Short answers, pulled from the story.
A canon is a musical composition technique where a melody begins in one voice called the leader or dux, then another voice enters to copy it exactly after a set delay. The copying voice follows as the follower or comes and must match the leader's rhythms and intervals with sometimes slight adjustments.
The earliest known non-religious canons are English rounds from around 1250 including the famous example Sumer is icumen in which appears in a manuscript source calling it a rota. Walter Odington gave this form the name rondellus at the beginning of the 14th century while the word round first appeared in English sources during the 16th century.
J.S. Bach incorporated canons into his major works by composing several table canons including examples in The Musical Offering and creating double canons that unfold two different themes simultaneously as seen in Bach's Cantata BWV 9. He also composed puzzle canons presenting only one voice with clues requiring solutions to determine remaining parts such as riddles K. 73r using Latin epigrams.
Specific types include inversion canons that move followers in contrary motion to leaders, retrograde or crab canons that accompany leaders backward using retrograde motion, and mensuration canons that alter rhythmic proportions by doubling values or cutting them in half. Other forms include phasing which applies modulating rhythmic proportions according to sliding scales, perpetual or infinite canons allowing voices to restart upon reaching their end, and table canons placing music between two musicians reading opposite directions from the same line.
Haydn wrote a Minuet for String Quartet in D Minor Op. 76 No. 2 that functions as a strict classical canon while Beethoven employed canonic passages throughout his symphonies including a delightfully naïve canonic passage in the first movement of Symphony No. 4. Schumann hid canonic devices subtly in his 1851 piano piece Vogel als Prophet from Waldszenen where the canon is absorbed into the texture so listeners do not easily hear it.