In 1912, C.I. Lewis wrote an article in which he alluded to what he saw as defects in the propositional logic at the basis of the system of Whitehead and Russell’s 1910 Principia Mathematica (PM). He objected to their truth-functional use of ‘imply’ and introduced what he saw as a more accurate and stricter sense of that word. The now standard strict implication is necessary material implication, material implication being that of PM. As long as truths of material implication are not confused with those of strict implication, the classical propositional calculus is contained in Lewis’s system. In his 1918 Survey of Symbolic Logic, he provides a first axiomatisation of modal logic in a system intended to replace that of PM. With his strict implication symbol J, he used two negation operators, the ordinary −, and ∼ for impossibility. But he failed to appreciate the modal nature of ∼, as he did later in his 1932 with C.H. Langford Symbolic Logic, where he uses for the first time the word ‘modal’ for his strict operation symbols. Modal logics as genuine logical systems will then develop from there in the works of Becker, Gödel, Wajsberg, Feys and Bayart.