June 26, 1914] 



SCIENCE 



931 



2 3 4 5 



6 7 8 9 10 II 12 



Fig. 2. E.-W. Mean of all observations, semi-daily lunar tide. Dotted curve, observed values; 

 full curve, 0.7 of calculated. 



01234567 



9 10 II 12 



Fig. 3. N.-S. Mean of all observations, semi- daily lunar tide. Dotted curve, observed values; 

 full curve, 0.5 of calculated. 



refinement has not heretofore been necessary 

 in work on the tides, but it introduces differ- 

 ences of as much as 3 per cent. In its final 

 form Professor Moulton's formula, including 

 harmonics of the third order, took account of 

 all factors which affected the result by as 



much as 1 per cent., the probable accuracy of 

 the experimental work. 



In comparing the observed and computed 

 tides in the E.-W. pipe. Professor Michelson 

 found that if the computed tides were multi- 

 plied by the factor Y/10 they agreed almost 



