386 XXVIII. THE ANDBCECIUM OF ANOIOSPERMS. 



two brighter homogeneous-looking spots (B). These are the 

 two nuclei, of which the one (vegetative) is elliptic, the other 

 (reproductive) is elongated and bent, and almost fills the re- 

 productive cell. The other contents of the pollen-grains are 

 pretty uniformly finely granular. The pollen-grains after some 

 time begin to flatten, whereby the nuclei, together with the 

 contents, are squeezed out. The two nuclei can be seen very 

 beautifully if the pollen-grains are crushed in a drop of acetic 

 iodine-green. The generative nucleus stains more deeply, and 

 in coming out often elongates considerably. If the pollen- grains 

 are placed in the reagents in question, but without crushing, 

 the nuclei show in their natural position inside the grain the 

 vermiform nucleus always staining very deeply; the elliptic, on 

 the other hand, somewhat more feebly. The rest of the pollen- 

 grain remains at the same time unstained. If the pollen-grains 

 in water have a drop of iodine solution added, \ve see, after 

 crushing the grain, numerous small blue starch granules in the 

 extruded yellow- brown contents. 



If we go back to the younger flowers, and remove the anthers 

 from a bud about \ inch long, and crush them in water, we shall 

 see part of the pollen-grains with one nucleus, and part, as in 

 Fig. 141, C f , with two nuclei lying close together. These two 

 nuclei are, however, separated by a convex partition wall, which 

 encloses one nucleus together with ;i littlr protoplasm. Thi> plano- 

 convex cell, in basal outline almost circular, lies always upon the 

 flatter side of the grain, which later on is opposed to the fold, 

 or furrow. In somewhat older flower-buds we can see that 

 this cell has separated from the wall of the pollen-grain, and 

 lies free in the contents of the grain. It has elongated, and 

 proportionately thinned, and at the same time tapered at both 

 ends ; with the exception of the two ends, it is filled by its 

 nucleus. It is this cell which, later on, effects fertilisation, 

 always previously undergoing bipartition in the pollen-tube, The 

 staining distinction between generative and vegetative nucleus 

 is as a rule much more marked than in Tradescantia. 



The species of Leucojum (Snowflake) agree with Tradescantia. 

 The Crown Imperial, Fntillaria imperialis, presents another 

 favourable object for investigation. In sections of alcohol mate- 

 rial, first placed in water, and then stained with logwood, the 

 two nuclei in each pollen-grain can be seen very beautifully. 



Polkn of (Enothtra. If we open a bud of (Enothera I 



