346 Colchicine 



One other fact is striking. There were, in all. 17-^ chromosomes 

 lost; and the largest type, known as the L chromosome, was missing 

 more often than other types. Previous data tor spontaneotisly occur- 

 ring Datura showed that the 1 + 2, or L chromosome was missing 

 more often than any other type. Special morphological traits are 

 fairly reliable for recording Datura progenies.^ 



Before these data were reported, missing chromosomes were known 

 in Drosophila. Nicotiaita^'^ heteroploids were obtained by other treat- 

 ments, and a genetic demonstration proved the loss of chromosomes 

 in a culture of Hyocyauius niger. Since the Datura work was pub- 

 lished, deficient types have been recognized in Nicotiana,^'' LiliumP 

 and Eruca.^^ There must be many that have escaped notice and also 

 records that are not specifically listed here. 



If one looks at the recovery stages from colchicine, the explanation 

 for the tetraploid deficient types can be seen easily. One or two 

 chromosomes are left outside the restitution tetrai)loid nucleus. The 

 causes of a diploid deficient case require additional examination be- 

 cause a c-mitosis leading to a tetraploid restitution nucleus would not 

 have taken place unless a distributed c-mitosis of unequal distribution, 

 23 and 25 respectively, occurred. The 23-chromosome cell would lead 

 to a deficient cell and the 25 to extra-chromosome types. There is 

 yet another explanation. When grasshopper neuroblasts were treated 

 at certain concentrations that did not completely destroy the spindle, 

 certain chromosomes were lagging. Presumably an incomplete in- 

 hibition could cause one chromosome to lag. The fact that the larg- 

 est chromosome of Datura was the one most often missing is of in- 

 terest.^ To assume that tetraploid deficient tyj^es and the diploid 

 deficiencies arose from a similar action on the spindle appears to be 

 oversimplification of the problem. 



Among the progenies of these treated plants there appeared also 

 extra-chromosomal types.^ The fifteen-year breeding record for Da- 

 tura showed that 0.16 -f .019 jier cent of the 2// jilants recorded were 

 extra-chromosomal types.'' Among the 2135 plants, 0.52 + .105 per 

 cent hatl one or more chromosomes. This value is 3.36 times the 

 probable error, and combining data for two years leads to a value 4.42 

 times the probable error.^ An increase caused by colchicine seems a 

 reasonable explanation. Of the extra-chromosomal types induced by 

 colchicine, ten plants had 2" + 1 chromosomes, one had 2n -U 1 + 1, 

 and three were 4» -(- 1. If colchicine increased the frequency, the 

 action had to occur at mitosis during treatment. A specific action on 

 the spindle directed to one chromosome is suggested. 



Aneuploids from treatments in Lilium lon<riflurufn were analyzed 

 from root tips and not the jjollcn mother cells.-" Out of 500 plants 

 treated and analyzed, 303 cases from roots were counted. Eight aneu- 



