THE TWO WAYS 37gardening, early-rising uncle' had lapsed; and it is one of Aunt
Leonie's most terrible nightmares that 'her poor Octave', who is
long dead, should tum out to be still alive and insist on her taking
a walk every day.l
It was necessary also that Comb ray and its chapter should be
the whole novel in miniature, and contain the germs of all its
themes and events. For the rest of the novel the Narrator follows
up the ways on which he first set foot at Combray: however far
he seems to be leaving it behind, he is really circling back, and
Time Regained is also Comb ray regained. Characters who belong
chiefly, as did their originals in real life exclusively, to Paris,
appear also in Combray. Sometimes, as in the case of Swann, the
Duchesse de Guermantes or Gilberte, their connection with
Comb ray is aided by the existence at IIIiers of persons in whom
Proust afterwards saw analogies with the corresponding person in
Paris: in each place there was a family friend, <In unapproachable
noblewoman, a little girl he loved. Sometimes he was helped by
coincidences of history and geography: he named the Narrator's
friend, Robert de Saint-Loup-en-Bray, from Saint-Loup-de-Naud
in Seine-et-Marne, whose church he visited with a group of young
noblemen who collectively suggested his hero; but there is also a
village of Saint-Loup eight miles east of IIIiers. He made Saint-
Loup marry Gilberte partly because one of the originals of Saint-
Loup in real life married one of the originals of Gilberte; but the
name was already linked with TansonvilIe, where the chateau
was occupied in 1710 by a certain Robert de Durcet after his
marriage with Claire de Saint-Loup. As for the suffix en-Bray, no
doubt it comes from Bourg-en-Bray near Saint-Loup-de-Naud;
but there is a River Braye which flows into the Loir some fifty
miles downstream from IIIiers.
Sometimes characters seen at Combray, such as Charlus,
Odette or Legrandin, have originals with no possible association
with IIIiers. Perhaps this is true of Vinteuil: lIliers was large
enough to possess a music-teacher, but there is no published
record of his existence. Legrandin can be identified with a person
who resembled him in every way, except that he had no link with
IIIiers and was a doctor instead of an engineer. Dr Henri Cazalis
(1840-1909) was a professional friend of Dr Proust; but he was
also, under the pseudonym of Jean Lahor, a symbolist poet of
1 I, 110