IQI9] STOUT— INTERSEXES II5 



been unsuccessful. Pollen grains have been removed from anthers 

 of various ages, anthers have been artificially dried to various 

 stages of dryness before pollen was removed, and many kinds of 

 media have been employed. In extensive tests of pollen from 4 

 dilTerent plants during 2 years of bloom only one germinating 

 grain was found, and this may have been accidentally introduced 

 from another plant. 



Accurate tests of the ability to set seed have not been made 

 for plants that are best classed with this form. From the evidence 

 at hand it appears that the pistils are very frequently functional, 

 so that the plants most typical of this class are functionally 

 female. 



It will be noted later that the flowers of numerous plants 

 which would ordinarily be classed with this form are found upon 

 more careful examination to present somewhat decided differences 

 indicative of various grades of maleness. 



Female or pistillate form (figs. 7-10, 58). — Plants that 

 may be grouped in this class have flowers with rudimentary and 

 rather reduced stamens, the tips of which only slightly or not at 

 all protrude above the corolla. There is much variation, however, 

 in the development of the stamens in such flowers. Frequently 

 there is a dift"erentiation of filaments and anthers as shown in 

 fig. 8, and in cases even some traces of the 4 anther sacs. In other 

 cases the stamens are more foliose, with no trace of anthers, as 

 shown in fig. 10. 



Numerous plants with this general type of stamen have what 

 may be termed "closed" flowers; that is, the corolla lobes do not 

 spread out and become reflexed, and when the flowers are fully 

 developed they appear as shown in figs. 7 and 9, a condition 

 decidedly in contrast to the reflexed corollas seen in such flowers 

 as shown in figs, i and 4. Such a reduction of corolla in pistil- 

 late plants has long been recognized in gynodioecious species, and 

 such a condition was recognized for P. lanceolata in the early 

 observations made by Darwin (7, p. 307) and Ludwig (19, 

 p. 322). Examination of such flowers shows that the blade portion 

 of the petals is well developed, but that the part below the blade 

 is shortened and often crumpled; the corolla lobes, therefore, are 



