686 KANSAS CITY REVIEW OF SCIENCE. 



about 24° while the sun's is 36°, accounts for the shorter time that the lunar 

 series can be seen from the earth. And the earth's shadow where the mo6n 

 passes it being smaller than the earth itself, accounts for the shorter ecliptic. 



As an example to illustrate the application of the Saros, let us take a total 

 eclipse of the sun which occurred in 1806, June i6th. It was a very large total 

 eclipse and visible in Ohio and New York about eleven o'clock A. M. But in- 

 stead of using our common twelve-hour division of the day, it is better in calcula- 

 tions of this kind to count the twenty-four hours through, and thus avoid the use 

 of A. M. and P. M., and for the convenience of western readers I will use Kan- 

 sas City time, which is about one hour earlier than eastern Ohio, etc. Then the 

 date to start with would be 1806 6mo. i5d. 22h. In determining the leap-years I 

 have found it convenient to write them in a vertical column, beginning with 4, 8, 

 12, 16, 20, etc., all through the century. This shows every year which has twen- 

 ty-nine days in February, and makes it quite easy to ascertain whether we must 

 use 10, II or 12 days in any particular Saros. The same column will do for any 

 century. 



We observe that the first period, 1806 to 1824, 

 has five leap-years; that is, the 29th day of February 

 occurs five times before the period ends, which is in 

 June. So we must add only ten days with the eight- 

 een years. It may be observed that if the Saros ends 

 before the 29th day of February, that leap-year is 

 not to be counted in that Saros; but if after that 

 day it must be. It may be noted also that the seven 

 hours and forty-three minutes is not always the exact 

 part, or fraction, of a day to the actual occurrence 

 of the eclipse. It is only the mean, or average, part 

 of the day for all periods. There are slight irregu- 

 larities in the moon's motions which can only be got at by a tedious process of 

 calculation. So that the nearest hour is quite sufficient for our purpose. Hence 

 we add eight hours twice (in two different periods), then seven hours once, and 

 so on. So we obtain 1824 6th mo. *26th day, six o'clock in the evening, Kansas 

 City time, for the date of the return of the 1806 eclipse. The sun not being 

 then set the eclipse must have been visible over northwestern America and the 

 Pacific Ocean. 



You may observe too, that, the eclipse being nearly eight hours later in the 

 day than the preceding one, it must be about one-third of the way round the 

 earth farther west than its predecessor. The next Saros having but four leap- 

 years, we add eleven days to six months and twenty-six days and have six months 

 and thirty-seven days, from which subtract the thirty days of June, and we ob- 

 tain for the next eclipse, 1842, July 7th, two hours after midnight. Of course it 

 could not be seen here, the sun being away below the eastern horizon. But in 

 the early morning of July 8th, the people of eastern Europe had a fine view of 

 one of the grandest eclipses of modern times. It was then that the red flames 



