The idea that memories could be transferred from one organism to another by administration of a "trained" donor brain to a naive recipient seized both scientific and public attention in the 1960's and early 1970's. Georges Ungar was one of the earliest and strongest proponents of this idea, and he provided it extensive theoretical and experimental support. This paper reviews Ungar's work on memory transfer (and in particular on the scotophobin molecule), with an analysis of its successes and failures.