TERRESTRIAL REFRACTION AND MIRAGE. 275 
the concavities. presented towards the denser layers. 
Some of these cases of extraordinary refraction appear when we 
come to study terrestrial refraction in the plains of India. I have taken 
a great deal of pains to obtain correct information in regard to this 
matter. I have been in communication, written and verbal, with 
General J. T. Walker, R.E., lately Superintendent of the Great Trigo- 
nometrical Survey of India, who has given me much information 
which I have drawn upon for this lecture in case ib might be useful to 
those serving in India. I have also had assistance from General 
Tennant, R.H., who was employed for many years on the Great Trigo- 
nometrical Survey of India. He has written me some interesting 
letters giving me his experience of refraction both vertical and lateral. 
General Walker carried out some experiments in India in the Plains 
of the Punjab for the express purpose of determining the extent of terre- 
strial refraction. He selected stations from about 9 to 13 miles apart ; 
at those stations he erected towers, and on the towers he placed the 
theodolites and heliostats, he determined the zenith distances and then 
computed the refractions at various hours of the day and times of the 
year. ‘The observations were very good as observations, because all 
the collateral circumstances were noted ; the indications of the baro- 
meter and of the dry and wet bulb thermometers at the stations, and all 
the climatic conditions were duly noted, so that the results were ex- 
tremely valuable. The circumstances that were found mainly to affect 
terrestrial refraction were in order of importance the following :— 
(1) The time of the day. 
(2) The temperature. 
(3) The aspect of the sky—whether cloudy or sunshine; and 
(4) The humidity of the air. 
The time of the year was late Autumn and Winter, that is'to say, the 
time of year in which there is much artillery practice in India. In 
these experiments the average height of the rays above the ground 
varied from 19 to 40 feet. From General Walker’s tables which are 
found in the account of the Great Trigonometrical Survey of India I 
have gathered the following indications :— 
(1) Negative refraction—that is to say, with the concavity pre- 
sented upwards—was only met with between the hours of 
1p.m. and 3.30p.m.; and the maximum negative refraction met 
with was ‘09, that is +25 of the contained arc between the two 
points, so that the maximum negative refraction met with was 
not of very great importance. 
(2) Positive refraction was met with from 3.30 p.m., and through 
the night to 10 o’clock in the morning ; it was greater during 
the night than in the day time, and the maximum was about 
sths of the contained arc ; that is much larger than anything 
we consider possible in the case of Great Britain, and that 
maximum positive refraction occurred at 7 a.m. After this 
hour of 7a.m. the refraction rapidly diminishes until at 
10 a.m. it amounts to only about 35 of the contained are. 
38 
