Each reign between 1153 and 1249 had within it a prolonged difficulty over the succession. Malcolm IV, the 'Maiden', never married and left no heirs. William I, although he had produced a string of bastards by the 1170s, did not marry until 1186, at the age of forty-three; and no legitimate male heir would appear until the birth of Alexander in 1198. There was a further prolonged period of uncertainty in the reign of Alexander II (1214-49). Although he married Joan, daughter of King John of England in 1221, the marriage had produced no heirs by the time she died in 1238, and it was only in 1241 - twenty-seven years into the reign - that an heir was produced by his second wife, Marie de Coucy. This was Alexander III (1249-86), who succeeded as a minor eight years later. He was the first macMalcolm king since his namesake Alexander I to show a sense of urgency as to one of the basic functions of kingship - he married Margaret, a daughter of Henry III of England, at the age of ten. The supreme irony of the dynastic crisis of 1286 and the succession of a four-year-old infant granddaughter was that it came in the first reign since 1152 which had seemed to escape early from the spectre of a failed succession. A daughter had been born in 1261, a first son in 1264 and another in 1271. It would take the deaths of all three within three years after 1281 to make Margaret, Maid of Norway, the sole legitimate successor in the macMalcolm line.

The twelfth and thirteenth centuries are sometimes seen as marking the 'triumph of primogeniture'. From the death of Edgar in 1107, primogeniture operated throughout the macMalcolm line, despite minors in 1153, 1214 and 1249 and a female minor in 1286. Yet this is to compress history with a vengeance. What marked out the macMalcolm dynasty from the later royal house of Stewart was the number of times it was challenged. The crisis of the 1090s left two possible lines of claimants. One was the macWilliams, descendants of William, son of Duncan II. The origins of the other - the macHeths - are obscure but their threat was nonetheless real, especially in the reign of William I. The accession of Malcolm IV in 1153 is the first firm evidence of primogeniture in operation, but the singular progress round the country made by the new young heir to the throne in 1152, accompanied by the Earl of Buchan, looks rather like an election campaign. William I tried and failed in 1195 to persuade his nobles to accept one of his two daughters as his heir. The fact that he had his son Alexander recognised twice as his heir - in 1201 and again on his deathbed in 1214 despite the fact that by then Alexander was sixteen - testifies to a lingering nervousness about the principle of primogeniture. It was not until 1249, when a boy less than eight years old succeeded his father, that the practice was secure.