260 



SCIENCE 



[N. S. Vol. XXV. No. 633 



age of 15 per cent. 0. lata plants, the re- 

 mainder being 0. Lamarckiana, the pollen 

 parent. 



It is found that pollen development in 

 0. lata may proceed to the formation of 

 tetrads. The later stages are very irregu- 

 lar and result in complete degeneration of 

 the pollen cells, usually long before ma- 

 turity. This degeneration may begin in 

 some cases as early as the synapsis stage. 

 Irregularities in the distribution of the 

 chromosomes in the reduction divisions 

 give rise to small extra nuclei in the pollen 

 tetrads, and other conditions similar to 

 those found by Juel in Remerocallis fulva 

 and Syringa Rothonagensis (a hybrid), by 

 Tischler in hybrids of Bibes, by Guyer in 

 hybrid pigeons, and others. The preva- 

 lence of the condition, especially among 

 sterile hybrids, suggests that it may prove 

 to be a distinguishing characteristic of cer- 

 tain classes of hybrids. 



The sporophyte number of chromosomes 

 in 0. lata, as determined in the prophase 

 of the heterotypic mitosis in the pollen 

 mother cell, is 14; while the sporophyte 

 count for the 0. Lamarckiana side of the 

 cross is at least 20. The conclusion from 

 this is that the pure 0. Lamarckiana itself 

 must have over 20 chromosomes. It is be- 

 lieved that differences among the chromo- 

 somes of 0. Lamarckiana and its mutants 

 may furnish a cytological basis for discon- 

 tinu!ous variation, and hence a count of the 

 chromosomes in 0. Lamarckiana and other 

 mutants is to be made at once to settle this 

 important point. 



In both sides of this cross the hetero- 

 typic mitosis in the pollen mother cell 

 shows one, or more commonly two, ring- 

 shaped chromosomes with a peculiar his- 

 tory. They may be designated heterochro- 

 mosomes. In 0. lata their origin has been 

 traced. They arise during synapsis, ap- 

 parently by the cutting off of a loop of the 

 spireme, for in earlier stages of synapsis 



the whole spireme is continuous. One or 

 two of these bodies may be formed as large 

 rings or closed loops some time before the 

 rest of the spireme breaks into chromo- 

 somes. Later they are found, much con- 

 densed in size, on the spindle ; but in meta- 

 phase they have generally wandered to- 

 wards the poles or out into the cytoplasm, 

 where they probably disintegrate. Their 

 possible significance in connection with the 

 phenomena of mutation will not be dis- 

 cussed now. The full paper will appear 

 shortly. 



On the Behavior of the Seedlings of Cer- 

 tain Hybrids of Viola: Ezra Beainerd, 

 Middlebury College, Vermont. 

 Many violet hybrids are found in the 

 second generation to revert in some cha**- 

 acters to one of the parent species, and in 

 other characters to the other parent spe- 

 cies. Illustrations are given of this in four 

 particulars: (1) in leaf-outline; (2) in 

 pubescence; (3) in size; (4) in color of 

 capsule and of seeds. 



The tendency of certain individuals to 

 recover from impairment of fertility is 

 shown. 



Illustrations are given of the way in 

 which species may arise by the attainment 

 in the more marked hybrid forms of fer- 

 tility and stability in reproduction. 



Origination of Species by Hybrids among 

 Wild Plants: D. T. MacDougal, Car- 

 negie Institution. (To be published in 

 the Botanical Gazette.) 



An Instance of Natural Hybridization: 



W. W. RowLEE, Cornell University. 



In 1896, the writer in conjunction with 



Dr. Wiegand, published an article in the 



Bulletin of the Torrey Botanical Club, 



describing as hybrids certain willows that 



grow in a small peat bog near Ithaca. 



The writer was impressed at the time 



