278 MEMOIRS OF THE NEW YORK BOTANICAL GARDEN 



States that the first division is completed by the formation of a 

 cell-plate across the isthmus. "The formation of the cell-plate of 

 the second mitosis is about in the same way" (p. 139). His figures 

 show constrictions of the protoplast, but in no case do they indicate 

 the semblance of a cell-plate. The most recent work on Magnolia 

 is that of Maneval (41). He makes no reference to the work of 

 Guignard, but says that the division is simultaneous, while his 

 figures are not entirely in harmony with this interpretation. 



In 1 9 14 also there appeared the paper of Samuelsson on 

 Anona and Aristolochia (60) in which he attempts to establish a 

 relationship between rtionocotyledons and dicotyledons on the 

 basis of the occurrence of both bipartition and quadripartition of 

 the pollen-mother-cells in these forms. In Aristolochia he reports 

 successive bipartition, resulting in a monoplanal square or a pyra- 

 midal arrangement of the spores. In Ano7ta he figures a mono- 

 planal square in which the mother-cell-wall is in thickness about 

 one fourteenth of the diameter of the cell. As Guignard found in 

 Magnolia, he reports for Anona that in the telophases of the 

 heterotypic mitosis the division of the cytoplasm begins as an 

 equatorial constriction on the periphery of the cell, which is 

 completed after the homoeotypic karyokinesis. The cytokinesis 

 after the homoeotypic karyokinesis is said to agree completely 

 with the first. He does not refer to a cell-plate in any connection 

 throughout the whole paper. He gives no description of his 

 drawings, which appear to be very poorly reproduced. Shibata 

 and Miyake (63) present a similar case in Houttuynia, which they 

 consider as associated with the abortion of pollen in that form. 



It thus appears that there are a number of dicotyledons which 

 display a type of cell division of the pollen-mother-cells which has 

 been regarded as intermediate between true bipartition and true 

 quadripartition. It is generally recognized, however, that the ma- 

 jority of dicotyledons form their microspores by a quadripartition 

 of the mother-cells; but the details of this process of quadriparti- 

 tion have not been accurately determined, and the data in the 

 literature regarding it is not entirely consistent. Some papers as 

 noted below give evidence of the existence of cell-plates in this 

 cytokinesis, while others indicate that it may be accomplished by 

 furrowing. 



In 1898 Lawson (3 8j i)ublishcd in the Proceedings of the Cali- 



