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

24 Lee figures the shape of the blade face has an appreciable effect, so that for a given area ratio they would deliver less thrust and absorb less power than those of the full ended type which I used. There is no data to show the exact shape Prof. Durand used, and there may have been some other minor differences 'in section which would affect his results, but the effect of thickness and blade shape is, in my opinion, enough to account for the difference between his results and mine, without assuming material er- ror in his results or mine. The effect is not only as regards the power and thrust, but there is a decided effect on the virtual pitch to which Prof. Mc- Dermott referred. Referring again to Fig. 5, these curves are not carried down to the negative nominal slip. If they were carried down to negative nominal slip, you would see would reach zero thrust at a much smaller negative slip than the thick - blades. If Col. Stevens' results were modified, showing the results were plotted in each case for the virtual pitch, instead of nominal pitch, you would find the curves of efficiency would be remarkably close. That has been practically done by Mr. McEn- tee, who has plotted results, not upon the nominal slip, but upon the thrust of propeller. That substantially re- duces them to the same virtual pitch and causes, as he shows by an exam- ple of curves, the efficiency to come very close togethen. As regards the maximum efficiency, I do not see how the change in method of plotting re- sults could affect the maximum eff- ciency. At the time of my 1904: paper, when I gave an efficiéricy of 78 and more for three-bladed screws, I hesi- tated a good deal before publishing such high efficiencies, but I could not get them lower from my experimental results. Since then Prof. 'Durand published his results last year with four-bladed screws, which naturally have less efficiency than three-bladed screws; he found efficiencies of 74 per cent. I have also reliable information to the effect that Mr. Froude made a number of experiments on propellers, later than his classical experiments, published twenty years ago, and also found efficiencies reaching as high as 75 per cent'for three-bladed screws. The efficiencies, as you will find. on referencé to my original paper two years ago, eliminate practically the ef- fect. of the hub as plotted by me. Prof., Durand excludes the effect of the hub, and as regards maximum effi- that the thin blades. TAE MaRINE REVIEW ciencies we may be said to be in close agreement. With regard to Capt. Hovgaard's remarks, I am sorry I cannot agree with him in concluding that the re- sults show that the thrust per square inch is the prominent factor in caus- ing cavitation rather than the acceler- ation, As a matter of fact I do not see there is much difference. The thrust per square inch is mofe or less proportional to the acceleration. In the experiments given in this paper, as you approach cavitation, the slip does not change very much, before the speed of the screw commences. You have comparatively little cavitation, but get enormous increase in thrust and acceleration of the water, owing to the natural increase in velocity of the screw. I have been personally in- clined to think that cavitation is large- ly affected by the blade section; in other words, that you might find a form of blade section which would produce much less acceleration of the water for a given thrust, or a very much less violent variation in acceler- ation of the water fon a given thrust, than in other forms. Cavitation ap- pears to be sought to be avoided now- adays by increasing the width of the blade, necessarily accepting the in- creased friction loss involved. As you increase the width of the blade, hold- ing constant the maximum thickness, you decrease the angle of entrance, if I may so express it, and-I think the gain is from decreasing the angle of entrance rather than from decreasing the thrust per square inch, as a result of the greater area. - 'I would invite attention' 'to the papen of two years ago, in which a model 'screw gave signs of cavitation at five knots; at six knots it gave dis- tinct cavitation, 'and at seven knots showed very 'pronounced © cavitation, but at a very much lower thrust than it began to show cavitation at five knots. I think the thrust was prob- ably not over half that. At five knots you can get twice as much thrust out of the screw, before you get cavita- tion, than at seven knots. The dif- ference was that at seven knots the velocity of the screw through the water was very much greater; it was clear that the cavitation was due. to the: blade actually splitting the water; whether that would apply to full size screws I would not like to say. As regards the method of calculat- ing the efficiencies referred to by Mr. Thomas, that was also fully described in my 1904 paper. The efficiency, however, was taken at the ratio be- tween the work absorbed in a given time, determined from the measure- ment of torque, and the useful work taken as the thrust multiplied by the distance traversed by the screw in the same time. The efficiency was simply the ratio between the useful work and the work absorbed by the screw. There is no other true efficiency I know of. We determined the efficien- cy without considering the slip, al- though it was afterward naturally plotted on the slip as this was con- venient. As regards Prof. Peabody's remarks, I would be glad some day to attempt to combine this data into more simple form and reduce it a little more to rule, but purposely in these papers we so far have given as nearly as possible the original data, so that any one who wished could modify it or transform it in any way he wished for design purposes. I am aware of several people who have used them for design purposes, having reduced them to rule, but each use a different method and each man thinks his own the best. Mr. McEntee's suggestion that experimental results be plotted in standard form might be useful, but it seems to me it would be more de- sirable to make sure that the experi- mental results are fully given, so that they can be reduced to any form de- sired. 'Col. Stevens also leans in that direction, evidently, because he is do- ing a good deal of work in comparing the two. Although Prof. Durand's re- sults and mine were apparently criti- cally different, Col. Stevens found a satisfactory method of reducing them to a common basis and getting a re- liable comparison. The President--Owing to the ab- sence of Col. Stevens and the declina- tion of my friend Cox to try to fill his place, the reply on behalf of Mr. Stevens' paper will have to be passed over. This completes the list of papers for today. We will not take up any fur- ther papers this evening, as those who could not get here today are counting on certain definite papers to be con- sidered tomorrow. We will now ad- journ. ' . Three steamships that have been in the coasting trade will at once sail from San Francisco to Central Amer- ica to enter the coffee traffic. These vessels are the Indiana, formerly in the Seattle-Alaska trade, the Barra- canta, and the Aztec. These vessels are controlled by the Pacific Mail Steamship Co., and will ply between the various coffee exporting ports and Ancon,