A name and a body of work
Bhāskara II, born in 1114, is also called Bhāskarācārya. The numeral distinguishes him from the earlier Bhāskara I. His writings include the arithmetic Līlāvatī and algebraic Bījagaṇita. These belong to the mathematical and astronomical programme associated with his Siddhāntaśiromaṇi, completed in 1150. 1 Colebrooke’s 1817 translation makes the arithmetic and algebra accessible in English, although its terminology and historical commentary require critical reading. 2
The most useful introduction is to work through a problem. Two examples below are modern teaching reconstructions, using present-day notation. They illustrate kinds of algebra treated in the Bījagaṇita; they are not quotations or transcriptions of Bhāskara’s calculations. 2
Finding another integer solution
Suppose we want whole numbers satisfying x² − 2y² = 1. The small pair x = 3, y = 2 works, because 9 − 8 = 1. A search through larger numbers would eventually find more pairs, but a transformation is more informative.
Starting with any solution, define X = 3x + 4y and Y = 2x + 3y. Expand the difference:
X² − 2Y² = (9x² + 24xy + 16y²) − (8x² + 24xy + 18y²) = x² − 2y².
The middle terms cancel. The new pair therefore solves the same equation. Beginning with (3, 2), we obtain (17, 12), then (99, 70). Direct checking gives 289 − 288 = 1 and 9801 − 9800 = 1.
This explains why a rule can outperform trial and error: it preserves an invariant, the value of x² − 2y². It also exposes the limits of our example. We have generated a sequence of solutions for one coefficient; we have not proved that we have found every positive solution, or explained a procedure for every nonsquare coefficient. Bhāskara’s treatment of the affected square and cyclic method addresses a wider setting. 2 For the simpler problem of a linear equation in integers, see the kuṭṭaka.
Taking apart a nested square root
Consider √(5 + 2√6). Look for an answer of the form √a + √b. Its square is a + b + 2√(ab), so we need a + b = 5 and ab = 6. Taking a = 2 and b = 3 gives
√(5 + 2√6) = √2 + √3.
Checking by squaring is essential. Because both expressions are positive, matching their squares establishes equality. Without that sign condition, squaring could hide a wrong choice of root. This is a useful reading habit for historical algorithms: distinguish the operation that produces a candidate from the argument that validates it.
Reading without retrospective labels
The arithmetic in these examples is exact, but the notation is modern. A compact symbolic identity is our explanation, not evidence that a medieval author wrote it in this form. Likewise, calling a historical technique “calculus” requires more argument than spotting a familiar-looking expression.
Read Līlāvatī for the relationship between a rule, a word problem, and a learner’s interpretation. Read Brahmagupta for an earlier mathematical setting. Comparing specific procedures gives a richer history than assigning a single modern title to an entire body of work.