CHAPTER LI.                                25

divisor of 720, we get 2,160,000 solar days, 2,226,389
lunar days, 66,389 days of the adhimdsa months. If
we, lastly, divide the universal solar, lunar, and civil
days of a caturyuga, each kind separately, by the uni¬
versal adhimdsa months of a caturyuga, the quotient
represents the numbers of days within which a whole
adhimdsa month sums up, viz. in 976^-i-i-i-^ solar days,
in ioo6-g%%%\ lunar days, and in 990-|-^f|- civil days.

These are the elements of the computation of the
adhimdsa, which we have worked out for the benefit of
the following investigations.

Kegarding the cause which necessitates the unardtra, Explanation
lit. tloe days of the decrease, we have to consider the f ol- °(M(«4ra^."^
lowing.

If we have one year or a certain number of years,
and reckon for each of them twelve months, we get the
corresponding number of solar months, and by multi¬
plying the latter by 30, the corresponding number
of solar days. It is evident that the number of the
lunar months or days of the same period is the same,
plus an increase which forms one or several adhimdsa
months. If we reduce this increase to adJiimdsa months
due to the period of time in question, according to the
relation between the universal solar months and the
universal adhimdsa months, and add this to the months
or days of the years in question, the sum represents the
partial lunar days, i.e. those which correspond to the
given number of years.

This, however, is not what is wanted. What we want
is the number of civil days of the given number of
years which are less than the lunar days; for one civil
day is greater than one lunar day. Therefore, in order
to find that which is sought, we must subtract some¬
thing from the number of lunar days, and this element
which must be subtracted is called Unardtra.

The Unardtra of the partial lunar days stands in the
same relation to the universal lunar days as the universal