I’ve spoken often about the spacing effect — that it’s better to spread out your learning than have it all massed in a block. A study in which mice were trained on an eye movement task (the task allowed precise measurement of learning in the brain) compared learning durability after massed training or training spread over various spaced intervals (2.5 hours to 8 days, with 30 minute to one day intervals). In the case of massed training, the learning achieved at the end of training disappeared within 24 hours. However learning gained in spaced training did not.
Moreover, when a region in the cerebellum connected to motor nuclei involved in eye movement (the flocculus) was anesthetized, the learning achieved from one hour of massed training was eliminated, while learning achieved from an hour of training spaced out over four hours was unaffected. This suggests that the memories had been transferred out of the flocculus (to the vestibular nuclei) within four hours.
However, when protein synthesis in the flocculus was blocked, learning from spaced training was impaired, while learning from massed training was not. This suggests that proteins synthesized in the flocculus play a vital part in the transfer to the vestibular nuclei.
 . Role of Cerebellar Cortical Protein Synthesis in Transfer of Memory Trace of Cerebellum-Dependent Motor Learning. The Journal of Neuroscience [Internet]. 2011 ;31(24):8958 - 8966. Available from: http://www.jneurosci.org/content/31/24/8958.abstract