60 ECONOMICAL CARGO SHIPS—SOME MODEL EXPERIMENTS. 
sion tables would be necessary if a change is made, but this only makes it the more 
necessary to act at once. 
I am glad to notice that Mr. Baker has developed further the effect of fullness of 
form on skin friction, and believe some such correction as he suggests is necessary. It 
must not, however, take the nature of a guess, or it would thereby largely vitiate the 
value of the tank tests to which this correction is applied. 
Regarding the use of the P value, I would be very glad to have further information. 
Where the prismatic coefficient is the same, the P value, of course, is the same, and there- 
fore for each of Figs. 15 to 20, Plate 23, the P would be identical for each curve. In 
Fig. 18, for example, P4 is at about 57 aaa and that gives us a hump on DX and a 
hollow on FV. It appears to me that the theory regarding the position of these humps 
is good, but is still incomplete. 
THE CHAIRMAN:—I am sure you will all desire me to express the thanks of the 
Society to Mr. Robertson for his valuable paper and the discussion which has resulted. 
We will now pass to the next paper, the last one for this afternoon, No. 5, entitled 
‘‘Notes on Rivets and Spacing of Rivets for Oil-Tight Work,’’ by Mr. Hugo P. Frear, 
Member. 
Mr. Frear presented the paper. 
