August 2026 Panchang

A tithi is not a day. It is the time the Moon needs to open another 12° of distance from the Sun, and because both bodies move on ellipses that takes a little under a day at some points in their orbits and a little over at others. The column above prints the tithi current at sunrise, so a long one can be caught by two sunrises in a row and appear on two consecutive dates — a vriddhi tithi — while a short one can begin after one sunrise and end before the next, in which case no date in the month carries it at all: a kshaya tithi. Neither is visible on a single day’s page, because both are properties of a pair of days.

The tithi at sunrise advances 30 places across the month's 30 nights, and one place on 28 of them. On one night a tithi holds through a second sunrise: Shukla Dwadashi, on the 24th and the 25th. On one night the column steps past a tithi no sunrise caught: Krishna Trayodashi, between the 10th and the 11th.

The two are not symmetrical, and one division is the reason. A lunation averages 29.53 days and holds 30 tithis, so the mean tithi runs 23 h 37 m — under a day. A column whose values average less than a day skips more often than it repeats, and across the 24 months this table covers it does: 32 kshaya tithis against 22 vriddhi. Every one of those months has at least one kshaya tithi. Only three months have no vriddhi tithi — April 2026, September 2026 and February 2027 — and August 2026 is not one of them.

Which civil date an observance takes when its tithi never touches a sunrise is a question for the tradition being followed, and this page does not answer it. What it does is print the arithmetic that creates the question, with the dates it happened on.

The nakshatra column stumbles for the same reason and comes out on the opposite side of it. A nakshatra is 13°20′ of the sidereal zodiac — one 27th of the circle rather than one 30th of an angle that opens and closes — and the Moon crosses one in 24 h 17 m on average, because a sidereal month of 27.32 days divided by 27 is slightly more than a day where a lunation divided by 30 was slightly less. So the same comparison runs the other way: across the same 24 months there are 28 vriddhi nakshatras against 18 kshaya, while the tithi column ran 22 against 32. One line of arithmetic, two columns leaning opposite ways, and both of them printed above where you can check the claim.

The nakshatra at sunrise advances 30 places across the month's 30 nights, and one place on 28 of them. On one night a nakshatra holds through a second sunrise: Swati, on the 18th and the 19th. On one night the column steps past a nakshatra no sunrise caught: Magha, between the 13th and the 14th.

Seven months of the 24 skip no nakshatra at all — February 2026, March 2026, September 2026, October 2026, February 2027, May 2027 and October 2027 — while every month in the table skips at least one tithi. Same arithmetic, opposite sign.

Read the Rahu Kaal column straight down and it looks as though the season is throwing the window around the day: in August 2026 it opens as early as 07:24 and as late as 17:30, a spread of 10 h 06 m. Almost none of that is the season. Rahu Kaal is one of the eight equal parts of daylight, and which part it takes is fixed by the weekday — the second part on Monday, the eighth on Sunday — so most of that spread is seven weekday assignments sitting in one column, not a moving sun.

Two quantities in each row really do move, and both can be read off the printed times. The first is the window’s own length, which is the eighth: across August 2026 it runs from 95 to 102 minutes. The second is the distance from sunrise to the window’s start, which is a whole number of those eighths — one of them on Monday and seven on Sunday — so one seasonal change is counted a different number of times on each weekday. In August 2026 that distance on Mondays moves 5 minutes across the month’s five Mondays, while on Sundays it moves 36 minutes across five Sundays.

Those two figures are not in the ratio of the two part numbers, and they are not meant to be: the Sundays and the Mondays fall on different dates, so each row samples the eighth’s drift over its own four or five days. The multiple is the mechanism, not the measurement. Subtract sunrise from the Rahu Kaal start on any two rows of the same weekday above, and what is left of the difference is the eighth, counted that weekday’s number of times.

This is not a fact about August 2026 alone. Monday's distance from sunrise is the least mobile of the seven in all 24 months this table covers, and the widest is always the sixth, the seventh or the eighth part of the day. A site publishing one flat Rahu Kaal for the whole country has the eighth wrong, because an eighth is a fraction of local daylight; a site publishing one flat window for a whole month has the multiple wrong as well.