Son ANNUAL REPORT SMITHSONIAN INSTITUTION, 1959 
whereas that absorbed by dry soil piled up at or near the surface. 
Any extra absorption by the black surface was more than offset by 
poor conductivity of the soil, and this, naturally enough, was reflected 
in the temperature of the overlying air. Readings taken at midnight 
and sunrise bear out this explanation. At those times, temperatures 
immediately over the pavement, still pouring out its stored-up heat, 
were from 5° to 8° higher than over the ground, which had radiated 
its near-surface heat rapidly. Just after sundown, the two were 
about equal. 
There is still left the question as to why the superheated air over 
the ground and so much of the pavement was not mirage forming. 
The ground had many depressions which seemed to have no value 
at all for that purpose. Pursuing the question a little further and 
not having cement paving available for testing, I took temperatures 
over the cement sidewalk. They ran fairly close to those over the 
asphalt. The relatively small number of these mirages over cement 
paving, even when blackened by oily exhaust, may be explained by the 
method of laying. The asphalt is rolled on, making quite possible 
the formation of across-street depressions. 'The cement is poured on 
and leveled while wet, eliminating most such inequalities. 
In warmer weather, the seeming water in the little strips fre- 
quently spreads until it covers the whole street, gutter to gutter, 
for several blocks. At about a hundred yards only the normal] strips 
will be seen; then at approximately double that distance, the spread 
area will begin. This will occur only where there are already 
strips—never, in my experience, in other areas of the pavement. It 
is evident that the angle of refraction for the responsible air lenses 
is only half as great for the spread area as it is for the strips. 
In the spring of 1959 use of 7X50 binoculars in the study brought 
added information on the makeup and behavior of these street 
mirages. Kneeling or sitting, to meet the light path lower down, 
one may cut the observing distance from them in half. The mirage 
itself, of course, gives us the shape and size of the air lens. In 
effect, the transmitted light colors the air forming the lens much as 
a stain brings out the details of a microscope slide. One might 
almost get the effect of looking at the mirage through a microscope. 
Particular attention is called to a study made early in June of one 
of the spread areas. The nearer strips of seeming water were in 
evidence, but the larger mirage covering the street from gutter to 
gutter and extending, with some breaks, for about three blocks was 
most interesting. A brisk south wind, 12 to 15 m.p.h., was blow- 
ing across the street. Street surface was asphalt about 35 feet wide 
with cement gutters 18 inches wide. The spread-area mirage and 
the air lens that caused it had irregular across-the-street boundaries 
but sharp lateral ones. On both sides, as it reached the line between 
