978 RECORD OF CURRENT RESEARCHES RELATING TO 



When the new nuclei begin to round themselves off, the protoplasm 

 again becomes more transparent ; and the granules again collect into 

 a peripheral zone, which includes the nuclei. The nuclear plate makes 

 its appearance suddenly as a dark granular streak, occupying at first 

 only the centre of the cell, and rapidly elongating on both sides until 

 it reaches the periphery of the cell, and the division is complete. It 

 is only after the complete division of the cell that the definite forma- 

 tion of the new nuclei commences, which gradually again attain their 

 original uniform finely granular structure. During the development 

 a considerable absorption of water must take place, the young nuclei 

 becoming less dense and increasing at the same time greatly in volume. 



The four pollen-cells which are formed from a single mother-cell 

 are, in Tradescantia, always produced by two successive bipartitions, 

 the processes being in all essential points the same in the second 

 aivision as in the first. 



The description now given applies to Tradescantia virginica, suhas- 

 pera, and pilosa. In T. discolor the pollen-mother-cells are small, and 

 filled with granular but slightly transparent protoplasm, and are 

 therefore not so favourable for observation. Some peculiarities are 

 presented by T. zehrina. The breaking-up of the nuclear threads into 

 short, nearly oval fragments takes place at a very much earlier period. 

 The nuclear plate has also more the character of a ring, its elements 

 almost disappearing from the centre and collecting near the periphery. 



The nuclear threads so often referred to can be crushed out and 

 their structure examined in water. They are then seen to be longer 

 or shorter, often vermiform threads, the ends of which are always 

 smoothly rounded ofi". Their substance is not homogeneous, but con- 

 sists of a less dense matrix, and a denser portion which assumes the 

 form of an elevated ridge running spirally round the length of the 

 thread, and which can be even separated from its mass. 



Baranetzky observed also the course of the division of the pollen- 

 mother-cells in other plants, both monocotyledons and dicotyledons, 

 especially in Aijapantlms umhellatus, Hemerocallis flava. Yucca gloriosa, 

 Hesperis matronalis, Lathyrtis odoratus, and Pisum sativum. None of 

 these present any differences, except in subordinate points. In Pisum 

 the process is extremely similar to that in Tradescantia zehrina, the 

 breaking-up of the nuclear threads taking place at an early period. 

 In Hesperis the substance of the nucleus is from the first differentiated 

 into rod-shaped, and in Pisum, Hemerocallis, and Yucca into perfectly 

 isodiametrical elements. The filiform nature of the dense elements of 

 the nucleus is therefore not a universal rule. 



In conclusion, the following may be stated as the three most 

 important points in the division of the mother-cells of pollen : — 



1. The differentiation of the mass of the nucleus; that is, the 

 gradual separation of the dense substance, which then assumes tlie 

 form, according to the species of plant, of long vermiform threads, 

 shortish rods, or roundish bodies, which may be called the elements 

 of the nucleus. 



2. The tendency of the nuclear elements to separate from one 

 another in the equatorial plane of the cell, or rather, to approach the 



