NOTICES OF BOOKS AND MEMOlKS. 281 



siuTOuncled antipodal vesicles in Delphinium, Gladiolus, and Crocus, 

 two germs in Santahnn and Sinniiujia, binucleolate germ in 

 Xothoscordium fragrans, occasional abortion to a greater or less 

 degree of one (very rarely of both) of the auxiliaries, and pro- 

 trusion of the embryo-sac into the micropyle in Torcnia asiatica. 

 Attentive consideration was also given to the development and 

 structure of the apical " filamentary j)rolongations " in Gladiolus, 

 Crocus, Bartonia, &c., and the thickening of the upper part of the 

 auxiliaries, which in other cases foreshadow them. 



For studying the act of impregnation and its immediate results, 

 the author recommends Torenia asiatica. The ovules were stripped 

 off the placenta thirty- six hours after pollination, and put in a 

 3 per cent, solution of sugar. Directly the pollen-tube comes in 

 contact with the top of the auxiliaries, it cleaves to them and 

 rebounds as though it would separate from them. At this stage 

 its contents are strongly refractive and uniformly finely granular. 

 The contents of one of the auxiliaries now appear cloudy, and its 

 nucleus and nucleolus disappear ; it then contracts itself a little, 

 and is uniformly filled with a finely granular substance, the second 

 auxiliary either following the example of its fellow or not, accord- 

 ing as it participates or not in the fertilizing lu'ocess. The apex 

 of the i)ollen-tube either remains at the top of the auxiliaries or is 

 prolonged between them, often directly reaching the germinal 

 vesicle. The now irregularly-contoured auxiliaries are next seen 

 clinging as a formless tenaciously fluid mass to the germ, round 

 which a cellulose membrane can be detected. The auxiliaries are 

 ultimately entirely absorbed, their contents serving, as the author 

 supposes, to nourish the young embryo, unless, as is sometimes the 

 case, the aid of one is dispensed with. 



Gloxinia Injhrida and Sinninrfia IJndleyana are also good types 

 for study, but especially Orchids (Orchis pallens, Gymnadenia 

 conopsea) Monotropa and Pi/rola, on account of the germ not being 

 in those i3lants covered by the auxiliaries, as is the case with 

 Torenia. In these the author saw that after the clouded plasma 

 has spread itself on the germ two close-lying nuclei are seen in the 

 latter. The second nucleus cannot be the last remaining nucleus 

 of the pollen-tube introduced as such into the germ, because 

 that always dissolves immediately the tip of the tube touches the 

 embryo-sac. What then is it ? Most probably a new nucleus 

 formed by the re-collection of the particles of the nucleus of the 

 pollen-tube. Both nuclei soon unite, and in one case there were 

 three nuclei in the germ of a Monotropa ovide. Strasburger's 

 opinion is, therefore, that the union of sperm and germ does not 

 take place by osmosis, but by direct passage of the contents 

 of the pollen-tube, first through their own membrane and then 

 through that of the embryo-sac. For this position he finds strong 

 support from Cornu's observation of the development of macrogo- 

 riidia in Xectria, and the processes of impregnation m Ascomycetes, 

 Floridea, and Lichens. 



As for the pollen-tube, it finds itself compressed after impregna- 

 tion by the narrowing of the micropyle, a phenomenon essentially 



2o 



