COMMON THINGS TIME. 



time, all longer intervals being expressed by its multiples, and 

 all shorter ones by its fractional subdivisions, it is above all things 

 indispensable that its absolute length should be understood with 

 perfect clearness, and ascertained with the most rigorous precision. 

 Like all standard measures, it is necessary that it should have 

 one invariable length, and that this length should be at all times 

 capable of verification by comparison with some natural pheno- 

 mena, observable at all times and places, and which, during 

 an endless succession of ages, past and future, is subject to no 

 change. 



19. It may perhaps be thought that such extreme precision and 

 permanency is needless, and that a departure of the standard from 

 exactness by a very minute fraction would be for all practical 

 purposes unimportant. If the standard, whatever it be, were 

 only applied to the measurement of quantities which are not large 

 multiples of itself, this might be admitted. But it is otherwise, 

 when it forms a very minute fraction of that which it is applied 

 to measure. An error of the ten-thousandth part of an inch in a 

 foot maybe unimportant, so long as short spaces as, for example, 

 the length of a room only are in question. But, if we attempt 

 to apply the foot to measure great distances, the small error 

 multiplied until it swells into one so great as utterly to vitiate 

 the results. Thus an error of the ten-thousandth part of an inch 

 in a foot becomes an error of more than an inch in two miles ; of 

 more than a foot in twenty -four miles ; of more than a mile in 

 120000 miles, and so on. 



If in the measurement of distance vast errors may thus arise 

 from the indefinite increase of small inaccuracies of the standard 

 units by multiplication and accumulation, it is much more so 

 with respect to the measures of time, errors in which, even of the 

 smallest amount, accumulating for ages, would involve not only 

 astronomy but history and chronology in complete confusion. It 

 will therefore be understood how important it is in many points 

 of view, that we should obtain clear, distinct, and settled notions 

 of the import of these terms, days, weeks, months, and years, 

 which constitute our chronometric nomenclature. 



20. "What is a day, the fundamental unit of all time ? In a 

 rough and general way we have defined it to be the interval of 

 time which elapses between two successive returns of the sun to 

 the same point of the firmament. But to observe and ascertain 

 with the necessary precision this interval, it is necessary to have 

 some means of marking a certain point of the firmament ; and, 

 when so marked, of observing the exact moment at which the 

 sun arrives at it. The sun, however, not being a mere point, but 

 a circular space or disc, as it is called, of considerable apparent 



122 



)pa,ruiii; 



