104 



DISPLACEMENT INTERFEROMETRY APPLIED TO 



for a half hour or more. The observations usually began at about 10 a. m. 

 and were finished after 6p.m.; it is thus convenient to distinguish a.m. and 

 p.m. excursions, as will be done presently. The date of each gioup is marked 

 on the graph. 



The first feature of the results is the gradual drift into larger figures, but 

 this may be incidental. For the present it is of no interest, as it would have to 

 be interpreted in the future and does not affect the triplets. 



The next feature is the enormous variability of the excursions. These are 

 given for a. m. and p. m. times in figure 1 2 1 , at first including the mean values. 



The needle of the apparatus pointed north-south ; consequently the weights 

 M took position east- west of the shot. On the west side was the large central 

 pier in the middle of the laboratory to which the case of the apparatus was 

 attached. On the west side, at a distance of 4.5 meters, was the heavy base- 

 ment wall of the laboratory, 2.5 feet thick, and this was illuminated on the 

 outside in the morning only, by the sun if shming. When M is on the east, the 



deflections recorded are toward large numbers. One notes the generally 

 greater variability of the top readings as compared with those at the bottom 

 (M west). This is particularly marked during the hot weather near the end 

 of July. 



A second apparatus (No. II, below) was afterward placed on the east-west 

 wall of the pier, the needle pointing in that direction, and the variability, 

 though still very marked, was less excessive. 



It is thus clear that in figures 119, 120, we are confronting a pervasive 

 temperature or radiation effect. To ascertain the extent of this, figure 121 

 may be consulted, as it exhibits the mean excursions Aj, a. m., p.m., and 

 their means for the successive dates. On the cooler days there is little 

 difference in the a. m. and p. m. readings, and these were therefore at first 



