200 Trans. Acad. Sci. of St. Louis. 



I. Pachystemonum. All stamens equal in length or nearly 

 equal. About 400 species. 



II. Lycianthes. Filaments of unequal length, one exceed- 

 ing the other. About 80 species. 



III. Leptostemonum. Filaments of equal length. About 

 400 species. 



IV. Lycopersicum. All stamens equal. About 10 

 species. 



V. Nycterium. Flowers zygomorphic, either with unequal 

 stamens only or with zygomorphic corolla in which the two 

 lower lobes are produced and envelop the large stamen, or 

 stamens, and pistil in the bud. 



According to this arrangement about 94 species would be 

 expected to show a dimorphism in the stamens. He fails to 

 note the frequent difference in size of anther as well as length 

 of filament. So far as I know he has published no list of 

 species which he would refer to his several sections and some 

 parts of his arrangement are not clear to me. So far as I 

 have been able to learn from examination of descriptions of 

 the species of the genus 69 species show more or less pro- 

 nounced structural differences in the stamens, the list furnish- 

 ing a very interesting study in incipient zygomorphy. 



Zygomorphy as a teratological phenomenon in three species 

 and its possible occurrence in a fourth, S. CaroUnense, is 

 certainly suggestive. Occurring in the forms in which it 

 does and in a generally actinomorphic genus precludes the 

 suffo'estion of atavistic reversion in these cases. The heredi- 

 tary nature of many malformations, or mutations, is well 

 known. The advantage of the projection of the lower be- 

 yond the upper stamens might be an immediate one in that 

 the. adaptation to visiting insects would become more perfect 

 and so the aid of natural selection in the fixation of the new 

 form would become immediately effective. Direct ecological 

 work has been done in only one of the strongly zygomorphic 

 forms in this genus* but this is sufficient to show the great 

 difference in favor of the zygomorphic as compared with the 



* Todd. Amer. Nat. 16:281-287. 1882. Harris and Kuchs, Kans. Univ. 

 Sci. Bull. 1 : 15-41. 1 pi. 1902. 



