249 
sile, upon a very thick filament not unlike the base of an ordinary 
ovary. This view of the centrally located stamen, sometimes actu- 
ally surrounded by the pistils, is supported by the observation 
of a modified form combining the characteristics of both stamen 
and pistil. For example, an organ was found with one lobe 
complete, while the other terminated midway, being replaced 
below by a row of ovules. 
Figure 1 shows an ordinary Petunia pistil ; at figure 2 isseen one much swollen, 
and figure 3 exhibits a well developed specimen of the remarkable doubling. The 
upper portion consists of protruding, highly colored petals, while in the center of all 
is seen the secondary pistil. At @ is the style and stigma of the primary pistil. 
Figure 4 exhibits one of these stamen-pistils with the anther lobe conspicuous 
upon the right hand, and a number of imperfect ovules to the left and below. 
In figure 5 is shown a section through a similarly deformed organ, with an ovule 
upon the left hand, and the anther lobe and pollen grains to the right. 
From the size and central position, it is evident that these organs are metamor- 
phosed pistils. 
Notes upon Epigza Repens. 
By Byron D. HALSTED. . 
Dr. Gray stated long ago* that the trailing arbutus flowers 
are of two kinds, each with two modifications. The two sorts 
are characterized by the perfection or abortive nature of pistils 
and stamens respectively, and the modifications resided in the 
varying lengths of the essential organs. The two kinds indicate 
a decided tendency towards dicecism, and the modifications sug- 
gest dimorphism. 3 
Of sixty flowers. from as many plants, examined recently, for- 
ty had the stamens predominating, and in these, sixteen had the 
style longest, fourteen with the stigmas and anthers even, and 
ten having short styles. In the remaining twenty, with strong 
pistils, eleven had the style shorter and nine longer than the fila- 
ments. It is evident that while there is a hint of dimorphism it 
is not far advanced. Many kinds of flowers exhibit equally great 
variations and are not dimorphic. More than this, if dimorphism 
was prevailing we should expect to find a difference in the size of 
pollen between the short and long stamens, but this is not appar- 
ent. There are not two sizes as found in dimorphic flowers. It 
seems evident that but little stress should be laid upon the sug- 
*Am, Jour. Science, July, 1876: Am. Nat., Aug., 1876, 
