96 CARNEGIE INSTITUTION OF WASHINGTON. 



A paper on inheritance in tetraploid Daturas now ready for publication 

 gives records of color in over 37,000 individuals and forms the basis for the 

 conclusion that the ratios of purples to whites obtained from combining the 

 various genetic types are brought about by a random assortment of chromo- 

 somes in the purple-white chromosome set. 



Dr. Blakeslee has reported further as follows: 



"Tetraploidy may have been of influence in evolution. An experimental 

 proof of this belief would probably demand the transformation of a tetraploid 

 with 12 sets of chromosomes each into a double diploid with 24 sets of 2 

 chromosomes each. A single plant last year appeared to be duplex for the 

 purple and armed factors and to give 15: 1 ratios for both these two factors 

 in its offspring, a result to have been expected if the plant had been a double 

 diploid. Plants in the next generation, however, gave normal ratios character- 

 istic of typical tetraploids and showed that, if the plant were in fact a double 

 diploid, the condition had not been transmitted to its offspring. 



"The assumption seems reasonable that competing plants in nature are 

 diploid, although few forms have been critically tested and proven to be 

 diploid rather than tetraploid either by breeding or cytological evidence. In 

 order to discover if tetraploids are capable of establishing themselves in nature, 

 we planted out on Goose Island this spring a score of tetraploid plants of 

 Datura, and plan to leave them to seed themselves in competition with other 

 species. 



"The most expeditious way to obtain the full range of different (2n-fl) 

 mutants in a given stock is by crossing triploids (3n) by diploids (2n) . Tri- 

 ploids we have never identified in the offspring when both parents were di- 

 ploids, but they may be obtained by crossing tetraploids with diploids. We 

 are interested, therefore, in having possibly discovered a method of identifying 

 tetraploids in the seed stage. Seeds of tetraploid plants average distinctly 

 larger than those of diploids and from the few large seeds out of many 

 thousands examined from our chief main line one already has given rise to a 

 tetraploid plant, the first that we have discovered in this line during the last 7 

 years in which it has been extensively planted. It is not impossible that by 

 a similar method of selecting large seed from tetraploids {An) we may be able 

 to discover an octaploid (8n) individual. 



" Our balanced series now stands In, 2n, 3n, An, and in general a correspond- 

 ing increase in size from the In to the An condition can be observed in external 

 parts such as leaf and flowers, as well as in their component cells. Such a 

 quantitative change is to be expected from the increased number of chro- 

 mosomes when the balance is not disturbed by an increase or decrease in 

 individual chromosomal sets. Less expected is the qualitative change in 

 shape from the In to the An condition. Considering the leaves of 2n plants 

 to be normal, In leaves are distinctly narrower, and An leaves are broader. 

 Further, the 2n capsules are ovate and An capsules are nearly spherical. Such 

 qualitative differences between balanced types may be due to quantitative 

 differences in the ease of division or of expansion of cells in the different 

 directions or may be due to the factors in certain chromosomes passing critical 

 points sooner than those in others when the balanced chromosomal number is 

 increased or decreased. 



"The condition of the pollen has been found to be a diagnostic feature of 

 considerable value in distinguishing the different balanced types, before the 

 formation of capsules. Thus diploids (2/i) have relatively good pollen; 

 tetraploids {An) have pollen only slightly less perfect than diploids, but the 



