329 
17 
Indeed, these contributions are continuously increasing when shocks are pres- 
ent, and they account for the degradation of energy according to (21), (22). 
The velocities produced by these oscillations are easy to estimate, 
G. 1h, 
Denoting such a velocity by ae amplitudes (maxima) by and 
averages by ial » clearly 
Se ee vy 
: 2 — ( it Pa VoNe) 
Z | = A, = eth) lew a bdvy 
uz 
(use (21), (20)), and assuming that the oscillations are essentially harmonic 
ies eae Niat 
ose 
From these 
Vv, 
ee 
V2 
Actually (29) should be corrected inasmuch as /\ s my not be 
entirely kinetic energy: If the oscillations are of finite size, the non- 
linearity of the potential energy Uz = U, (y) will cause we (Cea 
to be different from oh (x, Sey a ) , and while the hydrodynamical energy 
contains the first expression, (26) contains the second one. Thus 
Die a ed = OF =e too contributes to the specific 
dissipation NG . If the oscillation of V, = Xa - Xa is Vee 
, 
then this term is approximately 
ee SC? 2 
a ae (Vose) e a eb (y, 
where 
(30) Ca dp = V oa 
is the local sound velocity. The average is 
