52
 

ALBERUNTS INDIA.
 

Method of
the Arabic
canon
Al-harkan.
 

The divisor is the number of sevenths of the time of
one ctdhimdsa, which he reckons as 32 months, 17 days,
8 ghati, and about 34 cashctkct.

Further, we write down the lunar days in two diffe¬
rent places. The lower number we multiply by 11,
and add to the product 514. The sum is 2,147,164.
Dividing it by 703, we get the quotient 3054, i.e. the
Unardtra days, and a remainder of fff. We subtract
the days from the number in the second place, and
get the remainder 192,096, i.e. the civil days of the
date on which we base the chronological computations
of this book.

The theory of Varahamihira comes very near that of
Brahmagupta; for here the fraction at the end of the
number of the adhimdsa days of the gauge-date is J-f,
whilst in the calculations which we have made, starting
from the beginning of the kctlpct, we found it to be \%^,
which is nearly equal to W (cf. p. 29).

In a Muhammadan canon or calendar called the canon
Al-hctrkctn we find the same method of calculation, but
applied to and starting from another era, the epoch of
which must fall 40,081 (days) after that of the era of
Yazdajird. According to this book, the beginning of the
Indian year falls on Sunday the 21st of Daimah of the
year no of Yazdajird. The method may be tested in
the following manner:—

" Take seventy-two years, change them into months
by multiplying themi by 12, which gives the product
864. Add thereto the months which have elapsed
between the ist of Sha'ban of the year 197, and the
1st of the month in which you happen to be. Write
down the sum in two different places. Multiply the
lower number by 7 and divide the product by 228.
Add the quotient to the upper number and multiply
the sum by 30. Add to the product the number of
days which have elapsed of the month in which you
are.    Write down this number in two different places.