326 POWER AND SPEED TRIALS OF 



great this effect is in many cases. By the author's method of dealing in the first place with 

 curves of mean effective pressure plotted on revolutions, the wind effect can be reasonably well 

 eliminated if fairly close observations are taken of it. 



The further step to correct the speed curves for wind effect is not so obvious, but I 

 gather that, if we assume the sHp to be constant, the method is exact in theory which 

 makes it quite close enough for the actual conditions of practice. 



The efficiency of propulsion deduced would be a shade higher if the actual E. H. P. for 

 the trial displacement had been used. The assumption that in the neighborhood of the trial 

 displacement the E. H. P. varies with displacement is near enough for practical purposes, but 

 brings the E. H. P. slightly lower than if model trials had been made corresponding to the 

 actual trial displacement. 



I hope that in the future we may have more such papers in our Transactions, and fur- 

 ther investigations of the wind effect on trials. 



There is another effect, very real but hitherto not subjected to analysis, namely, the 

 effect of the rudder. We know that the rudder when put over constitutes a very powerful 

 brake, and there is a material difference between the speed and power curve with a good 

 helmsman and a poor helmsman. Possibly some of our scientific members may reduce this 

 matter to rule in time. 



The Chairman : — I will ask Professor Everett to close the discussion. 



Professor Everett : — With regard to Professor Bragg's request for additional data, I 

 will be glad to furnish it. The reason why I put a drawing of the propeller in the back of 

 paper is that I thought someone might wish to work up something in that connection. 



The cut-offs are given on Plate 68. The other data I will furnish. I will not be able 

 to furnish exact clearances, but they will be nearly exact, and I will also send in the data 

 requested with reference to the lines. The information in regard to wake is extremely 

 interesting. 



With reference to Mr. Rigg's remarks, may I take a moment to ask him to look at Plate 

 64 again, because the object of the paper was partly to get away from the method that he 

 mentions, namely, the averaging of groups of runs. 



As Admiral Taylor mentions in his discussion, some years ago he proposed two methods 

 of making trials, one by nmning groups of runs under as nearly the same condition as pos- 

 sible and taking the average, and the other, by starting at low speed and going up progres- 

 sively, but keeping the runs in sequence, down and back over the course. It is the latter method 

 that these trials are based on; that is, the trials were run consecutively, gradually increas- 

 ing speed and paying no attention to the averaging of runs. If anyone should prefer to 

 work up the runs, using the other method, the original data are given in Table I. I did it in 

 a casual way, and the result came very little different from the final curve shown on the last 

 page. I think the method used here has merit as being the logical procedure for getting rid 

 of both the wind effect and tidal effect. 



I did not mention, in the brief abstract just read, one point I should like to bring out. 

 On Plate 65, the lower right-hand curve is the E. H. P. taken from the model tank runs. 

 The trial of the ship was made at a displacement which was slightly different from that (370 

 tons less), due to the fact that the trials were nm in fresh water but at the salt water load 

 draught. To correct for the trial displacement, we derived the lower E. H. P. curve by 

 comparison directly on the displacement basis, and the work which Professor Sadler and Pro- 



