1262 
under the Rankine-Hugoniot rather than the adiabatic assump- 
tion. In each of these figures is also drawn (as a dashed 
curve) the acoustic approximation, 2 Pye When the acoustic 
approximationsare straight, the correct pressure-time curves 
are seen to be concave upward, and in the exponential cases 
the correct pressure-time curves are more concave than the 
acoustic ones. | 
10. Probably the most interesting feature of the 
curves in fig. 2 - 4 is the difference in behavior shown by 
air and water. In terms of the y defined in (7.1) the in- 
fluence of the fluid is shown in fig. 5. Here the impulse 
given to the wall is plotted as a function of y for three 
values of the strength, L§ » of the incident shock. In fig. 
5 the incident pulse is always of type (c), that is, it is 
progressive and its velocity decreases linearly behind the 
shock front. As remarked before, the pressure in this case 
is also approximately linear. The impulse in fig. 5 is 
normalized as in fig. 2 - 4. In this graph all gases lie 
between Y= 1 and Y= 5/3. The limiting value Y = 1 may 
be interpreted as relating either to an isothermal reflection 
or to reflection in a gas whose polyatomic molecules have in- 
finitely many inner degrees of freedom; ¥ = 5/3 describes a 
monatomic gas. The curves show that in gases and water the 
impulse given to a rigid wall is respectively more and less 
than acoustic theory predicts. The esssntial reason for this 
was indicated in par.5, namely, in gases the tail of the in- 
