EXPERIMENTAL EVOLUTION. 131 



in both epidermal and subepidermal layers, while in other purples the 

 color seems confined to the epidermis. The occurrence of sectorial and 

 periclinal chimeras and mutative stripes and spotting render the por- 

 tulacas a favorable species for the study of vegetative mutations. 



Datura.— In the jimson-weed, work has been brought to a close on 

 the graft-infectious disease "quercina," which causes profound mor- 

 phological changes in the plants affected; and the results of the in- 

 vestigations are embodied in a paper now in press. It is suggested that 

 the non-Mendelian behavior of ''rogues" in culinary peas may be due 

 to a similar type of disease. 



The investigation of the mutants obtained from the jimson weed 

 has been continued by Dr. Blakeslee, who reports as follows : 



"Most intensive work has been carried on with the 'Globe' mutant, since 

 this mutant can readily be recognized in the seed pans, but the same mode of 

 inheritance seems to be characteristic of all the mutants of this type. The 

 mutant character is transmitted to about one-quarter of its offspring, whether 

 the female parent is fertilized by its own or by foreign pollen. The mutant 

 character is transmitted not at all or but slightly through the pollen. Tests 

 made last fall and winter showed that all the normal races have good pollen, 

 with less than 5 per cent of the grains defective, while all the mutants of the 

 globe type have a relatively high percentage of bad pollen-grains. The 

 mutant 'New Species,' which differs from other mutants in breeding true and 

 being largely sterile with other hues, resembles normals in that its pollen is 

 relatively good. This mutant is one in which abnormal color ratios had been 

 observed. It was believed that chromosome relations might furnish a clue to 

 the abnormal behavior of the mutants in this species. Mr. John Belling has 

 cooperated in a cj'tological study of our mutant variants. The work is still 

 in progress, but some definite results have already been obtained. 



" The ' New Species' turns out to be tetraploid, having 24 in contrast to 12 

 pairs of chromosomes, characteristic of normal races. Sufficient breeding 

 work with this form has already been carried out to settle certain questions 

 in regard to the beha\dor of the tetraploid chromosomes at the formation of 

 gametes. Independent assortment of the chromosomes is demanded by the 

 breeding results. Starting with a tetraploid plant that is heterozygous for a 

 factor A and represented by the formula AA'aa', we expect its gametes to be 

 AA', Aa, Aa', A'a, A'a', aa', if the chromosomes assort independently. If we 

 disregard the primes and sum the types we have 1 AA 4- 4 Aa + 1 aa as the 

 formula for both the male and female gametes produced by a tetraploid 

 plant of the formula AA'aa'. Selfing such a plant, one should obtain 

 the following: 1 AAAA + 8 AAAa 4-18 AAaa -}- 8 Aaaa + 1 aaaa, or 

 a ratio of dominants to recessives of 35 : 1. The AAAA plants should 

 always breed true; the AAAa plants when selfed or back-crossed to 

 recessives should give only dominants, but in later generations should give 

 35 : 1 ratios; the AAaa plants when selfed should again give 35 : 1 ratios and 

 when back- crossed to recessives should give 5 : 1 ratios; the Aaaa plants when 

 selfed should give 3 : 1 ratios and 1 : 1 ratios when back-crossed; and in later 

 generations some 35 : 1 ratios should be expected from selfing individual 

 plants of a 3 : 1 pedigree. By using the purple color of stem, which is domi- 

 nant to green stem, we have been able to take records from the seed pans and 

 in consequence have obtained relatively large numbers in our pedigrees. The 

 results show that the inheritance of the purple color of stem in our tetraploid 

 jimson is in accord with the theoretical expectation already described for 

 independent assortment of chromosomes. Last j'ear a tetraploid plant arose 



