DEPARTMENT OF GENETICS. 97 



individual grains are distinctly larger; haploids have a high percentage of 

 empty grains (around 75 per cent), and the few good grains present are mainly 

 of the same size as those of diploids; triploids have not only a high percentage 

 of empty grains, but the full grains are of various sizes. Modified diploids, 

 triploids, and tetraploids have pollen characters resembling their balanced 

 types, but generally with a higher percentage of empty grains. Pollen counts 

 can not take the place of chromosomal counts, but they are useful in survey 

 work in determining in which group individual plants probably belong." 



Unbalanced Types of Datura Mutants. 



"A paper is nearly ready for publication on trisomic inheritance of the color 

 factor in the (2n+l) mutant Poinsettia, which will embody color records on 

 between 12,000 and 15,000 descendants of Poinsettia parents. The completed 

 mass of data fully substantiates the idea of random assortment of chromosomes 

 in the purple-white chromosomal set at the reduction division. This is true 

 when the purple and the white factors come from closely similar lines 

 (obtained from Washington, D. C.) . When the white factor, however, is derived 

 from a distinctly different line (obtained from Germany) , an excess of whites 

 regularly appears in the offspring from certain crosses. The matter is 

 receiving further experimental investigation. 



"The trisomic inheritance in the cocklebur mutant is similar to the more 

 typical behavior in Poinsettia, except that recessive inermis plants appear 

 where they would not be expected. A similar peculiarity in the inheritance of 

 the same factors occurs in tetraploid races and indicates that there is some 

 peculiar behavior of the armed-inermis set which is not evident in the purple- 

 white set. 



"It is obvious that if there are 12 sets of chromosomes there should be 

 expected 12 mutants of the type (2n+l). These we have called the twelve 

 apostles. At the present time we have, however, at least 20 distinct mutants 

 which, both from their breeding behavior and from counts of their chromo- 

 somes, are known to be of the (2n+l) type. In certain cases we have been 

 able to show that one of these mutants is a variety or "acolyte" of one of the 

 apostles. The matter is under investigation, but a provisional hypothesis for 

 which there is some evidence is that the difference between apostle and acolyte 

 is due to a Mendelian factor which produces a visible effect only when certain 

 of the chromosomes are in trisomes. 



"Apparently a new type of mutation has occurred in this year's plantings, 

 namely, a somatic mutation producing a sectorial chimera with one branch 

 distinctly abnormal in appearance and the rest of the plant apparently normal. 

 Pollen from the abnormal sector contains a high proportion of bad grains, and 

 chromosomal counts in pollen mother-cells show the tissue to have the con- 

 stitution (2/i—l). This fact suggests that some of the not infrequent bud 

 mutations found elsewhere in plants, such as in the citrus group, may be due 

 to somatic changes in chromosomal number. 



"It has been established, that of the modified tetraploids, selfed (4/1+1) 

 Globe plants produce over 50 per cent Globe offspring; that selfed (4n+2) 

 Globes produce around 90 per cent Globes; that pollen from a (4n+l) Globe 

 when used on a 4/i female produces about 15 per cent Globe offspring, while 

 pollen from a (4n+2) Globe used on a An female produces about 70 per cent 

 Globes. 



"Non-disjunctional mutations occur much more frequently in tetraploids 

 than in diploids, and in addition to simple (4n+l) and (4n+2) mutants, 

 double, triple, and quadruple mutants have been discovered. Due to the 

 comparatively slight unbalance produced by the change of a single chromo- 

 some in tetraploids, the types are not so readily identified by mere inspection. 



