CHAPTER XXXV. 349
For the purpose of finding these times they use the
following calculation :—
Multiply the elapsed lunar days of the month, if
they are less than 15, or, in case they are more, the
difference between them and 15, by the number of the
ghatis of the night in question. Add 2 to the product,
and divide the sum by 15. Then the quotient repre¬
sents the number of ghatis and minor fractions of time
between the first night, and either the setting of the
moon in the night in question, one of the nights of the
white half, or the rising of the moon in the night in
question, one of the nights of the black half.
This calculation is based on the fact that the space
of time between the first night and the rising or setting
of the moon in some following night of the same luna¬
tion varies by two minutes (ghati), and that the nights
vary, lasting either a little longer or a little shorter
than thirty minutes. If, therefore, you count thirty
minutes for each nychthemeron, and you divide the
product by half the number of the minutes, you get
two minutes for each nychthemeron. As these two
minutes, however, agree with the difference of the
nights, they multiplied the number of nychthemera
by the measure of the night, i.e. the number of its
ghatis (see above, 11. 6, 7), whilst it would have been
more accurate to multiply b}^ the half of the sum of
the ghatis of the night in question and of the first night
of the lunation. It is useless to add the two minutes, Page 177.
for they represent the moment when the crescent of
the moon first becomes visible, but if this moment were
adopted as the beginning of the month, the two minutes
would be transferred to the conjunction.
As months are composed of days, there are as many various
kinds of months as there are kinds of days. Each month.s.
month has thirty days. We shall here use the civil day
(Sdvana, v, chap, xxxiii.) as a standard.
In agreement with the Hindu calculation of the re-