364 
rorULAR SCIENCE REVIEW. 
And clearly, if we could only conceive such a state of things 
as that the ship should be really at rest and the whole mass of 
the rolling sea bodily transferred under the ship, we should get 
a similar result. If the transference were in the direction of 
the wave-motion (which would correspond to a motion of the 
ship against the direction of the waves), the result would be an 
under-estimate of the wave-lengths ; while in the reverse case 
the wave-lengths would be over-estimated. 
Now let us consider the case of sound-waves. These are 
somewhat less familiar to us (so far as our ordinary modes of 
perception are concerned); but, inasmuch as we can more 
readily make experiments on them than on light- waves (owing 
to the enormous velocity with which the latter travel), they will 
serve to give a convenient illustration of the property we are to 
deal with. 
Let fig. 2 represent a series of sound-waves generated by the 
vibrations of the tuning-fork A. When the right-hand prong is 
at a (the limit of a vibration), a is a place of aerial condensa- 
tion; the next such place is at b {ah being the wave-length 
corresponding to the vibrations of the tuning-fork), the next at 
c, the next at c?, and so on. The ivave-amplitude does not con- 
cern us, but I may mention in passing that it is measured by 
the amount of the aerial condensation at a, h, c, cl, &c. The 
tone of the sound depends on the wave-length cib, and a given 
tuning-fork will cause aerial waves of a particular length (that 
is, will give out its proper tone), if it be at rest 
But now suppose that the tuning-fork is being moved, and 
that with a velocity bearing an appreciable relation to the 
velocity with which sound travels. It will readily be seen that 
the tone now produced by the tuning-fork will be different from 
what may be termed its natural tone. 
Thus suppose that during the interval which sound would 
occupy in travelling from a to b, the tuning-fork has been 
moved so that the prong a is at a\ During the interval the 
prong has made one complete vibration, and a' is now therefore 
a region of condensation instead of a ; b is, of course, a region 
of condensation, just as it would have been if the fork had been 
at rest. Hence the wave-length has been reduced to cc'b ; and 
as all the waves proceeding from the neighbourhood of the 
vibrating fork are similarly affected, there results a series of 
waves, pf, fg, gh, &c., as in fig. 3. 
On the other hand, if the fork had been moved in the oppo- 
site direction, there would have resulted the series of waves 
kl, Ira, ran, &c., represented in fig. 4. 
In the former case the tone of the resulting sound would 
have been more acute, in the latter it would have been more 
