ECONOMICAL CARGO SHIPS—SOME MODEL EXPERIMENTS. 59 
is suggested for making calculations to take account of this effect, but the C constants 
given in recent papers of the tank have been corrected without any such modification. 
Its effect is not so large as might appear at first sight since it is to be made both on model 
and ship friction. 
In our reports of forms we work it out in some cases, and the total effect for large 
merchant ships is usually of the order of 4 to 6 per cent, and this percentage includes 
the effect of the adoption of the Teddington version of existing skin-friction data. One 
point of difference in Mr. Robertson’s correction and that adopted at Teddington is 
the standard temperature that is taken at 70° at Washington and is 55° at Teddington, 
which would have the effect of making our results appear a little worse than those of 
Washington. 
Thirdly, with regard to the “‘humps”’ which so often show on the C curves of ships. 
It is quite true that in some cases the humps due to the P values only are masked by 
other important developments, as suggested by Mr. Robertson. One of these important 
developments I have given in some detail in my book, namely, what is called an entrance 
wave. But when a hump occurs, it is very often quite easy to detect what it is due to, 
after the P values for the boat are worked out as suggested by Mr. Kent and myself. 
For example, in the figures published by Mr. Robertson there are no less than thirty 
humps, the majority of which can be explained by due and proper consideration of the 
P values at which they occur and the point at which the entrance or run will begin 
serious wave making in their own length. There are two or three cases in which we are 
not able to identify these humps, but that is all. I may illustrate the preceding by point- 
ing out that the bad result obtained by model CX in Fig. 17, CW in Fig. 16, model CW 
in Fig. 11, model DV in Fig. 10 (Plate 23), and model BX in Fig. 5 (Plate 19), are all due to 
bad combinations of their P3 or P4 hump with the entrance hump. Although plotting 
to a P base does not always show humps, yet when they occur we have found at 
Teddington that such a plotting enables one to detect the cause of the hump with 
comparative ease. 
Mr. Rosertson (Communicated) :—In reply to Mr. G. S. Baker, I am informed 
that shortly after the establishment of the Ann Arbor tank Dr. Sadler towed plane 
surfaces therein, the results of which when plotted differed so little from the resistances 
calculated by the Tideman coefficients that those coefficients were adopted as standard 
at Ann Arbor. 
At Washington, in 1901, tests were made with a 20-foot plane of 64.1 square feet 
area, and the results were tabulated and are constantly in use in the estimation of wetted 
surface friction. The formulae in use at these two tanks are shown on Fig. 4, Plate 18. It 
will be noticed that the frictional resistance for the ship is based on the same formula for 
both tanks, the C difference being due wholly to the estimation of the frictional resist- 
ance of the ro-foot or 20-foot model. The introduction of a temperature correction to 
the Ann Arbor models is an innovation in the results herewith, the correction being similar 
to that used at the Washington tank. 
I can see many reasons why a standard formula would be very desirable for all the 
tanks, both for temperature correction and for frictional resistance, and the first step 
towards this is to let some light into what the tanks are presently using. There is already 
a large body of information at each tank calculated by individual formulae, and conver- 
