598 



SCIENCE 



[Vol. LV, No. 1431 



e-alarged ovules were doubtless fertilized while 

 the tiny ovules which were probably but not 

 certainly fertilized were counted as fertilized 

 when they were located among the enlarged 

 ovules and seeds. 



Counts of the aborted ovules in well filled 

 seed capsules resulting from abundant hand 

 pollinations were as follows : 



A. Normal x Normal, 6-9 per cent, (exact 

 average for 5 capsules was 7.6). 



B. Normal x Globe, 10-15 per cent, (exact 

 average for 5 capsules was 12.8). 



0. Globe X Normal, 22-34 per cent, (exact 

 average for 5 capsules was 29.-5). 



D. Globe X Globe, 35-50 per cent, (exact 

 average for 4 capsules was 39.4). 



Since the Globe character is transmitted 

 through the pollen parent in less than 3 per 

 cent of the seeds, the discrepancy in the num- 

 ber of abortive ovules between A and B as 

 well as between C and D suggests that 4-10 

 per cent, of the (w-fl) pollen tutes enter 

 the ovary. More extensive studies will be 

 needed to justify this tentative conclusion but 

 the data at hand seems to indicate this and 

 that there is a much greater mortality of Globe 

 zygotes than of normals in embryonic devel- 

 opment. 



The style of Datura, as in many angiosperms, 

 contains a central core of conducting tissue 

 which is soft and fibrous, made up of narrow 

 linear shaped cells, extending lengthwise of 

 the style and terminating in the stigma where 

 these cells become papillate. The pollen 

 tubes, aided bj- a process of digestion grow 

 down to the ovary through this tissue. 



For a study of the pollen tubes, receptive 

 stigmas were pollinated with a single layer of 

 pollen in order to insure opportunity for uni- 

 form germination. This was done by applying 

 the pollen in moderate quantities and blowing 

 off the excess which was not immediately held 

 by the stigmatic fluid. The styles were re- 

 moved after a given period of time, scalded in 

 hot but not boiling water (about two min- 

 utes) their cortex slit lengthwise by passing 

 them through a groove in which the sharp 

 corner of a fragment of a razor blade pro- 

 truded slightly. This treatment facilitated 

 the removal of the cortical tissue by dissec- 

 tion, leaving only the central strand of eon- 

 ducting tissue with which the stigma is con- 

 tinuous at the end. These central cores were 

 stained in magenta (acid red), washed a little 

 in water and mounted whole on a slide using 

 concentrated lactic acid as a mounting me- 

 dium and clearing agent. Balsam mounts 

 were not found satisfactory but these lactic 

 acid preparations have kept for more than 

 six months. 



Pressure aj^plied to the cover glass will 

 spread tliis tissue out in a thin layer, and the 

 l^oUen tubes may be seen even under low 

 power (better after 12-24 hours) as dark red 

 streaks imbedded among the elongated pink- 

 stained cells of the conducting tissue. Ger- 

 minated pollen grains are transparent and 

 may be recognized only by their empt5' shells 

 (the exiue walls) while the ungerminated pol- 

 len grains will stain a deep red. This method 

 makes possible reliable counts of the number 

 of ungerminated pollen grains and the num- 



FiG. 1. Distribution of the pollen tubes in the styles of 18 Globes combined, and compared 

 with 11 normals. Stigma is at left and the pollen tubes were growing to the right. Values 

 plotted at distance represent the ungerminated pollen grains. 



