ibility and the exactitude of doctrine, and his indignation was
too strong to be pacified. He returned somewhat, I have heard, to his
old beliefs in later years, as old men will to the beliefs of their
youth, and his Christianity was too sincere and profound for a matter
of mistaken credence in mere formalities ever to affect its substance,
and the years which followed showed that in no essential trait had the
religious foundations of his character been moved. For myself, I was
still a sincere believer in the substantial accuracy of the body of
Christian doctrine, and the revolt of Ruskin from it gave me great
pain. My own entire liberation from the burdens of futile beliefs had
yet to come, and at that time he went further than I could go with
him. But we never discussed theological matters any more.

I finally found a subject which interested me in a view of the foot of
the Mer de Glace from the opposite side of the river, looking up the
glacier, with the bridge under the Brevent, and a cottage in the
foreground, and set to work on it energetically. Ruskin used to sit
behind me and comment on my work. My methods of painting were my own,
for I had never painted under any one except the few months with
Church, whose method had taught me nothing; and I had a way of
painting scud clouds, such as always hang around the Alpine peaks, by
brushing the sky in thinly with the sky-blue, and then working into
that, with the brush, the melting clouds, producing the grays I wanted
on the canvas. It imitated the effect of nature logically, as the
pigment imitated the mingling of the vapor with the blue sky; but
Ruskin said this was incorrect, and that the colors must be laid like
mosaic, side by side, in the true tint. Another discouragement! I
used to lay in the whole subject, beginning with the sky, rapidly
and broadly, and, when it was dry, returning to the foreground and
finishing towards the distance; and Buskin was delighted with the
foreground painting, insisting on my doing nothing further to it. In
the distance was the Montanvert and the Aiguille du Dru; but where the
lines of the glacier and the slopes of the mountain at the right met,
five nearly straight lines converged at a point far from the centre,
and I did not see how to get rid of them without violating the
topography. I pointed it out to Ruskin, and he immediately exclaimed:
"Oh, nothing can he done with a subject like that, with five lines
radiating from an unimp