THE GREAT ECLIPSE OF AUGUST 17, 1868. 
267 
Now, suppose that, in fig. 2, SFs'c represent the moon’s 
orbit, the broken curve near it the ecliptic. If the sun and 
moon are together between A and c, but close to c, there is an 
eclipse of the sun, but a very small part of his disc is hidden. 
The moon passes in advance, and about fourteen days or so later 
the moon is in opposition. Now, in this interval, the sun 
traverses about 14°, and is therefore near to B, so that the point 
directly opposite to him, or the centre of the earth’s shadow, is 
near to f. Sioce the moon and the earth’s shadow come 
together near F, there is a total eclipse of the moon. Next, the 
sun and moon come together near to A, and there is a partial 
eclipse of the sun. Thus there are three eclipses in this set, 
and there are no more eclipses for about five months. 
But suppose that the moon and the earth’s shadow are 
together near to n, first on the side of n towards f. Then 
there is a partial lunar eclipse. At the ensuing conjunction, 
the sun and moon are together beyond b (that is, a little 
towards a), and there is a total or annular, or at least con- 
siderable solar eclipse. At the ensuing opposition, the moon 
is beyond G, and there is no lunar eclipse. Hence, there are 
two eclipses in this set. But if the moon had been near to D, 
on the side away from r, at the first opposition, the next con- 
junction would find the sun and moon close to B, and there 
would be a total or annular eclipse. At the ensuing opposition, 
the moon would be beyond G, and there would be no lunar 
eclipse. Thus, at this passage of the “eclipse-month,” there 
would be but one solar eclipse, and that one total or annular. 
A little consideration of these extreme cases will show that 
during the eclipse-month there may be — (1) three eclipses, in 
which case two are solar and partial, the other lunar and total ; 
(2) two eclipses, in which case one or other is lunar, the other 
solar, and either may be total or partial (but both cannot be 
total) ; or (3) one eclipse, which must be solar, and total or 
annular.'’^ Also note that two sets of class 1 cannot succeed 
each other either immediately or closely. Now there intervene 
rather more than months between successive eclipse- 
months. Hence there may be three, and must be two, eclipse- 
months in the course of a year. If there are three, one may 
be of class 1, the other two of class 2, in which case there are 
seven eclipses — the greatest number which can possibly take 
* In this last case, the moon must pass through the penumbra at the epochs 
of full moon on either side of the solar eclipse. Thus, though the Nautical 
Almanac makes no record of the fact, the moon will he obscured by the 
penumbra at llh. 52m. a.m. on August 3, and at 3h. 67m. a.m. on the 
morning of September 2, 1868. The former phenomenon will, of course, not 
bs visible in England \ the latter will, The obscuration will last some time. 
