ECONOMICAL CARGO SHIPS—SOME MODEL EXPERIMENTS. 45 
Method of Experiment.—These tank tests were, in almost every case, carried 
out at the University of Michigan experimental tank because models used in that 
tank are of wax, and variations in such models are much more readily accom- 
plished and at a much lower cost than is possible with the larger wood models in 
Washington. At the same time, one single-screw model was tested at Washington 
in order to link up the Ann Arbor and Washington tests; and the twin-screw model, 
with and without bossing, was tested in Washington. 
It should be noted that the method of extension of experimental model tank 
tests at Ann Arbor is based on the coefficients of Tideman both for the model and 
the full-sized ship. These apparently coincide with the results of tests made at 
the tank some time ago on the frictional resistance of plane surfaces. At the 
same time it must be pointed out that Tideman’s formule for skin friction, when 
applied to the full-sized ship, give resistance results considerably higher than those 
attained by the use of the Froude coefficients in the region of speed of cargo ships 
such as are now under consideration. 
Method of Presentation—All the results are presented herewith in the form 
of C constant curves for a length of 400 feet, wherein— 
ca 271 Ns alles 
eC 
and S 
EH pack VX. 
427.1 
This is somewhat of an innovation on this side of the Atlantic, but the advan- 
tages of this form of presentation appear to the writer to be so important that he 
considers it well worth while to draw the attention of the Society to this system. 
It should be noted, however, that the speed ratio aa used in this paper is the one 
familiar to our designers, where V equals the speed in knots and L is equal to the 
actual displacement length of the ship. No use has been made of the speed con- 
stant P used largely by Messrs. Baker and Kent, because curves on that basis are 
not so readily comparable for estimating purposes and also because the writer has 
failed to detect any corroboration of Messrs. Baker and Kent’s theory regarding 
humps and hollows of resistance in the models whose results are presented to you 
to-day. It is probable that the theory advanced, that at certain fixed P values 
resistances are at a maximum, may not hold good where the variation in prismatic 
coefficient is procured by variation in ratio of entrance to run in conjunction with 
variation of parallel middle body, or possibly the speeds of these ships were too low 
to prevent the masking of Baker’s humps by other more important developments 
of resistance. I believe that the curves placed before you to-day would permit 
some investigation along these lines. 
In order to illustrate this form of presentation and to permit of ready com- 
parison of present results with previous investigations, diagrams of tests showing 
the effect of location of parallel middle body, as carried out by Mr. McEntee and 
