Fbbeuabt 19, 1909] 



SCIENCE 



283 



for surfaces of revolution as experimentally 

 determined, is known to vary not with any 

 constant power of tlie velocity, but by a 

 range of exponents from the first to the 

 cube, if not higher. For example, in the 

 enormous velocities attained by modern 

 artillery, where bodies weighing a ton or 

 more are hurled through the air at 2,000 

 feet per second, it is known that the phys- 

 ical phenomena become entirely different 

 in nature from those found when dealing 

 with moderate velocities such as are met in 

 transportation devices. 



If the rigid system be employed where 

 an internal frame prevents deformation of 

 the envelope, the stresses due to external 

 pressure are taken up by the framework 

 itself, and the gas required for flotation is 

 usually contained in several separate recep- 

 tacles or ballonets similar to compartments 

 employed in ships. In this system, there- 

 fore, we are concerned only in securing 

 such a shape of the rigid frame as will ful- 

 fill the condition of minimum total resist- 

 ance for a given displacement and velocity. 



Once the shape of the bag is determined 

 from the considerations already enumer- 

 ated, the dimensions become immediately 

 fixed when the tonnage is assumed, or con- 

 versely, if any linear dimension is assigned 

 the tonnage is thereby determined. 



In addition to the two general systems 

 above considered, there are various types 

 involving some of the principles of each, 

 which are classed in general as semi-rigid 

 systems. Such systems usually comprise 

 a rigid frame, to which is attached the gas- 

 bag above, and the load below. 



The next step is one of structural design 

 along strictly engineering lines. The aero- 

 dynamic features of airship construction 

 may be considered under the heads: (a) 

 static balance, (6) dynamic balance, (c) 

 stability, {d) natural period and oscilla- 

 tion. 



Static Balance.— The dimensions of the 



gas-bag being determined, the lift of each 

 transverse segment thereof is immediately 

 known, and the design of the frame may 

 proceed by approximate trial and correc- 

 tion as in other structural work. The 

 weight of each segment of the envelope 

 itself is readily computed, which, added to 

 the corresponding segment of the frame, 

 gives the total weight of each segment, and 

 this total subtracted from the lift of each 

 segment gives the net lift for that complete 

 segment. From the magnitude and posi- 

 tion of these net forces the position of the 

 resultant lift is known, and this determines 

 the vertical line through the center of 

 gravity. Such procedure evidently insures 

 static balance of the machine as a whole, 

 and an approximate distribution of the 

 load. 



Dynamic Balance. — The dynamic balance 

 must also be carefully considered ; and here 

 a difficulty has been experienced on account 

 of the inability to place the resultant thrust 

 coincident with the line of resistance of the 

 ship as a whole. Heretofore, it has been 

 customary to balance the thrust-resistance 

 couple by means of suitable horizontal rud- 

 ders or planes, so situated and at such 

 angles that the resultant moment of the 

 system should be zero at uniform speeds 

 of travel, though not necessarily zero for 

 accelerated motion. 



If, however, the line of thrust be made 

 coincident with the line of resistance, the 

 disturbing moment in question will be elim- 

 inated at uniform speeds. If, furthermore, 

 the center of mass be located on the line of 

 thrust and sufficiently forward to form a 

 righting couple with the resistance when 

 the wind suddenly veers, the evil effects of 

 a disturbing moment will be obviated for 

 variable as well as for constant speeds. 

 The ship is then dynamically balanced. 



This, of course, requires that the form of 

 hull be such that a quartering wind shall 

 exert a force passing to the rear of the 



