AMERICAN LEGION CLASS. 197 
reading 0.8 inch lower than the temperature reading. In another case, the vacuum corre- 
sponding to the temperature is 28.53 inches, whereas the recorded vacuum is 28.8 inches, 
which is 0.27 inch lower. On Plate 59 the vacuum corresponding with the temperature 
agrees exactly with the recorded vacuum. 
ApmiraL C. W. Dyson, U. S. Navy, Member of Council (Communicated ) :—The great 
point which stares the reader of this paper in the face is the large discrepancy between the 
shaft horse-powers for the different speeds as obtained from the Model Basin trials and those 
obtained on the actual trials of the vessels. 
Why this discrepancy? In the paper the author hints at the possibility of the trial 
course depth being the cause. I cannot agree with him as to this. I ascribe the entire trouble 
to the character of the hulls. These hulls are extended in the quarters to form a casing in 
which the propeller shaft is carried, and these casings extend clear aft to the bosses on the 
struts. The structure varies from about 5 feet in the external width at the forward end to 
the diameter of the strut boss at the after end. The depth is not shown on Plate 55, but 
it is sufficient to carry the stern tube and stuffing box and one coupling and one line bear- 
ing and must have clear area enough to give access to these. 
In my experience I have encountered several vessels where horizontal interferences to 
vertical flow of the water closing in around the stern of the ship exist, and others where 
these interferences have been with the horizontal flow of the closing-in water. In the former 
cases, with the interference extending clear out from the hull of the ship to the propeller 
shaft, as in the case of a spectacle frame with horizontal fin, the effect has been beneficial, 
the propeller being slowed down in revolutionsand the “Loss Factor’ K becoming VK. In 
the second case, however, where the horizontal flow is interfered with, incipient cavitation 
occurs due to the bad quality of the wake between the interfering member and the hull, 
speeding up revolution where the basic slip is 13 per cent or less, and causing K 
to become K?. 
As Model Basin trials do not give any indication of cavitation, of course, the results 
obtained by running the self-propelled model are those which may be expected when neither 
incipient nor full cavitation occurs. 
I have analyzed the propeller performances of the President Pierce and of the Ameri- 
can Legion on the assumption that incipient cavitation is actually occurring, and in place 
of K have used K”’. 
I include herewith these analyses and call your attention to the close agreement between 
estimated powers and revolutions and those obtained on the actual trials (page 198). 
Mr. AtFrepD J. C. Rosertson, Member:—Gentlemen, as you are no doubt aware, four 
of these ships have been on the South American route of the United States Shipping Board 
and operated under the Munson Steamship Line. I have had access to the logs, and it would 
have been a very great pleasure if I could, from these, give you really reliable data as to 
the performance of these ships. I have gone over the logs very carefully, but anyone who 
has not looked over logs of that kind, and gone into them a little bit, will realize how difh- 
cult it is to eliminate the various things which will prevent us getting reliable figures. 
These boats run on a schedule; the first day after they leave port they are retarded in 
their progress for various causes. Then they cross the Gulf Stream and later on get into 
the tidal streams of Cape Roque, so that altogether their sea speeds do not correspond in 
any way to measured mile conditions. 
