farr: cytokinesis of pollen-mother-cells 277 



karyokinesis, but disappears before cell-division is accomplished. 

 Their figures indicate a tetrahedral quadripartition ; but in only one 

 figure, 54, is a cell-plate indicated, and this after the second nuclear 

 division. The mother-wall in the drawing is about one twenty- 

 second of the diameter of the cell. During exine formation the 

 mother-cell-wall is shown to be much thicker. They explain (p. 

 450) their omission to figure a series of stages in the cell-plate 

 formation after the homoeotypic division as due to a condensation 

 of the cytoplasm about the nucleus, and the formation of a cell- 

 plate so quickly that the different stages could not be found. 



The pollen-formation of Magnolia has been investigated by sev- 

 eral; and apparently it is of another intermediate type. Guignard 

 {26h) found that the heterotypic spindle enlarges by the forma- 

 tion of new fibers. Before it reaches the plasma membrane the 

 fibers become very conspicuous and stain heavily in the equatorial 

 region. A granular plate is not shown to appear in the center, but 

 the fibers are grouped together, leaving clear spaces between them. 

 Soon a ridge appears on the inner surface of the mother-cell-wall, 

 which in optical section resembles a wedge projecting toward the 

 center of the cell. This furrow he likens to those in the cell- 

 division of Cladophora and Spirogyra. The invaginating ring is 

 heavier than in the latter, and the fibers do not thicken enough to 

 account for its ingrowth. As the ring advances the spindle fibers 

 on either side disappear. It does not regularly at this time con- 

 tinue its growth until cytoplasmic division is complete, but stops 

 when the depth of the furrow is about equal to the breadth of 

 the isthmus or protoplasmic bridge remaining between the two 

 halves of the mother-cell. In a few cases the division is 

 said to be complete. The fibers are reformed across the isthmus 

 after the homoeotypic division, and the cytokinesis previously 

 begun is completed usually slightly before the daughter-cells 

 divide. The second division is said to begin like the first, but is 

 more rapid and continues without interruption. He shows no 

 figures of these later stages, which, in fact, constitute the crucial 

 point in establishing the nature of this division. This work was 

 succeeded in 1901 by that of Andrews (2). In the title of his 

 paper he includes both Magnolia and Liriodendron, and he does 

 not indicate in his drawings from which plants they were taken. 

 He agrees in most particulars with Guignard, except that he 



