Prior to the experiments of Mr. Russell, J. Robison, Esq., afterwards Sir J. Robison, had instituted a series of experiments on the Forth and Clyde canal, with the view of ascertaining the best form of canal boats. Of these experiments (to which those of Mr. Russell may be considered as a sequel, as they were, we believe, undertaken at the suggestion, and made chiefly at the expense, of Mr. Robison) he has furnished an interesting account, in a communication to the Society of Aris, which appeared in the Transactions of the Society for 1833, from which we make the following extract: -

"Four models were prepared of the following dimensions -

And the weight of each 187 1/2 lbs.

"No. 1 was quite flat on the floor, rounded at the bilges, and perpendicular in the sides at the midships section, but with a fine entrance and run.

"No. 2 was made in the proportions of an ordinary coasting trader.

" No. 3 in the proportions of a sharp built schooner.

"No. 4 was a twin-boat, similar in its dimensions to No. 1, only that the breadth of each portion was half of the other breadth: the depth was the same.

"The weight of all the models being alike, their displacement of water was equal, although their draft or depth of immersion was necessarily different.

"The usual way of trying the resistance of floating bodies is by drawing them across a dock or basin by a cord, running over delicately hung pullies, on a high mast, and with certain weights attached; the time is accurately noted which each form requires to move through a certain space, and the comparative resistances are calculated from these elements. This method presents many difficulties and disadvantages. I was enabled, however, by a suggestion from an ingenious friend (Mr. Oldham, of the Bank of Ireland), to adopt a much more summary and satisfactory way of determining the comparative resistance of the different models; and, as it was the comparative resistance alone which required investigation, I had no occasion to go through the more tedious process of trying the resistance only, and of incurring the risk of error from mistake in reading off the indications of the dynamometer. I prepared, accordingly, a spar, or yoke, sixteen feet eight inches long, which was divided into 100 parts of two inches each: a small eye-bolt was fixed at each extremity, and a shifting hasp was fitted to the middle part.

With this yoke all the experiments were made, by the two following processes: - First, a model was attached, by a slender tow-line, to each eye-bolt, and the hasp was fixed exactly in the middle of the yoke, and linked to an outrigger on the steam vessel, which was then set in motion at the required speed. If it was found that one of the models preceded the other in consequence of its offering less resistance, the hasp was shifted along the spar towards the sluggish one, until the resistances were balanced, and the two models ran abreast of one another. The relative lengths of the arms of the yoke then gave an inverse measure of the comparative resistances of the models at that rate of speed; this being noted down, the hasp was brought again to the middle of the yoke, and the model which showed least resistance was, by degrees, loaded with weights, until it again exactly balanced the other, and swam abreast of it; the amount of the added weights being likewise noted, afforded a second measure of the difference of the resistance of the two models.

"Each of these forms of the experiment was gone through with different pairs of the models, and was frequently repeated through long spaces of the canal, as it was found that various circumstances interfered to render the resistances inconstant, such as approaching nearer to the one or other side of the canal, passing a loaded vessel, or making a turn round a projecting part of the bank. It was first attempted to conduct the experiments by towing the models astern; but it was immediately found that the ripple of the wake of the steamer disturbed the uniformity of the resistance of the models. Various modifications were then tried, with more satisfactory results, and finally the arrangement was made as follows: - A spar like a bowsprit, about twenty feet in length, was run out a little above the level of the water from the bow of the steamer, the hasp of the yoke being attached by a link to the point of the spar, the models were in this way kept ahead of the steamer in smooth water, and were altogether undisturbed by any ripple or wave."

The accompanying tables contain the results of these trials, from which the important inference may be drawn, that there is no form which will present a minimum resistance in all circumstances; and that the form which is easiest drawn through the canal at a low velocity, does not possess the same advantages at a higher rate of speed.

By looking into the table A, experiment 1st, we see that, although the resistance of No. 1 be to that of No. 2 as 13 to 12 when the velocity is three miles per hour, yet, when the speed is increased to six miles, the advantage which No. 2 had over the flatter vessel entirely disappears.

Again, in Table B, we see that in one experiment No. 2 carries two-ninths more weight than No. 1 with equal resistance when the velocity is three miles per hour; but that, when the rate is raised to six miles, the loads required to be the same in both, in order to equalize the resistance.

It appears, from numerous experiments made at intermediate speeds, that this change in the relative resistance is progressive: there is reason, therefore, to conclude, that if circumstances had admitted of carrying on the experiments at a higher velocity than six miles per hour, the flatter formed vessel would have attained a superiority over the sharper ones. This conclusion is corroborated by the fact that the swiftest going steam vessels which have been built in this country, are those which are nearly quite flat in the floor for a great proportion of their whole length.

The first practical inference which may be drawn from these experiments is, that all vessels which are intended to be tracked, or impelled by machinery through canals, at low velocities, should be built as sharp in their bottoms as circumstances will admit of, although this must necessarily increase their draught of water; the second inference is, that whenever vessels are intended to move in canals with a higher rate of speed than six miles per hour, the general form of the bottom should be quite flat.