Variable male sterility was encountered in a spontaneous mutant of Lycoper-sicon esculentum Mill. 'San Marzano.' The trait, conditioned by a single recessive ger4e vms, was proved by trisomic and linkage tests to lie on chromosome 8, probably in the longer arm between the markers bu and dl. Under field conditions, from June to October in the interior valleys of California, vms develops severely reduced, deformed stamens and defective corolla, being otherwise physiologically and morphologically niormal. Despite the production of a limited amount of stainable pollen under these conditions, it is effectively male-sterile. Earlier in the season in the field and throughout the entire year under normal greenhouse conditions, the flowers are normal except for a slight reduction in the width of corolla segments. Differences in temperature account for the varied response of vms to the environment, minimal temperatures of 30 C in the field and of 32 C in the greenhouse being required to evoke the sterile phenotype. Application of high temperatures to a single branch of vms elicits a response only in the subsequently developed flowers of that branch. The approximate minimal requirements for the greenhouse environmeiit are temperatures of 35 C or higher applied (1) for 4 hr/day for 2 weeks, (2) for 8 hr/day for 1 week, or (3) continuously for 48 hr. The vms response was also elicited by 29.5 C applied continuously for 96 hr and by 40.5 C for 12 hr. Depending upon the rate of growth, the response first appears 3-5 weeks after the start of treatment and continues for approximately the same length of time as that of the treatment. In the light of these findings and of the observed rate of development under the same environment, the sensitive stage is estimated to be flower primordia of 0.075 mm or smaller diam. Buds of this size appear very close to the shoot meristem, are still highly mitotic, and precede meiosis in PMC by at least 10 days-a relationship that is in keeping with the deformity observed in anthers of all dissectable sizes. THE MUTANT under consideratioii was discovered as an unfruitful plant (2-219) in a field of the cultivated tomato Lycopersicon esculentum Mill. 'San Marzano' inear Clarksburg, Calif., in October, 1951. At that time the flowers were observed to be partly cleistogamous with abortive anthers. Despite complete unfruitfulness of the majority of inflorescences, a few fruits with seeds had set near the base of the plant. The abrupt change in fruitfulness was surmised to be the result of virus disease acquired in midseason. Subsequent studies have demonstrated the real significance of this preliminary observation. Progeny from seeds set by the original plant were observed in various environments. Plants in the first cultures grown during the summer season of 1952 exhibited the same floral abnormalities and attendant infertility of the original mutant. Rooted cuttings of these plants grown in the greenhouse during the subsequent winter, 1 Received for publication July 11, 1966. Research supported in part by a grant from the National Science Foundation. The assistance of Dr. Leonard Morris in temperature regulation and loan of equipment and of Miss, Dora G. Hunt in preparing slides of the sectioned material is gratefully acknowledged. 2 Present address: Institute of Genetics, University of Copenhagen, Copenhagen, Denmark. however, developed flowers of normal structure and fertility. When the same clones were again propagated and grown in the field culture the following summer season, they reverted to the sterile condition. In subsequent years this differential response to the summer and winter environment was displayed consistenitly, regardless of propagation vegetatively by rooted cuttings or sexually by seeds from the original plant. Since this mutant displays in its fertile phase all the features characteristic of 'San Marzano,' it probably originated in that cultivar by spontaneous mutation. As a result of their iiearly exclusively self-pollinating mating system, tomato cultivars are essentially pure lities or collections of very closely related pure lines. Such an origin therefore provides ideal material for physiological tests because, as we have demonistrated, the mutant differs from the parent line solely in the change of a single gene; i.e., the mutant and parent are isogenic. For the following experiments 'San 1lMarzano' was accordingly utilized as the control line. In consideration of the highly diverse approaches followed in this project, the various methods used are more aDDroDriately Dresented in