IOWA ACADEMY OF SCIENCE 65 



present, the members of the outer cycle are always smaller than those 

 of the inner. Figures 20 and 21, being sectional drawings of the same 

 flower, illustrate this point well; figure 20 is the outer stamen opposite 

 the petal and figure 21, the inner one, opposite the sepal. 



The outer cycle of stamens is thus unusual in several respects: first, 

 its members are opposite the adjacent petals, which interferes with the 

 normal alternatiton of the parts; second, it is often entirely svippressed 

 while the inner whorl of staminodia is well developed; and third, when 

 both whorls are present, the outer one is always more rudimentary than 

 the inner. In regard to the explanation of the first of these, a number 

 of theories have been proposed in an attempt to bring this obdiplos- 

 temonous arrangement into harmony w^th the normal alternation of the 

 parts. According to Braun^ a row of leaves which he regarded as the 

 inner corolla has disappeared between the outer stamens and the petals, 

 and the rudiments of these lost parts could sometimes be seen in abnor- 

 mal flowers, as well as in the five glands w^hich in a geranium flower 

 next succeed and alternate with the petals. Payer and Sachs* have 

 adopted the hypothesis of St. Hilaire which assumes that the epipetalous 

 Ktnmens belong to the eorolline whorl through the transverse splitting 

 of the latter. More recently the hypothesis of Celakowsky'^ regards the 

 epipetalous stamens as really the inner or second circle, which, in the 

 course of development has become external by displacement. These, in 

 general are the three main theories of which there are many modifica- 

 tions. In the species under consideration there is no possible chance, 

 it seems, of applying either the first or the third of these, for there is no 

 evidence whatever of a circle being lost, or that there has been a dis- 

 ]>la('ement of circles. Both the arrangement and the fact that the 

 stamens as well as the staminodia of the outer whorl are smaller than 

 those of the inner, and also that they are the first to abort, might suggest 

 that the outer are younger, phylogenetically, than the inner. In the 

 opinion of the writer, all these irregularities may be explained by some 

 theory of interpolation. 



The Staminate Flower. 



Tlie staminate flowers are early characterized by the vigoi-ous growth 

 of tlie stamens. Here, as in the pistillate flower, the inner stamens arise 

 first, grow larger, and mature earlier than the outer (Fig. 15). The 



3. Gray's Structural Botany. 1879. p. 199. 



4. Goebel K., Outlines of Classification and Special Morphology. 1887, p. 418. 



5. Celakovsky. Uber den "eingeschalteten" epipetalen Staubgefasskreis. Flora 

 58. p. 481 ff., 1875. 



