322 BOTANICAL GAZETTE [APRIL 
We have therefore a theoretical right to expect such phenomena 
as chromatin extrusion, irregular lobing, etc. The supernumerary 
nuclei produced by such means are thus ee a secondary phase 
of pollen degeneracy. 
Summary 
1. Maturation in Taraxacum laevigatum differs from that previ- 
ously described for parthenogenetic species of Taraxacum in early 
prophase, chiefly by showing a split thréad from which twenty-six 
univalent chromosomes segment. 
2. Following segmentation, there may be any of four inter- . 
grading sequences instead of a single uniform sequence as described 
for other parthenogenetic species of Taraxacum. 
3. These sequences are: A, almost typical iastice division 
characterized by perfect end to end pairing of the univalents; 
B, a qualitative division resulting in diads from which the functional 
embryo sacs arise and for which the term ‘‘ameiosis”’ is proposed, 
and in which sequence the univalents are slow in pairing; C, 
a more or less irregular division in which pairing of univalents is 
variable, accompanied by premature elongation of the nucleus and 
defective orientation; D, amitosis in which the nucleus elongates 
very prematurely and the split thread persists after segmentation, 
giving twenty-six X and Y-shaped chromosomes. There is no 
spindle. 
4. These variations are not anomalous, but are traced to an 
increasing degree of inhibition of sex by other forces, to wit, 
chromosome individuality and polarity. 
5. JuEt’s (12) interpretation of maturation in T. officinale, that 
it begins as heterotypic and switches to homotypic, does not apply 
in the present case. 
6. Evidence for parasynapsis in Chicoraceae, so far as predicated 
upon the presence of a dual thread only in sexual species, must 
be reexamined. 
UNIVERSITY OF CHICAGO 
