This article develops an event-based control strategy to achieve finite-time consensus for a chain of nonholonomic multi-agent systems subject to unknown disturbances. First, the controller switching time is determined for the state coupling of the systems in the chain. Then, adaptive event-based integral sliding-mode control protocols with novel triggering functions are proposed, which can avoid the inherent singularity problem. To this end, adaptive laws are designed to deal with the disturbances without knowing exactly their upper bounds, which can adjust the gains of the controllers automatically. Moreover, an algorithm is developed to compute the sliding-mode coefficients, adaptive parameters as well as the controller switching time. It is shown that all states of the systems in the chain can achieve consensus in finite time against disturbances. Furthermore, the Zeno behavior is proved to be excluded by showing that the inter-execution time is lower-bounded. Finally, a numerical example is presented to demonstrate the effectiveness of the proposed framework and methodology.