The Dresden Codex—a 78-page Maya manuscript from the 11th–12th century—contains an eclipse table covering 405 lunar months (11,960 days). Researchers John Justeson and Justin Lowry show the table remained accurate by starting successor tables at month 358, with occasional adjustments at month 223. This procedure limits timing errors to about 2 hours 20 minutes (and under 51 minutes over 134 years), enabling reliable eclipse forecasts across 350–1150 CE. The work, published in Science Advances, underlines the Maya's advanced calendrical mathematics and ritual astronomy.
How the Maya Predicted Solar Eclipses for Centuries — Decoding the Dresden Codex
The Dresden Codex—a 78-page Maya manuscript from the 11th–12th century—contains an eclipse table covering 405 lunar months (11,960 days). Researchers John Justeson and Justin Lowry show the table remained accurate by starting successor tables at month 358, with occasional adjustments at month 223. This procedure limits timing errors to about 2 hours 20 minutes (and under 51 minutes over 134 years), enabling reliable eclipse forecasts across 350–1150 CE. The work, published in Science Advances, underlines the Maya's advanced calendrical mathematics and ritual astronomy.

How the Maya Predicted Solar Eclipses for Centuries
A medieval Maya handbook used to forecast solar eclipses has long puzzled Western scholars. Linguist John Justeson (University at Albany) and archaeologist Justin Lowry (SUNY Plattsburgh) now offer a convincing explanation for how a mysterious eclipse table in the Dresden Codex was actually used.
A precious surviving manuscript
The Dresden Codex, dated to the 11th–12th century, is one of only four hieroglyphic Maya codices to survive European colonization. This 78-page, bark-paper, accordion-style book is richly illustrated and records observations and rules on astronomy, seasonal cycles, astrology and medical knowledge.
Why eclipse prediction mattered
Forecasting solar eclipses—when the Moon briefly blocks the Sun—was a crucial practice in Maya society. Celestial events structured political and religious life: priests and nobles used eclipse timings to perform rites, make offerings and carry out bloodletting ceremonies intended to ensure cosmic renewal.
"Keeping records of events tied to celestial phenomena let leaders anticipate repeating cycles and take prescribed actions," wrote University of Texas historian Kimberley Breuer.
The puzzling eclipse table
One table in the Dresden Codex spans 405 lunar months—equivalent to 11,960 days—and appears to allow eclipse prediction for roughly 700 years. For centuries researchers debated how the table was intended to be continued or reset so it would remain accurate over many generations.
The new interpretation
Justeson and Lowry reject the simple-loop idea—where the table is reset back to month 1 after month 405—because it produces cumulative misalignments that would permit unexpected eclipses in successor tables. Instead, they show the table was likely continued by beginning a successor table at the 358th month of the current table. This procedure keeps predicted Sun–Moon alignments only about 2 hours 20 minutes early.
The researchers further propose occasional larger corrections: on some cycles the following table's first entry would be set at the 223rd month, approximately 10 hours 10 minutes later relative to the alignment. These intermittent adjustments compensate for small, gradually accumulating errors from repeated resets at month 358.
Accuracy and significance
By comparing the codex table with modern eclipse cycles, Justeson and Lowry found this method would let Maya daykeepers reliably predict every solar eclipse visible in their region between about 350 and 1150 CE. They estimate the scheme could be kept viable indefinitely with deviations under 51 minutes over 134 years.
The finding highlights the technical skill of Maya daykeepers and the sophisticated mathematical understanding embedded in their calendrical systems—knowledge developed in service of ritual, political and cosmological life.
Publication: This research is reported in Science Advances.
