TO PROPELLERS OF OCEAN-GOING MERCHANT VESSELS. 87 
thermometer or a difference between the superheat obtained on this trial and that for which 
the curves were drawn. 
Problem 4, Plate 21.—As there was no measured mile within reasonable distance, the 
run was made from Ambrose Light Ship to Fire Island Light Ship and return. The revolu- 
tions were held as near constant as was possible, and indicator cards were taken every fifteen 
minutes. While the results are not as accurate as if the run had been made over a measured 
mile, it is believed they are sufficiently accurate to warrant presenting them in this paper. 
There was a very slight swell and a breeze of not over 3 knots during the trial. 
Problem 5, Plate 22,—The data for this problem were taken from ‘“‘Notes on the Per- 
formance of S. S. Tyler,” by Mr. E. H. Rigg, published in Vol. XX of the transactions of this 
society. The data given as actual I. H. P. and R, are those observed on the trial runs. If 
those of the faired curves had been given the actual I. H. P. for 12.233 knots would have 
been 1,480 and the actual R, would have been 90.8, and for 9,876 knots the actual I. H. Pu 
and Ra would have been 790 and 73.2. 
Problem 6, Plate 22.—The power and revolutions as given by an analysis did not agree 
with those obtained on the trial. As the difference was so great, an investigation was made 
as to the causes. It was found that the wheel was working in cavitation and the effective - 
horse-powers were greater than those shown on the curves. An analysis was made for revo- 
lutions (using the I. H. P. from the trials), which check with those obtained on the runs. 
The effective horse-powers were then calculated and compared with those of Model 1013, 
the block and midship section coefficient and dimensions of which are the same as this ship. 
The length of after body was the same, while the middle body of Model 1013 was slightly 
less and the fore body slightly longer. The forward sections of the ship were more of the 
U shape, and the after sections had more of a club foot than the model. According to experi- 
ments conducted by Professors H. C. Sadlerand E. M. Bragg, the results of which were 
published in last year’s transactions, it would appear that Model 1013 would require slightly 
less power for these low speeds. This would only partly account for the difference; the 
remainder would probably be caused by the condition of the ship’s bottom. The vessel was 
drydocked and then loaded, after which the trial was run. The time between drydocking 
and the trial was about three weeks. 
Problem 7, Plate 23.—The results of the trials at two speeds of three ships of the same 
design, together with the analysis, are given in this problem. No effective horse-power curves 
were available. The revolutions were calculated using the indicated horse-power. The ef- 
fective horse-power was then calculated and compared with that estimated. 
GROUP 2. 
We will now pass to the second group; namely, ships fitted with more than one screw, 
the shafts of the wing propellers being carried in struts. The first two, Problems 8 and 9, 
are the navy colliers Jupiter and Cyclops. While these are not merchant ships, their forms 
are very similar to moderate-speed freighters or to slow-speed passenger and cargo ships and 
are included in this paper for lack of other available data. 
Problem 8, Plate 23.—As the writer was not in possession of the length of after body 
of this ship, an estimate was made using that of the Cyclops (Problem 9, Plate 24). 
Problem 10, Plate 24—This ship had been in the water for some time before the trial, 
