liv The Transactions of the South African Philosophical Society 
light in interstellar space, the velocity of the earth in its orbit becomes 
known, and hence the distance of the earth from the sun. In 1883 
Michelson showed conclusively by direct experiment, as the theory had 
before indicated, that the observed velocity of light in any medium has 
merely to be multiplied by the refractive index of that medium to 
reduce it to the velocity in vacuo. By a series of experiments in 
1880-82 Newcomb determined the velocity of light with an accuracy 
almost certainly within z,;55 part of its amount, a precision amply 
sufficient for all practical astronomical purposes. This most beautiful 
result has been further independently confirmed by Michelson, and the 
velocity of light may now be regarded as one of the most satisfactorily 
known constants. 
Had this address been delivered four or five years ago, I should have 
been tempted to add likewise that the constant of aberration was 
known with almost sufficient precision. But within that time it has been 
demonstrated by Professor Chandler that the axis of the earth’s rota- 
tion is a changing one ; consequently terrestrial latitudes are not fixed, 
but undergo periodic change to the extent of 0”-12 (or 12 feet on the 
earth’s surface) in a period of 427 days. Many investigators had pre- 
viously endeavoured in vain to prove from observations the existence of 
a change of latitude having a period of about 306 days. This latter. 
period results from the dynamical theorem that the ratio of such a period 
is equal to the ratio of the polar moment of inertia of the earth to the 
difference between the equatorial and polar moments. This ratio is 
obtained from the observed value of precession and nutation. But 
Newcomb has recently shown that this theorem is only true for an 
absolutely rigid body, and that if the earth yields under the action of cen- 
trifugal force in a degree slightly less than if it were composed of steel, 
there would result an increase of the period of the revolution of the in- 
stantaneous axis about the mean axis amounting to about 121 days, and 
thus theory and observation are so far reconciled. But all previously 
determined values of the constant of aberration depend upon the assump- 
tion that the latitude of the observatory does not change, and although 
Professor Chandler has sufficiently re-discussed all the existing cbserva- 
tions which bear with considerable weight on the constant of aberration 
to prove the fact of the change of latitude and approximately to deter- 
mine its law, it cannot be said that his discussion of the value of the 
constant is definitive. Chandler’s investigations go ‘to prove the exist- 
ence of a second periodical change of latitude, having a varying ampli 
tude from 0”°04 to 0:20 and a period of 365 days. Whether this is 
the true form of the expression seems to me somewhat doubiful. It is 
true that large masses of snow are piled up at the poles of the earth in 
opposite seasons, and in other seasons large masses of water are trans- 
