1892.] NATURAL SCIENCES OP PHILADELPHIA. 169 



carpel, is fertile, we may look for arrangements to insure self- 

 fertilization. But the anthers do not seem to mature before the 

 expansion of the petals, and do not appear to be well situated to 

 pollinate the stigmas. The flowers have no odor to attract insects, 

 nor do I find after many attempts to discover them, that any insects 

 visit the flowers. The flower stems droop at night-fall, and I have, 

 heretofore, thought that, in the act of drooping, pollen falls from the 

 anthers to the stigmas ; and I meet with no suggestions warranting 

 a modification of this belief. 



On the CHARACTER OF THE StAMENS IN OrNITHOGALUM 



UMBELLATUM. 



It need scarcely be noted that the floral whorls of Liliaceous 

 plants are in sets of three, though it is often difficult in the six- 

 leaved perianth to distinguish the three-petaled from the three- 

 sepaled series. 



In Ornithogalum umbellatum each three are well defined on the 

 outer and the inner whorl, though there is no difference between the 

 two except in the smaller size. The filaments are petaloid. and 

 the outer whorl of three follows the character of the petals in being 

 of a still smaller size. But when we come to the fourth series, or 

 interior set of three stamens, they are found to be larger in their 

 filaments than the three before them. 



It is so unusual to find an inner series of stamens more petaloid 

 than the outer, that the fact is certainly worth recording. 



Few of this genus are odoriferous : this is one of the exceptions. 



Note on Barbarea in connection with Dichogamy. 



So far as I am aware no botanist but myself has ventured 

 to explain the cause of dichogamy. I have shown that stamens are 

 called into active growth under a much lower temperature or a 

 less enduring warm temperature than pistils. Hence a flower which 

 may be proterogynous under a continuously warm period late in 

 spring would be proterandrous under the fitful advent of a few warm 

 days in other seasons. 



I have long ago called attention to the fact that the Barbarea 

 vulgaris is proterogynous while Barbarea prcecox is proterandrous. 

 At that time I had no clue to the reason for this great difference 

 between two species so closely allied that botanists have usually to 

 wait until the fruit is nearly mature before they can positively distin- 

 guish them. 

 12 



