No. 454.] STUDIES ON THE PLANT CELL. 737 



the multipolar structure, developing a distinct axis (Fig. 

 becomes bipolar (multipolar polyarch). There is no central 

 body at the poles and no place for a centrosome in this develop- 

 mental history. 



The first detailed study of spindle formation in Angiosperms 

 was, as before stated, that of Mottier ('97) which treated especi- 

 ally of Lilium, Podophyllum and Helleborus. This paper with 

 one by Juel ('97) on Hemerocallis and Osterhout ('97) on Equi- 

 setum effectually disposed of previous views very generally held 

 (Guignard, '91, followed by other authors), that spindle forma- 

 tion and mitotic phenomena in higher plants was involved with 

 the activities of centrosomes or other kinoplasmic centers. 

 Mottier found that the nucleus in the pollen mother-cell of the 

 lily became invested just before mitosis with radiating fibers that 

 shortly after increased in quantity to form a felted web around 

 the structure. Some of the fibers gathered into cones (Fig. 

 14 d] which pointed towards the periphery of the cell so that 

 there resulted, with the disappearance of the nuclear membrane 

 and the entrance of the fibers into the nuclear cavity, a multi- 

 polar spindle (Fig. 14 c). The poles gradually came to lie 

 parallel to one another in a common axis, some of them disap- 

 pearing, so that the spindle generally became distinctly bipolar 

 at metaphase (multipolar polyarch). Essentially the same his- 

 tory was repeated during the second mitosis in the lily. 



From this time on there have been a succession of papers 

 verifying the general conclusions of Mottier and Juel and extend- 

 ing these results to many other forms until now it seems to be 

 well established that centrosomes are never present in the pollen 

 mother-cell and that multipolar spindles, developed from felted 

 stages and changing to bipolar spindles, may be expected in 

 most if not all forms. Guignard ('97 and '98) described multi- 

 polar spindles in several types (Nymphaea, Nuphar, Limoden- 

 dron, etc.), and while he believed that these poles were occupied 

 by granules that sometimes fused to form typical centrosomes, 

 nevertheless he admitted that the multipolar spindle might be 

 formed independently of centrosomes. 



The most important papers on spindle formation in Angio- 

 sperms following those of Mottier ('97 and '98) 'and Juel ('97), 



