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



the spindle wall and the structures, more or less membran- 

 ous, which form at varying distances from the nuclear wall 

 in the pollen-mother-cells of Lilimn (Mottier, 1897), Cobcea 

 (Lawson, 1898), and Lavatera (Byxbee, 1900). These 

 structures consist at first of a rather loose weft of fibres 

 (which stain violet with the Flemming stain); later the 

 fibrous weft becomes condensed and compressed so as to 

 form in many cases a thick continuous membrane which 

 stains violet; other cases may usually be found in the same 

 anther, in which the membrane is not continuous or frays 

 out in places into a fibrous weft. These structures stand in 

 no such intimate relation with the nucleus and the spindle 

 as the spindle wall of Agave, and seem in comparison with 

 it irregular and ill defined. That they may have some 

 phylogenetic connection with the spindle wall is not improb- 

 able; if so,they are probably to be regarded as rudimentary 

 structures.^ 



On the other hand, it seems quite certain that many of 

 the important resemblances between animal and plant cells 

 are to be regarded not as homologous but as analogous, and 

 the same is probably true of resemblances in cell structure 

 between different groups of plants. We must therefore 

 exercise considerable caution in assuming homologies 

 between cell structures. 



The development of the genetic fibres is peculiar in that 

 there is no weft stage such as is characteristic of most 

 forms of higher plants hitherto studied; on the contrary, the 

 fibres are radial from the beginning. This may very likely 

 be connected with the presence of the spindle wall. Numer- 

 ous observations incline me to believe that the weft stage is 

 caused by the expansion of the nucleus, which at first 

 causes a tangential elongation of the meshes immediately 

 surrounding the nucleus; as the expansion continues the 

 meshes are broken and form free fibres. It is easy to see 

 how the turgidity of the genetic layer (contained within the 



1 An interesting case is that described by Eisen (1900) for Batt-aclwseps, where a "false 

 nuclear wall" is formed around the daughter chromosomes: within this the true nuclear 

 wall is formed. 



