In the middle of the 20th century, animal tumor viruses were heralded as possible models for understanding human cancer. By the mid-1970s, the molecular basis by which tumor viruses transform cells into a malignant state was beginning to emerge as the first viral genomic sequences were reported and the proteins encoded by their transforming genes were identified and characterized. This was a time of great excitement and rapid progress. In 1978, prompted by the discovery from Ray Erikson's group that the Rous sarcoma virus (RSV) v-Src-transforming protein had an associated protein kinase activity specific for threonine, my group at the Salk Institute set out to determine whether the polyomavirus middle T-transforming protein had a similar kinase activity. Here, I describe the experiments that led to the identification of a kinase activity associated with middle T antigen and our serendipitous discovery that this activity was specific for tyrosine in vitro, and how this in turn led to the fortuitous observation that the v-Src-associated kinase activity was also specific for tyrosine. Our finding that v-Src increased the level of phosphotyrosine in cellular proteins in RSV-transformed cells confirmed that v-Src is a tyrosine kinase and transforms cells by phosphorylating proteins on tyrosine. My colleague Bart Sefton and I reported these findings in the March issue of PNAS in 1980. Remarkably, all of the experiments in this paper were accomplished in less than one month.