7 o BOTANICAL GAZETTE. 



our streets and from its chenopodiaceous character, abundant seed-. 

 ing and the tenacity with which it holds its own in a fair stand-up fight 

 with the street weeds,bids fair to become thoroughly established. One 

 back street near the gas works, where it grew most abundantly, was 

 regraded last winter, and if the coining summer shows that instead 

 of being destroyed it has only been scattered, I think we may fairly 

 count it in as having come to stay. The plant is as symmetrical in 

 habit, when not crowded, as a young Juniper; the foliage at a dis- 

 tance appears soft and graceful (more so than would be imagined by 

 those who have seen only herbarium specimens), and, above all, pre- 

 sents a light yellowish green color which contrasts prettily with the 

 darker bluish green of its associates. — M. S. Bebb, Rockford, 111. 



Dichoo-amy of Ulllbellifeim— I am under obligations to Mr. 

 Trelease for his kind correction of my mistake in calling Umbelliferce 

 proterogynous. It was caused by Gray's Manual, page 187 ! "the 

 styles are protruded from the bud some time before the anthers de- 

 velop," also by personal observations of Erigenia bullosa, Nutt., the 

 only plant of this order which had been investigated by myself pre- 

 viously to Pastinaca. This I consider an unmistakable case of pro 

 terogyny. The stvles protrude from the just unfolding petals for the 

 length of 1 mm. and have an angular divergence of 1-2 mm. At a 

 later stage the stamens expand, the anthers successively dehisce, and 

 during this period the styles approach each other so as to avoid self- 

 fertilization apparently." Frequently this is not effected, and the 

 styles remain expanded. But where they do close during this period 

 they expand again as the stamens grow effete, gradually recurve, be- 

 come effete themselves, turn reddish-brown and wither away. The 

 stamens and petals drop off at the last expansion of the styles. 

 The different flowers vary greatly, but this is the type de- 

 velopment. Not to fall into any error I examined nearly 

 500 specimens. Therefore I consider our Erigenia to be a proter- 

 ogynous plant, generally fertilized before the development of anthers, 

 but capable of being fertilized by its own pollen, although this is 

 apparently avoided in the first part of the second or polleniferous 

 stage. The coming together of the styles is often indistinct and 

 ineffectual. Plants left unfertilized can then be fertilized in the 

 third stage by their own pollen or even by that of other plants still, 

 since that would be more effective. But I believe that it does not 

 take long for the styles to become effete and the carpels are appar- 

 ently fructified before this stage in most cases. Although 1 

 have caught during the last -1 days 5 species of Diptera near these 

 plants, f have never seen any at work, although I have looked at 

 hundreds of plants in their natiye state during the last few days. 

 but the days were stormy and cc/i*d. Still, from the presence of nec- 

 tar in large quantities, f am willing to admit that there may be 

 insects that frequent them. Finally. I am of the opinion (thanks 

 to Mr. Trelease) that this plant (deviating from the family type) 



