considerable channels were in existence. One of these, called the Tulya Bhága, led off from near the site of the anicut and ran in a fairly straight line to Cocanada, terminating in the salt creek there. In 1846 a branch was taken from it to Samalkot from near Dowlaishweram. These two channels were connected with the head-works of the eastern delta.
At the end of 1849 a start was made with the new distributary works, sanction being obtained to the cutting of the main canals in the eastern and central deltas, the first of which (see the map) leads along the river bank nearly as far as Yanam and the second runs past Ráli. In April 1851 the western delta main canal (now in the Kistna district) was sanctioned, and in February 1852 considerable extensions of the eastern main canal and large distributary works in the central delta, including the great Gannavaram aqueduct, were agreed to.
This aqueduct carries a large canal across a branch of the Gódávari to the Nagaram island, which is surrounded by the sea and two arms of the river and to which water can only be taken in this way. The aqueduct may be roughly described as an arched bridge of brick thrown across the branch of the river, upon which, in the place where the roadway of an ordinary bridge is laid, runs a channel from 22 to 24 feet broad and some four feet deep. Its total length between abutments is 2,248 feet, and it consists of 49 arches with 40 feet waterway and 48 piers 6 feet thick. Ordinarily, the water of the branch of the river across which it is thrown flows through the arches of the aqueduct, but in times of high flood it completely submerges the whole work and pours over the top of it. It was impossible to make the aqueduct higher, because of the expense and danger involved in raising the embankments of the channels connected with it to a corresponding height above the level of the surrounding country. The work had therefore to be made of sufficient strength to resist floods sweeping over it.
The most noteworthy fact about the work is the wonderfully short period within which it was built. The estimate was submitted by Sir Arthur Cotton in August 1851 but was not sanctioned till February 14th 1852. It was considered of paramount importance to finish the work before the floods of that year came down, and, to effect this, extraordinary efforts were necessary. Between the first preparation of the materials for the work and the completion of all its 49 arches only four months elapsed, and in another four it was ready for its work. 'In any part of the world,' says Mr. Walch in his book already cited,' this would have been a noteworthy achievement; in an
