310 THE INFLUENCE OF SHAPE OF 



draught of the ship is sufficient, but at the hghter draughts wave-making is set up and inter- 

 feres with the normal resistance of the ship. Before leaving Plate 56 it should be 

 pointed out that for the different curves there is a dififerent prismatic coefficient except at 



-^ = 2.26rf = .435. 



Passing to Plates 58, 59, 60 and 61, these will probably be considered the most valuable 

 part oi this very valuable contribution. They represent an enormous amount of work and 

 provide the greatest assistance to the designers of full-form ships, as they enable the de- 

 signer at once to select the fullness of model satisfactory for any given speed at any given 

 draught and provide him with resistance figures corresponding to that model. As for Plates 

 56 and 57, tlie prismatic coefficient varies throughout this series, and therefore the resist- 

 ance curves are composite ones representing an optimum model with constantly varying 

 fullness to suit its particular speed. 



The speaker is fairly familiar with the various tank tests made public in this country and 



elsewhere, but I must confess that I feel very much at sea in regard to these new methods 



of presenting tank tests. Doctor Sadler and Professor Bragg have used two methods in 



E.H.P. 

 these sets, Plate 55 being in the form of ' . ' ' curves, and the other diagrams use a similar 



form divided by the cube of the speed ratio, a new form which is supposed to assist in the 

 ready application of the results to ordinary ship work. 



The authors must be aware, however, that anyone who makes a serious study of the 

 correct form for a ship must take into consideration not only this particular paper but all pre- 

 ceding contributions to this subject to form a proper judgment of the value of these curves, 

 and must have a good basis of comparison with other work. Unfortunately the conversion 

 of the results of this paper to any other standard method of presentation is going to be quite 

 a serious matter. For C curves at 425 feet the ordinates would be multiplied by 



for these models |. 



427.1 \ 20.5 / 



You will notice that the resistances have been presented for a vessel 425 feet long, 

 whereas generally in other papers these had been presented for a standard ship 400 feet 

 long. The model resistances have been reduced to a constant temperature of 65° (the first 

 time which Doctor Sadler has used a temperature correction), and this temperature must 

 therefore be corrected to a temperature of 70° to make the results comparable with Wash- 

 ington, or a temperature of 55° to compare with the Teddington tank. I think it is a great 

 pity that the authors did not adhere to the 70° which is used in Washington. This tempera- 

 ture was also used in the paper last year by myself, and as the Ann Arbor tank temperature 

 frequently runs as high or even above it, no serious error could possibly have been involved 

 in adhering to this figure until such time as the important authorities on ship resistance de- 

 cide upon a standard method of presentation. 



The standard model used by Doctor Sadler evidently lies between model E.W. and the 

 model F.W. in the paper of last year presented by the speaker, showing an excellent s /p ratio 

 of 0.65, and the results of varying parallel middle bodies have been presented on a percen- 

 tage of parallel body. 



The authors must surely be aware that such a form is very misleading if any changes 

 are made in the curves of areas forward or aft. For that reason the speaker last year sug- 

 gested the use of a standardized method of measuring entrance and run designated e / p. The 



