levine: somatic and reduction divisions in drosera 135 



position of these bodies on the spindle, very close to the newly 

 formed nuclei, resembles that of the belated chromosomes shown 

 by Strasburger, Juel, Tischler, Beer, Gates, Davis, and others. 

 It is not at all probable that these bodies are chromosomes. 

 Their number is large in some cells and they show no tendency 

 to form central spindles or diminutive nuclei. One can count 

 from nine to ten chromosomes in the daughter nuclei at this 

 stage and as many as four equally large bodies lying in the spindle 

 (Fig. 23) in the same cell. It seems quite improbable also that 

 these bodies have a nucleolar origin. Their number and conse- 

 quently their volume is far greater than that of the nucleoli (see 

 Figs. 13, 34). Yet in the homogeneity of their substance and 

 their color with the triple stain they resemble nucleoles. That 

 they are precipitation products formed in fixation is also quite 

 out of the question. They appear in all three species mentioned 

 above with all the different fixative and staining methods used. 

 These bodies as a rule disintegrate shortly after the first division 

 but they may be left in the cytoplasm until the completion of the 

 second division, when the entire protoplast disintegrates. No 

 evidence appears that these bodies are present at any stage in the 

 second division earlier than the late telophase. 



In the second or homoeotypic division the spindles (Fig. 14) 

 are considerably smaller and are variously situated with respect 

 to each other. The poles are usually single-pointed and the 

 spindles in general are like those of the heterotypic division. 

 The chromosomes are equally distributed on the spindle in the 

 equatorial plate stage. In the early anaphase as shown in FiG. 14 

 they have split lengthwise and are beginning to move apart. The 

 attachment of the mantle fibers is almost in the middle of the 

 chromosome, so that U-, V- and L-shaped figures result. Although 

 these chromosomes are small, in polar view twenty monads may 

 be clearly seen. 



While the majority of pollen mother-cells in anthers at a slightly 

 older stage show three nuclei a considerable number appear with 

 two and four. It is quite evident that in pollen cells with three 

 or four nuclei simultaneous division or quadripartition of the pollen 

 mother-cell must follow. In the case where three nuclei alone 

 are visible the fourth nucleus lies either in the plane above or 

 below, forming a tetrahedron, each nucleus situated at one of the 



