heavy, and
works of construction sufficiently substantial, to support it.
Moreover, where streams or rivers are encountered they must be
bridged. In short, the subsidiary lines must be built in a manner
which would make them nearly as expensive as the main lines; in other
words, railways must not be introduced into any part of India where we
cannot afford to spend from 13,000_l_. to 15,000_l_. a mile upon them.
I am not prepared to accept this conclusion. I have been a good deal
in America, and I know that our practical cousins there do not refuse
to avail themselves of advantages within their reach, by grasping at
those which are beyond it. In 1854, I travelled by railway from New
York to Washington. We had several ferries to cross on the way, but we
found that the railway with the ferries was much better than no
Railway at all. In short, in America where they cannot get a _pucka_
railway, they take a _kutcha_ one instead. This, I think, is what we
must do in India. There are many districts where railways costing
3,000_l_. or 4,000_l_. a mile might be introduced with advantage,
although they would not justify an expenditure of from 10,000_l_. to
15,000_l_. a mile. We have only to be careful that _kutcha_ lines are
not mistaken for _pucka_ ones--that they are not allowed to set up a
rival system as against the main lines, or to occupy ground which
should be appropriated by the latter.

[Sidenote: Carriage dak to Allahabad.]

As the railway from Benares to Allahabad was not yet complete, Lord Elgin
and his suite performed this part of the journey by carriage dak. They
travelled by night; 'each individual of the party occupying his own
separate carriage, and being conveyed along at a hand gallop by a
succession of single ponies, relayed at stages of four to five miles in
length.' In the letter which describes this, he adds the characteristic
remark:

'These ponies do not lead very happy lives, and, here as elsewhere, a
diminution in the sufferings of the brute creation will be one of the
blessings attending the introduction of a railway system.'

At Allahabad he inspected, among other things, the works which were in
progress for making a railway bridge across the Jumna.

This is (he wrote) in some respects the most interesting of that class
of engineering operations which has been already mentioned: because
whe