BUOYANCY AND STABILITY OF TROOP TRANSPORTS. 143 



Suppose next that a ship is designed to the two-compartment standard where 

 the loss of one bulkhead does not endanger the safety of the ship. We have then, 

 with a minimum spacing of So: — 



or, with the value of So here adopted — 



90 feet ^ / ^ 135 feet. 



With a spacing oi s ^ So =^ 4S feet, we have again the probability of a bulkhead 

 being lost: c = 3.5, but gradually, as s is increased, c will fall off, until in the 

 limit when / is equal to 135 feet and s equal to say 65 feet, we have 



65-35 



that is, the chance of a bulkhead being lost is one-third of what it would be with 

 s = 4S feet. This advantage of the wider spacing is believed to outweigh the 

 drawback that when a bulkhead is lost and two compartments flooded there will be 

 a smaller margin of safety than with the shorter spacing. Hence it is best in a 

 two-compartment ship always to make the spacing approach one-half of the flood- 

 able length. 



When f >T,So or f >i35 feet, the three-compartment standard can be attained. 



Suppose that— 



135 feet </< 180 feet, 



then the best result is attained within the three-compartment standard by making .s 



equal to or closely approaching —, in which case the probability of a bulkhead 



3 

 being lost will range from 3.5 to 1.4. It might be argued that it would be better 



to retain the two-compartment standard and make s =—, since then the value of 



c would be much reduced and would range from 1.08 to 0.64, but in that case the 

 hold compartments would be unnecessarily, perhaps inconveniently large, and when 

 a bulkhead was lost the flooded length would be 50 per cent greater than with the 

 three-compartment ship. It is to be borne in mind also that a three-compartment 

 ship is better able to stand underwater damage caused by collision or gunfire which 

 is not so liable to destroy the bulkheads. Hence, when / is greater than t,So and 

 smaller than 4S0, the three-compartment standard should be adopted with s equal 



to or slightly smaller than -^. 



It is clear that the same reasoning will apply and an analogous conclusion will 

 be reached when the floodable length is greater than 4.yo and we arrive at the fol- 

 lowing general rules: — 



