When the shaft or tower of brickwork was completed up to the top, 42 feet in height, the next step was to remove the blockings on which it rested, and this being done the gravel was excavated and hoisted up, and the shaft descended by its own weight. The Rotherhithe shaft was only sunk forty feet in this manner; the remaining twenty feet, in order to leave the opening for the Tunnel, was constructed by under-pinning, or underlaying, as it was then termed. The underlaying was commenced in the beginning of June.[10] By July 4 they had got down to the level of the intended foundation of the shaft, having passed into a stratum of gravel, black pebbles embedded in greenish sand, with little or no water; from which circumstance Sir Isambard was of opinion that it was unconnected with the stratum of gravel above. _July 12._--Engaged on a general drawing for the great shield, and in preparing some instructions for moving the same (a very intricate operation!) _July 22._--Underlaying is a very laborious mode of proceeding. The sinking of a wall well bound as the first, would evidently be the best and cheapest mode for making another tower of 50 feet diameter. On the 28th Sir Isambard enters in his journal the following additional observations upon the success of his plans for sinking the shaft:-- Considering the great labour necessary for securing the ground for the underlaying, the waste of planking, and of shores, and the time necessarily taken up in moving about, in securing and in baling out the water, and the many causes of interruption, and the imperfect way that things are done in underlaying, it is quite conclusive that the original plan of making a shaft, by sinking the structure, is the safest and the most economical. What is done is sound, and when once in place, may be secured with foundations in a very easy manner. The brickwork of the shaft is remarkably hard. Had it been made with brick facings and rubble stone, it would certainly be water-tight, and almost impenetrable by ordinary ways. The _vertical ties_ and the _circular wall bands_ are not to be dispensed with in a structure destined to be moved as the present has been. On August 11 the underlaying was completed, and preparations were made for constructing a reservoir in the bottom of the shaft for receiving the permanent pumps. This was finished on October 11, with great difficulty, owing to the nature of the ground, which consisted of loose sand containing a large quantity of water.