18 BULLETIN 532, IT. S. DEPARTMENT OF AGRICULTURE. 
June 3, 1914, and the comparatively large contraction in the crack 
about 100 feet away. 
Figure 15 plots the results of that part of the Chevy Chase meas- 
urements taken over the uncracked portion of the concrete. The 
ordinates represent the unit deformations of the solid concrete and 
do not include the movement at the cracks, and the abscissas repre- 
sent age. At the bottom are plotted temperature and precipitation 
measurements for the vicinity of Washington, supplied by the 
Weather Bureau. The daily precipitation records were averaged for 
each month, and the average precipitation for the month is plotted 
on an exaggerated scale and is shown by the dotted line. It will be 
seen that there is no decided relation between the shape of the expan- 
sion and contraction curves and that of the precipitation curve. On 
Fig. 14. — The measuring instrument in use on the Chevy Chase Road. 
the other hand, there is a tendency for the expansion and contraction 
curves to conform with the temperature curve. Note that as the 
temperature decreases at the approach of winter the concrete con- 
tracts, and in the summer season, during the highest temperatures, the 
concrete expands. It would seem from this set of measurements that 
temperature changes have played a more important part than have 
atmospheric moisture changes in influencing the movements of the 
concrete. The moisture in the concrete may or may not bear any 
relation to the atmospheric moisture, but will be influenced by 
the degree of wetness of the sub-base. Therefore, no very great im- 
portance can be attached to the precipitation record. Note that at the 
beginning of the measurements, just after the concrete was laid, there 
was a contraction in every case, notwithstanding the fact that the 
temperature remained not far from constant. It would seem that this 
