Almost hundred years ago Jerzy Kuryłowicz, the well known Polish linguist, convincingly demonstrated that the Indo-European short vowel *a was secondarily formed by the interaction of PIE. *h₂ with the next vowel *e. There are some instances where this explanation does not apply. The most characteristic example of the secondary root a-vocalism is the Latin adjective magnus ‘great, large’, etymologically related to Greek μέγας adj. ‘great, big, large’ (< PIE. *méĝhh₂s, cf. Arm. mec ‘great’, OInd. mahi- adj. ‘great’, Hitt. mekkiš adj. ‘id.’, Goth. mikils adj. ‘id.’). Lat. magnus demonstrates an unexplained vowel [a] as opposed to PIE. *e. It is suggested that the Latin a-vocalism appeared as the effect of influencing the laryngeal *h₂ on the preceding syllable. In other words, the Latin adjective is created by the following regular development: PIE. *méĝh₂s adj. ‘great, large’ → *meĝh̥₂-no-s adj. ‘id.’ > Proto-Latin *megḀnos > *magAnos > Lat. magnus ‘id.’. The same phonological law (which can be called the magnus rule) should be postulated for other Italic and Indo-European languages, e.g. MIr. maige adj. ‘great’ (< Celtic *magjos ‘id.’ < *megAi̯os ‘id.’ < *meĝh̥₂-i̯o-s ← PIE. *méĝh̥₂s); Toch. A māk, B māka adj. ‘many’ (< CToch. *măkă < *megAs < PIE. *méĝh̥₂s).