1892] NATURAL SCIENCES OF PHILADELPHIA. 453 



notice that the corolla is about to wilt, the stamens have shriv- 

 elled up, but the style projects at full length from below the upper 

 lip of the corolla ; no pollen was noticeable on the lobes of the 

 stigma, but the lobes are widely separated. The next younger 

 flower, B, shows the corolla still fresh, the style not yet quite 

 exserted, the stamens have wilted, and the stigma lobes are sepa- 

 rated. The next younger flower, C^ has opened and the filaments, 

 which have a fold in the bud, are straightening out ; the style is 

 quite hidden under the upper lip of the corolla. Fig C" represents 

 the style taken out from the upper lip ; the lobes of the stigma 

 are closed. Fig. D, the next stage, represents a bud just opened, 

 and the stamens beginning to protrude. 



Fig. E^ represents a closed bud, ready to open. E^ represents 

 this bud forcibly opened, and E^ shows the anthers united. Fig. F 

 is about the same as E, but here in F" a portion of the corolla is 

 removed in such a manner as to expose the relative position of 

 anthers and style in the closed bud. It was found that the anther 

 cells had discharged their pollen, that the lobes of the style were 

 widel}^ separated, and, in the case represented in the drawing, the 

 longest lobe of the style was actually dipping into the upper cell of 

 one of the anthers. The style, it will be observed, curves over the 

 top of the anthers. Its shape at that time is best seen in Fig. F^ 

 which represents the style (upper portion) taken out from the bud 

 F. Fig. F* represents the same style much more highly magnified, 

 with the pollen attached to it. 



I have noticed pollen discharged in small quantity in extremely 

 young buds, but at present it is impossible for me to say just at what 

 period fertilization begins. The stigma lobes are separated at a 

 very early period. One fact which leads me to suspect that fertili- 

 zation begins at this early period is the presence of peculiar 

 masses which strongly resemble the protoplasmic contents of the 

 pollen grains. These I have represented in C^, the style of C^ highly 

 magnified. These masses could not be detected in fresh material, 

 but were invariably found in styles of flowers which had been kept 

 in alcohol. They are inside of the tissue of the style and usually 

 found throughout its entire length, although they are more or less 

 irregularly distributed. To the right of O^ are three pollen 

 granules, and on comparison with these the yellowish masses in the 

 style seem as a rule smaller, often more or less disorganized, and 

 frequently lumped together. A final decision as to the character 



