, he visited the area of the lake Pavin
for the fi rst time while he bore in mind the interpretation of
Julien and Giraud, his colleagues at the university, on the
presence of glacial formations. Actually, he noticed the exis-
tence, everywhere around the lake, of a deposit containing a
wide range of different rocks from the Monts-Dore and the
basement for which he admitted the glacial origin back then.
However, making the detailed stratigraphy of the lake edges,
he rejected fl atly the hypothesis of a dam and supported
instead that of a crater of explosion opened “cut-die” in the
preexisting formations. But he was not entirely convinced by
his own conclusions so that he wrote “ this needs to be
verifi ed ”on one page of his fi eld notes.
He returned to the site in August 1914 and in this oppor-
tunity, his observations left no doubt: the formations hitherto
considered as glacial are volcanic projections (Fig. 5.6 ). He
thus arrived at the same conclusions as Lecoq , but with more
precise evidence.
These results will quickly be published (Glangeaud 1916 ;
Fig. 6.2 in Leyrit and al, chapter 6) and will be defi nitely
accepted. However, one mistake persisted! Philippe
Glangeaud did not know how to recognize the lava at the ori-
gin of the eruption of Pavin because, just as he did at Tazenat,
he was still looking for bombs, ash and projections of basaltic
lava. As a fi rst step, his cross sections show that he drew a
pipe fi lled with basalt at the bottom of Pavin. But very quickly
he would explain Pavin as the result of multiple purely gas-
eous explosions (which was actually the original hypothesis
submitted by Montlosier!) because he did not observe magma
associated with the projections (Roubille 1916 ).
No new study documented the eruption of the lake Pavin
after 1916 to a point that the observations made by
P. Glangeaud will be “forgotten” by the defenders of the
“ collapse caldeiras ” (Bout and Brousse 1969 ; Brousse
1969 ). Nevertheless, the geochronological studies of both5 The Origin of Lake Pavin: Two Centuries of Controversy