Testing of Boat-Motors at the Northern Fishery Exhibition. GUSTAF SETH.

(Teknisk Tidskrift, Stockholm, 1905, Mech. and Elec. section, pp. 1-3.)

The chief object of this Exhibition, held in 1904 at Marstrand, Sweden, was to promote the interests of the fishing industry, by a competition among the leading makers in Sweden, Norway and Denmark, of motors suitable for the propulsion of fishing boats. Representatives of the three countries were appointed a committee, which included the Author. Their scheme for testing the competing motors comprised: (1) a brake test on shore of not less than 1 hour; (2) a continuous test at sea of 12 hours without stopping; (3) a speed trial at sea of not less than 1 hour, and connected therewith a manoeuvring test. Every motor exhibited, therefore, had to be shown in duplicate, for the trials ashore and afloat. The speed trial was made both ways, out and back, over a measured course of 1 nautical mile. The results are tabulated of brake tests on ten petroleum motors and three benzine motors, and of speed trials with sixteen boats; the dimensions and other particulars are given, both of motors and propellers, and of boats. The consumption per effective horsepower-hour ranged from 0.7 to 0·273 kilogram (1·543 to 0·602 lb.) of petroleum, and from 0.964 to 0.722 kilogram (2.125 lbs. to 1.592 lb.) of benzine. The average speed in the double trip ranged from 4 to 6.46 knots per hour in twelve petroleum boats, and from 5.08 to 6.28 knots in four benzine boats. The manoeuvring test included the time spent in warming the cold motor until it could drive the boat at full speed ahead; and also the ease of altering her course, bringing her alongside the quay, and getting her clear away therefrom. Besides economy of fuel, the jury had also to report upon simplicity, strength, and durability of construction, excellence of workmanship, freedom from smell and smoke, ability to continue running for a long time without stopping, and also power to go astern as well as ahead. A protest is offered by the Author against statements of "indicated" horse-power, which are not only apt to be misunderstood as meaning "effective," but are also inappropriate for explosion motors, as the power therein developed seldom continues the same for many strokes together, while the cooling of the cylinders further affects indicator-diagrams considerably. Brake trials of such small engines present no particular difficulties, and can readily be carried out with accuracy.

A. B.

Influence of Depth of Water on the Speed of Torpedo Craft.1

A. C.

(Le Génie Civil, l'aris, vol. xlvi. pp. 304-305.)

The comparative proximity of the bottom, in addition to rendering steering difficult, exercises great influence on the speed which can be got from a vessel with a given expenditure of power. Captain Rasmussen, of the Danish marine, published in 1894 a dissertation on the subject, having experimented with a vessel of a length of 42 metres (138 feet), a beam of 4 metres (13 feet) 2 inches), and a draught of 2 metres (6 feet 7 inches). Major Rota, an Italian, and Mr. Schütte, a German engineer, have since taken up the question, using for their experiments scale models. With a view of testing the correctness of the formulas thus obtained exhaustive experiments were undertaken with a German torpedoboat, during perfectly still weather, in September, 1903. In these experiments runs over a measured distance were taken in all depths of water between 7 metres and 60 metres (23 feet and 197 feet).

Every precaution was taken to eliminate errors due to extraneous causes, such as wind or currents. During each trip the revolutions of the engines were counted, cards were taken from each of the six cylinders, the mean speed and list of the vessel were noted, and also a profile of the surface of the water alongside was measured. Diagrams are given showing the results, and from these the following particulars are extracted.

At all speeds up to 13 knots per hour the power required to develop a given speed was the same in 7 metres (23 feet) as in 60 metres (197 feet). From 13 to 20 knots per hour the greater the depth the less was the power required, while, when the speed exceeded 20 knots, there appeared to be a critical depth, varying from 15 metres (49 feet) when the speed was 21 knots to 25 metres (82 feet) when the speed was 27 knots, when the power required to drive the vessel was a maximum. This singular fact seems to agree with the results obtained from the experiments of Messrs. Rasmussen, Rota and Schütte. Mr. Paulus, who carried out the above tests, has calculated from the formulas of Mr. Rota the results which should be obtained from his vessel, and has found them to agree singularly closely with the results of the actual tests.

I. C. B.

The subject treated in this abstract has been discussed in various Papers contributed by Captain Rasmussen, Major Rota, Mr. Harold Yarrow and Mr. W. W. Marriner to the Transactions of the Institution of Naval Architects (vol. xli. p. 18; vol. xlii. p. 239; and vol. xlvii. pp. 339 and 344). The two latter Papers, which appeared in July, 1905, summarize the results obtained up to that time by various experimenters, and add much information in regard to the critical depths of water appropriate to various speeds of ships.

Disregarded Points in the Stability of Masonry Dams.

L. W. ATCHERLEY, Stud. Inst. C.E., with some assistance from KARL PEARSON, F.R.S.

The Author gives the equations (1) connecting the breadth of the dam at any horizontal section with the distances of the load point and neutral axis from the centroid of the section; (2) the equations for maximum vertical compression and tension due to the total vertical thrust on the horizontal section; and (3) the formula for shearing stress on the section, on the assumption that it is distributed in the same way as in the cross-sections of a beam, i.e., that the shearing stress varies as the ordinates of a parabola the axis of which passes through the centroid of cross-section. He next considers the general equations for stresses in an isotropic and homogeneous dam, but without attempting to solve them, the object of the investigation being not to inquire into the validity of the usual treatment, but to try to indicate, supposing it to be correct, that its present partial application to horizontal crosssections only, involves the serious neglect of large tensions across the vertical sections.

To justify this statement two model dams were prepared, made of a fairly heavy wood cut to the profiles which had been used in an actual construction. In one case the model was divided into a series of horizontal strata and in the second case into a series of vertical strata. In the first case the pull representing the water-pressure was communicated by a cord to a stiff lath which bore on the ends of the horizontal strata through two longitudinal strips of india-rubber tube; the attachment of the cord being onethird up the lath and the cord adjusted to pull in a direction perpendicular to the front of the model, so that the resultant force was applied at a point corresponding to the centre of pressure of the

water.

In the case of the model stratified vertically, much the same arrangement was adopted, except that the pull of the cord was applied directly to the first vertical stratum, which included the battered front.

The angle of friction of the wood strips on each other varied from 25° to 30°, and a shearing strength more nearly corresponding to masonry was obtained by pasting tissue paper round the battered fronts and curved flanks of the models.

In the case of the actual dam on which these were modelled the weight of a strip of the dam 1 foot broad is 505,000 lbs. and the corresponding water-pressures 312,500 lbs. The weights of the two models, although they were of the same dimensions, were not identical, that of the horizontally stratified one being 12.40 lbs., and that of the vertically stratified one being 13.85 lbs. The corresponding values for the pulls to represent the waterpressures would be 7.70 lbs. and 8.57 lbs. respectively. Trusting

merely to the friction of the wood on wood, the horizontally stratified model slid on its base at 5.70 lbs. pressure, and the vertically stratified one opened up at the third section from the tail, and then the whole thing sheared with a pressure of only 3.00 lbs. When the models were strengthened for shearing resistance by tissue paper, as referred to above, the respective pulls corresponding to the water-pressure before collapse were 6.50 lbs. and 4.20 lbs.

The general results of these experiments were conclusive and obvious. They are as follows:

(a) The current idea that critical sections of a dam are the horizontal sections is entirely erroneous. A dam collapses first by the tension on the vertical sections of the tail.

(b) The shearing of the vertical sections over each other follows immediately on this opening up by tension.

(c) It is probable that the shear on the horizontal sections is also a far more important matter than was supposed.

The Authors lay it down as a rule for the construction of future dams that the stability of the dam from the standpoint of the vertical sections must be considered in the first place.

This is followed by a combined graphical and analytical treatment of the dam from which the models were taken, on the basis of the equations referred to in the first paragraph; from which the Authors arrive at the following conclusions:

(a) That the line of resistance for the vertical sections lies outside the middle third for more than half the width of the dam.

(b) That the tensile stresses in the tail are, for masonry, very serious, and amount to nearly 10 tons per square foot at the extreme tip and to 6 tons per square foot after passing the vertical section where the strengthening of the tail has ceased.

(c) That the maximum shearing stress amounts to 6 tons per square foot at the tip of the tail and to 5 tons per square foot after passing the vertical section where the strengthening of the tail ceases.

Even if, instead of a parabolic distribution, a uniform distribution of shearing stress be assumed, they still find:

(a) That the line of resistance falls well outside the middle third for about one-half the dam.

(b) That there exist considerable tensions in the masonry amounting to 3 to 4 tons per square foot.

(c) That the average shearing stresses on the vertical sections are greater than at the horizontal sections.

An Appendix is added giving Mr. Lévy's solutions of the damproblem.

A. W. B.

Resistance of Metallic Wires for High-Frequency Currents. BROCA and TURCHINI.

(Comptes Rendus de l'Académie des Sciences, Paris, 1905, vol. cxl. pp. 1238-41.)

The resistance of metallic wires of circular section for alternating currents is usually found by applying the well-known formula of Lord Kelvin. The Authors have undertaken experiments to test the applicability of this formula when very high frequencies, ranging from 142,000 to 3,800,000 per second, are used. The method of experiment is to measure the effective intensity of the high-frequency current at each instant by the deflection of the electrodynamometer, and the heating produced in the wire by the same current. The same heating is then produced by a continuous current, and the corresponding deflection of the dynamometer for this current is measured. Equations are thus obtained which reduce to a simple equality between the ratio of the deflections and the ratio of the resistances of the wire in the two conditions of experiment, and the value of this latter ratio may then be compared with that given by the Kelvin formula.

The results obtained show that for non-magnetic materials (copper and platinum) the variations from the Kelvin law are but small in the case of moderate frequencies, but they are greater than the errors of experiment, and follow a perfectly determinate law. The ratio between the value calculated by the Kelvin law and the value measured as described above is a function of the variable in the Kelvin formula. The results obtained for iron, on

the other hand, have nothing in common with those given by the formula. The measured values depend essentially on the intensity of the current. German silver gives results analogous to those of iron, but much less pronounced.

W. C. H.

Methods of Charging for Electric Current. B. BRONISLAWSKI. (L'Éclairage Électrique, Paris, 1905, vol. xliii. pp. 5–9.)

The Author discusses the principles underlying the supply of electric current in their relation to the determination of the proper charge per unit to be made to the consumer. Dealing with an actual French central station of 2,500 HP. in prime movers, in a town of 100,000 inhabitants, he shows how the working expenditure is divided among the different items, and discusses the extent to which new consumers are a source of profit and the point at which they become unprofitable. His conclusion is that to utilize in the best manner the potential capacity of the station, the price charged to the consumer should take into account the daily duration of his demand and the ratio between the power of his installation and his average demand. The price per unit sold