June 26, 1914] 



SCIENCE 



929 



and subtracting the corresponding difference 

 at the other end. As the water moved, e. g., 

 toward the east end of the pipe, in response to 

 the tidal forces, the difference (E) at the east 

 end increased, since the water was rising higher 

 above the pointer, and the difference (W) at 

 the west end diminished as the water fell there. 

 The difference E-W therefore increased from 

 hour to hour. When the water moved toward 

 the west end of the pipe, the difference (E) 

 decreased and (W) increased, and E-W dimin- 

 ished. The values of E-W plotted as ordinates 

 against the time as abscissas, gave curves which 

 represented the observed tides. Since the 

 water moved down at one end and up at the 

 other, and since the reflected images moved 

 twice as far as the water surfaces, a four- 

 fold magnification was secured. Moreover, 

 since any change in level which might be 

 caused by small changes in temperature or 

 leaking altered the level at both ends alike, 

 the errors caused by such disturbances dis- 

 appeared in the double difference E-W. 

 Changes in level due to the settling of the pipe 

 between the ends were also without effect. If, 

 however, one end settled more than the other, 

 the double difference would increase or de- 

 crease, depending on whether the east or west 

 end settled more rapidly, and the whole curve 

 would be given an upward or downward slope. 

 A similar effect would also occur if there were 

 an appreciable change in the slope of the 

 rock strata. But since such changes were not 

 periodic, they would not affect the result 

 seriously. 



About eighteen or twenty readings per day, 

 at intervals of from one to three or four hours, 

 were taken during August, and it became evi- 

 dent that the method was capable of yielding 

 very accurate results. A similar line was 

 therefore installed in a N.-S. direction. Dur- 

 ing October and November readings were 

 taken by the writer on both the N.-S. and 

 E.-W. lines, with the assistance of Mr. Harold 

 Alden, of the Terkes Observatory staff. Read- 

 ings were taken once an hour from 6 a.m. to 

 12 P.M. and once in two hours from 12 p.m. 

 to 6 A.M. About four minutes usually elapsed 

 between readings at one end of a line and the 



other. The mean time of the two was taken 

 as the time of the observation. The result was 

 the same as if the two ends had been read 

 simultaneously at the mean time. The obser- 

 vations began at 8 a.m., September 27, and 

 ended at 2 p.m., November 29, 1913. 



The curves representing the tides in both 

 the E.-W. and N.-S. pipes were very satis- 

 factory. They showed with great faithful- 

 ness tides of the expected form, including the 

 diurnal inequality, and spring and neap tides. 

 The actual change of level at each end at 

 spring tide amounted to about 0.001 inch. 



But securing these observations was not all 

 of the problem. It was necessary to know what 

 the tides in the pipes would have been if the 

 earth were absolutely rigid. Computations to 

 determine them were made by Mr. W. L. Hart 

 under the direction of Professor F. R. Moul- 

 ton. The tidal forces acting on the water in 

 the pipes depend on the positions of the sun 

 and moon relative to the observer. These 

 positions change in a very complicated man- 

 ner. In the first place, the moon rises in the 

 east and travels westward across the sky be- 

 cause of the earth's rotation. The moon has 

 a motion eastward among the stars, completing 

 a revolution around the earth in a month. Be- 

 sides this eastward motion, it makes each 

 month an excursion from 28° (at the present 

 time) north of the celestial equator to 28° 

 south of the celestial equator. Its distance 

 from the earth varies by about 10 per cent, 

 during the month, and its eastward motion 

 among the stars is far from being uniform. 

 In addition to all these things the attraction 

 of the sun on the moon causes its motion to 

 be more irregular than it would otherwise be, 

 and it never moves around the earth twice in 

 the same orbit. When all of these complex 

 changes are properly compounded with the 

 almost equally complex ones coming from the 

 sun, the variations in the actual tidal forces 

 are obtained. 



Fortunately the Ephemeris gives us the 

 positions and distances of the moon and sun 

 for every hour, thus saving an enormous 

 amount of computation. Even with this aid 

 the work of computing the tides in the pipes 



