286 TERRESTRIAL REFRACTION AND MIRAGE. 
In connection with the more theoretical part of the lecture, that is to say, that 
part which dealt with the curvature of light rays in passing through air, in which 
the refractive power gradually changes, we are able under ordinary conditions in 
England to observe a very similar phenomenon, one which physically is identically 
the same, in the case of sound. 
When the air is under ordinary conditions, it is well known that it is very 
difficult to be heard well when speaking out of doors. Now in the case of sound 
a rise of temperature irrespective of the density is the analogue of greater rarifi- 
cation in the case of light ; each causes the velocity of propagation to rise. Under 
ordinary conditions the air on the ground is warmer than the air above, and the 
result is that sound rays near the ground travel more quickly than those higher 
up; and so sound rays are formed with the concavity upwards, so that when 
a person is speaking on level ground the rays of sound which ought to go to 
the listener curl up and go over his head, in other words the flat ground between 
casts an acoustical shadow. On the other hand under extraordinary conditions, 
that is to say, on a clear quiet night when cooling by radiation is especially 
effective, the air on the ground may be colder than the air above, and then the 
rays of sound which start in a slightly upward direction travel more quickly 
than those going horizontally, curl round, and come down again, while those that 
go along near the ground bend downwards to the ground are reflected and go up 
again; so that you have on such occasions a concentration of sound, not only 
the sound that started in the right direction reaching the listener but other sound 
that ought to have passed over his head coming down again and reinforcing 
the other. Then again when there is an excessively gentle wind, so gentle as 
not to give rise to any turmoil, so that the motion of the air is that of a viscous 
fluid, the upper regions are moving more rapidly than the lower; and then even 
if the whole of the air were of the same temperature the upper regions in moving 
would carry the sound with them more than the lower, and therefore produce the 
same effect as a higher velocity of sound in the upper regions ; and thus it is that 
on these occasions sound can be heard clearly to the leeward at an enormous dis- 
tance, whereas to the windward it can hardly be heard at all. 
Lirut.-Cotonen J. R. J. Jocptyx—lI should like to say a word on behalf of the 
garrison artillery. I do not know that I quite share that feeling of reassurance, 
that the Lecturer gave us, because we have vertical bases in use all over the 
world, and most of our firing, if we ever have the honour to fire at an enemy, will 
probably be in the Mediterranean or in a tropical climate, and I cannot admit 
that the horizontal base instrument is the best way out of the matter. Of course 
the horizontal base position-finder is an excellent thing, if you cannot use a vertical 
base; but as you know, it entails more observers and more complication; and 
also our organisation (I am speaking of garrison artillery) is better suited to 
vertical base instruments. 
At the same time I think that the Lecturer has done very good service indeed, 
in drawing attention to this matter, in the way he has, because, beyond doubt, 
the lecture we have just heard should not be allowed to drop; it is much too 
good a lecture to be allowed to drop; and I hope that it will receive attention. 
I may say that a friend of mine who is known to most men here, Major Rarlow, 
when he came back from Rangoon, told me that there was some difficulty 
there in range-finding, and I believe there was a proposition to build towers all 
round the estuary in order to put depression range-finders up; and there was a 
great deal of discussion about the effect of refraction there. 
I think Major MacMahon said, he himself is absolutely in the dark, as to what 
may be the condition of affairs in tropical climates. 
Of course with regard to the plains of India that is a question for field artillery, 
and I hope that I do not cffend, when I say, that accuracy of fire is not so impor- 
tant for that branch, as it is for the garrison artillery, seeing that their normal 
projectile is time shrapnel, at comparatively speaking large targets, while the 
