The article discusses the definition of chronological corpora and lexical series generated thanks to them; moreover, it describes various conceptualizations of categories of time in corpus studies. Chronological corpora are defined in opposition to general and historical corpora. Time is presented in structural perspective (continuity, circularity), anthropological perspective (astronomical, cultural, political, civilizational time) and methodological perspective (models of time series analysis). Taking into account specificity of lexical and statistical data, a model of anomaly (catastrophe) expressed in sudden increase or decrease of lexeme frequency has been added to used models of sequential phenomena (linear trend, periodic oscillations and random variations). Basing on the analysis of lexical series, the method of automatic identification of potential keywords (they follow a model of anomaly) and neosemantisms (they follow a seriously disturbed model of a steady trend).