420 
G. Tliibaut— Con/ributions to the 
[No. 4, 
the thirtieth part of the nakshatra month = one periodical revolution of 
the moon, calculated in a corresponding manner, is 549 kalas. Let us now 
compare with this some of Somakara’s statements. The first half of the 
obscure verse 21 : he explains as stating the measure of 
the tithi to be fifty-six nadikas (or 562 f kalas) and the same sense he finds 
in the first pada of v. 39 : ^T^TTl^r both explanations being foun¬ 
ded on his supposing the number seven to mean seven certain divisions of 
time, which had been mentioned under the name of bhansah—parts of a 
constellation—in verse 15 and were said there to consist of eight parts or 
according to Somakara, eiijrht nadikas. That Somakara’s estimation of the 
duration of a tithi is much too short is manifest, and has been remarked by 
the editor of the Jyotisha already (p. 75) ; it is in contradiction not only 
with Garga’s statements, as Prof. Weber notices, but with the elements of 
the Jyotisha itself. It might perhaps be imagined that this estimation re¬ 
fers not to the average tithi but a tithi which is shortened for some special 
reason; but there is no mention made of such a shortening in any of the 
other verses of the Jyotisha. Leaving however v. 21 aside, we are able posi¬ 
tively to show Somakara’s mistake regarding v. 39. The words 
inn: he explains by ^ 
^frT according to which we should have to translate “ the 
moon together with a nakshatra accomplishes a day consisting of seven 
parts (at 8 nadikas each).” Now firstly, the measure of a tithi having been 
fixed, according to Somakara, in v. 21 already, it would be utterly useless to 
repeat this statement here ; secondly, it does not appear why the conjunction 
of the moon with a nakshatra should be mentioned here, the tithi having 
nothing to do with that matter ; thirdly, cannot well mean “ con¬ 
sisting of seven parts” but must be explained as “ that which is joined with 
a septad.” The right explanation is doubtless as follows : The periodical 
revolution of the moon is performed in 30 X 549 = 16470 kalas (see 
above) ; 16470 divided by 27, the number of the nakshatras, gives us 610 
kalas as the time during which the moon remains in one nakshatra. Now 
610 kalas exceed by 7 kalas the length of a (savana) day which amounts to 
603 kalas. We have therefore simply to supply a word for day (either 
from verse 38 Y or from the “ dyuni” of the second pada of 39 ; (I prefer 
the latter, a different verse preceding v. 39 in the Rik recension) and to 
translate “ the moon remains in a nakshatra for one day plus seven kalas.” 
This translation is moreover demanded by the words immediately following : 
^TiHTPfr (R. ^*Rfa) g which undoubtedly state 
the length of the time during which the sun remains in one nakshatra. 
The reading of the Rik recension is clearly the right one ; for the sun 
passing through all 27 nakshatras in 366 days remains in one nakshatra 
