404 BOTANICAL GAZETTE [December 



enters the pollen tube in advance of the male cells. He found 

 the tubes in the micropyle, and the two male cells, when they 

 reached the ovule, were still spherical and unaccompanied by the 

 tube nucleus. 



The pollen tubes penetrate the stigma in a mass and pass down 

 the hollow of the style. All penetrate into the same ovary. The 

 tissue seems to be no obstacle, for they could get into the stylar 

 canal without such penetration by a slightly less direct route. 

 In the ovary they spread out, fan-like, over the surface, forming 

 a mat. To reach the micropyle a tube grows up between the 

 ovules, or when the ovules are somewhat isolated it follows the 

 funiculus, or even passes directly across the cavity to reach 

 its destination. Reaching the sac it usually passes along the 

 inner surface of the wall until it gets above the synergids, 

 where it seems to burst, discharging the contents for some dis- 

 tance into the sac. Sometimes one of the synergids is destroyed, 

 but there seems to be no regularity about it. In A, tiiberosa the 

 male cells are in advance of the tube nucleus and near the tip of 

 the tube when it penetrates the stigma {^fig. 47) ^ differing in this 

 respect, according to Gager, from A, Cormiti. In their passage 

 down the tube the male cells advance more rapidly than the tube 

 nucleus, so that the latter reaches the upper part of the ovary 

 about the time the male cells enter the embryo sac. The tube 

 nucleus goes no further, and may be found there after the endo- 

 sperm has begun to form. Experiments with tubes grow^n in 

 weak sugar solutions show, as one might expect from their deli- 

 cacy, that the tubes are very sensitive to changes in osmotic 

 pressure. By slight changes one can cause the contents to flow 

 back and forth in the tubes without injuring the walls. The flow 

 can be watched under a compound microscope. When the tubes 

 are burst by too rapidly decreasing the osmotic pressure on the 

 outside the rupture is always at the tip. This raises the question 

 whether in killed material much dependence can be placed upon 

 the position of the nuclei in the tube, either in reference to one 

 another or to the end of the tube. The killing fluid might easily 

 cause flowing and change in position, and the sensitiveness of 

 the delicate tubes to changes in osmotic pressure makes it seem 



