JOINT DISCUSSION ON TWO PRECEDING PAPERS. 115 
work, to certain speeds of advance, and these speeds of advance were assumed to be the speeds 
of advance through the water behind the ship, from which the wake factor was readily deter- 
mined. When torque and thrust coefficients gave slightly different speeds of advance they 
were averaged. 
In pointing out the effect of the immersion of the wheel, I rather hoped to start some- 
thing more in the way of discussion than we have had. It is rather a new point, and I was 
glad that Prof. Sadler was able to sustain my somewhat radical position on that matter. 
We have had a good deal of discussion as to the basis of the formulae. Luke, as you 
know, who was the pioneer in this matter, and also Mr. Semple, who was also a pioneer, at 
first adopted the block coefficient, and Mr. Semple, two or three years ago, adopted the co- 
efficient of the after body. I do not believe there is any single coefficient that can be used to 
advantage to cover the whole range for this reason: When you consider what the wake is, 
you find that it is really made up of three components. In the first place there is a fric- 
tional wake produced by the frictional drag upon the vessel. That is fixed and does not vary 
very much with fullness any more than surface varies with fullness. Prof. Bragg’s experi- 
ments, with a 10-foot plane, confirm our 4 priori conclusions that it is an important factor. 
He got 20 per cent behind the 10-foot flat plane. Then we have the stream-line wake, due 
to the natural stream-line action around the ship, which means a forward velocity abreast the 
stern. That stream-line wake depends on the whole body of the ship, without any question, 
but, broadly speaking, the most important factor entering into it, it seems, is the fullness of 
the after water lines—that is the most important of the numerous factors entering into it. 
I agree with Prof. Sadler that the way to modify the stream lines is not to modify the 
coefficients, but modify the after lines. That was our intention, and the idea was, without 
making much general change in the body, to find the after body or rather the after part of 
the after body. 
The second component is very difficult to reduce to rule, but it seems to me that per- 
haps the block coefficient was inadequate. The vertical coefficient for the after body and the 
longitudinal coefficient of the after body upon which Semple plots his results would, I think, 
both be better guides. In view of the great influence of propeller position it is hardly worth 
while to strive for minute accuracy in determining the influence of the shape of hull. 
We have another factor due to the stern wave. That is a factor affected by speed. There 
is no question about that, as we all know. In destroyers we sometimes have a negative 
wake, and the reason is that the stern is in the hollow of the wave which they have created. 
That wake depends upon the speed-length ratio and is a varying factor, which is liable to 
confuse the case for high-speed vessels. That is why I speak of the effect of speed. In all 
the cases we have discussed I presume the speed was low. It was certainly so in the case we 
tried, where the maximum corresponded only to about 12 knots on the large vessel. When we 
see the peculiar and material variations in wake brought about by comparatively small ver- 
tical or horizontal changes in propeller position, it looks to me as if it would be rather diffi- 
cult, with our present knowledge, to reduce the matter torule. Pending this it looks as if the 
best course is to actually test for wake by model experiment using a current meter, if you are 
looking for the wake alone, or a model propeller, if you are attempting to determine the wake 
and the thrust production. 
Professor Bragg in the discussion desired further information as to the variation of effi- 
ciency of propulsion with hull efficiency. He is quite right in his inference that the efficiency 
of propulsion shown in the experiments did not vary as the hull efficiency, and though the 
