Integral therefore demonstrates the transformation of sea into 
swell without any effect of friction. 
The period increase of ocean swell 
One point which Sverdrup and Munk [1947] make in their studies 
of ocean waves is that the "significant period" of swell from a 
distant storm is higher than the "significant period" of the waves 
in the storm and that when the swell is highest the "significant 
period" is higher than the "significant period" in the storm. 
Sverdrup [1947] explains this observed fact by supvosed selective 
attenuation of low periods. Figures 15 and 16 show another possible 
explanation. In these figures, the peak of the power spectrum 
is at a higher period than the median value of the square root 
of the power spectrum. From the figure the "significant period" 
in the storm would be probably around seven seconds. Now, the area 
unger the power spectrum equals the square of the wave record; 
and at large x, when the band width is narrow, the highest waves 
occur with a "significant period" of ten seconds. Hence, this 
shows a period increase of ocean swell without selective attenu- 
ation. But the "significant period" of the swell does not in- 
crease indefinitely. It increases to ten seconds as the width 
of the filter narrows, and then stops increasing. It is known 
that the forecasted swell periods in the Sverdrup Munk theory 
fail for great decay distances,* and the above reasons could easily 
be an explanation. 
*Personal communication, R. S. Arthur 
- 168 - 
