170 CALIFORNIA ACADEMY OF SCIENCES. [Proc. 3D Ser. 



with difficulties. In spite of the fact that Guignard (1891) 

 and other writers have described centrosomes in the higher 

 plants, recent investigations have rendered their existence 

 very improbable. These investigations agree in showing 

 that the spindle in its earlier stages possesses several poles, 

 and later, by the fusion of these, becomes bipolar. In such 

 a process of spindle-formation it is difficult to understand 

 how a centrosome can have any part, nor has any observer, 

 Guignard excepted, claimed to discover centrosomes in 

 connection with it. 



Guignard (1898), however, in a recent article, while 

 admitting that the spindle goes through a multipolar stage 

 in the course of its development, maintains that this does 

 not prove the non-existence of the centrosome. 



In the pollen-mother-cells of Nynifhcea and Ntiphar he 

 describes centrosomes situated at the apices of the cones in 

 the multipolar figures, and also at the poles of the mature 

 spindle. But unfortunately he does not describe or figure 

 a series of stages that would illustrate the behavior of these 

 bodies. Moreover, Strasburger (1897, h) has already dis- 

 cussed this idea and concludes that it is highly improbable 

 that a centrosome plays any part in this process. 



Although multipolar spindles were previously described 

 by Belajeff (1894), Farmer (1893), and Strasburger (1896), 

 their significance was not fully understood until Osterhout's 

 (1897) observations on Equisetum were made known. In 

 this paper, which furnishes us with the most complete 

 series of stages of spindle-formation yet worked out, the 

 process is described as follows: — 



The first thing to be observed in the formation of the 

 spindle in Equisetum is a felted zone of kinoplasmic fibres 

 surrounding the nucleus. These fibres grow out from the 

 nuclear wall and take on a radial arrangement. By the 

 coming together of their free ends these threads form a 

 series of cones. The nuclear wall now breaks down and 

 the fibres composing the cones grow in and become 

 attached to the linin and chromosomes. The apices of the 

 cones now approach each other and arrange themselves in 



