The following text may have been generated by Optical Character Recognition, with varying degrees of accuracy. Reader beware!

20 of the propeller shaft and the power delivered to it with considerable ac- curacy, it seemed desirable to deal with the subject of bilge keels. So bilge keels were put on a boat which had been tested without bilge keels. The keels were carried to a depth of nine inches, which is three times more than the ordinary depth for the proportion. In order to justify the work as presented, I wish to call your attention to a diagram which is pre- sented on plates 2, 3, 4 and following. The first three are presented as sam- ples, every condition having been tested by itself, and the results being assembled in that manner. On plate 2, the first appearing in the paper, we have a diagram showing the relation of the speed in knots for revolutions of the shatt. You will notite that there are two sets of points, small black points and the small circles. The first represent experiments made on, July 28 and the other represent experiments made on Aug. 12, with conditions the same, except that be- tween these two dates there had been a change made in that the bilge keels were lengthened, and then the exten- sion pieces were removed, so that on those two dates the boats were sup- posed to be in the same condition. The boat had, meanwhile, been hauled up. Both sets of points are well rep- resented, I believe you will agree, by the line which is drawn. Time of Making Tests "You will find some dispersion of the points, that is due entirely to a Cutrent in the. basins,; which current was evidently produced by wind which had and had ceased blowing from eight to twelve hours before the _ test made. We had to choose an oppor- tune time for making the tests, when the wind did not blow at all, or when there: was only: a fair trick: "On plate 3 we had plotted a scale of revolutions per minute of thrust of the shaft in pounds. This represents one series of experiments, and we found, as would be expected, that this Same curve represented properly all of the several series of experiments, because there is no reason why small changes made in putting on and tak- ing off bilge keels should affect the been blowing somewhere Was thrust for a certain number of revolu- tions. "On plate 4 we have in like manner the power as measured at the input to electric motor. We have again the black dots and the circles, all of which lie on the curve, because this relation not the small current. It is interesting, in the here was influenced by THE MARINE REVIEW January, 1914 PROM. -CHARTE ES HH, PEABODY Passing: to, note that this is a. deli. cate way of measuring very small cur- rent, but I do not know of any other method, which is available for meas- uring such currents with such a gree of delicacy. de- Having these cases; it is presented sample but fair to present on plates 5 and 6 certain experiments, also the results of which were not so good, and these we may present as the worst of those which we secured. On plate 5, near the top of the curve, some of the circles do not lie so well like. At the time when these experiments were on the curve as we should made a breeze had sprung up and died away again. I cannot assert that the deviation is due to that. The devia- tion 1s not of serious importance. On fae other. hand, on plate 6, at the top of the curve, there is a deviation for which we offer no explanation. It is in this case to be charged only to unavoidable errors of our experi- the nation of the thrust curve, which is in ments; but since this is determi- a sense the best located curve, we may fairly attribute this to unavoid- able errors, and. not .consider that they indicate any defect in our test. 'On plates 8 and 9, certain results are assembled--on plate 8 thrust in pounds, as determined from the pre- ceding curves, and on plate 9 we have the power as delivered to the propeller shaft, allowance proper having been made.for the friction. "This work may be summed up as €iven ion. page 2: "For speed less than that corresponding to a _ speed length ratio 0.7 the increment in