THE LONGITUDINAL STRENGTH OF RIGID AIRSHIPS. 177 
This difficulty I propose to overcome without departing from the bending method, by 
conceiving the keel and the remainder of the ship as separate and independent girders, except 
in so far as they have the M girders in common. Imagine the M girder to be divided into 
two parts, proportioned to the horizontal shear forces in the KM and MN panels, and allotted 
respectively to the ship and to the keel, we may then regard the keel and the rest of the ship 
as entirely independent of each other and calculate the stresses in each according to the bend- 
ing method. A similar procedure may be followed in the longitudinals A to E. 
I do not claim that this solution is mathematically exact; it is, of course, an approxima- 
tion. Actually, I propose to take into account not only the bending but also the shearing de- 
flection of the keel, and I make allowance for any possible difference in stresses in the two 
parts of the M-girders as determined by this method. Without taking up too much of your 
time, I cannot go deeper into this question at present, but I hope to have proved that it is 
necessary to follow primarily the bending method and to use the shear method in a subsidiary 
capacity. I object to the latter method only when used as the sole means of calculating the 
stresses, especially because the elastic strains in the longitudinals and the horizontal shear 
forces thereby created are neglected. 
Applying the method of calculation here recommended to a modern rigid airship, it is 
found that the moment of inertia of the ship girder, considered apart from the keel, is in- 
creased by about 8 per cent due to the main and secondary wires, substituting for them the 
equivalent or fictitious bars. If, moreover, the wire netting is included with due allowance for 
its curvature and tension due to the pressure of the gas cells, the moment of inertia will be 
increased altogether by about 10 per cent. This is nearly twice the amount of that deter- 
mined by the ordinary mode of calculation as given by Commander Land. 
I wish to thank Mr. Sperry for his very appreciative remarks, and I quite agree with 
him that great caution should be exerted in carrying out the tests with the ZR-1, but I believe 
that the Bureau of Aeronautics is fully aware of this and that all possible precautions will be 
taken. In fact the program of the tests, which is very detailed and complete, is so planned 
as to pass systematically and with caution from the simplest and easiest tests to the more 
severe tests by gradual stages under most careful observations of the behaviour of the ship 
and of all its parts. Such limitations are imposed for the first tests on speed, and on the 
use of the machinery, rudders, etc., as to make serious accidents extremely unlikely. Minor 
failures may of course occur, but it is believed that such weaknesses, should they appear, will 
be readily corrected without further trouble. 
THE PRESIDENT :—Professor Hovgaard, on behalf of the Society, I wish to extend our 
thanks to you for this very valuable paper. 
