258 Transactions of the Royal Society of South Africa. 
data, and is thrown right to the north by Newcomb's data, and in all 
cases does not reach its greatest phase until after sunset. 
Eclipse of —956 May 31st. According to the systems of correction 
Nevill I. and Cowell, the shadow of this eclipse passes from 5° to 7° north 
of Babylon, but according to the systems Nevill II. and Newcomb it 
passes 2° to the south of Babylon. Hence a permissible system of cor- 
rections intermediate between the systems Nevill I. and Nevill II. can be 
deduced, which will make this eclipse total at Babylon, it being only 
necessary to reduce the assumed value of the secular acceleration in mean 
longitude from +7-43" of Nevill II. to +7-10". The day of the eclipse, 
May 31st, though early, is a possible date for the twenty-sixth Sivan ; but 
the year —956 presents difficulties, for Marquart and Lehmann-Haupt 
alone of all authorities would place the seventh year of Eulbar-shakin- 
shum at so late a date as this, nearly all the others placing it from forty 
to eighty years earlier. 
From these results it would appear that the only eclipses which can 
be held to correspond with the eclipse of the sun supposed to have been 
recorded at Babylon on the cuneiform Tablet No. 35968 are these occurring 
on — 
-1217 June 5th 
- 1123 May 18th 
- 956 May 31st 
and of these three, that corresponding best with the theory and observa- 
tion is that of - 1217 June 5th. 
So far it has been supposed that the data for the secular motion of the 
sun in mean longitude employed in the adopted solar tables requires no 
amendment, so that Al' may be regarded as zero. But if it be supposed 
that the motion of the sun may be undergoing an apparent secular accele- 
ration ; either a true acceleration due to a gradual increase in the mass of 
the sun, or a gradual change in its figure, or to the effect of a resisting 
medium ; or an apparent acceleration due to a gradual increase in the 
length of the day arising from the effects of tidal friction on the rotation 
of the earth ; then by such a change it is possible that other of the above 
eclipses may have been total at Babylon. 
Dr. Cowell assumes that the adopted theoretical expression for the 
mean longitude of the sun requires a correction of -\-4:'0d"T^, and the 
corresponding expression for the mean longitude of the moon requires an 
additional correction of -|-4-69"T2. Under these conditions the eclipse of 
- 1062 July 31st, becomes total at Babylon, though it does not satisfac- 
torily comply with the record owing to the day, July 31st, being too late 
to accord with the twenty-sixth day of Sivan of the record. A still 
smaller correction of A^'=+2-5" would render the eclipse of -1069 
