TRANSLATOR'S INTRODUCTION liii 



ance, especially on account of the irregular partition of the 

 solid and fluid on the surface of the Earth, is connected 

 with almost insuperable difficulties, whether on the stand- 

 point of Kant or of Mayer. Mayer has also attempted an 

 approximate estimation of the retardation, and has reasoned 

 from it, according to correct mechanical principles, to the 

 result for the lengthening of the Star day ; and in this he 

 has corrected Kant as much as in the theory itself. He 

 believes that the retarding pressure at the lowest positions 

 must be estimated at not less than 1,000 million kilograms, 

 and finds, according to the rules of mechanics, that the 

 lengthening of the time of the Earth's rotation in 2,500 

 years supposing that the volume of the Earth remained 

 unaltered during that period must amount to ~$ second, 

 and, therefore, in 2,000 years it would be ^ second for 

 which Kant had given 86 seconds. (Kant's " body of water " 

 reduced to weight gives 10,300 million kilograms. As, 

 according to Mayer, the east-westward pressure of the upper 

 tidal mass is*related to the west-eastward of the lower, as 

 14 to 13, the difference thereof is -^th of the total pressure 

 which Kant brings into calculation, and Kant's number 

 should be just 27 times, while it is only 10 times that of 

 Mayer. But notwithstanding this, Kant finds for the increase 

 of the Star day in 2,000 years 1,720 times as great a number 

 as Mayer, from which it appears most clearly that he made 

 a second essential error in passing on to the time. Mayer 

 determines the effect of the Earth's rotation always under 

 the assumption of a uniform density of the Earth at 

 25,840 quadrillion kilogram-metres, while its weight-metre 

 amounts to 6f quadrillion kilograms.) If we compare 

 Mayer's result with the empirically proved increase of the 

 Star-day, which amounts to only ^-th second in 2,000 years, 

 it is about four times the latter amount. Supposing now 

 that the theoretical result is only half correct, so that it is 



