DEPARTMENT OF GENETICS. 93 



Irregular Segregation in Chromosomes. 



In ordinary diploid plants in the reduction division, one member of each 

 pair of chromosomes goes to each pole, so that each daughter nucleus, and 

 hence each gamete, gets the same number of chromosomes. In triploid 

 plants there is a complication, either two chromosomes of each set go to one 

 pole and one to the other, or else the third chromosome might conceivably 

 be left in the middle and not participate in either nucleus. The former 

 alternative is the one actually realized. The question now arises whether 

 the cell into which the odd chromosome shall go is determined purely by 

 chance, or whether there is any sort of attraction between chromosomes such 

 that an excess or all of the extra chromosomes go into one of the daughter 

 nuclei. To decide this matter. Dr. Belling has made counts on the number 

 of chromosomes going to each pole of the spindle in pollen mother-cells, in 

 addition to those reported previously. Over 100 double counts have been 

 made in the regular triploid. The frequency with which each of the series 

 of extra chromosome Nos. to 12 occurs in this 100 agrees with the coefficients 

 in the binomial series (a+fe)^^; the middle numbers of the chromosome 

 series are those most commonly found, but there is a slight excess over expecta- 

 tion of the more unequal assortments. 



In haploid plants the first division has now been examined in several cases; 

 it consists usually in a segregation of the chromosomes into 2 centers, but 

 sometimes into 3. In the division into 2 centers the frequency of occurrence 

 in any center of the number of chromosomes in the series 1 to 12 is, so far as 

 observed, in close agreement with the coeflacients of the binomial (a+&)". 



Two plants from the cross of tetraploid by diploid had 35 chromosomes. 

 These plants provided nearly 100 cases for counting the distribution of 

 chromosomes in the dividing pollen mother-cells. The frequency with which 

 the excess over the normal 12 at any pole was represented in the series of 

 Nos. to 11 is represented by the corresponding series of coefficients of the 

 binomial {a-\-hy^. There was a sUght excess over expectation in the more 

 unequal segregations. 



One plant from the cross tetraploid by diploid had 37 chromosomes, or 

 24+13. Of these 13 extra chromosomes, the frequency with which the 

 extra numbers in the series to 13 went to either pole at mitosis agrees 

 with the series of coefficients of the binomial (a+6)^^, with a slight deficit in 

 frequency of the more unequal segregations. 



These results are concerned only with total numbers of chromosomes, 

 and do not regard the distribution of the sets of 2, 3, or 4 homologous chro- 

 mosomes. It will be possible to follow the fate of such sets in the future by 

 means of their size differences. In the cultivated hyacinth it has for some 

 time been known (de Mol) that three size classes are easily recognized among 

 its 8 pairs of chromosomes. These size classes have been followed in the 

 pollen-grains of a "hexaploid" hyacinth and gave results corresponding 

 closely with the distribution expected in each case if the chromosomes of 

 the three different sizes assorted without mutual attraction or repulsion. 

 Hemerocallis fulva, growing escaped from cultivation near the station, has 

 been found by Dr. Belling to be a triploid; and in this case, apparently, to 

 have the distribution of the segregation of the extra 11 chromosomes which 

 follows the law of chance. 



