The genus Mirabilis has attracted only slight attention cytologically. Hofmeister (1858) briefly described the structure of the tapetum in Mirabilis jalapa. Tischler (1908Tischler ( , 1929 described pollen development and gave chromosome counts for M. longiflora (see Gaiser, 1930), M. jalapa, M. tubiflora, and the hybrid, M. jalapa X M. tubiflora.. Roten (1927), in an embryological study of the Centrosperms, makes frequent reference to features in the micro-and megasporogenesis of M. jalapa. A thorough study of meiosis in a related form, Buginvillaea glabra, has been made by Cooper (1931).The present problem arose while attempting a genetic analysis of certain members of the genus Mirabilis, without evaluating its merits as cytological material. For the most part it is cytologically as undesirable as it is genetically superb for study. The chromosomes are small and numerous, the amount of spore tissue is small, and the time during which reduction division figures may be obtained is short. However, other features treated below, coupled with the meagre knowledge of its cytology, recommend the genus for study.
MATERIALS AND . PURPOSEThe present investigation began in 1930. Three species of Mirabilis and one species hybrid have been used in this study-namely, Mirabilis jalapa L.; Mirabilis longifiora L., Mirabilis multiflora Gray, and the hybrid, M. jalapa X M. longiflora. In obtaining the hybrid, M. jalapa was used as the female and M. longiflora as the male parent. The reciprocal of this cross could not be effected (see also Kolreuter [Roberts, 1929], and Correns, 1902). The hybrid is larger than either of the parent types, and is more resistant to cold. Only a small percentage of its flowers set seed. M. multiflora was studied for chromosome counts only.This investigation attempts (I) to make a comparative study of the chief features of microsporogenesis in these materials, (2) to determine the chromosome numbers of the species and hybrid, and (3) to determine the cause, if possible, of degenerating microsporocytes and microspores.To obtain proper division figures of the material, it was necessary to collect the buds at intervals over a period of approximately two and onehalf hours prior to and one-half hour after the time of anthesis in the mature buds. In the middle of July anthesis takes place about 6 p.m. at The 0594