This paper presents a framework for tracking short, distinctive phrases that travel relatively intact through online text, and develops scalable algorithms for clustering textual variants of such phrases. The framework is applied to the news cycle, providing a coherent representation of the daily rhythms in the news media. The study tracks 1.6 million mainstream media sites and blogs over three months, totaling 90 million articles, and identifies a set of novel and persistent temporal patterns in the news cycle. A typical lag of 2.5 hours is observed between the peaks of attention to a phrase in the news media and in blogs, with a "heartbeat"-like pattern in the handoff between news and blogs. A mathematical model is also developed to capture the kinds of temporal variation that the system exhibits. The study finds that information primarily propagates from news to blogs, and that in only 3.5% of cases, stories first appear dominantly in the blogosphere and subsequently percolate into the mainstream media. The approach to meme-tracking opens new opportunities for studying the dynamics of information propagation and the evolution of topics over time. The results suggest that the news cycle can be analyzed quantitatively, revealing patterns that were previously only qualitatively understood. The study also highlights the importance of independent media in the dissemination of information and the role of blogs in shaping public discourse. The findings have implications for understanding the dynamics of news and information diffusion in the digital age.This paper presents a framework for tracking short, distinctive phrases that travel relatively intact through online text, and develops scalable algorithms for clustering textual variants of such phrases. The framework is applied to the news cycle, providing a coherent representation of the daily rhythms in the news media. The study tracks 1.6 million mainstream media sites and blogs over three months, totaling 90 million articles, and identifies a set of novel and persistent temporal patterns in the news cycle. A typical lag of 2.5 hours is observed between the peaks of attention to a phrase in the news media and in blogs, with a "heartbeat"-like pattern in the handoff between news and blogs. A mathematical model is also developed to capture the kinds of temporal variation that the system exhibits. The study finds that information primarily propagates from news to blogs, and that in only 3.5% of cases, stories first appear dominantly in the blogosphere and subsequently percolate into the mainstream media. The approach to meme-tracking opens new opportunities for studying the dynamics of information propagation and the evolution of topics over time. The results suggest that the news cycle can be analyzed quantitatively, revealing patterns that were previously only qualitatively understood. The study also highlights the importance of independent media in the dissemination of information and the role of blogs in shaping public discourse. The findings have implications for understanding the dynamics of news and information diffusion in the digital age.