EXPANSION AND CONTBACTTON OF CONCRETE. 17 
tliem from closing when the expansion of the adjacent concrete took 
place. The section of road read on July 3 was adjacent to a stretch 
of bituminous concrete laid on a concrete base. On the very hot 
day preceding these readings there was an upheaval at the junction 
of the concrete section with the bituminous concrete section, and the 
concrete base of the bituminous pavement was sheared off to some 
extent. Note that one of the cracks had opened more than one-eighth 
of an inch when the readings were taken. 
Measurements made on August 19 are very little different from 
those of the preceding February, notwithstanding the high tempera- 
ture existing during the August readings. The September 5 and 
June 17 readings also are peculiar in that they show very little change 
from those of the preceding winter, and, moreover, the June 17 read- 
ings still show contraction, notwithstanding the hot weather. On 
June 18, with the temperature not greatly different from that of the 
preceding day, a slight expansion was shown. No definite conclu- 
sions can be drawn from the remaining readings of the third set. 
Some of the cracks which had opened in the preceding winter re- 
mained open during the summer, and others became smaller. 
In April, 1914, an incomplete set of readings was taken and the 
concrete in general showed decided contraction compared with the 
preceding summer. The contraction of the concrete was accom- 
panied by an opening of the cracks. The low temperature here seems 
to have played an important part in influencing the length of the 
concrete. 
The fifth set of readings, taken in the spring and summer of 1914, 
on the whole shows the same characteristics as the third set. In the 
previous July (1913) the section of road which had buckled was cut 
out and filled with three double courses of vitrified paving brick 
with tarred joints. In the spring of 1914 buckling again took place 
at the same spot, and the bricks were removed and the space filled 
with concrete. 
Note that at the fifth set of readings some of the cracks showed 
by actual measurement an opening of nearly one-quarter of an inch 
at the section where maximum expansion took place. The crack 
openings were even wider than during the preceding summer. Un- 
fortunately, no measurements were made over this section during 
the winter of 1914. It is probable, however, that such readings would 
have shown that the cracks had opened very wide at this season and 
then were prevented from closing again because of becoming filled 
with loose material. It will be seen that at other sections of the road 
cracks which were indicated as quite small by some of the previous 
readings had opened wide. Note the large expansion in the crack 
which opened in the plain-cement gravel-aggregate section, read on 
