The individuality of the chromosomes and their serial arrangement, etc. 135 



of organization in the chromosomes of the somatic nuclei in many species, 

 and that as Overton (1909) shows this may extend to a pairing of two 

 complete spirems. 



In Carex aquatilis there is, as has been shown above, a permanent 

 serial arrangement of the chromosomes, especially during the history of 

 the somatic nuclei. In synapsis, however, the series of individual chromo- 

 somes is changed to a slender continuous chromatic spirem. After the 

 contraction and pachyneme stage there emerge the bivalent chromo- 

 somes of diakinesis. The composition of these pairs as influenced by the 

 events of synapsis has already been discussed. 



In the thick spirem and the stages following it, I find no evidence 

 of a strepsineme stage. The rounded chromosomes appear in pairs. They 

 are nearly isodiametrical, as in the case of various insects, with no apparent 

 twisting or crossing of parts. Hence in Carex aquatilis there is no possi- 

 bility of such phenomena as Janssens (1909) describes under his "chiasma- 

 typie" theory. If, however, the loops of the bivalent spirem should not 

 straighten out as the spirem enters the heterotypic equatorial plate but 

 should cross each other, there would be opportunity at this stage for 

 chiasmatypic readjustment of parts of the paired spirems similar to that 

 which Janssens describes between parts of paired chromosomes. 



The double spirem appearing in diakinesis passes, into the equatorial 

 plate where two complete series part. It is, however, possible that in 

 the more or less twisted condition of the double spirem just previous to 

 the formation of the equatorial plate of the heterotypic division, there 

 may be a rotating of certain pairs in the double spirem so that a con- 

 stant association of the same chromosomes may not prevail. The marked 

 persistance of the serial relationship, however, suggests strongly that 

 each of the two series may remain as such throughout all stages of nuclear 

 development. Even though there be a rotation of various pairs there 

 are certain parts of the series in which the chromosomes remain in the 

 same alignment. This decreases the opportunity for a simple chance 

 distribution of the chromosomes in the reduction division. 



From these observations it would appear that the serial arrange- 

 ment of the chromosomes is an important factor influencing the degree 

 of chance distribution of the maternal and the paternal chromosomes. 



The studies here reported were pursued at the University of Wisconsin 

 during the years 1909—11, and at the New York Botanical Garden 

 dming 1911—1912, and throughout the investigation the writer has 

 had the helpful criticism and direction of Professor E. A. Harper. 



