10 THE PRESIDENTIAL ADDRESS. 
deforming action of the solar and lunar tidal forces as seriously to affect 
the amplitudes of the oceanic tides, which are a differential effect. Un- 
fortunately, the tides are so much complicated by the irregular distri- 
bution of land and sea that a comparison of the theoretical amounts 
which they would have on the hypothesis of absolute rigidity with the 
actual values is hopeless. The fortnightly tidal component, due to the 
changing declination of the moon, is probably an exception, but the difficulty 
here is to extract this relatively minute component from the observations, 
and the material is consequently imperfect. The problem was attacked in 
a different way by G. and H. Darwin in 1881. The horizontal component 
of the lunar and solar disturbing forces must deflect the apparent vertical, 
and it was sought to measure this effect by a pendulum. The quantities 
to be determined are so excessively minute, and the other disturbing 
forces so difficult to eliminate, that the method was only carried out 
successfully by Hecker in 1907, and subsequently by Orloff in Russia. 
The results on the whole were to the effect that the observed deflections 
were about three-fifths of what they ought to be if the earth were perfectly 
unyielding, and were so far in accordance with estimates previously made 
by Darwin and others, from the somewhat imperfect statistics of the 
fortnightly tide. There was, however, a discrepancy between the results 
deduced from the deflections in the meridian and at right angles to it, 
which gave rise to much perplexity. The question was finally set at rest 
by Michelson in 1916. He conceived the idea of measuring the tides 
produced in two canals (really two pipes half filled with water) of about 
500 feet long, extending one N. and §., the other E. and W. These tides 
are, of course, of a microscopic character, their range 1s of the order of one- 
hundredth of a millimetre, and they could only be detected by the refined 
optical methods which Michelson himself has devised. The observations, 
when plotted on a magnified scale, exhibit all the usual features of a tide- 
gauge record, the alternation of spring and neap tides, the diurnal and semi- 
diurnal lunar tides, and so on. The theoretical tides in the canals can, of 
course, be calculated with great ease, and the comparison led to the result 
that the ratio which the observed tides bore to the theoretical was about 
‘69, being practically the same in both cases. The whole enterprise was as 
remarkable for the courage of its inception as for the skill with which it 
was carried out, and was worthy of the genius which has accomplished so 
many marvels of celestial and terrestrial measurement. The perplexing 
discrepancy in the results obtained by Hecker at Potsdam is no doubt to 
be explained by the attraction of the tidal waters in the not very remote 
