276 TERRESTRIAL REFRACTION AND MIRAGE. 
(3) Between 10 in the morning and 1 p.m. no observations were 
taken, but it may be gathered that refraction is likely to be 
either positive or negative but not as a rule exceeding about 
goth of the contained arc. 
(4) During the afternoon the refraction passes through zero from 
negative to positive, the average time of this phenomenon 
being about 2.80 p.m.; that is to say about 2.30 p.m. there 
was a state of the atmosphere in which there was practically 
no refraction—refraction was on the point of passing from 
positive to negative. This time of zero refraction occurred 
earliest on cloudy days and when there was much moisture in 
the atmosphere, and latest on days of bright sunshine and 
little humidity. 
(5) I gather that absence of cloud has a tendency to diminish 
positive refraction, and frequently to convert it into negative. 
Supposing a cloudy sky and considerable positive refraction, 
and if the clouds clear away and the sun comes out, the effect 
generally is to enormously diminish the positive refraction, 
and frequently to change it to negative. Thus in General 
Walker’s observations in a certain instance on a cloudy day 
the mean decimal of the contained are was +:046 ; but under 
otherwise identical conditions on a cloudless day this decimal 
was changed to —‘01. On another occasion at 1.30 p.m. the 
disappearance of cloud instantaneously changed the decimal 
from +°'042 to —:036. 
(6) As the humidity of the atmosphere increases the negative 
refraction diminishes, and positive refraction increases. 
Humidity in fact seems to have the exactly opposite effect to 
the sudden appearance of sunshine. 
(7) Lateral refraction is negligible except under very excep- 
tional circumstances. 
I received a very interesting letter from General Tennant, R.E., he 
writes : “ When the ray passed over damp or marshy ground or water, 
the case was worse, and often especially in the morning the distortion 
and fluctuation of images both horizontally and vertically was enormous. 
I have seen an image wander several minutes with an intermittent 
motion to one side, and then return, till after long watching we were 
compelled to give up hope of obtaining trustworthy angles.” And 
again he writes: “I have a vivid recollection of trying to observe a 
heliostat over part of the Runn of Kutch one morning when my recol- 
lection is that I saw it move out of the field and the wanderings were 
quite 30' horizontally.” Such cases however are quite extraordinary, 
and not likely to be encountered by artillerymen. 
The seven indications drawn from General Walker’s tables, and in part 
quoted from his report, are likely to be accentuated in artillery practice 
on a plain when the average height of the ray is less than the 19’ to 40’ 
mentioned. 
I finally draw the following conclusions in regard to terrestrial re- 
