1196 
ue 
lear 
to ie tn? dy. Proceedina on these lines, we obtain the following 
K 
n 
final results, comparing this case with the one for uniform water:—- 
a= 10° No chanae in flow of energy due to stratification. 
oi ao Energy flow in top layer increased by factor of 2,5 
Energy flow in bottom layer reduced by about 15% 
w= 1.82 x 10* Energy flow in top layer increased by factor ty strati- 
of 2.9 fication 
Energy flow in bottom layer reduced by about 
20% 
These results were not obtained by a very accurate method, as it was only 
desired to calculate the order of magnitude of the effect. Since the pressure 
pulse from a large charge is represented approximately by 17" anda is 
of the order of 102, it is clear that only the higher frequencies, making 
up the head of the pulse, can be canalised. 
Extension to other Depths of Water. 
The difficulties that stand in the way of applying this theory to 
infinite depths are fairly obvious, Not only does the number of carrier 
modes become Infinite, so that we have an integral to deal with instead of 
a sum, and the processes we hav2 adopted (while clearly justiflable fora 
finite serles) may not te so for an integral, but we have to take account 
of the remalning modes as well, It can, however, be shown without difficulty 
that If wo are 1,000 feet away, the figures we have given for initial 
frequency and energy flow would hardly be altered up to depths of 1,000 feet, 
so that our Investilaatlon seems to cover most of the situations that areever 
fikely to occur at a river mouth, 
Conclusions, 
It appears that the canalisation nechanisn suggested by Professor 
Taylor is in principle capable of accounting for freak cases of "shock damage" 
In such placee as river nouths where fresh water is likely to lie atove salt. 
It also seens that this mechanism Is not capable of accounting for freak cases 
of camage to ships! plating, because only part of the pressure pulse Is 
transmitted In this special way. It also seems that stratification due to 
tempcrature olfferences is too small an effect to be of Importance In this 
conncctlon. The extent to which the conclusions must be modified in ordor 
to take account of the differences between sound and explosion pulses is 
uncertain. The factor found for the higher frequencies is about three in 
the anergy flow, which would give a factor three in the maxinum distance 
for damage caused by these frequencies, 
