58 THE DETERMINATION OF THE RESISTANCE OF SHIPS. 



placement went up from i8 to iS^^ knots. This increase in a vessel displacing 

 over ii,ooo tons was due solely to a change in the lines of the bossing. 



In battleships appendage resistances as high as 25 per cent of the total have 

 been recorded. The importance of reducing this where practicable is at once evi- 

 dent. 



For more detailed information the reader is referred to Taylor's Manual and 

 to Dyson's work on screw propellers. 



No important vessel should be laid down without stream line diagrams at crit- 

 ical speeds. 



The fitting of anti-rolling tanks has a small and indirect effect on resistance. 

 The result of omitting bilge keels is a saving of about 3 per cent, which is cut in 

 half by the extra displacement due to weight of the tanks and contained water ; 

 this saving should help to pay for the tanks and controlling gear. Anti-rolling 

 tanks, however, undoubtedly require skilled handling and constant watching whilst 

 at sea. 



References to bilge keel and other appendage resistances will be found in the 

 bibliography. 



Baker has an interesting chapter on appendages in his book just published. 



The importance of an allowance for air resistance should not be overlooked. 



SUMMARY. 



To sum up briefly the above notes, suggestions and recommendations, the fol- 

 lowing may be said: — 



1. The system of notation used by Taylor and Sadler in presenting their re- 

 sults is ample, clear and direct ; it is, therefore, preferable for general use. 



2. The model of practically every naval or passenger design should be tested 

 in the tank and enough models of cargo vessels similarly tested to cover a useful 

 range of vessels, so that each yard can know whether their "lines" are good and 

 make improvements from time to time as experience points out. This is the only 

 real way "lines" can be compared and improved, for trial conditions vary all the 

 time and service conditions and results are seldom accessible to builders except in a 

 very general way. In this connection the relative smooth and rough water per- 

 formances should be kept in mind. 



3. For large cargo ships and practically all motor ships, a twin-screw drive is 

 the most efficient ; the two, three or four-screw drive is so well established in pas- 

 senger and war ships as to need no further comment. 



4. Further light is desirable on vibration problems, especially for vessels with 

 machinery aft. 



5. Speed and stability are intimately associated and generally pull at cross pur- 

 poses ; the temptation to small beam, high speed and comfort at sea for passengers 

 may be yielded at the price of safety in accident. 



6. We are at present witnessing the passing of the very full cargo ship at sea 

 in favor of a somewhat finer type. This is a good step. 



