This article details an experiment that uses neutron radiation to induce variations in roses and discusses the results and their implications. The author irradiated rose buds from several varieties, including Queen Elizabeth x Red Delight, Queen Elizabeth x Cavalier, Pink Sensation, Garnette, and Queen Elizabeth. These varieties were selected because they have commercially desirable traits but lack other important features, such as petal number, color, or disease resistance. The author found that neutron radiation induced a remarkable range of variation in the roses, including increased petal number, elimination of the dominant magenta color factor, and changes in leaf morphology. However, he notes that many of the mutations were biologically defective, producing weaker or sterile plants. This article concludes by arguing that mutations cannot change the basic patterns of the varieties but can only alter phases of their expression, suggesting that truly unique varieties, like the Peace rose, would never arise from the accumulation of mutations. The author also argues that the fact that most mutations are harmful suggests a need to reexamine the evidence for evolution.