With the multitude of cloud clusters over tropical oceans, it has been perplexing that so few develop into tropical cyclones. The authors postulate that a major obstacle has been the complexity of scale interactions, particularly those on the mesoscale, which have only recently been observable. While there are well-known climatological requirements, these are by no means sufficient. A major reason for this rarity is the essentially stochastic nature of the mesoscale interactions that precede and contribute to cyclone development. Observations exist for only a few forming cases. In these, the moist convection in the preformation environment is organized into mesoscale convective systems, each of which have associated mesoscale potential vortices in the midlevels. Interactions between these systems may lead to merger, growth to the surface, and development of both the nascent eye and inner rainbands of a tropical cyclone. The process is essentially stochastic, but the degree of stochasticity can be... Abstract With the multitude of cloud clusters over tropical oceans, it has been perplexing that so few develop into tropical cyclones. The authors postulate that a major obstacle has been the complexity of scale interactions, particularly those on the mesoscale, which have only recently been observable. While there are well-known climatological requirements, these are by no means sufficient. A major reason for this rarity is the essentially stochastic nature of the mesoscale interactions that precede and contribute to cyclone development. Observations exist for only a few forming cases. In these, the moist convection in the preformation environment is organized into mesoscale convective systems, each of which have associated mesoscale potential vortices in the midlevels. Interactions between these systems may lead to merger, growth to the surface, and development of both the nascent eye and inner rainbands of a tropical cyclone. The process is essentially stochastic, but the degree of stochasticity can be...