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



view it appears round and in the other oblong in outline, 

 corresponding to the position and shape of the spindle. 



In the next figure we see that the spindles are more 

 nearly in the same plane. The mantle-fibres of the two 

 spindles have reached over and have apparently united with 

 one another. In fig. 24 we see three daughter-nuclei in 

 the same plane. By means of the continuous fibres and 

 mantle-fibres they are connected with one another. These 

 fibres now occupy almost the entire cell-cavity. Cell- 

 plates are now formed in the usual way. Swellings appear 

 on the connecting fibres; these increase in size, and finally 

 result in forming cell-walls which separate the daughter- 

 cells from one another. During this stage there appears to 

 be very little of the perikaryoplasm left and what is present 

 is scattered irregularly through the cell. 



Whether the method of spindle-formation observed in 

 Cobaa is of general or exceptional occurrence must be left 

 for future investigation to decide. From Belajeff's (1894) 

 description of Larix it may be supposed that a similar pro- 

 cess takes place here, but the origin of the network and the 

 manner in which the spindle-fibres subsequently arise 

 from it were not sufficiently investigated. The develop- 

 ment of spindle-fibres out of a network, as shown in figs. 

 9-13, recalls Wilson's (1895) description of the formation 

 of astral rays and spindle-fibres in Toxofneiistes. 



This method of spindle-formation differs decidedly in its 

 earlier stages from that observed by Osterhout (1897) in 

 Equisetum, although both agree in the formation of a num- 

 ber of poles which subsequently fuse to form the bipolar 

 spindle. We have in Cobcea a zone of granular substance, 

 the perikaryoplasm, whose function it is to take part in the 

 formation of a network from which the spindle-fibres are 

 developed; no such zone, however, is present in Equise- 

 tum. Similar zones have been figured by Belajeff (1894) 

 (fig. 6) in Larix and Mottier (1897, a) in Liliuni, but 

 they have not been described as taking any part in the 

 formation of kinoplasmic fibres. 



