April, 1921] Reversal of the Sexual State 193 



andrecium of a mature carpellate flower. The stamens are 

 vestigial because of the changed sexual state. Figs. 29-34 rep- 

 resent a series of spikelets from the staminate part of a branch 

 to the carpellate part. The length of the awn is determined in 

 the vegetative tissues, depending on the nature of the sexual 

 state present. Sex limited characters of this sort are not due 

 to presence or absence of factors but sexual states present in 

 the tissues during their development. 



Salix amygdaloides And. Peach-leaf Willow. 



Several years ago the writer discovered a number of remark- 

 able trees of this species in a grove in Kansas.* The trees in 

 question have monecious catkins, being staminate in the lower 

 part and carpellate in the upper, with a wide transition zone 

 from the one sexual state to the other. As was to be expected 

 under such conditions, the transition part of the axis is covered 

 with all sorts of abnormal flowers. Figs. 35-51 represent some 

 of the gradations and confusions of sexual expressions to be 

 observed. They exhibit something of the remarkable patch- 

 work of small areas of tissues in different sexual states to be 

 observed in many monecious infloresceness. They show that 

 not only is the sexual state reversed from the lower part of the 

 axis of the catkin to the upper part but that an organ may be 

 practically intermediate in morphological expression because 

 the tissue is "spotted" in respect to the sexual state. An 

 ovulary may be normally developed in respect to the character 

 of its wall or some of the stigmas and may even have normal 

 ovules which develop into seeds and, at the same time, have 

 certain areas developed into microsporangia. Or a stamen 

 with normal microsporangia may have an imperfect stigma or 

 its stalk may be more or less carpellate in nature, taking on 

 some of the characters of an ovulary. The meristematic tissue 

 in the neutral zone was either entirely neutral, in respect to 

 sex, or the sexual state was so weakly developed that reversals 

 were easily brought about. In either case the incepts of the 

 floral structures were developing according to the activity of 

 the hereditary factors present, determining the position, unity, 

 diversity, and other characters of the parts, but the change of 

 the sexual state in local cells and groups of cells of these incip- 



* ScHAFFNER, JoHN H. The Nature of the Diecious Condition in Morus alba 

 and Salix amygdaloides. Ohio Journal of Science; 19:409-416. 1919. 



