To the Editor.— It was a welcome relief to read Lipsett's article, "Estrogen Use and Cancer Risk" (237:1112), in which he supports the thesis that estrogens per se are not carcinogenic but are "preparative" for the development of cancer. There is little doubt that the cause of endometrial cancer is some carcinogenic agent (a virus or some other villain) that may well reach the endometrial surface initially as a "chance phenomenon." The important fact is that about ten years must now elapse, as shown by Gusberg 1 and others, before the earliest morphologic change eventuates in cancer. What chance does an endometrium have of developing malignant neoplasms if it is periodically shed, albeit incompletely, as in normal ovulatory cycles? The occurrence of such an event is exceedingly rare (less than 1% of endometrial carcinomas), and even then one wonders if the cancer did not arise in a persistent polyp. Hertig